-------
r—
03 O)
+J E
o
•a o
OJ _Q
0) S-
Lu R]
O
•k
S. C
O^»
tJ
rj .a
cr s.
•r- ra r—
CD
•a r— E
cu
•r- O
z:
u_
M
O
M- C
•r—
cu E
O r—
>> E
u
cu
cs:
Figure 29. Response characteristics with
and without F/M control —
high loading
600 r-
500
400
Control Simulation
Blank Simulation
1000
f\ C\f\
900
800
700
600
1,0
0,9
0,8
0,7
Or
Feed Flow 2325 ml/min
r~ Initial Recycle Ratio 0.3
Initial Loading 2.1 gm TCp/gm TCM«day
^^-^•^
-i^^^
^^^^
— s' ^^.-~ — — —
^^ _ m " ""^
^r __ - •""
"" J^'~'~~~~~
-
—
- Jv
~ \r~~~-—
- l^—^l
,6 ' j -: —
900 r- 10 32,5
800
700
_
1 \^^_
1 1 1 1 1
0 10 20 30 40 50
Time Since Step Change, hours
73
-------
Figure 30. Comparison of aeration basin
substrate concentration with
and without F/M control
(O
c
1,0
0,9
0,8
0,7
0,6
Digital Computer Simulation
Feed Flow 2325 ml/min
Initial Recycle
Ratio 0.3
— •**. Blank
F/M Control
C
O
•r-
-H
E
C O
4->
cd O
i- +J
-»->
l/J
70
60
50
40
30
20
10
0
I
Blank
F/M Control
I
I
I
I
0 10 20 30
Time Since Step Change, hours
50
-------
Figure 31. Response characteristics with
and without F/M control
(salt-propylene glycol system)
tO C7>
•»-> £
o
+-> •
c.
•a o
cu
.
O r—
>> E
o
(U
1000
800
0,9
0,8
0,7
0,6
0,5
1200
1100
1000
1 Control Simulation
Blank Simulation
Feed Flow
Initial Recycle Ratio
Initial Loading
1938 ml/min
0.55
1.5 gm TCp
gm TCM«day
I
I
I
!
0 20 40 60 80 100
Time Since Step Change, hours
75
-------
SECTION VI
NUTRIENTS ADDITION CONTROL SYSTEM
An industrial waste stream to be treated by biological
oxidation may not have sufficient concentrations of
nitrogen and phosphorus to supply the amounts necessary
for cell production and respiration requirements.
Constant nutrients addition, based only on flow, will
occasionally result in excess or insufficient nutrients
addition. Insufficient addition decreases bacterial
viability while excess nutrients addition, besides being
economically unsound, can cause downstream algae blooms
and subsequent pollution problems. High nutrient resid-
ual may also prevent recycle of industrial treated
effluents.
The solution to all of these problems lies in providing
a varying flow of nutrients to the aeration basin based
upon a measurement of the soluble organic substrate
concentration in the feed.
SYSTEM DESIGN
The designed nutrients addition control system was operated
on the basis of the addition of an ammonia and a phosphoric
acid solution proportional to an analysis of total carbon
in the feed stream. The feed sample was passed through
a 3-tf-icron Cuno filter and the total carbon content was
measured by an Ionics Model 1212 Total Carbon Analyzer
(TCA) operating on a six-minute analysis cycle. The
recorder peak height was converted to an equivalent electric-
al signal of sufficient magnitude to match a peak picker
76
-------
sensor. The output of this sensor was transferred to a
sample and hold amplifier by a programmer synchronized
with the analyzer. The electrical signal was converted
to a pneumatic signal which linearly adjusted the per-
cent energized time of a 30-second cycle timer that
powered a nutrients addition pump (see Figures 10 through
13).
The nutrients pump maintained a given flow until its con-
trols received an up-dated analysis signal. If the
period between analyses extended to hours or days, some
slow reduction in the pump rate was anticipated due to
decay in the sample and hold amplifier output.
Proportionality of nutrients addition was varied by
changing the influent flow, the carbon analyzer calibra-
tion, the full scale pumping rate, and the concentration
of the nutrients solution. If required, additional varia-
tion in proportionality was available by altering the
calibration of various intermediate hardware components.
The same total carbon analysis (TCA) that was used for
F/M control (Section V) was also used for the nutrients
control system. A flow chart and block diagram which
includes both the F/M and the nutrients control system
are given in Figure 10 and Figure 13, respectively. An
electrical schematic of the controls is given in Figure 12,
SYSTEM TESTING AND PERFORMANCE
The time lag from the peak on the TCA to the air signal
operating the nutrient pump was less than 10 seconds and
the resulting linear response is shown in Figure 32.
The control system performed reliably during the nine
months of miniplant operation.
77
-------
Figure 32. Response of nutrients control system
25
20
15
CO
I 10
i.
0
I I I I I I I I I
0 10 20 30 40 50 60 70 80 90 100
Total Carbon in Feed, % of scale
78
-------
Routine analyses for the ammonia and phosphorus concentra-
tions of the bio-settler overflow were made in order to
determine the best C:N:P ratio for the miniplant operation.
Ammonia and ammonium ion concentrations were determined by
the Kjeldahl Nitrogen method. Total phosphorus concentra-
tion was determined by a molybdate colorimetric method,
after co-precipitation of the phosphorus with iron.
Figure 33 shows the relationship between ammonia consump-
tion and microbial growth. Data were calculated by material
balance on a daily basis. Although the experimental data
reflects the ammonia consumption for cellular growth plus
the ammonia consumed by adsorption onto the bio-floe, the
results compare closely with the theoretical ammonia con-
sumption for cellular grwoth only. For a cell formula of
C5H702N, 0.15 grams of ammonia would be required for the
growth of one gram of new cells. This compares with the
experimental value of 0.133 gm NH3/gm cells formed.
Figure 3^ shows that the phosphorus consumption is directly
proportional to the ammonia consumption. The phosphorus
required is 0.06 gram per gram of ammonia. This translates
to 0.008 gram of phosphorus required per gram of microbial
cells formed.
The linear curves of Figures 33 and 3^ illustrate the
theoretical basis for feed-forward nutrients control.
Because microorganism growth is directly proportional to
the incoming feed concentration and because the nutrients
consumption is directly proportional to the microorganism
growth (Figure 33), the nutrient utilization will vary
directly with the incoming feed concentration.
79
-------
Figure 33. Ammonia consumption by microorganisms
0,7
0,6
0,5
n.
E
3
03
•I—
c
O
£
0,3
0,2
0,1
0,0
Slope = 0.133
°o
I
111!
0,0 1,0 2,0 3,0 4,0 5,0
Microorganism Growth, Ib vss/day
80
-------
Figure 34.
Relationship between phosphorous
and ammonia consumptions
Q.
£
C
o
o
o
5-
o
-C
ex
10
o
0,07
0,06
0,05
0,04
0,03
0,02
0,01
0,00
8
0,0 0,2 0,4 0,6 0,8
Ammonia Consumption, Ib/day
1,0
81
-------
In order to determine the minimum nutrients requirement,
the ratio of nutrients addition to feed total carbon
concentration was progressively lowered until analysis
of the bio-settler overflow indicated a low concentration
of ammonia and phosphorus. The results shown in Figure 35
indicate a large adsorptive capacity of the mixed liquor
bio-floe for values (nitrogen fed/carbon fed x 100)
greater than 20. Material balance calculations show that
without any bio-adsorption, for every ten units increase
in the abscissa value above the minimum value, a 40 ppm
increase in the bio-settler overflow ammonia concentration
would have occurred. Because of the high adsorption, if
the data of (NiC x 100) above 20 in Figure 35 were plotted
in Figure 33 they would lie considerably higher than the
linear curve presented.
The results in Figures 3^ and 35 indicate an optimum car-
bon-to-nitrogen ratio of 100 to 9 and an optimum carbon-
to-phosphorus ratio of 100 to 0.5^. Using this propor-
tionality the nutrients control system assured microbial
viability and the effluent concentration was maintained
at less than 10 ppm ammonia and less than 2 ppm phosphorus.
This control system could be modified to allow variations
in influent flow rate by flow measurement and increased
nutrients addition corresponding to the multiple of the
flow times the feed total carbon analysis.
82
-------
•M
c
O)
£
0)
s-
cr
01
S-
c:
o
E
E
(O
c
o
i-
cr>
O
S-
CD
LO
CD
CD
mdd
CD
CM
BLUOUlUiy
CD
83
-------
SECTION VII
CHEMICAL PLOCCULATION CONTROL SYSTEM
The overflow of the bio-settler Is usually too turbid for
release without some additional clarification. This is
especially true in the treatment of saline waste waters,
where overflow turbidities are unusually high.
In the pilot plant, a chemical flocculation control system
was used to remove suspended and colloidal matter from both
the fresh water and the saline biologically treated effluents
DESCRIPTION OF THE SYSTEM
The schematic in Figure 36 illustrates the flocculation
controls and process flows. The proportional control was
based on a turbidity measurement. Aluminum sulfate was
used as the flocculant, and the pH was controlled (propor-
tional-integral) at the level corresponding to the optimum
for alum flocculation.
Both feed-forward and feed-back modes of control were tested.
An Ecologic Model 204 Turbidimeter was used for feed-back
control but fouling of the sensor window surfaces was a
reoccurring problem. A timer-operated water flushing sys-
tem was therefore installed to clean the probe every 100
minutes. This somewhat improved the instrument operation,
but significant fouling still occurred.
The Ecological turbidimeter was replaced by a Hach Surface
Scatter No. 2^126 Turbidimeter in which there is no contact
between the sample and the optical surface. In both feed-
back and feed-forward control the entire settler overflow
was passed through the instrument.
-------
CO
o
s-
o
o
c
o
(O
O
o
o
10
85
-------
The pH of the flocculator effluent was measured with
conventional glass electrodes and recorded on a pneumatic
recorder-controller. The caustic added was a 2-percent
NaOH solution. A high pH alarm automatically closed the
control valve at values above 8.3 and aided in control
recovery under high pH-upset conditions.
The pilot plant chemical flocculation control system was
operated at a constant overflow rate in the range of
1,800 to 2,200 ml/min.
OPTIMUM pH AND ALUM DOSAGE
The stability of bio-colloids is dependent upon the surface
charge density. In general, the microorganisms carry a net
negative charge, within the pH-range of interest. The
charge is acquired through acid-base interactions of func-
tional ionogenic groups and, thus, the biota surface charge
density is strongly pH dependent (Tenney and Stumm, 1965).
Flocculation is not a result of the aluminum ion Al (III)
but of its hydrolysis products (Stumm and Morgan,
1962j Stumm and O'Melia, 1968). The solution pH is the most
important parameter in determining which particular polymeric
hydrolysis specie predominates. This in turn influences the
polymer bridging characteristics which cause flocculation.
Jar Tests
The optimum pH for alum flocculation was determined by a
series of jar tests at various pH values. A one-liter
sample of the bio-settler overflow was placed in a beaker
and 1M HC1 or NaOH was added to bring the sample to the
desired pH. The pH was held constant by adding a predeter-
mined amount of caustic simultaneous with the alum dose
86
-------
during a one-minute, rapid-mixing phase. After an additional
slow stirring at 30 rpm for 15 minutes, the sample was
allowed to settle for 20 minutes. The supernatant was
then .siphoned and analyzed for turbidity.
To determine the optimum pH for flocculation, the tests
were performed on samples having the same initial turbidity.
There was an increase in the optimum alum dosage for floc-
culation above pH of 6 and below pH of 5. Also, the height
of the settled floe subsidence line was much higher when the
pH was above 7 and when the alum doses were greater than
1 millimole per liter.
Optimum Alum Dose as a Function of Turbidity
The meaning of optimum alum dosage is illustrated in Figure
37. If the dosage of alum is less than the optimum,
effluent quality will diminish, while doses above the opti-
mum are economically unsound. Also, high alum doses were
found to cause light, fluffy floes, which sometimes resulted
in settler upsets because of poor settling and high sub-
sidence levels. This would result in less compacted
sludge and necessitate increased sludge dewatering.
