-------
NO denotes nitrate plus nitrite nitrogen
NO denotes biodegradable organic nitrogen
SRT * solids retention time, t
TKN - total Kjeldahl nitrogen, M/L3 .
TONinf = total organic nitrogen 1n Influent, ML
U - specific substrate utilization rate, t
V - reactor volume
WQ2 - mass rate of oxygen applied, M/t
X • particulate material concentration, ML
Subscripts Used with X
B denotes blomass In conventional model, COD
Bl denotes Influent blomass, COD
B2 denotes effluent blomass, COD
BA denotes active autotrophlc biomass, COD
BH denotes active heterotrophlc blomass, COD
P denotes decay products, COD
I denotes Inert organic matter concentration, COD
II denotes Influent Inert organic matter, COD
ND denotes biodegradable organic nitrogen, N
P denotes products arising from oxldatlve decay, COD
S denotes slowly biodegradable substrate, COD
yield coefficient, blmass COD/substrate COD
<
Subscripts Used with Y
a denotes autotrophlc yield coefficient
H denotes heterotrophlc yield coefficient
HH * correction factor for anoxlc hydrolysis
0C - SRT at steady state conditions, t
HG • correction factor for anoxlc growth of hcterotrophs
u - specific growth rate, t
Subscripts Used with u
c denotes conventional specific growth rate
cm denotes conventional maximal specific growth rate
hm denotes autotrophlc specific growth rate
m denotes maximal specific growth rate
am denotes autotrophlc specific growth rate
-------
ACKNOWLEDGEMENTS
Significant effort on the part of the professional staff at the various
wastewater treatment plants examined 1n this study was required to
coordinate the field studies and to supply operating data. Contributions of
the following individuals are gratefully acknowledged: Lee Hauswlrth for his
efforts at the Portage Lake plant, Michael Plerner, David Schauer, and Jack
Boex for their efforts at the Green Bay plant, Paul Nehm for his efforts at
the Madison plant, Read WarMner for his efforts at the Milwaukee Jones
Island and South Shore Plants, and Jerry ElUfson for his efforts at the
Monroe Plant.
This study utilized off-gas and plant operating data collected by other
investigators working on the EPA - ASCE Fine Bubble Diffused Aeration Design
Manual Project. Such data and Information contributed by David Redmon of
Ewlng Engineering, William Boyle of the University of Wisconsin, James Marx
of Donohue & Associates, and Read Warrlner of the Milwaukee Metropolitan
Sewerage Commission are gratefully acknowledged.
This work could not have been completed without the efforts of several
students at Michigan Tech. The contributions of Ronald Mauno, Kevin
Hopkins, He1d1 P1ao, Susan Doerr, Lu L1n, and Janette Lutz are especially
acknowledged. Finally, special thanks are In order to the EPA Project
Officer, James A. Heldman, and to C.P. Leslie Grady Jr. for their help with
the IAWPRC Model.
xl
-------
SECTION 1
INTRODUCTION
The activated sludge process 1s the most widely used method for
secondary wastewater treatment in the United States, and Us popularity 1s
Increasing. Provision of oxygen to the active organisms through aeration is
the most energy intensive aspect of activated sludge process operation and
consumes 60 to 80% of the total energy requirements in wastewater treatment.
Because of this, approximately 1.75 million horsepower of aeration equipment
is currently installed in wastewater treatment plants 1n the United States
and Canada. The energy cost for operating this equipment amounts to more
than $600 million per year. It is estimated that, through improvements 1n
design and operation of these aeration systems, savings of more than $100
million per year could easily be achieved. Furthermore, many municipalities
are replacing older, less efficient aeration systems with more energy-
efficient fine bubble diffused aeration systems, and additional savings 1n
repair and replacement costs could be realized.
For the efficient design and operation of aeration systems 1t is
necessary that the oxygen demands of the biological system and the oxygen
transfer capability of the aeration equipment be accurately predicted. Over
the past few years,^ significant progress has been made in measuring and
modeling the oxygen!transfer capability of aeration equipment (ASCE, 1984)
(Brown and BaiHod,' 1982)(Mueller and Boyle, 1988) under clean water and
process conditions. However, the progress 1n modeling equipment performance
capability has not been matched by improvements in modeling, simulating, and
designing for oxygen transfer requirements 1n full scale activated sludge
plants. The methods commonly used at the present time are adequate only for
estimation of the average oxygen utilization rate, and are of limited
practical usefulness 1n design of efficient new and replacement aeration
systems. With these conventional methods, the designer must make an
educated guess of the spatial and temporal variation 1n oxygen utilization
rates, and this can lead to over-design or under-deslgn of an aeration
system. An under-designed system produces a poorly treated effluent,
whereas an over-designed system has high Initial and operating costs.
The objective of the research described 1n this report 1s to apply recent
advances in activated sludge process modeling (Grady et al., 1986) (Bldstrup
and Grady, 1988) to the simulation of oxygen uptake rates 1n full scale
activated sludge treatment plants. This Is accomplished by calibrating the
International Association for Water Pollution Research and Control (IAWPRC)
Model and associated SSSP software to operating data at six full scale
activated sludge treatment plants.
1
-------
This Information 1s of value to engineers 1n the cost-effective design
and operation of wastewater treatment systems because 1t provides a data
base of applicable sto1ch1ometr1c and kinetic model parameters which the
engineer can utilize, with appropriate judgment, to simulate and predict the
behavior of oxygen transfer systems in wastewater treatment.
-------
SECTION 2
CONCLUSIONS
Field studies were conducted at six municipally owned and operated
activated sludge wastewater treatment plants 1n order to assess and enhance
the usefulness of a mathematical model and associated microcomputer software
package for simulating the oxygen utilization rate (OUR) of the activated
sludge process. The results of the field studies were used to calibrate the
key biological parameters contained 1n the model so that the oxygen
utilization rates, dissolved oxygen (DO) concentrations, mixed liquor
volatile suspended sol Ids concentrations, and process performance simulated
by the model matched the corresponding quantities observed in the treatment
plants. Based on the results of this study, 1t can be concluded that:
1. The International Association for Water Pollution Research and Control
(IAWPRC) Mode! and related SSSP (Simulation of Single-Sludge Processes
for Carbon Oxidation, Nitrification and Den1tr1ficat1on) microcomputer
software package provide a useful capability to analyze and simulate
the average oxygen utilization rate in municipal wastewater treatment
plants.
2. It was possible to obtain reasonable agreement between the average
measured and steady-state simulated values of oxygen uptake rate,
dissolved oxygen concentration, mixed liquor volatile suspended solids,
effluent ammonia, and effluent nitrate concentrations at most of the
plants studied. This agreement was achieved by adjusting or
calibrating only two of three key model parameters while keeping the
other 17 model parameters at their adjusted default values. The key
model parameters were the heterotrophlc yield coefficient,
heterotrophlc decay constant, and autotrophic maximal specific growth
rate constant.
3. When provided with a realistic value of the process water volumetric
mass transfer coefficient (K|_a), the calibrated model and software were
normally able to simulate the spatial and temporal range of dissolved
oxygen concentrations and oxygen utilization rates observed in the
operating treatment plants. Instances in which lack of agreement
between simulated and observed values occurred could be possibly
explained by causes other than model and/or software Inadequacy.
4. Under conditions of low DO concentration (below 1 to 2 mg/L) In vitro
OUR values Indicated by the traditional BOD bottle method tended to be
greater than the in situ OUR values determined from off gas analyses.
-------
These differences were most evident 1n the inlet sections of long
aeration tanks where OUR values were high and DO values were low.
5. Comparison of the IAWPRC model and conventional approaches to estimate
average process oxygen utilization rates showed that, for given values
of the yield and decay coefficients, the two approaches agreed
reasonably well. In this study, the IAWPRC model was calibrated to the
data and, therefore, was judged to give the better estimates of the
process utilization rates. Moreover, the IAWPRC model was judged to be
advantageous because of its ability to simulate nitrogenous oxygen
demand.
-------
SECTION 3
RECOMMENDATIONS
This study relied upon adjustment or calibration of key parameters to
show that the IAWPRC Model could be applied to analyze, simulate, and
predict oxygen utilization and other activated sludge process
characteristics. Such information 1s valuable because it provides a data
base of key parameters from which engineers can draw Information for process
design and operation. However, this data base should be expanded to Include
information from more than the six plants of this study. Moreover, it would
be of particular value to demonstrate the applicability of techniques for
key parameter estimation based on direct measurements.
Accordingly, 1t 1s recommended that additional 1n-depth studies applying
the IAWPRC Model and associated SSSP Software to well controlled full-scale
wastewater treatment plants be conducted. The plants should be selected so
that the process train studied treats a sizable fraction (at least one-
third) of the wastewater flow monitored, and Intensive data collection
should be carried out for a period of about two weeks at each plant. The
Intensive data collection should Include the Items measured In the 24 hour
plant studies described 1n this report, plus periodic off gas measurements
of K|_a,and repeated direct measurements of the key model sto1ch1ometr1c and
kinetic parameters and wastewater characteristics.
»
It is also recommended that design engineers be encouraged to apply the
IAWPRC Model and associated SSSP Software to activated sludge process
design. Although other Improved models and software jnlght be developed 1n
the future, this model and software are available now and are easily usable
by any design engineer. With the aid of engineering judgment and the
parameter data base given In this report, this approach will facilitate more
economical and effective design and operation of wastewater treatment
plants.
-------
SECTION 4
MODELS APPLIED TO OXYGEN UTILIZATION IN BIOLOGICAL WASTE TREATMENT SYSTEMS
REQUISITES FOR BIOLOGICAL PROCESS MODELS
Design and operation of biological waste treatment systems are aided by
application of predictive models which can mimic the performance of these
systems. Because biological waste treatment systems are inherently complex,
It 1s necessary that the model be somewhat simpler than the actual system.
Over the past several years, various conceptual models ranging from single
substrate, single biomass aerobic models (Herbert, 1956) to multiple
substrate, multiple biomass anaerobic models (Rozzi et al., 1985) have been
applied to biological waste treatment.
Successful modeling of a system requires:
1. A substantially correct conceptual representation of the biological
processes, stolchiometry, kinetics, and pathways Involved.
2. Definition of the physical system, I.e. tanks, pipes, flow rates, 1n
sufficient detail so that material balance equations can be written for
each component of Interest.
3. An analytical or numerical solution of the material balance equations.
4. Verification of the model principles and calibration of the model
parameters by application of the model to real treatment systems.
*
Early modeling efforts were hampered by a multiplicity of conceptual
representations and a lack of available computing technology for model
solution. In 1983, the International Association on Water Pollution Research
and Control (IAWPRC) established a task group on "Mathematical Modeling for
Design and Operation of Biological Waste-water Treatment*. The assignment
of this group was to develop a consensus model applicable to an activated
sludge system performing simultaneous carbon oxidation, nitrification and
den1tr1f1cat1on. The model which resulted from this effort 1s termed the
IAWPRC Model and 1s described by Grady et al. (1986), Henze et al. (1987),
and the IAWPRC Task Group (1986). Basic features of this model are
discussed in the following section.
In 1987, Bldstrup and Grady (1987) developed the SSSP (Simulation of
Single-Sludge Processes for Carbon Oxidation, Nitrification and
Oen1triflcation) computer software package based on the IAWPRC Model. This
-------
software, written for the IBM Personal Computer or compatible machines
performs steady-state and dynamic simulations of activated sludge systems
based on the IAWPRC Model. The program 1s versatile 1n that 1t allows the
user to define the system configuration by using up to nine completely
mixed reactors 1n series. Additional flexibility is provided by the
capability to define the influent addition, return sludge recycle, and mixed
liquor redrculatlon flow diagrams between the various reactors.
The work of the IAWPRC Task Group, coupled with the SSSP software
package, satisfied the first three Requirements for successful modeling of a
biological waste treatment system. One of the objectives of this research
effort is to help satisfy Requirement 4, I.e. to verify and calibrate the
IAWPRC model by applying 1t to simulate oxygen utilization 1n full-scale
activated sludge plants.
OXYGEN REQUIREMENTS IN ACTIVATED SLUDGE SYSTEMS
Oxygen requirements in activated sludge systems arise from:
* Biological oxidation of organic carbon
* Biological oxidation of nitrogen
* Chemical oxidation of inorganic substances
The rates of oxygen utilization by each of these processes are functions of
the process influent load and operating conditions. Efficient design of
oxygenation or aeration systems requires reliable estimates of the average,
minimum and maximal total oxygen utilization rates. In addition,
information on the spatial variation of oxygen utilization rate within the
process must be known. Also, aeration frequently 1s relied upon to mix the
fluid to promote mass transfer and to keep the biomass particles suspended.
In lightly loaded activated sludge or aerated lagoon systems, the-mixing
requirement may control.
Empirical Approaches
In the past, various empirical and rule-of-thumb approaches have been
used to estimate oxygen requirements for activated sfudge systems:
* 1,500 cubic feet of air per pound of 5-day BOD applied
* 1 pound of oxygen transferred per pound of 5-day BOD applied (Great
Lakes - Upper Mississippi River Board of State Sanitary Engineers,
1978)
* 0.5 to 2.0 cubic feet of air per gallon of sewage (Fair and Geyer,
1954)
* 500 to 700 cubic feet of air per pound of 5-day BOD removed with at
least 3 cfm per foot of tank length for mixing (Water Pollution
Control Federation, 1977)
Conventional Model Applied to Heterotroohlc Oxygen Utilization
More recently, rational approaches based on the overall process oxygen,
7
-------
substrate, and blomass balances have been developed to estimate carbonaceous
oxygen demand (Grady and Lim, 1980). The following development illustrates
the use of a matrix table (Grady et al. 1986) to formulate material balances
for the conventional carbonaceous oxygen utilization model described in
Figure 1.
Substrate
Pathway
Process
Mediated by
Organics,
Oxygen, SQ
Decay
Products, Xp<-
Aerobic
Heterotrophic
Blomass Growth
Biomass
Heterotrophic
Biomass Decay
(negative growth)
Biomass
Figure 1. Substrates, pathways and processes Included in the
conventional carbonaceous oxygen utilization model.
It should be noted that, throughout this report,.concentrations of
carbonaceous substrates and blomass are expressed as total carbonaceous
biochemical oxygen demand-(TCBOD). This is used synonymously with
biodegradable chemical oxygen demand (BCOD). In the literature, this is
often referred to as ultimate biochemical oxygen demand or as toUl
biochemical oxygen demand. Blomass volatile suspended solids are assumed to
have the composition Cs^NO?, and are converted to o*ygen equivalent by the
factor 1.41 g oxygen/gram YSS.
Explanation of Matrix Table for Model Representation—
For complex models in which several processes operate to transform
several components, it 1s convenient to use a matrix table to represent the
kinetics and stoichiometry of the model. Table 1 1s an example of a matrix
table for the model described in Figure 1. The model considers only aerobic
heterotrophic processes in biological waste treatment. The components of
Importance in the model are listed by symbol across the top of the table,
and their definitions are given at the bottom of the corresponding columns.
This model considers only two processes, aerobic growth of
heterotrophic blomass, and oxidative decay of heterotrophic blomass. The
decay process 1s termed "oxidatlve" because it utilizes oxygen. It Is
8
-------
0)
T3
C
O
a
N
3
C
4)
u
a
a
o
Process Ral
00 CO
^^ S^
.*
CO X*
o
CM CO
•-I CO
#
O
CO
o +
CO
5
CO
CO
CO
co +
CO
1
1
u u
3. .O.
X 1 O
>- 1 X »-
1 1 >• 1
<-l 1 rH
1 1
o
E « 5 X* ^
3 *> MO +J
E ID C IO
•«- at o -o "3
X 0 C T-
*f ^: s
— 5 <•» *> ^
So "u £ c
^o 2 tJ 5
SiT > "§ "c
«- co 4)
o o. '•a o
O CO X O
Soluble Inert Organic Matter
Concentration, COD
Soluble Substrate Cone., COD
Parti cu late Inert Organic
Matter Concentration, COO
Blomass Concentration, COD
Parti cul ate Products Arising
from 0x1 dative Decay,
Concentration, COO
Dissolved Oxygen Concentration,
negative COO
o
u
o
<4-
>!
CO
(O
to
U
I
*z
tt
5
V)
t/t
§
°s
P
ft C
li
Vf
§
-------
important to distinguish this decay process from the hydrolytic decay
process incorporated into the IAWPRC Model. The hydrolytic process uses no
oxygen, and because of this, the magnitude of the hydrolytic decay
coefficient is roughly 2.5 times the oxldatlve decay coefficient. Reasons
for this difference are discussed later. The model assumes that each unit
of biomass that undergoes decay 1s partially oxidized and partially
converted to partlculate products which are resistant to further decay. The
observed fraction of biomass converted to resistant partlculate products,
Xp, has been reported to be 0.2 (IAWPRC Task Group, 1986).
The processes are listed down the extreme left column, and the
corresponding rate equations are given 1n the extreme right column. It 1s
important to realize that the process rate equations are expressed 1n
dimensions of (biomass COD)/(t1me)/(volume). The sto1ch1ometr1c coefficients
listed in the body of the table have dimensions of (mass of component
COD)/(mass of biomass COO). Thus, multiplication of a sto1ch1ometr1c
coefficient times the process rate equation gives the component transformation
rate, as (mass of component COD)/(t1me)/(volume), for that process. Therefore,
proceeding down a column and summing the products of the sto1ch1ometr1c
coefficients times the process rate equations will give the rate of component
transformation for use in component material balances. A typical component
material balance can be stated as:
Rate of
Component In
By Flow
Rate of Rate of
Component Out + Component
By Flow Production
Rate of
Component (1)
Accumulation
The rate of oxygen production, for example, 1s given by summing the products
of the coefficients listed under SQ times their corresponding rate equations.
Rate of
Oxygen
Utilization
Rate of
- Oxygen
Production
I-YH
+ (l-f0hbcXBV (2)
where: V • reactor volume, and the other symbols are defined 1n Table 1.
Stolch1ometrlc Coefficients—
By definition,«the yield coefficient relates the biomass growth rate to the
substrate utilization rate.
Specific Biomass Growth Rate uc
Specific Substrate Utilization Rate U
(3)
where U - specific substrate utilization rate, t
-1
10
-------
Hence, the appropriate sto1ch1ometr1c coefficient by which to multiply the
biomass growth rate expression to obtain substrate utilization rate 1s the
negative reciprocal of the yield coefficient, I.e. - U/uc - - I/YH- This
factor, therefore, appears in row 1, column 2 of Table 1.
When substrate and biomass are both expressed as total carbonaceous
biochemical oxygen demand TCBOD, or biodegradable COD, the rate of substrate
TCBOD utilized for growth 1s equal to the rate of oxygen utilization for
biomass growth plus the TCBOO of the biomass produced, or
U
(4)
where r0 » specific oxygen utilization rate, 1/t
-1
The appropriate stolchlometric coefficient by which to multiply the biomass
rate expressions to obtain oxygen utilization rate 1s -r0/uc. Combination of
Equations 3 and 4 gives this coefficient as
- r0/uc = -( u/uc - i) - - (i-YH)/YH
Use of Matrix Table for Formulation of Component Material Balances
(5)
Consider the input to and output from the activated sludge process shown in
Figure 2.
INPUT WASTEWATER
Flow
Biomass Cone.
Substrate Con'c.
Oxygen (00) Cone.
Inert Part. Cone.
Xjj
INPUT OXYGEN TRANSFERRED
"I—
OUTPUT EFFLUENT
Q - Qw
*B2
SQ2
OUTPUT WASTE SLUDGE
Qw
XB
SC2
SQ2
Figure 2. Activated sludge process material balance.
The process material balances can be written with the aid of the matrix shown
In Table 1.
Biomass Balance
QXBi - (Q-
(VXB)
dt
(6)
11
-------
Substrate Balance
1 d
QSci " °wsC2 - (Q-Ow)Sc2 Vc XBV - — (VSC2) (7)
YH dt
Oxygen Balance
QSfji ~ Qw^02 ~ (Q~Qw)So2 ~ ( )uc XBV ~ bcXBV (l~^o)
YH
+ KLa V(C*. - S02) - -- (VS02) (8)
* dt
where, C . - DO saturation concentration approached at Infinite time 1n
the unsteady state oxygen transfer test (ASCE, 1984).
