WATER POLLUTION CONTROL RESEARCH SERIES • I608QFSNJO/ 71
ENGINEERING METHODOLOGY FOR
RIVER AND STREAM REAERATION
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the results 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 Office of Research and
Monitoring, 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, Xi&M, Environmental Protection Agency, Washington,
D. C. 20'4-eo.
- about our cover
The cover illustration depicts a city in which man's activities coexist
in harmony with the natural environment. The National Water Quality
Control Research Program has as its objective the development of the
water quality control technology that will make such cities possible.
Previously issued reports on the National Water Quality Control Research
Program include:
Report Number
16080 06/69
16080 10/69
16080DRX10/69
16080 11/69
16080D0007/70
16080DVF07/70
16080 10/70
16080DWP11/70
16080DUP12/70
16080FYA03/71
16080GPF04/71
16080GGP07/71
Title
Hydraulic and Mixing Characteristics of Suction Manifolds
Nutrient Removal from Enriched Waste Effluent by the
Hydroponic Culture of Cool Season Grasses
Stratified Reservoir Currents
Nutrient Removal from Cannery Wastes by Spray Irrigation
of Grassland
Optimum Mechanical Aeration Systems for Rivers and Ponds
Development of Phosphate-Free Home Laundry Detergents
Induced Hypolimnion Aeration for Water Quality Improve-
ment of Power Releases
Induced Air Mixing of Large Bodies of Polluted Water
Oxygen Regeneration of Polluted Rivers: The Delaware River
Oxygen Regeneration of Polluted Rivers: The Passaic River
Corrosion Potential of NTA in Detergent Formulations
Effects of Feedlot Runoff on Water Quality of Impoundments
-------
ENGINEERING METHODOLOGY FOR RIVER
AND STREAM REAERATION
by
JBF Scientific Corporation
2 Ray Avenue
Burlington, Massachusetts 01803
for the
ENVIRONMENTAL PROTECTION AGENCY
Project #16080 FSN
October 1971
-------
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
recommendation for use.
Forsnlo by llio Superintendent of Documents, U.S. Government Printing 0/licc, Washington, D.C. '.'WOi- I'riccSI.-.'.s
11
-------
ABSTRACT
Results of recent activities in river and stream aeration by artificial
techniques are reviewed, and a rational engineering methodology is
developed for future river and stream aeration projects.
The development of the methodology follows from a thorough review
of the oxygen dynamics in rivers and streams and the capabilities of
aeration systems within the present state of the art. The report
shows how the theoretical work can be simplified considerably and
applied to the solution of river and stream water quality problems.
It is assumed that aeration would only be used as a "polishing" action
after all identifiable waste sources have received at least secondary
treatment.
The results indicate that, with careful consideration of site factors,
artificial aeration can be applied successfully to raise dissolved oxygen
to 5 ppm, using mechanical surface aerators, diffusers, downflow
contactors, and sidestream mixing. However, since the transfer
of oxygen from air into water is relatively inefficient above 5 ppm DO,
the introduction of molecular oxygen through sidestream mixing,
U-Tubes, and possibly diffusers should be considered, depending on
the volume of water to be aerated. In cases where DO may be main-
tained at levels lower than 5 ppm, systems using air are competitive
with molecular oxygen, depending on site conditions.
This report was submitted in fulfillment of Project Number 16080 FSN
under the partial sponsorship of the Environmental Protection Agency.
