WATER POLLUTION CONTROL RESEARCH SERIES
17090 FQJ 09/71
BIOLOGICAL CONCEPTS FOR
DESIGN AND OPERATION OF THE
ACTIVATED SLUDGE PROCESS
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the re-
sults and progress in the control and abatement of pollution
in our Nation's waters. They provide a central source of
information on the research, development, and demonstration
activities in the water research program of the Environmental
Protection Agency, through inhouse research and grants and
contracts with Federal, State, and local agencies, research
institutions, and industrial organizations.
Inquiries pertaining to Water Pollution Control Research Reports
should be directed to the Chief, Publications Branch (Water),
Research Information Division, R&M, Environmental Protection
Agency, Washington, D.C. 20460.
-------
BIOLOGICAL CONCEPTS FOR DESIGN AND OPERATION OF THE
ACTIVATED SLUDGE PROCESS
by
Anthony F. Gaudy, Jr., and Elizabeth T. Gaudy
Bioenvironmental Engineering Laboratories
School Civil Engineering & Microbiology Department
Oklahoma State University
Stillwater, Oklahoma
for the
Office of Research and Monitoring
ENVIRONMENTAL PROTECTION AGENCY
Project #17090 FQJ
September 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.25
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommenda-
tion for use.
-------
ABSTRACT
Over the past decade, many of the thousands of experiments accomplished
in the principal investigators' laboratories were designed to permit
interrelated analyses and correlation of the 10-year research effort
into a general body of conceptual principles applicable to design and
operation of activated sludge processes. The aim of this report is to
present these biological concepts in terms readily understood by engi-
neering professionals. The document is by no means a design and oper-
ational manual, although some recommendations for design and operational
procedures are included.
Generalized concepts of BOD exertion, the use of ACOD as a design and
operational tool, the stoichiometry and mass balance concepts of treat-
ment, and kinetic,equations for microbial growth.-are. presented. Design
models are discussed, and a mode,! for completely mixed reactors holding
recycle solids, XR, constant is,recommended. Some guidelines for
accommodation ofLvarious types of shock .loadings.are included. Con-
cepts of oxidative assimilation and the multiple effects of solids con-
centration, nitrogen concentration, and detention time are related; a
new activated sludge process (continuous oxidative assimilation) for
nitrogen-deficient wastes is presented. Data supporting the concept
of total oxidation are presented, and a modification of the extended
aeration process incorporating chemical hydrolysis of portions of the
sludge is recommended. In the final chapter, some possible flow dia-
grams for complete aerobic treatment (purification and sludge disposal)
of metabolizable organic wastes are presented.
The report was submitted in fulfillment of project 17090 FQJ under the
sponsorship of the Environmental Protection Agency.
Key words: activated sludge, kinetic models, shock loads, nitrogen-
deficient wastes, ultimate disposal, oxygen demand, oxidative
assimilation, endogenous phase, purification efficiency.
111
-------
CONTENTS
Page
Abstract ill
Contents v
List of Figures viii
List of Tables xii
Conclusions l
Recommendations 3
I. Introduction 5
The Perspective 5
The "Technological Gap" 6
The Nature of the Report 7
II. Measurement of Purification 9
The Concept of Biochemical Oxygen Demand 9
Exertion of Biochemical Oxygen Demand 10
Theoretical COD and BOD 12
ACOD as a Parameter of Purification 14
Parameters for Design and Operation 16
Additional Information 17
Summary and Conclusions is
III. 'Fundamental Stoichiometry of Biological Treatment 19
"Equation" for Purification 19
Partition of Substrate Between Respiration and Synthesis 20
Energy Coupling Through ATP 21
Energy Balance 22
V
-------
Page
Summary and Conclusions 26
IV. Kinetics of Microbial Growth 29
Exponential Growth 29
Relationship Between u and S 32
Sludge Yield 40
Summary and Conclusions 46
V. Kinetics of Completely Mixed Reactors 49
Criteria for Complete Mixing 49
Growth in the "Steady State" 54
Cell Recycle with Constant Concentration Ratio 58
Cell Recycle with Constant Return Solids Concentration 70
Summary and Conclusions 80
VI. Factors Tending to Disrupt the Steady State - Shock Loadings 83
Quantitative and Qualitative Changes in Substrate 84
Qualitative Shock Loads 90
Changes in Chemical Composition of the Inflowing Waste
Other than the Organic Substrates 92
Changes in Environmental Conditions not Involving
Chemical Composition of the Waste 94
Summary 96
VII. Process Modifications for Nitrogen-deficient Wastes 99
Nitrogen Supplementation for Biological Treatment 99
Oxidative Assimilation 105
Proposed New Process for Treatment of Nitrogen-
deficient Wastes 108
vi
-------
Page
Summary and Conclusions H4
VIII. Extended Aeration Process 115
Recent Investigations of Extended Aeration 116
Engineering Control of Total Oxidation 124
Summary 129
IX. Summary and Recommendations 131
Acknowledgments 141
References 143
Patents and Publications 149,
Glossary of Terms, Abbreviations, and Symbols
Vll
-------
FIGURES
Number Page
1. Generalized plot of substrate concentration, bio-
logical solids concentration, and oxygen utilization
during exertion of biochemical oxygen demand.
Circles mark inflection points. il
2. Arithmetic and semi-logarithmic plots of microbial
growth during the substrate removal phase in a
batch system. 31
3. Effect of initial substrate concentration, Sg, on the
rate and total amount of microbial growth. The ver-
tical lines indicate the end of the exponential growth
phase. 34
4. Hyperbolic plot of the relationship between specific
growth rate, y, and initial substrate concentration,
S0 (upper graph), and a straight line plot, S0/y vs.
SQ, of the same data (lower graph). 35
5. Effect of the numerical value of the saturation con-
stant, IG, upon the degree of curvature of the hyper-
bolic relation between y and S0. Curves were calculated
using equation 18 with ym.Y = 0.4 hr"' and the values of
.. u J.L. j- ma A
KS shown on the figure. 37
6. Hyperbolic plot of the relationship between specific
growth rate, y, and initial substrate concentration,
S0, for a heterogeneous microbial population of sewage
origin growing on a concentrate prepared from the
soluble portion of municipal sewage. 41
7. Constancy of sludge yield throughout the substrate
removal period during growth on glycerol of a hetero-
geneous microbial population of municipal sewage origin. 43
8. Flow diagram for a continuous flow, completely mixed
reactor of the once-through type. so
9. Theoretical dilute-in and dilute-out curves for a com-
pletely mixed reactor calculated for a dilution rate
of 0.125 hr-1. 52
10. Comparison of theoretical and experimental dilute-in
curves for a completely mixed laboratory reactor
operated at various detention times. 53
Vlll
-------
Number Page
11. "Steady state" parameters measured in a completely
mixed laboratory reactor of the once-through type,
operated at a dilution rate of 0.33 hr"1 with nom-
inal influent substrate concentration, S<, of
1060 mg/1 glucose COD. 59
12. Flow diagram for a continuous flow, completely mixed
reactor employing cell recycle at a constant ratio,
c, of recycle solids, XD, to aeration solids, X\
Diagram represents the kinetic model proposed by
Herbert. 61
13. Comparison of predicted levels of X and S at various
dilution rates in a once-through reactor (broken
lines) and a reactor employing cell recycle concen-
tration ratio, c, as a system constant in accordance
with the kinetic model equations of Herbert. Values
used for calculation were: l^max* 0.5 hr~l, Ks, 75
mg/1; Y, 0.60; a, 0.25; c, 4.0, Sn-, 1000 mg/1. 63
14. Operational data for a laboratory reactor operated
according to the kinetic model of Herbert with cell
recycle at a constant recycle ratio, c, of 1.5.
S-j = 1060 mg/1 glucose COD, « = 0.25, aerator deten-
tion time = 3 hr. 66
15. Comparison of observed (solid lines) values of X and
S with values predicted (broken line curves) by the
kinetic model equations of Herbert using experimen-
tally determined values of the biological constants
(see text for details). 67
16. Flow diagram for a continuous flow, completely mixed
reactor employing cell recycle at a constant recycle
solids concentration, XR. Diagram represents the
kinetic model proposed by Ramanathan and Gaudy. 71
17. Comparison of values for X and S at various dilu-
tion rates predicted by the two kinetic models for
cell recycle. Curves A are those predicted by the
model equations of Ramanathan and Gaudy. Curves B
are those predicted by the model equations of
Herbert. 74
IX
-------
Number
18. Predicted levels of aeration solids, X, and effluent
COD, S, for a range of organic loadings, S-j, at var-
ious dilution rates. Curves were computed according
to the model equations of Ramanathan and Gaudy using
the following values for system constants: ymax>
0.5 hr-1; KC, 75 mg/1; Y, 0.60; a, 0.25; XR,
10,000 mg/1. 76
19. Expanded scale plot of the curves from Figure 18 for
organic loadings up to 1000 mg/1 COD. 77
20. Response of a continuous flow, completely mixed
reactor of the once-through type to a severe quanti-
tative shock load consisting of a change in S-j from
450 mg/1 to 1450 mg/1 COD. The dilution rate was
0.244 hr-1. 87
21. Response of a continuous flow, completely mixed
system of the once-through type to a quantitative
shock load consisting of a two-fold increase in S^.
The dilution rate was 0.125 hr*1. 89
22. Relation between effluent substrate concentration, S,
and NH3 - N concentration in the feed at three COD:N
ratios for dilution rates ranging from 1 to 1/12 hr-1.
Feed COD was 1060 mg/1. 101
23. Relation between effluent and influent concentrations
of NH3 - N at three COD:N ratios for dilution rates
ranging from 1 to 1/12 hr~!. 102
24. Relation between purification efficiency, nitrogen
content of the sludge, dilution rate, and feed COD:N
ratio. 103
25. Substrate removal, sludge accumulation, and sludge
composition in the absence of protein synthesis in
a batch reactor. 107
26. Proposed flow sheet for the treatment of nitrogen-
deficient industrial wastes by the continuous oxida-
tive assimilation modification of the activated
sludge process. 109
27. Laboratory-scale pilot plant employed in operational
feasibility studies for the continuous oxidative
assimilation process. ill
-------
Number Page
28. Operational data for treatment of nitrogen-free
acetic acid waste by the continuous oxidative assim-
ilation process with nitrogen supplementation in the
endogenous aerator at a COD:N ratio of 70:1. H3
29. Performance data for a continuous flow extended
aeration activated sludge in a laboratory pilot
plant operated with total solids recycle, showing a
period of accelerated autodigestion. 119
30. COD removal capability of extended aeration activated
sludge after the number of days of continuous oper-
ation with total solids recycle indicated on each
graph. 120
31. Effect of aging on endogenous oxygen uptake by an
extended aeration activated sludge during long-term
pilot plant studies employing total solids recycle. 122
32. Total oxidation, during the endogenous phase, of the
biological solids synthesized during the substrate
removal phase in a batch system using a heterogeneous
population developed from a sewage seed. 123
33. Metabolism of hydrolyzed activated sludge by an
extended aeration activated sludge. 126
34. Proposed extended aeration activated sludge process
incorporating chemical hydrolysis for control of
sludge concentration. i27
35. Possible flow diagram for activated sludge process
incorporating suggested modifications for operational
control of purification efficiency and including
sludge disposal by aerobic autodigestion aided by
chemical hydrolysis. This diagram is based on the
conceptual principles and research findings presented
in this report. 135
36. One of several possible modifications of the flow
diagram shown in Figure 35. See text for details. 138
XI
-------
TABLES
Number Page
I. Values of ymax and Ks for Heterogeneous Populations of
Sewage Origin Growing on Glucose 39
II. Statistical Summary of Sludge Yield Values for Heter-
ogeneous Populations of Sewage Origin Grown on Various
Carbon Sources 45
III. Operational Parameters for a Completely Mixed Contin-
uous Flow Reactor With Cell Recycle 68
XI1
-------
CONCLUSIONS
The general conclusion which emerges from the decade of research inves-
tigations briefly summarized in this report is that activated sludge
processes must be designed and operated with greater consideration of
the biological nature of the process. Concepts based on microbial phys-
iology can be adapted to systems employing heterogeneous populations,
and application of biological principles can lead to innovative improve-
ments in the basic process. Design of treatment processes must include
provision for operational controls which will ensure reliability of per-
formance, and operation must be based on daily assessment of purifica-
tion efficiency. Both design planning and operational procedures should
attempt to minimize disruptions of the process. Each treatment facility
should be designed, not according to a standard flow sheet, but as a
combination of unit processes chosen to accomplish specific objectives
in purifying a specific waste. The activated sludge process can and
should deliver reliably a high degree of purification of organic wastes,
but this requires both good design and careful, intelligent operation.
Some of the specific conclusions discussed in the body of the report
are listed below with reference to the pertinent chapter.
1. Insofar as design and operation of activated sludge processes are
concerned, the BOD test should no longer be used as the parameter of
pollutional potential or as a measure of purification efficiency. For
design and for daily assessment of operation, the ACOD measurement
should replace the BOD test (Chapter II).
2. Treatability studies should precede design and a stoichiometric
balance based on measurements of ACOD, oxygen uptake, and ACOD of the
biological solids is recommended for use in these studies (Chapter III).
3. Kinetic equations for microbial growth may be employed with heter-
ogeneous populations, but no precise values for the biological con-
stants can be expected. The values of these constants should be deter-
mined in treatability studies. In general, the following ranges seem
reasonable: ymax, 0.4 to 0.6 hr"1; Ks, 50 to 125 mg/1; Y, 0.4 to 0.6
(Chapter IV).
4. A kinetic model was developed specifically for use in designing
activated sludge processes. The model employs recycle solids concen-
tration, XR, as a system constant, and it is designed to maximize puri-
fication efficiency by maintaining a "steady state" as nearly as is
possible with a heterogeneous microbial population and with a waste
which may vary quantitatively and qualitatively. The model requires use
of an aerated sludge consistency tank and, for wastes subject to con-
siderable variability, a surge basin is recommended (Chapter V).
5. Allowable limits for several types of shocks commonly encountered
-------
in waste treatment were suggested. In general, it was recommended
that for aeration periods of 6 to 8 hours, it should be attempted
through both design and operation to prevent fluctuations in flow rate
of greater than 100% of the design flow, fluctuations in concentration
of metabolizable organic material in the waste of greater than 100%,
fluctuations in pH of greater than one unit and rapid changes in tem-
perature of greater than IQOC (Chapter VI).
6. The standard level of nitrogen supplementation based on a BOD:N
ratio of 20:1 should be reconsidered in design of treatment facili-
ties for wastes deficient in nitrogen. To avoid leakage of nitrogen
in the effluent, treatability studies should determine the optimum com-
bination of detention time, purification efficiency, nitrogen supple-
mentation and nitrogen concentration in the effluent, since all of
these factors are related. For treatment of wastes containing little
or no nitrogen, a new modification of the activated sludge process was
developed, and it is recommended that these wastes be treated by con-
tinuous oxidative assimilation (Chapter VII).
7. The extended aeration process was shown to be theoretically sound
and a modification designed to allow engineering control of solids
levels by chemical hydrolysis of a portion of the sludge was developed
(Chapter VIII).
8. Two possible flow sheets for the activated sludge process, incor-
porating the kinetic model discussed in Chapter V and the hydrolytic
assist to total oxidation described in Chapter VIII, were presented.
These flow sheets summarize the biological principles and research
findings presented in the report (Chapter IX).
-------
RECOMMENDATIONS
One of the inevitable recommendations resulting from the past year of
data analysis and the past decade of intensive research on activated
sludge processes is that those designing activated sludge processes
should pay greater attention to the biochemical and biological prin-
ciples responsible for the behavior, i.e., the success or the failure,
of the process. It has become apparent to us, and to some other
researchers as well, that the potential for biological processing of
waste organic substrates has barely been tapped, and that such proc-
esses will, as time goes by, become increasingly important for pollu-
tion control and possibly in the future for direct recycle of organic
materials. Also, in the ultimate sense, the rivers, reservoirs,
lakes, and oceans which environmental engineers and scientists must
learn to control are indeed biological reactors. Thus, no matter what
alternatives to biological treatment we may seek for the immediate
future (and there are some depending upon the nature of the organic
matter in a waste water) we eventually must face up to the need for
biological engineering since the environment of which we may claim to
be custodians is a biological one. And thus, as the reader scruti-
nizes this report and the principles, process modifications and
approaches to design outlined herein, he may also employ the report
in conjunction with, or use it as a springboard to, the basic micro-
bial and biochemical literature. It is also recommended that the
report be used as an aid in assessing the research literature in the
environmental field.
Concerning the technological biochemical concepts and process modi-
fications presented in the report and briefly stated in the preceding
conclusions, it is recommended that the next step include larger
scale developmental research relative to the design model herein sug-
gested (Chapter V), as well as pilot plant research or field-scale
testing of the continuous oxidative assimilation process (Chapter
VII). Also, developmental research on the hydrolytically assisted
extended aeration process (Chapter VIII), and pilot plant research on
the total process flow sheets (Chapter IX) would seem a fruitful
extension of the research. Also it is felt that the continued anal-
ysis of the experimental data from which tentative rough guideline
limits for various shock loadings were suggested (Chapter VI) can
considerably refine these guidelines.
In addition to the developmental research on principles and processes
recommended in this report and further conceptual analysis of the
shock load investigations we have already accomplished, there is, in
general, a great need for definitive, fundamental biological research
on the biomechanics of the bio-mass, i.e., heterogeneous or natural
microbial populations, and on the ecology of heterogeneous microbial
systems. Paradoxically, these are areas which have been shunned by
most basic biological scientists as being too applied, and by most
-------
engineers as being too basic and beyond the scope of the engineer's
background knowledge. Clearly, if we are to make successful inroads
toward technological control and protection of the environment, this
situation must be corrected. The recommendation to correct the sit-
uation by performing the needed research is made to those readers
who have a proclaimed professional dedication to pollution control,
particularly to those now beginning their training for this field
and who will be active in it well into the 21st century.
-------
INTRODUCTION
The Perspective
The activated sludge process has been increasingly employed for secondary
treatment of municipal and industrial wastes containing significant
amounts of organic material. The process has many "modifications," and
various flow diagrams and hydraulic regimes (e.g., completely mixed as
compared to plug flow), but its chief distinguishing characteristic is
that the heterogeneous microbial populations (biomass) of which the
sludge is largely composed, are grown in fluidized culture with provi-
sion for maintenance of the organisms in suspension by turbulent mixing
and for introduction into solution of sufficient oxygen for the respira-
tory requirements of the microorganisms. In these characteristics the
process is distinct from other secondary treatment processes, e.g.,
fixed bed processes such as trickling filters.
As with other secondary treatment processes, the objective of the acti-
vated sludge process is the removal of organic compounds from the waste
water. The term "secondary treatment" is used here simply because it is
one which has a specific significance to engineers, and it implies the
traditional employment of activated sludge as a "second line of defense"
regarding removal of organic matter from waste waters; primary treatment
signifies a first line of defense embodying removal of initially settle-
able organic matter in raw waste. The term "tertiary treatment," as
herein employed, is used to signify a third line of defense for removal
of organic matter not removed by secondary treatment. The terms "pri-
mary," "secondary," and "tertiary" are rather meaningless unless one
specifies that they are being employed to describe processes for removal
of biologically usable organic substrates. As employed herein, they
apply only to processes for removing organic waste materials which exert
a "demand" on the dissolved oxygen resource in natural waters into which
waste may be discharged. Even within this specified boundary, one must
impose some qualifications or exceptions as to the subject matter which
will be herein discussed, since it will be readily recalled that, as it
was originally introduced, the so-called extended aeration (total oxida-
tion) process not only combined secondary treatment with the sludge
disposal process, but also combined primary and secondary treatments as
well. Also, it should be recognized that fluidized biological processes
(as well as fixed beds) can be employed to remove inorganic pollutants
such as ammonia, which can cause a drain on the dissolved oxygen
resources in the receiving stream, and that the process may also remove
from waste water some organic matter which does not exert significant
biochemical oxygen demand; for example, cellulose particles, which are
not readily metabolized, may become entrapped in the sludge, etc. At
the outset it should also be recognized that the biological processes
are by no means the only ones which can be considered for removal of
-------
biologically available organic compounds from waste waters. Chemical
coagulation has been employed for years to remove colloidal organic
material, and in sequential combination with other processes, e.g.,
adsorption on activated carbon, many types of soluble organic material
can be removed as well. Such a combination of processes may become a
useful alternative to biological treatment, especially for relatively
low strength municipal waste waters.
The point to be emphasized here is that there is no one universally
recommendable group of unit processes for accomplishment of removal of
organic substrates. Water reclamation plants of the future will not be
describable by any sort of "standard" flow diagrams such as have been
used in the past, but by groups of unit processes joined in such a man-
ner as to accomplish rejuvenation of a specific waste water in accord-
ance with the purity requirements of a specific environmental situation.
Thus, while much stricter standards of effluent quality will undoubtedly
evolve, the required effluent quality will be attained by less_"standard"
arrangements of unit processes to treat the waste waters. It is inno-
vativeness rather than standardization which is needed, and it is inno-
vativeness which is the essence of engineering.
While it is highly desirable to seek alternatives to such biological
processes as the activated sludge process, it may be an even more fruit-
ful endeavor to seek innovativeness within the biological process.
It does seem fair to say that, in many situations, if biological treat-
ment processes could be designed and operated so as to ensure reliable
delivery of the degree of treatment for which they are "designed," much
of the need for treatment beyond the secondary stage might be obviated.
Some engineers, scientists, and technologists in the area of pollution
control would welcome a replacement for biological treatment as a means
of removing organic substrates from used waters. If the reasons for
seeking such replacement involve frustration in attempts to understand
the process, or the feeling that it is, perhaps, a process too compli-
cated to control, these reasons are not at all soundly based. It is
important for us to recognize that engineers will, in fact, be forced to
gain a better understanding of the kinetics and mechanism of biological
treatment systems because, regardless of how we decide to prepare water
for re-use, or regardless of who makes that decision, it must be remem-
bered that our natural receiving bodies are "in-stream" biological
reactors. Thus, if pollution control engineers are to play a leading
role in overall management of water quality control systems, they must
lead in understanding the systems, since understanding is the most
essential ingredient for technological leadership.
The'Technological Gap"
The first aim of researchers in any profession is to gain insight (under-
standing) regarding a natural process or phenomenon. In engineering
practice, the ultimate aim is to employ such understanding to accomplish
a practical goal, to satisfy a social need. When the distance between
-------
understanding and practice becomes large, there is said to be a techno-
logical gap, and it is said that there is currently such a gap between
research and practice in the water pollution control field, especially
with regard to activated sludge. Such gaps are natural; they always
exist, and they are helpful. They are caused as often by the fact that
practice gets out ahead of understanding as by the fact that understand-
ing gets out ahead of practice. Most seem to attribute the cause (blame
or credit) to the somewhat banal belief that engineers are simply
appliers of bodies of known scientific information. In many regards,
this is surely not always the case. For example, often because of a
pressing need of society, the "appliers" must devise processes and
employ them prior to the unveiling of the basic facts which permit full
understanding of the processes. Such was the case with the activated
sludge process. The trouble arises when the apparent success of a proc-
ess engenders a lassitude with regard to seeking understanding and when
theory or hypothesis (a product of the mind), often offered without the
benefit of penetrating experimental proof, becomes technological dogma
which is often overly simplified half-truth.
Thus, while causes for the technological gap cannot be laid on any par-
ticular doorstep, its existence is well recognized. It should be equal-
ly well recognized that researchers and practitioners have a joint
responsibility to lessen (but not to completely close) the gap. Fortun-
ately, its complete closure seems impossible to attain, and this is a
desirable situation because it is the gap which provides the driving
force for progress in the profession. The pressure for progress will
come sometimes from the practicing and sometimes from the researching
segment of the profession. While both groups share the responsibility
for lessening the gap, perhaps the greatest weight of responsibility
resides with researchers. They should not be content with presenting
their results, arguments, conclusions, and recommendations for the scru-
tiny and criticism of other researchers in their own and allied fields,
but should from time to time communicate some of the things they think
they understand (and can defend with sound practical data) to the wider
and ultimately more vital technological audience which includes both
researchers and practitioners, the total investigational level of pro-
fessional activity. It is to this aim which the present communication
is dedicated.
The Nature of the Report
In the following pages we will attempt to bring into focus some major
biological concepts and insights applicable to design and operation of
activated sludge processes and to further understanding of biological
treatment processes in general. This report is not intended to be a
manual of practice for functional design and/or operation of activated
sludge treatment plants. The aim is to transfer some of the concepts
developed through laboratory research investigations to practicing engi-
neers. Whether the concepts become accepted and incorporated into tech-
nological practice is, in large measure, dependent upon how well we
-------
communicate and defend them, and how well they are received by those
who scrutinize them. We present them for the scrutiny of fellow pro-
fessionals (both researchers and practitioners) in the interest of
advancing, but surely not of writing "finis" to, development of con-
cepts in a field of professional activity which is only now emerging.
The ideas and concepts herein presented have been developed with the
aid of a decade of intensive experimental activity and observation
relative to activated sludge processes in our laboratories. The pri-
mary purpose for undertaking the present research project, which is
culminating with the preparation of the following report, was to draw
together, correlate and analyze information obtained from various
individual, but related, research projects which we have conducted
over the past ten years, in an effort to derive practical unifying
biological concepts. We shall not attempt to cover all aspects of
the work in this short report, but we shall draw upon all of this
information in presenting and clarifying "Biological Concepts for
Design and Operation of the Activated Sludge Process."
-------
II
MEASUREMENT OF PURIFICATION
The Concept of Biochemical Oxygen Demand
The most widely employed measurement of effluent purification or effi-
ciency of activated sludge processes is the amount of biochemical oxygen
demanding material (BOD) which has been removed by the process. Bio-
chemical oxygen demand is also employed in assessing the "pollutional
potential" of a raw or treated waste before discharge to a receiving
stream. The assets and liabilities of employment of this parameter of
measurement have been argued for years by engineers, but the fact
remains that the concept of BOD is valid and usable, and certainly it
is used (and misused) in all segments of the water pollution control
field. Today it is nearly impossible and surely unwise to neglect con-
sideration of the idea of biochemical oxygen demand. It seems far bet-
ter for pollution control professionals to clarify understanding of BOD
to help ensure its enlightened application, rather than to join the
chorus decrying use of the concept because of the inadequacy of the BOD
test. It is necessary to make a distinction between the concept and
the test. While we can defend and recommend the concept, we shall not
defend the test for all of its present applications, but will recommend
a more satisfactory measurement for assessing efficiency of treatment
by the activated sludge process.
Through many years of expansion of population and industry, study and
observation have shown that most, if not all, of the many natural and
technological uses to which the water resource is put are seriously
impaired or totally negated if the dissolved oxygen (DO) in surface
waters is seriously depleted. It has also become widely known that the
prime reason for depletion of DO is its utilization in aerobic meta-
bolic processes by microorganisms feeding upon the organic food (i.e.,
carbon source or substrate) contained in municipal wastes and various
industrial wastes discharged to the stream. Thus a conceptual principle
has been established which recommends that the "pollutional load" should
be assessed, not by measuring its amount (i.e., amount of metabolizable
organic matter in the waste) but by estimating the magnitude of its
effect (i.e., the amount of oxygen that will be used because of the
presence of the organic matter). Such a principle is defensible in an
engineering sense, because it goes directly to the heart of the matter,
potential depletion of the DO resource in the receiving stream, and
because it measures a colligative effect of that Organic matter in a
waste water which is readily available as organic carbon source to
microorganisms without requiring determination of either the total
amount or types of that organic matter.
On this basis, the use of the standard BOD test for estimating the
"pdllutional potential" of a waste, either raw or treated, which is to
-------
be discharged to a receiving stream can be defended to some extent.
However, even for this application of the concept, other techniques
may be preferable because of the time required for standard BOD incu-
bation and because of inaccuracies introduced by the dilution tech-
nique itself (1).
The use of the BOD test in measuring purification efficiency cannot be
defended, and it is primarily this application for which we shall
recommend a different approach. It should be remembered that the pur-
pose of biological treatment is removal of the biologically available
organic matter (carbon source) and that any means of measuring the
amount removed during treatment can be employed. The choice of a
method for this measurement should be based upon two criteria: (1) the
method used should actually measure the percentage of metabolizable
(biochemical oxygen-demanding) organic matter which has been removed
by treatment, (2) the method used should be sufficiently rapid so that
results may be available when needed, i.e., while the opportunity
exists to exert operational control over a malfunction leading to
decreased purification efficiency. The standard 5-day BOD dilution
bottle technique meets neither of these requirements (2).
Exertion of Biochemical Oxygen Demand
As stated above, the concept, or principle, of biochemical oxygen demand
is a vitally important one and much research effort in our laboratories
and elsewhere has been devoted to study of the details of BOD exertion.
An examination of the sequence of events which make up the total
course of BOD exertion is pertinent to a discussion of the proper appli-
cation of the concept.
Figure 1 is a generalized plot based on data obtained in many experi-
ments of various types in our laboratory. This figure shows the course
of microbial growth (increase in biological solids, X) which would be
observed if one placed a small inoculum (seed) of microorganisms in a
vessel in the presence of a biologically available source of carbon
(substrate, S) to which the microorganisms were acclimated. The "ves-
sel" might be a BOD bottle, a batch reactor, a reach of stream, or a
pond or lake. The figure also shows a plot of accumulated oxygen
uptake (BOD, y), and a plot of the total amount of organic matter
remaining in solution measured by a general chemical test involving the
digestion of organic matter to C02 and HpO in the presence of potassium
dichromate under acid conditions, i.e., the chemical oxygen demand (COD)
test. The total amount of substrate removed is A COD. This will be
discussed in more detail below.
The type of data diagrammed here can be generated experimentally by
plotting results of analyses of samples taken at various times and
analyzed for biological solids, X (using the membrane filter technique),
and for COD, S, determined on the filtrate. The course of accumulated
10
-------
SUBSTRATE
REMOVAL
"ENDOGENOUS" PHASE
CARBON .SOURCE, S
(COD)
2 UPTAKE, y
BOD)
BIOLOGICAL
SOLIDS.X
TIME'
Figure 1. Generalized plot of substrate concentration, biological
solids concentration, and oxygen utilization during exertion of
biochemical oxygen demand. Circles mark inflection points.
11
-------
0? uptake, BOD exertion or y, can be measured in a number of ways,
efg., by a manometric technique, or if the system is very dilute, by
DO depletion in BOD bottles. Thus we have three very important para-
meters :
1. The amount of biological solids, X, existing in the
system at any time.
2. The amount of organic substrate (COD) in the system
at any time.
3. The amount of oxygen used (y), the BOD expressed at any
time.
As an aid to analysis, we will assume that the system contains no nitri-
fying bacteria (or any other aerobic autotrophs), and therefore all of
the oxygen uptake is due to the respiration of aerobic microorganisms
which require organic carbon (aerobic organotrophs); thus the oxygen
uptake represents only carbonaceous BOD. The experiment is terminated
at the point where 02 uptake may be continuing very slowly but has for
all practical purposes stopped, and therefore all of the carbonaceous
BOD has been expressed. This point along the "BOD curve" has been
known for some time as the so-called ultimate BOD, and is designated by
the familiar symbol, L .
As shown in the figure, Q£ uptake (BOD exertion) goes on long after the
organic matter (COD) has been removed. It is apparent from the plot of
biological solids, X, that Op uptake, after the substrate removal phase
has ended, goes on at the expense of the biological solids which were
produced during that phase. The figure also shows a decided break or
pause in the curve at the termination of substrate removal. This pause
or "plateau" is not always manifested; the Q£ uptake sometimes continues
on a smooth ever-declining curve (see dotted lines). Also, it has been
demonstrated that "plateaus" exist which do not coincide with the ter-
mination of the substrate removal phase (3). However, in most cases
where the plateau is observed, it serves as a "marker" for the termina-
tion of the substrate removal phase, and its significance as such was
first recognized by A. W. Busch (4). Subsequent research by Wilson and
Harrison (5), and Gaudy and his co-workers (6-10) has helped to clarify
understanding of BOD exertion. Much of this work has been brought
together and analyzed in recent literature (2, 10, 11), and its inclu-
sion here would add unnecessary length to this report.
Theoretical COD and BOD
In the research literature one encounters the terms "theoretical COD"
and "theoretical BOD." These are useful terms, but have been subject
to considerable misinterpretation and misuse. The term "theoretical
COD" is synonymous with "calculated COD." It represents the
12
-------
stoichiometric amount of oxygen which would be required to oxidize
(chemically) all of the organic matter to carbon dioxide and water.