Particulate Products Balance
d
0(0) - QwXp - ((MM XP2 + f0bcXBV « —- (VXp) (9)
dt
Inert Particulates Balance
d
OXn - OwXj - (0 - Ow) XI2 - 0 - (VXj) (10)
dt
From Table 1, the rate of oxygen utilization, Rc, 1s defined by,
I-YH
- (l-fo)bc*BV (11)
YH
Consideration of the relative magnitudes of the terms 1n the oxygen balance
shows that the dissolved oxygen advectfon terms (the first three 1n Equation
8) are small compared to the other terms and can be neglected. Likewise, 1t
Is assumed that the partlculate products Xp2, and Inert partlculate, Xj2, 1n
the effluent can be approximated by zero. Steady state conditions are
assumed so that all derivatives with respect to time are zero. With these
simplifications, the blomass, substrate, oxygen, and partlculate products
balances can be combined with Equation 11 to express the oxygen utilization
rate as,
i + bcec - YH -
- SC2) ( -------------------- - ------ ) (12)
i + bcec
where:
Rc • mass of oxygen required per unit time for the carbon oxidation
processes, M/t
o
0 - flow rate entering and leaving the process, L /t
12
-------
SC2 * total carbonaceous oxygen demand (ultimate BOO) of the flows
entering and leaving the process, M/L
bc - conventional decay coefficient for oxldatlve decay of blomass,
1/t
YH - yield coefficient, mass of volatile suspended solids COD
produced per unit mass of carbonaceous oxygen demand utilized
9C - sol Ids retention time (SRT) at steady state conditions,
defined as the blomass 1n Inventory divided by the rate of
blomass wasting, t
f0 - fraction of blomass yielding nonblodegradable partlculate
products upon oxldatlve decay, generally taken as 0.2
The decay coefficient, bc, and yield coefficient, YH, are biological
parameters characteristic of the blomass Itself, whereas the solids retention
time, 8C, is the key parameter which controls the process operation and
performance. Typical values of sol Ids retention time range from 3 to 12 days.
Numerical values of Y depend on the units in which blomass and substrate are
expressed. Typical values for Y range from 0.45 to 0.7 mass of VSS COO
produced per unit mass of feed COD utilized. Typical values of the oxldatlve
decay coefficient, bc, range from 0.04 to 0.4 per day (Grady and Urn, 1980). A
variation of this model (Grady and Urn, 1980) neglects the partlculate
products component and becomes equivalent to this model when f0 1s taken as
zero.
In the general process analysis and design solution for this problem, (a
single completely mixed aeration tank with recycle) another blomass balance 1s
written on the reactor Itself to include the recycle stream, and use 1s made
of the kinetic relationship for uc shown in Table 1. For fixed values of the
input quantities, output quantities (Sc2 1s very small but 1s not fixed),
Sto1ch1ometr1c parameters, and kinetic parameters, the design problem has four
degrees of freedom. Typically, the blomass concentration, Xjj, the solids
retention time, 8C, the reactor dissolved oxygen concentration, $02, and the
blomass concentration 1n the recycled sludge are chosen as desigfi variables.
*
This model may also be applied to a more complicated flow diagram such as
tanks 1n series, contact stabilization, or step feed. However, the material
balance equations become more difficult to represent and solve. Furthermore,
there 1s little reason to do so as the SSSP software package already supplies
solutions for a more realistic model applied to a wide variety of process
configurations.
Conventional Model for Nitrogenous Oxygen Utilization
The overall nitrogenous oxygen utilization rate can be calculated based on
(EPA 1975):
1. The amount of nitrogen available for nitrification, NJA, mass/time,
determined as the total KJeldahl nitrogen (TKN) 1n the Influent less the
nitrogen in the waste sludge.
13
-------
2. The amount of total nitrogen converted to nitrate, Ny^. mass/time,
determined by subtracting the TKN of the effluent from the total
available nitrogen.
3. The amount of nitrate 1n the effluent, NNo»mass/t1me, measured or
estimated.
4. The amount of nitrogen denitrified, NQN, mass/time, determined by
subtracting the nitrate 1n the effluent from the total nitrogen
converted to N03.
The nitrogenous oxygen demand 1s then determined based on the stolchiometric
requirements that:
* 1 mg of ammonia or Kjeldahl nitrogen requires 4.57 mg of oxygen for
conversion to nitrate, and
* denltrificatlon of 1 mg of nitrate nitrogen provides an equivalent oxygen
credit of 2.86 mg.
Therefore, the nitrogenous oxygen demand, Rn, can be expressed as:
Rn - 4.57 NTN - 2.86 NDN (13)
And the total carbonaceous and nitrogenous oxygen utilization rate, Rt 1s then
given by,
Rt - Rc + Rn (14)
Evaluation of Conventional Models for Oxygen Utilization
The conventional modeling approach described above is useful and certainly
represents a great Improvement over the empirical methods. However, 1t has
the following disadvantages:
1. The model Itself 1s deficient 1n that it does not adequately distinguish
between slowly and readily degradable substrate*; nor does 1t allow for
replenishment of soluble substrate through hydrolysis.
2. The conventional nitrogenous model requires the user to estimate the
amount of nitrification and denltrificatlon attained by estimating the
TKN and nitrate concentrations 1n the effluent. It does not provide any
capability to simulate or predict these concentrations based on
wastewater or process parameters. Consequently, the nitrogenous oxygen
demand calculations are only as precise as the estimates.
These disadvantages are largely eliminated by the IAWPRC Model and the SSSP
Software Package.
14
-------
International Association for Water Pollution Research and Control (IAWPRC)
Model Applied to Heterotrophic and Autotrophlc Oxygen Utilization
The substrates, pathways, and processes included in the IAWPRC Model are
Substrate
Pathway
Process
Mediated by
Soluble
Organic S$
Oxygen SQ
Amnonla
Alkalinity
Nitrate SNO
Soluble
Organic N
Partlculate
Organic N
Partlculate
Organic X$
Decay
Products
Aerobic
Heterotrophic
Biomass Growth
Anoxlc
Heterotrophic
Biomass Growth
Autotrophlc
Biomass Growth
Hydrolysis and
Ammonlfication
(non-growth)
Heterotrophic
Biomass Decay
(negative growth)
Autotrophlc
Biomass Decay
(negative growth)
Heterotrophic
Biomass
Heterotrophic
Biomass
Autotrophlc
Biomass
Heterotrophic
Biomass,
Heterotrophic
Biomass,
Autotrophlc
Biomass,
Figure 3. Substrates, pathways and processes Included 1n the IAWPRC model.
15
-------
described 1n Figure 3. Comparison of Figure 3 with Figure 1 shows the
comprehensive nature of the IAWPRC Model. The model contains 9 substrates
which are involved in 8 processes mediated by 2 active biomass fractions. The
increase 1n complexity of this model over the conventional carbonaceous model
is caused by:
1. Inclusion of autotrophlc nitrification and other nitrogenous pathways.
This adds four processes (autotrophlc growth, autotrophlc decay,
ammonlfication, and hydrolysis of organic nitrogen) and five components
(autotrophlc biomass, nitrate, ammonia, particulate organic nitrogen,
and soluble organic nitrogen).
2. Recognition of two classes of carbonaceous substrate, readily degradable
(soluble) organlcs, and slowly degradable (particulate) organics. This
adds one additional component and one additional process.
3. Inclusion of den1trif1cat1on by anoxlc growth of heterotrophs. This
adds one additional process.
Table 2 shows the component/process matrix for the IAWPRC Model. The
corresponding process rate equations are given 1n Table 3, and the definitions
of process symbols are given 1n Table 4. Table 5 summarizes the kinetic and
stolchlometric parameters contained 1n the model.
The matrix table 1s straight-forward and can be applied to formulate the
reaction terms for system material balances 1n the same fashion that Table 1
was applied to formulate material balances. The stolchlometric coefficients
listed for the heterotrophlc growth processes are similar to those explained
earlier for Table 1. The factors 4.57 and 2.86, respectively, represent the
nitrification demand per unit ammonia nitrogen and the den1trif1cation credit
per unit of nitrate nitrogen. They appear because both ammonia and nitrate,
are expressed as concentration of nitrogen rather than as COD, and because the
autotrophlc yield 1s expressed as autotrophlc biomass COD produced/mass of
nitrogen utilized.
An Important difference between the IAWPRC and conventional carbonaceous
models arises because of the hydrolytlc (as opposed to oxldatlve) nature of
the decay processes built Into the IAWPRC Model. Note that 1n the
conventional model described by Table 1, the decay process utilizes oxygen to
oxidize (1-fo) fraction of biomass COD and converts f0 fraction of biomass
Into resistant particulate products. However, 1n the IAWPRC Model, the decay
processes use no oxygen. Instead, (1-fp) fraction of biomass 1s hydrolyzed to
slowly degradable substrate, X$, which, 1n turn, 1s hydrolyzed to readily
degradable substrate, S$. Utilization of oxygen Is accounted for only by
aerobic growth of heterotrophs on readily degradable substrate and by growth
of autotrophs. Therefore, to attain the same heterotrophlc oxygen utilization
as the conventional model, the IAWPRC Model has to process more readily
degradable substrate through the synthesls/hydrolytlc decay cycle. Because of
this, the heterotrophlc decay coefficient, bu> and maximal specific growth
rate constant, pm applicable to the IAWPRC Model, are larger (by a factor
of 2 to 3) than the corresponding values for the conventional model.
16
-------
I-H
CM
t-4
i-t
r-l
0
r-t
at
CO
rx
to
in
m
(M
»*
^
1
—I
Irt
O
z
X
O
i
i
I/)
O
i/i
Q-
X
*.
I"
fcO
X
>?
Irt
«r
Process
~
CD 1 «T
X 1 <-!
^~
1
X
T
^. T*
1 >•
r-4
1
t-l
- ^
1
^
£
1 2
5 £
Aerobic gr
heterotrop
-
>
<4
i <>
i ff
9-t *«
^
•T
I
>-
1
r-l
••
1-1
«.
c
f.
1
i
c
••>
«
i M
: £
$
i •?
I
i <
^-i i >-
I iv
CD 1
X 1 «•
1
^ ! >?
i
^
f-H t <
1 >-
>-
in >-
9
i
rl
•5
£
•tf
Aerobic gr
autotrophs
«*»
8-
a.
i
a.
n-
r-1
1
a.
i
M
Decay of
heterotrop
<*
^
a.
j
^
a.
-
a.
i
M
Decay of
autotroph
m
.
i —i
i
c
5
8
17
-------
TABLE 3. RATE EXPRESSIONS FOR PROCESSES INCLUDED IN THE IAWPRC MODEL
4
5
6
7
8
heterotrophs
heterotrophs
autotrophs
Decay of
heterotrophs
Decay of
autotrophs
Ammon1f1cat1on
Hydrolysis
Hydrolysis of
organic N
Process rate, ML~ T~
S 0
ii / \ ( \ Y
^nm i A MBH
KS + ss KS+SO
ss KOH SNO
Mhfli v A A 'HG^BH
KS + $s KOH •*• SQ K^Q > SQ
SNH s0
II 1 ___W__ ™_Mf»»
Warn v A ;ABA
KNH + SNH KOA + SNO
bHxBH
bAH
>M.
XS/XBH s0 KOH SNO
KX * (XS/XBH) KOH * so KOH + so KNO * SNO
r
(process 7 rate) (XNO/XS)
Another difference between the IAWPRC and conventional models arises 1n the
fractions of blomass decay converted to resistant partlculate products. In
the conventional model, this fraction 1s based on a "once through" decay and
Is generally taken as 0.2 (IAWPRC Task Group, 1986). However, the growth-
decay processes are Incorporated Into a cycle 1n the IAWPRC Model, and fp
represents the fraction of decay converted to partlculate products on each
pass through the cycle. By considering the quantity of readily degradable
substrate utilized and recycled through the growth/hydrolytlc decay cycle, It
can be reasoned that, for 1 gram of S$ used for growth,
YH(b/u)(l-fp) grams of Ss
are replenished through hydrolytlc decay, giving a net substrate use of
1 - YH(b/u)(l-fp)
18
-------
TABLE 4. DEFINITION OF FOR THE IAWPRC MODEL
Component Component
number symbol Definition
1
2
3
4
5
6
7
8
9
10
11
12
13
Assuming
gives,
be
b
Si
ss
*I
xs
XBH
XBA
Xp
so
SNO
SNH
SND
XND
SALK
Soluble inert organic matter — M(COD)L~
Readily biodegradable substrate — M(COD)L"3
Particulate Inert organic matter ~ M(COD)L
Slowly biodegradable substrate ~ M(COD)L~3
Active heterotrophic biomass — M(COD)L
Active autotrophic biomass ~ M(COD)L
Particulate products arising from biomass decay — M(COD)L
Oxygen (negative COD) M(COD)L~3
Nitrate and nitrite nitrogen -- M(N)L"3
NH4+ + NHa nitrogen -- M(N)L"3
Soluble biodegradable organic nitrogen — M(N)L
3
Particulate biodegradable organic nitrogen — M(N)L
Alkalinity— Molar units
that growth conditions are such that u 1s approximately equal to b,
«
-
fp
---- - ( 1 - YH ( 1 - fp )) (15)
fo
Taking the conventional f0 as 0.2, and the heterotrophic yield as 0.66, (based
on biomass COD and substrate COD), gives fD • 0.078, and 0.38 for the ratio of
bc/b.
Solution of the IAWPRC Model Using the SSSP Software Package
This software package was developed by Bldstrup and Grady (1987) and 1s
available from Professor C.P.L. Grady at Clemson University, Clemson, South
Carolina. A general description 1s given by Bldstrup and Grady (1988), and
detailed Instructions are given in the Users' Manual (Bldstrup and Grady,
19
-------
TABLE 5. SUMMARY OF THE KINETIC AND STOICHIOMETRIC PARAMETERS CONTAINED
IN THE IAWPRC MODEL
Kinetic parameters Symbols
Heterotrophic growth and decay uhm , KS, KQH, KNO«
Autotrophlc growth and decay uam » KNH» KOA» bA
Correction factor for anoxlc growth of
heterotrophs HQ
Ammonification K/\
Hydrolysis k^, KX
Correction factor for anoxic
hydrolysis HH
StoichiometMc parameters
Heterotrophic yield YH
Autotrophlc yield Y/\
Fraction of biomass yielding partlculate
products fp
Mass N/Mass COD in biomass
Mass N/Mass COD in products from biomass
1987). It 1s written for the IBM Personal Computer or compatible machines and
performs steady-state and dynamic simulations of activated sludge systems
based on the IAWPRC Model. The program Is user-friendly and versatile 1n that
it allows the user to define the system configuration by using up to nine
completely mixed reactors 1n series. Additional flexibility 1s provided by
the capability to define the Influent addition, return sludge recycle, and
mixed liquor redrculatlon flow patterns between the various reactors.
The program contains "default" values of the stolchlometMc and kinetic
parameters required by the IAWPRC Model. These serve as reasonable starting
points for simulations and calibration. One valuable feature of the program
Is that 1t allows the user to specify either reactor dissolved oxygen
concentration or the reactor volumetric mass transfer coefficient, KLa. The
mode 1n which the K^a value 1s specified 1s particularly useful for
calibrations in which the reactor K^a 1s known. However, when K|_a 1s
20
-------
specified, the program uses Identical values for the dissolved oxygen half-
saturation constants for autotrophs and heterotrophs. This 1s a limitation
because the value for autotrophs 1s generally believed to greater than the
value for heterotrophs.
21
-------
SECTION 5
TWENTY-FOUR HOUR PLANT STUDIES: DATA ACQUISITION AND MEASUREMENTS
ACTIVATED SLUDGE PLANTS USED FOR MODEL CALIBRATION
Six municipally owned and operated wastewater treatment plants were
selected for calibration of the oxygen uptake model. All six of the plants
were concurrently being studied as part of the ASCE-EPA Fine Bubble Diffused
Aeration Design Manual Project (EPA, 1985) and, because of this, background
information on process configuration and operation was readily available.
Moreover, frequent off-gas measurements of oxygen transfer were being made
at five of the six plants as part of the ASCE-EPA project. These
measurements were particularly valuable because they produced accurate
estimates of the process volumetric mass transfer coefficients for oxygen
(KLa).
Twenty-four hour field studies were conducted at the plants described
below. Additional information on the actual and modeled process flow
diagrams and sampling locations 1s given 1n Section 7, Results and Model
Calibration.
* Portage Lake Plant, Houghton, Michigan: This 1s an 8,700 m /d (2.3
MGD) contact stabilization activated sludge plant with aerobic sludge
digestion and no primary sedimentation. It receives almost no «
industrial waste and was of value for model calibration because of the
relatively low temperature and large partlculate organic load 1n the
Influent. The plant was operated at a solids retention time of 10.6
days and produced a partially nitrified effluent. It was the only one
of the six plants at which no off-gas measurements of oxygen transfer
were available.
* Green Bay, Wisconsin: This 1s a modern 182,000 m /d (48 MGD) contact
stabilization activated sludge plant receiving paper mill wastes which
account for 30% of the flow and 50% of the 5 day BOD. The plant has a
thermal sludge conditioning system and recycles the thermal sludge
conditioning liquor to the activated sludge process. Other notable
features of the plant Include a high Influent total Kjeldahl nitrogen
concentration (40 to 60 mg/1) and warm process temperature. The plant
operated at a solids retention time of 3.1 days and achieved very
Uttle nitrification. Off-gas measurements were made during the 24-
hour study period.
* Madison, Wisconsin: This 1s a 151,000 m3/d (40 MGD) activated sludge
22
-------
plant receiving municipal wastes containing an appreciable (6% of flow,
15% of 5 day BOD) meat and cheese processing component. The plant
operated at a solids retention time of 16.4 days and produced a
nitrified effluent.
* Monroe, Wisconsin: This 1s an 8,330 m3/d (2.2 MGD) step-feed activated
sludge plant receiving a significant Industrial load (17% of flow, 50%
of 5 day BOD) consisting primarily of soluble cheese processing and
brewing wastes. However, during the twenty-four hour study, an aerated
in-line equalization basin was employed between the primary settling
and activated sludge processes, and this significantly reduced the
soluble COD fed to the activated sludge process. The plant operated at
a solids retention time of 8.4 days and produced a nitrified effluent.
Off-gas measurements were made during the 24-hour study period.
* Jones Island East Plant, Milwaukee, Wisconsin: This 288,000 m3/d (76
MGD) plant receives an Industrial load (11% of flow, 38% of 5 day BOD)
dominated by the brewing, food processing and tanning Industries.
Primary treatment 1s by fine screening rather than sedimentation. The
plant operated at a sol Ids retention time of 2.8 days and experienced
very little nitrification.
* South Shore Plant, Milwaukee, Wisconsin: This 371,000 m3/d (98 MGD)
plant receives a relatively light industrial load (6% of flow, 18% of 5
day BOD) from glue processing, food processing and machinery
industries. Primary treatment 1s by sedimentation. The plant operated
1n a step feed configuration at a solids retention time of 4.3 days and
achieved partial nitrification.
EXPERIMENTAL DESIGN OF 24-HOUR STUDIES
The 24-hour studies conducted at each of the plants were designed to
acquire Information necessary to calibrate the IAWPRC model. The steady-
state model solution, based upon flow weighted average conditions, was
employed for calibration of key parameters, and these parameters were
subsequently used 1n the dynamic solution to simulate1 diurnal variations 1n
oxygen uptake rate. Comparison of the simulated and measured diurnal ranges
allowed assessment of the utility of the dynamic solution for design
purposes. Accordingly, 1t was necessary to obtain plant operating data
which would allow application of both the steady-state and dynamic
solutions.
The dynamic solution contained 1n the SSSP software package requires
Information on process Input concentrations and flows at Intervals which are
multiples of 15 minutes. It then uses linear Interpolation to determine the
values at the 15 minute sub-Intervals. The sampling program was designed to
develop the maximal amount of Information within the limits of time
available to the three person field study team. Key features of the
sampling program Included:
* activated sludge process Influent and clarified effluent samples at two
23
-------
hour Intervals analyzed for: total COD, soluble COD, ammonia, organic
nitrogen, nitrate, pH
* measurements at various positions 1n the aeration tank at hourly or
longer intervals for: oxygen uptake rate (OUR), dissolved oxygen (DO),
mixed liquor volatile suspended sol Ids (MLVSS), temperature
In addition, the wastewater and recycle flow rates and the aeration rate were
recorded at hourly Intervals. The model predictions require that the process
sol Ids retention time (SRT) be established. This was determined based upon the
sludge wasting practice 1n effect during the month preceding the 24-hour
study. Normally, the SRT was fairly constant and an average of the dally
values was used. However, at the Green Bay Plant, the sol Ids wasting
pattern varied considerably during the month preceding the study and the
transient SRT was determined following the technique outlined by Balllod et
al. (1977).