There exists a stoichiometric relationship between the
bio-settler overflow turbidity and the optimum alum dose
for flocculation, at constant pH (Figure 38).
The effect that the physiological condition of the bacteria
has on the optimum alum dosage for flocculation is illustra-
ted in Figure 38. Data point 5 represents a series of jar
tests on a sample of bio-settler overflow with an artificially
increased initial turbidity. The turbidity of the overflow
87
-------
Figure 37. Alum flocculation jar tests
33
30
27
24
21
r
-Q
i.
15
12
9
0,0
Constant pH 6
Initial Turbidity 31
Optimum
Dose, 0.26
1,0 2,0 3,0
Alum Dose, mil 1imoles/1iter
-------
Figure 38. Optimum alum dose as a
function of initial turbidity
pH 6
Fresh Water Biota
CD i-
l/> (U
O -M
Q T-
3 10
i— o;
-------
sample was increased from 25 JCU to 44 JCU by adding
aeration basin mixed liquor to the sample. Thus the sample
represents larger bio-colloids of older microorganisms. On
the other hand, data point 6, where a higher optimum alum
dose was required, represents a case of finely dispersed
microorganisms. Point 6 is taken from continuous flow
miniplant data. Here the alum flow was increased just to
the level of good flocculation. The high, bio-settler over-
flow turbidity was the result of a very high growth rate
of microorganisms in the aeration basin.
SYSTEM OPERATION AND PERFORMANCE
The Hach Surface Scatter Turbidimeter worked well in both
modes of control. When operating in the feed-forward mode
it was necessary to drain the instrument every two to four
hours in order to flush a buildup of bacteria on the in-
side walls of the turbidimeter body cylinder.
Because of the lag time of several hours associated with the
flocculator and final settler, control stability was a
major problem in feedback control.
Since in feedback control an increase in the floe-settler
turbidity was necessary before the controller would increase
the alum flow, final settler upsets were not uncommon. High
alum additions because of a slight settler upset would some-
times result in a fluffy, poor-settling floe which would
further upset the condition. High alum doses sometimes
caused a milky white effluent, increasing the overflow
turbidity and causing total instability.
The feed-forward mode of proportional control resulted in a
high quality effluent. Stability was not a problem.
90
-------
The feed-forward control was tested in both the propylene-
glycol fresh water system and the glycol-salt waste water
system. Typical performance of the control system is
given in Table 4 and Table 5.
The calibration of the optimum alum dose versus overflow
turbidity shown in Figure 38 was used for the feed-forward
control for the fresh water system. A ten-percent higher
calibration was necessary for the salt system.
Dosages of alum were sufficient to remove residual phosphate
(about 2 ppm phosphorus) due to excess nutrient addition.
Higher concentrations of phosphorus required additional alum
in stoichiometric proportions.
An average of 98 percent of the suspended solids were removed
by the flocculation step in the fresh water system and 85
percent were removed in the salt system.
The chemical flocculation control system increased process
efficiency in terms of TOD removal. (TOD data was taken
from samples passed through a 25-micron Cuno filter.) For
the fresh water system, an average of 25 percent of the
bio-settler TOD was removed in the flocculation step. This
increased the overall process efficiency from 90.9-percent
TOD removal to 93-4 percent. A 37-percent reduction in
effluent TOD resulted in an increase in the average, overall
process efficiency from 7^.5 percent to 85.4 percent for
the saline propylene glycol system.
The turbidity of the final fresh water effluent was reduced
to 2 to 3 Jackson Candle Turbidity Units (JCU); whereas,
the bio-settler overflow turbidity was typically 15 to 25 JCU.
91
-------
Table 4. ALUM FLOCCULATION SYSTEM PERFORMANCE,'
GLYCOL-FRESH WATER SYSTEM
Effluent Before
Chemical Flocculation
Effluent After
Chemical Flocculatlon
Day of
Operation
1
2
3
4
5
6
7
8
9
10
MLSS,
mg/1
45
22
20
20
11
36
52
86
34
55
TOD Removal,
%
88.8
95.5
93.5
89.3
92.0
95-2
84.1
90.7
88.3
91.7
MLSS,
mg/1
4
0
0
0
0
1
0
0
0
0
TOD Removal,
%
89.7
96.0
95.8
93.0
93.9
95.4
87.7
94.9
94.0
93.9
a proportional control with optimum alum dosage according
to Figure 38
effluent turbidity maintained at 2 to 3 JCU
92
-------
Table 5. ALUM FLOCCULATION SYSTEM PERFORMANCE,'
GLYCOL-SALT WATER SYSTEM
Effluent Before
Chemical Flocculation
Effluent After
Chemical Flocculation
Day of
Operation
1
2
3
4
5
6
7
8
9
10
MLSS,
mg/1
91
164
52
43
23
214
66
43
52
40
TOD Removal,
%
76.2
78.9
63.2
66.6
90.3
75.8
71.5
69.0
75.4
78.4
MLSS,
mg/1
24
17
0
3
0
18
0
19
21
18
TOD Removal,
%
91.9
91.9
65.9
90.8
96.3
80.3
81.0
aalum dosage at optimum (10$ greater than doses in
Figure 38)
93
-------
The bio-settler overflow turbidity for the saline system
was generally 35 to 80 JCU and the turbidity of the final
effluent was lowered in the same proportion as the suspended
solids removal, about 85 percent.
This control system could be modified to allow variations
in the flow rate by measuring the flow and increasing the
alum addition by the product of the flow times the turbidity.
-------
SECTION VIII
BIOLOGICAL INHIBITOR DETECTOR
The kinetic reactions within a biological oxidation system
proceed at rates defined by the nature of the chemical con-
stituents in the waste water and the activity of the biomass.
A real-time measure of enzyme activity to detect the presence
of inhibitors or toxins in the feed to an activated sludge
plant is required for optimization of the process.
The main objective of this study was to develop an upstream
sensing device for toxic loads to an activated sludge process,
Such a feed-forward control system is essential in an indus-
trial or combined plants to prevent the loss or inhibition of
the acclimated biomass. If the presence of a toxin in the
feed to the plant can be detected rapidly enough on a
repetitive basis, the feed could be diverted to a holding
pond until the toxin is identified and eliminated.
BACKGROUND INFORMATION
Microbiological activity is a dynamic concept that expresses
the ability of the microorganisms to interact with their
environment. Inhibition of the activity of an acclimated
culture may result from an adverse effect on the enzyme
catalytic ability, the cell membrane permeability, or a
general systemic effect by destroying the cellular integrity.
Toxic, heavy metal ions affect enzymes by binding reversibly
or irreversibly at the enzyme-active site or by causing
conformational changes in the enzyme. Some organic inhibit-
ors alter the cell membrane permeability, preventing the
organisms from maintaining an environment favorable for
proper metabolic activity. Heat and chlorine have a general
systemic effect through denaturation of protein and other
cellular components.
95
-------
The enzyme theory of Mlchaelis and Menten (1913), used by
biochemists to adequately describe the toxic effects on
purified enzymes, was applied to describe the inhibition
of activated sludge (Hartmann and Laubenberger, 1968).
The application of enzyme inhibition equations to a mixed
culture represents the overall kinetics of the biomass
since it is not known if the toxicity effects are identical
or different for various species of organisms. An inhibitor
at a certain concentration may be toxic to some species but
may stimulate the growth of others. The varying composition
of the feed to an activated sludge plant adds to the problem
in the study of toxicity. Some components may be syner-
gistic or antagonistic effects with the toxin being studied
(Poon and Bhayani, 1971).
The determination of different types of inhibitions of
enzyme-catalyzed reactions with more than one substrate
and product involves kinetic equations that are considerably
more complex than the basic Michaelis-Menten equation
(Cleland, 1963). Three types of inhibition patterns are
commonly recognized from the kinetic analyses of experi-
mental data:
1. Competitive inhibition results when inhibitor and sub-
strate compete directly for the same enzyme site. Thus
increased substrate concentration would reduce the
effect of the inhibitor.
2. Uncompetitive inhibition is reaction of the inhibitor
with the enzyme which can still undergo another reaction
with the substrate. This type results in decrease of
reaction velocity with increased inhibitor concentration
since there is no direct competition between inhibitor
and substrate for the free enzymes.
96
-------
3. A mixed or noncompetitive inhibition is a mixture of
the two mechanisms and is what takes place in a mixed
culture such as the activated sludge.
The biological effects of toxic wastes have been convention-
ally evaluated by standard sanitary tests that measure the
unit operation efficiency (BOD and COD reduction), biological
population density (MLVSS and SVI), and biological oxidation
capacities (sludge age, sludge yield, and rate of BOD exer-
tion) .
These measurements have been used by many investigators to
assess the effects of toxic substances on the gross unit
efficiency of an activated sludge system (Barth, et al.,
1965; Banerji, et al., 1968; Ghosh and Zugger, 1973;
Salatta, et al., 1964; Busch and Kalinske, 1956; Patterson,
et al., 1969; Stack, 1956; Hermann, 1959; Smith, 1953).
None of these tests are appropriate for monitoring the
activity of biomass in a treatment unit because of the
time required to conduct any of these tests and the slow
response of parameters to a reduction in the activity of
the biomass caused by a toxic slug introduced into the treat-
ment unit.
Biological waste treatment processes can be monitored through
selective biochemical tests. Cells may exist in either of
two metabolic states, the presence of utilizable substrate
or in the absence of substrate. The endogenous metabolic
level is indicative of the microbial population while the
metabolic activity in the presence of substrate represents
the oxidative activity of the biomass.
97
-------
The microbial concentration if deoxyribonucleic acid (DNA)
has been shown to be invariant with the physiological
state and fairly constant among bacterial species
(Genetelli, 1967). The concentration of DNA does not indi-
cate the activity of a biomass and thus cannot be used as
a measurement of toxic effects.
The hydrolysis of adenosine triphosphate (ATP), common to
all cellular metabolism, is the basis of a method of measur-
ing the physiological state of microbial cultures. The
endogenous ATP concentration represents a relative measure
of biomass while the rate of increase, or peak ATP concen-
tration after substrate addition, reflects the biomass
activity (Forrest, 1965 and Patterson, et al., 1969). Assay
procedures for ATP are based on its bioluminescent reaction
with luciferin and luciferase enzyme extracted from firefjy
lanterns. Briefly, ATP analysis requires rapid killing of
the cells and extraction of ATP into aqueous solution.
Many investigators utilize conventional liquid scintillation
counters for measurement of luminscence. Others have
developed their own light-measuring devices, and one company
(E. I. DuPont) is now marketing a special instrument for
ATP analysis. Concentration of ATP seems an attractive
parameter to study the response of cells to their environment.
Few attempts have been made to apply it to heterogenous
systems such as an activated sludge, and its application as
a continuous monitor does not seem practical.
The dehydrogenase activity test uses an electron acceptor
dye 2,3,5-Triphenyl tetrazolium chloride (TTC), which is
reduced to an insoluble red precipitate, formazan. The
drhydrogenase activity is reported as micrograms of formazan
produced during the incubation period per mg of cell material
(Lenhard, 1965). The results of this test are greatly
98
-------
affected by incubation time, temperature, mixing rate, and
the presence of oxygen (Patterson, et al., 19&9).
Measurement of dehydrogenase activity in sludges was used
to assess the oxidative capacity of activated sludge units
(Ford, et al., 1965). The measurement was found to respond
to significant changes in plant loading and sludge age,
although no correlation between BOD removal efficiency and
average dehydrogenase activity was found.
Many of the disadvantages of the dehydrogenase activity
method are obviated by the direct measurement of the oxygen
uptake, a measurement of the natural terminal electron
acceptor without interrupting the normal biochemical processes,
Traditionally, the Warburg respirometer has been used in
both biochemical and sanitary engineering research to
measure the oxygen uptake rate of microorganisms. The
Warburg technique requires special equipment and skilled
technicians and is inappropriate for routine plant monitor-
ing. Other manometric respirometers have been developed
for waste treatment application (Arthur, 1964; Tool, 1967).