This value can be calculated if one knows which organic compound(s)
are being oxidized or if one knows or can estimate the composite empir-
ical formula for this organic matter, and if the concentrations are
known. For example, if one wished to know the theoretical COD of a
sample which contained 200 mg/1 of sucrose (table sugar), it could be
easily calculated from the following balanced equation:
C12H22°11 + 12 °2 >~12 C02 + 1] H2° (1)
(342) (384)
The theoretical COD is 200 (384/342) = 224 mg/1.
Such calculated values are useful as checks against the values obtained
by the COD test, and they can be important for some research investiga-
tions. The term is not a useful one for the engineer dealing solely
with whole waste problems, however, since the determination of an
empirical formula for a whole waste is indeed an exercise in futility
and a wasteful expenditure of investigational efforts. ;
The term "theoretical BOD" is sometimes employed to signify the total
amount of oxygen which would be required to oxidize (biologically) to
C02 and F^O all of the organic substrate which is available to the
organisms as food material if it were possible for organisms to do this.
In accordance with this definition and with reference to Figure 1^ the
theoretical BOD (i.e., theoretical L ) would be equal to A.COD. For
this to be so would require total biological oxidation of the organic
substrate in the waste. Oddly enough, some who have used this defini-
tion for theoretical BOD would also argue that it is theoretically
impossible for total oxidation to occur.. Total oxidation of the orig-
inal organic matter in the waste>would require that all of the biolog-
ical solids produced during the substrate removal phase (Figure 1) were
autodigested during the subsequent "endogenous" phase; i.e., the bio-
logical solids returned to the initial concentration. Such an occur-
rence is not a theoretical impossibility, as will be discussed in a
later chapter, but the chance of its occurrence in any ecosystem in a
short-term experiment (e.g., incubation of BOD's) is not very high.
The degree of autodigestion of the biological solids can be expected to
vary; the relative decrease in the solids level may at times be greater
and at other times be less than that shown in Figure 1. The term
"theoretical BOD," like "theoretical COD," is not a particularly useful
one to incorporate into the practicing jargon of the field; however,
it is important to know what these terms mean and what they imply when
one analyzes research information (a responsibility shared equally by
practitioners and researchers).
13
-------
ACOD as a Parameter of Purification
The curves shown in Figure 1 could also be considered as representing
the course of waste purification rather than BOD exertion, i.e., the
emphasis may be placed upon substrate removal rather than 02 uptake.
If the waste were being purified in an activated sludge aerator, the
initial solids would, of course, be much higher and the kinetics of
substrate removal and solids accumulation would be affected by this dif-
ference, as will be discussed in a later chapter. However, the same
reactions would be involved; carbon source would be removed, biological
solids concentration would increase, and oxygen would be taken up.
In an activated sludge system of any type other than extended aeration,
however, only the reactions at the left of Figure 1 are important. In
the usual activated sludge aerator, the reactions occurring during the
"endogenous" phase (right-hand portion of Figure 1) do not occur or, if
the system approaches plug flow, may occur only briefly toward the exit
end of the aerator, because the solids are separated in the clarifier.
Depending upon the efficiency of flocculation and settling, the effluent
contains the non-metabolized organic material (COD) which remains in the
water, but little of the biochemical products of purification (biolog-
ical solids). At any rate, this is the aim: retention of the sludge
at the treatment plant.
Thus, at the end of the substrate removal phase the plant has accom-
plished its primary purpose, removal of organic carbon source which
would have caused an oxygen uptake (BOD exertion) in the receiving body
if this organic carbon source had not been removed from the waste. The
amount of organic matter removed in the aerator is the difference between
initial (influent) and terminal (effluent) COD values:
CODi - CODg - ACOD ' (2)
The term " ACOD" may be more generally defined as the amount of COD
removed at any time, i.e., the difference between the COD present at the
time of measurement and the COD initially present. When the amount
remaining cannot be further reduced biologically, ACOD is a measure of
the amount of organic matter in the waste sample which was available to
the microorganisms. This is precisely the interpretation which many
people place on BOD but, as can be seen in Figure 1, the BOD of the
waste sample, L0, is not equal to ACOD. It is loosely related to it,
since both result from the presence of the same amount of organic mat-
ter, but the relationship can vary because the measurement of LQ
requires that one consider not just removal of the carbon source, but
the subsequent fate of the biomass which was produced from the organic
carbon source. A significant portion of L0 is exerted after the orig-
inal carbon source of the waste has served as microbial food, and the
14
-------
extent of this oxygen uptake is dependent upon the degree of aerobic
digestion of the biological solids produced. The degree of digestion
is dependent upon many factors, some of which will be discussed in a
later chapter. For now it is sufficient to state that the degree of
aerobic digestion will vary, and because of this, L can be expected
to be more variable than ACOD.
For applications in which BOD is measured (either as BOD5 or L0) and
is interpreted as a measure of usable organic carbon source, it seems
reasonable to use a more direct, rapid, and reliable test for this
parameter, i.e., ACOD. This recommended concept is a simple and
straightforward one, but can lead to some confusion unless it is
clearly delineated that we are not comparing the COD test and the BOD
test. We are concerned here with ACOD, and the COD test is to ACOD
what the DO test is to BOD. It should also be emphasized that inso-
far as testing for the amount of organic carbon source available to
microorganisms is concerned, the COD remaining after the growth
period (i.e., CODe) is not really significant from the standpoint of
assessing the amount of biochemical oxygen-demanding material present.
The determination of ACOD is a measurement of only that portion of
the COD of the waste which is available as biological substrate for
acclimated microorganisms (or which may be otherwise removed due to
the presence of the biomass, e.g., the small amount of colloidal COD
which could be adsorbed on the cell surfaces, etc.). The residual
COD, if sufficient aeration time has been allowed, is composed of
material not utilizable by a microbial population. Thus one can employ
ACOD to assess the amount of biochemical oxygen-demanding organic mat-
ter present, regardless of the presence of some non-biodegradable COD
in the waste. For example, a waste may contain a significant amount
of lignin, which is subject to chemical attack in the COD test but is
not attacked to any extent biologically. Thus COD would be high, but
would not represent a serious stream liability with regard to deple-
tion of dissolved oxygen. It may be a detriment to the stream because
of color, but not because it will be used by aerobic microorganisms at
an appreciable rate. It is important to re-emphasize here that in the
biological treatment process, the concern is primarily with removal of
the carbon source in the waste that is used as "food" for the growth
of microorganisms which are in turn removed by sedimentation (i.e., up
to the time when the substrate removal phase is ended), and that ACOD
measures this purification. On the other hand, in the BOD bottle,
oxygen uptake proceeds well beyond this phase; in fact, if the curves
of Figure 1 are applied to the BOD bottle, the substrate removal phase
ends in one or two days. While it is a simple matter to standardize
the time of incubation of BOD's, e.g., BOD5, the actual rate of 02
uptake is not subject to such arbitrary standardization, and use of
the 'standard 5-day BOD test" as a measure of plant efficiency is there-
fore inferior to the ACOD method.
15
-------
Parameters for Design and Operation
The use of ACOD in design and operation of activated sludge processes
is rather straightforward. In the experimental investigation which
should precede design calculations, one can acclimate a heterogeneous
population of microorganisms (e.g., from a sewage seed) by repetitive
feeding cycles of the organic waste whose treatability is being inves-
tigated. During this time one can determine the extent to which the
organic matter in the waste serves as microbial food (i.e., COD-j -
CODe = ACOD). Through repetitive runs a usable base level for CODe
can be established, and the fraction of the total organic matter which
is amenable to biological attack can be determined. This fraction,
ACOD/COD-j, can be employed as a useful parameter, efficiency of COD
removal. This term is not equivalent to efficiency of purification
since, depending upon the waste water under consideration, CODp may
include some organic matter which does not serve as usable food mater-
ial for microorganisms. Efficiency of treatment, based upon quality
of the effluent in terms of its effect on the receiving stream, would
be much higher than efficiency of COD removal for such a waste.
In day-to-day operation of the process, the efficiency of COD removal
can be used as a guide to assessing performance. Also, the CODe is an
important operational parameter which should be continually compared
to the base line CODg established in the treatability study which was
made in the design stage. This routine check can help detect possible
changes in the nature of the waste between design and operational per-
iods and throughout the life of the plant.
The use of ACOD as the primary operational parameter should do much to
enhance intelligent and careful daily control of the process, since
information on plant efficiency is available immediately rather than 5
days later as it would be if BOD^ is used. If CODe remains relatively
constant, there is little possibility that the actual treatment effi-
ciency, i.e., percent removal of metabolizable organic matter, has
changed significantly. However, the COD test run on the effluent does
not distinguish between metabolizable and non-metabolizable organic
matter remaining in the effluent, and it is therefore absolutely neces-
sary that frequent checks on the metabolizable organic content of the
effluent be made. A hypothetical example may illustrate a possible
operational problem which would not be detected by use of ACOD as the
sole criterion of efficiency. A waste containing 2000 mg/1 COD, of
which 1000 mg/1 was not biologically metabolizable, would be completely
"purified," i.e., would exhibit no further oxygen demand on discharge
if CODe were 1000 mg/1. A ACOD of 1000 mg/1, or a 50% efficiency of
COD removal, would be established as the baseline for 100% removal of
biochemical oxygen-demanding material. However, a subsequent change in
the character of the waste, reducing the content of non-metabolizable
material and proportionately increasing the metabolizable organics,
e.g., 2000 mg/1 total COD including only 500 mg/1 non-metabolizable COD,
might overload the plant. If the same total COD removal were continued,
16
-------
i.e., ACOD = 1000 mg/1, the COD of the effluent would remain the same
but the plant would now be discharging 500 mg/1 of utilizable, BOD-
exerting organic matter, and the actual purification efficiency would
be 1000/1500 = 67%. Such a series of coincidental alterations, yield-
ing exactly the same results for ACOD and CODp is, of course,
extremely unlikely, but such a possibility emphasizes the necessity
for biological testing of the effluent. For this purpose, the stand-
ard BOD test could serve to detect significant changes in purification
efficiency. However, from the standpoint of rapidity of measurement
and of obtaining a more credible estimate of the oxygen demand which
will be exerted by the treated effluent in the receiving stream, meas-
urement of oxygen uptake in an open, stirred jar reactor is preferable
(1, 2). In addition, the plant operating engineer can and should
periodically determine the ACOD of a sample of the waste; i.e., a
sample of aeration tank influent. This can be done in an aerated
batch laboratory reactor using settled activated sludge from the plant
as a source of initial biological solids. Aeration of this reaction
liquor should be continued until the COD of filtered samples becomes
essentially constant. Comparison of the ACOD obtained in the labora-
tory experiment with the current ACOD through the activated sludge
tank allows a direct assessment of the actual purification being
achieved in the treatment plant aeration tank. A record of the values
of ACOD and CODe obtained in these periodic laboratory experiments and
in similar experiments which were made during the treatability studies
preceding design will allow detection of significant changes in the
nature of the waste or in the operational efficiency of the plant.
Design engineers and plant operating engineers can work more as a team
in devising changes and modifications to keep the plant "alive" if
close operational supervision is initiated and maintained on a daily
basis. Such a view may seem somewhat naive to some design engineers
because of past and present experiences both on the design and opera-
tional sides of the coin. However, when one addresses himself to the
future, it is becoming more and more apparent that design and opera-
tional technologies must merge. From the standpoint of prolonging his
career, the rapidity with which pollution control plants are being
designed today should be enough proof to convince any consulting engi-
neer that his consulting capacity should include operational consider-
ations as well as design.
Additional Information
We have in this chapter attempted to present a useful conceptual prin-
ciple for measurement of purification. Much productive thinking and
experimental investigation which amplify this concept have been accom-
plished by various researchers. Some articles in the research liter-
ature which will be useful for further study of the information pre-
sented here are listed in the reference section. Symons, et al. (12)
described the use of ACOD as a means of measuring the efficiency of
treatment, and also outlined some procedures for making treatability
17
-------
studies. The concept of ACOD was also described in an article by
Hiser and Busch (13). More recently, the concept of ACOD as well as
overall consideration of biochemical oxygen demand as applied to the
treatment plant and the DO profile in receiving streams have been
treated extensively by Gaudy (2). Exertion of BOD and methods for the
estimation of assimilating capacity of the receiving stream have been
discussed by Isaacs and Gaudy (14, 15). Jennelle and Gaudy have
studied the effect of dilution of the sample upon exertion of BOD, and
have recommended alternative procedures for measurement of BOD (1).
Summary and Conclusions
As. a parameter for measuring organic loading or the degree of purifi-
cation in treatability studies for the design of biological treatment
facilities, ACOD, i.e., CODi - CODe, represents the most straightfor-
ward estimate of the amount of biologically available organic matter.
Since the BOD test has been used to estimate what ACOD actually meas-
ures, use of the standard BOD test as a design and operational para-
meter does not seem warranted. Furthermore, since L0 (ultimate BOD)
can only approach ACOD as an upper limit, the ACOD parameter gives a
more conservative estimate of the ultimate biochemical oxygen demand of
the waste sample. For daily operation, ACOD is recommended as the
primary Operational parameter for assessing plant efficiency. Periodic
vdeterminations of ACOD of the waste in a laboratory reactor allow
detection of changes in the nature of the waste or in operational
efficiency of the plant.
18
-------
Ill
FUNDAMENTAL STOICHIOMETRY OF BIOLOGICAL TREATMENT
I
"Equation" for Purification
the, stoichiometry of purification is often represented by "pseudo-
chemical" equations in which the organic matter in the waste is shown as
the prime reactant being converted to another form of organic matter
(microorganisms) plus C02 and [-LO. Most engineers are familiar with
this representation:
Omanir Microorganisms
matter + °2 3*" Microorganisms + C02 + H20 (3)
The above equation is a gross simplification of a "chemical" equation
(and of microbial growth); it is only qualitative, since no numerical
coefficients are included, and significant elements have been omitted.
For example, sources of nitrogen and phosphorus and small amounts of
many other elements are needed on the left side of the equation, because
they are part of the microorganisms produced and are needed both in
reactions which form new cellular components and in the reactions by
which the waste is degraded. The organic matter in the waste serves pri-
marily as the carbon source, but depending upon the waste in question, it
may also contain some nitrogen and phosphorus which are metabolically
available to the cells. The microorganisms shown above the arrow are the
biocatalysts required to make the reaction proceed.
The reaction is irreversible, although the microorganisms produced are
themselves organic matter and could conceivably be used as a source of
organic carbon for other microorganisms. Thus, the product can be
recycled through the forward reaction but the process cannot be reversed.
Reversal of the reaction can be brought about with respect to C02 and
HpO as reactants and organic matter (e.g., algae) and 02 as products
tnrough the process of photosynthesis (photo-autotrophic metabolism).
This is essentially a reversal of the purpose of biological treatment,
and it is now being realized that because of this fact the problems one
causes are perhaps equal to or worse than the problems one solves by
attempting to rely on photosynthetic oxygen production in biological
waste water treatment (e.g., quiescent oxidation ponds). The "swapoff"
in organic matter actually conserves organic carbon rather than dissi-
pating it. Conservation of organic matter from waste water sources by
such direct recycle may become important as a source of food supply for
human beings and lower animals in the long range, but in the intermediate
future, at any rate, all modes of disposal of organic carbon in wastes
involve processes which in the main lead ultimately to C02 and H20, which
are recycled eventually through the carbon and the hydrologic cycles.
19
-------
Returning again to the irreversible equation 3, it could be made more
mathematically precise if the elements and their combining weights
(empirical formulas) for "organic matter" and "microorganisms" (or
activated sludge) were known. The chemical composition of the organic
matter in the sample is seldom known to such an extent that it can be
reduced to a representative chemical formula. There are various empir-
ical formulas which describe the approximate elemental composition of
bacterial cells with regard to C, H, 0, and N, and these have proven of
some use to researchers in analyzing data from investigations wherein
readily metabolizable specific compounds of known chemical composition
comprised the organic matter being treated. Such a simple stoichiometric
chemical equation for purification does represent a simplifying prin-
ciple; however, it is not easily applied, as a chemical equation, to
practical concepts for design of treatment processes. The point to be
kept in mind is that the equation represents (as does any other chemical
equation) a mass and/or energy balance, and this principle, as we shall
see shortly, can be usefully employed to provide quantitative information.
Partition of Substrate Between Respiration and Synthesis
Another representation of purification of wastes is often employed and is
perhaps more familiar to water pollution control engineers than is the
equation given above. The diagram shown below actually represents the
same reactions shown in equation 3, but shows separately the two general
types of reactions in which the organic matter participates. That is,
it represents the partition of the waste into (1) energy-yielding, C^-
consuming reactions (respiration), and (2) energy-requiring biosynthetic
reactions (synthesis of new cells).
Organic matter
utilization (respiration)
Sludge production (synthesis)
The portion of the organic matter which is used for respiration is oxi-
dized to C02 and HpO; i.e., this portion of the organic matter accounts
for essentially all of the CL shown on the left-hand side of equation 3
as a reactant and the C02 ana hLO shown on the right-hand side as prod-
ucts. Thus, this diagram might be more easily reconciled with equation
3 in the form shown below.
Organic matter
+ HO (respiration)
New cells (synthesis)
20
-------
The amount of Op which would be required for complete oxidation of the
organic material varies with the nature of the compounds present. One
quantitative measure of the relative amounts of oxygen required is the
respiratory quotient (RQ) which is based upon the relative amounts of
C02 (totally oxidized carbon) which can be produced from different
classes of organic compounds by use of the same amount of 0?. The res-
.piratory quotient, moles C02 produced per mole Q£ used, is 1.0 for car-
bohydrates, 0.8 for proteins, and 0.7 for fats. Thus, for a waste con-
taining primarily carbohydrates, which may be represented as C^O, the
balanced equation for oxidation (RQ =? 1.0) can be written as
CH0 + 0 - 5»-C0 + H0 (+ xATP) (4)
For proteins or fats, a proportionally greater amount of $2 would be
required for that fraction which was oxidized.
Another term, ATP (adenosine tri phosphate) has been introduced on the
right-hand side of equation 4 to represent the amount of chemical energy
which was trapped from that released during the oxidation. There are
other compounds which organisms can use to trap chemical energy, but ATP
is the major one, and the others can be expressed in equivalents of ATP,
in any event. It is this trapped chemical energy which now enables the
microorganisms to use some of the organic matter (Ch^O, as one example)
to make the cellular components needed for new cells:
CHJD + ATP - s—New microorganisms (5)
Thus, the diagram and the equation may be combined:
C00 + H00 + ATP (respiration)
2 2 (6)
New cells (synthesis)
The number of moles of 02 used is approximately equal to the number of
moles of C02 produced (for carbohydrates) and the 02 which is used is
involved in the complex process of trapping some of the energy released
during the oxidation.
Energy Coupling Through ATP
We cannot, in this brief presentation, attempt to explain in simplified
terms the mechanics of the processes of biological energy generation and
utilization, but a few facts concerning the role of ATP should be men-
tioned briefly. The precursor of ATP is ADP (adenosine diphosphate).
The ATP is thus formed by adding inorganic phosphate, iP, to ADP.
21
-------
ADP + IP 5»-ATP (7)
When the free energy which is contained in ATP is used to drive (to
"power") a biochemical reaction requiring energy (a synthesis), the
reaction is reversed and ADP and iP are regenerated. It is because of
this regeneration of ADP and iP that the phosphorus requirement for cell
growth is very low compared to the concentration of carbon source in the
waste, and that phosphorus is said to be reused in the system. The ATP
thus acts as a "shuttle" which couples oxidation of the waste to synthe-
sis of biological solids.
The efficiency with which microbial cells can use ATP to power synthetic
reactions is only approximately known, and it can be expected to vary
from species to species. Therefore, a numerical coefficient could not be
given for the ATP produced in equation 4. It should be remembered that
the reaction by which ATP is made from ADP and iP is a synthesis that
requires energy (which is provided from that energy released by the oxi-
dation of the organic matter in the waste). An oxidation of an organic
compound in the waste is represented by a loss of an electron. This
electron thus "donated" by the carbon source is "accepted" by another
compound which in turn passes it along to another until in the final
transfer, oxygen accepts the electron and water is formed. A certain
amount of dissolved oxygen thus disappears from solution, or one can say
that a certain amount of oxygen uptake has occurred. In the metabolic
process of passing each electron along from each temporary donor-acceptor
system, energy is released in a stepwise fashion and at various points
along the way ATP is produced from ADP and iP. The ATP harvest from each
passage of an electron from carbon source to oxygen is not the same for
all species of microorganisms, nor can it be said with certainty that
each time CU is used ATP is generated; i.e., oxidation is not necessarily
always coupled to ATP generation. Thus the efficiency of the process is
not 100%, and efficiency of ATP generation can be expected to vary some-
what from species to species. For this reason, exact numerical coeffi-
cients cannot be assigned to any of the terms in the equations given
above. This is true even when known pure compounds are metabolized by a
heterogeneous population of microorganisms. For wastes of unknown chem-
ical composition, it is even more futile to attempt to theorize regarding
quantitative relationships to fit the oversimplified equations used to
represent the stoichiometry of treatment. The practical approach to use
of these relationships lies in treatability studies, in small-scale pilot
plants, in which the quantitative estimates may be based upon actual
measurements. Equation 3 then becomes useful as a check on the methods
used for measurement.
Energy Balance
Although equation 3 is not suitable for direct use with the majority of
wastes (because the chemical formulas are unknown), the sum of the
amounts of carbon in the COz and cells produced should balance the amount
-------
of organic carbon which has disappeared from the waste due to the growth
process (except for wastes containing compounds sufficiently volatile to
be stripped in the aerator). Thus, we can make use of the mass and/or
energy balance which the equation (or any other chemical equation) rep-
resents. One approach requires the capture of all of the C02 produced
and use of a carbon analyzer to determine the organic carbon content of
the waste and of the cells produced. A carbon balance can then be made
in which the sum of the carbon produced should be equal to that which
has been used.
However, a more readily facilitated method for determining the stoichio-
metric balance for any specific waste is the energy balance, in which all
components of the balance are expressed in terms of oxygen, i.e., as oxy-
gen uptake and COD. If a certain amount of oxygen would be required to
completely oxidize soluble organic matter in the waste, and only part of
it is oxidized, then the remaining chemical oxygen demand should be equal
to the original minus that which has been expressed as 0- uptake, or:
COD + COD. = 09 uptake + CODQ + COD, (8)
W 1 S L. 6 TS
where CODW, CODis, CODe, and CODfs are the COD's of the waste, the ini-
tial biological solids, the effluent, and the final biological solids,
respectively. Since CODW - CODe = ACOD as defined in Chapter II, and
CODfs ~ cODjs represents the increase in biological solids expressed as
COD, or ACOD, . , ,.0-iids» equation 8 may be written as:
ACOD =02 uptake + ACODbioli SQlids (9)
This balance represents the partition of the substrate between respir-
ation and synthesis, in accordance with the diagram shown previously
and in accordance with Figure 1, Chapter II.
By returning to Figure 1, we may see the stoichiometric balance depicted
graphically. At the end of the purification phase, the amount of or-
ganic matter removed is represented as ACOD in accordance with concepts
presented in Chapter II. The amount of COD which has been oxidized is
represented by the accumulated 02 uptake at this point. This is the
amount of "BOD," y, which has been expressed. All of the so-called
"BOD" has now been removed from solution. The metabolizable organic
matter which has not been oxidized (i.e., expressed as 02 uptake or BOD)
has been incorporated into new biological solids.
By using ACOD as the measure of biologically available organic matter in
the waste, we have expressed this quantity in useful terms; i.e., in
equivalent units of oxygen. Also, D£ uptake is measured directly in such
terms. The increase in biological solids (A cells) can also be expressed
in terms of oxygen equivalents, either by a calculation or by direct
23
-------
measurement. Using an empirical formula for the cells, the 02 required
to oxidize them completely to C02 and H20 could be calculated, thus
yielding a "theoretical COD" of the cells. One such empirical formula
is employed below in a balanced equation for total oxidation of the
cells to C02 and
C5H?N02 + 5 02 -- 2— 5C02
(113) (160)
The ratio of combining weights for cells and 02 is 160/113 = 1.42; thus
each mg of biological solids is equivalent to 1.42 mg 02, or the calcu-
lated COD of the cells = 1.42 x the dry weight of the cells. Alter-
natively, one can measure the COD of the cells directly by employing
the COD test (16).
The mass balance inherent in equation 6, in the diagrams showing par-
tition of substrate, and in the graph of Figure 1, can now be very
usefully employed in treatability studies to determine whether the
measurements are adequate, i.e., whether all of the organic matter
removed from the waste can be accounted for. From equation 9
Accum. 0? uptake + ACOD, , , -, -d
%n T r\r\ £ DlOI. SO II Qb / 1 i \
Recovery = 100 x - - - (11)
All measurements may be expressed as mg/1 .
There are also other ways in which one may approach the calculation of
the materials balance. Various methods, including that described above,
have been discussed by Gaudy, et al . (16). However, the energy balance
method, equation 11, employing the experimental determination of ACOD
of the biological solids is straightforward and readily facilitated and
is applicable to whole organic wastes of complex and largely unknown
chemical composition. Thus there is a relatively easy way to check on
the validity of the substrate partition between that which has been
dissipated as inorganic carbon (C02) in the reactor and that which has
been converted to microorganisms (biological sludge). Since these
organisms are to be separated in the clarifier, the "BOD" has at this
point on the curve of Figure 1 been "purified."
The balance is not absolutely complete, since some of the substrate car-
bon may be physically stripped from the aerator, depending upon the
nature of the original compounds in the waste or the compounds the
microorganisms may make during metabolism (17,18). Also, a small
amount of the C02 produced may be reincorporated into the microorgan-
isms since there are some biochemical syntheses which require a "one-
carbon fragment" derived from C02> but these possible causes for dis-
turbances in the simple proportion described by equation 9 are not
usually of sufficient magnitude to negate the fundamental validity of
24
-------
the principle set forth in the stoichiometric equation or the diagram.
The materials and energy balances they imply and the method of making
such a balance herein recommended provide an engineer with a rela-
tively easy "bookkeeping" procedure to check the validity of experi-
mental data obtained on the waste under study.
If one is willing to assume that none of the carbon in the waste is
unaccounted for, the energy balance given above, equation 9, can be
used to estimate oxygen requirements for metabolism. By measuring
ACOD of the waste and of the sludge, 62 uptake during the purification
phase can be approximated as ACOD minus ACOD of the biological solids.
The sludge yield, Y, is an important parameter which is usually calcu-
lated at the termination of the purification phase. It is best defined
as the ratio of the amount of biological solids produced (Abiol. solids
= final - initial) to ACOD:
Y _ Abiol. solids _ AX ,,9x
ACOD AS u '
The importance of this parameter is obvious in regard to considerations
for subsequent handling of the sludge. The sludge or cell yield will
be considered in greater detail in subsequent chapters.
The materials or energy balance could be applied at any point along the
curve, either before or after the peak in biological solids and level-
ing off of the substrate removal curve. The ratio between that which
has been respired and that which has been channeled into synthesis
(i.e., y/AX) at any time may change; for example, it can be seen that
to the right of the purification phase in Figure 1 the ratio y/AX
increases. However, the balance would still be expected to apply; i.e.,
ACOD is still accountable for largely as 0^ uptake plus the oxygen equi-
valent of the net amount of biological solids synthesis.
Occurrences which follow the purification phase can be identified as
occurrences in the autodigestive or endogenous phase, or the phase of
aerobic digestion, and a term accounting for this phase is sometimes
included in stoichiometric equations. The term "endogenous respira-
tion" is one which has been used quite often, and it has been subjected
to various interpretations. In a strict microbiological sense, it is
employed to signify the metabolism of its internal carbon stores by a
microbial cell. This type of metabolism may proceed concurrently with
metabolism of the external (exogenous) carbon source or after the
exogenous substrate is exhausted, or both, but if measured, rather than
theoretical, values are used in stoichiometric calculations, this
metabolism is included and need not be considered separately.
The most practical way to define "endogenous metabolism" insofar as
biological treatment is concerned, is as that metabolism which occurs
after removal of the primary substrate, i.e., after the purification
25
-------
phase has ended. It represents respiration (C»2 uptake) at the expense
of the accumulated biological solids (see Figure 1); i.e., a net
decrease in biological solids concentration. This decrease may be due
to oxidation of internal carbon resource in individual cells but, more
importantly, with natural populations in waste treatment processes it
represents also cannibal ization of the microorganisms previously syn-
thesized by predator cells (other bacteria or higher microorganisms such
as protozoa, etc.). To the organisms performing the cannibal ization,
the cells on which they are feeding represent an external, not an inter-
nal, carbon source. The stoichiometry of this phase can be written as
a grossly simplified chemical equation in which the organic carbon of
the cells is oxidized to C02 and water:
+ 02 Cannibalizing ^^ + ^ + m1cro0rganisms (13)
Since the cells contain considerable protein which is deaminated during
the autodigestive process, some inorganic nitrogen (NH-) will also
appear on the right side of the equation as well as some inorganic
phosphorus.
In this phase, cells of one type are converted to cells of another type
and to C02 and H20. The similarity of this equation and equation 3 is
obvious, and as stated previously this amounts simply to recycling of
the products through the forward reaction with the result that the net
solids production is decreased and 02 utilization and CO? production
are increased. In short, as the endogenous phase is prolonged, the
partition of the initial substrate between utilization for energy pro-
duction and for synthesis is shifted further toward total utilization
for energy. Whether this reaction can approach completion depends upon
the ecology of the microbial system, i.e., the right feeding organisms
must be present for sequential cannibal ization to occur (proper food
chain). This important aspect will be discussed in a later chapter on
the extended aeration process. It is important here to note that the
stoichiometry of this phase is more variable than that of the purifi-
cation phase; it is less well defined and is, in any event for all but
the extended aeration process, a minor consideration in activated sludge
processes.
It is important next to consider description of the course of growth
during the purification phase to develop useful relationships between
growth and substrate removal, since these are phenomena which proceed
in the activated sludge reactor and accomplish the real purpose of the
activated sludge process.
Summary and Conclusions
Purification proceeds by utilization of the organic matter in the waste
26
-------
for two general types of reactions: (1) oxidatipn, or respiration,
which degrades the organic material to C0£ and HjjO by oxidaitive reac-
tions which produce energy in the form of ATP, (2) synthesis i>f new
cells from the organic material by reactions which require -energy
supplied by ATP. Determination of the fractions of the organic matter
utilized for these two purposes is important in treatability studies
carried out prior to designing treatment facilities. Ati energy bal-
ance based upon expression of organic matter in the waste and the syn-
thesized sludge in equivalents of oxygen and requiring only measure-
ments of COD and of 02 uptake is the most readily facilitated method
for all types of organic wastes.
27
-------
IV
KINETICS OF MICROBIAL GROWTH
In the previous chapter, the growth of microorganisms, or synthesis of
new cells, was treated from a thermodynamic point of view. That is,
the primary concern was determination of the distribution of substrate
between energy-producing and energy-consuming reactions, since this
distribution corresponds to the portion of organic matter which is
effectively removed from the waste (as C02 and HpO) and that which
remains as synthesized sludge requiring further disposal. Because
sludge production is a primary concern in both design and operation of
biological treatment systems, it is important to examine in some detail
the kinetics of microbial growth and the factors which control growth
and the concomitant removal of organic substrates from the waste.
There are two primary relationships between growth and substrate re-
moval . One relationship concerns the amount of growth (i.e., biolog-
ical solids accumulation) which one could expect from the utilization
of a given amount of substrate. This relationship was defined as the
cell or sludge yield in the previous chapter. The second relationship
is one between the rate of growth and the concentration of substrate.
It should be noted that growth rate can be affected by many experimen-
tal or operational factors. When the supply of any required microbial
nutrient is limited, it will become the critical factor determining
rate of growth or total amount of growth, or both. For example, nitro-
gen and phosphorus are well known essential nutrients, and when either
of these is present in insufficient amount it, rather than the organic
carbon source, becomes the growth-limiting component of the waste.
Nitrogen and phosphorus are usually added to the waste in sufficient
amounts so that they are not factors which exert a determinative effect
on growth rate. In short, pollution control engineers usually aim to
make the carbon source the growth-limiting nutrient. This principle
of design is based upon the use of biological treatment primarily for
removal of compounds which exert a biochemical oxygen demand. Consid-
eration of other limiting factors in biological treatment may be neces-
sary to avoid addition of other nutrients to the receiving water, and
this aspect will be discussed in Chapter VII. For most organic wastes,
however, the carbon source is the limiting nutrient and the most
important factor in determining the kinetics of microbial growth dur-
ing the purification phase. The following discussion will therefore
be directed toward the relationship between growth and removal of
organic substrates. First, it is important to define exponential growth.
Exponential Growth
In Figure 1 it may be observed that the course of substrate removal
takes the approximate form of an inverted S-shaped curve and that the
curves for 02 uptake and biological soltds concentration Cwhich depict
29
-------
the "fate" of the removed substrate) also follow S-shaped curves.