EXPERIMENTAL METHODS
Chemical Oxygen Demand (COD) was measured 1n the laboratory on preserved
samples by the dlchromate reflux method (Standard Methods, 1985).
Organic Nitrogen was measured 1n the laboratory on preserved samples by the
macro-kjeldahl method (Standard Methods, 1985)
Ammonia Nitrogen was measured 1n the field using an Orion, Model 95-10,
ammonia specific electrode.
Nitrate plus nitrite nitrogen was measured 1n the field following the
recommended Hach procedure using Hach Nitraver V and a portable Hach DR-2
spectrophotometer.
V
Dissolved Oxygen (DO) was measured 1n the field using a Yellow Springs
Instrument Company Model 54 dissolved oxygen probes. A submersible probe was
employed to determine 1n-s1tu DO values 1n the aeration tanks.
*
Filtration for soluble COD and MLVSS measurement was performed using glass
fiber filter papers as specified for suspended solids measurement (Standard
Methods, 1985).
Oxygen Uptake Rate
Oxygen Uptake Rate was measured 1n the field using a batch BOD bottle
technique.
a. A sample was withdrawn from the aeration tank using a weighted bucket and
quickly transported to the field laboratory.
b. The sample was contacted with pure oxygen for a period of 5 to 20 seconds
to raise the DO to about 10 mg/1.
24
-------
c. An aliquot of the oxygenated mixed liquor was transferred to a 300 ml BOO
bottle containing a magnetic stirring bar. A self-stirring YSI DO probe
was inserted, and the bottle was submerged in a 2 liter container of
mixed liquor to minimize temperature drift during the uptake measurement.
d. The two liter container was placed on a magnetic stlrrer, and both the
magnetic stlrrer and self-stirring probe were turned on to provide
mixing. The probe output of DO versus time was recorded using a Cole
Farmer Model 8376-30 strip chart recorder. The oxygen uptake record was
usually begun within two minutes of sample withdrawal from the aeration
tank.
e. The OUR was determined from the slope of the strip chart record. The
slope was determined from the earliest linear portion of the strip chart
record. Normally this was during the first three minutes of record.
Interpretation of Oxygen Uptake Rates Measured by. the BOD Bottle Method
The BOD Bottle Method described above has historically been used to
measure oxygen uptake rates in activated sludge. The method Itself gives an
accurate and precise 1n-v1tro measure of the oxygen utilization rate that
occurs 1n the BOD bottle. The problem, however, 1s that this rate may not
represent the in-situ oxygen utilization rate occurring 1n the reactor from
which the sample was withdrawn. Two conditions, oxygen limitation and soluble
substrate limitation, can cause the OUR measured by the BOD bottle method to
differ significantly from the 1n-s1tu value.
Oxygen limitation arises when the 1n-s1tu DO 1s near zero and causes the
in-situ oxygen uptake rate to be limited by the availability of oxygen. A
sample subjected to the BOD bottle method 1s exposed to high DO concentrations
and will respire at a higher rate. Consequently, for 1n-s1tu DO
concentrations near zero, the OUR Indicated by the BOD bottle method will be
greater than the 1n-s1tu OUR. Substrate limitation arises when the 1n-s1tu
exogenous substrate concentration 1s near zero and causes the OUR of a
withdrawn sample to decrease between the time of withdrawal and the time at
which the BOD bottle OUR 1s measured.
«
Mueller and Stensel (1987) compared 1n-s1tu OUR values estimated by non-
steady state process water tests with those Indicated by the bottle method at
several activated sludge plants. The results showed that, at low DO values,
the BOD bottle method produced Indicated OUR values which tended to be higher
than the 1n-situ values. At 1n-s1tu DO values below 1.5 mg/1, the Indicated
bottle uptake rates ranged from about 90% to 190% of the estimated 1n-s1tu
OUR. The results also showed that soluble substrate depletion could produce
Indicated bottle OUR values significantly lower (by as much as 50%) than the
estimated 1n-sltu values. This was most prevalent 1n measurements made on
completely mixed aeration tanks when the OUR was greater than about 250
mg/1/day. This was less prevalent 1n measurements made near the effluent end
of long narrow aeration tanks.
In this study, the following strategy was used to cope with the
Inaccuracies Inherent 1n using the BOD bottle method to estimate 1n-s1tu OUR
25
-------
values:
* Off-gas measurements made at 5 of the 6 plants were used to estimate the
process water volumetric mass transfer coefficients (
-------
SECTION 6
MODEL CALIBRATION
Calibration of the IAWPRC Model required that:
* the wastewater feed components required by the model be measured or
estimated from the data collected during the plant studies,
* the kinetic and sto1ch1ometric parameters required by the model be
estimated,
* criteria for calibrating the model to the plant data be established.
The IAWPRC Task Group suggested methods for establishing values for the
wastewater, stolchiometric, and kinetic parameters to be used 1n the model
(IAWPRC Task Group, 1986)(Ekama et al., 1986). However, most of these
methods are well suited only to laboratory pilot studies where the process
1s well controlled and significant effort can be spent on replicate
measurements to assess and Improve precision. The data 1n this report were
based on 24-hour studies at six operating, full-scale municipal wastewater
treatment plants. Although direct measurements of the wastewater feed
components and model parameters following the recommendations of the Task
Group were attempted, these techniques often were not practical for this
study (e.g. time was not available for replicate measurements) or did not
yield meaningful results (e.g. the anticipated step change 1n OUR required
for 85 measurement did not occur). Because of these problems, modified
methods were developed for measurement of feed components, and more reliance
was placed on model calibration for parameter estimation.
DETERMINATION OF FEED COMPONENTS
The 13 components considered 1n the IAWPRC Model are defined 1n Table 4.
Except for Component 7, partlculate products arising from blomass decay, all
of these components may be present 1n the feed wastewater. During the 24-
hour plant studies, samples of the wastewater fed to and clarified effluent
from the activated sludge processes were analyzed for total COD (TCOD),
soluble COD (SCOD), ammonia, total organic nitrogen (TON), and nitrate. The
following assumptions and logic were applied to the measured quantities to
estimate the concentrations of the components required by the model.
Several components from Table 4 were assumed to be absent 1n the feed:
27
-------
Component 5, XBH> active heterotrophlc biomass
Component 6, XBA. active autotrophlc biomass
Component 7, Xp, participate products arising from biomass decay
Component 8, SQ, dissolved oxygen
Component 9, SNQ, nitrate plus nitrite
Measured values were used directly for Component 10, ammonia. Component
13, alkalinity, was taken as the annual average value for the plant. If the
average annual value was not available, the default of 4 moles/m was used.
Lack of measured alkalinity values was not a serious drawback because the
model calibration was not Intended to Include alkalinity. Moreover, the
model does not include any relationships between alkalinity or pH and
reaction rates.
The remaining six components, consisting of four COD fractions and two
organic nitrogen fractions, were estimated from the measured data. It was
assumed that the process removed essentially all of the biodegradable COD,
so that the effluent soluble COD was taken as a measure of the Inert soluble
COD in the Influent, Sj « SCODeff.
The inert partlculate COD was assumed to account for between 10% and 20%
of the Influent partlculate COD,
Xj - (0.1 to 0.2) (TCOD1nf - SCOD1nf) (17)
The Task Group Report (IAWPRC, 1986) recommends that this component be
estimated to fit solids production (MLVSS) data. Simulated values of the
MLVSS concentration are sensitive to Xj, and higher percentages of the
Influent particulate COD (within the range of 10% to 20%) were selected as
required to improve agreement between the simulated and measured
concentrations.
The readily biodegradable COD, S$, was«est1mated 1n two ways. First, 1t
was assumed that U could be treated as 1f it were all soluble, thus,
SS - SCOD1nf - Si (18)
«
The second method (IAWPRC Task Group Report, 1986, Ekama et al., 1986) 1s
based on the observation that, following a batch feeding of activated sludge
with a wastewater containing both readily and slowly degradable substrates,
the oxygen uptake rate (OUR) 1s maintained at a relatively high, constant
level before abruptly declining to the lower endogenous level. The Increase
1n OUR 1s assumed to be caused by the uptake of S$ at the rate U. Combining
Equations 3 and 4 relates the Increase 1n OUR to the substrate uptake rate
In terms of the yield coefficient as,
A r0 A OUR /XBH
U . ----- . ------------- (19)
i - YH i - YH
28
-------
If the OUR is maintained at an elevated level for a time, t, before
declining to the endogenous level, the readily biodegradable substrate, S$,
utilized, 1s given by
A OUR t
SS - U XBH t - ---------- (20)
i - YH
Although the second method based on the batch reactor oxygen uptake data was
applied to each of the six plants studied, only the Green Bay, Jones Island,
and Portage Lake plants showed the anticipated step decrease 1n OUR and
allowed estimation of 5$ by this method. The other plants showed a gradual
decrease 1n OUR and did not give meaningful results. Consequently, $5 was
estimated by both methods for the Green Bay and Jones Island Plants and a
separate model calibration was performed for each feed fractlonatlon.
Because both methods gave nearly Identical (within 5%) 5$ values for the
Portage Lake Plant, only the first method, based on COD fractions, was used
for the calibration. In all cases, the slowly degradable COO was determined
by difference,
XS = TCOD1nf - Sj - SS - Xj (21)
Similar logic was used to fractionate the organic nitrogen 1n the feed
into four components (soluble biodegradable, SND» partlculate biodegradable,
XNQ, soluble inert, SMI. and partlculate Inert, XNI). It was assumed that
the soluble effluent organic nitrogen consisted of Inert substances. Thus,
the Inert soluble organic nitrogen 1n the Influent, SNI, was estimated by
multiplying the total effluent organic nitrogen by the ratio of the
effluent soluble COD to effluent total COD. The Inert partlculate organic
nitrogen In the Influent, XNI, was estimated by multiplying the total
organic nitrogen 1n the Influent by the ratio of Inert partlculate COD 1n
the Influent to total COD 1n the Influent. Thus, the Influent biodegradable
organic nitrogen was determined by subtracting the estimated values of the
Inert soluble and Inert partlculate organic nitrogen from the measured
value of the total organic nitrogen. This was subdivided Into soluble and
partlculate fractions by assuming that the organic nitrogen was divided 1n
the same proportion as the COD.
5$
- ----- ( TON1nf - SNI - XNI ) (22)
The partlculate biodegradable organic nitrogen 1n the Influent was
determined by difference,
XND - T°Nmf - SNI - XNI - SND (23)
DETERMINATION OF MODEL PARAMETERS
The effort here focused on the parameters to which the model simulations
29
-------
are sensitive. To verify the predictive capability of a model, parameter
estimates should be based on direct measurements. However, the IAWPRC model
is based on fundamental Sto1ch1ometr1c, kinetic, and conservation principles
that have been previously verified. Furthermore, 1t was pointed out earlier
that, for these field studies, direct measurement was not practical.
Consequently, sensitive parameters were determined by calibrating the model
to the data. Parameters to which the model predictions were Insensitive
were determined from the literature and set at constant values.
This study focused mainly on the oxygen utilization aspects of the
model. However, the model simulates many other system responses (e.g. mixed
liquor volatile solids, dissolved oxygen, nitrate) 1n addition to oxygen
uptake rate. Calibration of the model using only oxygen uptake rate data
could be misleading. Consequently, an effort was made to calibrate the
model so that it simulated MLYSS, dissolved oxygen, nitrate, and ammonia as
well as oxygen uptake rate. Primary emphasis was placed on matching the
oxygen uptake rate and DO. Lesser weight was placed on matching the MLVSS
and nitrogen concentrations.
Sensitivity Analysis
Table 6 shows sensitivity coefficients for simulation of OUR, DO,
MLVSS, ammonia, and nitrate to changes in key parameters. The parameters
listed are the ones to which the simulations were most sensitive. This
table was developed based on the simulation of the first reactor at the
Monroe Plant for the default parameter estimates. The sensitivity
coefficients were determined by changing each parameter by +20% from the
default value while holding all other parameters fixed at their default
levels. The dimenslonless sensitivity coefficient 1s then expressed as
percentage change 1n simulated value
Sensitivity Coefficient -
percentage change 1n key parameter
Positive sensitivity coefficients Indicate an Increase 1n the simulated value
with an Increase 1n the parameter value, whereas negative sensitivity
coefficients indicate a decrease 1n the simulated value with an Increase 1n
the parameter value. Specific values of the sensitivity coefficients depend
on features unique to the system being simulated. Table 6 pertains to reactor
1 of the Monroe Plant, and, because this plant had very high mixed liquor DO
values, the simulated values were not sensitive to changes In either K^a or
the half saturation constants for DO. Table 6 Indicates that, of the blomass
parameters, the OUR 1s most sensitive to the heterotrophlc and autotropMc
yield and decay coefficients. Overall, the heterotrophlc yield, YH, and
decay, bn, showed the most Impact on the simulated values of OUR, DO, and
MLVSS. The autotrophlc yield, Ya, decay, ba, and maximal specific growth
rate, uam, had the most Impact on simulated values of ammonia and nitrate.
The last line in Table 6 shows the Impact of the sol Ids retention time (SRT),
6C. Because SRT is an operational parameter and not a blomass parameter, 1t
1s shown separately. Values of SRT were not estimated as were the blomass
parameters but were known from the operational data at each of the plants
30
-------
TABLE 6. SENSITIVITY OF MODEL SIMULATIONS TO KEY PARAMETERS
Key
Parameters
Phm
YH
bh
Mam
Ya
ba
8c
Sensitivity Coefficient for
OUR
+0.01
-1.1
+0.1
+0.1
-0.2
-0.05
+0.3
DO
-0.1
+0.7
-0.8
-0.1
+0.1
0.0
0.0
MLVSS
0.0
+1.8
-0.3
0.0
+0.04
-0.01
+0.6
Simulation of
SNH
0.0
0.0
+0.4
-1.2
+0.4
+.8
-0.5
SNO
-0.03
-0.5
-0.1
+0.1
-0.6
-0.1
+0.15
studied. Table 6 Indicates that the simulated values are sensitive to SRT,
and, therefore, 1t 1s Important that values of SRT be known accurately and
precisely for the model simulations.
Parameter Estimation
The model parameters were estimated by calibrating the model to the
average plant data measured during the 24-hour studies. The procedure was
as follows:
1. Flow weighted average concentrations of the1 feed components
required by the model were determined following the procedure
described above.
2. The process flow diagram was modeled as realistically as possible
using a combination of completely mixed reactors and perfect
Clar1f1er/th1ckeners. Tank volumes and flow rates were
established based on plant data. Generally, preliminary decisions
on reactor configurations for modeling were made 1n the field as
data were collected. Further refinements took place during data
analysis. The following criteria were collectively applied to
divide the aeration tanks Into completely mixed reactors for
modeling:
a) Variation of OUR and DO along the tank. When OUR and DO
changed appreciably along the length of an aeration tank,
31
-------
efforts were made to divide the tank Into sections so that
the actual values were within approximately 10% of the
average value.
b) Points of feed addition. In cases where primary effluent was
added at points along the aeration tank, 1t was convenient
and logical to treat portions of the tank as separate
reactors.
c) Aeration pattern. When tapered aeration was employed, 1t was
convenient and logical to divide the tank Into sections based
on aeration Intensity.
d) Length/width ratio. Long tanks were divided Into Individual
reactors 1n an effort to keep the length/width ratio of the
individual reactors at 5 or less.
3. A steady-state simulation was performed using the SSSP Software
package with default values of all parameters. Table 7 summarizes
the default values and approximate ranges (IAWPRC Task Group,
1986) for the parameters required by the model. A second steady-
state simulation was performed using temperature adjusted default
values for the kinetic parameters contained 1n the model. The
expression,
kinetic parameter at T2 (T2-Ti)
. 1.02 (24)
kinetic parameter at TI
was used to adjust all kinetic parameters to the average measured
process temperature at each plant. The coefficient 1.02 1s
typical for the activated sludge process (Metcalf & Eddy, 1979).
4. Measured average values of reactor OUR, 00 and MLVSS as well as
effluent values of nitrate and ammonia were^calculated for each of
the plants.
5. The model simulated values of OUR, DO, MLVSS, nitrate and ammonia
based on the temperature adjusted default parameters were compared
with the measured average values. The model was then calibrated
to the measured data by adjusting selected key model parameters.
These adjustments were based on the sensitivity coefficients
shown in Table 6, and the adjusted parameters were kept within the
approximate range Indicated 1n Table 7. Except for the Green Bay
Plant, calibration adjustments were limited to the heterotrophic
yield and decay parameters and the autotrophlc maximal specific
growth rate. In addition, the autotrophlc half saturation
constant for DO was made equal to the default value of the
heterotrophic half saturation constant for DO (0.1 mg/1). The
SSSP Software package requires that these two half saturation
constants be equal when the simulation 1s based on given values of
32
-------
TABLE 7. SUMMARY OF DEFAULT VALUES AND APPROXIMATE RANGES FOR MODEL
PARAMETERS
gyag^g;s;sagagss~T?^rMMs»g^wwggg»Jsg3KgMM»M^afttM»aMK>igMaiwa^M3BiKwa»»»g»w>ia»MttaEB^ga
Heterotrophlc Parameter Default Approximate
Value Range
Maximal Specific .
Growth Rate, unm, d" 4.0 3 to 13
Half Saturation Constants
Soluble Substrate, K$, TO coo/1 10. 10 to 180
Dissolved Oxygen, KQH. mg/1 0.1 0.1 to 0.15
Nitrate, KNQ, mg N/l 0.2 0.1 to 0.2
Yield Coefficient, YH, g COD/g COD 0.67 0.46 to 0.69
Decay Constant, bn, d"1 0.62 0.13 to 4.2
Anoxic Growth Factor, n.G« 0.8 0.6 to 1.0
Hydrolysis Rate, k^, d" 2.2
Hydrolysis Saturation Ratio,
KX, g COD/g COD 0.15
Anoxic Hydrolysis Factor,nn 0.4
Anmonification Rate, KA» 1/mg COD/d 0.16
Fraction of Biomass Yielding
Particulate Products (Xp), fp 0.08
Fraction N in Biomass, IXB» 9 N/g COD 0.086
Fraction N in Xp, 1Xp, g N/g COD 0.06 - ,
Autotrophic Parameter
Maximal Specific Growth Rate, Ham, d'1 0.65 0.34 to 0.65
Half Saturation Constants
Ammonia, KNH» mg N/l 1.0
Dissolved Oxygen, KQA. TO/1 1.0 0.1 to 2.0
Yield Coefficient, YA, mg COD/mg N 0.24 0.07 to 0.28
Decay Constant, DA, d"1 0.12 0.05 to 0.15
33
-------
K|_a 1n each reactor. The average DO saturation value for the tank
was either determined from previous clean water oxygen transfer
tests conducted at the plant or calculated from Equation D.2 of
the ASCE Clean Water Standard (ASCE, 1984) with an effective
saturation depth equal to one-third of the tank depth.
6. A dynamic simulation was performed using the SSSP Software package
with the calibrated parameters determined from the steady state
simulation. This gave simulated diurnal profiles for OUR and DO.
These were then compared with the measured profiles to judge the
utility of the model for simulating reasonable variations 1n
oxygen utilization.
34
-------
SECTION 7
RESULTS OF PLANT STUDIES AND MODEL SIMULATIONS
DATA DIRECTORY
The field study data and model calibration results for each of the six
plants studied are presented and summarized in four types of tables. These
are:
Plant Process Summary Tables: These contain information describing the
wastewater characteristics and process configuration for each of the
plants studied. (Tables 9, 11, 14, 16, 18, and 21)
Model Calibration Tables: These summarize and compare the average
measured and steady-state simulation values of OUR, DO, ammonia,
nitrate, and MLVSS along with the default, temperature adjusted
default, and calibrated values of the model parameters. One model
calibration table is given per plant, except for the Green Bay and
Jones Island Plants where two different methods of feed COD
fractionation resulted in two model calibration tables for each plant.
