United States
Environmental Protection
Agency
Robert S. Kerr Environmental
Research Laboratory
Ada OK 74820
Research and Development
EPA-600/S2-83-073 Jan. 1984
Project Summary
Determination of Activated
Sludge Biokinetic Constants
for Chemical and Plastic
Industrial Wastewaters
Don F. Kincannon and Enos L Stover
The most widely used method of
wastewater treatment is biological
treatment. The use of kinetic models to
describe the behavior of a biological
wastewater treatment process has
become widely accepted practice. The
most often used kinetic models include
those developed by Eckenfelder, McKin-
ney, Lawrence and McCarty, and Gaudy.
Eckenfelder (first model) and McKinney
use first order kinetics for substrate
removal; Lawrence and McCarty, and
Gaudy use the empirical Monod kinetics
for substrate removal; and Eckenfelder
(second order) relates the substrate
removal rate as a function of the
remaining substrate concentration to
the initial concentration. All of these
models contain kinetic constants and
the usefulness of each model is a
function of the reliability of the kinetic
constants. However, there has.not been
enough information available to establish
reliable values for these kinetic constants
for industrial wastewaters.
Cooperative Agreement CR 806843-
01-02 has determined the biokinetic
constants and fate for 24 toxic organic
pollutants when present in a highly
biodegradable wastewater. Biokinetic
constants were also determined for
eight groups of three chemicals. The
biokinetic constants were determined
based upon biochemical oxygen demand
(BODs), total organic carbon (TOC),
chemical oxygen demand (COD), and
specific organic chemicals. The normal
approach for determining biokinetic
constants has been to plot the average
values. This approach masks the actual
scatter of the data and provides an
average value or 50 percent probable
value for the biokinetic constants to be
used in design. This study has produced
a methodology for analysis of the test
results from biological activated sludge
systems for determining the biological
variability inherent in these types of
systems.
This study also investigated the
possibility of predicting the fate and
effluent concentrations of the various
priority pollutants. It was found that
good predictions for the priority pollu-
tants can be made. In general, it can be
predicted that nitrogen compounds,
phenols, oxygenated compounds, poly-
nuclear aromatics, and phthalates will
be removed by biological degradation.
Aromatics will be removed by both
biological degradation • and stripping,
except when the Henry's Law Constant
is in the 10~5 atnvmVmole range, then
that chemical will be removed by
biological degradation only. Halogenated
hydrocarbons present a special problem.
Some are removed by stripping, whereas,
others are removed by combined biode-
gradation and stripping. No parameter,
such as Henry's Law Constant or
partition coefficient, was found that
would predict which removal mechan-
isms would persist. Therefore, biode-
gradability studies must be conducted.
This Project Summary was developed
by EPA's Robert S. Kerr Environmental.
Research Laboratory. Ada. OK. to
announce key findings of the research
project that is fully documented in a
-------
separate report of the same title (see
Project Report ordering information at
back).
Introduction
Provisions in the Clean Water Act
clearly require the U.S. Environmental
Protection Agency to promulgate regulations
to protect streams and publicly owned
treatment works (POTW's) and to see to it
that regulatory practices are in place to
protect the vast investment of public
funds. This law can affect nearly all
industrial manufacturing plants now
discharging or planning to discharge into
POTW's or a stream. An important
category for which effluent guidelines are
needed is waste components from
chemical and plastic industries. For this
category, the problem is extremely
difficult because of the thousands of
different chemicals produced at plants
across the nation. The product lines of
these plants, and thus the nature of the
wastes, are not fixed and change with
market needs. This industrial activity
consequently produces wastewaters of
great and changing variety in the types of
pollutants they contain. Furthermore, a
significant number of the growing list of
priority pollutants is appearing in various
waste streams from the manufacturing
of chemicals and plastics.
Because of the immensity and tremen-
dous variety of products and process
routes from production, it is logical that a
computer programming procedure would
be developed to handle the definition of
wastewater composition as well as
alternate processing routes for the
treatment of wastewater. Computer
programs have been developed that can
combine waste load information on each
product, select various product treatability
routes, combine them, design and cost-
out a plant's treatment system, and,
finally, allocate the estimated treatment
costs to the contributing product process
route. However, before plant costs can be
modeled, the treatment design models
must be established. Design models are
available but the biokinetic parameters
for the chemical and plastic industries are
not available. Therefore, the determination
of the biokinetic parameters becomes of
vital concern to the success of the cost
model.