The S-curves are not necessarily symmetrical, and in most cases they
are not. For example, the inflection point in the biological solids
(X) curve usually will be located (as shown in Figure 1) somewhat
higher than the mid-point of the curve. The inflection point in the
62 uptake curve may or may not come at about the same relative posi-
tion as that for X. Generally, an inflection point can be observed
in the substrate curve in a position approximately directly above the
one in X. The precise locations of all of these inflection points
are not overly important, but it is emphasized that when such an
experiment as shown in Figure 1 is run in the laboratory or in nature,
both the disappearance of the initial growth resource (in this case,
the carbon source) and the increase in biological solids usually pro-
ceed at an increasing rate, followed by a period of decreasing rate,
until the resource is depleted. That is, in such an experiment as
that shown, one kinetic order does not prevail throughout the course
of the experiment whether one examines substrate removal, 02 uptake,
or cell growth.
The period of increasing rate of growth can very often be described
mathematically as one in which X increases exponentially. Such a
kinetic situation can be expressed in the form of a first order dif-
ferential equation:
= WX (14)
The rate of increase in X, i.e., dX/dt, is proportional to the amount
of X present at any instant. The proportionality constant, u, is
known as the specific growth rate, or the exponential growth rate
constant.
Equation 14 is a useful description of growth, but is more valuable
in its integrated forms:
Xt = XQeyt (15)
or
In Xt - In XQ = ut (16)
From the latter form it can be seen that a plot of In X vs. t would
form a straight line of slope y.
In Figure 2, the biological solids (X) curve of Figure 1, up to the
point of maximum solids accumulation, is replotted on both arithmetic
and semi-logarithmic coordinates. It can be seen that the extent
30
-------
BIOLOGICAL SOLIDS CONCENTRATION, X
_ r\j -fc- — ro w 4^
OiOOO^OOOO
^ ooooooooo
-
-
-
^
L^^
^E>
y
1 f
rs
/
r
,s
^
(PONEN
GROWT
/
f
f
/
'
\
\
>
/
/
/
v
TIAL — ^-j
H
/
/
,
J
1
r
i
) 2 4 6
TIME
i
.
-
-
-
i
8 1C
Figure 2. Arithmetic and semi-logarithmic plots of
microbial growth during the substrate removal phase
in a batch system.
31
-------
(duration) of the exponential phase of growth can be assessed by
noting the extent of the straight line portion of the semi-log plot.
The end of the exponential phase corresponds to the point of inflec-
tion on the arithmetic plot. The numerical value of y can be deter-
mined as the slope of the line in accordance with equation 16, but
is most easily computed by determining from the semi-log plot the
time required for X to double in value (i.e., tj) by extrapolation, if
necessary, since the exponential phase may be shorter than one doub-
ling time. Substituting these values in equation 16, y is defined as:
y = 1n 2 = 0.693 (17)
*d td
It is important to examine the concept of exponential (logarithmic)
growth, since various interpretations and significances have been
assigned to "log" growth by basic microbiologists and by engineers as
well. It is important to remember that, regardless of any particular
physiological significance of this phase of growth, it is unequivo-
cally, by definition, growth which is mathematically describable by
equations 14 through 17. It is that growth for which the doubling
time t^, is constant regardless of the magnitude of tj or of y . It
is that period of growth for which one's data will be described by a
straight line on a semi-logarithmic plot of X vs. t. In short, log-
arithmic or exponential growth is simply a mathematical description,
and the symbol y is a proportionality factor which is empirically
determined from experimental data. The importance of the mathematical
definition of exponential growth will become apparent when continuous
flow cultures are considered (Chapter V).
Relationship Between y and S
We have defined y simply as a mathematical descriptor which is a rate
constant allowing one to assign a numerical value to the rate of
increase in X when the doubling time for X is constant. There are
some workers in both the fields of microbiology and biological engi-
neering who maintain that the doubling time cannot be constant, and
therefore logarithmic or exponential growth cannot exist, unless the
substrate, and all other required nutrients, are present in excess
concentration (e.g., see reference 19). This is an old concept which
has been passed along as an axiom for so many years that it has become
dogma. We state here that the concept is false, and we will not bur-
den the reader with all of the proven reasons for our statement. A
more detailed discussion is provided in the research literature (20,
21). We will state here simply that the weight of experimental data
indicates that exponential growth identifiable by specific values for
y , its descriptor, can occur at concentrations of substrate substan-
tially below those for which y approaches some maximum value at which
addition of higher concentrations of substrate causes no further
increase in y.
32
-------
This statement can be verified by a relatively simple experiment. If
one were to run a series of growth experiments concurrently in separ-
ate reaction flasks into which he placed small inocula of cells accli-
mated to a particular substrate, and in each reaction flask he had
placed increasing concentrations of the substrate covering a wide
range of concentrations, he would observe growth curves (arithmetic
plots of X vs. t) which would be different in two respects. First, he
would find that the more substrate he had in the flask, the more total
growth (i.e., higher concentration of X) he would have at the end of
the substrate removal phase, i.e., the increase in biological solids
should vary with initial substrate and the proportionality factor, Y,
should be constant. He would also observe that the rate at which X
attained its final value would be lower for the systems in which he
had placed the lower substrate concentrations. He would also observe
that the growth curves would become more alike and would become prac-
tically superimposable for the flasks into which he had placed higher
and higher substrate concentrations.
If one now plotted the growth curves on semi-logarithmic coordinates
(i.e., In X vs. t), the data for the period up to the inflection point
on the arithmetic curves would plot as straight lines, the slope
increasing with increasing initial substrate concentration, until a
maximum slope was attained. Such an hypothetical experiment is shown
in Figure 3. From this figure, one observes an apparent dependence of
the value of y on the initial substrate concentration in which the y
was generated. The mathematical description of this apparent rela-
tionship may be examined initially by plotting the values of y against
the corresponding values of initial substrate concentration, S . Such
a plot is shown in the upper portion of Figure 4. It is seen that as
the initial substrate concentration approaches higher and higher
values, y approaches a maximum value ( Vpiax^ anc' father increases in
initial substrate concentration cause no further increase in y; i.e.,
the value for y eventually becomes independent of substrate concentra-
tion.
It has been observed that most growth data of this type can be satis-
factorily fitted with a hyperbolic function, as given below:
- "max
This equation (the "Monod equation") has appeared in innumerable places
in both the microbiological and the pollution control literature with
the symbol S instead of S0, and such "familiarity" is in large measure
responsible for the varying "theoretical" interpretations which have
been placed upon equation 18. These aspects are described elsewhere
(20, 21). It is important here to note that the equation is a useful
relationship which was derived empirically and is simply an analytical
33
-------
to
Q
n
o
to
o
t5
o
o
00
300
250|—
200
30O
TIME
Figure 3. Effect of initial substrate concen-
tration, S0, on the rate and total amount of
microbial growth. The vertical lines indicate
the end of the exponential growth phase.
34
-------
160
X
^
^
00
80
40
0
So _ So KS
LL Mrr,nv Mrr,«v
' ' max /max
A
/A
J ^max
/
/
/
/
&
/c
^
r
v\
-SLC
max
Ks
/
)PE
Pr
/
'
1
nax
= 0.6 hr~'
= 125m
g/&
,
'
0
8
S0 x 10'
-2
10
Figure 4. Hyperbolic plot of the relationship
between specific growth rate, y, and initial
substrate concentration, S0 (upper graph), and
a straight line plot, S0/y vs. S0, of the same
data (lower graph).
35
-------
expression which has been found to fit a large amount of experimental
data. Like ymax, the term Ks is a constant in the equation. Whereas
ymax represents the constant value of y which is not exceeded in the
presence of higher substrate concentrations, Ks (the "saturation" con-
stant) determines the sharpness of curvature of a plot of V- vs S0-
The higher the value of Ks, the flatter will be the curve and the more
slowly will y become asymptotic to ymax as substrate concentration is
increased.
This behavior of the expression is shown in Figure 5 (taken from ref-
erence 22) in which are plotted curves for y calculated from equation
18 for a range of values of SQ. For each curve, the indicated value
was used for Ks and ymax for all curves was 0.4 hr"1. It can be seen
from the figure and equation 18 that if K$ were infinitely small, there
would be very little dependence of y on S, and for all practical pur-
poses only one exponential growth rate constant (umax) would describe
growth. When Ks is very small, the curve rises very sharply and breaks
to the right very sharply. In this instance, y can be said to be very
sensitive to a small change in substrate concentration at very low
concentrations only. For systems which exhibit a very low value of KS,
the sharply rising portion of the curve could possibly be fitted to a
straight line which intersected the horizontal asymptote, ymax- How-
ever, there are ample data in the literature (22-26) to indicate that
the relationship between specific growth rate and substrate concen-
tration is such that it can be most easily fitted to the continuous
hyperbolic function, equation 18.
It can be seen from equation 18 that the substrate concentration at
which yis equal to half the value of yma is the numerical value of
1C. Thus from a plot of the data (y vs. S ) one could estimate ymax,
then locate the intersection of 0.5 y with the curve and read tne KS
value on the abscissa (see Figure 4). A better procedure is to arrange
equation 18 in one of its straight line forms. Two of these forms are
given below:
ir = ir- + ii—V 09)
max max o
S S K
IT = y-9" + y~^~ (2°)
max max
In the lower portion of Figure 4, the y and S0 values shown in the upper
portion are plotted in accordance with equation 20. These forms of
equation 18 simply allow one to make a more accurate estimate of the
values of the kinetic "constants" y and Ks because one usually has
only a small number of values of y in the low substrate ranges, and by
using a straight line form one is not required to make a graphical
estimate of curvature.
36
-------
-------
Equation 18 has been employed by biochemical (fermentation) engineers
and microbiologists, most of whom deal with pure cultures of micro-
organisms. For any specific pure culture, ymax» Ks> and Y are con"
stants for that organism under constant cultural conditions. In the
activated sludge process, natural microbial populations exhibiting a
rather high degree of heterogeneity and variability are used (anything
but "pure" cultures). Values for the kinetic constants have been
shown with pure cultures to vary with species, substrates, tempera-
ture, pH, etc.; therefore, in the pollution control field one must
evaluate these constants in repetitive experimentation in order to
determine if there is a reasonably narrow range of variation for engi-
neering use. Also, there could be some valid doubt as to whether
equation 18 is the most logical one to employ for relating y and SQ for
systems such as activated sludge.
For these reasons, we carried out many experiments over a considerable
period of time using different heterogeneous populations developed from
various sewage seeds. All experiments were run at 25 ± 2°C and at
neutral pH. A synthetic waste was used in all experiments, and a num-
ber of different carbon sources of different types were used as sub-
strates. However, in order to determine the variability in the con-
stants due only to differences in the populations used in different
experiments, a large number of repetitive runs were made with a single
substrate. Glucose was used in these experiments, since it is easily
metabolized by most bacteria and should allow the greatest heterogen-
eity possible in a microbial population growing on a single organic
compound. Values for ymax and Ks determined in twenty-two separate
series of growth experiments using glucose as carbon source are shown
in Table I.
As these data show, even under constant conditions and with a simple
synthetic waste, there is no single value which can be used as "the"
kinetic constant for a heterogeneous population. Values obtained for
these constants with other substrates were also variable (26), but
only a few experiments were made with each and the range obtained with
glucose is probably more representative of the range to be expected
with heterogeneous populations at moderate temperatures and neutral pH.
Based upon the values given in Table I and upon other experiments, it
would appear that ranges of 0.4 to 0.6 hr-1 for ymax and 50 to 125 mg/1
for Kg might be chosen as representing reasonable estimates for use in
applying these kinetic constants to heterogeneous populations such as
activated sludge.
In all of these studies in which ym,x and K$ were evaluated for differ-
ent heterogeneous populations, the fit of tne data to the hyperbolic
equation of Monod (equation 18) was also determined. This equation
(using initial substrate concentration, S0) was found to provide a
more satisfactory fit of the experimental data than did other equa-
tions and relationships tested with the same data (22, 27).
Municipal sewage represents, perhaps, one of the most complex mixtures
38
-------
TABLE I
VALUES OF y AND K FOR HETEROGENEOUS POPULATIONS OF SEWAGE ORIGIN
3 S GROWING ON GLUCOSE.
u
max
hr-1
0.49
0.38 :
0.42
0.38
0.59
0.72
0.56
0.77
0.60
0.53
0.44
Ks
mg/1
29
. 11
68
87
91
145
97
181
116
30
37
y
max
h^1
0.40
,0.72
5.50
0.45
0.63
0.63
0.48
0.40
0.74
0.31
' 0.46
Ks
mg/1
93
105
108
18
13
58
52
58
111
23
85
1
Averages: ym,v = 0.53 hr , K = 74 mg/1.
IN a X S
39
-------
of carbon sources found in nature. It is such a weak substrate that it
is difficult to use in growth experiments requiring a gradation of
initial substrate concentration. Therefore, the soluble portion of
a municipal sewage was concentrated in a flash evaporator, and growth
on various dilutions of this carbon source was studied (26). Values of
y were determined from semi-logarithmic plots of optical density vs.
time, and values for one such experiment are plotted vs. the correspond-
ing substrate concentrations (measured as COD) in Figure 6. Average
values for sewage were umax = 0.46 hr"' and K§ = 52 mg/1. These are
in general accord with the range previously given for glucose. Data
such as these provide some assurance that these relationships indeed do
exist for complex populations and for complex substrates, and that there
is a reasonably narrow range of expected values.
In the pure culture area, very little experimentation regarding the
curvature factor (Ks) of equation 18 has been done. It is generally
believed that the value of Ks will be much below 50 mg/1 of substrate,
largely because in the early work by Monod (23) low values were found.
It is often overlooked that these studies involved only two bacterial
species and that other early studies yielded a very high value for K$
(24). These aspects and their ramifications for use of equation 18 in
research need not be dealt with here but are covered elsewhere (20,
21). However, before passing on to another important "kinetic con-
stant" it is appropriate to mention that the analytical geometry of
equation 18 is the same as that of the Langmuir adsorption isotherm
and that of the Michaelis-Menten equation for dependence of enzyme
velocity on substrate concentration. It is necessary to point out that
similarity in kinetic form does not imply similar mechanism.
The point to be emphasized here is that equation 18 is usable as a
kinetic description for determining the effect of initial substrate
concentration on y. It is an entirely empirical one which can be
recommended for use because it provides a good fit to growth data. It
will be used in a later chapter in the development of a kinetic model
for describing biological behavior in a continuous flow reactor, i.e.,
an activated sludge tank.
Sludge Yield
In the previous chapter, sludge yield was defined as the amount of bio-
logical solids produced per unit of organic material used. Within the
context of the mass balance equation, it is therefore most commonly
determined at the completion of the reactions, i.e., at the point of
maximum biological solids accumulation, which normally corresponds to
the point of termination of substrate removal. It is important to note
that while its determination at the end of the substrate removal phase
is valid, the factor Y is applicable throughout the growth curve, i.e.,
during the exponential and the declining phases of growth. Thus Y is
useful as a kinetic constant, and we can write
40
-------
0.50
0.40
0.30
0.20
0.10
0
SEWAGE
Pm= 0.46 h
Ks = 55
O
200 400 600 800
S0(mg/;COD)
1000
Figure 6. Hyperbolic plot of the relationship between
specific growth rate, y, and initial substrate concen-
tration, S0, for a heterogeneous microbial population
of sewage origin growing on a concentrate prepared
from the soluble portion of municipal sewage (26).
41
-------
In large measure, the above statement is true and can be shown to
apply to heterogeneous microbial populations. The relative constancy
of sludge yield throughout the substrate removal period is demonstrated
in Figure 7. These results were obtained in an experiment in which
glycerol at relatively high initial concentration was fed to a hetero-
geneous population at low initial solids concentration (28). Biolog-
ical solids and filtrate COD concentrations were measured and consti-
tute the primary (raw) data shown in Figure 7A. The substrate removal
curve was constructed by subtracting the COD curve from the initial
COD. In Figure 7B, this curve and the biological solids curve are
plotted on semi-logarithmic coordinates, and in Figure 7C, the sub-
strate removal curve is plotted vs. the biological solids curve. It
is apparent that the ratio of solids produced to substrate removed,
i.e., AX/AS, is relatively constant throughout the substrate removal
period. The calculated correlation coefficient, r, is 0.975. The
sludge yield determined in the usual way at the end of the substrate
removal period is thus also applicable to the exponential phase of
growth; i.e., Y is a constant during any particular experiment.
It should be emphasized that this constancy of Y does not imply that
one would obtain precisely the same value for the next experiment.
Sludge yields, likeM.,,^ ancl KS, can be expected to vary, depending
largely on the species of cells present and upon the substrate. The
substrates one encounters in waste waters are, except for certain
highly specific industrial wastes, not so selective that one could
expect the same species to predominate all of the time; therefore, Y
will vary not only from waste to waste (substrate to substrate) but
even for the same substrate at different times, and we must seek to
determine a range of expectable values. Thus it is important to
orient our thinking, not to precise constants based upon theoretical
reasoning (energy trapping, thermodynamics, energetics, etc.), but to
a probable range of values which will be observed for naturally
developed populations. This is, incidentally, the type of thinking
which engineers should be readily able to perceive. It may be more
convenient to use one value, but it is unrealistic to do so for the
natural biological populations with which pollution control engineers
must deal. For an engineer to accept a certain value for a biological
"constant" to two significant places is as much of a mistake on his
part as is the insistence on the part of some basic microbiologists
that valid conclusions can be drawn only from studies in which pure
cultures were used. The engineering mechanics of the bio-mass, which
is indeed a rather new engineering material, are just now beginning
to evolve. At this stage of the development of this vital engineer-
ing field, one cannot afford to replace empiricism and experimentation
with rather weakly based simplifying theory. The importance of micro-
bial ecology and the fact that shifts in predominance continually
42
-------
2400,
(COD) SUBSTRATE
REMOVED (S,
12 16
TIME, hrs.
10,000
o
UJ
1000
a:
<
a.
100'-
f^^
~\§J~
(COD) SUBSTRATE
REMOVED (Sr)
p - 0 3^ hr~ '
/
Q
8 A
.A I
/
A-
/ ,
/ /
J /
y
7
/n
//.
/
/
jf
^-BIOLOG
SOLID
1
D
A
ICAL
5{X)
~> hr'1
12
16
TIME, hrs
1200
1000
. 800
c/r
o
o
<" 600
©
END LOG—•
GROWTH PHASE
g 400
O
CD
200
Y X/Sr -- 0 48
r 0.975
'0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
(COD) SUBSTRATE REMOVED , mg/M
Figure 7. Constancy of sludge yield throughout the sub-
strate removal period during growth on glycerol of a
heterogeneous microbial population of municipal sewage
origin (28).
43
-------
occur, mean that there is an inbuilt variability in the properties of
the bio-mass, and the engineering profession has to gain some defin-
itive insight regarding the range of variability and its effect on
working theories for design and operation before it can embrace any
theory.
Statistical ranges for cell yields on a variety of carbon sources are
given in Table II. For more in-depth discussion of cell yield in het-
erogeneous microbial populations, the reader is referred to recent
articles in which values obtained in our laboratories during the last
decade are summarized (28, 29). It is important to note that most of
the sole carbon sources shown in Table II do not exert a very selec-
tive effect on the population. Many types of organisms can metabolize
these compounds. All of the organisms employed in the many studies
for which Y was determined were grown up (acclimated) on the compound
tested from original sewage seeds all taken from the same municipal
sewage treatment plant (Stillwater, Oklahoma). All values were obtain-
ed under the same operational (experimental) conditions. All values
were measured in the same manner, at the end of the substrate removal
period (see Figure 1). In cases where eight or more experiments were
run, the standard deviation, coefficient of variance (CV), and 95%
confidence limit (CL) were calculated. For example, the average
yield, 7, in twelve experiments employing lactose was 47.1%. The
lowest yield obtained was 30, and the highest, 61%. The standard
deviation of the mean was 7.5%, i.e., two-thirds of the Y values were
between 39.6 (47.1 - 7.5) and 54.6 (47.1 + 7.5). Expressed as a per-
centage of the mean Y value (i.e., expressed as the coefficient of
variance, CV), two-thirds of the values fell within ± 15.9% of the
mean Y of 47.1, and there is a 95% chance that if the experiments were
run again, the ? would fall between 42.4 and 51.8% (95% CL).
The fact that a wide range of values can be obtained for any given sub-
strate attests to the variability of the bio-mass and to the engineer-
ing need to work within a range of reasonably expected values, i.e., a
statistical range. Data such as these and other data as well (28, 29)
indicate that a reasonable range of ? for engineering calculation is
40 to 60% for most carbon sources consisting largely of sugars and
sugar alcohols and combinations thereof.
The cell yield can also be obtained from continuous flow reactor data
and although we have detected some differences in average values (27,
29) for batch and continuous flow systems, most fall within the range
cited above. Also, while there is some evidence that reactor deten-
tion time in continuous flow processes may exert some effect on Y
(i.e., somewhat lower values may be observed at very high detention
times), the data available indicate that at reactor detention times
below 24 hours, the effect is not readily observable.
44
-------
TABLE II
STATISTICAL SUMMARY OF SLUDGE YIELD VALUES FOR HETEROGENEOUS
POPULATIONS OF SEWAGE ORIGIN GROWN ON VARIOUS CARBON SOURCES
Carbon Source
Arabinose
Cellobiose
Fructose
Galactose
Glucose
Glycerol
Lactose
Maltose
Mannose
Ribose
Sorbitol
Sorbose
Sucrose
Xylose
?
44.8
50.3
53.0
51.9
61.9
46.5
47.1
51.7
52.2
45.7
50.6
52.0
53.1
50.4
n
6
4
8
24
118
31
12
7
8
9
39
12
12
18
Range
32-51
34-61
34-69
36-76
36-88
31-61
30-61
39-86
36-70
36-56
35-59
23-73
33-77
25-71
S
-
-
10.4
11.3
12.5
9.4
7.5
-
13.4
6.4
6.0
16.7
14.1
11 .4
CV
-
-
19.6
21.8
20.1
20.2
15.9
-
25.6
14.0
11.8
32.1
26.5
22.6
CL
-
-
44.4-61.6
47.1-56.7
59.6-64.2
43.1-49.9
42.4-51.8
-
41.1-63.3
40.8-50.8
48.7-58.5
41.4-62.6
44.2-62.0
44.8-56.0
Y is expressed as percent substrate used for synthesis, i.e.,
100 x AX/AS.
45
-------
Summary and Conclusions
In this chapter we have considered the relationship between substrate
removal, the desirable result of biological treatment, and microbial
growth, which accompanies substrate removal . Each species of micro-
organism is characterized by an absolute maximum growth rate which can
be achieved under optimum conditions with all nutrients in excess.
Under these conditions, growth is exponential and is described mathe-
matically as
g-=yX (14)
At maximum growth rate, y = y .
ffluX
Growth rate may be limited at values below y,,,^ by sub-optimal con-
ditions of temperature, pH, aeration, etc., or by limiting concentra-
tion of any required nutrient. Since the objective of biological
treatment is removal of the organic carbon source with the greatest
possible efficiency, the process is normally designed so that the car-
bon source is the limiting nutrient, e.g., by addition of nitrogen and
phosphorus, if necessary. Therefore, the specific growth rate, y,
becomes a function of substrate concentration, i.e., y = f(S). With
substrate concentrations below those which allow growth at the maximum
rate for the cultural conditions used (pH, temperature, etc.), expon-
ential growth occurs, i.e., growth which plots as a straight line on
semi -logarithmic coordinates. The proportionality constants (y) for
exponential growth can be evaluated from the slopes of such plots of X
vs. t for various initial substrate concentrations using equation 17
07)
where t. is the time required for a doubling in mass. A plot of these
values of y vs. initial substrate concentration for heterogeneous
microbial populations has been shown to fit a hyperbolic function, the
"Monod equation" (equation 18).
o
+ S
s o
Evaluation of the kinetic constants, y and K , may be made using
one of the straight-line forms of the hyperbolic equation (e.g.,
equation 19 or 20).
46
-------
For heterogeneous populations, affected by ecological shifts which
cause changes in predominating species, precise values for the kin-
etic "constants" cannot be expected. However, repeated determina-
tions of these constants for populations selected from municipal sew-
age have demonstrated that a usable range of values can be defined
for such populations. For heterogeneous populations of sewage origin,
a reasonable range for ]J (at neutral pH and 25 + 2°C) appears to
lie between 0.4 and 0.6 nr-1 and KS values are generally between 50
and 125 mg/1 . The fact that values within the same range were found
for growth on the soluble portion of municipal sewage lends added cre-
dence to these results. Values outside these ranges may be observed,
but the ranges appear adequate to cover most situations in which a
readily metabolized carbon source is the limiting nutrient for natural
populations at approximately neutral pH and moderate temperatures.
The rate of growth, dX/dt, is also related to substrate concentration
by the following expression:
_ .
dt dt
The kinetic "constant," sludge (cell) yield, Y, applies in general
throughout the growth cycle and substrate removal period, i.e., the
exponential, or logarithmic, phase and the declining phase of growth.
Its determination is usually most easily facilitated by determining
the amount of biological solids produced and substrate (COD) removed
at the end of the substrate removal period:
Y=§ (12)
Generally, expectable values for Y will lie between 0.40 and 0.60
(40-60%).
Thus, during exponential growth, both the rate of accumulation of
solids and the total amount of biological solids synthesized are
relatable to the concentration of carbon source in definable fashion,
which can be quantitatively estimated by three biological systems
constants, Vmax> Ks> and Y.
In the following chapter, these biological kinetic constants which are
determinable in batch systems, are to be related to physical and
hydraulic characteristics of continuous flow systems, e.g., activated
sludge.
47
-------
KINETICS OF COMPLETELY MIXED REACTORS
It is a well known fact that throughout much of its history, the acti-
vated sludge tank has been designed as a long, narrow tubular reactor
through which the waste passes under aeration. The reactor has usually
been considered as a plug, or piston, flow system into which a waste
containing a relatively high concentration of organic carbon enters,
and from which a low concentration of organic carbon exits at the
effluent end. Thus X and S are envisioned as changing, X increasing
and S decreasing, from influent to effluent ends.
It has long been known also that this does not actually happen, because
such perfect piston flow is extremely difficult to obtain. The longer
and narrower the tank, and the less turbulent the mixing, the closer
will such a hydraulic regime be approached. It is roughly approximated
in some rivers. It is difficult to attain in laboratory growth react-
ors, and seldom, if ever, does it obtain in activated sludge tanks.
At the other end of the spectrum of hydraulic regimes which are theo-
retically possible in biological reactors is one in which the inflowing
material is instantaneously and homogeneously, i.e., completely, mixed
with the liquid and suspended particles in the reactor. While perfect
plug flow is very difficult to attain, conditions approaching perfect
complete mixing are less difficult to achieve. Since even activated
sludge tanks which are not designed as such are in the main closer to
complete mixing than to plug flow, the kinetic aspects of the former
are of most critical importance in considering the activated sludge
process.
We can say with considerable certainty that growth of heterogeneous
microbial populations in batch systems does, in general, follow the
"rules" covered in the previous chapters. In this chapter we will
examine the behavior of microbial populations in a completely mixed
reactor under continuous flow conditions.
Criteria for Complete Mixing
A simple reactor of the continuous flow, once-through type is diagram-
med in Figure 8. The batch reactor has been converted to continuous
flow by the continuous pumping of medium into the tank at a constant
rate of flow, F, and concurrent withdrawal by overflow at the same
rate so that the volume of mixed liquor, V, is constant. The inflowing
concentration of carbon source is S-j. The concentrations of substrate
and cells in the reactor and in the reactor effluent are S and X,
respectively. If the reactor is completely mixed, the concentrations
of substrate and cells are the same in all parts of the reactor;
therefore, these concentrations in the mixed liquor exiting the reactor
49
-------
AERATION
TANK
F
^
Si
S
X
V
F
S
X
D = F/V = l/t
Figure 8. Flow diagram for a continuous flow, com-
pletely mixed reactor of the once-through
type.
are also the same. Also, if the reactor is completely mixed, the
incoming substrate is "instantaneously" mixed; i.e., diluted in the
reactor, the dilution factor being the ratio of the volumetric rate
of inflow, F, to the volume of reaction liquor in the aeration tank,
V. Thus the dilution factor, or_the dilution rate, D, is equal to
F/V, and D is the reciprocal of t, the mean residence time (detention
time) in the reactor. This hydraulic parameter is an extremely impor-
tant one, as will be shown below.
Before the Mnetics of completely mixed systems are discussed, it is
necessary to consider ways to ascertain whether a reactor is, or is
close to being, completely mixed. The criterion of the sameness of X
(biological solids concentration) in the reactor and the reactor
effluent can be checked (approximated) by measuring the transmission
of light (OD, optical density) by the mixed liquor in the reactor and
in the reactor effluent. The optical density provides a quick estimate
of biological solids concentration. In cases where X is so high that
50
-------
correlation between optical density and biological solids is not pos-
sible, the samples of reactor mixed liquor and effluent can be equally
diluted before the comparison is made. Also, for best results the floe
must be vigorously dispersed before reading optical density. The cri-
terion of complete mixing with respect to substrate could be checked by
sampling at the same places, but this is more time-consuming and, in
any event, if the mixing in the reactor is vigorous enough to satisfy
the criterion with respect to suspended particles (X), it can be
assumed that the criterion will also be satisfied with respect to sub-
strate molecules (S).
Complete and instantaneous mixing characteristics of the reactor can
also be checked by performing either dilute-in or dilute-out (washout)
studies. A washout study could be performed by filling the reactor
tank to the operating volume with water containing a known concentra-
tion, C0, of dye (for example), or suspended particles (e.g., clay).
If clear water is now pumped into the tank and the concentration in the
effluent is measured, it can be seen that, if complete mixing obtains,
the concentration will be continually decreased (diluted) by the factor
F/V, i.e., D. Thus we may write
£ - -DC (22)
or, after integration:
Ct - Coe- (23)
If, on the other hand, the reactor had been filled with clear water
and at zero time water containing dye, clay, or another marker material
at C0 had been pumped into the tank, the concentration Cj. in the efflu-
ent at any time thereafter would be given by the dilute-in equation:
Ct - CQ(1 - e~Dt) (24)
Theoretical dilute-out and dilute-in curves calculated for a dilution
rate of 0.125 hr~' are shown in Figure 9.
A comparison of theoretical and observed dilute- in curves obtained for
one of the laboratory-scale continuous flow reactors employed in
research in our laboratories is shown in Figure 10. Experimental val-
ues for two dilution rates, 1/4 and 1/8 hr , are shown. Circular,
well-agitated tanks of considerable depth in relation to surface area
appear to possess the geometry favorable to complete mixing.
51
-------
Ul
to
0
0
8 10
TIME, hr.
14
16
Figure 9. Theoretical dilute-in and dilute-out curves for a completely
mixed reactor calculated for a dilution rate of 0.125 hr-1.
-------
100
8-HOURJ
HEORETICAL EQUATION
= c0(.-e-D<)
THEORETICAL VALUES
EXPERIMENTAL VALUES a a 4-HOUR
A A 8-HOUR
6 8 10
TIME, HOURS
14
Figure 10. Comparison of theoretical and experimental
dilute-in curves for a completely mixed laboratory
reactor operated at various detention times (41).
53
-------
Growth In the "Steady State"
Growth under continuous flow conditions may be studied in a reaction
vessel fulfilling the requirements of complete mixing as outlined above,
i.e., growth in a reactor such as shown in Figure 8. To start the sys-
tem, one may fill the reactor with waste, add seed from a source such
as sewage, and allow the cells to grow under batch conditions. The
pumps may be started approximately when the solids have reached the
maximum level (i.e., at the end of the substrate removal phase) or be-
fore. There are a number of ways to initiate continuous flow opera-
tion. The essential fact is that the system will, after a time, attain
a condition (provided F is not exceedingly high or low) wherein values
for X and S will remain essentially constant. The concentrations
attained will be close to those exhibited in batch systems at the end
of the substrate removal phase in instances in which SQ in the batch
reactor was equal to S-j in the continuous flow reactor. However,
unlike the batch system (Figure 1), the biological solids concentra-
tion, X, would not decrease below this point; it would remain relative-
ly steady, as would S. A continual feed stream allows the cells dis-
placed from the reactor (by the factor F/V or D) to be replenished and
X remains approximately constant. The substrate concentration remains
approximately constant because cell yield, Y, remains approximately
constant. In general, then, the system "approaches" a steady state
with respect to X and S.