(Tables 10, 12, 13, 15, 17, 19, 20, and 22)
Raw Data Tables: These contain the influent, effluent, and reactor
data obtained for each plant during the 24-hour studies. They appear
in Appendix A as Tables Al, A2, A3, A4, A5, and A6.
SSSP Steady State Simulation Tables: These summarize the essential
Input and output Information for the steady-state simulations performed
with the calibrated parameters. They appear in Appendix B as Tables
Bl, B2, B3, B4, B5, and 86.
SUMMARY OF MODEL CALIBRATION RESULTS
It was possible to obtain reasonable agreement between the average
measured and steady-state simulated values of OUR, DO, MLVSS, nitrate, and
ammonia at most of the plants studied. This agreement was achieved by
calibrating only two out of three key parameters. The other parameters were
set at their temperature adjusted default values. Notable lack of agreement
between the simulated and measured values was evidenced at the Jones Island
Plant where the simulated and measured effluent ammonia concentrations did
not agree well, and at the South Shore Plant where the simulated and
measured MLYSS values did not agree well.
35
-------
Table 8 summarizes the values of the key parameters which resulted from
model calibration. The key parameters were: the heterotrophlc yield
coefficient, YH, the heterotropMc decay constant, b\\, and the autotrophic
maximal specific growth rate, uam.
The Green Bay plant differs from the others 1n that 1t is dominated by
the heavy industrial load (pulp and paper). The soluble COD concentration
in the feed to the activated sludge process at Green Bay was 718 mg/1. This
was more than four times the average of the soluble COD concentration fed to
the other five plants. Nevertheless, calibration was achieved by adjusting
only two parameters when the feed fractions were based on COD data.
TABLE 8. SUMMARY OF MODEL CALIBRATION
All Parameters Were Set at Temperature Adjusted
Default Values Except for:
Plant Heterotrophic Heterotrophlc Autotrophic
Yield Decay Maximal Specific
Coefficient Constant Growth Rate
g COD/ g COD 1/d 1/d
Default (20°C)
Madison
Monroe
Portage Lake
Jones Island
South Shore
Green Bay
0.67
0.57
d
0.69
d
0.46
0.57
d signifies
0.62
d
0.80
1.10
0.35
0.50
d
temperature adjusted default
0.65
0.34
0.34
d
0.62 "
d
0.45
value
Interpretation of Steady State and Dynamic Simulations
In the following sections, steady state simulations are employed to
calibrate the model parameters to the operating data collected at the six
plants. The calibrated model 1s used to simulate dynamic behavior at three
of the plants. In comparing the "calibrated" simulated values of DO, OUR,
NLVSS, nitrate, and ammonia to the measured values perfect agreement should
not be expected. Four reasons for this are: (1) the Initial conditions
upon which the simulation 1s based; (2) the fact that the plant surveys
typically examined only one of several parallel aeration tanks contained In
the total plant; (3) Inaccuracies Inherent 1n using the BOD bottle uptake
36
-------
method to measure 1n-s1tu OUR; and (4) changes 1n the process water value of
To perform either a steady-state or dynamic simulation, the program must
begin the simulation from a known starting or initial condition. The SSSP
Software assumes that the flow and concentration values (1n the case of the
steady-state solution) or pattern (in the case of the dynamic solution) are
constant from day to day. That 1s, the simulation assumes that the plant
experienced conditions Identical to the measured study conditions during the
previous day, previous week, and previous month. In reality, inputs to
municipal wastewater plants vary markedly with day of the week and season of
the year. This will account for some lack of agreement between the measured
and simulated values.
In the smaller plants studied, e.g. Portage Lake or Monroe, the aeration
tanks studied accounted for one-third to one-half of the total blomass
inventory and flow treated. However, in the large plants, such as Jones
Island or South Shore, the aeration tanks studied accounted for only 4% to
5% of the biomass inventory and flow. Thus, the measurements in the
smaller plants could be expected to be more representative of the average
plant conditions.
In Section 5, 1t was pointed out that oxygen limitations arising when
the 1n-situ DO value is near zero can cause the BOD bottle OUR method to
yield erroneously high values (Mueller and Stensel, 1987). The simulations
were based on values of the process water K^a measured by the off gas
technique. For conditions where the simulated and measured DO values are
equal, the simulated OUR value 1s derived entirely from the measured values
of the process water K|_a. Ifc 1s subsequently shown that, for low DO
conditions at the head of long aeration tanks, the SSSP Software correctly
simulates 1n-s1tu OUR values much lower than the in-vitro values measured by
the BOD bottle method.
It was also pointed out in Section 5 that soluble substrate limitations
arising in completely mixed tanks could cause OUR values measured, by the BOD
bottle method to be somewhat lower than the true 1n-s1tu values. For these
conditions, the model simulations were generally slightly higher than the
OUR values measured by the bottle method.
At five of the six plants, K|_a values based on off gas measurements
(Redmon et al. 1983) were Incorporated Into the simulations. However, the
process water K|_a Is not constant and may vary with time of day. Thus, even
though the K(,a measurements were sometimes made on the same day as the plant
tests, the dally average process water K|_a could have been different from
the measured values. These differences would be reflected as differences
between the measured and simulated OUR and DO values.
Portage Lake Plant
The plant process summary for the Portage Lake Plant 1s given 1n Table
9, the aeration tank configuration 1s shown 1n Figure 4, and the process
flow diagram as modeled 1s shown 1n Figure 5. Table 10 shows the model
37
-------
Table 9. PORTAGE LAKE PLANT PROCESS SUMMARY
= = = = = = = = = = = = = = i: = t3 = = *i: = = = = = = = = = = = = = = = « = = = = = = = = = = = = = = = = = = = = = = = :
Average Dally Flow: 8,700 m'/d (2.3 MGD)
Average Raw Influent BOO,: 150 mg/L
Major Industrial Contributors: none
Primary Treatment: coarse screening
Nominal SRT at Time of Study: 10.0 days
Recycle Ratio at Time of Study: 1.03
Fraction of Flow Treated by Aeration Basin Studied: 0.50
Process Configuration: This contact stabilization plant consists of 2
separate circular modular units, each of which Includes a contact tank,
reaeratlon basin, clarifier, and aerobic digestor. Both units were in
operation during the study.
Prinopy
Effluent
• Indicate location of neasurenents
for OUR, Da MLSS/MLVSS. t Tenp.
Rli Reactor I, Reaeratlon Basm <530 n3>
R2i Reactor 3, Aeration Basin (350 n3>
Depth 'of nixed liquor m Aeration Basin * 4.7 n
Depth of nixed liquor In Reaeratlon Basm = 3.8 n
SCALD 1 cm * 6 n
WAS
Secondary
Effluent
Figure 4. Portage Lake Plant plan sketch of'basin one.
Primary
Effluent
Secondary
Effluent
Avg. Primary Effluent Rewrote = 3547 n3/day
Avg. RAS Flowrate = 3653 r»3/day
NO SCALE
Figure 5. Portage Lake Plant process flow diagram as modeled.
38
-------
calibration. For this plant, the default and temperature adjusted default
values of the parameters produced remarkably good agreement between the
measured and simulated values of OUR, DO, and MLVSS. The heterotrophlc
yield and decay coefficients were adjusted slightly to Improve the agreement
between the measured and simulated nitrate concentrations. The simulated
values of the OUR tend to be greater than the measured values. However, for
the well mixed contact tank, substrate depletion may have caused the BOD
bottle method to underestimate the 1n-s1tu OUR.
Figure 6 compares the measured and dynamically simulated values of OUR
and DO for the Portage Lake contact tank. The agreement between the OUR
values is surprisingly good. However, the DO comparison suggests that the
model may simulate changes before they actually occur. Nevertheless, the
model did simulate the range and trend of OUR values and the range of DO
values.
Green Bay Plant
Table 11 gives the Green Bay Plant Process Summary. Figure 7 shows the
aeration tank configuration and Figure 8 shows the process flow diagram as
modeled. A notable feature 1s that 50% of the BOD results from paper mill
wastes. The model calibration shown in Table 12 Indicates that the default
and temperature adjusted default simulations gave values of OUR close to the
measured values, and values of MLVSS somewhat higher than the measured
values. Values of ammonia were considerably out of agreement with the
measured values. In addition, the simulation results showed considerable
den1tr1f1cation. The simulated MLVSS values were reduced by decreasing the
heterotrophic yield, and the simulated ammonia values were Increased by
decreasing the autotrophlc maximal specific growth rate.
The average of the dally SRT values during the previous month was 8.3
days. However, this was Influenced by a week of very high (13d to 62d)
daily SRT values which occurred early 1n the preceding month. During the
ten days preceding the field study, the dally SRT values ranged from 2.4 to
3.1 days. The actual transient SRT at the time of the field study was
estimated as 3.1 days following the method given by Balllod et al. (1977).
Green Bay was one of the two plants 1n which the method based on the
Increase and subsequent decrease 1n OUR observed 1n a batch reactor
following feeding gave meaningful results for estimation of readily
degradable substrate (IAWPRC Task Group, 1986). Table 13 shows the model
calibration for the feed COO fractlonatlon based on the batch reactor data.
Although the same values of the model parameters were used for both Tables
12 and 13, differences 1n the feed fractions (Table 12 1s based on S$ - 510,
and X$ « 278, whereas Table 13 1s based on 85 • 253 and X$ • 536.) caused
slight differences 1n the simulated values.
In both Tables 12 and 13, the simulated values of the OUR are
considerably greater than the measured values. However, these measurements
are 1n a large we11-mixed tank, a condition that would be conducive to
substrate depletion which would cause the BOD bottle method to give low
measured values for OUR. Moreover, the good general agreement between the
39
-------
TABLE 10. MODEL CALIBRATION FOR THE PORTAGE LAKE PLANT
SSSP STEADY-STATE SIMULATION
SRT -10.6 days
Temperature « 18.0 C
SIMULATED VALUES
Measured Default w/
Values Default Temp. Adj. Calibrated
OUR (rag 02/L/day):Contact Tank 598 600.6
DO (mg/L):
Contact Tank
Reaeratlon Tank
Effluent Ammonia (mg-N/L)
Effluent N03/N02 (mg-N/L)
MLVSS (iig COD/L): Contact Tank
K(L)a (I/day): Contact Tank
Reaeratlon Tank
0.5
1.5
2.6
4.5
2361
72
72
0.6
0.6
1.9
0.0
2438.6
72
72
663.6
2.0
1.6
0.9
0.0
2344.9
72
72
665.2
1.2
1.5
0.7
1.7
2156.7
72
72
VALUES OF MODEL PARAMETERS USED FOR SIMULATIONS
H
E
T
E
R
0
T
R
0
P
H
I
C
A
U
T
0
u max
Ks COO
Ks 02
yield
b decay
1/d
mg COD/L
mg 02/L
mg cells/mg COD
1/d
anoxlc growth factor
Ks N03
hydrolysis rate
hydrol satur ratio
mg-N/L
1/d
g COD/g COO
anoxlc hydrol factor
a mmon1f1 cation
frac. part. prod.
N 1n blomass
N 1n part. prod.
02 saturation cone
u max
Ks NH-H
Ks 02
yield
b decay
L/ng COO/d
mg COO/ma COO
mg-N/mg COD
mg-N/mg COD
mg-02/L
1/d
mg-N/L
mg-02/L
mg cells/mg COD
1/d
4.000
10.0
0.55
0.67
0.620
0.8
0.2
2.200
0.150
0.4
0.160
0.080
0.086
0.060
9.0
0.650
1.000
0.550
0.240
0.120
3.845
9.612
0.596
2.115
0.156
0.154
11.0
0.625
-
3.845
9.612
0.10
0.69
1.100
2.115
0.156
0.154
IP.'O
0.625
0.10
measured and simulated DO values (I.e. the measured and simulated DO driving
force for mass transfer) means that the OUR simulation 1s 1n agreement with
the off-gas measurements upon which K|_a 1s based.
Madison Plant
The Madison Plant Process Summary 1s shown 1n Table 14. Figure 9 shows
the configuration of the basins studied and Figure 10 shows the process flow
diagram as modeled. A distinguishing feature of this plant 1s the long SRT
(16.4 days) and resulting high degree of nitrification. The plant was modeled
as four reactors 1n series, and only reactors 3 and 4, where the average DO
values were above 3 mg/1, were used to calibrate the model. Table 15
Indicates that the temperature adjusted default parameter values gave high
40
-------
1000.0
Model Simulated, Contact Tank
BOO Bottle Method, Contact Tank
6.0
12.0
Time (hours)
o
o
£ 3.0
*
o
o
t.o
O 1.0
0.0
Simulated. Contact Tank
Measured, Contact Tank
o.o
4.0
a.o 12.0 ie.o
Time (hours)
20.0
24.
Figure 6. Portage Lake Plant dynamic simulation: comparison of measured
and simulated OUR and DO values.
41
-------
Table 11. GREEN BAY PLANT PROCESS SUMMARY
Average Dally Flow: 182,000 m'/d (48 MGO)
Average Raw Influent BODS: 375 mg/L
Major Industrial Contributors: paper mills
Fraction of Flow: 0.30
Fraction of BOD»: 0.50
Primary Treatment: sedimentation
Nominal SRT at Time of Study: 3.1 days
Recycle Ratio at Time of Study: 0.66
Fraction of Flow Treated by Aeration Basin Studied:
0.33
Process Configuration: This plant consists of 4 parallel contact
stabilization processes, each of which Includes an aeration (contact) basin,
approximately 74 m. (245 ft.) long and 22 m. (74 ft.) wide. Three of these
processes were In operation during the study.
Pnwory
EffliwnV
Bostt
ftoocrtor 1 (3607 n3>
Atration Basin
Rtactor Z O0443 n3>
RAS
Mxcd Liquor
>
JL>VAS
Air
1
JL
R2
1 ! Secondary
^^
j L»VAS
Air
j ^ tttiuent
Ava Prmary Effluent Flowrate » 58879 r»3/day
AVQ. RAS Flowratt « 38840 *3/day
NO SCALE
Figure 8. Green Bay Plant process flow diagram as modeled.
42
-------
TABLE 12. MODEL CALIBRATION FOR THE GREEN BAY PLANT
FEED FRACTION BASED ON COD FRACTIONS
SRT » 3.1 days
Temperature - 26.5 C
SIMULATED VALUES
Measured Default w/
Values Default Temp. Adj. Calibrated
OUR (ing 02/L/day) Contact Tank 1840
DO (rng/L):
Reaeratlon Tank
Contact Tank
0.3
1.5
30.4
Effluent Ammonia (mg-N/L)
Effluent N03/N02 (mg-N/L)
MLVSS (ng COD/L): Contact Tank 2792
K(L)a (I/day): Reaeratlon Tank 118
Contact Tank 311
1752
0.4
3.3
19.6
0
3197
118
311
1879
0.4
3.6
13.5
0.2
3066
118
311
2201
0.1
2.5
32.2
0
2446
118
311
VALUES OF MODEL PARAMETERS USED FOR SIMULATIONS
H
E
T
E
R
0
T
R
0
P
H
I
C
V
A
U
T
0
y iax
Ks cod
Ks 02
yield
fa decay
1/d
mg COD/L
mg-02/L
mg cells/mg COD
1/d
anoxlc growth factor
Ks N03
hydrolysis rate
hydrol satur ratio
mg-N/L
1/d
g COO/g COO
anoxlc hydrol factor
ammonlflcatlon
frac. part. prod.
N In blomass
N In part. prod.
02 saturation cone
y max
Ks NH-N
Ks 02
yield
b decay
L/mg COO/d
mg COO/mg COD
mg-N/mg COD
mg-N/mg COD
mg 02/L
1/d
mg-N/L
mg-02/L
mg cells/mg COO
1/d
4.000
10.0
0.55
0.67
0.620
0.8
0.2
2.200
0.150
0.4
0.160
0.080
0.086
0.060
9.0
0.650
1.000
0.550
0.240
0.120
4
11
0
2
0
0
0
.549
.374
.705
.502
.132
.182
9.7
.739
4.
11.
0.
0.
0.
2.
0.
0.
9.
0.
0.
549
374
1
57
705
502
132
182
7
45
1
simulated values for DO and MLVSS, and low simulated values for OUR.
Calibration required that the yield coefficient be decreased to Increase the
OUR and decrease the MLVSS. Simultaneously, the maximal autotrophlc specific
growth rate was decreased. The simulated DO values agreed fairly well (within
0.5 mg/1) with the measured values, whereas the simulated OUR values for
Reactors 3 and 4 were 5% to 10X below the measured values.
Figure 11 shows the dynamic simulation of the OUR and DO values along with
the measured values for Reactors 3 and 4. Except for the last four hours of
the study, the Measured and simulated values agree remarkably well. Moreover,
the model accurately simulated the range of diurnal DO change 1n Reactors 3
and 4. The simulated and measured profiles of OUR and DO along the tank
length are given 1n Figure 12. The OUR profile clearly shows the simulated
values 1n Reactors 1 and 2 to be well below the corresponding values measured
by the BOD bottle method. The general agreement between the simulated and
43
-------
TABLE 13. MODEL CALIBRATION FOR GREEN BAY PLANT WITH
FEED FRACTIONS BASED ON BATCH REACTOR DATA
SRT - 3.1 days
Temperature » 26.5 C
SIMULATED VALUES
Measured Default w/
Values Default Temp. Adj. Calibrated
OUR (mg 02/L/day) Contact Tank 1840
DO (mg/L):
Reaeratlon Tank
Contact Tank
Effluent Ammonia (mg-N/L)
Effluent N03/N02 (mg-N/L)
0.3
1.5
30.4
MLVSS (mg COO/L): Contact Tank 2792
K(L)a (I/day): Reaeratlon Tank 118
Contact Tank 311
1452
0.3
4.2
21.4
0
3578
118
311
1599
0.3
4.4
19.1
0.1
3365
118
311
2201
0
2.5
32.7
0
2439
118
311
VALUES OF MODEL PARAMETERS USED FOR SIMULATIONS
H
E
T
E
R
0
T
R
0
P
H
I
C
A
U
T
0
u max
Ks cod
Ks 02
yield
b decay
1/d
mg COO/L
mg-02/L
mg cells/mg COO
1/d
anoxlc growth factor
Ks N03
hydrolysis rate
hydrol satur ratio
mg-N/L
I7d
g COD/g COD
anoxlc hydrol factor
anmonlflcatlon
frac. part. prod.
N In blomass
N 1n part. prod.
02 saturation cone
u max
Ks NH-N
Ks 02
yield
b decay
L/Mg COD/d
mg COO/nig COO
mg-N/mg COO
mg-N/mg COO
•g 02/L
1/d
mg-N/L
mg-02/L
in cells/mg COO
1/d
4.000
10.0
0.55
0.67
0.620
0.8
0.2
2.200
0.150
0.4
0.160
0.080
0.086
0.060
9.0
0.650
1.000
0.550
0.240
0.120
4
11
0
2
0
0
0
.549
.374
.705
.502
.132
.182
9.7
.739
4
11
0
0
0
4
0
0
9
0
0
.549
.374
.1
.57
.705
.110
.132
.182
.7
.45
*
.1
Measured DO values 1n Reactors 1 and 2 Indicates that the OUR 1s being
correctly simulated based on the measured K^a, and that the apparent
discrepancy between the simulated and measured OUR values 1n Reactors 1 and 2
is an artifact of 1n-s1tu oxygen limitations on the BOD bottle uptake rate
Measurement method.
Monroe Plant
The Process Summary for the Monroe Plant 1s given 1n Table 16. Figure 13
shows the configuration of the aeration basin studied and Figure 14 shows the
process flow diagram as modeled. Distinguishing features of this plant are
the dominant Industrial contributions of the brewing, dairy, and food
processing Industries and the ln-11ne, aerated flow equalization basin. This
basin functioned to reduce the high soluble COD and to Increase the
participate COD 1n the wastewater fed to the activated sludge process. Another
44
-------
Table 14. MADISON PLANT PROCESS SUMMARY
Average Daily Flow (East and West Plants): 151,000 m*/d (40 MGD)
Average Raw Influent BOD,: 170 mg/L
Major Industrial Contributors: meat and cheese processors
Fraction of Flow: 0.06
Fraction of BOD,: 0.15
Primary Treatment: sedimentation
Nominal SRT at Time of Study: 16.4 days
Recycle Ratio at Time of Study: 0.67
Fraction of West Plant Flow* Treated by Aeration Basin Studied: 0.45
* Average flow for the West Plant during the study was 45,370 m*/d (11.99
MGD).