The current best practical technology
(BPT) for treatment of chemical industrial
and plastic manufacturing wastes usually
involves activated sludge, since biological
treatment is often found to be the most
cost-effective in dealing with relatively
low concentrations of the types of organic
materials in such wastes.
Modern approaches for the design of
activated sludge employ mathematical
process models depicting the relationships
among factors affecting the kinetics of
wastewater purification. All design models
require quantitative assessment of
numerical values for biokinetic constants
for the activated sludge in question.
These values are, in general, a function of
the type of carbon source comprising the
carbonaceous biochemical oxygen demand.
Various models are available for use in
designing activated sludge processes.
The primary objective of this study was
to determine the biokinetic constants
needed in process modeling for predicting
effluent quality and design configurations
of activated sludge processes treating the
waste streams from chemical and plastic
manufacturing. Secondary objective was
to refine the method of combining the
biokinetic constants of components in a
mix in order to determine the overall
biokinetic constants of the mix and to
establish the fate of the priority pollutants
in the wastewater.
Conclusions
High treatment efficiencies (in terms of
biochemical oxygen demand, BODs; chemi-
cal oxygen demand, COD; and total organic
carbon, TOC) can be achieved by biological
treatment of wastewaters containing
toxic priority pollutants. This study has
shown that priority pollutants can be
accommodated in biological wastewater
treatment processes. In addition, it was
found that many of the priority pollutants
were biodegraded to very low concentra-
tions. In other cases, the priority pollutants
were stripped to low levels without
interfering with the biodegradation of the
other organics.
The biokinetic constants of the waste-
waters containing the priority pollutants
are representative of a readily biodegrad-
able .wastewater. The biokinetic constants
are subject to variability, and a frequency
analysis should be used in determining the
biokinetic constants. A new design model
provides an alternative method for
accounting for the variability of the
system. This new model provides designs
comparable with the other design models
without the variability of the biokinetic
constants.
The fate and effluent concentrations of
priority pollutants can be predicted with a
high degree of accuracy. Parameters,
such as Henry's Law Constant and Log
partition coefficient, are key elements in
predicting the fate of priority pollutants.
Some priority pollutants are both biode-
graded and stripped. These are more
difficult to predict. Some means of
estimating the biodegradability of priority
pollutants is needed.
Recommendations
This study has investigated the biological
treatment of 24 toxic organic chemicals.
The study also included eight groups of
three chemicals as a mixture. It is
recommended that additional research
be conducted using chemicals that would
enhance sorption as a removal mechanism.
Biodegradability and fate of additional
chemicals could allow the creation of a
structure/activity correlation for predict-
ing the degradability and fate of toxic
organics. Cost impacts for the correction
of existing problems and the ability to
anticipate adverse environmental impacts
from the manufacture of chemicals
would be made available through the use
of such correlation.
It would also be of benefit to investigate
the effects of factors other than the
priority pollutants. Surface tension,
sulfur compounds, inorganics, etc. may
be much more detrimental to biological
treatment than, so called, toxic organic
chemicals.
It is recommended that the water
pollution control field recognize that
"toxic" organic chemicals can be treated
in a biological wastewater treatment
plant with no detrimental effects to the
treatment process.
Experimental Methods
The general experimental plan specified
bench-scale continuous flow activated
sludge reactors be used to treat a
synthetic wastewater containing selected
chemicals normally present in chemical
industry wastewaters. The bench-scale
activated sludge system is shown in
Figure 1. The system consisted of a
stainless steel internal recycle reactor.
The activated sludge reactor had a
volume of 3.0 liters and the settling
compartment had a volume of 3.23 liters.