It is not necessary to ascribe a physiological significance to the
term "steady state" since, like exponential growth, it may be defined
on a strictly mathematical basis; i.e., dX/dt = 0, and dS/dt = 0. It
has been experimentally demonstrated for pure cultures, where there is
no chance for change in the predominating species except by mutation
and, therefore, no change in the biological constants, that relative
steadiness in X and S can be attained. Also, after considerable expen-
diture of experimental effort (22), it has been shown that a "pseudo"
steadiness in X and S can be observed with heterogeneous populations
of microorganisms, where there is much chance for changes in the pre-
dominating species, and therefore opportunity for change in the bio-
logical constants.
Having established the possibility of steady state growth experimen-
tally, we may consider the mathematics of such growth. In Chapter IV
it was stated that equation 14 describes the behavior of X in an expon-
ential phase of growth characterized by a constant proportionality
factor, y, the specific growth rate. We can determine whether, in the
steady state continuous growth situation, the cells are in the expon-
ential phase by removing this constraint from v as we have defined it
and by assuming that the specific growth rate, u, is dX/dt)(l/X) in
any phase of growth, exponential or otherwise. Considering X in the
reactor, there is a continual pressure for biological solids to
increase due to growth (dX/dt = uX) and to decrease due to cells being
diluted out of the reactor (dX/dt = -DX). Since the total weight of
54
-------
cells is the volume multiplied by the concentration, the change in
mass can be expressed in terms of a mass balance;
Rate of change Rate of Rate of
in biological = increase due - decrease due
solids to growth to outflow
V ^| = VyX - FX (25)
Since - = D» dividing through by V and substituting D yields:
= yx - DX (26)
However, we know from experimental observation that a state is attain-
ed wherein dX/dt =0, or at least approximately so. Thus,
y = D (27)
The identity above is a very important one. It states that the spe-
cific growth rate is equal to the dilution rate, D, i.e., F/V; there-
fore, it is subject to hydraulic control. The incoming flow rate to
the reactor is then a very important physical parameter which can
exert a considerable effect on the biological system. This fact has
obvious engineering significance, since F is a system parameter which
is more subject to engineering control than are the biological para-
meters Umax5 KsJ ancl Y<
Equation 27 has another significance which may not be generally appre-
ciated. It states in essence that the specific growth rate, u, is in
fact a proportionality factor during exponential growth, and that
completely mixed systems operating under conditions approaching a
steady state with respect to X (i.e., dX/dt = 0) are operating in an
exponential phase of growth. This conclusion follows from the fact
that D was held constant; therefore y is constant, and if y is con-
stant, by definition, growth is exponential. The biological solids
concentration could not be held steady (dX/dt = 0) unless the cells
were replenished at a constant specific rate, y = D = In 2/td, Again,
this condition satisfies the previous definition of exponential growth,
and if a once-through system such as that shown in Figure 8 is said to
be in steady state (dX/dt =0), it must also be said that exponential
growth obtains.
It is now possible to write a mass balance for the reactor with
respect to change' in substrate concentration, S, and then to let
dS/dt approach zero in the steady state:
55
-------
Rate of Rate of Rate of Rate of decrease
change in = increase due - decrease due - due to consump-
substrate to inflow to outflow tion in reactor
(28)
As with the balance for biological solids, the rate of change in total
mass of substrate, (V)(dS/dt), in the reactor is the sum of the fac-
tors causing an increase and a decrease. The term for substrate con-
sumption, i.e., yX/Y, comes directly from the equation
dS/dt = }(dx/dt) (21)
and the equation
4 = yX (14)
As in the case of the mass balance for biological solids, we can sim-
plify the substrate balance by dividing by V and substituting D for
F/V:
«= DS. - DS -^ (29)
Also, since y = D, the dilution rate can be substituted in the con-
sumption term, and letting dS/dt = 0 as steady state is approached
provides an equation for the average steady state value of X (i.e., X),
which can be written in terms of the average steady state value of
S (i.e., S)
X = Y(S. - S) (30)
This equation for the steady state value of X indicates that the sol-
ids concentration is dependent on the biological constant, Y, and on
ACOD, wherein ACOD = S. - S. This is very similar to the expression
for X in a batch system.
The equation for X is not really useful as a predictor for X unless
5 is known or can be predicted by an independent relationship not con-
taining X. Again, equation 27 (y = D) provides a useful relationship.
Combining equations 27 and 18 (the Monod equation), but using the
steady state concentration of S, i.e., S, instead of S , we can write
-------
from which
5 = Ks iT-~T (32)
max
an equation for S in terms of the biological system constants, Ks and
^max> anc' the selected value of D (i.e., F/V) which is subject to engi-
neering control^ It is seen by equation 32 that the effluent substrate
concentration, S, is dependent, not on X, but on D and the biological
constants, in the simple once-through (no cell recycle) system shown
in Figure 8.
Again, it is emphasized that equations for X and for S can be shown
experimentally to provide a reasonable assessment of the biological
solids and substrate concentrations exiting the completely mixed re-
actor. These equations were presented as theory as early as 1950 by
Monod (30). They were further developed and tested by Herbert, et al.
(19) [also see Herbert (31)]. In all of these cases, the equations
evolved from the assumption that very small changes in S caused essen-
tially instantaneous changes in y; i.e., that S in the Monod equation
(equation 18) was not S0 in batch or S, the steady state substrate con-
centration in a completely mixed continuous flow reactor, but S at any
time and place. It can easily be shown that y is actually not imme-
diately sensitive to small changes in S (20, 21). For example, in a
batch system, a considerable amount of substrate is removed during the
time y is constant, i.e., during exponential growth. This happens at
initial substrate concentrations below those which will allow develop-
ment of ymax but it is readily observable only in systems for which Ks
is not extremely small. Therefore, when equation 18 is written using
S, rather than SQ, to describe growth in a batch system, the equation
is not really correct. Our studies with both pure cultures and a. heter-
ogeneous population have shown that y and S are not tightly coupled and
that "slippage" in the equation may amount to 50% (20, 21). The workers
cited above observed very low values of KS in their studies with pure
cultures, and this contributed to a theoretical development based on
some incorrect assumptions which led to some misleading interpretations
by subsequent researchers. Thus, there is an obvious inadequacy in the
"theoretical" basis of equations 30 and 32. For this reason we have
derived them as approximate empirical models for depicting "steady
state" values of X and S. In brief, we have taken a practical approach
which should be more appreciated by engineers than by biological kin-
etics researchers. Experimentally, a relatively steady state develops,
and these equations can be used to model it and to predict levels of X
and S; the steady state does not inevitably develop because of these
57
-------
equations or because of any theory on which they are based. The fact
that a process can be described quantitatively with formulas of pre-
dictive value is not necessarily ascribable to correctness of theory
or hypothesis. The equations are best looked upon as an outcome of
enlightened empiricism. They are important equations because they
are basic to description of the behavior in X and S of completely
mixed, continuous culture reactors.
For further discussion of once-through reactors, demonstration of the
pseudo-steadiness in X and S which develops using heterogeneous popu-
lations of sewage origin, and levels of X and S at various dilution
rates, the reader is referred to an article by Gaudy, Ramanathan, and
Rao in the research literature (22). Figure 11 is reproduced from
this publication and shows the typical degree of steadiness developed
in the laboratory reactor. It is apparent that S approaches a steady
state value much more closely than does X. Using equations 30 and 32,
it is possible to select values for ymax, KS, and Y (e.g., 0.5 hr ,
75 mg/1, and 0.6, respectively) for any value of S^ (e.g., 1000 mg/1)
and to calculate values for X and S at various dilution rates
(beginning with 1/24 and progressing toward D = ymax). These calcu-
lations have been made and will be shown in Figure 13 in the follow-
ing section. On a plot of X and S vs. D it will be seen that X and 5
remain about the same over a fairly wide spread of D values (or deten-
tion times, D = 1/t); the substrate concentration will remain rather
low and will sweep upward sharply as D approaches ymax- The biologi-
cal solids concentration will fall rapidly (wash out). Comparison of
such calculated values with those observed for heterogeneous popula-
tions in laboratory reactors validates use of these equations for once-
through continuous flow systems (22).
Cell Recycle with Constant Concentration Ratio
If cells (sludge) are recycled to the reactor after being separated
from the mixed liquor, as is the general practice with activated
sludge processes (see Figure 12), there are, in addition to the engi-
neering control possible through control of F and V, i.e., D, two
other important parameters which are subject to engineering control.
These are the hydraulic rate or ratio of sludge recycle and the con-
centration or concentration ratio of biological solids in the recycle
flow. Mass balance equations for steady state operation of such a
system have, as with the once-through reactor, been developed by
workers in the field of microbial kinetics for pure cultures and, as
with the once-through system, much research effort in our laboratories
has been expended to determine their applicability to systems in
which heterogeneous populations are employed.
The kinetic model to be discussed first is one which has been devel-
oped by Herbert (31). This model is useful, but we have not found it
well adapted for use with activated sludge systems, and will present
58
-------
Ul
LU
O
Q
Z
o
h-
1200
1000
800
600
^ 400
o:
UJ
LxJ
Q
200
BIOLOGICAL SOLIDS
CARBOHYDRATE
EFFLUENT COD
16 20 24
TIME, DAYS
28
32
36
Figure 11. "Steady state" parameters measured in a completely mixed laboratory reactor
of the once-through type, operated at a dilution rate of 0.33 hr-1 with nominal influent
substrate concentration, S-j, of 1060 mg/1 glucose COD (22).
-------
a modification of it. First, it is important to show the equations of
Herbert, since they provide a direct and simple expression of kinetic
behavior in systems employing cell recycle.
Figure 12 shows a flow sheet for a reactor with recycle. After leaving
the reactor, the mixed liquor is passed through a concentration step
(clarifier, centrifuge, etc.); in Herbert's case, a centrifuge was used.
Cells are concentrated to the desired degree for recycle. The parameter
which is employed as a systems constant in Herbert's model is the sludge
concentration factor, c, which is the ratio of cell concentration, XR, in
the recycle to cell concentration, X, in the reactor. The parameter for
recycle flow is termed the recycle ratio, <*, which is the ratio of recycle
flow rate to the rate of flow, F, of the incoming medium. Exiting the
system is the flow, F, containing the excess or waste cell concentration,
Xe. The dilution rate, D, for the system as a whole is as before, F/V,
but the dilution rate for the reactor, Dr, is now increased (detention
time decreased) because of the recycle flow and is equal to (1 + a)D.
The two new system controls (constants), a and c, effect changes in the
mass balance for the reactor. The materials balance for the rate of
change of biological solids in the reactor is:
Rate of change in _ recycle + growth _ outflow
biological solids ~ rate rate rate
v = aFcX + Vy X " F(1 + a)X
The materials balance for rate of change of substrate in the reactor is:
Rate of change _ inflow + recycle _ outflow _ consumption
in substrate rate rate " rate " rate
V at = FSi +aFS "F(1 +a)S " V T (34)
Dividing both equations by V, substituting D for F/V, letting dX/dt
and dS/dt equal zero (i .e. ,_steady state conditions) and substituting
the Monod relationship for S, equations 33 and 34 can be solved simul-
taneously, yielding the following equations for steady state concen-
tration of biological solids, X, and substrate concentration S:
(35)
K [)(1 + a - uc)
S - . - Dd + a - -.c) (36)
ma x
60
-------
AERATION
TANK
F..SL
— ^
S
Y
/\
V
af
^
(l*a)F
c ^
o
x
i_Mr\
r inr
r
•^
F
y«
Ae
XD= CX
Figure 12. Flow diagram for a continuous flow, completely mixed reactor
employing cell recycle at a constant ratio, c, of recycle solids, XD, to
aeration solids, X. Diagram represents the kinetic model proposed by
Herbert (31).
-------
Letting the factor (1 + a - ac) = A
X = I (S1 - S) (37)
K AD
S ' r-^AD (38)
max ttu
Equations 37 and 38 are those developed by Herbert (31). They provide
a fairly accurate description of the steady state levels of biological
solids and substrate concentrations for completely mixed continuous cul-
ture devices employing cell recycle; they help delineate important kin-
etic principles, and extend those developed in arriving at equations 30
and 32 for once-through systems. Selecting the same values for the bio-
logical constants as those used in making calculations for the once-
through system and selecting <* = 0.25 and c = 4.0, the calculated levels
of X and S for S^ = 1000 mg/1 are plotted for various values of D in
Figure 13. The values of X and S for the once-through system of Figure
8 are plotted in broken lines on the same figure to illustrate the
effect of recycling solids. The values of 0.25 for a and 4.0 for c
were chosen because they are values which are close to those typical of
many field installations. Design hydraulic recycle flow rates of 0.25
times the design flow are commonly used, and it is not uncommon to
observe recycle sludge concentrations at least four times the concen-
tration of biological solids in the aeration tank. Thus, the values
chosen are realistic ones (as are those for the biological constants).
The term A in equations 37 and 38 permits selection of a wide combina-
tion of values of c and a. It is obvious that selection should be such
that A does not go to zero. For example, one could not select a = 0.25
and c = 5. If one were to concentrate the return solids much beyond
c = 4, a could be reduced.
As seen from the figure, cell recycle permits the existence in the re-
actor of considerably higher concentrations of cells, X, than is pos-
sible in a once-through reactor, and consequently allows greater utili-
zation of substrate at lower detention times (higher dilution rates).
In both cases, the biological constants are the same, but with cell
recycle the engineering or physical constants, a and c, exercise a con-
trol over the system over and above that in a once-through system which
seeks the natural steady state substrate and solids levels governed
solely by the biological constants, Y, KS, and vm and the dilution
rate, D. It was shown previously that for a once-through reactor,
P = D (equation 27). However, for a system employing cell recycle,
V is less than D. According to Herbert's derivation of the equations
for this kinetic model (31), n is related to D by a factor depending
upon the values chosen for a and c, i.e., u = AD. From this it fol-
lows that the specific growth rate, u, in a system employing cell
recycle can be considerably lower than that in a once-through system
62
-------
3000
2500
2000
U>
en
«S)
IX
1500
1000
500
DILUTION RATE, hr
Figure 13. Comparison of predicted levels of X and S at various
dilution rates in a once-through reactor (broken lines) and a
reactor employing cell recycle concentration ratio, c, as a system
constant in accordance with the kinetic model equations of Herbert.
Values used for calculation were: umax> °-5 hr~l, Ks, 75 mg/1;
Y, 0.60; a, 0.25; c, 4.0, Si =1000 mg/1.
-------
with the same system dilution rate. For example, in the system shown
in Figure 13 where 1 + a - ac = 0.25, u = 0.25D. Thus, at a D of 0.2,
y is only 0.05 hr"1 rather than 0.2 hr'1 as in the once-through sys-
tem. Therefore, on the average, the probability that a cell will Divide
during a pass through the reactor with recycle is much less than in the
once-through system. However, as pointed out previously, according to
the mathematical definition of exponential growth (y is constant), if A
and D are kept constant and if dX/dt = 0, growth is exponential even
at the very slow specific growth rates of recycle systems.
We have in Herbert's model, a simple and direct, albeit empirical, pic-
turization of completely mixed continuous microbial growth systems in
steady state conditions. One could also use these model equations in
X and S to assess the effect on effluent substrate levels of various
values of the biological constants as was done for values of D in Fig-
ure 13. It was shown in Figure 13 that S remains fairly low over a
wide range of dilution rates; i.e., that these systems are capable of
delivering rather low effluent concentrations of S at high dilution
rates. However, operation at very low detention times cannot now be
recommended because of problems with sludge flocculation and settle-
ability at high specific growth rates, and because in this area even
more than in other engineering areas of functional design and opera-
tion, safety factors are needed. In any event, it can be seen through
study of these model equations and in charting the behavior of X and S
that biological treatment systems of this type possess potential cap-
abilities for removing substrate loadings much in excess of present
day loadings.
One cannot expect the experimental data (lab or field) to trace a per-
fect fit to such a model for the simple reason that for heterogeneous
populations the biological constants vary somewhat even if one could,
as in the lab, operate with S-j, D, a, and c steady. Herein lies
an important point, or engineering concept, which deserves emphasis.
One could use these equations in design with the assumption of constant
values of Sj, D, a, and c, in accordance with the model. But if the
design provided no means for attempting to hold these physical (engi-
neering) parameters constant, the operator could not hope to attain
the designer's goal. In brief, if designers use a model, the design
must attempt to accommodate the model; this is a major requirement of
"engineering control" of processes. An equalization or surge basin
could attenuate swings in Sj and D. If it were aerated, it could also
provide some degree of treatment as an unsteady once-through reactor.
The recycle flow is fairly easily controlled. Control of c requires
sensing for X in both the reactor mixed liquor and the clarifier under-
flow, and a means of adjusting XR to hold c approximately constant at
the chosen value of Xp>/X. This latter requirement is not very easily
attained, especially if X is subject to much variation. As with the
once-through model, the recycle model was originally proposed on the
assumption that pure cultures would be employed, for which Ks, ymax,
and Y are constants, and we have undertaken experimental studies to
64
-------
determine the applicability of this recycle model for heterogeneous
(natural) populations.
Experimentation with the once-through model had indicated some varia-
tion in S and somewhat more variation in X (e.g., Figure 11), but
average values of S and X during operation at various values of D did
produce dilute-out patterns in S and X in general accord with equa-
tions 30 and 32 (22). To test the applicability of Herbert's recycle
model to heterogeneous populations, a laboratory pilot plant was set
up in accordance with the flow diagram for the model which was shown
in Figure 12 (27). The biological sludge was not recirculated direct-
ly from the bottom of the clarifier, but was channelled to another
tank in which the recycle sludge concentration, XR, was adjusted in
accordance with c = XR/X. Data obtained for one set of values of the
operational constants are shown in Figure 14 (taken from reference 27).
The concentration factor, c, was purposely held low (1.5) in these
experiments in order to ensure that we could always obtain the required
XR by diluting underflow. The sludge recirculation ratio, a, was
maintained at 0.25. Figure 14 shows the variation in X and XR for S-j
of 1000 mg/1 glucose (1060 mg/1 COD)(top portion) and in substrate,
S, in terms of COD and carbohydrate (lower graph). In the example
shown, the detention time in the activated sludge tank was three
hours; i.e., Dr = 1/3 hr-1, and the overall dilution rate, D, was
1/3.75 hr-1 i.e., Dr/(l + a ) . It is seen that in general, c was
maintained close to the selected 1.5 ratio. Also, it is apparent that
X varied; the reactor did not attain a "steady state." The values for
S (COD) also varied considerably, but the scale is somewhat expanded
(8 to 1, as compared with the top graphs), and the effluent substrate
concentration, S, could be adjudged as fairly steady over the two-
month period of operation at this detention time. The values of S
measured as carbohydrate (using the anthrone test) showed only slight
variation. Both the COD and carbohydrate measures for S shown in the
figure were obtained by analysis of the membrane filtrate. However,
it should be noted that during this run, had COD been plotted for the
clarifier supernatant (actual effluent), the values would be only
slightly higher because the sludge settling characteristics were rather
good. In general, the supernatant COD was 25-30 mg/1 higher than the
filtrate COD. The average value of aeration solids during this period
was 753 mg/1, and the average effluent COD was 79 mg/1. These values
along with other parameters for operation at this and other values of
D are given in Table III. In Figure 15, the average experimental
values for X and S from Table III are plotted as the "steady state"
levels (X and S) for the four dilution rates examined. The heavy
lines through these experimental points describe a portion of a dilute-
out in X and rise in S as D increases. This experimental behavior
agrees with the dilute-out pattern which is the expected behavior
according to equations 37 and 38^ The_dotted curves labelled X-| and
S^ show the predicted levels of X and S calculated from equations 37
and 38 using values of the biological constants, ymax' KS' anc' ^
determined for the cells present at each dilution rate. The ym:3V and
lildX
65
-------
1000
"10 20 30 40 50 60 70
DAYS OF OPERATION
250
^
o>
Q 200
UJ
h-
Q 150
-CHEMICAL
OXYGEN
DEMAND
20
30 40 50
DAYS OF OPERATION
60
70
Figure 14. Operational data for a laboratory reactor
operated according to the kinetic model of Herbert
with cell recycle at a constant recycle ratio, c, of
1.5. S-j = 1060 mg/1 glucose COD, a = 0.25, aerator
detention time = 3 hr (27).
66
-------
1000
BIOLOGICAL SOLIDS
(X)
CARBOHYDRATE (S
0.10
0.20
0.30
0.40
0.50
DILUTION RATE, D, HOURS"
Figure 15. Comparison of observed (solid lines) values of X and S with values
predicted (broken line curves) by the kinetic model equations of Herbert using
experimentally determined values of the biological constants (see text for
details)(27).
-------
00
TABLE III
OPERATIONAL PARAMETERS FOR A COMPLETELY MIXED CONTINUOUS FLOW REACTOR WITH CELL RECYCLE
Mean
Rate, hr" Time, t, hr
1/7.5 6
1/5 4
1/3.75 3
1/2.5 2
3 . A ™
b. c^ =
c. S =
d. 3, =
e. S_ =
f. c
a =
c =
Si =
a bed
X cX S" 5i
mg/1 mg/1 mg/1 mg/1
837 1295 60 1040
797 1236 87 1060
753 1121 79 1067
500 861 277 1065
mean biological solids
mean biological solids in
mean effluent COD
mean influent COD
mean effluent carbohydrate
COD removal rate
0.25
1.5
1060 mg/1 glucose COD
e
5C
mg/1
6
15
17
156
(srs)f
Xt
mg COD/
mg SS/hr
0.195
0.305
0.438
0.788
Loading
Factor
Ib COD/
Ib SS/day
9.94
16.0
22.6
51.1
COD Removal
Efficiency
94.2
91.7
92.6
74.0
the recycle
-------
Ks values were obtained by taking a sample of sludge as seed material
at various times during the operational period and using these samples
for batch growth studies run as previously described. The value for Y
was calculated directly from the pilot plant data. The curves labelled
Xp and So were calculated using average values of ymax, Ks> and Y for
all the dilution rates examined. In general, it can be concluded that
the model does describe (predict) fairly well the behavior of these
systems in which the microbial population was not a pure culture but a
heterogeneous bio-mass, i.e., an activated sludge.
After a large expenditure of research effort we are thus able to con-
clude that the model proposed by Herbert does describe the observed
biological behavior. It is appropriate to examine the model, and to
defend and criticize it with regard to the reasonableness of the
approach before going further. It is first and most importantly a
simple one unencumbered by a multitude of biological constants which
need to be determined; only V-max, K§, and Y are required. It may be
argued that in setting up the materials balance there was no term for
substrate consumption for maintenance of existing solids, i.e., sub-
strate was considered as being consumed only for respiration and growth
of the bio-mass. The amount of exogenous carbon source used by a
microbial cell simply to maintain itself intact is very, very small
compared to the amount needed for growth. It is so small that it is
not a significant factor and, indeed, is very difficult to measure
except for extremely high cell concentrations. Also, in the mass bal-
ance for X, no term was inserted for a decrease in X due to "endogen-
ous" degradation of the bio-mass. This omission is also defensible.
In the presence of exogenous substrate (S,- coming into the reactor con-
tinually), the net effect is an increase in the sludge concentration.
If this were not so, there would be no excess sludge to dispose of.
(The special situation in which endogenous metabolism, i.e., autodiges-
tion, is extremely important, the extended aeration process, will be
discussed in Chapter VIII.) At any rate, it is impossible to deter-
mine whether, or how much, endogenous metabolism occurs in the pres-
ence of exogenous substrate, and non-measurable quantities have no
practical value. Since net increase in solids is measurable, endogen-
ous metabolism is automatically accounted for if it does occur, and
there is really no good reason for complicating general, working-
model equations for activated sludge by insertion of a term for endog-
enous degradation. The question posed is: Why overcomplicate for the
sake of not missing anything which could possibly affect X and S?
There seems to be no justifiable reason for doing so. One can surely
be cognizant of the fact that these phenomena can occur, but if they
do not play a significant role (maintenance) or are not measurable as
separate reactions (endogenous metabolism), they can and should be
omitted from the general model.
Thus, it can be seen that the model presented could be used as an
approach to plant design, an approach which is in accord with biolog-
ical principles relating the major parameters which govern the function
69
-------
(operation) of the plant. Equalization tanks could help even out
fluctuations in D and Si; a can be controlled fairly easily, but a
constant sludge concentration factor causes difficulty in operation.
It was observed in once-through systems (e.g., Figure 11) that while S
attained a relatively steady level, a steady state with respect to X was
only roughly approached. When sludge recycle is practiced (e.g.,
Figure 14), X is subject to great fluctuation. If one operates in
accordance with the model, it is required that c remain constant.
In order to accomplish this, one is required to change the recycle
sludge concentration, X^, when he detects a change in X in the
reactor. This mode of operation requires rather close monitoring of
the plant (which should be done in any event). A more serious defect
of this model results from solids variations due to heterogeneity of
the population, so that operation at a constant c actually militates
against steadiness in X. The operator may take a sample of mixed
liquor and find that since the last sample (e.g., the previous day or
shift), the aeration solids concentration has increased. To hold c
constant, he then must increase X^. This will not help to steady X,
but will tend to cause a further increase in X. Thus, while the model
may work well with pure cultures and while it offers an excellent
approach to kinetic picturization of the behavior of S and X for the
activated sludge process and could be employed as a design tool,
operation in accordance with this model is difficult, and militates
against steadiness in X. One could let X vary, so long as S remain-
ed steady, since the prime concern is naturally with the effluent
quality and with reliable delivery of as low an S as is economically
possible. However, since constancy of c is the most difficult oper-
ational requirement of the model and militates against predictability
of X in activated sludge systems (heterogeneous microbial popula-
tions), we asked the following question: What happens to the model
if, instead of holding c as an operational constant, we hold X^ con-
stant (an engineering expedient which we could facilitate in practice
much more easily than constant c)?
Cell Recycle with Constant Return Solids Concentration
A flow sheet for the system operated with XR constant is given in
Figure 16. It is essentially the same as Figure 12, but an aerated
sludge consistency and holding tank (aeration tank 2) is placed in
the line. Exiting the process is the effluent substrate concentra-
tion S and excess sludge, Xe. The total flow leaving the system is
F (part of the flow is the supernatant and part is the flow of waste
sludge).
Assuming, as in previous calculations, that the biological solids in
the incoming waste flow is negligible compared to solids in the
reactor, the balance equation for the rate of change in reactor sol-
ids can be written as follows:
70
-------
AERATION
TANK 1
SL
S
X
V
CLARIFIER
S
X
AERATION
TANK 2
aF
Figure 16. Flow diagram for a continuous flow, completely mixed reactor
employing cell recycle at a constant recycle solids concentration, XR.
Diagram represents the kinetic model proposed by Ramanathan and Gaudy
(32).
-------
(recycle) (growth) (outflow)
VdX/dt = aFXD + V MX - (1 + a)FX (39)
K
Proceeding as before, and setting dX/dt = 0:
y= D [0 +a) - (aXR/X)] (40)
The symbol X has now been replaced by the average steady state value
X, in accordance with attainment of steady state in X; i.e.,
dX/dt = 0.
The materials balance equation for the change in S through the
reactor is:
(inflow) (consumption) (outflow)
VdS/dt = FSi - yXV/Y - (1 + a)FS (41)
from which
X = (YD/u) [S1 - (1 + a)s] (42)
Substituting_y from equation 40 in equation 42, an expression for X
in terms of S is obtained, and one for S in terms of X is_pbtained
by substituting the Monod relationship for u in terms of S in equa-
tion 40. Solution of these simultaneous equations leads to the
quadratic form in which X and S are given as follows (for more detail-
ed presentations of the intermediate steps in the derivation, see
reference 32) :
a)S"] +aX
J
X = (1+a) - (43)
5 - b ± N b2 - 4ac ....
S = - - - (44)
where
»' "max' (' +a'D (45)
b-D[S,- (l+.)ig - [s. t o,xR/Y] (46)
72
-------
c = KsDSi (47)
These equations can be employed to depict the behavior of X and S
for various values of VLax, K$, Y, S., D, a, and XR. It is readily
appreciated that the calculations can be facilitated by use of a
computer. It is interesting to compare the values of X and S for
this model with those for the recycle system shown in Figure 13.
Such a comparison is presented in Figure 17. The curves labelled B
are those representing Herbert's model and are replotted from Figure
13. The curves labelled A were computed from equations 43 through
47. The same values for all system parameters except recycle solids
were used in calculating the two sets of plotting points. The only
difference in the two systems is that c, i.e., XR/X, is held con-
stant at a value of 4.0 for curves B, while XR is held constant at
a value of 10,000 mg/1 for curves A. The latter value for recycle
solids concentration was chosen as one which is readily attained by
quiescent settling. Two things are particularly important to note
in this figure. First, operation at constant XR flattens the
dilute-out pattern of the system, i.e., provides more stability at
high dilution rates; secondly, at dilution rates which would usually
be considered for use in activated sludge systems (certainly those
below 1.0), the values for S are the same for both models.
The model in which c is a system constant is more in keeping with
justifiable kinetic theory, because it contains no assumption which
prevents a system from diluting out. In the model we are recommend-
ing for activated sludge, total dilute-out is impossible (XR is a
constant), and X can reach only a lower level of XR (a /I + a); also
the assumption of zero substrate concentration in the recycle flow
(see substrate balance equation 41) is unrealistic at extremely high
values of D because S cannot rise to Sj but can only approach
S.j(l/l + a). However, at realistic values of D, the consequences of
these assumptions are immaterial; thus the engineering expediency in
making XR a system (design) constant is entirely defensible and
advisable. These points are developed more fully in reference 32.
The behavior of the kinetic model in X and S which we are herein
recommending has been examined computationally for various values of
%ax> Ks, Y, a, XR, and S-j at various values of D (.32). All of these
factors are important kinetic parameters which affect both X and S.
However, at a reasonably high value for S^ (e.g., 1000 mg/1), changes
in the biological constants ymax» Ks> and Y do not appear to exert as
much effect on X and 5 as do the physical constants a and XR (32).
In general, selection of a = 0.25, which is a value widely employed
traditionally, appears to be a rather good one for XR values which
can be reasonably attained in practice (e.g., 10,000 mg/1). The
values for these parameters which are in one case (i.e., a) nomin-
ally attained in the field, and in the other case (i.e., XR) are
usually attained fortuitously, are rather satisfactory for dilution
73
-------
2500
2000
o>
CO
tr
LU
3E
<
tr
1500
1000
o
o
z
500
0.4
O8 1.2
DILUTION RATE (D), hr"1
Figure 17. Comparison of values for X and 5 at various dilution
rates predicted by the two kinetic models for cell recycle.
Curves A are those predicted by the model equations of Ramanathan
and Gaudy (32). Curves B are those predicted by the model equa-
tions of Herbert (31).
74
-------
rates of less than 1/2 hr" and organic substrate loadings of 1000 mg/1
and below.
In general, the model we have presented tends to explain why acti-
vated sludge plants have, on the basis of experience, gained a good
reputation as substrate removers. Also, the model shows that unless
one does control a, XR, S-j, and D, they can give variable results in
S (the most critical parameter). It is rather fortunate that the
biological parameters, umay, KS, and Y, which are not really subject
to engineering control (unless predominance of species can be con-
trolled at some point in the future), can be expected to exert less
effect on performance than do the parameters which can be subject to
engineering control (a, XR, 85, and D).
It is emphasized again that the operating engineer stands little
chance of operating the system steadily and reliably unless the means
for exerting control over these selectable variables are provided for
by the design engineer. Conversely, if they are provided in the de-
sign, they are of little avail unless used by the operating engineer.
In Figure 18, the computed steady state levels of S and X at various
dilution rates for S^ levels up to 5000 mg/1 are plotted in accordance
with our kinetic model (equations 43 through 47). The values employed
for the biological constants as well as the engineering constants are
those used previously (e.g., for calculating the curves shown in Fig-
ure 17), since they appear to represent values that are reasonably
expectable on a biological basis and reasonably attainable from an
engineering viewpoint, respectively. It can be seen that the effi-
ciency of substrate removal remains fairly high over a wide range of
feed substrate (S-j) concentrations, especially for dilution rates of
0.5 hr~l and below. This dilution rate corresponds to a reactor
dilution rate, Dr, of 0.625 hr"1, i.e., a reactor detention time of
1.6 hours. At higher dilution rates, as Sn- is increased, substrate
leakage increases significantly. In Figure 19, the curves for D val-
ues of 0.1, 0.5, 1.0, and 2.0 hr'1 for Si up to 1000 mg/1 are plotted
in expanded scale. Figure 18 was intended to show the overall behavior
of the model equations, but the range of S-j up to 1000 mg/1 is prob-
ably the one of most interest, since for many types of waste water the
biologically available organic matter (measured as ACOD) is within
this range.
A graph such as this could be employed as a design "guide," but its
real value is that it shows the potential capability of the process.
Estimates for design should be revised upward from this baseline guide.