The Arthur respirometer recycles the waste culture and air
counter-currently through a large, closed, reaction chamber.
The air-stream is passed through a sodium hydroxide solution
to absorb carbon dioxide. As oxygen is utilized by the cul-
ture, the partial pressure of oxygen decreases within the
aeration chamber, which is monitored by an oil manometer,
and a pressure transducer. The AC signal from the trans-
former is then rectified and recorded as the change in
partial pressure in the aeration chamber withtime, which
represents the oxygen demand curve.
99
-------
Continuous manometric respirometers have been developed for
BOD monitoring (Clark, 1961; Pipen, 1973). In these
devices, the oxygen uptake of a large, biological sample
is continuously replaced by a manometrically triggered
electrolysis reaction. The amount of electrical energy
required to generate the required oxygen is recorded with
time and is a measurement of the oxygen demand of the sample.
Since the development of membrane electrode systems to
measure dissolved oxygen, such devices have been used to
measure the oxygen uptake rate replacing the manometric
methods.
The Simcar Respirometer (Abson, 35 al., 196?) utilizes a
galvanic cell oxygen electrode to obtain a continuous record
of dissolved oxygen concentration in an aerated culture to
obtain the oxygen uptake of large samples over a period of
10 to 18 hours.
An automated respirometer (Wallace and Tierman, 1968)
measures the oxygen uptake in a biological sample with an
inlet probe and an outlet probe as the sample is pumped
from one to the other in a predetermined transit time. The
differential output of the two probes is representative of
the oxygen uptake rate of the sample flowing through the
device. A plot of the oxygen uptake rate versus time is
then produced and the area under the curve represents the
BOD of the sample tested. Results could be obtained within
four hours for each test.
Oxygen uptake rates, both Warburg and oxygen electrodes,
have been used to measure the effect of toxins and inhibitors
on activated sludge (Ayers, et al., 1965; Ghosh and Zugger,
1971; Sato, 1971; Hartmann and Laubenberger, 1968).
100
-------
The oxygen utilization rate shows more relative inhibition
than the ATP or the dehydrogenase activity measurements,
and it better approximates the actual activated sludge
process (Patterson, et al., 1969). The oxygen utilization
rate response rapidly to induced disruption of metabolic
activity. It is sensitive, reliable, and can be easily
automated with oxygen electrodes. A quantitative evaluation
of commercially available oxygen probes has been recently
published (Pijanowski, 1971).
OXYGEN UPTAKE MEASUREMENT CYCLE
Laboratory tests were conducted to determine an oxygen uptake
measurement cycle that would be sensitive to the presence
of a toxic substance in the feed to an activated sludge plant
and that could be automated to rapidly detect the toxic
effect on a repetitive basis.
A New Brunswick Scientific Chemostat Model C-30 was used
as a batch reactor in which the temperature, mixing speed,
and rate of aeration were controlled. It was also equipped
with an analyzer and recorder to follow the dissolved oxygen
(DO) level in the liquid phase in the reactor. A procedure
was followed to measure the oxygen uptake rate and total oxy-
gen utilized when a sample of a degradable carbon source was
added to an acclimated and stabilized bacteria sample ob-
tained from the activated sludge pilot plant:
About 500 ml of the activated sludge was introduced in
the reaction vessel and aerated to reach an equilibrium
DO level that represented the endogenous equilibrium
oxygen concentration. A 10-ml sample of a standard
propylene glycol solution or a test sample containing a
toxic substance was introduced in the reaction vessel
101
-------
and the airflow turned off at the same moment. The
initial slope of the decrease in DO concentration was
the oxygen uptake rate due to the oxidation of the
organic substrate plus the endogenous uptake rate.
If a low airflow was maintained as the sample was
introduced in the reactor, the oxygen concentration
followed a curve as shown in Figure 39. The area under
this curve was a measure of the oxygen consumed in the
oxidation of the organic substrate in the sample added.
The volume of the test sample of 10 ml had negligible dilu-
tion effect on the DO concentration in the reaction vessel,
and for a 1000-mg/l propylene glycol sample, complete degra-
dation time was less than 15 minutes.
The oxygen uptake rate and the total oxygen consumed by
propylene glycol standard solutions containing various con-
centrations of toxic substances were compared to the measure-
ment obtained in the absence of the toxin. The sensitivity
and reproducibility of the total oxygen consumed, as deter-
mined from the area under the oxygen uptake curve, was
superior to that of the oxygen uptake rate, as determined
from the initial slope of the oxygen concentration curve
when the aeration was cut off. The rate of oxygen uptake
was more dependent on the concentration of viable biomass
than the total oxygen consumed. The automation of a measure-
ment cycle utilizing the area under an oxygen uptake curve
was simpler since it did not require cutting of the air
every time a test sample was added.
For these reasons, the total oxygen consumed by a test
sample added to an aerated sludge under controlled conditions
was selected as the parameter to monitor toxic and inhibitory
effects of the feed to an activated sludge process.
102
-------
Figure 39.
Schematic of oxygen concentration
during the bio-oxidation of a
degradable carbon source
6 8
Time, minutes
10
12
103
-------
Mathematical Model of the Oxygen Uptake Curve (Figure 39)
When a sample of a degradable carbon source was added to a
stabilized and aerated sludge, the dissolved oxygen
decreased rapidly due to the oxygen consumed by the bacteria
to oxidize and metabolize the substrate added. The DO
continued to decrease as long as the oxygen uptake was
greater than the oxygen transferred to the solution by the
controlled aeration. The time required to reach the minimum
oxygen level depended upon the concentration and activity
of the biomass, the organic loading, the rate of aeration,
and the oxygen transfer characteristics. As most of the
substrate was degraded and the rate of oxygen uptake became
less than the oxygen supply, the DO started increasing till
it leveled off at the original endogenous respiration
equilibrium concentration.
The concentration of dissolved oxygen in the liquid phase,
C, at any time is given by:
|| = KLa (Cs - C) - rr (16)
where KLS = mass transfer rate of oxygen, min-1
Cs = saturation DO concentration, mg/1
rr = oxygen uptake rate by the microorganisms,
mg/l«min
Before the addition of the test sample containing the degrad-
able substrate, the oxygen uptake by the microorganisms is
the endogenous uptake and is proportional to the biomass
concentration, X.
T£ = KLa (Cs - C) - KeX (17)
dt
104
-------
At equilibrium -rr = 0 and therefore
U. 0
C = Cs -
where Ke = endogenous reaction rate constant, min"1
X = concentration of the biomass, mg/1
When a substrate is added to the biomass, the oxygen uptake
will be the sum of the endogenous uptake and the uptake due
to the substrate degradation. In a system that is not oxy-
gen limited, the substrate utilization can be represented
as a first order reaction by:
|| = -k'XS (19)
where k1 = substrate biodegradation rate constant,
min-1 (mg/1)-1
S = substrate concentration in oxygen equivalent
units, mg/1
Assuming the biomass concentration, X, to be constant during
the degradation of a small sample of a substrate, Equation
(19) can be integrated toi
s = S0 e-k'XS (20)
The DO concentration after the addition of the substrate is
given by:
|§ = KLa (Cs - C) - KeX - k'XS (21)
Substituting Equations (18) and (20) in (21) and rearranging!
105
-------
If + KLaC = KLaCo - k'XSo e~k'xt (22)
This differential equation is readily solved using an inte-
grating factor, exp. -/KLa^t and the initial condition of
Equation (18) to obtain the following expression for the DO
concentration as a function of time:
C = °o + k'X-KLa (e-k'Xt _ e-KLat) (23)
The area of the oxygen uptake curve of Figure 39 is given by:
Area = / (Co - C)dt
t=0
ce-k'Xt _ e-KLat )dt (24)
-J.-Q
Performing this integration yields
Area = °- (25)
The area of the oxygen uptake curve can be used to determine
the biological oxygen demand (So) of the test sample if the
oxygen mass transfer coefficient KLE is known. The value of
KLa depends on the rate of aeration and mixing and the
characteristics of the stabilized culture used in the test.
Development of the Measurement Cycle
In order to utilize the area under the oxygen uptake curve to
monitor the toxic effects of a feed to an activated sludge
plant, it is required to be able to distinguish between the
changes due to the varying concentrations of the substrate in
the feed and the presence of toxic substances. The oxygen
uptake of a feed sample could be calculated per unit total
106
-------
organic carbon in the sample to eliminate the change due to
varying substrate concentration, but this requires the use
of another instrument to monitor the total organic carbon in
the feed and a complicated, synchronized, electronic system.
A comparison of the oxygen uptakes of two standard propylene
glycol samples before and after the exposure of the bacteria
to a feed sample was a measure of the toxic effects of the
feed to the activated sludge process. The oxygen uptake
of the first standard established a reference activity
of the bacteria sample, which was then exposed to the
feed to the process for long enough time to degrade the sub-
strate in the feed. Then the same bacteria sample received
an equal volume of the standard and its oxygen uptake, com-
pared to that of the first sample, was a meausre of any change
of activity due to the feed sample.
The procedure developed for the automated biological inhibitor
detector was as followsi
1. About 500 ml of the mixed liquor was withdrawn from the
aeration basin to the reaction vessel of the instrument.
2. The bacteria sample was aerated at an adjusted rate of
mixing and aeration for 16 minutes to reach an equilibrium
endogenous respiration level.
3. A 10-ml sample of a 600 mg/1 propylene glycol solution
was introduced in the reaction vessel. During a period
of 12 minutes, the oxygen uptake curve was recorded and
its area, AI, was determined and stored.
4. A 10-ml sample of the feed was then introduced to the
reaction vessel and the DO level recorded over a period
107
-------
of 18 minutes, enough to degrade feed sample containing
substrates equivalent to about 1000 mg/1 of propylene
glycol.
5. Step 3 was repeated and the Area, A2, of the oxygen
uptake curve of this second standard solution was deter-
mined and stored.
6. The ratio of A2 to Ai was the activity ratio and was a
measure of the toxic effects of the feed sample.
7. At the end of the cycle, the bacteria sample was drained
from the reaction vessel and within two minutes a second
cycle started. The total cycle time was 60 minutes.
Figure 40 is a schematic illustration of the measurement cycle
when the feed contained no toxic substance. The activity ratio
would be approximately equal to 1.0.
Figure Ul illustrates the measurement cycle when the feed
contained a toxic substance. The activity ratio would be
much less than 1.0.
LABORATORY TOXICITY TESTS
The developed method for the measurement of an activity ratio
was used to study the effects of concentration and time of
exposure of various toxic substances on a propylene glycol-
acclimated fresh water bacteria. The purpose of the laboratory
batch testing was to determine the response of the activity
ratio measurement to the inhibitory effects of heavy metal
ions and toxic organic substances that may be present in a
petrochemical waste water treatment plant.
108
-------
o
c
I 3:
I >
o o
G
S-
-------
o -a
•i- OJ
4-> 4->
CU T-
•a 3
i- CU
O r-
•«-> O
••- >1
JO O
c: c
•r- CU
E
r- a;
(O S-
o n
•i— «/)
CD fO
o cu
r- E
O
••-
CD
cn
CD
00
CD
CD
UD
CD
LA
* uo
110
-------
The bacterial culture, obtained from the activated sludge
miniplant operating on a fresh water, propylene glycol
feed, had a mixed liquor volatile suspended solids (MLVSS)
of 1700 rr.g/1 to 2800 mg/1 during these tests. Direct
plating of the acclimated culture indicated a microorganism
composition of 77 percent Pseudomonas, 13 percent Bacillus
subtilis, and 7 percent Alcaligenes as the main species.