Process Configuration: The plant 1s divided into two sub plants, East and
West. The West Plant, 1n turn 1s divided Into Units 3 and 4. This study was
conducted on Unit 3 which Includes 2 three-pass aeration tanks, with each pass
80.8 m. (265 ft.) long and 9.2 m. (30 ft.) wide. Effluent from both tanks of
Unit 3 1s combined before clarification.
Primary
Effluent
X
A >
\
RAS
ABC D
* * t *
Reactor 1 (1733 n3> 1 Reactor Z (4008 n3> X
/'Reactor 3 (3827 n3>lp *
\ JH I*
~Q I Reactor 4 (1914 n3>
E
\ Mixed
^ Liquor
• Indicate location of measurements for OUR,
Da MLSS/MLVSS S, Temp.
Depth of mixed liquor In Aeration Basins = 5.2 m
SCALD 1 cm = 10 n
Figure 9. Madison Plant plan sketch of basins 22, 23 and 24.
Primary
Effluent —
RAS
->
v
/
o£
m
^
/
o4
R2
*
/
J,
R3
^
/
<=L
R4
i i Secondary
1 1 > Effluent
\/
T«— >VAS Tl— >VAS 1'— >VAS T L_^.vAS
Air Air Air Air
Avg. Prlnary Rowrate * 20346 n3/day
Avg. RAS Flowrate = 13552 m3/day
NO SCALE
Figure 10. Madison Plant process flow diagram as modeled.
45
-------
TABLE 15. MODEL CALIBRATION FOR THE MADISON PLANT
SIMULATED VALUES
iKi » 10. «» aays
Temperature » 20.5 C
OUR (ng 02/L/day) Reactor 3
Reactor 4
DO (mg/L) Reactor 3
Reactor 4
Effluent Ammonia (mg-N/L)
Effluent N03/N02 (mg-N/L)
Measured
Values
447
363
3.0
3.9
0.2
12
Default
337.4
302.4
2.1
2.8
0.2
6.3
Default w/
Temp. adj.
330.7
295.4
3.7
4.5
0.2
8.1
Calibrated
404.8
341.3
2.4
3.4
0.6
11.7
MLVSS (ng COD/L) Reactor 1
2194 2280.1
2263.5
1897.8
K(L)a (I/day):
Reactor 1
Reactor 2
Reactor 3
Reactor 4
45-55*
68-77*
43-50*
37-51*
45
68
50
51
45
68
50
51
45
68
50
51
VALUES OF MODEL PARAMETERS USED FOR SIMULATIONS
H
E
T
E
R
0
T
R
0
P
H
I
C
A
U
T
0
u max
Ks COD
Ks 02
yield
b decay
1/d
mg COD/L
mg 02 /L
mg cells/Rig
1/d
anoxlc growth factor
Ks H03
hydrolysis rate
hydro 1 satur ratio
mg-N/L
I7d
g COO/g COD
anoxlc hydrol factor
anmonlfl cation
frac. part. prod.
N In blomass
N 1n part. prod.
02 saturation cone
u max
Ks NH-N
Ks 02
yield
b decay
L/mg COD/d
•g COD/mg CO
mg-N/mg COO
mg-N/mg COO
mg 02/L
1/d
mg-N/L
mg 02/L
mg cells/mg
1/d
4.00
10.0
0.55
0.67
0.620
0.8
0.2
2.20
0.150
0.4
0.160
0.080
0.086
0.060
9.0
0.650
1.000
0.550
0.240
0.120
4.04
10.1
0.626
2.222
0.149
0.162
10.5
0.656
•*
4.04
10.1
0.10
0.60
0.626
2.222
0.149
0.162
10,. 5
0.34
0:10
notable feature Is the high DO levels (greater than 6 mg/1) maintained 1n the
shown 1n Table 17 gave simulated values of OUR which were somewhat high 1n
zone 1 and somewhat low In zone 2. Agreement between the simulated and
measured values was Improved by slightly Increasing the heterotrophlc decay
rate and decreasing the autotrophlc maximal specific growth rate.
The dynamic simulation given 1n Figure 15 shows good general agreement
between the measured and simulated values. Moreover, the 1n-11ne equalization
basin effectively attenuated the diurnal variations.
Jones Island Plant
Table 18 shows the process summary for the Milwaukee Jones Island East
Plant. Figure 16 shows the configuration of the tank studied and Figure 17
46
-------
BOO.O
—» 700.0
cd
15 eoo.o
O
600.0
400.0
3
0 800.0
a
£ 200.0
X
O 100.0
0.0
Simulated, Reactor 3
Simulated, Reactor 4
-+- Measured, Reactor 3
-»- Measured, Reactor 4
0.0
4.0
6.0 12.0 16.0
Time (hours)
O
6.0
4.0
X
O
O
«
*
2.0
1.0
0.0
Simulated, Reactor 3
Simulated, Reactor 4
Measured, Reactor 3
Measured, Reactor 4
12 16
TIME (hours)
20.0
24.
20
24
Figure 11. Madison Plant dynamic simulation: comparison of measured and
simulated OUR and DO values.
47
-------
1400.0
- 1200.0
3
o
1000.0
— 800.0
O
aa
a.
eoo.o
400.0
O 200.0
0.0
Simulated
—•- Measured
o.o
50.0
6.0
4.0
8
3.0
X
X
o
O 1.0
0.0
100.0 160.0
Tank Length (m)
200.0
250
— Simulated
-•- Measured
o.o
60.0
200.0
Figure 12. Madison Plant:
tank length.
100.0 160.0
Tank Length (m)
variation of average OIR and DO values along
260
48
-------
TABLE 17. MODEL CALIBRATION FOR THE MONROE PLANT
SIMULATED VALUES
SKI « o.<» aays
Temperature - 23.1 C
OUR (mg 02/L/day) Zone 1
Zone 2
00 (»g/L): Zone 1
Zone 2
Effluent Ammonia (mg-N/L)
Effluent N03/N02 (mg-N/L)
MLVSS (mg COO/L):Zone 1
Zone 2
K(L)a (I/day): Zone 1
Zone 2
Measured
Values
478
418
6.1
6.5
0.1
11.5
1822
1491
155
115
Default
412.8
292.9
6
6.4
0.4
14.9
2229.9
2112.4
155
115
Default w/
Temp . adj .
523.2
352.7
6.1
6.7
0.4
17.1
1686.7
1587.7
155
115
Calibrated
497.9
398.8
6.3
6.3
2
19.1
1613.9
1516.8
155
115
VALUES OF MODEL PARAMETERS USED FOR SIMULATIONS
H
E
T
E
R
0
T
R
0
P
H
I
C
A
U
T
0
M NX
Ks COD
Ks 02
yield
b decay
1/d
mg COO/L
mg-02/L
mg cells/mg COO
1/d
anoxlc growth factor
Ks N03
hydrolysis rate
hydrol satur ratio
mg-N/L
1/d
g COO/g COO
anoxlc hydrol factor
an*on1f1 cation
frac. part. prod.
N In blomass
N In part. prod.
02 saturation cone
|i MX
Ks NH-N
Ks 02
yield
b decay
L/mg COO/d
mg COO/M COO
mg-N/ng COO
mg-N/ng COO
mg-02L
1/d
mg-N/L
mg-02/L
mg cells/mg COO
1/d
4.000
10.0
0.55
0.67
0.620
0.8
0.2
2.200
0.150
0.4
0.160
0.080
0.086
0.060
' 9.00
- 0.650
1.000
0.550
0.240
0.120
4.253
10.633
0.659
2.339
0.141
0.170
9.82
0.691
4.253
10.633
0.10
0.67
0.800
2.339
0.141
0.170
9.82
0.340
0.100
shows the process flow diagram as modeled. Notable features of this plant are
the relatively high (38% of BOD) Industrial waste component, high average
Influent concentration (300 mg/1 5 day BOD), and short SRT (2.8 days). The
long, single pass tank studied was modeled as a series of 5 completely mixed
reactors In series.
Tables 19 and 20 show the model calibrations for feed COD fractions based
on measured soluble and partlculate components and batch reactor data
respectively. (For Table 19, Ss - 337, and XQ - 139; for Table 20, S$ - 151,
and Xs • 325.) Oxygen uptake rate (OUR) and DO data from reactors two and
four were used for model calibration. It was apparent that the extremely high
bottle OUR values measured for reactor one were not representative of the 1n-
sltu values at the prevailing low DO conditions 1n that reactor. The values
of OUR, DO, and MLVSS simulated by the default parameters for reactors 2 and 4
agreed reasonably well with the measured values, and this agreement was
50
-------
600.0
500.0
CM
9 400.0
« aoo.o
oc
o
JC
a
a 200.0
100.0
0.0
Simulated, Reactor 1
Simulated, Reactor 2
Measured, Reactor 1
Measured, Reactor 2
o.o
4.0
a.o 12.0
TIME (hours)
20.0
24.0
Figure 15. Monroe plant dynamfc simulation:
simulated OUR values.
comparison of measured and
Improved by adjusting the heterotrophlc decay coefficient and maximal
autotrophlc specific growth rate.
However, the measured and simulated ammonia concentrations did not agree.
This plant had relatively high concentrations of ammonia and organic nitrogen
In the feed (16.7 and 30.6 mg/1 respectively) but the effluent ammonia
concentration was measured to be only 5 mg/1. At the operating SRT of 2.8
days, the model simulated effluent ammonia concentrations ranged from 12 to 21
mg/1. The simulation with feed fractions based on the batch reactor data
(Table 20) simulated more ammonia loss through n1tr1f1cat1on-den1tr1f1cat1on
and consequently produced better agreement between the measured (5 mg/1) and
simulated (12 mg/1) ammonia values.
Figure 18 shows the diurnal profile for observed and measured values of
OUR and DO. The observed and simulated values of 00 agree reasonably well,
and this means that the simulated OUR values should be realistic. (Recall
that, since the simulation 1s based on K|_a measured by off-gas, perfect
agreement between the measured and simulated 00 values means that the OUR
51
-------
***.,,-
Table 18. JONES ISLAND EAST PLANT PROCESS SUMMARY
Average Dally Flow: 288,000 m'/d (76 MGD)
Average Raw Influent BOD,: 300 mg/L
Major Industrial Contributors: breweries, food processors, tanneries
Fraction of Flow: 0.11
Fraction of BODS: 0.38
Primary Treatment: fine screening
Nominal SRT at Time of Study: 2.8 days
Recycle Ratio at Time of Study: 0.33
Fraction of Flow Treated by Aeration Basin Studied: 0.04
Process Configuration: The East Plant Includes 20 aeration tanks each divided
Into 2 parallel one-pass basins, with each pass 111 m. (360 ft.) long and 7.5
m. (24.5 ft.) wide. Effluent from a group of basins flows to a cluster of
common claMfiers. There were 14 tanks (28 basins) 1n operation during the
24-hour study, and the north basin of Tank # 6 was monitored.
Mixed
Liquor '
Si
R2
1 R3 j
1 I
R4
• i
RS
Mixed
Liquor
(to clarifler)
i Prtwry CffluMrt
i HAS nfirtd »t
hf«l of tte
•trwwport
Rli Reactor 1 (240 n3>
R2> Reactor 2 (601 n3>
R3i Reactor 3 (601 n3>
R4> Reactor 4 (1803 n3>
R5< Reactor 3 (962 n3>
Indicate location of
neasurenents for
OUR. Da MLSS/MLVSS,
L Tenp.
Depth of nixed Uquor
m Aeration Basin * 4J n
SCALE) 1 en - 12 n
Figure 16. Jones Island Plant plan sketch of north basin of tank six.
Primary
Effluent
RAS
Secondary
» Effluent
Avg. Prmary Effluent Flowratc - 12380 n3/day
AVQ. RAS Rowrate - 3460 n3/okxy
NO SCALE
Figure 17. Jones Island Plant process flow diagram as modeled.
52
-------
TABLE 19. MODEL CALIBRATION FOR THE JONES ISLAND EAST PLANT
WITH FEED FRACTIONS BASED ON COD FRACTIONS
SRT - 2.8 days
Teraperature - 20.8 C
SIMULATED VALUES
Measured Default w/
Values Default Temp. adj.
Calibrated
OUR (ng 02/L/day)
DO (iig-02/L)
Effluent Ammonia
Effluent N03/N02
MLVSS (mg COD/L):
MLVSS (Mg COD/L):
K(L)a (I/day):
: Reactor 2
Reactor 4
Reactor 2
Reactor 4
(mg-N/L)
(Mg-N/L)
Reactor 2
Reactor 4
Reactor 1
Reactor 2
Reactor 3
Reactor 4
Reactor 5
1146
712
1.6
4.9
5
1
2470.8
2669.6
61
105
98.5
105
105
916.1
843.3
0.2
0.9
23.3
0
2423.4
2484.6
61
105
98.5
105
105
1032.2
839.6
0.3
1.9
22.7
0.1
2364.5
2410.3
61
105
98.5
105
105
1061.2
717.1
0.0
3.3
21.1
0.5
2584.7
2627.3
61
105
98.5
105
105
VALUES OF MODEL PARAMETERS FOR SIMULATIONS
H
E
T
E
R
0
T
R
0
P
H
I
C
A
U
T
0
U MX
KS COD
Ks 02
yield
b decay
1/d
ng COD/L
mg 02/L
mg cells/mg COD
1/d
anoxlc growth factor
Ks N03
hydrolysis rate
hydro 1 satur ratio
MQ-N/L
I7d
g COO/g COO
anoxlc hydrol factor
aamonlflcatlon
frac. part. prod.
N In blomass
N 1n part. prod.
02 saturation cone
u Max
Ks NH-N
KS 02
yield
b decay
L/Mg COD/d
Mg COD/MQ COO
mg-N/Mg COO
Hg-N/mg COD
Mg 02/L
1/d
Mg-N/L
Mg 02/L
MQ cells/mg COD
1/d
4.00
10.0
0.55
0.67
0.620
0.8
0.2
2.20
0.150
0.4
0.160
0.080
0.086
0.060
9.0
0.650
1.000
0.550
0.240
0.120
4.06
10.2
0.630
2.235
0.148
0.163
*
10.15
0.660
*
4.06
10.2
0.10
0.350
2.235
0.148
0.163
10.15
0.62
0.10
simulation 1s realistically based on the measured K|_a and D0 driving force).
The measured and simulated OUR profile for reactors two and four also agree
reasonably well. However, for reactor two, the measured values tend to be
higher and show nore variation than the simulated values. This Is probably a
result of oxygen limitation carried over from reactor one. The measured and
simulated OUR values for reactor one do not show good agreement, and this 1s
an artifact caused by the effect of oxygen limitation on the BOO bottle uptake
measurement.
South Shore Plant
Table 21 shows the process summary for the Milwaukee South Shore Plant.
53
-------
TABLE 20. MODEL CALIBRATION FOR THE JONES ISLAND EAST PLANT
WITH FEED FRACTIONS BASED ON BATCH REACTOR DATA
SRT - 2.8 days
Temperature - 20
OUR (mg 02/L/day):
00 (mg-02/L)
Effluent Ammonia (
Effluent N03/N02 (
MLVSS (Mg COO/L):
MLVSS (Mg COD/L):
K(L)a (I/day):
SIMULATED VALUES
.8 C
Reactor 2
Reactor 4
Reactor 2
Reactor 4
mg-N/L)
mg-n/L)
Reactor 2
Reactor 4
Reactor 1
Reactor 2
Reactor 3
Reactor 4
Reactor 5
Measured
Values
1146
712
1.6
4.9
5
1
2470.8
2669.6
61
105
98.5
105
105
Default
904.4
723.6
0.3
2
19.2
0.1
2685.9
2633.9
61
105
98.5
105
105
Default w/
Temp. adj.
1015.3
778.2
0.4
2.6
16.2
0.7
2598.4
2534.5
61
105
98.5
105
105
Calibrated
1056.9
709.7
0.1
3.1
12.5
4.7
2887
2820.9
61
105
98.5
105
105
VALUES OF MODEL PARAMETERS USED FOR SIMULATIONS
H
E
T
E
R
0
T
R
0
P
H
I
C
A
U
T
0
\i max
Ks COD
Ks 02
yield
b decay
1/d
mg COD/L
mg 02 /L
mg cells/mg COD
1/d
anoxlc growth factor
Ks N03
hydrolysis rate
hydrol satur ratio
mg-N/L
I7d
g COO/g COD
anoxlc hydrol factor
amonlfl cation
frac. part. prod.
N 1n blomass
N 1n part. prod.
02 saturation cone
u Max
Ks NH-N
Ks 02
yield
b decay
L/mg COO/d
mg COD/Mg COD
mg-N/Mg COO
mg-N/mg COD
mg 02/L
1/d
mg-N/L
mg 02/L
Mg cells/mg COD
1/d
4.00
10.0
0.55
0.67
0.620
0.8
0.2
2.20
0.150
0.4
0.160
0.080
0.086 ,
0.060
9.0
0.650
1.000
0.550 '
0.240
0.120
4.06
10.2
0.630
2.235
0.148
0.163
10.15
0.660
t
4.06
10.2
0.10
0.230
2.235
0.148
0.163
10.15
0.59
0.10
Figure 19 shows the configuration of the basin studied and Figure 20 shows the
process flow diagram. Notable features of this study were Us limited
duration (11 hours) and difference 1n DO conditions between the basin studied
(Basin 17) and the other parallel basins feeding the same group of clarlflers.
These limitations make these data more difficult to Interpret.
The model calibration 1s shown 1n Table 22. It was not possible to
calibrate the model to reasonably simulate both the measured OUR and measured
MLVSS concentrations. Simulations based on the default parameters gave OUR
values much lower than the measured values, and MLVSS values somewhat lower
than the measured values. Improved agreement between the simulated and
measured OUR values was obtained by decreasing the heterotrophlc yield and
54
-------
4000.0
_. S600.0
SI
Simulated, Reactor 1
Simulated, Reactor 2
.Simulated, Reactor 4
Measured, Reactor 1
Measured, Reactor 2
Measured, Reactor 4
o.o
o.o
a.o
12.0
TIME (hours)
16.0
20.0
24.0
7.0
Simulated, Reactor 1
Simulated, Reactor 2
Simulated. Reactor 4
Measured. Reactor 1
Measured. Reactor 2
Measured, Reactor 4
e.t
1.0 it.o ie.0
TIME (hours)
te.o
24.0
Figure 18. Jones Island Plant dynamic simulation: conparison of measured
and simulated OUR and DO values.
55
-------
Table 21. SOUTH SHORE PLANT PROCESS SUMMARY
Average Dally Flow: 371,000 m'/d (98 MGD)
Average Raw Influent BOD,: 162 mg/L
Major Industrial Contributors: glue processors, food processors, and machine
Industries
Fraction of Flow: 0.06
Fraction of BOD,: 0.18
Primary Treatment: sedimentation
Nominal SRT at Time of Study: 4.3 days
Recycle Ratio at Time of Study: 0.19
Fraction of Flow Treated by Aeration Basin Studied: 0.05
Process Configuration: This plant consists of 28 single-pass basins, each 60
m. (196 ft.) long and 9.1 m. (30 ft.) wide. Mixed liquor from a group of
basins 1s combined before clarification. There were 20 basins 1n operation
during the study and many of these became anoxlc during the night. The
aeration rate and 00 concentrations of Basin 17 were considerably higher than
those of the other basins, and, because of this, H was selected for study
midway during the study period. Thus, only 11 hours of data were taken.
45X of
Prinary
Effluent
Rl
i R2 i
i • i
R3
RAS
A B A
28X of
Prinary
Effluent
ZT/. of
Prinary
Effluent
SCALE' 1 en - 15 n
Mixed
Liquor
R2i Reactor S (1112 n3>
« R3i Reactor 3 <2506 n3>
Depth of nixed liquor ft
Aeration Bash * 4.6 n
Figure 19. South Shore Plant plan sketch of basin 17.
45% of
Prinary •
Effluent
RAS
2S7. of
Primary
Effluent
875C of
Prinary
Effluent
—>
i
/
o£
Rl
Air
"V
* /
J*
RS
-»
Air
4-/
2
R3
i i Secondary
i -i ••'> Effluent
\y
'M ^
[L»VAS
Air
Avp, Prinary Effluent Flowrate « 17498 n3/day
Avg. RAS Rovrate - 4467 n3/day
NO SCALE
Figure 20. South Shore Plant process flow diagram as modeled.