111
-------
CONTENTS
SECTION Pag
I CONCLUSIONS 1
II RECOMMENDATIONS 3
III INTRODUCTION
Scope and Purpose 5
Background 6
Approach 7
IV DISSOLVED OXYGEN DYNAMICS IN RIVERS
AND'STREAMS
Oxygen Balance 9
Evaluation of Stream Parameters in Oxygen-
Balance Equation 14
Saturation Values of Cs 14
Aeration Constant--Ka 14
Ct
Biochemical Oxygen Demand (BOD) 21
Benthal Demands S 28
Net Photo synthetic Oxygen Production _P 30
Longitudinal Dispersion 31
Simplification of the Oxygen-Balance Equations 34
V AERATION SYSTEMS FOR RIVER AND STREAM
APPLICATIONS
Introduction 39
Mechanical Surface Aerators 39
Diffuser Systems 50
Diffuser Systems Tests Using Air 51
Diffuser System Tests Using Molecular
Oxygen 54
The Effect of Flow on Dispersion of
Oxygen from Diffusers 54
Downflow Contactors 55
Sidestream Pressurization 72
Use of Pure Oxygen 75
Hybrid or Mixed Systems 78
v
-------
SECTION Page
VI ENGINEERING METHODOLOGY FOR RIVER AND
STREAM AERATION
Problem Recognition 83
Preliminary Assessment 83
Determination of Oxygen Requirements 85
Selection of Aeration Units 94
Engineering Considerations Affecting the
Selection of an Artificial Aeration System 94
Economic Considerations Associated with
an Aeration System 96
Design and Cost Examples 99
Summary 103
Site Factors 103
Final Selection of an Aeration System 104
VII ACKNOWLEDGEMENTS 105
VIII REFERENCES 107
IX APPENDICES
Appendix A - Diffusers and Mechanical Aerators 1 13
VI
-------
FIGURES
Figure pag,
4. 1 Typical Rate- of-Diffusion Plot 11
4.2 Typical Plot of Photosynthetic Oxygen Production 11
4. 3 K vs Flow, as Calculated by Various Prediction
a
Equations 18
4.4 Grand River BOD Curve 22
4. 5 Total Carbonaceous and Nitrogenous BOD as a
Function of Time 25
4.6 BOD Curve, Lansing Wastewater Treatment
Plant, 24-hr Composite Sample, Taken
July^lZ, I960 26
4.7 First-Order Representation of a Nitrification
Process 29
5. 1 Example of Velocity Profiles at Various Distances
from a 75-hp Aerator 41
5. 2 Mean Rate of Oxygen Absorption at Steady-State
Operation Under Standard Conditions 46
5. 3 Oxygen Transfer Rate vs Flow for Mechanical
Aerator 47
5. 4 Graphical Results of Cross-Sectional Area
Distribution 56
5.5 DO Deficit in 40-Ft-Deep U-Tube 58
5.6 Air Blower Injection Modification of U-Tube
System 59
5. 7 Cascade Air Injection 60
5.8 Venturi Air Injection 61
5.9 Schematic Installation of U-Tube Oxygenation of
Stratified Impoundment Releases 62
5. 10 Increase in DO vs Per Cent A/W for Diffusion of
Oxygen and Air 65
5. 11 Effect of Depth on Oxygen Transfer Economy in
60-in. U-Tube 66
5. 12 Effect of Nominal Velocity on Transfer Economy
in 60-in. U-Tube, 40-ft Deep 67
5. 13 Comparison of Pressure Drop Across Center-Plug
and Venturi Aspirators 70
5. 14 Zone of Influence of the Downdraft Bubble Contactor
on the Surrounding Water 71
vn
-------
Figure Page
5.15 Cost of Pure Oxygen 77
5. 16 Total Oxygen Absorption vs Depth of Injection 79
5. 17 Effect of Percentage Oxygen in Aerating Gas on
the Rate of Oxygen Transfer 81
6. 1 Outline of Engineering Methodology 84
6. 2 Example DO Profile 86
6. 3 Relative "Work Requirements for a Unit Increase
in DO in Water Initially Having Different DO
Levels 88
6.4 DO Profile Before and After Oxygen Addition at
Three Locations 91
6. 5 DO Profile Before and After Oxygen Addition at
Two Locations 92
6. 6 DO Profile Before and After Oxygen Addition at
One Location 93
Vlll
-------
TABLES
TABLE
4. 1 Saturation Values of Oxygen in Clean Water at One
Atmosphere Pressure 15
4.2 Aeration Parameters for Determining K 17
3,
4. 3 Transfer Rate of Oxygen into Water from Air
for Various Flow Conditions 20
4.4 Percentages of Carbonaceous BOD as a Function
of Time for Kc = . I/day 23
4. 5 Typical Longitudinal Dispersion Coefficients for
Rivers 32
4.6 Sample^Calculation of Dispersion Coefficient from
Field Data Test 35
5. 1 Surface Aerator Pumping Rates 42
5. 2 Zone of Outward Aerator Influence 43
5. 3 Sample DO Profiles at Various Distances from Aerators 44
5.4 Average Oxygen Transfer Efficiencies, in lbsO£/hp-hr 48
5. 5 Average Oxygen Transfer Rate Obtained by Different
Investigators for Surface Aerators 49
5.6 Diffuser Results from the Passaic and Delaware
Rivers 52
5.7 Reaeration Data for the Pearl River, Louisiana, Using
a Double Aeration Header with 1/32-in. and 3/61-in.