The laboratory procedure was similar to the procedure
developed for the monitoring instrument, in whifch the feed
sample was a synthetic 600 mg/1 propylene glycol containing
varying concentrations of the toxic substance. The oxygen
uptake of the second standard was measured at different
time intervals after the exposure of the bacteria sample
to the test solution to determine the accumulative toxic
effects or if the bacteria recovers with time.
The oxygen uptake, calculated per unit MLVSS, for some
selected inorganic toxins is plotted versus the concen-
tration of the toxin in the test sample in Figure ^2.
The activity ratios for the same toxins are given in Table 6
for one concentration at varying exposure time. Mercury was
the most toxic of the inorganic toxins tested, reducing the
oxygen uptake to less than half at 3 ml/1 and to nearly
zero at 16 mg/1 concentration. At an intermediate concen-
tration of 8 mg/1, mercury exhibited an immediate reduction
of the activity ratio to 0.27 that was further reduced with
longer exposure times. The same chronic toxicity effect
of the other inorganic toxins tested is illustrated in
Table 6.
The oxygen uptake for samples containing varying concentra-
tions of phenol and dichloroisopropyl ether is shown in
Figure ^3 and the effect of the exposure time on the activity
ratio of the two substances is given in Table 7. Phenol was
111
-------
Figure 42. Effects of inorganic toxins
on the oxygen uptake of
qlycol-acclimated fresh
water bacteria
X Cadmium
A Sodium cyanide
© Sodium hypochlorite
O Copper
• Mercury
0
10 20 30 40 50
Concentration of Toxin, mg/1
112
-------
Table 6. EXPOSURE TIME EFFECT OF INORGANIC TOXINS ON
GLYCOL-ACCLIMATED FRESH WATER BACTERIA
Toxin
Concentration
mg/1
Exposure
Time
minutes
Activity
Ratio
(A2/Ai)
Mercury
Copper
Sodium Cyanide
8
25
3.0
Sodium Hypochlorite
Cadmium
60
50
immediate
16
48
immediate
19
64
immediate
20
40
50
immediate
12
25
immediate
14
24
0.27
0.20
0.12?
0.31
0.206
0.204
0.607
0.502
0.353
0.316
0.294
0.187
0.102
0.70
0.337
0.276
113
-------
Figure 43.
Effects of organic toxins on the
oxygen uptake of glycol-acclimated
fresh water bacteria
• Dichloroisopropyl ether
0 Phenol
0
20 40 60 80 100
Concentration of Toxin, mg/1
-------
Table 7. EXPOSURE TIME EFFECT OF ORGANIC TOXINS ON
GLYCOL-ACCLIMATED FRESH WATER BACTERIA
Exposure Activity
Concentration Time Ratio
Toxin mg/1 minutes (A2/Ai)
Dichloroisopropyl
Ether 100.0 immediate 0.498
20 0.649
40 0.649
60 0.628
Phenol 111.5 immediate 0.394
13 0.438
27 0.417
39 0.410
115
-------
slightly more toxic to this culture than the ether, which
occurred in low concentrations in the waste water from the
production of propylene glycol. Also the recovery of the
bacteria, as indicated by an increase of the activity
ratio with time of exposure, was better for the ether than
for the phenol. The immediate reduction of activity could
have been due to a shock loading effect, followed by accli-
mated and, thus, recovery of the biomass.
DESIGN AND OPERATION OF THE AUTOMATED INSTRUMENT
The measurement cycle, developed and tested in the laboratory
on standard solutions containing various toxic substances,
was used as the basis for the design of the automated instru-
ment.
The compact table top model, as built, is shown in the
photographs of Figure 44 and Figure 45. The bacteria sample
entered from the bottom connection to the reaction vessel
and drained through the same connection. The reaction
vessel was placed on a variable speed magnetic stirrer and
an air sparger, sample lines, the oxygen probe, and a ther-
mometer were introduced through the stopper. The air supply
was controlled and measured by a rotameter while the exact
volumes of the standard solution and feed samples were intro-
duced through a calibrated, volumetric syphon system. A
tracing of the measurement cycle as operated by a 10-cam timer
is shown in Figure 46. A schematic of the electrical controls
for the instrument is shown in Figure 47. A block diagram,
Figure 48, traces the path of the measurement signal through
the various electronic components. The recorder may be
connected to any of six points indicated by the numbers in
parenthesis in Figure 48 and listed in Table 8, the legend.
116
-------
3
O
-a
CO
0)
S-
-o
£_
(O
(SI
o
o
S-
u
01
I
I
t.
o
-l->
o
0)
+->
OJ
s-
o
U
CT
O
O
•r~
OQ
QJ
S-
117
-------
CD
Q.
£
ro
S-
o
4->
O
to
Ol
r»
O
01
dJ
s_
o
4
nj
O
cr
O
o
CO
HI
C7>
u_
116
-------
D n
s-
CO
o
s.
o.
O)
u
I
I
I
S-
o
4->
u
0>
4J
o
TJ
LO
CD
LT*
CD
CT
CD
m
a
o
o
•f—
CO
CD
CD
vo
ai
s-
a>
-M
O
(U
O)
E -a
•r- S-
i— — c
r— rt
•r- -»->
u_ co
o
a;
u
0)
•o
CM CO
>
-a •—
i» fO
(O >
•a
c: a.
s-
OJ
X
o
E
cu
•a
S-
-M
C. to
> s-
X CO
CD a.
119
-------
Figure 47. Biological inhibitor detpctnr
schematic of electrical controls
CAM TIMER FOR
115 VOLTAGE AC
MC-1
1-2
h-
MC-2
MC-4
MC-3
HI—
JO.
MC-5
&-±
MC-6
MC-7
CAM TIMER FOR
+18 VOLTAGE DC
R-l
R-l
MC-10
MC-9
MC-8
©
O
LEGEND
Relay, P-B KRP-11AG
Timer, ATC 305, 0-15 seconds
MX A Mixer, magnetic, 600 rpm
MC Timer, 10 Cam, ITC MC-7
Solenoid valve, 115 V AC coil
INT) Integrator, Bell & Howell
19-407A
[S-H) Sample & Hold, Bell & Howell
20-419
120
-------
Figure 48. Biological inhibitor detector-
signal block diagram
Dissol ved
Oxygen
Analyzer
I
MV/V
Pre-Amplifier
tor
I
(2)
Integrator
(3)
Sample & Hold
3
Comparator
(6)
Relay
Sample & Hold
1
Sample & Hold
I
(4)
Coefficient
(0-1.0)
Legend in Table 8
121
-------
Table 8. LEGEND FOR SIGNAL BLOCK DIAGRAM OF FIGURE 48
Position of
Selector Switch Recorded Parameter
1 dissolved oxygen
2 endogenous equilibrium
dissolved oxygen
3 integration of oxygen uptake
curves of standard 1 and 2
*J area of oxygen uptake curve
of standard 1 (Ax)
5 area of oxygen uptake curve
of standard 2 (A2)
6 [(Ax)(coefficient) - A2]
comparator output
signal recorded is AI (coefficient)
122
-------
The operation of the automated cycle is as follows:
Cam 1 activated a 0- to 15-second timer which opened
a solenoid valve for a preset time interval to intro-
duce 500 ml of the bacteria into the reaction vessel.
Cam 6 switched on the mixer 15 seconds after filling
the reaction vessel. The bacterial sample was aerated
to an airflow rate of 36 ml/min and mixed at a rate of
200 to *JOO rpm for a period of 16 minutes. This was
enough time for the DO level to reach equilibrium at
80 to 90 percent of saturation, representing the con-
stant endogenous oxygen uptake in equilibrium with the
oxygen transfer from the air supply. The rate of mixing
and airflow were kept constant throughout the measurement
cycle. At this point, Cam 2 activated a three-way
solenoid valve allowing 10 ml of the standard propylene
glycol solution (600 mg/1) into the reaction vessel.
It also activated a relay to start the integrator and
caused sample and hold unit No. 1 to store the endogenous
DO value. Cam 9 activated sample and hold unit No. 2
to sample the integrator 11 minutes after the standard
sample was introduced. At the end of 12 minutes, Cam
3 opened a three-way solenoid valve on the feed sample
line allowing 10 ml of the waste vmter feed into the
reaction vessel. The feed was continuously flowing
through the solenoid valve into a stand pipe which
held 10 ml to the overflow level. After 18 minutes of
exposure time (enough to degrade the feed samples),
Cam 4 closed and repeated the operation of Cam 2 by
adding another sample of the standard propylene glycol
solution. Cam 10 activated sample and hold unit No. 3
to sample the integrator 11 minutes after the second
standard sample was introduced. At the end of 12 minutes
123
-------
Cam 5 opened the solenoid valve on the bacteria line
to dump the contents of the reaction vessel, while
Cam 6 turned off the mixer during the two-minute
period that the reaction vessel was empty so that the
mixer did not lose its magnetic coupling. The total
cycle time was 60 minutes and the cam timer ;\ras a
continuous mechanism by which the same cycle was
repeated every 60 minutes.
The integrated area of the oxygen uptake curve of the
first standard sample, which was stored in sample and
hold unit No. 2, was multiplied by a preset fraction,
usually 0.5, and compared to the integrated area due
to the second standard sample, which was stored in
sample and hold unit No. 3. The comparator operated
a relay that set off an alarm or diverted the feed to
a holding pond when the oxygen uptake of the second
standard was less than 0.5 times the oxygen uptake of
the first standard.
The instrument was designed for flexible operation and had
many features for easy trouble shooting and maintenance.
A recorder-selector switch located in front of the instru-
ment allowed recording of any of six different outputs
that are shown in Figures 48 and its legend (Table 8) . A
dump button on the front of the instrument actuated the
solenoid valve on the bacteria sample line to drain the con-
tents of the reaction vessel at any time for the purpose
of routine cleaning of the reaction vessel or resetting the
measurement cycle.
The electrical controls were quite versatile. The total
measurement cycle time was varied by simply changing a
gear in the 10-cam timer, and each cam switch was adjustable
for duration and phasing.
124
-------
The electronic components were also very versatile since
the gain of each component was adjustable.
INSTRUMENT TESTING ON MINIPLANT OPERATION
The automated biological inhibitor detector was installed
at the activated sludge miniplant to test its operation
both under normal conditions and withtoxic substances
addded to the feed to the plant.
The bacteria sample to the instrument was obtained from
the aeration basin mixed liquor or the sludge recycle
stream from the biosettler. The higher concentration of
the biomass in the sludge recycle gave a higher initial
slope of oxygen uptake, but the time required for recording
the oxygen uptake curve was not significantly different
from that obtained with the lower concentration of biomass
in the aeration basin liquor. The sludge recycle sample
was deficient in DO and required a much longer time of
aeration to reach the equilibrium endogenous level, as com-
pared with the aeration basin liquor. The biomass concen-
tration in the aeration basin did not vary as much as that
of the sludge recycle, and more reproducible oxygen uptake
data were obtained with bacteria samples from the aeration
basin.
Continuous operation of the instrument to monitor the oxygen
uptake of fresh water bacteria under normal operating condi-
tions for a period of three months showed a measurement pre-
cision of ±5 percent. The area of the oxygen uptake curve
as integrated by the instrument was linear for the range
of 200 to 1000 mg/1 standard propylene glycol solutions. For
the same concentration of viable biomass in the bacteria
sample, the sensitivity of the oxygen uptake signal was
dependent on the rate of mixing, airflow rate, and other
125
-------
factors that affect the overall oxygen transfer coefficient
KL&. Sensitivity was varied by controlling the airflow
rate and. adjusting the gain of the electronic parts of
the instrument.
The instrument was tested for the detection of toxic sub-
stances in the feed while it was continuously monitoring
the activity ratio (A^/Ai) of the activated sludge miniplant.
Batch tests were conducted by adding varying concentrations
of the toxic substance to the feed sample entering the re-
action vessel of the instrument. Continuous feeding of a
selected toxin in the feed was the basis of dynamic tests
to evaluate the effects of accumulation of the toxin in the
bacteria.