56
-------
TABLE 22. MODEL CALIBRATION FOR THE SOUTH SHORE PLANT, BASIN 17
SIMULATED VALUES
SKI » i.j oays
Temperature • 18.5 C
OUR (mg 02/L/day)
00 (mg-02/L):
Effluent Ammonia
Effluent N03/N02
MLVSS (mg COO/L):
Kla (I/day):
: Reactor 1
Reactor 2
Reactor 3
Reactor 1
Reactor 2
Reactor 3
(mg-N/L)
(mg-N/L)
Reactor 1
Reactor 3
Reactor 1
Reactor 2
Reactor 3
Measured
Values
993
818
644
1.1
2.1
3.5
3.6
5.8
1852
1505
121.0
82.0
70.0
Default
753.9
524.6
361.7
2.5
2.5
3.6
1.6
8.3
1597.9
909.9
121.0
82.0
70.0
Default w/
Temp. adj.
781.8
544.7
364.1
3.8
3.7
5.2
1.5
9.6
1607.6
915.6
116.8
79.1
67.6
Calibrated
990.8
715.8
471.4
2.1
1.8
3.6
2.7
12.3
952.7
582.1
116.8
79.1
67.6
VALUES OF MOTEL PARAMETERS USED FOR SIMULATIONS
v
H
E
T
E
R
0
T
R
0
P
H
I
C
A
U
T
0
u max
Ks COD
Ks 02
yield
b decay
1/d
mg COD/L
mg 02/L
mg cells/mg
1/d
anoxlc growth factor
Ks N03
hydrolysis rate
hydrol satur ratio
mq-N/L
I7d
g COO/g COO
anoxlc hydrol factor
amonlflcatlon
frac. part. prod.
N In blomass
N In part. prod.
02 saturation cone
u max
Ks NH-N
Ks 02
yield
b decay
L/mg COO/d
mg COD/mg CO
mg-N/mg COO
mg-N/mg COO
mg 02/L
1/d
mg-N/L
mg 02/L
mg cells/mg
1/d
4.00
10.0
0.55
0.67
0.620
0.8
0.2
2.20
0.150
0.4
0.160
0.080
0.086
0.060
9.00
0.650
1.000
0.550
0.240
0.120
3.88
9.7
0.602
2.136
0.146
0.155
10.84
0.631
3.88
9.7
0.10
0.46
0.602
2.136
0.146
0.155
10.84
0.500
0.100
decay coefficients, but this further accentuated the difference between the
measured and simulated MLVSS concentrations.
The South Shore Plant uses a relatively complex process control strategy
1n which an on-line settleometer 1s used to maintain the MLVSS concentration
at a set value by controlling the return sludge flow rate (GMnker, 1986).
Over the short period of study, this control system may have caused a change
1n MLVSS Inventory 1n the claMflers. This might explain the lack of
agreement between the simulated and measured MLVSS concentrations 1n aeration
basin 17.
57
-------
TOTAL PROCESS AVERAGE OXYGEN UTILIZATION RATES
An estimate of the total daily process average oxygen utilization rate
can be made by multiplying the estimated OUR values by the corresponding
reactor volumes. In the absence of direct measurements of OUR values for all
tank regions, the best estimates of OUR for the various reactor regions are
given by the SSSP simulation based on the calibrated parameters. These
estimates are consistent with the measured OUR, DO, off-gas and sol Ids data
used to calibrate the model. An example calculation for the portion of the
Madison plant studied 1s shown in Table 23.
TABLE 23. CALCULATION OF TOTAL PROCESS AVERAGE OXYGEN UTILIZATION RATE
FOR MADISON PLANT BASED ON CALIBRATED IAWPRC MODEL
Reactor Volume m OUR g/m /day Utilization Rate kg/day
1
2
3
4
1,733
4,008
3,827
1,914
468.7
541.1
404.8
341.3
Total
812.3
2,168.7
1,549.2
653.2
5,183.4 kg/day
Values of the best estimates of the total process average oxygen utilization
rates calculated 1n this manner for all six plants are shown 1n column 2 of
Table 24. These quantities apply to the portion of the plant Included in the
24 hour studies. Consequently, the magnitudes of the utilization rates
reflect both the size of the plant and the fraction of flow treated by the
portion of the plant Included 1n the study. Thus, the utilization rates
reported for the Green Bay plant where the portion under study treated 33% of
the plant flow are much greater than the rates reported for the larger Jones
Island plant where the portion under study treated only 4% of the plant flow.
TABLE 24. COMPARISON OF TOTAL PROCESS AVERAGE OXYGEN UTILIZATION RATES
ESTIMATED BY THE IAWPRC AND CONVENTIONAL MODELS
Plant
Madison
Monroe
Green Bay
Portage Lake
Jones Island
South Shore
Oxygen Utilization Rate, kg/day
IAWPRC
Model
Total
5,183
822
29,350
593
2,544
3,079
Conventional
Model
Total
5,001
845
30,600
633
2,925
3,532
3,797
581
30,448
592
2,803
2,352
1,204
264
152
41
122
1,180
Percent
Difference
+3.5
-2.9
-4.3
-6.8
-15.
-15.
58
-------
A similar more conventional estimate may be made by viewing the process
aeration tank as a single completely mixed reactor and applying the
conventional model given by Equation 12 to calculate the carbonaceous oxygen
utilization rate, Rc. The values of the heterotrophic yield coefficient,
YH, used in the IAWPRC (SSSP) and conventional models are identical.
However, the appropriate value of the oxidative decay coefficient, bc, for
use 1n the conventional model must be calculated based on the IAWPRC
hydrolytlc decay coefficient, b, using Equation 15. The total oxygen
utilization rate 1s then calculated according to Equation 14 after
determining the nitrogenous oxygen demand from Equation 13. An example
calculation for the Madison process 1s shown 1n Table 25.
TABLE 25. CALCULATION OF TOTAL PROCESS AVERAGE OXYGEN UTILIZATION RATE
FOR THE MADISON PLANT BASED ON CONVENTIONAL MODEL
From Equation 15
From Equation 12
be
• b ( 1 - Yu
0.626( 1 - 0.60
0( i - fp n
( 1 -0.2))
- 0.28 d"1
b8 - Y - f
Rc -
i + bcec
1 + 0.28(16.4) - 0.60 - 0.2(0.6)0.28(16.4)
Rc » 20,346 (116.6 + 119.8 - 1.4)( ------------------------------------------ )
1 + 0.28(16.4)
Rc - 3797 kg oxygen/day
From the definition of available nitrogen, NT A,
NT* - Influent Nitrogen - Nitrogen In Waste Sludge
NfjJ » 417.1 - 91.0 • 326.1 kg/day
(quantities are calculated based on Measured Influent data, SRT and
IAWPRC simulations calibrated to Measured effluent and process data)
From the definition of nitrogen nitrified to nitrate, NTN,
NTN - Available Nitrogen - TKN 1n Effluent
N{N • 326-1 - 20-4 • 305»7 kg/day
From the definition of nitrogen denitrified from nitrate to nitrogen gas, NHM,
NHN - NTH - Nitrate Remaining 1n Effluent
NON - 305.7 - 238.0 - 67.7 kg/day
From Equation 13 «•
Rn - 4.57 NTN - 2.86 NnN
Rn - 4.57 (385.7) - 2.86 (67.7) - 1,203.6 kg/day
From Equation 14
R* - R. + Rn
Rt - 3,797.6 V 1,203,6 - 5,001.2 kg/day
Column 3 of Table 24 summarizes the total process average oxygen
utilization rates calculated for all six plants using the conventional
model. For most of the plants, the agreement between the IAWPRC and
conventional models was excellent as the estimates generally agreed within
10%. The exceptions to this were the two Milwaukee plants, Jones Island and
South Shore where the conventional approach estimated an oxygen utilization
rate 15% greater than the IAWPRC approach.
59
-------
It can be concluded that, for given values of the yield and decay
coefficients, the two approaches produce estimates of the average process
oxygen utilization rate which agree reasonably well. In this study, the
IAWPRC model was calibrated to plant data and, therefore, was judged to give
the better estimates of the process oxygen Utilization rates. There 1s some
indication, based on the Milwaukee plants, that the conventional approach may
slightly overestimate the oxygen utilization rate.
These results reinforce the observations made earlier 1n Section 4, namely
that the conventional approach 1s useful but has several disadvantages when
compared with the IAWPRC model and SSSP software package. Both approaches
need applicable values of the heterotrophlc yield and decay coefficients.
However, the conventional model does not distinguish between slowly and
readily degradable substrates and cannot easily be adapted to simulate spatial
and temporal variations of OUR. Moreover, the conventional approach can
neither simulate nor predict nitrogenous oxygen demand. It can only calculate
nitrogenous oxygen demand based on observed Influent and effluent nitrogen
concentrations. On the other hand, 1t was observed 1n this study that the
IAWPRC model needs only one additional calibrated parameter, the maximal
autotrophic specific growth rate, to simulate the entire process. In almost
all cases, default values of all but two or three parameters (hetertrophlc
yield, hetertrophlc decay, and autotrophic maximal specific growth rate) were
adequate 1n model calibration.
SPATIAL AND TEMPORAL VARIATION IN OXYGEN UPTAKE RATES
Figure 12 shows that the IAWPRC model was able to correctly simulate the
spatial variation in the temporal average DO and OUR for the Madison plant.
Similar results were obtained for other plants 1n which several reactors were
used in the simulation model. Notice that the measured OUR values 1n Figure
12 refer to the 1n-v1tro values measured11 by the BOD bottle uptake method.
These represent potential maximal values which could occur at high DO
concentrations. In regions of low DO concentrations, the actual 1n-s1tu OUR
values which are simulated by the model based on the off-gas K|_a values are
considerably lower than the potential maximal values measured by the bottle
uptake method. This 1s because the bottle uptake method does not reflect
oxygen limitations which begin to occur at DO values below approximately 1.5
mg/1.
In quantltatlng the spatial variation of OUR, 1t 1s useful to understand
the relationship between OUR and DO and how the phenomenon of oxygen
limitation smoothes the variations in OUR and distributes the oxygen
utilization More uniformly over the tank volume. It was observed 1n studying
the long, narrow aeration tanks at Madison and Jones Island that a zone of low
DO and high potential maximal OUR was present in the mixed liquor distribution
channel and aeration basin entrance zone. During times of high load, this
zone moved Into the aeration basin and progressively occupied more aeration
volume until H receded under conditions of lower load. Figure 12 represents
a situation in which the zone of high potential maximal OUR occupies the first
third of the tank. If additional aeration capacity were provided In this
60
-------
region, the DO would Increase and the OUR could approach 1,200 mg/l/day, the
potential maximal value, at the tank entrance. However, the faster rate of
readily biodegradable COD removal would cause the OUR to decrease more rapidly
with tank distance and to fall below the current, correctly simulated values.
This is because the total amount of oxygen utilized would remain approximately
constant. In this case, oxygen limitation smoothes the spatial variation 1n
OUR.
From the viewpoint of aeration energy economy, it 1s beneficial to operate
under conditions of low DO. However, because the DO concentration can
influence many facets of process performance, considerations other than energy
economy may also be important. The DO concentration and presence or absence
of anoxlc conditions are known to Influence the selection of microorganisms
and this, in turn, can Influence blomass settleab111ty and process efficiency.
Figures 6, 11, 15, and 18 Illustrate the temporal variation 1n the
measured and simulated spatial average OUR and DO for several plants. The
IAWPRC model was able to simulate these temporal variations reasonably well.
In regions of low DO, the model again correctly simulated the actual OUR
significantly lower than the value measured by the BOD bottle technique.
Table 26 shows the estimated average OUR values for the various plants
along with the ratios of the estimated maximal and minimal values to the
average. Columns 3 to 5 describe spatial variations of the temporal dally
TABLE 26. SPATIAL AND TEMPORAL VARIATION OF OXYGEN UPTAKE RATES
Plant Estimated
Average OUI
mg/l/day
Madison
Monroe
*
Green Bay 2
*
Portage Lake
Jones Island
South Shore
contact
451
454
,200
665
705
651
tank
Spatial Variation
? RatTos to Average OUR
*
Potential
Maximum
2.8
1.1
1.0
1.0
3.0
1.5
only
Actual
Maximum
1.2
1.1
1.0
1.0
1.4
1.4
Minimum
0.8
0.9
1.0
1.0
0.6
0.6
Temporal Variation
Ratios to Average OUR
Maximum
1.6
1.1
1.3
1.3
1.6
1.1
Minimum
0.7
0.8
0.7
0.8
0.8
0.9
61
-------
average determined from the IAWPRC model calibrated to the measured data.
Columns 6 and 7 describe temporal variation of the spatial reactor average
determined directly from the measured data for reactors which were not oxygen
limited. It should be emphasized that these ratios are based only on the 24
hour studies conducted on the plants and represent the maximal and minimal
occurrences observed during a single 24 hour period. For a longer observation
period, such as a month or a year, more variation 1n uptake rates would be
expected to occur.
62
-------
REFERENCES
American Society of Civil Engineers (ASCE), "A Standard for the Measurement of
Oxygen Transfer in Clean Water," ASCE, New York, 1984.
Baillod, C.R., Cressey, G.M., and Beaupre, R.T., "Influence of Phosphorus
Removal on Solids Budget," Journal WPCF, 49: 131-145, 1977.
Bidstrup, S.M., and Grady, C.P.L., Jr, "A User's Manual for SSSP," Clemson
University, Clemson, South Carolina, 1987.
Bidstrup, S.M., and Grady, C.P.L., Jr., "SSSP - Simulation of Single Sludge
Processes," Journal WPCF, 60:351-361, 1988.
Brown, L.C. and Baillod, C.R., "Modeling and Interpreting Oxygen Transfer
Data," ASCE, Jour. Environ. Engr. Div., I08(EE4):607-628, 1982.
Ekama, G.A., Gold, P.L., and Marais, G.v.R., "Procedures for Determining
Influent COD Fractions and the Maximum Specific Growth Rate of Heterotrophs in
Activated Sludge Systems," Wat. Sc1. Tech., 18:91-114, 1986.
Fair, G.M., and Geyer, J.C., Water Supply and Wastewater Disposal. Wiley and
Sons, New York, 1954.
Grady, C.P.L., Jr, Gujer, W., Heffze, M., Marais, G.v.R., and Matsuo, T., "A
Model for Single Sludge Wastewater Treatment Systems," Wat. Sd. Tech., 18:47-
61, 1986.
Grady, C.P.L, Jr., and Urn, H.C., Biological Wastewater Treatment. Theory and
Applications, Marcel Dekker, Inc., New York, 1980.
Great Lakes - Upper Mississippi River Board of State Sanitary Engineers,
•Recommended Standards for Sewage Works," 1978.
Henze, M., Grady, C.P.L., Jr., Gujer, W., Marais, G.v.R., and Matsuo, T.,
•Model for Single Sludge Wastewater Treatment," Water Res. 21:505, 1987.
Herbert, D. "A Theoretical Analysis of Continuous Culture Systems," S.C.I.
Monograph No. 12, p.21, Society of Chemical Industry, London, 1961.
IAWPRC Task Group on Mathematical Modeling for Design and Operation of
Biological Wastewater Treatment, M. Henze, Chairman, "Final Report - IAWPRC
Activated Sludge Model No. 1," IAWPRC Scientific and Technical Reports, 1986.
63
-------
Metcalf & Eddy, Inc., "Wastewater Engineering: Treatment, Disposal, Reuse,"
Second Edition, McGraw-Hill, 1979.
Mueller, J.A., and Boyle, W.C., "Oxygen Transfer Under Process Conditions,"
Journal WPCF, 60:332-341, 1988.
Mueller, J.A., and Stensel, H.D., "Biologically Enhanced Oxygen Transfer 1n
the Activated Sludge Process - Fact or Folly?," paper presented at the WPCF
Annual Conference, Philadelphia, 1987.
Redmon, D., Boyle, W.C., and Ewing, L., "Oxygen Transfer Efficiency
Measurements in Mixed Liquor Using Off-Gas Techniques," Journal WPCF, 55:1338-
1347, 1983.
Rozzl, A., Merl1n1, S., and Passino, R., "Development of a Four Population
Model of the Anaerobic Degradation of Carbohydrates," Environmental Technology
Letters, 6:610-619, 1985.
Standard Methods for the Examination of Water and Wastewater. 16th Edition,
APHA, AWWA, WPCF, 1985.
U.S. Environmental Protection Agency, "Process Design Manual for Nitrogen
Control," 1975.
U.S. Environmental Protection Agency, "Summary Report: Fine Pore (Fine Bubble)
Aeration Systems," EPA/625/8-85/010, 1985.
Wastewater Treatment Plant Design. Water Pollution Control Federation Manual
of Practice No. 8, WPCF, Washington, D.C., 1977.