The reactor and settling compartment
both had a stainless steel cover. The
wastewater was pumped from a sealed
feed tank to the reactor. The effluent from
the settling unit flowed by gravity to a
collection tank. The off-gas was pulled by a
vacuum pump through a purge trap
which contained 6 inches of Tenax and 4 •
inches of Silica Gel. The influent air and
off-gas were measured with air-flow
meters. The influent wastewater flow
was regulated to provide a hydraulic
-------
Off-Gas
D^^~~^F,ow VK""m
Purge Trap Meter ™mp
Effluent
Collection Tank
Wastewater
Feed Tank
Figure 1.
Pump
Bench-scale activated sludge pilot plants.
detention time of 8 hours in the activated
sludge reactor. All parts of the experimental
apparatus were constructed of materials
to minimize contamination.
Two types of experimental studies
were conducted with the previously
described stainless steel reactors. A
nonbiological system was operated to
determine the strippability of the chemical
compound. In this system, the reactor and
settling compartment were filled with
distilled water. The feed tank contained
the wastewater. The wastewater was
pumped to the reactor and the level of
TOC and specific compound present in
the reactor was observed as a function of
time. When the wastewater had been
completely diluted-in, influent, effluent,
and off-gas samples were collected for
specific compound analyses.
The second experimental system was a
biological activated sludge system.
Activated sludge for initial seeding was
obtained from a local municipal activated
sludge plant. Three individual systems
were acclimated to each synthetic
wastewater containing the pollutant to be
evaluated. The activated sludge systems
were operated at mean cell residence
times (SRT) of 2, 4, and 6 days. After
acclimation, influent, effluent, mixed
liquor and off-gas samples were collected
over a 60-day period for analyses. All
three reactors were operated at a
hydraulic detention time of 8 hours.
The synthetic wastewater referred to
as the "base mix" contained ethylene
glycol, ethyl alcohol, glucose, glutamic
acid, acetic acid, phenol, ammonium
sulfate, phosphoric acid, and salts. The
specific chemical compoundto be studied
was added to this base mix. The specific
compound and the corresponding initial
BODs, TOC, and COD are shown in
Table 1.
Results and Evaluation
Treatment efficiencies when measured
in terms of BOD5, TOC, and COD were
very high for all systems in this study. The
mean BODs did not exceed 10.0 mg/l for
any system. Generally, the mean BOD5
was below 5.0 mg/l. This indicates that
the "base mix" was highly biodegradable.
This factor did create problems in
obtaining data for determining biokinetic
constants. The ideal situation is one in
which the effluent BODs varies over a
wide range. However, with the wastewa-
ters used in this study, even an SRT of 2
days produced a very low effluent BODs.
Therefore, SRT values of 2,4, and 6 days
produced effluent BODs's with very little
difference. The high treatment efficiencies
also indicated that the priority pollutants
used in this study had very little or no
effect on the biological treatment of the
wastewater. Even when the priority
pollutant was not biodegraded, it did not
affect the treatment efficiency.
It was found that the specific chemicals
were removed by several mechanisms.
These included stripping, biodegradation,
and sorption. The percent removed by
each mechanism is shown in Table 2.
The normal approach for analyzing and
presenting biological system data is to
plot the average values at each //„ or SRT.
This approach masks the actual scatter of
the data and provides an average value or
50 percent probable value for the
biokinetic constants to be used in design.
Use of these biokinetic constants will
provide a system capable of achievingthe
desired effluent quality only 50 percent of
the time. The allowed statistical variability
of the effluent discharge criteria must be
evaluated to determine the appropriate
biokinetic constants to use for design.
Designing to meet a specific effluent
quality 95 percent of the time will require
a significantly larger reactor volume than
that required for 50 percent of the time
and will require use of the 95 percent
probable biokinetic constants, as shown
in the following example.
Using Eckenfelder's design model and
the following design conditions the
impacts of data variability for the acryloni-
trile - containing wastewater can be
demonstrated.
F = 10MGD
S, = 200 mg/l BOD5
Se = 5mg/l BODs
X = 1000 mg/l volatile suspended
solids
The activated sludge reactor volumes are
determined as follows:
= F
, - Se)
X Ke' Se
Using the 50 percent line or Ke' = 31, the
required reactor volume will be:
v = 10x106x200x(200-5)
1000x31 x5
V= 2,516,000 gallons .
This reactor volume of 2,516,000 gallons
should provide an effluent soluble BOD5
(Se) of 5 mg/l 50 percent of the time.