A constant balancing of "theory" vs. "engineering judgment" or
"experience" is as necessary (or even more so) in this area as it is
in any other area of engineering. , For example, in Figure 19, the
"chart" says that at a feed COD of 1000 mg/1, just over 60 mg/1 COD
ought to appear in the effluent at a D of 1.0 (reactor detention time
of 0.8 hour)- Thus well over 90% efficiency is predicted by the
model at this low detention time. In determining whether the
75
-------
4ooo
3600 .-
42OO
3600
500
2000 Z500 3000
FEED COO (S, I . mg/l
3500
4000
4SOO
Figure 18. Predicted levels of aeration solids, X, and
effluent COD, S, for a range of organic loadings, S-j, at
various dilution rates. Curves were computed according
to the model equations of Ramanathan and Gaudy (32) using
the following values for system constants: vimax, 0.5 hr-
Ks, 75 mg/l; Y, 0.60; a, 0.25; XR, 10,000 mg/l?
76
-------
2500
2400
2300
en
E
IX
2200
2100
2000
180
160
140
120
100
en
E
i of
80
60
40
20
DILUTION RATE, D, hr".
O.I
7
2.0/
.0
200
400
600
800
1000
Figure 19. Expanded scale plot of the curves from Figure 18
for organic loadings up to 1000 mg/1 COD.
77
-------
prediction is reasonable, e.g., in checking performance in treatabil-
ity studies, one needs to consider first that the substrate COD of
the chart is just that: not total COD, but the portion of organic
matter in the waste which is capable of being utilized as carbon
source by an acclimated heterogeneous microbial population, i.e., the
parameter ACOD developed in Chapter II. For example, a waste con-
taining biologically resistant organic matter may exhibit a very large
residual COD even after the biologically usable (i.e., biologically
oxygen-demanding) organic matter has been removed. Thus one could
not expect to register COD values as low as shown, even when these
values, on the basis of utilizable carbon source, were attained. Use
of a test for residual biologically degradable material in the efflu-
ent, as recommended in Chapter III would, however, demonstrate the
high efficiency of the process. It should also be realized that CODs
run on potable tap water can sometimes be registered as high as 25
mg/1. Also, a very important point to make concerning this model
(which we believe to be the most realistic one available) is that it
is an engineered one, i.e., D,a, Sj, and XR, the engineering para-
meters which can be controlled or for which engineering control can
at least be approached have, in accordance with the model, been con-
trolled. One cannot approach these results unless one approaches
this degree of control. Also, it should be re-emphasized that the S
refers to organic substrate in solution. Any kinetic biochemical
model yet devised is addressed only to this aspect, i.e., the assump-
tion is made that the biological solids have been separated from the
mixed liquor before the effluent is discharged. If the separation
is not essentially complete, the COD of the cells which remain in
suspension will add to the values for effluent COD predicted by the
model. Thus, although we can recommend the graphs of Figure 19 as a
guide in selection of reactor detention time at rather steady loads
of S-j, the model cannot replace the judgment inputs of the engineer.
It can enable more enlightened and reliable judgment; it provides a
way to balance rational kinetic theory and experimental data (lab
and field experience).
Employing the model equations (43-47), one could construct his own
"design" charts using biological constants for the specific waste
being investigated. These can be determined as previously described
(Chapter IV), using populations acclimated to the waste in question.
Also, during the treatability study, the expected amount of biolog-
ically available carbon source (ACOD) is determined. Then the
behavior of X and § for various dilution rates and engineering con-
trol design parameters can be computed and charted. The resulting
charts can be employed as the design guides. One of the most impor-
tant things to be considered is that there needs to be some provision
in the design to attempt, or approach, control of the controllable
parameters, i.e., S-j, F, a, and XR.
Possible ways and means of gaining some engineering control over
variation in S. and F will depend upon analysis of the individual
78
-------
waste water situation. For some industrially produced wastes, con-
trol of S-j and F may be possible at the various individual sources
comprising the total waste. This is not always feasible, and in the
case of municipal wastes, this type of control may be even more dif-
ficult to attain. Control at the treatment plant would seem to
dictate use of some sort of surge basin. For a waste containing
settleable material, it would be best to place the settling tank be-
fore the surge basin. There seems little to gain in using a surge
basin as a settling basin. Depending upon the size of the surge
basin required, the cost of providing for removal of sludge from the
basin could be prohibitive. Also, depending upon the size of the
basin, relative seeding population and strength of the waste, etc.,
some slight degree of mixing and aeration in the surge basin may be
required. Thus, it seems best to let the settling tank ride on the
flow in the usual manner and to even out F and S-j to the activated
sludge tank. Microbial growth in the surge basin can be somewhat con-
trolled by providing a paucity of aeration, e.g., enough to keep it
from going anaerobic but not enough to allow the carbon source to
limit growth. These are aspects which are naturally subject to engi-
neering investigation in the treatability study stage which precedes
the design stage. Some wastes will be grossly deficient in essential
nutrients such as nitrogen and phosphorus, and this situation may
obviate the need for providing aeration in the surge basin. If there
is much growth in the surge tank, the biological solids concentration,
X.,-, entering the activated sludge reactor might be included as a term
in the materials balance equation. This would lead to some modifi-
cation in the resulting model equations (equations 39 through 47).
In most cases, however, the amount of biological solids which grew in
the surge tank would be small in relation to those in the aeration
tank (which contains the recycle solids) and the assumption of neg-
ligible X in the aeration tank influent would appear to be satisfactory.
When the use of a surge basin is contemplated, it is advisable to
determine ymax» Ks, Y, and ACOD, using the effluent after the needed
period of storage in the surge tank. Metabolic activity in the surge
tank even under severe 62 limitation can lead to chemical alteration
of the carbon source in the waste. Thus, although significant sub-
strate removal may not take place, the original organic carbon source
may, in some degree, be converted to metabolic products (e.g., some
fermentation products). This may have the effect of reducing the
variability of the kinds of organic compounds in the waste. This
slight amount of pre-treatment may be highly desirable (more research
is needed here) but the point emphasized is that one should determine
the characteristics of the system as they apply to the reaction ves-
sel, and operations on the raw waste preceding the reactor can affect
these characteristics; thus it is not necessarily the raw waste the
designer is concerned with when considering the design of the acti-
vated sludge process.
Surge or equalization basins may or may not be considered in design
79
-------
depending upon the expected variation of S^ and waste flow rate in
any particular situation. As we shall see in the next chapter,
completely mixed systems are inherently provided with some degree of
internal stability and do resist changes in S in response to changes
in Si and F. Control of XR can be facilitated by the incorporation
in the return sludge line of an aerated sludge consistency and
storage tank from which sludge at constant concentration XR is
pumped at flow rate aF.
It might be argued that prolonged holding of sludge in the consis-
tency tank might affect the biological constants. Some aerobic
digestion could and would take place in this tank in which concen-
trated sludge is aerated. In this case, portions of sludge itself
represent substrate, and preliminary studies indicated that even
aerated storage up to 12 days did not cause larger variations in
^max (when the sludge was used as seeding material in growth studies)
than those normally obtained with freshly grown populations. A
greater amount of experimental data on this aspect would be desir-
able, but it does not appear that the placement of an aerated sludge
consistency tank in the return line will cause any drastic change in
the biological constants.
Summary and Conclusions
Equations have been developed in this chapter for three kinetic
models describing the steady state in continuous flow, completely
mixed, microbial growth reactors. The first model considered was
that for the simple once-through reactor generally used in microbio-
logical research with pure cultures. Experimental evidence for the
applicability of this model to "steady state" growth of heterogen-
eous populations was presented. Two models describing systems more
operationally similar to the activated sludge process, i.e., contin-
uous flow reactors with cell recycle, were then developed and com-
pared with each other and with the once-through model.
Herbert's model, in which the ratio of recycle solids to aerator
solids is maintained constant, was originally developed for use
with pure cultures. Experimental testing of the model showed that
it provided a fairly accurate prediction of the behavior of hetero-
geneous populations. However, operational control, even in the
laboratory, was difficult and a steady state in solids concentration
could not be achieved.
The kinetic model which we have proposed in equations 43 through 47
is more operationally feasible since the system constant employed
is the concentration of recycle solids. While this model is not
theoretically correct in that prediction is poor for conditions
approaching dilute-out, it is realistic and is applicable over the
range of values of the various kinetic parameters which might be
80
-------
encountered in practice in biological treatment. This model is
recommended for use in the design of activated sludge processes,
with the caution that provision for control of the parameters S-j,
F,'a, and XR must be made in the design and must be utilized in
operation. The flow sheet for this model includes a sludge consis-
tency tank required for maintaining constant recycle solids concen-
tration, and a surge tank is recommended for systems expected to
be subjected to significant variations in volume rate of flow or in
waste concentration.
Both the surge basin and the sludge return consistency tank repre-
sent departures from "standard" or "traditional" practice. They
represent changes which are in accord with more sound conceptual
principles of microbial metabolism and accommodate the particular
needs for engineering control of the process for the heterogeneous
bio-mass, i.e., an activated sludge.
The kinetic model herein recommended and the suggested changes in
unit processes which are desirably included in the flow sheet in
order to best accommodate the model are of practical utility to engi-
neers and operators. The model is by no means the "last word" con-
cerning activated sludge, nor do we claim that it comprises the
ultimate in kinetic theory. It is recommended simply as a practical
advance in conceptual understanding and in design and operational
control of the biological process.
81
-------
VI
FACTORS TENDING TO DISRUPT THE STEADY STATE - SHOCK LOADINGS
In the previous chapter we presented a kinetic model for activated
sludge of the completely mixed type in which the system is designed so
that it can approach a steady state condition. The steadiness is
encouraged by engineering actions in the design phase which are intended
to smooth out fluctuations in S^ and F, and operationally the system is
encouraged toward steadiness by controlling a and holding XR, rather
than c, constant. These are things engineers can do to make an acti-
vated sludge process operate more in keeping with the premise on which
it was designed.
Even with the incorporation of these steadiness-enhancing modifications
in the design of the system, there can be times when the activated
sludge may be subjected to changing environmental conditions which tend
to disrupt the steady state. Any environmental change tending to do this
can be termed a shock load. Those environmental changes which cannot be,
or have not been, smoothed by preventive engineering expedients must be
accommodated solely by successful biological response or by combined bio-
logical and engineering remedial responses and/or measures. It is some-
what paradoxical that although the response of natural (heterogeneous)
microbial populations to environmental changes has been one of the prime
research interests of the senior author for more than a decade, the
space which can be given to this subject in this particular document is
the short chapter which follows. This chapter is intended to serve pri-
marily as an introduction to biological responses to environmental
changes, or shock loads. One can begin to appreciate the vastness and
depth of this area by remembering that the biological world as we know
it today evolved in response to environmental change. Often the envi-
ronmental change was one which was imposed externally on the biological
system and often the change in environmental conditions was produced by
species within the "bio-mass." Adaptation and selection of species occur
in response to environmental change. Also individual species can make
an internal (molecular level) adjustment and become acclimated to a
change in environmental conditions. An activated sludge, or for that
matter any naturally selected microbial population, represents a micro-
cosm which in itself can be said to serve as a crude model of the
evolutionary process. Thus it possesses many of the complexities which,
if ultimately resolved through continued investigation, could add in
great measure to attainment of the ultimate goal of the life sciences,
i.e., understanding of the biological world.
Whether or not one agrees with these statements there can be little
argument concerning the mechanistic complexity of these responses, and
oversimplification of the situation would be misleading and a dis-
service to the profession. On the other hand, in-depth discussion at
the molecular and ecological level would require a voluminous document
83
-------
and would probably necessitate greater amounts of study and reference
reading in the area of biology than most practicing engineers would be
able to devote to it. While one could surely give valid arguments for
such a course of study by pollution control engineers, it is not appro-
priate for the present document. We shall, however, attempt to intro-
duce this important subject and set forth some rough and tentative
guidelines which engineers can use. These are based primarily upon
research in our laboratories, much of which is yet to be published.
Quantitative and Qualitative Changes in Substrate
A quantitative change in S^ (concentration of influent carbon source)
is the type of perturbation usually and traditionally envisioned as a
shock loading. In the usual jargon, a shock loading is envisioned as
an increase in "BOD loading," although it could also be a decrease. A
decrease in organic loading, if not too severe, e.g., a sudden halving
of the loading, would precipitate some adjustment in X (concentration
of biological solids) but would not be expected (in our experience) to
cause a severe leakage in substrate (i.e., increase in S) in the
effluent. Such shocks cannot be discounted, but our experience indi-
cates that they are less deleterious than a comparable increase in
loading.
An increase in loading can come on the line in various forms. There
can be a rise in concentration of metabolizable carbon source with no
change in F, i.e., a quantitative shock load. There can be a rise in
organic substrate concentration which is accompanied by a change in F,
i.e., a combined quantitative and hydraulic shock loading. Also there
may be a rise in the total concentration of organic substrate coming
into the system, and the additional organic matter may be of an entirely
different type than the steady state substrate. Alternatively, the
entire organic loading may change from one type to another, probably
also involving a change in total organic concentration. In brief, the
quantitative shock loading in S-j may often be accompanied by a quali-
tative shock loading in S-j. Also the change in Sj may be more or less
gradual and of variable duration. Or it may be applied as a relatively
short-lived pulse or slug dose. Slug doses to the sewer are commonly
experienced and, where the run to the treatment plant is short, the
activated sludge aeration tank may receive a relatively unattenuated
slug. The use of completely mixed reactors is helpful in this regard
because of the instantaneous dilution of the inflowing material (see
Figure 10).
Thus it can be seen that even the "simple" quantitative shock loading
(e.g., an increase in BOD loading) is not really a simple concept for
which it would be easy to provide equations, "models," and/or rules of
thumb of unquestionable predictive value. However, some approaches to
conceptual analysis can be initiated, by considering the system which
might be expected to be the most susceptible to deleterious changes in
84
-------
S-j. Some conceptual insight into the ability of such systems to respond
successfully to changes in S^ may form a basis for making conservative
estimates for a more shock-resistant system. The least resistant system
is the once-through system with no cell recycle. Activated sludge
processes are not designed to operate as once-through systems, but the
study of such systems provides conservative insight. The recycle of bio-
logical solids would be expected to provide resistance to the leakage of
substrate in a shock-loaded process and, in our experience, it does.
Therefore, systems without recycle are more suitable for studies design-
ed to determine how each type of shock affects the steady state and
limits within which the effects are acceptable.
A quantitative shock load occurs in a once-through system operating under
steady state conditions of X and S CSf and D constant) when the inflowing
feed, S.j, is increased. If the increment of increase in S-j were not
metabolized by the sludge, substrate concentration in the reactor (and
the effluent) would increase along a dilute-in curve (Figure 10) until
the new effluent_substrate level would be equal to S + AS-j, i.e., the
former level of S plus the new increment in substrate concentration due
to the increase in S-j. The system would be said to have undergone a
decided functional failure, i.e., the effluent substrate concentration
would have increased to an unsatisfactory level. For a successful
response to occur, S in the reactor would have to be held approximately
at the steady state level of S by increased growth (increased X) due to
consumption of the new amount of carbon source. Thus X would increase
to a new steady state level, while S would remain the same or, in any
event, the same percent substrate removal efficiency would be maintained.
In accordance with the hyperbolic equation (Monod equation, 18) the
increase in substrate can be expected to cause an increase in the specific
growth rate, y, thus increasing X. The increase in X uses up the new sub-
strate (Y remains "constant"), preventing S from rising. The behavior of
X and S, i.e., rise in X to maintain S constant, forms the transient
stage of successful response as the system approaches the new steady
state in accordance with the new concentration of Sn-.
In an approximate way such a response can be shown to occur, but one
must use "conceptual caution" tempered with empiricism and experimental
observation in analyzing this important response. The hyperbolic (Monod)
equation is in itself an approximation. It will be noted that in equa-
tion 18 we have employed SQ as it refers to batch (transient system)
kinetics and have indicated that the substrate value, S, could be used
provided it was understood that it referred to S in continuous flow sys-
tems under steady state conditions (equation 31). In a continuous flow
system when S-j in the incoming flow is changed, steady state conditions
no longer exist (because S is changed) and it has been shown that y does
not change instantaneously in response to a change in S (20, 21, 33).
Therefore, equation 18 really cannot be expected to predict precisely
the concentration of S and X during the transient resulting from an
increase in S-j, although this relationship does provide rather good pre-
diction of the new steady state levels of S and X.
85
-------
Figure 20 shows a response in X (upper portion) and in S (as COD and
carbohydrate) when a once-through system, which was operating previously
in a "steady state" with S, of 450 mg/1 COD, was shock-loaded by chang-
ing S-j to 1450 mg/1 COD. The dilution rate, F/V, was 0.244 hr~l, i.e.,
a reactor detention time of slightly over four hours. The response to
this rather severe shock loading was an increase in X during the tran-
sient phase as X approached its new steady state level, but X did not
rise rapidly enough to prevent substrate leakage (see bottom portion of
the figure). Approximately 1.5 to two detention times were required
before the system recovered, i.e., regained the steady state level of S.
Obviously this shock was greater than the system could successfully accom-
modate, unless one accepts a six-to-eight hour functional disruption.
This severe shock loading was purposely applied so that COD would leak
from the system in order that the Monod equation could be tested during
the transient state. The computations involve numerical integration
techniques applied to the balance equations in X and S (equations 25
and 28) for once-through systems (i.e., dX/dt and dS/dt not equal to 0
in the transient state). These calculations are discussed in reference
33. It is important here to note that the dotted curves for S and X
labeled "predicted by the Monod equation" provide at best a rough pre-
diction of the observed transient response.
An important consideration in response to shock loads is illustrated in
the plot of filtrate carbohydrate COD (values designated on the figure
by triangles). These values were obtained by analyzing filtered sam-
ples for carbohydrates (the synthetic waste being fed was a carbohy-
drate), and it is seen that during the transient rise in COD a very sig-
nificant portion of the substrate which leaked from the system was not
the original carbon source being fed but consisted of organic compounds
produced by the organisms from the original carbon source. Thus, during
the quantitative shock loading there was also a qualitative change in S.
Even a "simple" quantitative shock can cause complications which can
affect the mechanistic and kinetic response to the change in the envi-
ronment.
It could be argued that attempts to provide mathematical models for
depicting or predicting transient responses are, at this stage of the
knowledge, somewhat futile because to cover all of the metabolic con-
tingencies, they would need to be very complicated and unwieldly to em-
ploy. Any relatively simple model would have to be such a gross over-
simplification of the system that it would not provide meaningful output.
Obviously, there is need for much data obtained by study of defined and
controllable environmental changes before satisfying mathematical models
for shock loadings can be produced. Such endeavors are an important
area for continued research. However, from the standpoint of the con-
sulting engineer interested in the design of the process, mathematical
models for the steady state operation are of more immediate utility, and
what are vitally needed are guidelines regarding the extent of change
in S. which can be accommodated without severe disruptions of the
86
-------
800
RESPONSE PREDICTED BY
THE MONOD EQUATION
m
600
345
TIME, HOURS
400
<
DC
\-
co
m
lo 200
o FILTRATE COD
A FILTRATE CARBOHYDRATE COD
0
\
x
RESPONSE PREDICTED
BY THE MONOD
EQUATION
345
TIME, HOURS
Figure 20. Response of a continuous flow, completely mixed
reactor of the once-through type to a severe quantitative
shock load consisting of a change in S-; from 450 mg/1 to
1450 mg/1 COD. The dilution rate was 6.244 hr-1 (33).
87
-------
designed plant efficiency. Obviously, a 200% increase in S-,- under oper-
ational conditions extant for the system of Figure 20 would not be
desirable. The prime question which should be answered now is how much
change in Sj could be accommodated without a rise in S. Even this infor-
mation, while more easily provided than a mathematical model depicting
the transient, is not easily given. We know from our own experiments
that the change in S-j which can be accommodated is dependent upon the
reactor dilution rate (or reactor detention time) and upon the biological
solids level in the system. At reasonably selected reactor dilution
rates, i.e., those not close to Mmax> even a once-through system will
undergo little if any leakage of substrate during the transient state
when S-j is increased 50% and the reactor detention time is approximately
four hours. At a detention time of eight hours, a 100% increase in Si
can be expected to cause only a minor and short-lived increase in
effluent S. With sludge recycling (maintaining X higher than in once-
through systems), the range of increase in Sn- which can be expected to
be accommodated without a significant increase in S can be revised up-
ward. We have obtained a large amount of experimental data for quanti-
tative shock loadings at various detention times and recycle sludge
levels and will be able, after more complete analysis of these data, to
refine the ranges of increase in S-j which can be reasonably expected to
be accommodated without system upset. For the present it seems reason-
able, although somewhat conservative, to conclude that at detention times
of six to eight hours one can allow increases in S-j of approximately 100%
without expectation of significant loss of efficiency during the tran-
sient stage.
In Figure 20 the immediate metabolic response to a shock load in S-j was
shown. This is the portion of the response which may eventually be
depictable by some sort of kinetic model (for a recent attempt see the
article by Young , et al., 34). There is, however, the possibility of
ecologically important aftereffects which are manifested, not during the
transient, but after the system has reached an apparent new steady state.
These were uncovered during a recent investigation by Thabaraj and Gaudy
(35), and Figure 21 is reproduced from that report. The data of par-
ticular interest in this figure are those in the lower graph showing
biological solids concentration and substrate. This completely mixed
continuous flow system (once-through) was operated at a dilution rate of
0.125 hr-' (detention time of eight hours) with S-j = 1000 mg/1. It had
attained a relatively steady condition and S-j was then changed to 2000
mg/1. A successful response to the change was registered, i.e., the
biological solids "rose to the occasion" and S remained steady. The
immediate (primary) response was one of successful accommodation to the
100% increase in S^, and it appeared that a new steady state condition
was being approached. However, between the thirtieth and fiftieth hours
after changing S-s, there was a severe disruption in the steadiness of X
and S, and the C6D in the effluent (filtrate) rose to nearly 800 mg/1.
The situation was not self-correcting until the seventieth hour after the
shock (i.e., nearly nine reactor detention times). This secondary
response was accompanied by a noticeable shift in predominating species
88
-------
- 300
00
LLJ c
O E
IS
z a:
200
z *:
uj -
DO-
'-CARBOHYDRATE
T TV-IT-
PROTEIN
—o-o—taWratrrn |-
-o-o-t-
- FEED = 2000mg/C GLUCOSE
20
40
60
80
100 120
140 160 ISO 200 220 240 260
TIME, hrs
280 300 322
Figure 21. Response of a continuous flow, completely mixed system of the once-
through type to a quantitative shock load consisting of a two-fold increase in
Si. The dilution rate was 0.125 hr-1 (35).
-------
of microorganisms. The secondary response is one for which there may be
little hope of developing any sort of reasonably reliable mathematical
or "mechanistic" models, at least in the relatively near future. The
possibility of this type of response does emphasize the need for caution
and some conservatism with regard to engineering guidelines for allow-
able limits of increase in S-j.
In many cases the approximate allowable limit we have tentatively recom-
mended, i.e., up to 100% increase in S^, does not seem unduly difficult
to attain by a combination of on-line procedures and installation of an
equalization basin at the plant site.
Qualitative Shock Loads
Over the years, pollution control engineers have become accustomed to
thinking of the incoming organic loading, i.e., BOD, as a specific enti-
ty. When it changed, as under shock loading conditions, one had more
(or less) of the same kind of organic matter. Also, because of the diver-
sity of the microbial population, as well as the diversity in the kinds
and amounts of organic compounds in the wastes, one could aptly envision
the ecosystem as one in which each species metabolized that portion of
the incoming waste to which it was acclimated. Thus all component sub-
strates in the waste were thought to be metabolized concurrently.
However, early studies employing mixed (heterogeneous) microbial popu-
lations in experimental systems in which two compounds comprised the
organic substrate indicated that the presence of one substrate could
prevent or hinder the metabolism of the other compound even though the
population had previously been acclimated to it (36). In accord with
the general beliefs previously outlineds this would not have been
expected to occur with mixed populations.
It was known from basic research investigations with pure cultures that
certain enzymes required to metabolize specific substrates are not pres-
ent in the cell at all times (i.e., they are not constitutive enzymes)
but are produced only when needed (i.e., they are inducible enzymes).
The required inducible enzymes are thus produced by the cell in response
to the presence of the particular organic substrate for which they are
needed. The time required to turn on the genetic synthesizing machin-
ery represents the acclimation period. If the organism is not geneti-
cally coded for this particular machinery, it cannot make the requisite
enzymes and it cannot use the substrate. This is one of the ways in
which changes in predominance of species are brought about in natural
populations when changes in the available substrates occur. Whether
an organism has the ability to acclimate biochemically determines its
fitness for survival in the adaptation process. When one develops an
"acclimated sludge," both selection of species (adaptation of the mix-
ed population) and acclimation can occur. One of these processes, bio-
chemical acclimation or synthesis of inducible enzymes, is an intra-
cellular response, whereas the other, adaptation, is an inter-cellular
90
-------
or inter-species phenomenon. Both are important processes which are set
in motion in response to qualitative shock loadings. In response to the
qualitative change in S-j, the cells in the reactor can acclimate, or
there may be an adaptation of the population or a combination of both.
Both mechanisms of response may overlap or go on concurrently, with or
without a disruption in system efficiency, i.e., leakage of organic
substrates. On the other hand, these responses may occur in sequence.
Waste streams usually do not completely "turn over" during shocks; some
of the former substrates are present with the new, and one would think
that the former carbon stream would continue to be removed while the sys-
tem attempted to accommodate to the new. This is, of course, one possi-
bility, but experimental observations indicate that the reverse situation
can also occur. That is, the new substrate may prevent (repress) con-
tinued synthesis of the enzymes required for metabolism of the former
carbon source(s). The control mechanism by which the synthesis of one
group of enzymes is blocked and another triggered has become known as
metabolite repression, and a detailed discussion of this phenomenon is
beyond the scope of the present report. However, at this point one may
wonder concerning the functioning of the enzymes which have already been
synthesized; cannot these continue to function? There is, after all,
quite a large amount of acclimated sludge in the reactor when the change
in Si occurs. Should not these organisms be expected to continue func-
tioning on the old substrates even though the new substrates may have
prevented the continued production of enzymes? This was a question we
posed as investigators and, in the process of experimentation, we came
to the conclusion that there was, in addition to the blockage of the
synthesis of enzyme systems, i.e., metabolite repression, another more
immediately acting mechanism by which the functioning of the already
existing enzyme system may be inhibited (either partially or completely)
by the presence of new incoming substrate.
While the discovery of this type of control mechanism for enzyme sys-
tems in oxidative metabolic reactions contributed to the knowledge in the
basic field, it also was of practical significance in the area of waste
water pollution control, and a considerable experimental program was
undertaken in our laboratory in efforts to elaborate both basic and
applied insight into the phenomenon and the factors which might tend to
attenuate or to exaggerate the magnitude of disruption ensuing because
of changes in composition of the incoming waste. A considerable amount
of these results have been published (3, 8, 36-46). However, much of
the work remains to be published, and even more research is needed on
this type of shock.
In view of the infinite variety of qualitative changes in inflowing sub-
strate which can occur, it is impossible to predict the effects of such
changes upon purification efficiency. It is, however, possible to pro-
vide some guidelines regarding factors which seem to attenuate disruption
of system efficiency due to qualitative shock loadings. In some cases,
the age or maturity of the sludge seems to provide an attenuating effect
91
-------
on the degree of interference which one substrate may exert on the removal
of another or, in any event, on the manifestation of deleterious effects.
Thus freshly grown, highly active cells, while admittedly exhibiting in-
creased substrate removal capability, may be more susceptible to a change
in the substrate composition than the older ecosystem which evolves upon
longer retention of cells in the system. The employment of completely
mixed reactors also helps to attenuate qualitative shocks since the change
in incoming substrate is diluted into the system thus tending to permit
more time for the system to generate a response (metabolic and ecologi-
cal). The equalization basin, which was recommended for smoothing out
variations in flow rate and concentration of substrate, will also be use-
ful in preventing abrupt changes in the composition of the waste.
Changes in Chemical Composition of the Inflowing Waste Other Than the
Organic Substrates
The incoming waste can undergo significant chemical changes involving, not
the amount or kind of carbon source, but other chemical constituents which
may affect metabolism of the carbon source. Among changes of this type,
the effects of introducing toxic components, either inorganic or organic,
are readily appreciated, and it is also readily appreciated that the best
way of handling such situations is to require treatment of waste streams
containing toxic components at the source of the waste water production
prior to their entry to the waste stream influent to the biological treat-
ment plant. Various wastes which would in general be toxic may be sub-
jected to biological treatment if the toxfcant stream is steady or con-
trollable. For example, several decades ago phenolic wastes were consid-
ered toxic, but it is now well known that such compounds are amenable to
biological treatment. An activated sludge can be developed for phenol
metabolism, but the population "natural" to this environment can be ex-
pected to be a rather restricted one and would be predominant only if
phenol was the predominant carbon source in the waste. If phenol were
not the predominant steady carbon source, but were to come on and off the
waste stream as a shock loading, it could inhibit many of the species
developed in response to the dominant carbon source(s) and it would not
be readily metabolized because the relatively fewer species which can
metabolize phenol would not likely be present in the sludge when the shock
occurred, i.e., the sludge would have to re-adapt to it each time. The
same generality might be applied for considering inorganic toxicants, e.g.
cyanides, chromium, etc. Sludges can be developed which exhibit increas-
ed tolerance to some of the inorganic toxicants, but again only when the
system is fed a rather steady diet of the toxicant.
Certain normally nontoxic inorganic components, when applied to the sys-
tem in high concentration as shock loads, can also cause severe disrup-
tion of substrate removal efficiency before the system can adjust. So-
dium chloride is an example of compounds in this category. We have ob-
served that activated sludges developed on relatively fresh water are con-
siderably more resistant to shocks of high sodium chloride concentration
92
-------
than are sludges developed at high sodium chloride concentration to
shocks with fresh water. Here again, it might be expected that the
sludges developed in high salt concentration contain a narrower selec-
tion of species and, in general, a selection of species which may
require salt (rather than, or in addition to, being tolerant to it).
Thus, when the salt concentration is drastically reduced, e.g., to the
fresh water level, these species die and undergo lysis, and during the
interim transient period of adaptation, substrate leakage occurs. Since
sodium chloride is not disruptive to biological metabolism at relatively
low concentrations, only a rather drastic change would cause a deleter-
ious transient response. For example, in a once-through, completely
mixed system operating with an 8-hour aerator detention time, the effi-
ciency of substrate removal was disrupted for nearly two days when the
sodium chloride concentration in the reactor was increased to 30,000
mg/1 over a period of one day (see Figure 1, reference 47). However, an
increase to 8000 mg/1 sodium chloride in the same period caused no dis-
ruption in substrate removal efficiency (see Figure 5, reference 47).
It cannot be said that changes of the same magnitude, or even smaller
changes, involving other inorganic constituents of the waste would not
have more deleterious effects. In general, the effects of changes in
the concentration and types of inorganic constituents of the carriage
water should be the subject of continued investigation as the neces-
sity for exertion of greater technological control of the aqueous
environment increases.
Changes in pH also constitute chemical alterations of the growth envi-
ronment which can cause severe transient responses. It is rather well
known that systems can be adapted to pH levels considerably below or
above neutral (i.e., pH 7.0) when the pH is held relatively steady.
Here again it should be kept in mind that most biological forms function
best at or near a neutral pH. When one develops an activated sludge at
either a high (e.g., 8.5-9.5) or low (e.g., 4-5) pH, he restricts or
narrows the diversity of species which can exist in the system and
thereby may be narrowing the diversity or range of successful ecological
response to all types of changes in environment, i.e., to al 1 types of
shock loadings, which that system can accommodate. If one can be sat-
isfied that the particular waste stream in question will be subject to
only small variations in chemical composition, pH, etc., operation of
activated sludge processes at either high or low pH values may save
some costs of neutralizing chemicals and devices for their addition,
etc. However, pH control and the monitoring of pH in the process are
among the more readily facilitated engineering controls which can be
applied either manually or automatically, and the control of pH either
at or near neutrality is certainly recommended from the standpoint of
optimum function of the plant in steady state and in response to shock
loading as well as from the standpoint of possible deleterious inter-
action between the waste and the materials of construction. A system-
atic study of response of completely mixed systems to changes in pH has
been made in our laboratories and these results will be prepared for
publication. The present aim is to provide guidelines for allowable
93
-------
variation in pH which should not produce serious disruption of system
efficiency. In this regard, our studies indicate that for systems oper-
ated in the "steady state" at neutral pH (6.5-7.0) a change in pH, i.e.,
a shock load, consisting of either an increase or decrease of one pH
unit can be accommodated with little or no disruption of the metabolic
efficiency of the system. Systems can recover from more severe pH
shocks but a transient leakage of substrate can be expected to occur as
well as retardation of the flocculation and sedimentation of the bio-
mass, especially in the case of acid shocks. In general, acid shocks
can be expected to be more deleterious than alkaline shocks.