64
-------
APPENDIX A
RAW DATA FROM THE 24-HOUR PLANT STUDIES
TABLE Al. RAW DATA FOR THE PORTAGE LAKE PLANT
1
N
F
L
U
£
W
T
£
f
F
L
U
£
N
T
C
0 T
M A
T N
A K
C
T
1
N
F
L
U
E
M
T
E
F
F
L
U
E
•
T
C
OT
• A
T N
AK
C
T
TIME
TOTAL COO (on/I)
SOLUBLE COO Tug/L)
AMMONIA (iQ-NA)
ORGANIC-N (ig-NA)
ALKALINITY (iOlA)
FIOHRATE (M3/day)
TOTAL COO (ugA)
SOLUBLE COO (igA)
AMMONIA (ig-NA)
ORGANIC-N (ig-NA)
N03/N02 (ng-HA)
OUR (ig 02A/day)
00 (*gA)
HVSS (tg/l)
TIME
TOTAL COD (iQA)
SOLUBLE COO (MA)
AMMONIA (•Q-NA)
ORGANIC-N (on-NA)
ALKALINITY (ODlA)
FLONRATE (kC/day)
TOTAL COD UBA)
SOLUBLE CODOn/l)
AMMONIA (onHIA)
ORGANIC-N (HHi/L)
N03/N02 (•9-JI/L)
OUR (ig 02A/day)
DO(^)A)
HVSS (igA)
(PM)
12
325
140
7.8
4.1
3638
54
26
4
528
1.8
(AM)
2
171
70
2.8
2426
35
26
0.3
600
0.9
6-15-87
1 2
255
131
7.8
10.1
3638 4374
48
26
2.7
1.7
2.3
528 595
0.5
1440
6-16-87
3 4
V
100
50
4.7
3.1
2426 2426
32
28
2.2
1.6
6
557 577
1.3
1950
345
448
253
5 5.6
5
4115 3798 3925
54
39
3.8 3.2
1
3.3 4.2
706 626 667
0.5
1450
567
200
93
2.8 6
7.7
2876 2426 3658
34
30
1 0.8
2.3
9.6 7.5
600 551 523
2.4
1660 1810
6
342
183
7.7
3611
64
43
1.2
695
0.5
8
225
110
7.8
4115
32
28
2.5
513
3.5
7 8
349
207
6
4.8
5069 2923
50
34
3.8
0.5
3.3
758 706
0.5
1430
9 10
250
120
7.2
5.8
5778 *3638
38
33
1.3
2.6
7
470 527
4
1600
9 10 11
250
120
6 5.2
5.6
3638 3638 3638
44
34
3.4 3.6
0.7
2.1 1.8
675 600 643
0.7
1540 1770
(AM)
11 Average F
266.8
134.3
7 5:7
5.8
4.1
3638 3547.5
Average F
43T
31.4
2.5 2.6
1.6
4.5 4.5
Average
514 597.6
1.5
1663.0
(AM)
12 1
287
135
4.9
1.1
4.1
2679 3050
38
30
2.5
0.8
2.9
602 582
0.8
1980
N Average
274.9
138.9
5.8
6.1
N Average
44.5
31.6
2.6
1.7
4.5
65
-------
TABLE A2. RAW DATA FOR THE GREEN BAY PLANT
C
0
N
T
A
C
T
C
0
N
T
A
C
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
T
A
N
K
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
T
A
N
K
TIME
TOTAL COO (iig/L)
SOLUBLE COD (fflgA)
AMMONIA (ig/L)
ORGANIC-N (ig/L)
ALKALINITY (l»IA)
FLCfRATE (M3/day)
TOTAL COD (ugA)
SOLUBLE COD (*gA)
AMMONIA (n/L)
ORGAN IC-H (ng-NA)
N03/N02 (ng-NA)
OUR (ng-02/L/day)
DO (mg-02/L)
HVSS (ugA)
RAS(M3/day)
TIME
TOTAL COO (KI/L)
SOLUBLE COD (MA)
AMMONIA (n/L) ,
ORGANIC-N (M/L)
ALKALINITY (»IA)
FLONRATE (M3/day)
TOTAL COD (H/L)
SOLUBLE COD fo/L)
AMDNIA (MA)
ORGANIC-N (M-NA)
N03/N02 OV-HA)
OUR (ig-02A/day)
DO (M-02A)
M.VSS (MA)
RAS(M3-day)
(PtO
12
6-22-87
1 2 3
932
626
32.5
19.1
5
69303 66919
280
176
3.2
31.5
1510 1210
2.9
1880
39004 39046
(AM)
2
1112
724
43
14.6
51098
293
211
30
3.1
1846
1
2000
38898
905
610
32.5
25.8
65443 60409
289
201
30
4.9
1363 1426
3.3
1960
39061 39035
6-23-87
345
52687
28.4
1146
895
42
21.8
47918 47313
250
200
25.5 23.9
3
1793 2026 1762
1.3
2060
38732 38721 38766
4
5 6
789
10 11
900 1105 1269 898
700 773 854 610
36 40.5 36 40
26 27.7 24.1 19.6
65632 62679 62642 59727 61847 60598 61598
298
220
3.1
289
235
32
2
250
221
40 29 36
3.9
246
211
2.8
1474 2031 1884 2124 2280 2066 2177 2117
1 0.7 0.8 0.9
1760 1880 1680 1960
38751 38859 38796 38857 37850 38588 38633 38679
6
1300
1106
18.5
45042
263
221
3.1
1877
1.2
2080
38872
7 8
1000
750
19
45458 53444
266
210
27
2.5
1726 1721
2.3
1960
3884238888
(AM)
9 10 11
997
529
29
65821 66729 67940
*
263
206
32.5
3.2
2040 2326 1685
1.9
1820
38974 38959 39020
Average
1038.7
733.1
37.2
21.7
58878.9
270.0
208.5
30.4
3.1
1842.2
1.5
1908.3
38807
(AM)
12
1
900
620
40
22.9
5
58592 55375
253
190
30
2.8
1829 1920
0.5
1860
38635 38898
F N Average
1027
718
36
22
270
206
30
3
.6
.3
.8
.0
.7
.1
.8
.2
-------
TABLE A3. RAW DATA FOR THE MADISON PLANT
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
R
E
A 0
C N
T E
0
R
R
E
A T
C N
T 0
0
R
R
E T
A H
C R
T E
0 E
R
R
E F
A 0
C U
T R
0
R
TIME
TOTAL COO (ugA)
SOLUBLE COO (in/L)
AMMONIA (mg-N/L)
ORGAN IC-N (mg-N/L)
FLOKRATE (M3/day)
TOTAL COD (MA)
SOLUBLE COD (ig/L)
AMMONIA (n-NA)
ORGAN IC-N (•g-N/L)
N03/N02 (ng-NA)
OUR (mg-02A/day)
DO (mg-02A)
MLVSS (igA)
RAS (M*3/day)
(POINT B)
OUR (mg-02/L/day)
DO («g-02A)
MLVSS (ig/L)
(POINT E)
OUR («g-02A/day)
DO (H-02A)
MLVSS (tQ/L)
(POINT 6)
OUR (ig-02A/day)
DO (ig-02A)
MLVSS (ig/L)
(POINT 6 & POINT 1)
(PM) 7-13-87
678
162
39
20.5
10.8
22448 22223 22789
8
20
0.2
1.0
1183 996
0.5
1670.0
13569 13524
968 913
1.8
325 336
3.9
9
278
191
11.7
21580
16
7
1.9
1084
0.5
0.9
3.8
10
41.5
21104
0.3
1123
0.7
1600.0
13558
903
0.9
, unf
2.2
NOT
351
3.6
NOT
(AM)
11 12
271
29
10.0
10.8
22901 19940
16
7
0.2
1.7
15
1267
0.4
1480.0
13549
996 1006
0.9
428
2.2
324 387
3.9
UFA9JRFD
7-14-87
1 2
363
65
10.0
10.8
20459 15432
16
7
0.0
1.3
14.6
1340 1250
0.4 0.2
1570.0
13535
1071
0.9 1.0
387
2.4
338
3.7 4.0
3
265
158
9.1
14851
24
10
1.3
982
374
335
67
-------
TABLE A3. RAW DATA FOR THE MADISON PLANT CONTINUED
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
R
E
A 0
C N
T E
0
R
R
E
A T
C N
T 0
0
R
R
E T
A H
C R
T E
0 E
R
R
E F
A 0
C U
T R
0
R
TIME
TOTAL COD (MA)
SOLUBLE COD (ig/L)
AMMONIA (w-NA)
ORGAN IC-N (ng-N/L)
FLOHRATE (M3/day)
TOTAL COO (igA)
SOLUBLE COO (ig/L)
AMMONIA (•g-NA)
ORGAN IC-N (ng-N/L)
N03/N02 (ng-NA)
OUR (mg-02/L/day)
DO (mg-02A)
MLVSS (ngA)
RAS (MA3/day)
(POINT B)
OUR (mg-02A/day)
DO (mg-02A)
MLVSS (igA)
(POINT E)
OUR («g-02A/day)
DO (ng-02A)
MLVSS (ugA)
(POINT E)
OUR do-CK/l/day)
00 (ig-02A)
MLVSS (MA)
(POINT G (POINT 1)
(AM)
4 5
245
123
14.0
13.2
14956 12779
8
20
0.2
1.3
1210 1084
0.3
1860
13542
968
0.8
406
3.2
348
6 7
172
11.0
9.9
14371 16654
16
16
0.3
1.3
12.4
1115
0.3
1500
13533
871 871
1.1
324 357
3.9
335 309
4.6
8
16.0
17217
0.4
13.4
1130
0.3
1480
13592
893
1.0
- NOT
4.4
MOT
4.6
NOT
9 10
210
172
12.0
20.7
22482 23076
20
23
0.3
1.5
11.6
1250
0.5
1450
13574
1052
1.4
yrictiprn
319 348
4.2
UPAQ Drn
329 336
4.0
if ignrn
(PM)
11 12
167
97
6.5
9.9
24936 24341
24
7
0.1
1.2
12.6
1233
0.6 0.3
1550
13561
1045
1.4 0.8
*450
3.7 3.2
374
4.2 3.2
1
340
246
19.2
23442
8
7
0.7
1858
1115
416
386
68
-------
TABLE A3. RAH DATA FOR THE MADISON PLANT CONTINUED
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
R
E
A 0
C N
T E
0
R
R
E
A T
C N
T 0
0
R
R
E T
A H
C R
T E
0 E
R
R
E F
A 0
C U
T R
0
R
TIME
TOTAL COO (M/l)
SOLUBLE COO (mg/L)
AMMONIA (IKQ-N/L)
ORGAN IC-N (ng-N/L)
FLOHRATE (M3/day)
TOTAL COD (•g/l)
SOLUBLE COD (m/L)
AMMONIA (ng-N/1)
ORGAN IC-N (•g-N/L)
N03/N02 (»g-N/l)
OUR (mg-02/L/day)
DO (18-02/1 )
MLVSS (ing/l)
RAS (MA3/day)
(POINT B)
OUR (mg-02/L/day)
DO (§9-02/1)
MLVSS (ig/L)
(POINT E)
OUR (ig-02/L/day)
DO (ig-02A)
MLVSS (»g/L)
(POINT G)
OUR (•B-02/L/day)
DO (ig-OZ/l)
MLVSS 0*|/l)
(POINT 6 & POINT 1)
(PM)
2
6.0
24182
0.0
9.6
1711
0.5
1500
13540
3 4
353
91
7.0
14.1
22992 21623
8
20
0.2
1.4
8.4
1583
0.8
1340
13548
1230 1226
1.2 1.3
< NOT
615
3.1 1.8
< MOT
3.7
448
3.4
NOT
5 6
246
194
12.4
21537
16
20
1.2
1417
0.7
996
1.8
MEASURED
739 654
1.5
MEASURED
462 441
3.2
yrtqDCn
Average
264
131
14.1-
12.7
20346
15
14
0.2
1.3
12.4
1284
0.45
1545
13552.0
1006
1.15
447
3.0
363
3.9
Flow Weighted
Average
263.1
130.0
14.0
12.9
15.0
13.4
0.2
1.3
12.0
69
-------
TABLE A4. RAW DATA FOR THE MONROE PLANT
(AM) 8-18-87 (PM) 8-19-87
R
E
A
C
T
0
R
R
E
A
C
T
0
R
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
0
N
E
T
N
0
TIME 8
TOTAL COO (ig/L) 249
SOLUBLE COO (ig/L) 88
AMMONIA (ig-N/L)
ORGAN IC-N (ig-N/L) 21.3
FLOHRATE (M3/day)
TOTAL COO (ig/L) 39
SOLUBLE COD (ng/L) 20
AMMONIA (ng-N/L)
ORGAN IC-N (ng-N/L) 3.1
NITRATE/NITRITE (ng-N/L)
OUR (ng-02/L/day)
DISSOLVED OXYGEN (ig/L)
* MLVSS (ugA)
RAS (M'3/day)
(POINT C)
OUR (ng-02/L/day)
DISSOLVED OXYGEN (ig/L)
MLVSS (ig/l)
(POINT F)
9 10 11 12 1 2
323 290 303
115 81 67
11.0 12.0 10.0
18.4 18.4 19.6
2905 3278 3105 3105 3278 3278
32 32 26
25 35 35
0.13 0.10 0.08
3.1 3.9 3.6
13.0 12.0 11.5
404 445 419 450 457
6.5 5.6 6.0
1400 1300
3860 3888 3894 3924 3897 3888
371 371 395 389 402
6.9 7.0 6.5
1200
*
3 4 5
337
67
11.3 11.0
22.4
3105 3105 3105
26
35
0.11
3.4
11.5 11.0
470 457 489
1400
3906 3906 3888
400 411 415
800
* MLVSS data frot Points A. C. and D Here used for this reactor.
70
-------
TABLE A4. RAH DATA FOR THE MONROE PLANT CONTINUED
(PM)
R
E
A
C
T
0
R
R
E
A
C
T
0
R
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
0
N
E
T
N
0
6
TOTAL COO (nig/l) 343
SOLUBLE COD (BgA) 23
AMMONIA (•g-NA)
ORGAN IC-N (§g-NA) 23.5
FLWRATE (MA3/day) 3105
TOTAL COD (iigA) 20
SOLUBLE COD (to/I) 35
AMMONIA (ng-NA)
ORGAN IC-N (mg-NA) 3.1
NITRATE/NITRITE (nig-NA)
OUR (•g-OZA/day) 536
DISSOLVED OXYGEN (iigA) 5.5
* MLVSS (qjA)
RAS (rs/day) 3906
(POINT C)
OUR (•g-02A/day)
DISSOLVED OXYGEN (igA) 6.0
MLVSS (q/L)
(POINT F)
789
364
63
10.5 9.2
23.5
3105 3105 3105
39
35
0.08 0.10
4.2
14.0 13.5
485 509 485
6.2
3888 3888 3888
456 480 449
6.6
(AM)
10 11 12 1 2
364 343 263
54 43 43
10.0 11.0
24.1 23.5 25.2
3105 3105 3105 3105 3105
32 39 35
45 25 20
0.10 0.10
3.9 3.9 3.1
10.5 11.0
452 477 489
6.6 6.3
900 1400
3888 3888 3888 3888 3888
402 414 441 432
6.5 6.6
1300
-
3
9.7
3105
0.12
11.3
521
3888
433
* MLVSS data froi Points A, C, and 0 were used for this reactor.
71
-------
TABLE A4. RAN DATA FOR THE MONROE PLANT CONTINUED
R
E
A
C
T
0
R
R
E
A
C
T
0
R
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
0
N
E
T
N
0
TOTAL COD (lig/L)
SOLUBLE COD (tg/L)
AMttNIA (ng-NA)
ORGAN IC-N (ig-NA)
FLOHRATE (M3/day)
TOTAL COD (ig/L)
SOLUBLE COD (ing/L)
AMtiNIA (ng-NA)
ORGANIC-N (ng-NA)
NITRATE/NITRITE (mg-N/L)
OUR (ng-02/L/day)
DISSOLVED OXYGEN (igA)
« ItVSS (BgA)
RAS (IT3/DAY)
(POINT C)
f
OUR (ig-02/L/day)
DISSOLVED OXYGEN (igA)
ILVSS (igA)
(POINT F)
(AM)
4567
263 188
60 53
9.5 11.7
23.5 20.2
3105 3105 3016
35 25
15 15
0.11 0.12
3.6 4.2
10.0 8.2
481 494 495 543
6.2
1300
3888 3888 3888 3888
409 419 449 419
6.5
900
t
Flow Weighted
8 Average Average
303
63
10.6
22.0
3115
30 32
20 28
0.10
5.0 3.7
11.5
489 478
6.3 6.1
1283
3888 3891
434 418
6.0 6.5 ,
900 1050
•
319.2
61.9
10.6
22.2
31.6
30.5
3.6
11.5
* MLVSS data fra Points A, C, and D Here used for this reactor.
72
-------
TABLE A5. RAW DATA FOR JONES ISLAND EAST PLANT
TIME
1
N TOTAL COO (ig/L)
F SOLUBLE COO (tg/l)
L AMMONIA (ig-N/L)
U ORGAN IC-N (mg-N/L)
E FLOKRATE (M3/day)
N
T
E
F TOTAL COO (ifl/L)
F SOLUBLE COO (Kg/I)
L AMMONIA (ig/L)
U ORGAN IC-N (nig-N/L)
E N03/N02 (ig-N/L)
N
T
R OUR (ig 02/L/day)
E DO (ng-02/L)
A 0 MLVSS (ugA)
C N RAS (MA3/DAY)
T E
0 (POINT A)
R
R OUR (ig 02/L/day)
E 00 (ig-02/L)
A T MLVSS (ig/L)
C N
T 0
0 (POINT B)
R
R OUR U* 02/L/
-------
TABLE A5. RAW DATA FOR JONES ISLAND EAST PLANT CONTINUED
TIME
1
N TOTAL COD (ig/L)
F SOLUBLE COO (19/1 )
L AMMONIA (ig-N/L)
U ORGANIC-N (ng-NA)
E FLOWRATE (M3/day)
N
T
E
F TOTAL COD (ig/L)
F SOLUBLE COD (ig/L)
L AMMONIA (§g/L)
U ORGANIC-N (ng-N/L)
E N03/N02 (ig-N/L)
N
T
R OUR (gg 02/L/day)
E DO (»g-02/l)
A 0 MLVSS (igA)
C N RAS (M"3/DAY)
T E
0 (POINT A)
R
R OUR (ig 02/L/day)
E DO (ig-02/l)
A T MLVSS (ig/L)
C «
T 0
0 (POINT B)
R
R OUR (ig 02/L/day)
E F DO (ig-02/l)
A 0 MLVSS (m/l)
C U
T R
0 (POINT C)
R
(PM) (AM) 7-29-87
8 9 10 11 12 1 2
439 547 496 519
342 332 402 444
12
24.1 39.1 52.3 53.2
13590 13970 11700 11130 9420
50 61 61 50
53 47 55
3.6 4.5
4.2 4.48 4.48 3.92
0.8 2 0.8
2798 2304 1267
0.3 0.25 0.2 0.3
1570 1728
ONLY AVERAGE DAIIY Fl OH WAS
— mn i nfunnuLi UH i L i r mil FWW
1260 1029 905 1306
1 0.8 0.4 1.7
MOT UFi^BFR
*
768 757 741 743 743
4.8 5 4.8 5.1
3 4 5
639
323
14.5 14
39.8
9240 9240 8860
53
53
4.9 4.3
3.92
0.7 0.8
1234
0.4 0.8
1912
DETERMINED
1248 998
1.8 3.2
691 640
4.8 5.2
74
-------
TABLE A5. RAW DATA FOR JONES ISLAND EAST PLANT CONTINUED
TIME
1
N TOTAL COD (iQ/L)
F SOLUBLE COD (ig/L)
L AMMONIA (ig-NA)
U ORGAN IC-N (ig-NA)
E FLOHRATE (M3/day)
N
T
E
F TOTAL COD (Kg/1)
F SOLUBLE COD (Kg/I)
L AMMONIA (ng/L)
U ORGAN IC-N (ig-N/l)
E M03/N02 (mg-N/L)
N
T
R OUR (ng 02/L/day)
E DO (ig-MA)
A 0 MLVSS (ig/l)
C N RAS (M"3/DAY)
T E
0 (POINT A)
R
R OUR (ng 02/l/day)
E DO 0*1-02/1)
A T MLVSS (ig/L)
C III
T 0
0 (POINT B)
R
R OUR (m 02/L/day)
E F 00 (ig-02A)
A 0 MLVSS (m/l)
C U
T R
0 (POINT C)
R
(AM)
6 7 8 9 10 Average
512 611 505 534
246 330 309 377
25 17 16
31.9 30.8 35
9046 9800 10750 14530 12384
65 47 71 53
59 41 53 44
3.8 3.7 5
3.92 4.61 3.78 4.1
0.7 0.8 1
1324 2496 1234 2145
1 0.7 1 0.5
1598 1740
04CA
942 1280 929 1146
3.2 2 3 1.6
.-. NDT tfi^GFTI
&
665 653 683 712
4.7 5.5 5.4 4.9
1880
Flow Weighted
Average
532.8
378.5
16.7
30.6
52.9
41.1
5.3
4.2
1.0
75
-------
TABLE A6. RAW DATA FOR THE SOUTH SHORE PLANT BASIN 17
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
R
E
A
C
T
0
R
R
E
A
r
I/
T
0
R
R
E
A
C
T
0
R
(AM) 7-30-87 (PM)
TIME 9 10 11 12 1 23
TOTAL COD (ng/L) 201 217 223 201
SOLUBLE COD (iig/L) 121 149 121 103
AMMONIA (ng-N/L) 15.5 26 29
TOTAL ORG-N (ng-N/L) 6.72 7.28 9.52 8.96
FLOKRATE (iS/day)
BASIN » 17 WAS NOT MONITORED UNTIL AFTER 10:00 PM
TOTAL COD (us/I) 35 41 50 18
SOLUBLE COD (ig/L) 20 14 32 14
AMMONIA (ng-N/L) 2.8 4.5 2.7
TOTAL ORG-N (ig-NA) 2.24 3.92 2.8 1.96
N03/N02 (iig-N/L) 5 77
OUR (ig 02/L/day)
0 DO (M-02A)
N MLVSS (MA)
E RAS (ITS/DAY)
(POINT A)
OUR (tj 02/L/day)
T DO (IQ-02A)
NIJVS5 (MA) NOT UFA5URFD
0
(POINT B)
T OUR (B 02A/day)
H DO (•Q-02A)
R MLVSS (MA)
E
E
(POINT C AND D)
4 5
214
126
34
8.96
38
20
2.2
2.8
6.5
c
76
-------
TABLE A6. RAW DATA FOR THE SOUTH SHORE PLAKT BASIN 17 CONTINUED
1
N
F
L
U
£
N
T
E
F
F
L
U
E
N
T
R
E
A
C
T
0
R
R
E
A
C
T
0
R
R
E
A
C
T
0
R
0
N
E
T
N
0
T
H
R
E
E
(PM)
TIME 678
TOTAL COD (ig/L) 175
SOLUBLE COD (iiigA) 69
AMMONIA (ig-N/L) 31 36
TOTAL ORG-N (»g-N/L) 6.72
FLOHRATE (§3/day) R 1:
R 2:
R 3:
TOTAL:
TOTAL COD (nig/L) 29
SOLUBLE COD (Bg/L) 14
AMMONIA (ng-N/L) 2.1 3.4
TOTAL ORG-N (iig-N/L) 1.96
N03/N02 (ig-N/L) 5 1.5
OUR (ig 02/L/day)
DO (ig-02/L)
MLVSS (igA)
RAS (ra/DAY)
(POINT A)
OUR (tg 02/L/day)
00 (•Q-02A)
MLVSS (ig/L)
(POINT 8)
OUR (*J 02/L/day)
DO 09-02/1)
MLVSS OV/L)
(POINT C AM) 0)
(AM) 7-31-87
9 10 11 12 1 2
210 243 227
103 132 109
10.6 8.4 7.84
8274 8524 8421 8395
5102 5212 5193 5142
5102 5212 5193 5142
18478 18948 18807 18679
29 29 32
20 26 17
2.24 2.24 2.52
3 7.5 5.5
1029 960
1.8 1 0.8
2.8
NOT If AflBFD _
662 683 620
4.7 3.2 3.3
1068
3
207
143
6.16
8346
5114
5114
18574
17
17
4.98
995
77
-------
TABLE A6. RAW DATA FOR THE SOUTH SHORE PLANT BASIN 17 CONTINUED
(AM) 7-31-87
1
N
F
L
U
E
N
T
E
F
F
L
U
E
N
T
R
E
A
C
T
0
R
R
E
A
C
T
0
R
R
E
A
C
T
0
R
TIME
TOTAL COD (IQ/L)
SOLUBLE COD (ilig/L)
AMMONIA (•g-N/L)
TOTAL ORG-N (ing-N/L)
FLOWATE (US/day) R 1:
R 2:
R 3:
TOTAL:
TOTAL COD (tQ/L)
SOLUBLE COD (ig/L)
AMMONIA (ig-N/L)
TOTAL ORG-N (ng-NA)
N03/N02 (HJ-N/L)
OUR (ig 02/L/day)
0 DISSOLVED OXYGEN (Kg 02/L)
N MLVSS (ugA)
E RAS (MA3/DAY)
(POINT A)
OUR (*) 02/l/day)
T DISSOLVED OXYGEN (qj 02/L)
N MLVSS (HJA)
0
(POINT B)
T OUR (ig 02/L/day)
H DISSOLVED OXYGEN (IQ 02A)
R MLVSS (igA)
E
E
(POINT C AMD 0)
5
148
HO
8.4
7948
4897
4897
17742
28.5
17
3.08
5
988
1
1.8
—
638
3
1052
6789
148 128
122 87
11.2 21.3
8.4 7.84
7646 7233 6942 6442
4694 4454 4289 3976
4694 4454 4289 3976
17034 16141 15520 14394
17
5.04 5.6
1.68
5 6
1005 981
1 1.2
1304
2 2.3
NOT MEASURED
648 614
3.1 3.9
Flow Weighted
Average
195.5
117.3
25.5
8.1
7847.6
4825.4
4825.4
17498.4
30.3
19.2
3.5
2.7
5.3
993
1.1
1304
4444
*v
"*
818
2.1
643.9
3.5
1060.0
Average
186.9
123.4
7.8
25.0
19.3
2.9
5.8
78
-------
APPENDIX B
STEADY-STATE SIMULATIONS
TABLE B.I. PORTAGE LAKE PLANT SSSP STEADY STATE SIMULATION WITH
CALIBRATED PARAMETERS
PROCESS CONFIGURATION AND FLOW DISTRIBUTION
OVERALL PLANT SPECIFICATIONS:
Number of Reactors (up to 9) - 2
Solids Retention Time (days) - 10.6
Average Flow Rate (m3/day) - 3547
INDIVIDUAL REACTOR SPECIFICATIONS:
Reactor Volume (m3) -
Feed Fraction (0 to 1)
Mass Transfer Coeff for 02 (day-1) -
Recycle Input (m3/day)
Reclrculat Ion Input (m3/day) -
Reclrculat Ion originated from reactor -
STEADY-STATE SOLUTION WITH FEED FRACTIONATION 'BAS.ED
CONSTITUENTS
Heterotroph ic Organisms
Autotrophic Organisms
Partlculate Products
Inert Partlculates
Partlculate Organlcs
Soluble Organlcs
Soluble Ammonia N
Soluble Nitrate/Nitrite
Soluble Organic N
Blodegrad Part Organic
Oxygen
Alkalinity
MLVSS
02 Consumed
FEE&
g
a
g
g
g
g
g
N g
g
N g
g
cod
cod
cod
cod
cod
cod
n
n
n
n
02
mole
g
g o2
Nitrate Consumed g no3-n
cod
m-3
ro-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
d-1
d-1
•
-
-
-
•
•
-
•
•
-
•
-
.