Using the 95 percent line or Ke' = 11, the
required reactor volume will be:
v= 10 x 106x200x(200-5)
1000 x 11x5
V= 7,091,000 gallons
-------
Table 1. Wastewater Concentrations
Group
Chemical
Wastewater concentrations, mg/l
BOD5 TOC COD
1
II
III
IV
V
VI
VII
VIII
Tetrachloroethane
Nitrobenzene
2,4 dichlorophenol
Combined
Acrylonitrile
Acrolein
1,2 dichloropropane
Combined
Benzene
Methylene chloride
Ethyl acetate
Combined
1,2 dichlorobenzene
1,2 dichloroethane
Phenol
Combined
2,4 dinitrophenol
1,3 dichlorobenzene
1,1,1 trichloroethane
Combined
Pentachlorophenol
Trichloroethylene
Bis (2-ethylhexyl) phthalate
Combined
Toluene
Ethylbenzene
Phenanthrene
Combined
Chloroform
Carbon tetrachloride
Naphthalene
Combined
244
255
298
314
192
219
247
279
212
287
259
248
162
201
277
268
245
178
244
227
265
295
271
270
180
236
270
217
240
257
219
317
224
253
234
246
189
190
199
197
132
170
145
133
152
187
163
176
203
136
183
143
153
179
159
169
129
143
166
147
155
162
163
170
496
668
572
574
483
477
480
565
441
480
450
400
416
460
486
454
551
379
530
429
470
530
528
574
416
520
510
470
363
387
390
409
Control
"Base Mix"
266
170
460
This reactor volume of 7,091,000 gallons
is almost three times the previous reactor
volume and should provide an effluent
soluble BOD5 (Se) of 5 mg/l 95 percent of
the time. Using the 75 percent line or Ke' -
19, the required reactor volume will be
4,105,000 gallons. If the reactor volume
was designed at 2,516,000 gallons, the
following probability of occurrence of Se
could be expected:
Probability Level %
50
75
95
Se (mg/l)
5
8
14
The variability analysis approach
presented indicated that Kd was a true
biokinetic constant, //max was also treated
as a true biokinetic constant, and since
/AT»X = K-Yt, and Yt was found to be a
variable biokinetic coefficient, jumax and K
could not both be treated as constants. If
Yt = yUmax/K varies, yumax, K, or both must
vary. If both /jmax and K vary, the
prediction of these coefficients becomes
more difficult. The decision to make//max a
biokinetic constant and K a variable
biokinetic coefficient depending on the
variable biokinetic coefficient Yt simplifies
the prediction of (ime* and K. The following
reasoning was used in the selection of
/Umax as a biokinetic constant.
K is simply the maximum U, and U was
found to vary at constant//,, or/j. At/Umax, U
would be equal to K and could be
expected to be variable. Also in a
heterogeneous population such as in the
activated sludge process treating a
particular wastewater, several different
kinds of bacteria will actually be in
predominance; however, shifts in the
ratios of these specific micro-organisms
in predominance can change. As these
population dynamic shifts occur Yt
changes, and thus, the mixed liquor
solids (X) changes. The specific substrate
utilization rate (U) could be expected to
change as X changes; however, /umax
should not change. As //„ or /u approaches
A/max, the microorganisms with lower /umax
values than the existing fjn or ^ will be
washed out of the system. Therefore, the
microoranisms with the greatest /Umax will
predominate at the /umax condition of the
system. At this absolute //max which could
be considered to be a constant, many
microorganisms will have been washed
out of the system, but several microorgan-
isms will still remain. These few organisms
may still provide a variable maximum U or
K value but with lower variability since
the few remaining organisms should
provide less variability. All the other
substrate utilization terms (Ks and Ke' or
Ke) were also found to be variable in a
manner similar to the true yield. Even at
very constant controlled /un's or SRT's, all
the substrate utilization characteristic
descriptors were found to be variable.
Based on this reasoning, Yt Ke', Ke, Ks and
K are all considered to be variable
biokinetic coefficients, while Kd and fjmax
are considered to be biokinetic constants.