Changes in Environmental Conditions not Involving Chemical Composition
of the Waste
Two major types of system shock loadings in this category are thermal
shocks and hydraulic shocks. These, like the environmental changes dis-
cussed previously, engender biochemical and ecological responses within
the activated sludge ecosystem which, if the shock is great enough, can
lead to substrate leakage from the process. Concerning these types of
shock, systematic studies of the transient behavior of completely mixed
systems have been accomplished in our laboratory (more extensively with
regard to hydraulic shocks), and these results will be readied for pub-
lication in the research literature. In this report we will attempt to
provide some approximate guidelines for allowable limits of change which
can be expected to cause minimal or no leakage of substrates. As with
all types of shock loading, it is rather difficult to provide simple
dependable rules of thumb because shock loadings may be imposed in combi-
nations and what one observes concerning any specific type of shock may
not hold when that type of shock comes on to a system in combination
with another. Also, differences in the immediate past history of the
system can engender different responses to the same shock. For example,
the dilution rate (or detention time) at which the system is operating
prior to a certain change in temperature may certainly affect the re-
sponse. Also the steady temperature at which the system was operated
prior to the change may affect the response. Furthermore, for a given
dilution rate the rapidity with which the change in the reactor temper-
ature occurs can affect the response. Thus the response to environ-
mental changes is an area which cannot be overly simplified, and there
are many facets to construction of a reasonably reliable set of concep-
tual guidelines for design and/or operation.
Regarding changes in temperature, we have examined the responses of
completely mixed systems operated at detention times of four and eight
hours at a steady state temperature of 25°C. Increases and decreases
in temperature were applied. At an 8-hour detention time, the system
could accommodate, with no excessive leakage of substrate, an increase
of 10° (25 to 35°C) and with some, but not severe, substrate leakage,
a decrease from 25 to 18°C. These were not immediate changes in tem-
perature. On the downward side the drop in temperature was slightly in
94
-------
excess of 1 per hour and on the upward side somewhat less than 1 per
hour. At the lower detention time, a change from 25 to 35°C could be
accommodated without substrate leakage, but a change in temperature from
25 to 18°C caused a significant and long-lived leakage of substrate.
Thus the bio-mass can respond somewhat more successfully to shocks of
increased temperature than to those leading to temperature decreases.
This seems reasonably predictable since increases in temperature (within
reasonable biological limitation) increase the rate of metabolic reac-
tions, thereby engendering more rapid response. Also, microorganisms
capable of growth between 25 and 40° (mesophiles) are much more common
than those capable of growth below 20°C (psychrophiles). Shocks con-
sisting of increases to temperatures above 35°C could be assimilated by
the system but not without severe transient leakage in substrate. As a
rough tentative guideline, it would seem reasonable to recommend for a
system operating normally at or around 25°C that the reaction liquor
should be protected as much as possible from daily variations (i.e.,
changes in temperature taking place over a 24-hour period) of ± 10°C,
with particular precautions being necessary to prevent sudden drastic
decreases in temperature.
Since the reactor volume V is fixed, a change in F causes a change in D
(or in detention time) and, as was seen previously, D is related to the
exponential growth rate, y. Thus a change in the rate of flow of incom-
ing waste imposes a change in y. How the system adjusts itself to this
hydraulic change determines the response, i.e., the loss or maintenance
of substrate removal efficiency during the transient period between
initial and final steady states. There is a limit to the change in F
from which the system can recover, since if D exceeds Uraax for a suf-
ficient period of time, the solids will be completely diluted out. If
a system which was formerly operating at a given (steady) flow rate
with substrate concentration, S^, is now shocked by an increase in F
while S-j remains constant, it can be seen that not only has D been
changed, but the overall average (daily) organic loading has also been
increased simply because the rate of inflow of organic matter has been
increased. The system has not undergone a quantitative shock loading
(change in substrate concentration) but the daily organic loading has
definitely increased. This is a rather severe type of "purely" hy-
draulic shock loading. Often hydraulic shocks are accompanied by a
concomitant decrease in S^, and at times the shock may be accompanied
by an increase in S^. We have examined responses of completely mixed
systems to both general types of hydraulic shock under a variety of
conditions (with various synthetic wastes employing specific carbon
sources and mixtures of carbon sources, changes in dilution rate, etc.).
In accordance with our purpose of providing tentative guidelines for
accommodation of shocks on the basis of our experimentation, it is
recommended that the process be protected from fluctuations in F (with
no change in substrate concentration) larger than 100% for design
reactor detention times of approximately eight hours. That is, if a
system is being operated at a detention time of eight hours on the basis
of the design average daily flow, protection (by way of an equalization
95
-------
basin or a bypass of flow to another unit) against a two-fold increase
in F (leading to a decrease in detention time from eight to four hours)
should be provided in the interest of ensuring reliable and steady
delivery of the design substrate removal level. This precaution cannot
be expected to provide absolute protection against disruption of the
system but it can be expected that in this range of change in F (pro-
vided there is no change in S^, temperature, etc.) the transient leakage
of S will be rather short-lived and the system will rather rapidly
recover its former steady state level of S. Again, as a rough guide for
design detention times less than eight hours, changes in F somewhat less
than 100% should be protected against, and for design detention times
greater than eight hours, changes in F somewhat higher than 100% may be
allowable.
Summary
This chapter has served as an introduction to various types of shock load-
ings involving chemical changes, both quantitative and qualitative, of
the incoming carbon source, changes in the chemical composition of waste
water constituents other than the carbon source and physical changes such
as temperature and volumetric rate of flow. The behavioral response of
the bio-mass (activated sludge) in completely mixed continuous culture
reactors is a complex subject, and in-depth understanding of transient
kinetics may rightfully be a subject of continuing basic environmental
engineering research. Before any sort of useful mathematical models for
the transient response will be forthcoming, the molecular and ecological
physiology of these systems will need to be penetrated. These problems
are of enormous future interest to the entire area of engineering manage-
ment of the aqueous environment. However, the most immediate concern of
the design engineer is not the ways and means to predict transient re-
sponses, but ways and means to prevent deleterious transient responses
leading to substrate leakage and guides to the extent of environmental
change (shock loading) the process can accommodate without the need for
preventive procedures. Thus we have attempted to provide rough but
quantitative (and conservative) guideline limits for changes which the
system can be expected to accommodate without serious metabolic upset.
From our experimental observations, it appears that at reasonable design
aerator detention times (six to eight hours), periodic increases in S^
of 100% of the average daily design concentration can be accommodated
without serious or long-lived leakage of substrate in the effluent. Con-
cerning changes in F, one can expect that hydraulic overloads represent-
ing a change of 100% in the average daily design flow (with no change of
organic carbon concentration, S^) can be accommodated at detention times
of approximately eight hours. Where variations larger than these are
expected or cannot be controlled in-line or at the waste source, the use
of an equalization tank ahead of the aeration tank seems advisable to
smooth out peaks above this range. Equalization tanks or basins for
high volume plants (e.g., those in large metropolitan areas treating
combined municipal and industrial discharges) may not be practical,
96
-------
depending upon the variation in flow. However, the range of change in
flow rate for larger systems is considerably less than for smaller ones.
Effective control of recycle solids concentration, XR, can also provide
a useful tool for attenuating effects of shock loadings. One relatively
reliable and easily understood concept is that, in general, the higher
the concentration of cells, the more organic matter can be removed in a
fixed time. Activated sludge plants are usually operated with much
higher concentrations of biological solids than of substrate. For
example, an organic concentration in the waste of 200 mg/1 "BOD" and a
concentration of biological solids in the aeration tank of approximately
2000 mg/1 are not uncommon. Even allowing for the fact that we cannot
say that all of the biological solids represent substrate users, a sub-
strate to cell ratio of this magnitude indicates that activated sludges
are essentially run under starvation conditions. This situation repre-
sents a potential reserve capability which can be drawn on during
increases in S-j. The reserve capacity that is found in excess metabolic
capability might also be expected to be beneficial during increases in
F which decrease the reactor detention time. However, it must be
remembered that the hydraulic shock loading which is applied to the
aeration tank is also passed along to the secondary clarifier, and the
decreased settling time may be more of a problem than the decreased
aeration time. It may not only cause loss of solids in the effluent,
but may also decrease the solids concentration in the sludge recycle
flow. Thus the combined use of equalization basins and facilities to
control XD (see Figure 16) can help to steady the system under both
normal and shock loading conditions.
97
-------
VII
PROCESS MODIFICATIONS FOR NITROGEN-DEFICIENT WASTES
Nitrogen Supplementation for Biological Treatment
In previous chapters, removal of organic material from waste waters and
growth of microorganisms have been treated as interdependent processes.
It was stated that it is the usual practice in biological treatment to
add nutrients such as nitrogen and phosphorus to wastes deficient in
these elements to ensure that the carbon source is the limiting nutrient.
This procedure has been recommended for maximum purification efficiency
and it is a logical procedure since removal of the oxygen-demanding
organic components of the waste has traditionally been, and still is,
the foremost goal of biological treatment. The recommended levels of
supplementation are based upon the ratios of carbon, nitrogen, and
phosphorus required by microorganisms for "balanced" growth. This means
simply that if one considers synthesis of microbial cells, as described
in Chapter III, as the essential mechanism by which organic material is
removed from the waste, the amounts of nitrogen and phosphorus required
are dependent upon the amount of organic carbon which is to be removed.
Based upon studies of optimum nutritional conditions for purification,
ratios of 20:1 for BOD:N and 100:1 for BOD:P are commonly accepted and
supplementation is provided to achieve these ratios if necessary.
The addition of nutrients represents a continuous operational cost which
may amount to a significant fraction of the total cost of operation.
Any means of reducing the supplementation required could therefore allow
a significant reduction in the operational cost. An additional reason
for re-examination of the practice of adding excess nutrients is the
recent concern over stream enrichment, i.e., the concern that leakage of
nitrogen and phosphorus from treatment plants may contribute to exces-
sive algal growth in receiving waters. While much of the nitrogen and
phosphorus in many waters is undoubtedly contributed by natural and
agricultural runoff and while no nitrogen is required for growth of
blue-green algae which are capable of using atmospheric nitrogen, it is
important that all possible sources of nutrient addition be curtailed.
For wastes which contain nitrogen in excess of the required BOD:N ratio,
e.g., municipal wastes, the problem is removal of the excess nitrogen.
However, wastes which are initially deficient in, or devoid of, nitrogen,
e.g., many industrial wastes, present a different problem. For these
wastes it is necessary to avoid adding excess nitrogen which may leak
from the system while maintaining maximum purification efficiency with
regard to removal of organic material. In a laboratory pilot plant or
an industrial fermentation process using a pure culture and medium of
known and constant composition, it would theoretically be possible to
exactly balance the amounts of carbon and nitrogen sources so that each
would be maximally consumed. However, on a practical basis using whole
wastes which vary in composition both qualitatively and quantitatively,
99
-------
this is an impossible achievement, and it is necessary to consider other
means of avoiding leakage of added nutrients.
Studies of the applicability of the Monod equation (equation 18) to the
relationship between nitrogen concentration and growth rate of hetero-
geneous populations have shown that the relationship provides a good
approximation of the data (48). Values for maximum growth rate, ymav?
were comparable to those measured in studies of carbon source limitation
(see Chapter IV). The values of Ks ranged from 1.5 to 4.0 mg/1. Based
on comparison of amounts of carbon source and nitrogen source required
for maximum specific growth rate determined in these and other studies,
a ratio of BOD:N of 25:1 appeared to be optimum for maximum growth rate.
This indicates a nitrogen requirement only slightly less than that found
in earlier studies using different methods (49).
The studies cited above were performed in batch systems. Rather exten-
sive studies of the effect of nitrogen level upon the performance of
once-through, continuous flow, completely mixed systems have also been
made in our laboratory (50). The laboratory pilot plant was operated
at various dilution rates with a synthetic waste containing 1060 mg/1
glucose COD, and three levels of nitrogen supplementation were employed.
These studies showed that purification efficiency and leakage of nitrogen
in the effluent depended not only upon the level of nitrogen supplemen-
tation, but also upon the dilution rate. The nitrogen content of the
sludge was also determined in these studies since this parameter appears
to be related to the purification capability of the sludge under some
conditions (49, 51).
Figure 22 shows the variation in effluent COD at three levels of nitro-
gen supplementation and various dilution rates, and Figure 23 shows
ammonia-nitrogen levels in the effluent under the same conditions. It
is apparent from Figure 22 that at longer detention times there is much
less effect of the COD:N ratio on purification efficiency than there is
at shorter detention times. At a detention time of twelve hours, fairly
good purification was achieved, even in a once-through system, at a
COD:N ratio as high as 70:1. (In a system with solids recycle, one
might expect greater purification efficiency under otherwise equal con-
ditions.) Very little additional purification efficiency was gained by
decreasing the COD:N ratio from 40:1 to 25:1. With increasing detention
time, nitrogen leakage increased sharply at a COD:N ratio of 25:1, while
no leakage was observed at any of the dilution rates at a COD:N ratio
of 70:1.
Figure 24 shows the relationship between nitrogen content of the sludge
and purification efficiency at four COD:N ratios and various dilution
rates. As these data show, even with severe nitrogen limitation, i.e.,
at a COD:N ratio of 70:1, the nitrogen content of the sludge can vary
depending upon dilution rate. It is also apparent from this figure
that the same purification efficiency may be achieved with sludges of
100
-------
1000
CP
Q
O
O
LJ
h-
<
DC
\-
00
GO
z>
CO
h-
2
UJ
U.
LU
800
600
400
200
0
15 20 25 30 35
NH3-N CONCENTRATION IN THE FEED,
40
45
Figure 22. Relation between effluent substrate concentration, S, and NH3 -
concentration in the feed at three COD:N ratios for dilution rates ranging
from 1 to 1/12 hr'1. Feed COD was 1060 mg/1 (50).
-------
o
to
<*
\
E
UJ
Z)
_J
u_
u_
UJ
UJ
X
8.0
6.0
o
h-
cr
LU
O
I
I
ro
X
4.0
2.0
0
10
COD/N=70/lb'
|COD/N = 40/I
7
p
D=I/I2 HR
D= I/8HR
-i
D=I/4HR"'-
fCOD/N=25/l
15 20 25 30 35 40 45
NH3-N CONCENTRATION IN THE FEED, mg/Ji
50
Figure 23. Relation between effluent and influent concentrations of NH3 - N at three
COD:N ratios for dilution rates ranging from 1 to 1/12 hr~l (50).
-------
UJ
345678
NITROGEN IN BIOLOGICAL SOLIDS, %
Figure 24. Relation between purification efficiency, nitrogen content of the sludge,
dilution rate, and feed COD:N ratio (50).
-------
quite different nitrogen content when other design parameters are var-
ied. For example, 90% COD removal might be chosen as the desired
purification. From the figure, one can determine that this degree of
purification could be achieved with: a 12-hour detention time at a
COD:N ratio of between 70:1 and 40:1 and a sludge nitrogen content of
2.5%; an 8-hour detention time, COD:Nratio of 40:1, and sludge nitrogen
content of slightly above 4%; or a 4-hour detention time, COD:N ratio
between 25:1 and 10:1 and sludge nitrogen content of 9%. However, con-
cern over stream enrichment dictates that consideration also be given to
efficient utilization of nitrogen, and the data shown in Figure 23 indi-
cate that the lower COD:N ratios which enhance sludge nitrogen content
also enhance nitrogen leakage in the effluent. Therefore, it should be
emphasized that no single factor such as a standard ratio of nitrogen to
BOD or COD or sludge nitrogen content can be chosen as the basis for
design. Both COD and nitrogen utilization must be considered in choos-
ing the design detention time and the level of nitrogen supplementation
to be employed. It should also be pointed out that a system in which
solids are recycled would be expected to require less nitrogen than a
once-through system where cells are growing more actively and that
nitrogen leakage would probably be greater in a system with recycle.
Treatability studies carried out in the design stage should include
determination of effluent nitrogen levels. Especially for industrial
wastes which are not devoid of nitrogen but contain available nitrogen
at a level below that usually considered optimum, i.e., wastes for
which the BOD:N ratio is higher than 20:1, treatability studies should
be designed to minimize or possibly eliminate nitrogen supplementation.
For wastes which contain little or no nitrogen, a completely different
mode of treatment may be preferable.
Our studies of nitrogen requirements for the activated sludge process
suggested the possibility that substrate removal might be accomplished
by oxidative assimilation without the addition of nitrogen. Concurrent
studies on the kinetics of purification in relation to initial biolog-
ical solids level (52, 53) showed that the COD: solids ratio is impor-
tant in determining the mechanism of substrate removal. These findings
led us to suggest a possible means of utilizing oxidative assimilation
to effect a savings in nitrogen supplementation and avoid the possi-
bility of nitrogen leakage while maintaining purification efficiency
(54). The new modification of the activated sludge process which we
have proposed for nitrogen-deficient wastes is described below. This
rather radical departure from traditional methods of treatment of such
wastes has been described in a series of publications (54-60) and some
of the data obtained in laboratory pilot plant studies will be pre-
sented herein. Before discussing the process itself, it is desirable
to describe oxidative assimilation and to discuss some of the basic
kinetic studies which led to the proposal of the new process.
104
-------
Oxidative Assimilation
In a system containing metabolizable organic compounds and other required
inorganic nutrients, but devoid of an external supply of nitrogen, rep-
lication of microbial cells cannot occur because synthesis of essential
proteins, nucleic acids, and some lipids requires nitrogen. However,
many bacteria are capable of converting a large fraction of the avail-
able organic material into internal storage products, while oxidizing
the remainder of the organic matter to provide energy for the synthesis
of the storage product and for other activities of the cell. The spe-
cific product synthesized varies with the bacterial species. The most
common internal storage product is a polymer of glucose which is very
similar to glycogen, the storage product synthesized by the human liver.
Other organisms synthesize a lipid polymer, poly-3-hydroxybutyric acid.
These storage products may accumulate to a level of 50% of the dry
weight of the cells. Some organisms, if inorganic phosphate is present
in excess, also form a polyphosphate which serves as a store of both
phosphate and energy rather than as a carbon source store. The storage
of polyphosphate is undoubtedly responsible for much of the biological
phosphate removal which can occ'jr on prolonged aeration of activated
sludge if the organisms which possess this ability are present in suf-
ficient numbers. Other organisms may synthesize external polysaccharide
which is deposited as a capsule or slime layer outside the cell wall.
Upon addition of nitrogen to the system, internal storage products are
utilized for growth.
Gaudy and Engelbrecht (61) studied removal of COD by activated sludge
in the presence and absence of an external nitrogen source and showed
that oxidative assimilation could occur in such systems, i.e., COD was
removed in the absence of a nitrogen source, and the organic material
was stored primarily as carbohydrate. These studies were carried out in
batch at a solids:COD ratio of approximately one, using sludge harvested
from the effluent of a continuous flow system. Interestingly, analyses
of protein and carbohydrate content of the sludge throughout the period
of aeration in batch showed that both systems,,with and without nitrogen
added, initially synthesized carbohydrate. In the system containing
nitrogen, the carbohydrate content of the sludge later decreased while
the protein content increased.
In later studies of the effect of initial biological solids concentra-
tion upon the kinetics of COD removal and solids synthesis, it was
found that as initial solids level was increased in batch systems, the
autocatalytic curves for both COD and solids typical of low initial
solids systems (see Figure 1) became more nearly linear. That is, re-
actions which exhibited first-order increasing kinetics at low initial
solids concentrations approached zero-order kinetics or, in some cases,
first-order decreasing kinetics, when high initial solids were used
(52, 53). Since replication of microbial cells is an autocatalytic proc-
ess, simply because replication occurs by division of one cell to form
two cells, the kinetics observed in these studies were atypical for
105
-------
replicating systems. This suggested that replication (cell division)
actually does not occur to an appreciable extent during the period of sub-
strate removal in a batch system when the ratio of initial solids to COD
is high. Krishnan and Gaudy (53) showed that, at an initial COD:biolog-
ical solids ratio of approximately 0.5, COD was efficiently removed in
the presence of an inhibitor which completely prevents replication of
cells by preventing synthesis of proteins. Figure 25 (taken from this
publication) shows the course of COD removal and synthesis of biological
solids, both of which approximate zero-order kinetics during most of the
substrate removal period. These data demonstrate that replication of
cells is not necessary for COD removal. Sludge synthesis in this case
does not represent synthesis of new cells but rather an increase in the
weight of the cells due to synthesis of storage products, primarily car-
bohydrate as shown by the analysis of the sludge. More detailed studies
of rates of COD removal and synthetic reactions during the period of sub-
strate removal in relation to the initial COD:solids ratio, provided more
complete evidence that as the initial COD:solids ratio is decreased, COD
removal and sludge synthesis follow more similar patterns during the
period of COD removal in systems with and without nitrogen (62). In
short, the mechanism of COD removal in batch systems containing high
initial levels of biological solids is primarily oxidative assimilation,
not cell growth. Subsequently, if nitrogen is present, the stored car-
bohydrate disappears and protein is synthesized, i.e., the storage prod-
ucts are converted to new cellular components required for replication
(see Figures 3 and 4 of reference 53, and Figure 3 of reference 62).
This behavior of high solids batch systems may be explained as follows:
When microbial cells are placed in fresh medium, replication is not
initiated immediately unless the cells are from a population actively
growing in the same medium. In most cases, e.g., when a culture is
grown to maximum numbers and a portion is transferred to fresh medium,
a period of adjustment is required before replication can begin. Dur-
ing this period, the lag period, the cell synthesizes the proteins and
other cellular components required for growth under conditions which
differ from those in which the cells had previously existed. The longer
the period between cessation of growth and transfer to fresh medium,
the greater is the amount of adjustment required, in part because
enzymes no longer used once the food supply has been exhausted may be
degraded by endogenous metabolic reactions. The cells used in the
batch experiments described above were taken from batch activated sludge
systems maintained by feeding once in twenty-four hours with one-third
of the sludge and two-thirds of the supernatant wasted each day before
feeding (52, 53). The wasted sludge was used for experimentation.
This sludge was certainly not composed of young, actively growing cells.
Thus, when the sludge was contacted with fresh waste, it was not capable
of immediate replication. However, it was capable of removing COD, as
our studies showed, and indeed it is necessary that COD be removed to
supply energy for the formation of cellular components required for rep-
lication. As our studies have also shown, the sludge is, in addition,
capable of forming storage products which it may later use as a carbon
106
-------
2400
BIOLOGICAL SOLIDS
TOTAL COD
GLUCOSE COD
SLUDGE
CARBOHYDRATE
SLUDGE PROTEIN
6000
5000 ?\
o>
E
4000
3000
2000
1000
2.0
TIME, HOURS
0
Figure 25. Substrate removal, sludge accumulation,
and sludge composition in the absence of protein
synthesis in a batch reactor (53).
UJ
&
ir
Q
>-
x
o
DO
o:
<
o
UJ
\-
cc
CL
co"
Q
Zi
o
CO
107
-------
source for synthetic reactions. Thus, sludge synthesis (increase in
weight of biological solids, X) is not impeded but replication of cells
is delayed. If only small numbers of cells are present (low initial
biological solids), a significant amount of COD cannot be removed dur-
ing the lag phase and COD removal is dependent upon formation of new
cells, giving rise to the autocatalytic curves for substrate removal
typical of a replicating system. However, if a sufficient number of
cells is present in relation to the available carbon source, all of the
carbon source can be removed even before replication begins. If large
numbers of cells are present, the carbon source can be exhausted during
the oxidative assimilation phase which precedes replication, and there-
fore replication may not occur, and is not required, for COD removal at
low COD:solids ratios. However, for this process to be repeated, it is
necessary that a nitrogen source be available so that the storage prod-
ucts may be utilized for synthesis of normal cell components. In batch
systems containing very high solids, this conversion can take place
after the purification phase is complete, i.e., after the carbon source
(COD) has been taken up by the cells.
Since the purification of the waste, i.e., removal of COD, and the syn-
thesis of nitrogen-containing cellular components can occur sequentially
even in the presence of nitrogen, it should be possible to separate these
phases physically. It seemed that several advantages might be realized
by such a treatment scheme for nitrogen-deficient wastes. Accordingly,
studies were undertaken to determine the feasibility of the process
described in the following section.
Proposed New Process for Treatment of Nitrogen-deficient Wastes
The flow sheet for the proposed method of treatment is shown in Figure
26. After primary settling or other required pre-treatment, the
nitrogen-deficient waste flows into the feeding aerator, or activated
sludge tank, where it is contacted with a high concentration of recycled
sludge in a completely mixed reactor. In contrast to the usual mode of
operation, no nitrogen source is added, although phosphorus should be
added if necessary. The waste is purified by oxidative assimilation;
that is, the organic material is removed from the waste and converted
to intra-cellular storage products. The mixed liquor exiting the
aerator is settled in the usual manner (clarifier) and the supernatant
is discharged as plant effluent. At this point, the settled sludge
differs from the usual activated sludge in that it contains a higher
percentage of carbohydrate and possibly lipid, and a lower percentage of
protein. The effluent should be free of nitrogen, since the nitrogen-
starved cells would remove any trace of nitrogen present in the waste.
The waste has been purified without nitrogen having been added.
Before the sludge can be recycled to the aerator, its ability to con-
vert external organic carbon to internal storage products must be regen-
erated by allowing the cells to utilize the material already stored for
108
-------
NITROGEN DEFICIENT WASTE
RECYCLED SLUDGE
(HIGH PROTEIN CONTENT)
FEEDING
AERATOR
CLARIFIER
FINAL
EFFLUENT
ENDOGENOUS
AERATOR
SETTLED SLUDGE
(LOW PROTEIN CONTENT)
EXCESS SLUDGE
NITROGEN SUPPLY
Figure 26. Proposed flow sheet for the treatment of nitrogen-deficient industrial
wastes by the continuous oxidative assimilation modification of the activated sludge
process (54).
-------
synthesis of the nitrogen-containing compounds (proteins, nucleic acids,
and some lipids) which are required for continued metabolic activity and
for replication. However, only that portion of the sludge which is to
be recycled requires regeneration, and the amount of nitrogen required
is therefore less than that usually added to allow unrestricted synthe-
sis of nitrogenous compounds by all of the sludge in the aerator. The
portion of sludge to be recycled is pumped to a second aerator, called
the endogenous aerator because no external carbon source is present.
Nitrogen is added, as NH/1", and the stored carbon source is used by the
cells for synthesis of nitrogenous compounds, largely protein, and for
replication. This regenerated sludge, now depleted of storage products
and high in protein content, is recycled to the feeding aerator, where
it again removes COD by oxidative assimilation.
Initial tests of the feasibility of this mode of treatment were carried
out in batch systems using a synthetic waste containing glucose as the
sole organic carbon source (54). Parallel systems were operated in an
identical manner except that nitrogen (as NH4+) was added to the feed-
ing aerator in one system and to the endogenous aerator in the other.
The necessity for regeneration was also tested by recycling sludge from
two endogenous aerators, to only one of which nitrogen was added. All
sludges were analyzed during each aeration period (feeding or endog-
enous) for carbohydrate and protein content. These studies showed that
the physical separation of waste purification and replication of cells
was possible in practice. COD was removed with approximately equal
efficiency in the presence and absence of nitrogen in the feeding
aerators, and was stored primarily as carbohydrate by the sludge in the
aerator to which nitrogen had not been added. In the endogenous aeration
period, the stored carbohydrate was converted to protein in the presence
of ammonia-nitrogen, and the recycled regenerated sludge removed COD as
efficiently as the recycled sludge which had been treated in the usual
manner by addition of nitrogen to the waste before treatment. Without
the regeneration step, the sludge rapidly lost its ability to remove COD.
Since glucose is rather easily polymerized to form the glycogen-like
carbohydrate storage product, it was important to test the feasibility
of the process for other carbon sources. Batch studies (56) were carried
out using synthetic wastes containing a non-carbohydrate, acetic acid,
and lactose, a compound which is metabolized by inducible enzymes. The
use of lactose was designed to determine whether inducible enzymes neces-
sary for oxidative assimilation would be degraded during endogenous aer-
ation in the absence of the carbon source. The process operated suc-
cessfully with both synthetic wastes.
The feasibility of the process for continuous flow operation was studied
in a laboratory-scale pilot plant (Figure 27) using two synthetic wastes
and a whole waste. In the synthetic wastes, glucose (57) and acetic
acid (58) were used as carbon sources, and the whole waste was a sugar
refinery plant effluent obtained from the Imperial Sugar Company refin-
ery at Sugarland, Texas (55). The nitrogen supplement for the endogenous
110
-------
.Pump
Waste Holding Tank
Sludge
Recycle Pump'
Feeding Phase
Aerator
Regenerated
Sludge Holding
Tank
Waste Sludge
Effluent
Diluent for sludge
consistency
NH/
Endogenous
Aerator
Figure 27. Laboratory-scale pilot plant employed in operational feasibility
studies for the continuous oxidative assimilation process (55).
-------
aerator was calculated as a ratio of the COD added to the feeding
aerator, and various ratios were used to determine the minimum require-
ment for the process. These data have been published in detail in the
references cited above. With nitrogen-free glucose synthetic waste
(1060 mg/1 COD), 95% removal of COD was accomplished with a 4-hour
detention time in the feeding aerator at a COD:N ratio of 70:1 (approx-
imately equivalent to BOD:N = 50:1). Sludge wasted was 17%, and no
NH4+ could be detected in the effluent. With acetic acid (1060 mg/1 COD)
at the same COD:N ratio and feeding aeration time, COD removal was 90%.
Under the same conditions, 94% of the COD was removed when the COD:N
ratio was 50:1. The sugar refinery waste varied considerably in COD
content, and no attempt was made to adjust its concentration since this
variation allowed assessment of the ability of the process to withstand
quantitative shock loads. The COD of the waste varied from 380 to 1375
mg/1, and BOD from 271 to 841 mg/1. Approximately 35% of the BOD of the
waste was carbohydrate. At an aerator detention time of five hours, a
COD:N ratio of 60:1 and a COD:P ratio of 140:1, good purification effi-
ciency was maintained. No serious disruption resulted from the quanti-
tative shock loads applied to the system, and no MM/*" was detected in
the effluent. Sludge wastage was approximately 32%.
In all continuous flow operations, as shown in Figure 27, two aerators
were used for the regeneration and recycling of sludge. These were
operated as batch systems. Each twelve hours, sludge was withdrawn from
the clarifier underflow, diluted with effluent to a constant concentra-
tion (to maintain XR constant), and a portion was wasted. The remainder
was aerated for twelve hours in the presence of NH^+ at a concentration
based on the amount of COD fed to the feeding aerator during the pre-
ceding twelve hours. The regenerated sludge was then pumped to the aer-
ated holding tank, from which it was recycled at a constant rate during
the succeeding twelve hours while a new batch of sludge was being regen-
erated. Preliminary studies of required regeneration time indicated that
the detention time in the endogenous aerator could be reduced.
Figure 28 shows typical data for effluent analyses during continuous flow
operation of the oxidative assimilation process with synthetic waste con-
taining acetic acid as the sole carbon source (58). The hatched area
represents the synthesis of storage products. The synthetic waste con-
tained no nitrogen, and the ratio of COD:N used as a basis for nitrogen
supplementation in the endogenous aerator was 70:1. The efficiency of
solids removal in the clarifier was approximately 96%. These data dem-
onstrate the potential of the proposed process for treatment of nitrogen-
deficient industrial wastes.
There is some obvious similarity in the flow sheets for the oxidative
assimilation process and the biosorption, or contact stabilization
process, and it might be possible in some cases to convert biosorption
plants to operation as oxidative assimilation processes. Both types of
treatment involve dual aeration tanks; however, relative detention times
for the first and second aeration periods may be quite different. The
112
-------
M
U)
FEED FLOW = 400 mi/hour
RECYCLE FLOW = 200 mi/hour
MIXED LIQUOR DETENTION TIME = 4hrs
BIOLOGICAL SOLIDS
BIOLOGICAL SOLIDS DUE TO RECYCLE
CELL CARBOHYDRATE
CELL PROTEIN
EFFLUENT COD (FILTRATE)
i i i i i i
8 10
TIME, DAYS
12
14
16
18
Figure 28. Operational data for treatment of nitrogen-free acetic acid waste by continuous
oxidative assimilation process with nitrogen supplementation in the endogenous aerator at a
COD:N ratio of 70:1 (58).