Batch Toxicity Tests
The same inorganic and organic toxic substances tested in
the laboratory were used for testing the instrument while the
miniplant was operated on fresh water-glycol feed and the
high salt waste water from the glycol production plant.
The activity ratio, as recorded by the instrument, is plotted
as a function of toxin concentration in Figures 49 through
Figure 52. The effects of the inorganic toxins on the fresh
water bacteria, Figure 49, were very similar to the resutls
obtained in the laboratory tests, Figure 42, in which the
order of toxicity was Hg > CN > CIO > Cu > Cd. In the case of
the glycol-acclimated salt bacteria, Figure 50, the order
of toxicity was CN> Cu>Hg> Cd. The less toxic effect of
copper and its leveling off at concentrations higher than 10
mg/1 in case of the fresh water bacteria may have been due
tp the multiplicity of the species as compared to the salt
water acclimated bacteria, where one major species predominated,
126
-------
Figure 49. Effects of inorganic toxins on
activity of alvcol-acclimated
fresh water bacteria
/\ Cadmium
X Copper
D Sodium
Hypocnlorite
O Sodium Cyanide
Mercury
0
Cd
10 20 30 40 50
Concentration of Toxin, mg/1
60
127
-------
Figure 50. Effects of inorganic toxins on
activity of qlycol-acclimated
salt water bacteria
Cadmium
O Mercury
X Copper
O Sodium Cyanide
0
10 20 30
Concentration of Toxin, mg/1
60
128
-------
Both phenol and dlchloroisopropyl ether were quite toxic to
the fresh water bacteria, Figure 51, while the ether had
negligible effect on the salt water bacteria, Figure 52.
This was clearly due to the presence of the ether in the
waste water feed to the salt water bacteria, which was
acclimated to concentrations of less than 100 mg/1 of the
ether.
Continuous Toxicity Tests
Dynamic toxicity tests were conducted by adding copper to
the glycol waste water feed in the activated sludge mini-
plant. The copper was added as cupric chloride to the
equalization tank and fed continuously to the salt water
bacteria. The area of oxygen uptake curve for a standard
propylene glycol solution (Ai) was determined before the
copper addition. The same area (A2) was recorded by the ins-
instrument once an hour after the addition of copper for
a period of 24 hours. The activity ratio (A2/Ai) for the
test runs with 6.6 mg/1 copper and 3*0 mg/1 copper are
shown in Figure 53 and Figure 5^, respectively. The data
scattering of Figure 53 was due to the malfunction of the
stirring mechanism of the instrument; that was corrected
during the second test.
Both tests showed no, or slight, reduction of activity for
a period of six to nine hours followed by a sharp decline
of the activity and reached a nil value after 16 hours and
21 hours for the 6.6 mg/1 copper and 3.0 mg/1 copper,
respectively. A material balance of the copper during
test B is shown in Table 9, and the concentration of copper
in the bacteria and the liquid of the aeration basin are
plotted as a function of time in Figure 55. During the 24
hours of feeding 3 mg/1 of copper to the activated sludge
129
-------
Figure 51.
Effects of organic toxins on
activity of glycol-acclimated
fresh water bacteria
1,0
0,9-
0,8-
- 0,7-
to
GC.
0,6
O
0,5-
0,4-
0.3
Q-Phenol
n-Dichloroisopropyl
^ Ether
I
I
0 10 20 30 40 50 60 70 80
Concentration of Toxin, rng/1
130
-------
Figure 52.
Effects of organic toxins on
activity of glycol-acclimated
salt water bacteria
ODichloroisopropyl Ether
X Phenol
1,0
0,9
>
+J
£ 0,7
o
«=c
0,6
I
I
I
I
I
I
I
0 10 20 30 40 50 60
Concentration of Toxin, mg/1
70 80
131
-------
O i.
(U
c o
•i- «3
•M
« cj
ra
I
>>•—
•4-> O
•r- O
a >>
X CT.
O
4-> O
E >o
Q O
un
(U
s_
3
a>
CXI
oo
in
i.
o>
E
u:
CXI
I
I I
I
oo
CD
CXI
CD
CD
CD
132
-------
Figure 54.
Dynamic toxldty test B,
continuous feed of copper
to glycol -accl imated
salt water bacteria
(copper concentration in feed
bacteria concentration in feed
3.0 mg/1
1855 mg/1
MLVSS)
1,0
0,8
o" 0,6
-M
(O
•M
O
-------
Table 9. DYNAMIC TOXICITY TEST B —
MATERIAL BALANCE ON COPPER
(BASIS 24 HOURS)
Copper ,
mg
Copper In 10,044
Copper Out
Overflow, liquid 2,253
Overflow, bacteria 956
Wasted, Bacteria 1,043
Total Out 4,252
In-Out 5,792
Copper
Accumulation
In Liquid
In Bacteria
Total
Accumulation
Copper.
mg
848
5*048
5,896
134
-------
to
£ 4000
O
CO
Figure 55.
Dynamic toxicity test B,
copper accumulation in
aeration basin mixed liquor
0 3000
o>
• 2000
Copper in Bacteria
03
s 1000
c
O
-------
miniplant, about 5800 mg of copper were accumulated In
the bacteria, resulting in an increase of copper concen-
tration from 656 to more than 4000 mg of copper per Kg
of dry bacteria. The copper uptake rate by the bacteria
was very rapid during the first eight hours, reaching a
level of 3000 mg per Kg of dry bacteria. A plot of the
rate of copper uptake versus the amount of copper adsorbed
by the bacteria (Figure 56) indicates a second order
reaction. This shows that the reaction with copper was
controlled by the amount of respiratory enzyme available,
which was large at the start of the test, showing no
reduction in activity, then, at a point where most of the
enzyme sites were complexed with copper, sharply decreased
in activity.
136
-------
Figure 56.
Reaction order kinetics of copper
adsorption on glycol-salt bacteria
700
600
oi
4J
«J
10
500
-a
CD
£400
c:
o
£-300
£.200
Q.
O
c_>
« 100
-------
SECTION IX
EFFLUENT QUALITY MONITORING AND INSTRUMENTS EVALUATION
An Ionics Model 225 Total Oxygen Demand (TOD) Analyzer and
a multiple stream selector were used to monitor the feed
to the activated sludge miniplant and the effluent before
and after chemical flocculation.
A schematic of the monitoring system is shown in Figure 57-
The five-stream selector was equipped with a four-stream
manifold for the three streams to be monitored and for a
standard solution used to calibrate the instrument response
once a day. The feed stream was diluated 1:3 with distilled
water using a cassette peristaltic pump. Gravity flow of the
effluent samples to the instrument resulted in plugging the
sample valves. Filters were added in the sample lines and
pumping at a rate of 20 ml/min was necessary to maintain
the sample flow to the instrument. The instrument was
calibrated daily with a 700 mg/1 TOD standard solution and
the gain adjusted for a 70-percent full-scale response.
The response that included the diluted feed and the effluent
samples concentrations was linear in the range of 100 to
700 mg/1. The linear range of the calibration curve was
checked once a week.
The Total Oxygen Demand Analyzer and the Total Carbon
Analyzer, used for the control of nutrients addition and
F/M control, were quantitatively evaluated for their per-
formance during the periods of monitoring both fresh water
and salt water streams (8 to 10 percent NaCl).
138
-------
5-
O
O
E
CO
3
O
S-
o
1— •»->
(T3 ro
O i—
O
O
OJ
JC
u
LT>
(U
s-
ISO
O) U
J= O
O •—
CM CO
139
-------
PERFORMANCE OP THE TOTAL OXYGEN DEMAND ANALYZER
The instrument was operated in the laboratory for several
months before being installed in the miniplant. The
precision of measurement of standard solutions was ±4 per-
cent in the linear range of 0 to 500 mg/1 TOD. The rotary
sampling valve required frequent adjustments to prevent
leakage around the injection port and the drive gears for
the motor turning the valve failed early during the second
month of operation. The gears were replaced under warranty
by the manufacturer at a cost of four hours labor and four
days downtime.
The instrument and the multiple stream selector were installed
in the miniplant and operated on fresh water samples for a
period of four months and on salt water samples for one month.
During the second week of operation, the printed circuit
card in the stream selector, containing a stepping switch,
failed and was replaced under warranty, resulting in three
days downtime and four hours labor. This was followed by
several problems with the instrument, a failure of a control
card (two days of downtime), a failure of a power relay in
the combustion heater circuit (two days downtime), and a
failure of the balancing motor of the recorder (four days
downtime). Numerous, short-duration downtimes occurred
for cleaning the rotary sample valve and adjusting its 0-ring
injection seal. The oxygen sensor required adjustment of
electrolyte level (that affects sensitivity) as frequently
as twice a week and was disassembled for electrode cleaning
once a month.
A precision chart and a record of the time out of control,
downtime, and hours of maintenance labor were kept during
the continuous operation of the instrument. A sample of
the precision control chart is shown in Figure 58 and a summary
140
-------
Figure 58.
Precision control
TOD analysis
chart
Range - 45-65% Scale
* = 0.05 0 = 0.05
Full Scale = 1034 mg/1 TOD
S = 1.90% Scale
CM
-o
-50
23456
Day of Duplicate Sample, M
-------
of the performance of the instrument is given in Table 10
for the fresh water and salt water sampling. The last month
of operation on salt water (8 to 10 percent NaCl) samples
proved to be very difficult. The combustion tube was coated
with a salt glaze and carbon precipitated in the lines
between the tube and the oxygen sensor, plugging the sample
passage in less than four days of operation. The combustion
tube was then slowly cooled to prevent its rupture and
replaced every four days.
PERFORMANCE OF THE TOTAL CARBON ANALYZER
The instrument was operated in the laboratory for a period
of three months, without incident. The precision of measure-
ment of standard solutions was ±1 percent of the range of 0
to 300 mg/1 organic carbon.
Operated on the miniplant feed stream was once interrupted
by the failure of a heater element around the combustion
tube that was replaced under warranty. For the purpose of
monitoring the total carbon of a homogenized sample from the
aeration basin, several controls were added to the analyzer.
The added controls included a sampling switching circuit, a
memory control, and a signal retransmission circuit. Oper-
ation of the analyzer on the homogenized bacterial samples
caused frequent plugging of the injection valve until the
sample handling system was improved to include a backwashing
with the continuous feed sample. Replacement of the injec-
tion tube below the slide valve with one of larger bore
resulted in significant loss of sensitivity.
A sample of the precision control chart is shown in Figure 59,
and a summary of the instrument performance is given in
Table 11 for the fresh water and salt water operation.
During the four months operation on fresh water sampling one
142
-------
Table 10. PERFORMANCE OF THE TOTAL OXYGEN DEMAND ANALYZER
(Ionics Model 225-Serial No. 357)
Fresh Water
4-mon. Per.
Salt Water
1-mon. Per,
Time in Control, %
Time Out of Control, %
Down Time, %
Maintenance Labor, hrs
62.5
35.4
4.1
95
52.6
41.6
5.8
20
143
-------
Figure 59. Precision control chart
TC analysis
Range - 44-55% Scale Full Scale = 690 mg/1 TC
<* = 0.05 6 = 0.05 Sd = ± 3.67% Scale
CM
"O
1X1
- 50-
-100-
0 1
23456789 10 11
Day of Duplicate Sample, M
144
-------
Table 11. PERFORMANCE OF THE TOTAL CARBON ANALYZER
(Ionics Model 1212-Serial No. 1196-12)
Fresh Water
4-mon. Per.
Salt Water
1-mon. Per,
Time In Control, %
Time Out of Control, %
Down Time, %
Maintenance Labor, hrs.
82.2
14.3
3.5
60
97.5
2.0
0.5
145
-------
Incident resulted in most of the downtime and 50 hours of
maintenance labor. The loss of sensitivity and the mal-
functioning of the automatic zero electronic unit attached
to the infrared detector resulted in a need to overhaul
many parts of the instrument, including the infrared unit
adjustment to restore its optical balance. Satisfactory
operation was restored only after replacement of the com-
bustion tube. Apparently, the catalyst inside the tube had
fallen below the high temperature region and was the main
cause of the erratic results.