•
-
0
0
0
13
122
107
5
0
2
2
0
5
.0
.a
.0
.6
.4
.3
.8
.0
.2
.5
.0
.0
1
2
530 350
0.00 1.00
72.0 72.0
3653 0
0 0
* *
ON TOTAL & SOLUBLE COD
1
1466
26
1392
731
496
3
0
2
0
35
1
4
4114
695
12
.9
.5
.8
.1
.8
.3
.5
.8
.6
.0
.2
.4
.1
.0
.8
2
766
13
705
375
294
9
0
1
0
18
1
4
2155
662
9
.9
.4
.1
.2
.9
.1
.7
.6
.7
.8
.5
.5
.5
.8
.8
79
-------
TABLE 8.2. GREEN BAY PLANT SSSP STEADY-STATE SIMULATION WITH
CALIBRATED PARAMETERS
PROCESS CONFIGURATION AND FLOW DISTRIBUTION
OVERALL PLANT SPECIFICATIONS:
Number of Reactors (up to 9) - 2
Solids Retention Time (days) - 3.1
Average Flow Rate (m3/day) - 58879
INDIVIDUAL REACTOR SPECIFICATIONS: 1_
Reactor Volume (m3) - 5607 10443
Feed Fraction (0 to 1) - 0.00 1.00
Mass Transfer Coeff for 02 (day-1) - 118.0 311.0
Recycle Input (n»3/day) - 38840 0
RecirculatIon Input (m3/day) -00
RecirculatIon originated from reactor * *
STEADY-STATE SOLUTION WITH FEED FRACTIONATION BASED ON TOTAL 4 SOLUBLE COD
CONSTITUENTS
Particulate Products
Inert Part iculates
Partfculate Organics
Soluble Organics
Soluble Ammonia N
Soluble Nitrate/Nitrite
Soluble Organic N
FEED
9
9
9
9
9
N g
9
Biodegrad Part Organic N g
Oxygen
Alkalinity
MLVSS
9
cod
cod
cod
cod
n
n
n
n
o2
mole
9
02 Consumed g o2
Nitrate Consumed g no3-n
STEADY-STATE SOLUTION WITH FEED
CONSTITUENTS
Partlculate Products
Inert Part Iculates
Particulate Organics
Soluble Organ fcs
Soluble Ammonia N
Soluble Nitrate/Nitrite
Soluble Organic N
Biodegrad Part Organic N
Oxygen
Alkalinity
MLVSS
02 Consumed g
cod
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
d-1
d-1
—
-
•
-
-
-
-
-
•
«.
m
-
0
30
278
510
36
0
12
6
0
3
V
.0
.9
.4
.2
.7
.0
.2
.7
.0
.9
FRACTIONATION BASED ON
10
FEED
9
9
9
9
9
N g
9
9
9
cod
cod
cod
cod
n
n
n
n
02
mole
9
02
Nitrate Consumed g no3-n
cod
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
d-1
d-1
.
•
•
•
•
m
m
m
m
m
m
m
m
0.
30.
536.
253.
36.
0.
6.
12.
0.
3.
0
9
0
0
8
0
1
8
0
9
1
678
567
704
2
39
0
0
48
0
4
5670
1133
0
.7
.3
.8
.6
.6
.0
.3
.2
.1
.1
.2
.1
.0
BATCH
1
679.
567.
681.
4.
37.
0.
0.
41.
0.
3.
5648.
1139.
0.
1
3
0
9
7
0
2
8
0
9
9
5
0
2
272
238
307
7
32
0
0
18
2
3
2445
2201
., 0
.1
.0
.0
.9
.2
.0
.7
.1
.5
.6
.8
.5
.0
REACTOR DATA
2
272.
238.
299.
7.
32.
0.
0.
14.
2.
3.
2438.
2201.
0.
2
0
0
9
7
0
7
4
5
6
7
2
0
80
-------
TABLE 8.3. MADISON PLANT SSSP STEADY-STATE SIMULATION WITH
CALIBRATED PARAMETERS
PROCESS CONFIGURATION AND FLOW DISTRIBUTION
OVERALL PLANT SPECIFICATIONS:
Number of Reactors (up to 9) - 4
Solids Retention Time (days) - 16.4
Average Flow Rate (m3/day) - 20346
INDIVIDUAL REACTOR SPECIFICATIONS: 1 2 3 4
Reactor Volume (m3) - 1733 4008 3827 1914
Feed Fraction (0 to 1) - 1.00 0.00 0.00 0.00
Mass Transfer Coeff for 02 (day-1) - 45.0 68.0 50.0 51.0
Recycle Input (m3/day) - 13552 000
Reclrculation Input (m3/day) -0000
RecirculatIon originated from reactor - * * * *
STEADY-STATE SOLUTION WITH FEED FRACTIONATION BASED ON TOTAL & SOLUBLE COD
CONSTITUENTS
Heterotrophic Organisms
Autotrophic Organisms
Partlculate Products
inert Partlculates
Particulate Organics
Soluble Organics
Soluble Ammonia N
Soluble Nitrate/Nitrite
Soluble Organic N
Biodegrad Part Organic N
Oxygen
Alkal inlty
MLVSS
O2 Consumed g
FEED
g
g
g
g
g
g
g
N g
g
g
g
cod
cod
cod
cod
cod
cod
n
n
n
n
02
mole
g
02
Nitrate Consumed g no3-n
cod
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
d-1
d-1
0
0
0
- 13
- 119
- 116
• 14
0
- 5
- 5
0
3
.
-
-
.0
.0
.0
.3
.8
.6
.0
.0
.5
:e
.0
.9
1
668
46
544
386
130
11
7
0
0
8
0
3
1774
469
100
.0
.5
.3
.0
.0
.7
.7
.2
.7
.4
.1
.4
.8*
.2
.0
2
671.3
47.2
548.4
386.0
93.3
2.0
3.4
5.6
0.5
6.6
2.3
2.7
1746.2
537.9
3.8
3
663
47
552
386
69
1
1
9
0
5
2
2
1718
400
2
.1
.5
.2
.0
.6
.5
.2
.3
.4
.3
.5
.3
.4
.0
.7
4
657.5
47.6
554.1
386.0
60.2
1 .4
0.6
10.7
0.4
4.8
3.5
2.2
1705.4
340.4
1.8
81
-------
TABLE B.4. MONROE PLANT SSSP STEADY-STATE SIMULATION WITH
CALIBRATED PARAMETERS
PROCESS CONFIGURATION AND FLOW DISTRIBUTION
OVERALL PLANT SPECIFICATIONS:
Number of Reactors (up to 9) -
Solids Retention Time (days) -
Average Flow Rate (m3/day) -
2
8.4
3115
INDIVIDUAL REACTOR SPECIFICATIONS:
Reactor Volume (m3)
Feed Fraction (0 to 1)
Mass Transfer Coeff for 02 (day-1)
Recycle Input (m3/day)
Recirculation input (m3/day)
RecirculatIon originated from reactor -
•
«
.
-
m
m
1
852
0.90
155.0
3891
0
*
2
996
0.10
115.0
0
0
*
STEADY-STATE SOLUTION WITH FEED FRACTIONATION BASED ON TOTAL & SOLUBLE COD
CONSTITUENTS
Heterotrophic Organisms
Autotrophic Organisms
Paniculate Products
Inert Particulates
Particulate Organ Ics
Soluble Organ Ics
Soluble Ammonia N
Soluble Nitrate/Nitrite
Soluble Organic N
Blodegrad Part Organic N
Oxygen
Alkalinity
MLVSS
02 Consumed g
FEED
g
g
g
g
g
g
g
N g
g
g
g
cod
cod
cod
cod
cod
cod
n
n
n'
n
o2
mole
g
o2
Nitrate Consumed g no3-n
cod
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
d-1
d-1
«
•
m
m
m
m
m
m
m
m
m
-
.
•
-
0
0
0
25
231
31
10
0
2
14
0
5
.0
.0
.0
.7
.6
.4
.6
.0
.0
.9
.0
•0.
1
686
31
372
375
145
3
3
14
0
11
6
3
1610
486
1
.3
.7
.2
.4
.2
.4
.3
.4
.7
.0
.4
.4
.8
.7
.5
2
658
30
361
359
103
2
2
17
0
8
€
3
1513
389
1
.4
.7
.5
.7
.4
.6
.0
.7
.6
.2
.4
.1
.8
.9
.2
82
-------
TABLE B.5 JONES ISLAND EAST PLANT SSSP STEADY-STATE SIMULATION WITH
CALIBRATED PARAMETERS
PROCESS CONFIGURATION AND FLOW DISTRIBUTION
OVERALL PLANT SPECIFICATIONS:
Number of Reactors (up to 9) - 5
Solids Retention Time (days) - 2.8
Average Flow Rate (m3/day) - 12380
INDIVIDUAL REACTOR SPECIFICATIONS:
Reactor Volume (m3)
Feed Fraction (0 to 1) -
Mass Transfer Coeff for 02 (day-1)
Recycle Input (m3/day)
Recirculation Input (m3/day) -
Recirculation originated from reactor -
1
240
1.00
61 .0
3460
0
*
2
601
0.00
105.0
0
0
*
3
601
0.00
98.5
0
0
*
4
1203
0.00
105.0
0
0
*
5
962
0.00
103.0
0
0
*
STEADY-STATE SOLUTION WITH FEED FRACTIONATION BASED ON TOTAL & SOLUBLE COD
CONST I TUENTS
Particulate Products
Inert Part iculates
Part iculate Organics
Soluble Organics
Soluble Ammonia N
Soluble Nltrate/Nitr)
Soluble Organic N
g
g
g
g
g
te N g
g
Biodegrad Part Organic N g
Oxygen
Alkal inity
MLVSS
02 Consumed
Nitrate Consumed g
g
cod
cod
cod
cod
n
n
n
n
02
mole
g
g o2
no3-n
cod
m-3
m-3
m-3 -
m-3 -
m-3 -
m-3 «
m-3 -
m-3 -
m-3 -
m-3 -
m-3-
m-3-
m-3-
d-,1 .
d-1 .
FEED
0.0
15.4
138.9
337.4
16.7
0.0
18.8
7.7
0.0
5.0
1
158
148
226
239
28
0
2
15
0
5
2523
616
6
.1
.0
.5
.7
.3
.0
.6
.3
.0
.8
.3
.7
.9
2
160
148
230
138
24
0
0
16
0
5
2584
1061
1
.3
.0
.7
.2
.9
.0
.3
.1
.0
.6
.4
.2
.5
3
162.6
148.0
234.1
44.4
20.0
0.0
0.1
16.8
0.0
5.2
2639.7
995.7
1.5
4
167.3
148.0
163.7
2.0
19.0
0.2
0.3
12.2
3.0
5.2
2627.0
715.3
3.0
5
17}. 2
148.0
123.1
1.2
20.4
0.4
0.3
9.6
5.3
5.2
2599.5
459.6
1.6
STEADY-STATE SOLUTION WITH FEED FRACTIONATION BASED ON BATCH REACTOR DATA
CONSTITUENTS
Particulate Products g cod m-3 -
Inert PartIculates
Particulate Organics
Soluble Organics
Soluble Ammonia N g
Soluble Nitrate/Nitrite N g
Soluble Organic N g
Biodegrad Part Organic N g
Oxygen g o2
Alkalinity mole
MLVSS
02 Consumed
Nitrate Consumed
FEED
114.5 116.1 117.8 121.2 124.0
g cod m-3- 15.4 148.0
g cod m-3 - 325.0
g cod m-3 - 151.0
n
n
n
n
m-3-
m-3-
m-3 -
m-3-
m-3-
m-3 -
g cod m-3 -
g o2 m-3 d-1-
g no3-n m-3 d-1-
16.7
0.0
8.4
18.1
0.0
5.0
368.5
88.9
19.5
0.0
1.1
22.3
0.0
5.2
148
344
11
15
0
0
21
0
4
.0
.5
.1
.5
.0
.2
.3
.1
.9
148
279
2
14
0
0
17
1
4
.0
.3
.2
.3
.5
.4
.7
.0
.8
148
180
1
13
2
0
12
3
4
.0
.3
.5
.0
.5
.3
.0
.2
.6
148.0
124.5
1.1
12.1
4.4
0.2
8.7
4.3
4.4
2847.92886.1 2864.52819.9 2791.5
616.81056.9 871.6 702.1 581.1
67.4 13.1 14.8 4.6 2.8
83
-------
TABLE 8.6. SOUTH SHORE PLANT SSSP STEADY-STATE SIMULATION WITH
CALIBRATED PARAMETERS
PROCESS CONFIGURATION AND FLOW DISTRIBUTION
OVERALL PLANT SPECIFICATIONS:
Number of Reactors (up to 9) - 3
Solids Retention Time (days) - 4.3
Average Flow Rate (m3/day) - 17498
INDIVIDUAL REACTOR SPECIFICATIONS:
Reactor Volume (m3)
Feed FractIon (0 to 1)
Mass Transfer Coeff for 02 (day-1)
Recycle Input (m3/day)
Reclrculation Input (m3/day)
Recirculation originated from reactor
•
•
-
•
.
•1
1
1112
0.45
121.0
4467
0
*
2
1112
0.28
82.0
0
0
«
3
2506
0.28
70.0
0
0
*
STEADY-STATE SOLUTION WITH FEED FRACTIONATION BASED ON TOTAL & SOLUBLE COD
CONSTITUENTS
Heterotrophic Organisms g
Autotrophic Organisms
Particulate Products
Inert Particulates
Particulate Organics
Soluble Organics
Soluble Ammonia N
Soluble Nltrate/Nitrl
Soluble Organic N
Biodegrad Part Organ i
Oxygen
Alkalinity
MLVSS
02 Consumed
Nitrate Consumed g
g
9
g
g
g
g
te N g
«g
c N g
g
cod
cod
cod
cod
cod
cod
n
n
n
n
o2
mole
g
g o2
no3-n
cod
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
m-3
d-1
d-1
-
-
-
-
-
-
-
-
-
-
-
•
.
m
-
FEED
0.0
0.0
0.0
6.3
57.1
104.1
15.0
0.0
3.5
1.9
0.0
5.0
1
620
46
117
128
93
3
4
10
0
5
2
3
«
1005
1023
8
.0
.6
.2
.7
.0
.4
.2
.5
.6
.7
.2
.5
.5
.6
.8
2
450
33
84
93
68
3
4
10
0
4
2
3
731
707
6
.8
.7
.9
.4
.8
.1
.0
.6
.6
.2
.1
.5
.6
.9
.2
3
353
26
67
73
48
2
2
12
0
3
3
3
570
464
2
.6
.8
.6
.7
.7
.2
.6
.5
.5
.1
.9
.2
.4
.7
.1
84
-------
TECHNICAL REPORT DATA
/Plcost read Instructions on the reverse before completing'/
I. REPORT NO.
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE ANOSUBTITLE
Oxygen Utilization in Activated Sludge Plants:
Simulation and Model Calibration
5. REPORT DATE
June, 1988
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
C. Robert Baillod
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Michigan Technological University
Houghton, Michigan 49931
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
CR813162-01-2
12. SPONSORING AGENCY NAME AND ADDRESS
Water Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final, June 1986 to June 1988
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The objective of the research described 1n this report 1s to apply recent advances
in activated sludge process modeling to the simulation of oxygen utilization rates in
full scale activated sludge treatment plants. This 1s accomplished by calibrating the
International Association for Water Pollution Research and Control (IAWPRC) Model and
associated SSSP micro-computer software to operating data at six full scale activated
sludge treatment plants. Field data were used to calibrate the key biological
parameters contained 1n the model so that the oxygen utilization rates, dissolved
oxygen concentrations, mixed liquor volatile suspended sol Ids concentrations, and
process performance simulated by the model matched the corresponding quantities
observed 1n the treatment plants.
The results showed that the model and associated software package provide a
useful capability to analyze, simulate, and predict oxygen utilization rates. It was
possible to obtain reasonable agreement between the measured and simulated values of
oxygen uptake rate, dissolved oxygen concentration and other process parameters at
most of the plants studied. The key model parameters were the heterotrophlc yield
coefficient, heterotrophlc decay constant, and autotrophlc maximal specific growth
rate constant.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DcscnirroRS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
»•»-«* 18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report/
unclassified
21. NO. OF PAGES
20. SECURITY CLASS fThit pagt)
unclassified
22. PRICE
EPA Perm 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE
-------