Orifices 55
5.8 Air-Water Ratio Required for Oxygen Saturation 63
5.9 Air-Water Ratio Required for Nitrogen Saturation 63
5. 10 Results of Chemical Analysis for Dissolved Nitrogen 68
5. 11 Vertical Area Affected by Bubble Contactor 72
5. 12 Dissolved Oxygen Added to the Pearl River at an
Oxygen Addition Rate of 30, 000 Ib/day by Sidestream
Oxygenation 74
5. 13 Comparison of Reaeration Studies with Air and Oxygen 80
6.1 Characteristic Features of Aeration Systems 97-9!
A. 1 List of Diffuser Manufacturers 114
A. 2 Diffuser Efficiencies at Various Locations in Tank
Under Standard Conditions 116
A. 3 List of Mechanical Aerator Manufacturers 117
IX
-------
SECTION I
CONCLUSIONS
1. A review of the present state of the art in river and stream
aeration indicates that the performance of aeration systems is
strongly related to the minimum DO level set as a standard on
any particular river or stream.
2. When a minimum level of 5 ppm is set, any system not using
molecular oxygen will be relatively inefficient and, unless
unusual site conditions prevail, will probably not be suitable.
This is particularly true in summer months, when the
saturation concentration for DO is at its lowest value.
3. The most efficient location for mechanical surface aerators
and diffusers using air is at the point of maximum oxygen
deficit. In order to maintain some established minimum DO
level, however, these devices must be located wherever that
minimum is approached, resulting in a significant loss of
transfer efficiency. Based on presently available transfer rate
data, mechanical surface aerators and diffusers using air are
not efficient for maintaining DO levels above 4 ppm.
4. Downflow contactors provide a higher transfer efficiency than
surface aerators or diffusers but can only be used where
sufficient water depth is available. Nitrogen supersaturation
may also be a problem but can be avoided by limiting the depth
of the down leg of the tube. Since maximum transfer efficiency
occurs at depths greater than 40 feet, there will always be some
loss in efficiency if nitrogen supersaturation is to be avoided.
5. The lack of mobility and dependence on water depth limit the
usefulness of downflow contactors in problems requiring the
maintenance of a minimum DO level. These systems may be
considered only if an injection point for oxygen occurs at a
compatible location.
6. A relatively simple methodology can be developed for treating
river and stream aeration problems. In this methodology the
DO profile for "worst conditions" is used to determine alternative
locations for aeration units in order to maintain some specified
minimum DO level. Loss in natural aeration due to the oxygen
addition must be taken into account. Alternative locations are
required because site conditions may preclude the use of some
types of aeration systems. There may also be cases where a.
trade-off is necessary between the injection of large amounts
of oxygen at one location versus smaller amounts at several
locations .
-------
7 . For large rivers where more than 50, 000 Ibs /day of oxygen is
required, the use of molecular oxygen applied through side
stream mixing should definitely be considered. Whether or
not the oxygen is supplied from a generating plant located on
the site, or delivered to the site as liquid oxygen will depend
on the length of time over which oxygen is required. For con-
tinuous year-round operation, gaseous oxygen should be
generated on the site. For intermittent periods and for volume
requirements of less than 50,000 Ibs/day, liquid oxygen trucked
to the site may be more economical.
8. The transfer processes of various aeration systems in rivers and
streams are not as well understood as in treatment facilities or
even lakes and impoundments. The superimposed flow field
creates some uacertainty in handling the analytical details, and
a wide range of efficiencies has been reported. The range for
surface aerators after conversion to standard conditions is 1.2
to 4.5 Ibs O2/hp-hr. Thus, depending on the number chosen,
cost estimates may be off by as much as a factor of three.