-------
fundamental difference between the oxidative assimilation process and
biosorption lies in the point of addition of nitrogen. In the bio-
sorption process, nitrogen is added to the waste before the carbon
source is removed, and in oxidative assimilation, the carbon source is
removed before nitrogen is added. In the latter process, nitrogen is
added only to the sludge, not to the waste, and only to that portion
of the sludge which is to be recycled. In the biosorption process,
nitrogen cannot be used efficiently and much of the added nitrogen is
likely to be lost in the effluent. This occurs because, as our studies
have shown, at very high solids:COD ratios, such as those used for bio-
sorption, the carbon source is removed and stored in the cells before
synthesis of nitrogen-containing compounds occurs, even if nitrogen is
present. Thus, it is likely that most biosorption plants actually
remove carbon source largely by oxidative assimilationand that much of
the added nitrogen is wasted because it is made available at a time
when the cells cannot utilize it efficiently. The oxidative assimila-
tion process eliminates leakage of nitrogen in the effluent and allows
a considerable reduction in the cost of nitrogen supplementation.
Summary and Conclusions
Studies of the kinetics of substrate removal showed that at high ini-
tial ratios of biological solids to COD, nitrogen utilization and cell
replication occur primarily after the COD has been removed. In these
systems COD removal is largely accomplished, even in the presence of
nitrogen, by oxidative assimilation. The carbon source is removed and
converted to internal storage products, which are then used for syn-
thesis of the nitrogen-containing compounds necessary for formation of
new cells. Since carbon source assimilation and nitrogen utilization
occur sequentially in the cell, it is possible to separate these
processes physically in a new flow sheet proposed for treatment of
nitrogen-deficient industrial wastes.
The proposed flow sheet, shown in Figure 26, utilizes two aerators. In
the first, or feeding, aerator, nitrogen-free waste is treated by oxi-
dative assimilation. A nitrogen-free effluent is produced since no
nitrogen is added to the waste. The sludge to be recycled is regener-
ated by addition of ammonia-nitrogen in a second (endogenous) aerator.
A considerable savings in the cost of nitrogen supplementation is
realized, since a COD:N ratio much higher than that used in traditional
methods of treatment can be employed. Laboratory-scale pilot plant
studies using several synthetic wastes and a sugar refinery waste have
demonstrated the feasibility of the proposed continuous oxidative
assimilation process for treatment of nitrogen-deficient industrial
wastes.
114
-------
VIII
EXTENDED AERATION PROCESS
The present document is concerned with conceptual principles relevant to
the•purification of wastes, i.e., with metabolic occurrences taking
place largely in the aeration chamber. It is not our intention to cover
the topic of sludge disposal, although ultimate disposal of organic or
biological sludges is a vital concern in the water pollution control
field. Most engineers are, however, aware of the "extended aeration"
process or the "total oxidation" process, in which biological treatment
of the incoming waste is effected via an activated sludge and, concur-
rently with this function, 'the activated sludge is subjected to aerobic
digestion—the end result supposedly being net conversion to C02 and HoO
of all of the organic matter originally in the waste stream or at least
of that portion of it which is biologically treatable. Concurrent waste
water purification and aerobic digestion are accomplished by returning
all sludge to the aeration chamber. The organic matter in the sludge is
subject to wet biochemical combustion, and is dissipated as C02 and F^O
just as when it is ultimately disposed of by chemical combustion or
anaerobic digestion and burning of methane. Thus, if the approach to
ultimate disposal is recycle of the waste organic carbon by converting it
to innocuous inorganic carbon, C02> combustion, i.e., consumption of
oxygen, is inevitable. Since in an activated sludge process the oxygen-
ation energy which is supplied to the system is much in excess1 of the
metabolic requirement, the attractiveness of accomplishing both treat-
ment of the waste and wet combustion of the sludge developed during
treatment can be readily appreciated.
This unique modification of the activated sludge process was suggested
by Forges and his co-workers as a consequence of their research on the
biological treatment of dairy wastes (63). It was suggested that if the
biological solids were continually returned to the aeration chamber (no
wasting of excess solids), endogenous respiration would tend to decrease
the solids concentration by about the sa'me amount that the new incoming
carbon source in the waste would tend to increase it, and eventually a
state of equilibrium might be approached at which there would be no net
solids accumulation. Thus, the dual function of biological waste water
purification and sludge digestion (aerobic) could be accomplished con-
currently and without recourse to disposal of any biologically produced
sludge.
At the time the process was suggested, the term "endogenous respiration"
was subject to various interpretations; it was at that time a rather new
term to the field of biological waste treatment. Interpreted solely as
the intra-cellular oxidation of organic carbon, one could argue against
the theoretical possibility of total oxidation, since it is obvious that
a given microbial cell cannot oxidize itself totally to inorganic car-
bon. Also, it could be reasoned that if one cell died, not all of its
115
-------
organic carbon might serve as available sources of carbon for another
species; thus the validity of the concept was suspected. In any event,
the suggested process modification precipitated much interest and
engendered much research activity by other investigators (e.g., 64, 65).
Two general conclusions emanated from these investigations. First, it
was concluded that certain component parts of the biological cells syn-
thesized in the purification phase were nonmetabolizable. Thus there
was a fraction of organic matter which was inert from two points of
view: it was dead organic matter, therefore incapable of removing sub-
strate, and it could not serve as food material for live cells. As time
went on and this material was retained in the system, the bio-mass would
thus consist of an ever-increasing inert fraction, and eventually the
system must fail. The other conclusion, which was somewhat related to
the first, was that a condition of biological solids equilibrium could
not be attained, i.e., the biological solids concentration would con-
tinually build up. It was generally felt that biological solids had to
be wasted, or the system could not function. The concept of total oxi-
dation was generally concluded to be theoretically untenable, and bio-
logically unsound. Regardless of these conclusions, many extended
aeration plants have been built, and operational experience has varied
regarding their worth and the need to waste biological sludge period-
ically.
Recent Investigations of Extended Aeration
A few years ago, based upon some of the experimental observations made
during investigations pertinent to our research on "kinetics and mech-
anism of activated sludge processes" there was, we believed, some cause
to doubt the finality of the previous conclusions concerning the possi-
bility of total oxidation. Also, the paradoxical situation in the
field wherein extended aeration plants were being installed in opposi-
tion, as it were, to the conclusions of some engineering researchers,
certainly seemed to require some reconciliation. Also, there were no
definitive data on how long an extended aeration activated sludge might
continue to function until it experienced failure, i.e., until it no
longer could function as an effective means of removing soluble organic
matter from waste waters. It was reasoned that even if, after a year or
so, some sludge had to be wasted and new bio-mass developed, the process
would still be advantageous since continuous operation of a sludge treat-
ing and disposal facility would not be necessary. Thus it might have
considerable utility and applicability to treatment of various industrial
wastes. All of these considerations contributed to a decision to enter
upon long-term investigations of the process.
In order to test the theoretical possibility of attaining total oxida-
tion, i.e., to determine whether continual buildup of biological solids
containing an ever-increasing fraction of biologically inert solids
occurred, it was necessary to devise a laboratory pilot plant scheme
which could guarantee that all biological solids were retained in the
116
-------
system. The problem of inadvertent loss of biological solids in the
effluent had been experienced by other workers, and was one of the fac-
tors which originally caused us to doubt the validity of some of the
conclusions which had been drawn from their results. The pilot plant
and operational techniques we employed are described in detail in the
research literature (66-68). Briefly, positive control over retention
of biological solids in the pilot plant system was maintained by centri-
fuging all effluent from the clarifier and returning the solids to the
aeration tank. Thus, even at times when solids could have exited a more
usual pilot plant, they did not exit the one which we operated. A
small amount of solids was removed for analysis. Such removal amounted
to no more than 0.2 per cent of the bio-mass.
The organic feed to the unit consisted of a "synthetic waste" in which
the carbon source was a soluble carbohydrate (glucose). It is impor-
tant to consider why a simple soluble compound such as glucose was
employed, since it has been the experience of researchers that some in
the pollution control field have a tendency to discount work with "syn-
thetic wastes." Some of the pros and cons of this and other approaches
to research in the environmental field have recently been presented by
Gaudy (69). In the present instance, it is important to point out that
the crux of the question concerning the theoretical unsoundness of the
total oxidation process was the supposed non-biodegradability of various
fractions of the microbial cells of which the activated sludge is com-
posed, e.g., the extra-cellular polysaccharide slime layer (64). It is
essential, then, to make sure that the activated sludge one is examining
does consist wholly of microbial cells. Obviously, entrapped organics
such as coffee grounds, and inorganics such as silt, etc., are relatively
or completely inert to aerobic biological treatment. It is emphasized
that the previous conclusion that the principle of total oxidation was
unsound was based upon experimentations employing simple soluble carbon
sources, e.g., acetic acid (64), and glucose (65). While we are in dis-
agreement with the conclusions of these workers, we agree that their use
of simple soluble carbon sources was wholly justified.
If one retained all of the biological solids in the system, and if there
were a permanently inactive organic fraction, the biological solids con-
centration should continue to increase; i.e., one would expect continual
accumulation of biological solids. Also, if bacterial polysaccharide in
the cell capsule or slime layer were biologically inert, the carbohydrate
content of the sludge would gradually increase. The ability of the acti-
vated sludge to remove the carbon source in the waste would gradually
decrease. Another indication of an increasing inactive fraction of the
sludge might be a decrease in basal rate of respiration (02 uptake rate
per unit of sludge under endogenous conditions, i.e., in the absence of
an extra-cellular source of carbon). These and other aspects were inves-
tigated during a three-year period of continuous pilot plant operation
with centrifugation and total solids retention.
It was found that the biological solids concentration did not reach an
117
-------
equilibrium, or steady, value in which "endogenous respiration" balanced
new synthesis. However, neither did the biological solids concentration
continually increase. There were periods of solids increase and periods
of decreasing biological solids concentration. During a decreasing
cycle in biological solids concentration, the organic matter (COD) in
the effluent did not rise, as would have been the case if some cells had
undergone lysis and the organic matter thus released were not metabo-
lized by the remaining intact cells. One of the most dramatic downward
cycles occurred during the latter part of the first year of operation
(67). Figure 29 shows the operational behavior of the system during this
cycle. From day 285 to 307, the biological solids concentration decreased
from approximately 8500 to 2400 mg/1. Despite this rather sharp drop,
the filtrate COD (effluent) did not rise appreciably. In fact, there was
only a slightly perceptible rise with a rapid return to the previous
level. The COD of the synthetic waste was approximately 530 mg/1 (500
mg/1 glucose), and the filtrate COD was for the most part less than 40
mg/1 except for the short-lived transient as the solids began to de-
crease, when it rose to approximately 100 mg/1.
Since solids were not wasted (purposely or accidentally), and since they
decreased without any concomitant rise in COD, and since no external
operational changes were imposed upon the system, this downward cycle in
biological solids was a natural one brought about by the system itself.
During such a decreasing cycle in biological solids concentration with
no rise in effluent COD, the most apparent explanation is that a por-
tion of the sludge (microbial population) is serving as a source of car-
bon (substrate) for another portion of the microbial population. In
short, total oxidation of the organic material in the waste must occur
by a combination of the reactions shown in equations 3 and 13, with
recycling of the carbon through the latter. It cannot be expected that
the amount of new growth on the incoming substrate will always be bal-
anced by cannibalism of the past growth so as to produce a steady or
equilibrium level of solids. The predominance of species is constantly
shifting, and it is necessary that the ecology of the system be such
that some of the cells extant in the system act as food material for
other cells which play the role of feeders. Neither the periodicity nor
the extent of cyclic accumulation and de-accumulation of biological sol-
ids can be predicted, but the results shown in Figure 29 surely attest
to the occurrence of such periods.
It should also be pointed out that during more than 1000 days of oper-
ation there was no evidence for the buildup of carbohydrate content in
the sludge. Samples taken for measurement of protein and carbohydrate
content of the sludge close to the end of the pilot plant operation
[e.g., see Figure 5, reference (68)3 snow that the carbohydrate content
was approximately 20%, which represents a normal value for microbial
cells.
Many batch feeding tests to determine the substrate removal capability
of the sludge were periodically performed (67, 68). Figure 30 shows
118
-------
EFFLUENT CHARACTERISTICS
O SUPERNATANT COD
VBIOLOQICAL SOLIDS
D FILTRATE COD
255
260
265
270
275
280
290
300 305 310
TIME ,DAYS
335
340
345
350
355
Figure 29. Performance data for a continuous flow extended aeration activated sludge in a
laboratory pilot plant operated with total solids recycle, showing a period of accelerated
autodigestion (67).
-------
Q
O
O
CTJU
600
(
400,
200
o
DAY
I
N
i
722 MAR 21,1969 |
r'
JIOLOGIC
E.TRATE COD
-JL_,
AL SOL
1 >
IDS
^
j
CO*
g
O
CO
_J
o
1
g
CD
to
O
800
600
o>
. 400
o
o
o
200
DAY
k
N
N
899 SEPT. 14. 1969 |
^-B
( I
N
OLOGIC
(
\L SOL
ILTRATE COD
^, A '
IDS
4
1
15 30 45 60 75
TIME, MINUTES
X
o>
E
CO
g
o
co
<
o
o
3
g
CD
800
600
o>
o
O
o
1400
200
DAY 987: DEC. II, I969~|
B.IOLOGICAL SOLIDS
A I
FILTRATE COD
10
'O
16
CO
Q
14
12 g
o
o
DO
15 30 45 60 75
TIME, MINUTES
Figure 30. COD removal capability of extended aeration activated sludge
after the number of days of continuous operation with total solids
recycle indicated on each graph (68).
-------
the results of four such experiments. Even after 987 days of oper-
ation with no sludge being wasted, the system removed the daily feed,
added as a slug dose, within thirty minutes. Considering the fact that
the retention time in the aeration chamber was sixteen hours under con-
tinuous flow conditions, the capacity for substrate removal was far in
excess of that required even after nearly three years of aging. This
would not have been possible had the sludge not possessed the ecological
and biochemical capacity for self-renewal, i.e., internal feeding. It
is important to point out that the basal respiration rate (endogenous 0?
uptake rate; mg Oz/hr/gm sludge) was considerably lower than for a
younger microbial population. This is an indication that the sludge was
somewhat less active than young cells on the basis of unit weight. For
example, the endogenous uptake rate for this sludge appeared to level off
between 1 and 2 mg/hr/gm, as is shown in Figure 31. Young populations
freshly grown from a small inoculum of cells from the aeration chamber
manifested basal respiration rates approximately ten times higher. The
rising trend in the values between days 300 and 350 occurred immediately
after the period of accelerated autodigestion (decline in solids) shown
in Figure 29. During and after such a period of declining biological
solids concentration, there is a larger fraction of younger cells in the
sludge because of the higher substrate concentration (i.e., the canni-
balized cells), and as a consequence of this physiological condition,
higher endogenous 0^ uptake rates were manifested.
The facts that biological solids did not continually accumulate, that
the carbohydrate content did not attain abnormal proportions, and that
the system continued to deliver excellent substrate removal efficiency
after three years of continuous operation with total cell recycle, bring
us to the conclusion that the conceptual principle of the total oxidation
or extended aeration process is not theoretically unsound, and work
designed to gain a more definitive insight into the metabolism of cell-
ular components is continuing.
In addition to the long-term studies just described, we have run other
experiments in closed batch systems which demonstrate that the total
oxidation of the bio-mass is possible. Results of one such experiment
are shown in Figure 32 (62). In this experiment, a relatively low ini-
tial concentration of acclimated sludge was fed 2280 mg/1 glycerol COD
and various parameters were monitored during the ensuing growth and
"endogenous" periods. The top portion of the figure shows changes in
protein and carbohydrate content of the sludge as well as ammonia
nitrogen, which was the only source of nitrogen in the system. The
lower portion of the figure is particularly important for present con-
siderations, because it can be seen that the concentration of biologi-
cal solids, for all practical purposes, returned in the prolonged endog-
enous phase to the initial concentration; thus there was no "permanent"
synthesis product. Essentially total oxidation of the assimilated car-
bon was effected in this closed batch system.
121
-------
(O
to
b
UJ
S5
3
_l
W)
. 4
E
o
tt.
£ 3
(T
§2
«
§
£
V
' A
^
A
/ V
- / x
/ \
/ - *
V
\
A ^4
^>«.
A
*.^. A
A A
A
•
^
" A
A
A
••a— A A
—d-*-*--*-
~K-
250 300 400 500 600 700 800 900 1000 110
TIME, DAYS
Figure 31. Effect of aging on endogenous oxygen uptake by an extended aeration acti-
vated sludge during long-term pilot plant studies employing total solids recycle (68).
-------
1200
12
20 60 100
TIME, hrs
400 600 800
Figure 32. Total oxidation, during the endogenous
phase, of the biological solids synthesized during
the substrate removal phase in a batch system using
a heterogeneous population developed from a sewage
seed (62).
123
-------
Engineering Control of Total Oxidation
Because the total oxidation process has been shown to be conceptually
sound does not automatically confer upon it practicability or engineer-
ing feasibility, but removal of the onus of conceptual unsoundness from
the process allows engineering considerations to bring theory to prac-
tical fruition. No one can predict how long a total cell recycle sys-
tem will continue to accumulate biological solids before the ecological
(predator-prey) relationship is such as to initiate a period of accel-
erated autodigestion. No one can predict the extent of the ensuing
decrease in biological solids concentration. All that can be said is
that if the cells can be retained in the system, periods of de-
accumulation will occur. At times, solids may accumulate to very high
concentrations, thus making retention difficult with currently employed
field procedures. It seemed essential, therefore, to explore engineer-
ing possibilities for attaining control of biological solids concen-
tration, especially ways to initiate the autodigestive process.
When cells are totally oxidized, all portions of the cell must be con-
verted to simple inorganic compounds which include C02» H20, NH3, and
small amounts of inorganic salts. Even for readily metabolized com-
pounds, approximately five passes through the food chain would be
required for reduction of the original amount of carbon source to the
desired level of purification. This is based on an assumption of 50%
average yield for metabolism, i.e., the assumption that approximately
50% of a substrate is oxidized and 50% converted to new cell material.
When the newly formed cell is oxidized by other cells, if the yield is
again 50%, the amount of original carbon source still present as cell
material is 25%, and each succeeding pass reduces this by a factor of
one-half. For components of the cell which are metabolized with diffi-
culty, long periods of time may be required for complete oxidation.
Initiation of oxidation of cellular components may depend upon access to
them by other cells which are capable of utilizing them. Some of the
carbon in a microbial cell is in soluble form, but a large part of it is
incorporated in larger insoluble macromolecules in the cell membrane, cell
wall, and extra-cellular capsular material, e.g., polysaccharide,-and in
the proteins and nucleic acids within the cell. Some of the predatory
activity which occurs in the sludge involves the ingestion of particulate
organic matter, e.g., ingestion of some species of bacteria by some spe-
cies of protozoa and other predatory microorganisms, and other activity
involves the use of organic carbon of some species of bacteria by other
species of bacteria. This organic substrate must be transported into the
cell if it is to be metabolized. Therefore, the first and probably the
most difficult metabolic task is the breaking down of the organic macro-
molecules, i.e., their hydrolysis to smaller fragments which can be taken
into the cell. It is indeed a very complex ecology which is needed for
all of these steps to occur. Our results have demonstrated that they do
occur; i.e., these biological occurrences are not theoretically impos-
sible. However, one cannot "engineer" the ecology of the system, at
124
-------
least not at present. We are continuing studies to gain insight into
understanding this useful natural occurrence (a prerequisite to attempts
to engineer the natural ecosystem).
There are ways to approach control of the system other than learning to
exert environmental control of the ecosystem. One such channel, which
we felt warranted investigation, was to combine aspects of chemical and
biological digestion of the sludge. One of the mildest digestive chemi-
cal operations and the easiest to perform is hydrolysis, the preparatory
step to biological digestion. If one provided such an engineering assist
to the system, the autodigestive process might be subjected to direct
engineering control. With the hydrolysis portion of the digestion proc-
ess subject to engineering manipulation, the heterogeneous microbial pop-
ulation might be better able to metabolize that portion of the sludge
which had been hydrolyzed and returned to the aeration compartment.
This process, like the one discussed in the previous chapter, would rep-
resent a rather radical departure from current practice or concept, but
it seemed a plausible one in view of our previous findings and was con-
sistent with the general aim and need for innovation in the pollution
control field. Experiments were initiated to test this concept.
Figure 33 shows the course of COD removal when hydrolyzed sludge was
fed to the extended aeration pilot plant in a batch experiment similar
to those shown in Figure 30 in which the synthetic waste (glucose) was
fed. It is seen that the hydrolyzed sludge was metabolized rather rap-
idly (compare with Figure 30) by the extended aeration activated sludge.
These hydrolyzed cells had been rendered essentially soluble by acid
hydrolysis (pH = 1) for five hours at 15 psi (1.05 kg/sq m) and 121°C.
The hydrolyzed sludge was neutralized to pH 7 and fed as a slug dose to
the pilot plant.
On the basis of such results, we have proposed the modification of the
extended aeration process embodying the concept of the hydrolytic assist
shown in Figure 34 (68). The type of extended aeration process shown is
the compartmentalized aeration chamber-clarifier with internal sludge
recycle (dotted arrow). The lower portion of the flow diagram shows the
hydrolysis process. This portion of the plant can be considered as an
aid to aerobic sludge digestion in which a portion of the total return
sludge is solubilized before return. In the extended aeration process,
the return sludge itself is really a waste stream being fed to the sys-
tem like the incoming waste. The modification proposed in Figure 34
simply prepares a portion of the sludge so that it becomes a more read-
ily available substrate, thus allowing some engineering control over the
biological process. With the chemical hydrolysis modification, one is
provided with a sludge disposal aid which need be used only when requir-
ed, and for which there is no long startup time. The hydrolysate could
be stored without neutralization, thus preventing odors, etc., and
neutralized as it was channeled back to the aeration chamber at a rate
controlled and selected by the operator. Such a system uses both chem-
ical and biological methods of sludge disposal to their best advantage.
125
-------
to
600
500
BIOLOGICAL SOLIDS
Q
0
40 50
TIME, MINUTES
20
480
Figure 33. Metabolism of hydrolyzed activated sludge by an extended aeration
activated sludge (68).
-------
RAW
WASTE
AERATION
TANK
SETTLING
TANK
^HYDROLYSATE
RECYCLE
HYDROLYSATE
NEUTRALIZATION
TANK
EFFLUENT
EXTENDED AERATION
PROCESS
HYDROLYSIS PROCESS
HYDROLYSIS
UNIT
Figure 34. Proposed extended aeration activated sludge process incor-
porating chemical hydrolysis for control of sludge concentration (68).
-------
It performs chemically a function which is biologically difficult, and
performs biologically a function which is difficult and costly to perform
chemically. Additional investigational work is warranted and is in
progress; to date we have successfully operated the bench scale pilot
plant with periodic sludge hydrolysis, and shortly a report on these
operational experiences will be prepared for publication in the techno-
logical literature.
It may be noted in Figure 34 that "raw waste" flows into the aeration
tank. The figure was drawn in this way because it is generally the
case that where extended aeration is employed, an attempt is made to
require the process to perform the function of primary treatment as well
as secondary treatment and sludge disposal, i.e., there is no primary
settling tank. In the "traditional" flow sheet for treatment of muni-
cipal wastes, secondary and primary sludges are combined and channeled
to anaerobic digestion, so it is probably natural that with the extended
aeration process, both types of sludge would be expected to be combined
for aerobic digestion in the aerator; i.e., it is natural to look upon
the non-soluble organic material in primary sludge as a substrate just
as the return sludge, the microorganisms, are considered as a particu-
late substrate in the extended aeration process. However, one does not
expect such panaceas with other activated sludge processes, and there
is no real justification for expecting an extended aeration activated
sludge to totally oxidize primary sludge. The probability is that it
will not. The extended aeration process is best considered as a sec-
ondary biological treatment process. Also, it is ideal for waste con-
taining primarily dissolved organic matter. In this regard, it is bet-
ter suited to treatment of some industrial wastes than to municipal
sewage (where it has been largely employed). However, with the hydro-
lytic assist process we are recommending, it may be possible that some
particulate organic matter in the raw waste which becomes entrapped in
the microbial sludge (and, indeed, is partially metabolizable) can be
hydrolyzed under the same conditions which solubilize the cells. In
this case, some of the organic matter which normally would be removed by
primary settling is treatable "chemo-biologically." Some organic com-
pounds in "primary" municipal sludge can undoubtedly serve as excellent
microbial substrates with or without chemical hydrolysis. However, con-
stituents such as coffee grounds, for example, would be difficult to
hydrolyze chemically or to metabolize aerobically; thus for municipal
sewage some organic residue should be expected, i.e., total oxidation of
such sludges may not be possible. However, the residual organic sludge
which would have to be disposed of should be low in amount and, by
virtue of its resistance to both chemical hydrolysis and biological
attack, should be relatively non-putrescible. Thus, addition of the
hydrolytic assist should allow the extended aeration process to be very
usefully employed even for municipal wastes. The non-hydrolyzed sludge
would have to be separated from the hydrolysate prior to channeling the
hydrolysate to the aeration chamber. Removal of finely divided suspend-
ed organic matter in the hydrolysate could be accomplished with chemical
addition (coagulation, flocculation, and settling) following the hydrolysis
128
-------
unit. This would also remove finely divided inorganic matter which may
pass the grit chamber and be entrapped in the sludge. It should be
apparent that suspended matter which is initially inert, i.e., non-
biodegradable, should remain so after passage through the aeration cham-
ber, after possible entrapment in the bio-mass and after hydrolysis.
Indeed, the hydrolysis modification including chemical precipitation of
permanently inert material originally in the waste stream provides a
useful way of separating the pollutional (treatable) from the non-
pollutional material in the waste.
Summary
In summary, based upon long-term and thorough studies, we are able to
conclude that the extended aeration or total oxidation process is not
conceptually unsound. The proposed modification of the process embody-
ing the hydrolytic assist offers an effective way to combine ideal
features of chemical and biological treatment for concurrent treatment
of the waste water and essentially total aerobic digestion of sludge
developed in the process.
129
-------
IX
SUMMARY AND RECOMMENDATIONS
At the outset it was emphasized that this document was not intended to
serve as a manual for design and operation of activated sludge process-
es, but as a conceptual guide for updating and unifying mechanistic and
kinetic principles which bear strongly on design and operation. The
real aim was to help close the so-called "gap" between research and
practice, and this we have attempted to do by relating, largely in
readily understood engineering terms, many of the findings of our funda-
mental research effort and the concepts which have been developed over
the past decade. While our research has been, and in large measure will
always be, designed to gain better basic understanding of biological
processes, our ultimate aim is the useful application of the informa-
tion. Thus, throughout this document, presentation and discussion of
conceptual principles have been accompanied by embodiment of these prin-
ciples in recommended procedures, models, process innovations, etc., for
the consideration of the profession.
Beginning with generalized concepts of exertion of biochemical oxygen
demand and use of ACOD as the measure of biologically treatable organic
material, through discussion of the stoichiometry of the treatment proc-
ess and development of growth kinetics, incorporation of the kinetic
relationships into a relatively simple model for description of com-
pletely mixed growth reactors and, finally development of a modified
kinetic model specifically for the activated sludge process, the approach
has been to foster better understanding and to provide recommendations
leading to engineering control over the biological treatment process.
The model finally evolved in Chapter V, equations 43 through 47, embodies
the concepts developed in the previous chapters, and represents an engi-
neering modification of basic kinetic theory. The modification, i.e.,
holding XR rather than c as a system constant, complicates the kinetic
equations somewhat but allows for practical attainment of greater stead-
iness in X and S. The goal in biological treatment is attainment of
steady, reliable delivery of the needed degree of process efficiency,
and the recommended model for design and operation should foster this
aim. The graphs of Figure 18 or 19 can be used as a rough guide to de-
sign, although the engineer should ideally determine values of the bio-
logical constants ^^v, KS» anc' Y for the particular waste in question.
In computing the graphs shown in Figures 18 and 19, values of the phys-
ical constants chosen were XD = 10,000 mg/1, and a = 0.25, which are
reasonably selected values; however, the design engineer can, with the
model, select values of his own choice and can test the effect of his
selection on X and S.
It has been emphasized that if one desires a reasonable biological engi-
neering model for a process, he must provide as best he can in his de-
sign the facilities for control of the parameters included in the model.
131
-------
Thus, means of controlling XR nearly constant at the selected value
should be provided. This requires the incorporation of a sludge con-
sistency tank, i.e., a revision of the "traditional" flow sheet. The
sludge consistency tank is somewhat analogous to the make-up tank for
chemical dosage used in chemical treatment. The sludge consistency
tank is one in which the biochemical "dosage" to the activated sludge
aeration tank is prepared. Also, some attempt to even out fluctuations
in F and S^ may be required either on the line or at the treatment
plant. The use of an equalization basin to accomplish this is recom-
mended where practicable. While equalization basins have been used in
some installations, their use is a departure from the general practice.
Thus, the newer concepts we have introduced for the "normal" activated
sludge process, involve some changes in the usual or traditional flow
diagram. These changes are intended to help the system approach the
"steady state" assumption of the kinetic model.
One should realize that regardless of which processes one employs in
the treatment of waste waters—chemical, physical, or biological, or any
combination thereof—an effort must be made to hold the system steady or
to protect against drastic changes in the operational environment.
After all, chemical-physical treatment plants are just as likely to
undergo stress leading to fluctuations in efficiency, due to changes in
S-j, F, or the amount of coagulant or adsorbent chemicals added, as are
biological treatment plants. For example, one would not think of chem-
ical treatment without control of chemical dosage. In like manner, one
should also provide control of the biological dosage, i.e., XR, or, in
any event, should not let XR fall below a selected level. Thus, build-
ing in the engineering control to make the process function in accord
with the preconceived notion (some sort of design model) is not incon-
gruous with accepted engineering practice. It is probably more true of
biological treatment systems than of chemical-physical ones, that the
system possesses an internal resistance and responsiveness to external
change. For example, the agents for removal of organic matter (the
microorganisms) are self-producing, and do essentially "rise to the
occasion" of high organic loadings, i.e., there is a self-correcting
feature in biological systems which, if not overloaded by abusive exter-
nal change, serves to protect the system and to provide some internal
regulation and restoration of efficiency when disruption does occur.
While the material in Chapters II through VI can be said to pertain to
general understanding of activated sludge processes, and in these chap-
ters some engineering innovations which depart from traditional practice
were introduced, in Chapters VII and VIII more radical departures from
the norm were covered. In Chapter VII, the effect of nitrogen as limit-
ing nutrient and the effect of biological solids concentration on the
predominant biological mechanism of substrate removal were delineated.
The new process modification which we have termed "continuous oxidative
assimilation," has been demonstrated to be feasible, at least in labor-
atory scale pilot plant studies. The process has unique possibilities
for nitrogen-deficient industrial wastes and, in addition, should serve
132
-------
to demonstrate that some of the older concepts and nutritional guide-
lines are by no means inviolable.
It should perhaps be pointed out that the finding that COD may be
removed primarily by oxidative assimilation in batch or continuous flow
systems containing high concentrations of biological solids does not
negate the use of kinetic models for nitrogen-containing wastes, based
on equations for microbial growth, such as those presented in Chapter V.
The continuous oxidative assimilation system is one in which substrate
removal and cell replication are separated artificially. The cells are
forced to carry out oxidative assimilation and protein synthesis in a
completely sequential manner by addition of carbon source and nitrogen
source to separate aerators. The sequential synthesis of storage prod-
ucts and nitrogen-containing cellular components observed in batch sys-
tems, even in the presence of nitrogen, would undoubtedly be observed in
activated sludge systems closely resembling plug flow reactors. It is
not possible to determine whether such sequential synthesis occurs to an
appreciable degree in completely mixed reactors with high solids, i.e.,
in reactors employing cell recycle, because one cannot examine synthesis
in individual cells and it is the newly recycled cells which would be
most likely to carry out oxidative assimilation. We consider it likely
that oxidative assimilation does occur to some degree in these reactors.
Activated sludge is not completely comparable to the usual growing sys-
tems of the microbiologist or fermentation engineer. In the first
place, as was brought out in Chapter V, the recycling of cells forces a
very slow rate of replication, a rate which is only a fraction of the
system dilution rate. Secondly, the recycled cells have an immediate
past history of existence in an environment essentially devoid of car-
bon source (the clarifier). Because of these two reasons, as well as
the general starvation condition of the system, these cells cannot be
classified as physiologically young, actively growing microorganisms.
Therefore, when sludge is returned to the aerator, the cells may remove
carbon source initially by oxidative assimilation, possibly to an extent
depending primarily upon the retention time in the secondary clarifier.
However, in systems in which the sol ids:COD ratio is high, i.e., sys-
tems in which 1 + a - ac is small, the doubling time for the population
may extend over several aeration periods. Therefore, in high solids
systems, a very long period of time is available to the cell to prepare
for replication. For maintenance of a steady state in X in systems with
recycle, it is only necessary that cells replicate at a very slow rate
or, in other words, that a relatively small percentage of the cells in
the aerator replicate during one detention period. Since batch studies
have shown that in high solids systems the rate of increase in X (meas-
ured in units of mass, not cell numbers) is not very different for cells
which are utilizing carbon source in the presence of nitrogen or in its
absence, the occurrence of oxidative assimilation during part of the
cell division cycle of individual cells should have no noticeable effect
on the kinetics of the system.