Operation on the high salt (8 to 10 percent NaCl) sampled
at 18-minute intervals was excellent for a period of three
weeks followed by a slow loss of sensitivity to the end of
the one-month operation. This was due to accumulation of
salt within the combustion tube.
The slide injection valve was smoothed by grinding twice
during the one-year operation of the instrument. Electrical
maintenance was limited to replacement of one wire in the
infrared unit, while the recorder and associated signal
conditioning circuitry required no maintenance during the
entire operating period.
Comparing the performance of the two analyzers, it was con-
cluded that;
1. The slide injection valve of the TCA was more reproducible
and required less maintenance to prevent leaking than
the rotary valve of the TOD analyzer.
2. The infrared unit of the TCA was far more accurate and
precise than the oxygen sensor of the TOD analyzer.
146
-------
3- The Hastelloy combustion tube of the TCA was not sub-
ject to thermal fracture as was the case with the
quartz combustion tube of the TOD analyzer.
4. Samples containing salt were difficult for both analyzers
to handle. Somewhat better service was expected with
the TCA.
147
-------
SECTION X
PROCESS DESIGN CONSIDERATIONS
The development of on-line control systems for the activated
sludge treatment process opens up many possibilities for
flexible and innovative process design.
The two most important automatic control systems that
could be incorporated in new plant designs are the biological
inhibition detection (BID) system and the F/M control system.
This is especially true for the design of facilities for the
treatment of combined domestic and industrial wastes, where
variations in the loading and the chemical composition of the
waste stream may be very large. A typical domestic waste
may have daily variations in BOD from 10 to 250 percent of
the average value and flow rates may vary from 50 to 150
percent (Andrews, 1971). When combined with the possibility
of slugs (or continuous flow) of industrial wastes where a
toxin to the activated sludge biota may be present, the need
for an inhibition detection system and an F/M control system
becomes readily apparent.
Figure 60 illustrates a conceptual design of an activated
sludge treatment system using these controls,
Equalization of the feed is very helpful for F/M control
and is a necessary part of the plant design. Equalization
dampens fluctuations in the organic content, in flow, in pH
and in order physical-chemical characteristics of the feed
to the aerated basin.
As illustrated in Figure 60, the equalization tank is also
part of the feed-forward inhibition detection control system.
Equalization provides a time delay, together with dilution
of a possibly toxic feed during the biological activity
analysis period.
-------
c
O)
E
•M
o
s-
-M
C
O
o
QJ
C7>
-------
A toxic feed would ordinarily be diverted to a holding basin
by the BID control system. After determination of the chem-
ical nature of the toxin it could either be bled back into
the system or treated by other means. Pre-treatment could
include chelation of toxic metals and sorption of toxic
organics.
The F/M control system would include measurement of
flow and measurement of soluble organic concentration. The
control system includes an aerated waste stabilization tank
which supplies additional microorganism recycle in periods
of high loading. The stabilization tank may be located as
shown in Figure 60 or it may be located in the sludge re-
cycle line. A pilot study would be useful in determining
which site is more adequate in terms of microorganism viability
and the effect on bio-flocculation in the aeration basin.
For equal-sized stabilization tanks, a tank in the recycle
line would give the shorter hydraulic residence time, which
would indicate a higher biota viability. Having the tank
located in the waste line would produce a more stabilized
sludge, possibly aiding in bio-flocculation.
If the post-treatment step of chemical flocculation is used,
another alternative in F/M control is possible. If the sludge
from the floe-settler is viable, it may be used as the source
of additional recycle in periods of high loadings. It has
also been shown that both the treatment efficiency and the
compactability of bulking sludge are improved significantly
with the addition of aluminum hydroxide to the mixed liquor
(Hsu and Pipes, 1973).
150
-------
Two possible sources of the feed sample to be used for
F/M control are shown in Figure 60. The location used
for the experimental work in this study was in the feed
line to the aeration basin (see Section V). Sampling from
this location allowed easy proportional control.
If the feed sample for F/M control is taken before the
equalization tank, F/M control will be more anticipatory
in nature with very short F/M response times. The control
system could then include a mini-computer which would be
programmed to calculate the hydraulic effect that the equal-
ization tank will have on the feed flow and composition and
would affect F/M control on this anticipatory basis. Hence,
extra sludge recycle could begin before the higher loadings
affect the aerated biomass.
Another available control system that needs to be incorporated.
into the plant design is an aeration-basin oxygenation sys-
tem. Besides optimizing oxygen allocations, the system should
assure that proper agitation and micro-turbulence occurs so
that the various mass transfer processes can be optimized
(Kalinske, 1971). The supply of oxygen to the stabilization
tank would also be controlled.
Nutrients and pH control would be provided for those processes
requiring them. Tertiary treatment is necessary for
colloidal and suspended solids removal from the effluent.
Turbidity control would be provided for designs that employ
flocculation as the tertiary treatment.
OPTIMUM DESIGN OF AN AUTOMATED ACTIVATED SLUDGE PLANT
The design of the aeration basin volume needed for a specified
treatment efficiency is greatly affected by the incorporation
of the F/M control system.
151
-------
The choice of a steady-state model and the determination of
the kinetic coefficients and other parameters pertinent to
the chosen model are necessary in determining the aeration
basin volume. A continuous pilot system is recommended for
determining this data.
The steady-state model may then be used to calculate the
required volume of the aeration basin as a function of other
design variables. For examples a computer program using
the Monod model steady-state calculations (completely mixed
reactor) is presented in Appendix D and some of the design
relationships for the fresh water, propylene glycol system
are illustrated in Figure 61. The sludge compaction data
(SVI) and the assumed recycle sudge concentration (Xr) are
obtained from the pilot study. The treatment must be at an
F/M within acceptable SVI limits. The aeration basin volume
(the hydraulic residence time multiplied by the influent flow
rate) is then a function of the recycle ratio.
The steady-state program also predicts the concentration of
microorganisms in the aeration basin and the cell production
rate, which at steady state is set equal to the wastage rate.
The aerated basin design should include evaluation of a
transient-state model of the system to determine the effects
that the suspected operational variations in loading will
have on the process. The transient model permits optimum
design of the F/M control system, both in terms of control
action and stabilization tank volume. System response times
can be determined to evaluate the adequacy of a given design.
The design should incorporate mixing and dispersion parameters
obtained from actual equipment in order to insure that the
residence time distribution in the constructed system will be
152
-------
a6pn[s
C
o
fO
OJ
i.
01
to
0)
•o
0)
•4->
(O
-)->
to
I
>•
T3
(O
CD
•t->
OO
O)
S-
3
O1
CD
CD
I^O
1
CD
LO
CVJ
1
CD
CD
CNJ
1
CD
tn
i— i
1
CD
CD
«— 1
1
CD
I
Ts
001
CD
CD
CD
CSJ
CD
CD
tn
CD
CD
CD
CD
CD
LO
s/Cep
Bouapisay
153
-------
the same as was assumed for the designed system (Irvine,
et al., 1973). This may entail further transient-state
modeling to consider the possibility of a combination of
two or more biological reactors.
The final design of the aeration basin volume is dependent
on an economic evaluation of the alternative systems. Items
to be considered include the reactor capital cost, recycle
pumping costs, aeration and mixing costs, and control systems
costs. The F/M ratio control costs include the stabilization
tank capital and the cost of aeration. Benefits of stabiliza-
tion, such as decreased sludge handling costs because of
sludge digestion, must be taken into account. The benefits
of increased process removals must also be weighed.
It is clear that the availability of developed control systems
and of sufficiently accurate predictive modules will enable
more efficient, more economic, and more controllable activated
sludge systems to be built and to be operated at more closely
regulated effluent limits. Sophisticated plants will require
more sophisticated operation.
Future plant designs will have to incorporate the best avail-
able technology. Future plant operation will have to reflect
the increased level of technology. Both design and operation
depend upon the availability of sensors and the instrumenta-
tion necessary to use the sensed data to control the processes,
-------
SECTION XI
REFERENCES
Abson, J. W., C. D. Furness, and C. Howe. 1967. Development
of the Simcar Respirometer and Its Application to Waste Treat-
ment. J. Inst. Water Poll. Cont. (British) 6:607-621.
Andrews, John F. 1971. Kinetic Models of Biological Waste
Treatment Processes. Biotechnol. and Bioenp;., 2:5-33•
American Public Health Association. 1971. Standard Methods
for the Examination of Water and Waste Water, 13th Ed. APHA,
New York.
Arthur, R. M. 1964. An Automated BOD Respirometer. Purdue
Univ. Eng. Exten. Series, Eng. Bull. 117:628-637.
Ayers, K. C. , K. S. Shumate, and G. P. Hanna. 1965. Toxicity
of Copper to Activated Sludge. In: Proc. 20th Indust. Waste
Conf., Purdue Univ., Lafayette, Indiana.
Banerji, S. K., B. D. Bracken, and B. M. Garg. 1968. Effect
of Boron on Aerobic Biological Waste Treatment. Purdue Univ.
Eng. Exten. Series, Eng. Bull. 132:956-965.
Earth, E. P., M. B. Ettinger, B. V. Salotto, and G. N. McDermott
1965. Summary Report on the Effects of Heavy Metals on the Bio-
logical Treatment Processes. JWPCF 37:86-96.
Benedek, P. and I. Horvath. 1967. A Practical Approach to
Activated Sludge Kinetics. Water Res. 1:10.
Bisogni, James J. and Alonzo VI. Lawrence. 1971. Relationships
Between Biological Solids Retention Time and Settling Character-
istics of Activated Sludge. Water Research 5:753-763.
Black, A. P., G. P. Singley, G. P. Whittle, and J. S. Maulding
1963. Stoichiometry of the Coagulation of Color-Causing Organic
Compounds with Ferric Sulfate. JAWWA 12:13^7-1366,
Busch, A. W., and A. A. Kalinske. 1956. The Utilization of the
Kinetics of Activated Sludge in Process and Equipment Design.
In: Biological Treatment of Sewage and Industrial Wastes.
Reinhold Publishing Company, New York.
Chen, Gilbert K., L. T. Fan, and Larry E. Erickson. 1972.
Computer Software in Waste Water Treatment Plant Design.
JWPCF 44(5) :746-762.
155
-------
Chiu, S. Y., L. T. Fan, I. C. Kao, and L. E. Erickson. 1972.
Kinetic Behavior or Mixed Populations of Activated Sludge.
Biotechn. and Bioeng. 14:179-199.
Clark, J. W. 1961. BOD for Plant Operation. Water and
Sewage Works 108(2): 61.
Cleland, W. W. 1963. The Kinetics of Enzyme-Catalyzed
Reactions with Two or More Substrates or Products. In:
Inhibition: Nomenclature and Theory. Biochem. Biophys.
Acta 67:173-187.
DeVillaret, Foulques and David M. Himmelblau. 1973. Kinetic
Modeling of Aeration Basins. JWPCF 45. (2):292-302.
Eckenfelder, W. W., and D. L. Ford. 1970. Water Pollution
Control. Jenkins Publishing Company, New York.
Erickson, Larry E., and L. T. Fan. 1971. Optimization of the
Hydraulic Regime of Activated Sludge Systems. JWPCF 40(3):
345-362.
Fan, L. T., L. E. Erickson, P. S. Shah, and B. I. Tsai. 1970.
Effect of Mixing on the Washout and Steady-State Performance
of Continuous Cultures. Biotech, and Bioeng. 12:1019-1068.
Ford, D. L., J. T. Yong, and W. W. Eckenfelder. 1966. De-
hydrogenase Enzyme as a Parameter of Activated Sludge Activities
Purdue Univ. Eng. Exten. Series., Eng. Bull. 121:53^-5^3.
Forrest. W. W. 1965. Adenosine Triphosphate Pool During the
Growth Cycle in Streptococcus faecalis. J. Bacteriol. 90:1013-
18.