The variability observed in these biokinetic
coefficients during the treatability study
can and should be considered during
scale-up design to achieve specific
effluent discharge criteria.
In addition to the variability analysis for
the existing models, a new model has
been presented. This model provides an
approach in which there is no variability
of the biokinetic constants. All variability
can be handled as influent variability and
normal procedures can be used. The new
design model provides design values
comparable with the other design models.
This study has also looked at the
possibility of predicting the fate and
effluent concentrations of the various
priority pollutants. It has been found that
good predictions for the priority pollutants
can be made. In general, it can be
predicted that nitrogen compounds,
phenols, oxygenated compounds, poly-
nuclear aromatics, and phthalates will be
removed by biological degradation.
Aromatics will be removed by both
biological degradation and stripping,
except when the Henry's Law Constant is
in the 10~5 atm-rnVmole range, then that
chemical will be removed by biological
degradation only. Halogenated hydrocar-
bons present a special problem. Some are
removed by stripping, whereas, others
are removed by combined biodegradation
and stripping. No parameter, such as
Henry's Law Constant or partition coeffi-
cient, was found that would predict which
removal mechanisms would persist. It
was also found that the type of reactor
could be a factor. Therefore, biodegradabil-
ity studies must be conducted. If the
biodegradability is high, the chemical will
be both stripped and biodegraded.
It was also found that biodegradability
and stripping constants could be developed
for each priority pollutant and used very
successfully in predicting the effluent
-------
Table 2. Removal Mechanisms of Toxic Organics
Percent Treatment Achieved
Single Units
Combined Units
Compound
Strip. * Sorption Biol,H
Strip. Sorption Biol.
Nitrogen Compounds
Acrylonitrile 99.9
Phenols
Phenol 99.9
2,4-DNP 99.3
2,4-DCP 95.2
POP 0.58 97.3
Aromatics
1,2-DCB 21.7 78.2
1.3-DCB
Nitrobenzene 97.8
Benzene 2.0 97.9
Toluene 5.1 0.02 94.9
Ethylbenzene 5.2 0.19 94.6
Halogenated Hydrocarbons
Methylene Chloride 8.0 91.7
1,2-DCE 99.5 0.50
1.1.1-TCE 100.0
1,1,2.2-TCE 93.5
1,2-DCP 99.9
TCE 65.1 0.83 33.8
Chloroform 19.0 1.19 78.7
Carbon Tetrachloride 33.0 1.38 64.9
Oxygenated Compounds
Acrolein 99.9
Polynuclear Aromatics
Phenanthrene 98.2
Naphthalene 98.6
Phthalates
Bis(2-Ethylhexyl) 76.9
Other
Ethyl Acetate 1.0 98.8
2.0
6.2
7.5
2.0
75.0
99.7
96.2
99.5
69.7
15.0
28.0
0.30
1.36
0.03
90.0
99.9
99.2
77.1
97.8
99.9
33.8
97.2
93.7
92.4
97.2
25.0
30.1
82.8
77.7
90.0
94.7
78.6
85.7
98.9
*Strip. = Stripping.
**Biol. = Biological.
concentrations for the pollutants when
combined with other priority pollutants or
when the influent concentrations varied.
Don F. Kincannon and Enos L Stover are with Oklahoma State University,
Stillwater. OK 74078.
Thomas E. Short is the EPA Project Officer (see below).
The complete report consists of two parts, entitled "Determination of Activated
Sludge Biokinetic Constants for Chemical and Plastic Industrial Wastewaters,"
(Order No. PB 83-245 233; Cost: $14.5O) and
"Determination of Activated Sludge Biokinetic Constants for Chemical and
Plastic Industrial Wastewaters (Appendix A—Raw Data)," (Order No. PB 83-
245 241; Cost: $25.00, costs subject to change)
The above reports will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Robert S.Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
P.O. Box 1198
Ada, OK 74820
-------
United States
Environmental Protection
Agency
Official Business
Penalty for Private Use $300
Center for Environmental Research
Information
Cincinnati OH 45268
mi ^^>
111!
III!
|m
III!
U.&OFRCIALMAJL
Ill/ . „ \ ™* --^
IK ^ F:3 : "E!f j Sm:|5^^
{, U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/825
------- |