Chapter VIII deals with the extended aeration process, which incorpor-
ates the important aspects of sludge disposal and secondary treatment of
133
-------
the waste. It was shown that this process is not biologically unsound
and a new modification of the process (the hydrolytic assist), which
allows engineering control of the process, was proposed. This modifi-
cation is a union of chemical and biological treatment which can pro-
vide a useful alternative or replacement for anaerobic digestioir-end/or
incineration of waste organic matter from treatment plants. It accom-
plishes the same thing as these processes, i.e., conversion of organic
carbon to inorganic carbon. In the long range, any method of disposal
which provides for recycle of carbon via conversion of waste organic
carbon to CO? for ultimate reconversion via the photosynthetic process,
may become obsolete, since this process is a slow recycle and the
organic carbon in the waste may need to be recycled for man's use by
more direct means; e.g., it may eventually be a usable byproduct of water
rejuvenation or a unique raw material for production of useful commodi-^
ties. However, in the near future, the conversion of these waste organic
sludges to C02 by whatever means, and the dissipation of C02 into the
atmosphere represents, if not the ultimate, the "penultimate" sludge
disposal method. This is by any means a combustion process, and the
chemically-assisted wet aerobic biological combustion we have herein
recommended for consideration of the profession should be accomplishable
without the remedial and often costly expedients to abate air pollution
problems usually required with other means of disposal.
Figure 35 shows a possible flow diagram for an activated sludge plant
which embodies many of the conceptual principles discussed in this
report. Since there are really no standard wastes, there is no standard
way of treating all wastes. Accordingly, all activated sludge process
schemes should not be expected to look the same, but each should be
tailored to the unique requirements of the particular situation. Unique-
ness and engineering innovation rather than standardization should be the
rule, provided that the engineer can defend the reasonableness of the
concepts and has performed pre-design investigations to justify omission
or addition of any particular unit operation or process in his overall
plant design. While we feel that the flow diagram shown in Figure 35
may offer considerable improvement over the more or less standardized
flow diagrams which have traditionally become associated with activated
sludge processes, we do not offer it as a new standard flow sheet but
as the general embodiment of many of the conceptual principles we have
developed in the previous chapters. It is in a design sense, a tenta-
tive "diagrammatic" conclusion which seems a meaningful result of the
transference of conceptual principles and research findings to function-
al design. It is readily appreciated that this flow diagram differs
from the traditional representation of an activated sludge plant in a
number of ways.
The essential units are shown in solid lines and those which may or may
not be needed, depending upon the particular situation, are shown in
broken lines; the surge basin falls into the latter category. If the
waste flow and strength can be regulated within the guideline limits
previously given without a surge basin, and if drastic changes in the
134
-------
SURGE
BASIN
PRIMARY
CLARIFIER
r—
SECONDARY
CLARIFIER
i—i i i
GRIT L-I-J
REMOVAL '
POLISHING
TREATMENT
N, P, ETC.
| PRECIPITATION
L_
I
INERT
SOLIDS
Figure 35. Possible flow diagram for activated sludge
process incorporating suggested modifications for oper-
ational control of purification efficiency and including
sludge disposal by aerobic autodigestion aided by chem-
ical hydrolysis. This diagram is based on the conceptual
principles and research findings presented in this report,
135
-------
types of organic compounds in the waste are not expected to occur, the
surge basin is not needed. The main aerobic reactor (aeration tank 1),
the clarifier, and the sludge consistency tank (aerator 2) are in accord
with Figure 16, and permit design (and operation) in accordance with the
model equations (43 through 47) presented in Chapter V, holding XR as a
system constant. Also, the flow sheet accommodates the principle of
total (wet biological) oxidation and employs the hydrolytic assist dis-
cussed in Chapter VIII. Thus, disposal of biological solids is effected
via the concept of extended aeration, but the substrate removal and the
autodigestive phases are separated by design in order that biological
solids concentration in aerator 1 and subsequently in the clarifier, can
be controlled within operational limits selected by the design. Also,
separation of the purification (aerator 1) and "endogenous" phases
(aeration tank 3) should enhance control of the latter (i.e., the auto-
digestive) phase.
Into aeration tank 1 flow the major portion of the influent waste, the
recycle sludge XR from aeration tank 2, and a significant portion of the
hyrolysate, and all mixed liquor exits to the clarifier. Into aerator
2, the sludge consistency tank, flow a portion of the clarifier under-
flow, and three additional inflows. One consists of biological solids
from aerator 3; these may be bled in continuously, or as needed to help
maintain constant XR. Both of the remaining inflows consist essentially
of substrates (some of the waste inflowing to aerator 1 and some of the
hydrolysate). These bleeder lines of inflowing carbon source function
essentially as sludge "fresheners," and help pre-condition the biological
solids for reentry to aeration tank 1. The sludge has been existing
outside of the growth environment of aeration tank 1 for some time, and
addition of some of the substrates with which it will shortly be con-
tacted should help to reduce the readjustment or lag period in aerator 1.
A lag period at high solids concentration can force oxidative assimi-
lation and, for wastes which contain ample amounts of nitrogen, it is
best to curtail oxidative assimilation because this process conserves
nitrogen and could possibly increase nitrogen leakage in the effluent.
Aeration tank 3 receives the excess sludge from the clarifier. This
tank can be looked upon as a modified aerobic digester which receives,
in addition to biological solids, some of the incoming waste and some
hydrolysate. The dual function of the aeration tank of an extended
aeration plant is here accomplished in two separate tanks, substrate
removal in aerator 1 and autodigestion of the excess solids in aeration
tank 3. The major portion of the outflow from aeration tank 3 is cycled
through the hydrolysis unit, and a small portion may be recycled through
aeration tank 2 to aeration tank 1.
The hydrolysis unit receives mixed liquor from aerator 3. Also, pro-
vision can be made to pump some clarifier underflow directly to the
hydrolysis unit. For wastes containing settleable organic solids, the
primary sludge can be channeled to the hydrolysis unit, where the
hydrolyzable portion can be liquified and recycled to the treatment
136
-------
plant, i.e., to aeration tank 1, 2, or 3. For waste containing non-
hydrolyzable participate organic matter or suspended inorganic matter,
effluent from the hydrolysis unit can be subjected to chemical treat-
ment (shown in the lower portion of Figure 35) to precipitate this
material prior to recycling hydrolysate to the treatment plant. These
precipitated solids, i.e., non-hydrolyzable particulate organic matter
and suspended inorganic solids are labeled "inert solids." They are not
inert solids produced in the treatment process, i.e., they are not pro-
duced from the biochemically Op-demanding organic matter in the waste
(ACOD), but are materials which were initially present as biologically
inert organic or inorganic matter. It can be expected that municipal
sewage and some industrial wastes would contain some such materials and
the chemical treatment unit could prove useful in effecting separation
of these "inerts" while the chemical hydrolysis-assisted aerobic digestion
is used to provide "ultimate" disposal of the biological solids. The
type of flow diagram shown in Figure 35 is, to be sure, a radical de-
parture from current picturizations of activated sludge plants. This
flow sheet permits incorporation of new principles and there are various
modifications of the flow sheet which could easily be devised.
Figure 36 shows one such modification which uses essentially the same
core unit processes in a somewhat different mode of operation. In the
flow diagram of Figure 35, the hydrolysis unit was set up for continuous
operation. Alternatively (Figure 36), aeration tank 3 and the hydrolysis
unit could be operated essentially as a separate extended aeration proc-
ess which receives a portion of the waste flow, diverted from the inflow
to aeration tank 1, and all of the excess sludge from the clarifier. The
primary difference between the two flow sheets is the inclusion of a
settling compartment ir. aerator 3 in Figure 36. In this case, the hydro-
lytic unit could be used intermittently or continuously for control of
solids level in aerator 3, as discussed in Chapter VIII (see Figure 34).
In addition, aerator 3 could be used to accept increases in flow, thus
helping to buffer aerator 1 against hydraulic shocks (changes in F).
In both flow diagrams (Figures 35 and 36), the portion of the plant
which includes aerator 1, the clarifier, and aerator 2 is designed to
operate according to the kinetic model developed in Chapter V (equa-
tions 43-47). It should also be pointed out that recycling of hydro-
lysate to aerator 1 can be used to conserve nitrogen for wastes which
are low in nitrogen.
While these flow schemes may seem innovative or imaginative, the concep-
tual scheme does not require a "far stretch" of the engineering imagi-
nation. All new conceptual principles require scrutiny and question by
the field and considerable developmental work before being put into gen-
eral practice. The extrapolation of research findings represented by
the flow diagrams in Figures 35 and 36 is presented in order that its
pros and cons can be considered by the whole of the profession.
In closing, let us re-emphasize what was said at the beginning: This
report is presented as an advancement of conceptual principles, surely
137
-------
SECONDARY
CLARIFIER
AERATION TANK 1
SETTLING
COMPARTMENT
Figure 36. One of several possible modifications
of the flow diagram shown in Figure 35. See text
for details.
138
-------
not an attempt to write finis to principles applicable to biological
treatment. The essence of engineering is an ordered and justifiably
cautious change based upon definitive scientific fact. We have docu-
mented the concepts with investigational findings, and extrapolated
the concepts to possible process modifications for further scrutiny by
the field. Our aim has been to improve and, if we can hope to write
finis to anything at all, it is to the "bio-stasis" and dogmatized
concepts with which engineering of the biological process has been
traditionally approached.
139
-------
ACKNOWLEDGMENTS
We are grateful to the Environmental Protection Agency for its sup-
port of this project (17090 FQJ). Moreover, we gratefully acknow-
ledge its support and that of the federal agencies preceding it
(FWPCA and USPHS-NIH) for support of three major research projects
(WP-00075, WP-00325, and WP-00786) which in considerable measure
made possible the present research undertaking. Also, some of the
experimental work which contributed to the concepts herein developed
was supported under project A-017, WRRI-USDI, and this support is
gratefully acknowledged. Our research was accomplished in associa-
tion with many of our MS and PhD graduate students. Some students
were employed as research assistants on the projects cited above,
and others received financial support as graduate trainees (WP-19)
or were institutionally supported. In addition to the technological
information and principles represented by this report, the graduates,
who have helped accomplish and have been helped by accomplishment of
this research, are themselves no mean by-product of this effort. We
gratefully acknowledge their help as students and as colleagues in
the work, and they join us in acknowledging the supporting agencies.
We are also grateful to Mrs. Grayce Wynd for her invaluable assist-
ance in preparing the final manuscript and for the contributions she
has made to the research team over the past decade as our research
secretary.
141
-------
REFERENCES
1. Jennelle, E. M., and Gaudy, A. F. Jr., "Studies on the Kinetics
and Mechanism of BOD Exertion in Dilute Systems." Biotech.
Bioeng., 12, 519-539 (1970).
2. Gaudy, A. F. Jr., "Biochemical Oxygen Demand." In Water Pollu-
tion Microbiology, John Wiley (in press)(1971).
3. Bhatla, M. N., and Gaudy, A. F. Jr., "Sequential Substrate Removal
in a Dilute System by Heterogeneous Populations." Appl. Microbiol.,
13, 345-347 (1965).
4. Busch, A. W., "BOD Progression in Soluble Substrates." Sewage Ind.
Wastes, 30, 1336-1349 (1958).
5. Wilson, I. S., and Harrison, M. E., "The Biochemical Treatment of
Chemical Wastes." J. Inst. Sewage Purification, 3, 261-275 (1960).
6. Gaudy-, A. F. Jr., Bhatla, M. N., Follett, R. H., and Abu-Niaaj, F.,
"Factors Affecting the Existence of the Plateau During the Exertion
of BOD." J. Water Pollution Control Fed., 37, 444-459 (1965).
7. Gaudy, A. F. Jr., Bhatla, M. N., and Abu-Niaaj, F., "Studies on the
Occurrence of the Plateau in BOD Exertion." Proceedings 18th
Industrial Waste Conf., Purdue Univ., 183-193 (1963).
8. Bhatla, M. N., and Gaudy, A. F. Jr., "Studies on the Causation of
Phasic Oxygen Uptake in High Energy Systems." J. Water Pollution
Control Fed., 38, 1441-1451 (1966).
9. Bhatla, M. N., and Gaudy, A. F. Jr., "Studies on the Plateau in
Oxygen Uptake During Exertion of Biochemical Oxygen Demand by Pure
Cultures of Bacteria." Biotech. Bioeng., 7, 387-404 (1965).
10. Bhatla, M. N., and Gaudy, A. F. Jr., "Role of Protozoa in the
Diphasic Exertion of BOD." J. San. Eng. Div., Proc. ASCE, 91,
SA3, 63-87 (1965).
11. Gaudy, A. F. Jr., and Gaudy, E. T., "Microbiology of Waste Water
Purification." Ann. Rev. Microbiol., 20, 319-336 (1966).
12. Symons, J. M., McKinney, R. E., and Hassis, H. H., "A Procedure
for Determination of the Biological Treatability of Industrial
Wastes." J. Water Pollution Control Fed., 32, 841-852 (1960).
13. Hiser, L. L., and Busch, A. W., "An 8-Hour Biological Oxygen Demand
Test Using Mass Culture Aeration and COD." J. Water Pollution
Control Fed., 36, 505-516 (1964).
143
-------
14. Isaacs, W. P=, and Gaudy, A. F. Jr., "Comparison of BOD Exertion
in a Simulated Stream and in Standard BOD Bottles." Proceedings
22nd Industrial Waste Conf., Purdue Univ., 165-182 (1967).
15. Isaacs, W. P., and Gaudy, A. F. Jr., "Atmospheric Oxygenation in
a Simulated Stream." J. San. Eng. Div. Proc. ASCE, 94, SA2, 319-
344 (1968).
16. Gaudy, A. F. Jr., Bhatla, M. N., and Gaudy, E. T., "Use of Chem-
ical Oxygen Demand Values of Bacterial Cells in Waste Water
Purification." Appl. Microbiol., 12, 254-260 (1964).
17. Gaudy, A. F. Jr., Turner, B. G., and Pusztaszeri, S., "Biological
Treatment of Volatile Waste Components." J. Water Pollution Con-
trol Fed., 35, 75-93 (1963).
18. Goswami, S. R., and Gaudy, A. F. Jr., "Removal of Aldehydes and
Ketones by Stripping and by Combined Stripping and Microbial
Metabolism." Proceedings, 4th Mid-Atlantic Industrial Waste Con-
ference, University of Delaware (in press)(1970).
19. Herbert, D., Elsworth, R., and Telling, R. C., "The Continuous
Culture of Bacteria; a Theoretical and Experimental Study."
J. Gen. Microbiol., 14, 601-622 0956).
20. Gaudy, A. F. Jr., Yang, P. Y., Bustamante, R., and Gaudy, E. T.,
"'Slippage1 in the Monod Equation." (Submitted for publication.)
21. Gaudy, A. F. Jr., Obayashi, A., and Gaudy, E. T., "Control of
Growth Rate by Initial Substrate Concentration at Values Below
Maximum Rate." (Submitted for publication.)
22. Gaudy, A. F. Jr., Ramanathan, M., and Rao, B. S., "Kinetic Behavior
of Heterogeneous Populations in Completely Mixed Reactors."
Biotech. Bioeng., 9, 387-411 (1967).
23. Monod, J., "Recherches sur la Croissance des Cultures Bacteriennes."
Hermann et Cie, Paris (1942).
24. Schaefer, W., "Recherches sur la Croissance du Mycobacterium
tuberculosis en Culture Homogene. Ann. Inst. Pasteur, 74, 458-
463 (1948).
25. Monod, J., "The Growth of Bacterial Cultures." Ann. Rev. Microbiol
3, 371-394 (1949).
26. Peil, K. M., and Gaudy, A. F. Jr., "Kinetic Constants for Aerobic
Growth of Microbial Populations Selected with Various Single Com-
pounds and with Municipal Wastes as Substrates." Appl Microbiol
21, 253-256 (1971).
144
-------
27. Ramanathan, M., and Gaudy, A. F. Jr., "Effect of High Substrate
Concentration and Cell Feedback on Kinetic Behavior of Heter-
ogeneous Populations in Completely Mixed Systems." Biotech.
Bioeng., 11, 207-237 (1969).
28. Ramanathan, M., and Gaudy, A. F. Jr., "Studies on Sludge Yield
in Aerobic Systems." Proceedings 26th Industrial Waste Conf.,
Purdue Univ. (in press)(1971).
29. Gaudy, A. F. Jr., and Ramanathan, M., "Variability in Cell Yield
for Heterogeneous Microbial Populations of Sewage Origin Grown
on Glucose." Biotech. Bioeng., 13, 113-123 (1971).
30. Monod, J., "La Technique de Culture Continue. Theorie et Appli-
cations." Ann. Inst. Pasteur, 79, 390-410 (1950).
31. Herbert, D., "A Theoretical Analysis of Continuous Culture Sys-
tems." In Continuous Culture of Micro-organisms. Soc. Chem.
Ind. Monograph No. 12, 21-53 (196TT
32. Ramanathan, M., and Gaudy, A. F. Jr., "Steady-State Model for
Activated Sludge with Constant Recycle Sludge Concentration."
Biotech. Bioeng., 13, 125-145 (1971).
33. Storer, F. F., and Gaudy, A. F. Jr., "Computational Analysis of
Transient Response to Quantitative Shock Loadings of Heterogen-
eous Populations in Continuous Culture." Environ. Science Technol.
3, 143-149 (1969).
34. Young, T. B., Bruley, D. F., and Bungay, H. R. Ill, "A Dynamic
Mathematical Model of the Chemostat." Biotech. Bioeng., 12, 747-
769 (1970).
35. Thabaraj, G. J., and Gaudy, A. F. Jr., "Effect of Dissolved Oxygen
Concentration on the Metabolic Response of Completely Mixed
Activated Sludge." J. Water Pollution Control Fed., 41, R322-
R335 (1969).
36. Gaudy, A. F. Jr., "Studies on Induction and Repression in Acti-
vated Sludge Systems." Appl. Microbiol., 10, 264-271 (1962).
37. Gaudy, A. F. Jr., Gaudy, E. T., and Komolrit, K., "Multicomponent
Substrate Utilization by Natural Populations and a Pure Culture
of Escherichia coli." Appl. Microbiol., 11, 157-162 (1963).
38. Gaudy, A. F. Jr., Komolrit, K., and Bhatla, M. N., "Sequential
Substrate Removal in Heterogeneous Populations." J. Water
Pollution Control Fed., 35, 905-922 (1963).
145
-------
39. Gaudy, A. F. Jr., Komolrit, K., and Gaudy, E. T., "Sequential
Substrate Removal in Response to Qualitative Shock Loading of
Activated Sludge Systems." Appl. Microbiol., 12, 280-286 (1964).
40. Komolrit, K., and Gaudy, A. F. Jr., "Substrate Interaction
During Shock Loadings to Biological Treatment Processes." J.
Water Pollution Control Fed., 38, 1259-1272 (1966).
41. Komolrit, K., and Gaudy, A. F. Jr., "Biochemical Response of
Continuous-flow Activated Sludge Processes to Qualitative Shock
Loadings." J. Water Pollution Control Fed., 38, 85-101 (1966).
42. Krishnan, P., and Gaudy, A. F. Jr., "Studies on the Response of
Activated Sludge to Shock-loadings." Biotech. Bioeng., 7, 455-
470 (1965).
43. Kincannon, D. F., Gaudy, A. F. Jr., and Gaudy, E. T., "Sequential
Substrate Removal by Activated Sludge After a Change in Salt Con-
centration." Biotech. Bioeng., 8, 371-378 (1966).
44. Grady, C. P. L. Jr., Gaudy, A. F. Jr., and Gaudy, E. T., "Control
Mechanisms Operative in a Natural Microbial Population Selected
for Its Ability to Degrade L-Lysine. I. Effect of Glucose in
Batch Systems." Appl. Microbiol., 18, 776-784 (1969).
45. Grady, C. P. L. Jr., and Gaudy, A. F. Jr., "Control Mechanisms
Operative in a Natural Microbial Population Selected for Its
Ability to Degrade L-Lysine. II. Effects of Fructose and
Ribose in Batch Systems." Appl. Microbiol., 18, 785-789 (1969).
46. Grady, C. P. L. Jr., and Gaudy, A. F. Jr., "Control Mechanisms
Operative in a Natural Microbial Population Selected for Its
Ability to Degrade L-Lysine. III. Effects of Carbohydrates in
Continuous-flow Systems Under Shock Load Conditions." Appl.
Microbiol., 18, 790-797 (1969).
47. Kincannon, D. F., and Gaudy, A. F. Jr., "Response of Biological
Waste Treatment Systems to Changes in Salt Concentration."
Biotech. Bioeng., 10, 483-496 (1968).
48. Goel, K. C., and Gaudy, A. F. Jr., "Studies on the Relationship
Between Specific Growth Rate and Concentration of Nitrogen Source
for Heterogeneous Microbial Populations of Sewage Origin."
Biotech. Bioeng., 11, 67-78 (1969).
49. Sawyer, C. N., "Bacterial Nutrition and Synthesis." In Biolog-
ical Treatment of_ Sewage and Industrial Wastes, edited by J. McCabe
and W. W. Eckenfelder, Jr., 3-17. Reinhold Publishing Corp.,
New York (1956).
146
-------
50. Goel, K. C., and Gaudy, A. F. Jr., "Regulation of Nitrogen Levels
for Heterogeneous Populations of Sewage Origin Grown 'in Completely
Mixed Reactors." Biotech. Bioeng., 11, 79-98 (1969).
51. Helmers, E. N., Frame, J. D., Greenberg, A. E., and Sawyer, C. N.,
"Nutritional Requirements in the Biological Stabilization of
Industrial Wastes. III. Treatment with Supplementary Nutri-
ents." Sew. Ind. Wastes, 24, 496-507 (1952).
52. Rao, B. S., and Gaudy, A. F. Jr., "Effect of Sludge Concentration
on Various Aspects of Biological Activity in Activated Sludge."
J. Water Pollution Control Fed., 38, 794-812 (1966).
53. Krishnan, P., and Gaudy, A. F. Jr., "Substrate Utilization at
High Biological Solids Concentrations." J. Water Pollution Con-
trol Fed., 40, R54-R66 (1968).
54. Komolrit, K., Goel, K. C., and Gaudy, A. F. Jr., "Regulation of
Exogenous Nitrogen Supply and Its Possible Applications to the
Activated Sludge Process." J. Water Pollution Control Fed., 39,
251-266 (1967).
55. Gaudy, A. F. Jr., Goel, K. C., and Freedman, A. J., "Activated
Sludge Process Modification for Nitrogen-deficient Wastes."
In Advances in Water Pollution Research, Pergamon Press, New York,
613-623 (1969T.
56. Goel, K. C., and Gaudy, A. F. Jr., "Regeneration of Oxidative
Assimilation Capacity by Intracellular Conversion of Storage
Products to Protein." Appl. Microbiol., 16, 1352-1357 (1968).
, *,
57. Gaudy, A. F. Jr., Goel, K. C., and Gaudy, E. T., "Application of
Continuous Oxidative Assimilation and Endogenous Protein Synthe-
sis to the Treatment of Carbohydrate Wastes Deficient in Nitro-
gen." Biotech. Bioeng., 11, 53-65 (1969).
,58. Gaudy, A. F- Jr., Goel, K. C., and Gaudy, E. T., "Continuous
Oxidative Assimilation of Acetic Acid and Endogenous Protein
Synthesis Applicable to Treatment of Nitrogen-deficient Waste
Waters." Appl. Microbiol., 16, 1358-1363 (1968).
59. Goel, K. C., and Gaudy, A. F. Jr., "Development of a New Method
for Treatment of Nitrogen-deficient Wastes." Proceedings 2nd
Mid-Atlantic Industrial Waste Conf., Drexel Institute of Tech-
nology, Philadelphia, 194-216 (1968).
60. Gaudy, A. F. Jr., and Goel, K. C., "Oxidative Assimilation of
Nitrogen^deficient Industrial Waste." Proceedings 16th Ontario
Industrial Waste Conf., 111-123 (1969).
147
-------
61. Gaudy, A. F. Jr., and Engelbrecht, R. S., "Basic Biochemical Con-
siderations During Metabolism in Growing versus Respiring Sys-
tems ." Advances in Biological Waste Treatment. Vol. II, Pergamon
Press, New York, 11-26 (1963).
62. Thabaraj, G. J., and Gaudy, A. F. Jr., "Effect of Initial Biolog-
ical Solids Concentration and Nitrogen Supply on Metabolic Pat-
terns During Substrate Removal and Endogenous Metabolism."
J. Water Pollution Control Fed., 43, 318-334 (1971).
63. Porges, N., Jasewicz, L., and Hoover, S. R., "Aerobic Treatment
of Dairy Wastes." Appl. Microbiol. 1, 262-270 (1953).
64. Symons, J. M., and McKinney, R. E., "The Biochemistry of Nitrogen
in the Synthesis of Activated Sludge." Sew. Ind. Wastes, 30,
874-890 (1958).
65. Busch, A. W., and Myrick, H. N., "Food-Population Equilibria in
Bench-scale Bio-oxidation Units." J. Water Pollution Control Fed.,
32, 949-959 (1960).
66. Ramanathan, M., Gaudy, A. F. Jr., and Ragthaidee, W., "Responses
of Extended Aeration Activated Sludge to Quantitative Shock
Loads." Proceedings 19th Oklahoma Industrial Wastes and Pollution
Control Conf., Oklahoma State Univ. (1968).
67. Gaudy, A. F. Jr., Ramanathan, M., Yang, P- Y., and DeGeare, T. V.,
"Studies on the Operational Stability of the Extended Aeration
Process." J. Water Pollution Control Fed., 42, 165-179 (1970).
68. Gaudy, A. F. Jr., Yang, P. Y., and Obayashi, A. W., "Studies on
the Total Oxidation of Activated Sludge With and Without Hydro-
lytic Pretreatment." J. Water Pollution Control Fed., 43, 40-54
(1971).
69. Gaudy, A. F. Jr., "Comparative Utility of Research Employing Model
Synthetic Systems and Natural Field Systems." Biotech. Bioeng.
(in press)(1971).
148
-------
PATENTS AND PUBLICATIONS
Although several of the process modifications described herein would
seem to constitute patentable art, no patents have been applied for.
Publications completed during the tenure of this research project and
supported in part by it include numbers 20, 21, 28, and 32 of the
preceding list of references.
149
-------
GLOSSARY OF TERMS, ABBREVIATIONS, AND SYMBOLS
Page
Definition of Terms introduced
Autotroph - an organism capable of utilizing C0? as the
sole carbon source for growth 8
Carbon source - a compound which furnishes carbon for syn-
thesis of cellular components 5
Carbonaceous BOD - biochemical oxygen demand due to
organic carbon compounds 8
Endogenous phase - metabolism in the absence of exogenous
carbon compounds. In a strict sense, for a
pure culture this metabolism involves oxida-
tion of its storage materials or other cellu-
lar components by an individual cell. In a
mixed population, cells may feed on other
cells 9
Organotroph - an organism which requires organic compounds
as carbon source 8
Oxidative assimilation - utilization of carbon source only
for respiration and synthesis of storage
products, i.e., without cell replication 100
Photosynthesis - synthesis of organic compounds from C02
using light as a source of energy 15
Substrate - any compound which is acted upon by an enzyme;
therefore, a compound which is used by the
cell for energy and/or synthesis 1
ACOD - a measure of the amount of biologically available
organic matter in a waste sample 6
Abbreviations and Symbols
A A collective constant replacing the term
(1 + a - ac) 58
BOD Biochemical oxygen demand 5
Biochemical oxygen demand which has been express-
ed in five days, i.e., y5 11
151
-------
Page
Abbreviations and Symbols introduced
c Sludge recycle concentration factor, equal
to the ratio between the recycle solids
concentration, XR, and the biological solids
concentration in the reactor, X 56
C Initial concentration 47
o
C. Concentration at any time, t 47
COD Chemical oxygen demand; measure of total chem-
ically oxidizable organic matter 6
COD Effluent COD exiting the biological reactor
e system, or COD remaining at the end of a batch
experiment 10
CODf The COD of biological solids at the end of an
experiment 19
COD. COD at the beginning of a batch experiment, or
1 influent COD to aerator 10
COD. COD of the biological solids at the beginning
1 of an experiment 19
COD COD of the waste, not including biological
solids if present 19
D Dilution rate. Ratio of the rate of inflow, F,
and the volume of liquor in the aeration tank,
V. It is equal to the reciprocal of the mean
residence time in a completely mixed reactor 46
D The dilution rate for the reactor alone in sys-
tems employing sludge recycle. It is always
larger than the dilution rate for the entire
system, i.e., Dr = D(l + a), a being the ratio
between the hydraulic rate of recycle flow and
the flow of incoming waste, F 56
DO Dissolved oxygen concentration 5
e Base of Naperian logarithms 26
F Rate of flow of incoming substrate or waste
water 45
152
-------
Page
Abbreviations and Symbols introduced
K A biological "constant" used in the hyper-
bolic expression relating specific growth
rate to substrate concentration. It is known
as the saturation constant. It is numeri-
cally equal to the substrate concentration at
which the specific growth rate is equal to %
the maximum specific growth rate for the system 29
LQ Ultimate BOD 8
S Substrate concentration 6
S Steady state concentration of substrate in a
completely mixed continuous flow reactor 52
S Initial concentration of substrate in batch
systems 29
S. Concentration of substrate in the inflowing feed
in continuous flow operation 45
t Time 47
t Mean residence time in a completely mixed contin-
uous flow reactor, V/F 46
t. Doubling time, time required for one doubling in
weight of the bio-mass growing in an exponential
phase 28
V Volume of reaction fluid under aeration 45
X Biological solids concentration, weight per volume 6
X Steady state biological solids concentration in a
completely mixed continuous flow reactor 52
X Initial biological solids concentration in a
0 batch reactor 26
X Excess biological solids produced in a process
e (sludge wasted) 56
XR Biological solids concentration in the recycle
solids flow to an aerobic reactor in a continuous
flow system 56
153
-------
Page
Abbreviations and Symbols introduced
X Biological solids concentration at time t 26
y BOD expressed, i.e., accumulated oxygen uptake 6
Y Sludge or cell yield; mg biological solids
produced per mg organic substrate used 21
a Recycle flow ratio in a continuous flow bio-
logical reactor system employing cell recycle.
It is the ratio between the rate of flow of
recycle solids and the rate of flow of incoming
waste water 56
A Change in value of a parameter, or difference
between initial and final values 6
y Specific growth rate in an exponential phase
of growth 26
u The maximum specific growth rate for a system
in exponential growth 29
154
.NG Off ICS:
-------
i 1 Accession Number
w
f. Organization
n Subject Field & Group
05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Oklahoma State University. Cpnt.pr for Ulatpy Rpcpavrh in Fnn-i napvi nn
Ti«e
Biological Principles for Design and Operation of the Activated
Sludge Process
•J Q Authors)
Anthony h. baudy. Jr.
and
Elizabeth T. Gaudy
16
Project Designation
17090 FQJ
21 I Note
22
Citation
23
Descriptors (Starred First)
Activated Sludge*, Biological Treatment*, Aerobic Treatment*, Ultimate Disposal*
Purification*, Oxidation, Oxygen Demand '
25
Identifiers (Starred First)
Kinetic Models*, Shock Loads*, Purification Mechanisms*, Nitrogen Requirements*,
Oxidative Assimilation*, Endogenous Phase*, Purification Efficiency, Metabolism
27
Abstract
Generalized concepts of BOD exertion, the use of ACOD as a design and operational
tool, the stoichiometry and mass balance concepts of treatment, and kinetic
equations for microbial growth are presented. Design models are discussed, and
a model for completely mixed reactors holding recycle solids, X^, constant is
recommended. Some guidelines for accommodation of various types of shock load-
ings are included. Concepts of oxidative assimilation and the multiple effects
of solids concentration, nitrogen concentration, and detention time are related;
a new activated sludge process (continuous oxidative assimilation) for nitrogen-
deficient wastes is presented. Data supporting the concept of total oxidation
are presented, and a modification of the extended aeration process incorporating
chemical hydrolysis of portions of the sludge is recommended. In the final
chapter, some possible flow diagrams for complete aerobic treatment (purifica-
tion and sludge disposal) of metabolizable organic wastes are presented.
Abstractor
Anthony F. Gaudy, Jr.
IfiHtitiilion
Oklahoma State University
WR;!02 (REV. JUUY 19*9)
WRSIC
SEND, WITH COPY OF DOCUMENT, TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 20240
GPO: 1970 - 407 -891
------- |