Forster, C. F., and N. M. Choudhry. 1972. Physico-Chemical
Studies on Activated Sludge Bioflocculation. Effl. and Water
Tmnt. J. 3:127-131.
Gaudy, J., M. Ramanathan, and B. S. Rao. 1967- Kinetic
Behavior of Heterogeneous Populations in Completely Mixed
Reactors. Biotechn. and Biceng. 9:387-411.
Genetelli, E. J. 1967. DMA ana Nitrogen Relationships in
Bulking Activated Sludge. JWPCF 39:R37.
Gill, S. 1951. A Process for the Step-by-Step Integration
of Differential Equations in an Automatic Digital Computing
Machine. Proc. Cambridge Phil. Soc. 47:96.
Ghosh, M. M., and P. D. Zugger. 1973. Toxicity Effects of
Mercury on the Activated Sludge Process. JWPCF 45:425-433.
156
-------
Grieves, Robert B., William Milburg, J. K. Pipes, and 0.
Wesley. 1964. A Mixing Model for Activated Sludge. JWPCP
36:5.
Hartmann, L., and G. Laubenberger. 1968. Toxicity Measure-
ments in Activated Sludge. J. ASCHE San Div. 94:247-256.
Hattingh, W. H. J. 1963. Influence of Nutrition on the
Respiratory Rate of the Mocroorganisms. Water and Waste
Tmnt. 9(9) :424-426.
Hattingh, W. H. J. 1963. The Nitrogen and Phosphorus
Requirements of the Microorganisms. Water and Waste Tmnt.
8(8):380-386.
Herman, E. R. 1959. A toxicity Index for Industrial Wastes.
Ind. and Eng. Chem. 51:84A-87A.
Hsu, Deh Y., J. K. Pipes, and 0. Wesley. 1973. Aluminum
Hydroxide Effects on Waste Water Treatment Processes. JWPCF
45(4):68l-697.
Irvine, Robert L., Robert T. Keegan, William D. Langley,
and Ronald C. Catchings. 1973. Specific Removal Patterns
in Activated Sludge System Design. JWPCF 45(8):1771-1782.
Jones, P. H., and D. Prasad. 1969. The Use of Tetrazolium
Salts as a Measure of Sludge Activity. JWPCF 4l:R44l-449.
Kalinske, A. A. 1971. Effect of Dissolved Oxygen and Sub-
strate Concentration on the Uptake Rate of Microbial Suspen-
sions. JWPCF 43(1):73-80.
Kornegay, B. H., and J. F. Andrews. 1970. Characteristics
and Kinetics of Biologically Fixed Film Reactors. Env. Syst.
Eng. Dept., Clemson Univ., Clemson, S.C.
Lenhard, G. 1965. Dehydrogenase Activity as Criterion for
the Determination of Toxic Effects on Biological Purification
Systems. Hydrobiologia 25:1-8.
McLellan, J. C., and A. W. Busch. 1969- Proc. 24th Ind.
Waste Conf., Purdue Univ., Lafayette, Indiana.
Monod, J. 1949. The Growth of Bacterial Cultures. Annual
Review of Microbiology, Vol. III.
Mulbarger, M. C., and J. A. Castelli. 1966. A Versatile
Activated Sludge Pilot Plant Its Design, Construction and
Operation. Purdue Univ., Lafayette, Indiana. Exten. Series
No. 121:322-337.
157
-------
Ottengraf, S. P. P., and K. Rietema. 1969. The Influence of
Mixing on the Activated Sludge Process in Industrial Aeration
Basins. JWPCF 41(8):R282-293.
Patterson, J. W., P. L. Brezonik, and H. D. Putnam. 1969.
Sludge Activity Parameters and Their Application to Toxicity
Measurements in Activated Sludge. In: Proc. 24th Ind. Waste
Conf., Purdue Univ., Lafayette, Indiana.
Pippen, D. L., and G. R. Kramer. 1973. Metal Toxicity to
Sewage Organisms. JASCE San Div. 97:161-169.
Ramanathan, M., and A. P. Gaudy, Jr. 1969. Effect of High
Substrate Concentration and Cell Feedback on Kinetic Behavior
of Heterogeneous Populations in Completely Mixed Systems.
Biotech, and Bioeng. 11:207-237.
Ramanathan, M., and A. F. Gaudy, Jr. 1971. Steady-State Model
for Activated Sludge with Constant Recycle Sludge Concentration.
Biotech, and Bioeng. 13:125-145.
Salotto, B. V., E. F. Earth, W. E. Tolliner, and M. B. Ettinger.
1964. Organic Load and the Toxicity of Copper to the Activated
Sludge Process. Purdue Univ. Eng. Exten. Series, Eng. Bull.
117:1025-1034.
Sato, T. 1971. The Toxicity of Metallic Ionics on the Activated
Sludge and the Detoxication Effect of EDTA. Gifu Yakka Daigau
Kiyo 20:1-8.
Sawyer, C. N., and M. S. Nicholes. 1939. Activated Sludge
Oxidations. I. Effect of Sludge Concentration and Temperature
upon Oxygen Utilization. Sew. Works. J. 11:51-59.
Schaezler, D. J., W. H. McHarg, and A. W. Busch. 1971. Effect
of the Growth Rate on the Transient Responses of Batch and
Continuous Microbial Cultures. Biotech, and Bioeng. 2:107-129.
Sherrard, Joseph H., and Edward D. Schroeder. 1972. Importance
of Cell Growth Rate and Stoichiometry to the Removal of Phosphorus
from the Activated Sludge Process. Water Res. 6:1051-1057.
Smith, D. B. 1953. Measurements of the Respiratory Activity of
Activated Sludge. Sew. and Ind. Wastes 25:767-778.
Smith, R., and R. G. Eilers. 1969. A Generalized Computer Model
for Steady-State Performance of the Activated Sludge Process.
Federal Water Quality Administration, U.S. Dept. of Interior,
Washington, DC.
158
-------
Stack, V. T., Jr. 1957. Toxicity of Alpha, Beta-Unsaturated
Carbonyl Compounds to Microorganisms. Ind. and Eng. Chem.
49:913-917.
Storer, F. F., and A. F. Gaudy, Jr. 1969- Computational
Analysis of Transient Response to Quantitative Shock Loadings
of Heterogeneous Populations in Continuous Culture. Eng. Sci.
and Tech. 3(2) :143-149.
Stumm, Werner, and James J. Morgan. 1962. Chemical Aspects of
Coagulation. JAWWA 5*1:971.
Stumm, Werner, and Charles R. D'Melia. 1968. Stoichiometry
of Coagulation. JAWWA 60:514.
Tenney, Mark W., and Werner Stumm. 1965. Chemical Flocculation
of Microorganisms in Biological Waste Treatment. JWPCF 10:1370-
1387.
Tool, H. R. 1967. Manometric Measurement of the Biochemical
Oxygen Demand. Water Sew. Works J. 114:211-218.
Tracy, Kenneth D., and Thomas Keinath. 1973. Dynamic Model
for Thickening of Activated Sludge. Presented at the 74th
National Meeting AIChE in New Orleans, Lousiana, March 11-15.
Wellens, V. H., and R. Zahn. 1971. Untersuchungen iiber die
Toxizitatsbestimmung van Abwassern und Absasserinhaltstoffen
nach der Dehydrogenasen-aktivitat (TTC-method) Chemiker-
Zeitung 95:472-478.
Zeitoun, M. A.,^W. F. Mcllhenny, C. A. Roorda, J. C. Williams,
R. W. Murray, and W. D. Spears. 1971. Treatment of Waste
Water from the Production of Polyhydric Organics. U.S. Environ-
mental Protection Agency, Water Quality Office, Water Pollution
Cont. Res. Series 12020.
159
-------
SECTION XII
LIST OF INVENTIONS
The following inventions are disclosed herein to the United
States Government.
1. Biological Inhibitor Detector
The invention concerns the design and automation of a measure-
ment cycle to detect the presence of toxic or inhibitory mater-
ials in waste water fed to a biological process. The measure-
ment cycle is based on measuring the dissolved oxygen in a
controlled biological reactor as a standard sample, a waste
water feed sample, and another standard sample are consecu-
tively added to the reactor. A toxicity index or activity
ratio is calculated by comparing the oxygen uptake of the two
standard samples before and after the bacteria sample has been
exposed to the test sample. The invented measurement cycle is
automated for repetitive monitoring of streams fed to a bio-
logical exidation process and can be used as a feed-forward
control to divert a toxic feed before the loss or inhibition
of the acclimated biomass.
2. Sampling System for a Homogenizable Solid-Liquid Mixture
Solid-liquid mixtures, such as an acclimated bacteria in an
activated sludge aeration tank containing 2000 to 4000 mg/1
suspended solids, are diluted as necessary and homogenized
with a high-speed mixer. The homogenized sample then enters
a stand pipe to be de-aerated, then flows by gravity to the
sample injection valve of a monitoring instrument, and into
the second leg of the U-shaped stand pipe. This flow flushes
the sample valve, and upon injection of the homogenized sample
all of the solenoid valves on the U-shaped stand pipe are
de-energized to drain the excess sample. This restores the
flow of a cobinuous sample of a clear waste water, e.g.
the feed through the system, thus cleaning the lines from
any residual from the batch sample.
160
-------
SECTION XIII
GLOSSARY
ABBREVIATIONS
COD chemical oxygen demand
DO dissolved oxygen
JCU Jackson candle turbidity units
MLTC mixed liquor total carbon
MLVSS mixed liquor volatile suspended solids
N:C nitrogen to carbon ratio
SVI sludge volume index
TC total carbon
TCp total carbon of the feed
TC]y[ total carbon of the mixed liquor
TCA total carbon analyzer
TOD total oxygen demand
TODr total oxygen demand removed
VSS volatile suspended solids
MATHEMATICAL SYMBOLS
A waste calibration constant in equation 10
B waste calibration constant in equation 10
C concentration of DO in a bacteria sample, mass/volume
Cs saturation DO concentration in a bacteria sample,
mass/volume
E percent efficiency of substrate removal, 100 (S0-Si)/
So
F food concentration, mass/volume
F/M food to microorganisms ratio or loading
(F/M)r F/M "removed" or (S0-Si)/Xavg
k substrate utilization kinetic constant—maximum
rate of substrate utilization per unit weight of
microorganisms, time"1
161
-------
kr combined substrate biodegradation rate constant,
volume/mas s • time
microorganism decay coefficient, time"1
Ke endogenous reaction rate constant , time"1
KLa oxygen mass transfer coefficient, time"1
Ks half velocity coefficient, equal to the substrate
concentration when the specific substrate utiliza-
tion equals (1/2 )k, mass/volume
M microorganism concentration, mass/volume
n bio-settler efficiency, percent
Q influent waste flow rate, volume/time
R recycle sludge flow rate, volume/time
rr oxygen uptake rate by microorganisms, mass/volume* time
S0 feed soluble organic substrate concentration,
mass/volume
Si soluble organic substrate concentration in aeration
basin, mass/volume
T total bio-settler underflow rate, volume/time
t time
V aeration basin volume
W waste sludge flow rate, volume/time
Wmax maximum waste flow rate, volume/time
X concentration of microorganisms in aeration basin,
mass/volume
Xavg average microorganism concentration in aeration basin
over a chosen time interval, mass/volume
Xe effluent microorganism concentration, mass/volume
Xr recycle microorganism concentration, mass/volume
X growth yield kinetic coefficient, mass microorganisms
produced per mass of substrate utilized
Y growth yield kinetic coefficient, mass microorganisms
produced per mass of substrate utilized
162
-------
SECTION XIV
APPENDICES
Appendix Page
A Operation and Maintenance Manuals 165
B Calculations of Plant Performance 175
C Transient State Activated Sludge Computer
Simulation 190
D Steady-State Activated Sludge Computer Model 196
163
-------
APPENDIX A
OPERATION AND MAINTENANCE MANUALS
The automated activated sludge miniplant is operated
unattended but the instrumentation and control systems re-
quire certain routine maintenance and trouble shooting
practices that are detailed in this appendix. The operational
procedures include some sampling collection and laboratory
analyses to check the control systems and assess the overall
performance of the process.
The details of the control systems applied to the operation
of the aeration basin are shown in Figure 12. The automatic
shutdown system is detailed in Figure 62. Operation and
maintenance of these systems are described under the approp-
riate subsection of this appendix.
START-UP PROCEDURE
The influent flow to the equalization tank was started a
day or two before plant start-up depending on the volume
and residence time in the equalization tank. The stepwise
start-up procedurewas as follows:
1. Prepare the required concentration of chemicals and fill
the alum, caustic, acid, and nutrients tanks. Turn the
pov/er on to heat the furnaces of the total carbon and
total oxygen demand analyzers.
2. Add enough sludge to the aeration basin to obtain the
desired mixed liquor suspended solids (MLSS) concentra-
tion when the basin is full. The source of the sludge
could be from the sludge recycle of a sewage treatment
plant or a batch-acclimated culture previously prepared
-------
2. Adjusting the set points for the alum and caustic flow
rates
3. Checking the liquid level of the alum and caustic tanks
The surface scatter turbidimeter required infrequent replace-
ment of the electrical fuse, the lamp, and the meter relay
lamp.
The Total Oxygen Demand Analyzer, monitoring the TOD in the
feed and effluents required the following routine maintenance:
1. Changing filters to sampling system once a week
2. Checking rubber tubing in the peristaltic pumps
3. Checking leaks or plugging in sampling lines
4. Checking the rotary sampling valve for leaks or plugging
5. Checking the stream selector for proper operation sequence
6. Adjusting the nitrogen flow and electrolyte level in
the oxygen detector of the instrument
7- Adjusting the level in the instrument scrubber, the
detector current, and the automatic zero of the recorder
for proper range of the signal
8. Replacing the scrubbing and detector cell solutions once
a week
9- Adjusting sample valve 0-ring clearance as necessary
10. Replacing furnace tube with freshly prepared catalyst
at intervals found necessary by the operation
11. Changing the nitrogen cylinder when empty
12. Replacing burned out light bulbs
165
-------
in drums by a fill-and-draw procedure. Open the air
supply valve and adjust the air rotaraeters to the re-
quired flow so as to maintain 2 to 3 mg/1 dissolved
oxygen in the aeration basin. A yellow springs oxygen
meter is used to measure the DO in the aeration basin.
When the head in the equalization tank is enough to
prime the feed-recycle pump, open all valves in the
recycle loop. Turn the pH control on bypass and press
the start switch. When recycle has started, set the pH
at the control point midway between the upper and lower
shutdown limits. Slowly close the ball valve on the
pump suction until five inches of mercury vacuum is
shown on the gauge. Open the acid valve and turn the
pH controller to automatic to start the acid flow.
Observe the rotameter to check the acid flow and when
the pH, as recorded, is in control, turn the pH bypass
off.
Start the feed to the aeration basin by turning the
set point on the pneumatic proportional controller at
the panel. This controller is calibrated for maximum
flow of one gpm and is usually started at 0.5 gpm
depending on the concentrations of the influent and
MLSS in the aeration basin.
Check the temperature of the influent in the feed line as
it enters the aeration basin. Control the temperature by
turning on one or several of the electrical heaters around
the feed line or diverting the feed through the cooling
tower to maintain the temperature required for the test
being conducted.
167
-------
As the overflow from the aeration basin fills about half
of the biosettler, start the sludge recycle pump.
The variable drive on the pump is adjusted to the desired
recycle flow rate. The rate is checked at the inlet
to the aeration basin using a stopwatch and a one-liter
graduated cylinder.
Observe the biosettler as it overflows to the flocculator
and the last settling tank. Control the overflow weirs
in the two settling tanks and remove the air trapped in
the lines to obtain a smooth gravity flow through the
flocculator and the last settling tank. The total over-
flow from the biosettler is directed through the Hach
Surface Scattering Turbidimeter to the flocculator.
Put the Total Carbon Analyzer in service by turning on
the recorder switch, the timer switch, and the stream
selector switch. Keep the stream selector switch on
Stream I (the feed) for a one cycle, then observe the
instrument and the sampling system operation through a
complete cycle. The sampling system includes Stream II
(the bacteria sample) that is diluted, homogenized, and
pumped through to the instrument slide valve.
Check the nutrient pump for proper operation and flow.
The required flow rate of the nutrients to obtain the
right ratio of organic carbon in the feed to the
nitrogen and phosphorus in the nutrient solution is
controlled by two variables. One is the speed of the
nutrients pump that is adjusted manually. The other is
the fraction of a (30 seconds) cycle that the pump is
on. This is automatically controlled by the feed total
carbon signal from the analyzer.
168
-------
10. Operate the plant with total recycle and no sludge waste,
bypassing the F/M control system, till the required
steady-state concentration of MLVSS is reached. Adjust
the controller, at the panel, that receives the F/M
signal from the total carbon analyzer to obtain the
required flow rate of sludge waste. This signal operates
a three-way solenoid valve on the sludge recycle line
and opens it to the waste line for a fraction of a 60-
second cycle. The sludge waste flow rate is measured
in a graduated cylinder at the wasted sludge drum. The
maximum flow of wasted sludge is controlled by a limit
switch that should be set so the sludge wasted during
a 24-hour period would not exceed a precalculated
maximum value. This limit switch guards against a
washout of the bacteria in the event that the F/M
control system fails.
11. Put the Hach Surface Scatter Turbidimeter in service by
switching it on and setting the range to record the
turbidity of the biosettler overflow. Open the manual
valves on the alum and caustic tanks, and start the
flocculator stirrer. The flow of the alum solution is
controlled at the panel by adjusting the set point
depending on the optimum dosage found by experiment. The
flow of the caustic solution is controlled by setting
the pH control system at the predetermined optimum pH
(about 6.5).
12. Calibrate the Total Oxygen Demand Analyzer using standard
solutions setting the instrument zero at 5-0 on the
sensitivity dial. Put the stream selector in service by
starting the sampling pumps and adjusting the stream
169
-------
selector switches 1, 2, and 3. Check the flow of all
samples and the signals en the recorder through one
complete cycle. Readjust the sensitivity of the instru-
ment to obtain a feed TOD signal around 75 percent of
chart.
PLANT OPERATION, SAMPLING, AND ANALYSES
The operation of the automated miniplant was limited to
daily and weekly routine checks of the control systems.
The extent of sampling and analyses depended on the experi-
mental design and the data required. The objective of the
fresh water and salt water glycol feed tests was to evaluate
the control systems. Accordingly, data collection was limited
to those data necessary for this evaluation.
The influent pH-controller had a limit switch on high and
low pH that closed the feed valve to the aeration basin if
the pH got out of control and opened it when the pH control
was regained. The control points were set at 5.5 and 9.0
with the set point at an intermediate value of 7.8- The
pH electrodes were checked with buffer solutions as frequently
as needed. A twice weekly check was found to be sufficient.
The nutrient control system was checked by running daily
analysis of ammonia in the final effluent from the plant.
The dial setting on the nutrient pump was then readjusted
to minimize the residual ammonia in the effluent. A daily
check of the amount of nutrients used was made by reading
the calibrated sight glass on the nutrients tank, where
each 3 mm was equal to one liter of solution. Using a
nutrient solution of 1 percent ammonia and 0.1 percent phos-
phoric acid, normal operation of the plant consumed one tank
of 55 gallons every two weeks of operation.
170
-------
The food to microorganisms (F/M) control system required
daily attention and service. The wasted sludge, collected
in a 55 gallon drum, was measured and samples for analysis
every 24 hours of operation. The sludge recycle rate was
measured manually every day to check the performance of
the recycle pump and to verify the recycle and waste sludge
flow rates. The sampling system to the total carbon analyzer
was cleaned daily to prevent any solids accumulation in the
lines that could result in irregular analyses or even plug-
ging of the sample flow.
The flocculation control system required very little
attention. If the turbidity readings were recorded, the
24-hour results were obtained; otherwise, the indicator of
the instrument should be read and entered in the data book
daily.
The Total Oxygen Demand Analyzer, monitoring the feed and
effluents, had no controls that affected plant operation.
Calibration curves for this instrument and the Total Carbon
Analyzer were run once a week. Daily, one standard was run
to check of the two instruments, and the span control was
reset if necessary.
Laboratory analyses were made daily as follows:
1. Feed grab sample: propylene glycol by VPC
2. Aeration basin mixed liquor: sludge volume index (SVI),
Mixed liquor suspended solids (MLSS), and volatile
suspended solids (MLVSS)
3. Biosettler overflow composite: MLSS, MLVSS, TOC,
and ammonia
171
-------
4. Final settler overflow composite: MLSS, MLVSS, and
TOC
5. Sludge waste: MLSS and MLVSS
A plant data sheet was posted daily showing the data
recorded by the instruments and the results of the labora-
tory analyses.
MAINTENANCE OF INSTRUMENTS AND CONTROLS
The instruments used for control and monitoring the oper-
ating parameters of the activated sludge miniplant required
strict routine maintenance for good operation. Sample
preparation and pretreatment was emphasized to minimize
downtime for non-routine maintenance. Regular routine
maintenance practices for recorders and support equipment
are not covered in this text.
The influent pH-control system required the following routine
and non-routine maintenance:
1. Inspecting and cleaning the pH-electrodes (once a week)
2. Calibration against high and low pH-buffer solutions
(once a week)
3. Checking the acid line and rotameter for any leaks (daily)
4. Replacing defective electrodes (when necessary)
5. Cleaning acid flow rotameter (once a month)
The Total Carbon Analyzer (Ionics Model 1212) used to
measure the TC in the feed and mixed liquor of the aeration
basin required daily checking of the sampling system besides
the following routine maintenance:
172
-------
1. Checking the liquid level of the liquid gas separator
2. Checking the gas flov; to the infrared analyzer
3. Adjusting wattmeter for correct operating wattage
(185 watts)
4. Inspecting reaction chamber for oven heat
5. Checking the sample slide valve for leaks
6. Adjusting zero on the infrared analyzer
7. Back-flushing of sample flow to analyzer (once every
two days)
8. Calibrating with one or two standards (daily)
9. Running all of the calibration curve (once a week)
10. Replacing nitrogen cylinder when empty
11. Replacing distilled water for water rinse (once a week)
In cases of out-of-control instrument response the main
defective parts to check are:
1. Sample injection valve - which may have to be dis-
assembled, cleaned, and reground to prevent leaking.
2. Reaction tube - which may have to be removed, and the
catalyst and the injection tube cleaned.
3. Infrared detector - which may require removal and clean-
ing of the sample cell.
The flocculatlon control system, comprising the turbidimeter
and the alum and caustic addition, required very little
special maintenance:
1. Draining the turbidimeter once a day to clean the solids
accumulated at the surface.
173
-------
APPENDIX B
CALCULATIONS OP PLANT PERFORMANCE
A computer program was developed to facilitate routine
data handling and to calculate parameters which describe
the miniplant performance. The program was written in
Fortran IV language and a Xerox Sigma 5 batch time-sharing
monitor was used.
A sample data input form (#12) and the resulting computer
printout displaying the summary of plant performance are
presented. The logic of the program is also given (#13 and
Definitions of symbols in the program can easily be related to
the data input sheet and the corresponding READ statements
and to the output printout and the corresponding WRITE state-
ments. Material balances and parameters calculated have been
described in the literature (Zeitoun, et. al., 1971).
-------
•M r-
0)
on ^x^
0 J J >
O *r
" 1 •
3 o_^ — h o
< s *V — 1 u-
.4
£ h\ ^
<" b i t*
°? VX
i X.
•r-
CO
•el-
to
4J C
E 0
•1- +>!-
S- re