9. The mathematical models of stream processes are generally
adequate for the engineering design of a river or stream aeration
system. The major inadequacy lies in the prediction of natural
aeration coefficients from empirical methods. If measurements
of the reaeration coefficient cannot be made at the site and an
empirical method must be selected, the stream conditions from
which that particular derivation was made should be reviewed to
determine suitability for the particular application.
10. A review of current techniques for measuring rate processes
which control DO and BOD in rivers and streams indicates that
the present state of the art offers sufficient accuracy for the
design of stream aeration systems. Additional refinements can
be made; however, the results may offer a precision greater
than the variations in the processes being measured.
11. Differences in longitudinal dispersion coefficients for a stream
have been reported. Such differences may be due to whether or
not measurements were made with the aeration system in place
and operating. Knowledge of the dispersion characteristics are
particularly important for large river systems where the aerator
system is a relatively localized source. Measurements of the
dispersion coefficient should be made with the device in place.
-------
SECTION II
RECOMMENDATIONS
Artificial aeration of rivers and streams should not be used as a direct
substitute for waste treatment at the source. There may be cases,
however, where aeration can be used as a "polishing" action during
periods of high temperature and low flow. Until advanced waste treat-
ment methods are fully developed and implemented, public opinion may
require an interim solution. Artificial aeration is at a state of devel-
opment now where it can be applied to specific river and stream
problems.
Although artificial aeration is technically feasible, cost estimates are
still not sufficiently accurate, due to problems in predicting the trans-
fer efficiency of_a system in a stream. This deficiency will probably
not be corrected by performing more tests. The primary need is to
define the conditions under which aeration devices should be tested if
they are to be used in a river or stream environment. A laboratory
program should be undertaken to standardize measurement conditions
for a variety of hydraulic parameters which might be encountered in
the field. The results of this program should then be cross-checked
to determine compatibility with field conditions.
The results of this study have indicated that the maintenance of a 5 ppm
DO requirement will be difficult unless molecular oxygen is employed,
and methods for delivering molecular oxygen to the water need additional
development. The sidestream pressurization method is oriented toward
large rivers where only a portion of the flow can be diverted. Work
should be done on using gaseous oxygen, delivered to a site by truck, to
supply a fine-bubble-size diffuser. The gas would be delivered under
pressure, and this might reduce the possibility of internal clogging. A
system like this would have application in small, shallow rivers or
streams.
The conclusion has been made that surface aerators will not be effec-
tive in maintaining a DO level above 4 ppm. In cases where a level
less than 4 ppm is acceptable, the design of the aerator should be modi-
fied to promote mixing of air and water downstream of the aerator.
The symmetrical radial mixing zone of present aerators can be made
asymmetrical by adding baffle plates or flow guides.
The design of an aeration system for a river or stream requires con-
sideration of a number of factors. The system can be designed suc-
cessfully by following an orderly procedure which highlights all of the
considerations. The procedures should be formalized to the extent
that design errors due to a lack of knowledge about river and stream
processes are minimized.
-------
In summary, the following areas are most in need of additional
research and development:
I. Uniform standards for the measurement of transfer
efficiencies of various aeration devices under flow
conditions.
2. Improved devices for the diffusion of gaseous oxygen
in the water.
3. Enhancement of surface aerator performance through
the use of flow guides or baffles.
-------
SECTION III
INTRODUCTION
Scope and Purpose
Present aeration technology for rivers and streams has developed
primarily from waste treatment applications and from a limited
number of field tests in lakes and rivers. The purpose of this study
was to review the results of recent activities in river and stream
aeration by artificial techniques and to assess the present state of
the art. A direct result of this review has been the development of
an engineering methodology for river and stream aeration systems.
The major objective in any aeration system design is to add oxygen to
the water. In waste treatment applications this is done to satisfy a
high biochemical oxygen demand (BOD), while in lakes and rivers one
wishes to maintain a high-enough dissolved oxygen level to support a
healthy aquatic population. There are other differences affecting
system design, including the need to promote mixing of suspended
solids in an aerated lagoon versus the desire to minimize excessive
turbulence in lakes for aesthetic reasons. In rivers and streams
natural flow conditions provide mixing, and new water is always
being exposed to the aeration system. In lakes where there is no
natural flow condition the most efficient aeration system is one which
provides continual recirculation with minimum power. This results
in the exposure of new water surfaces to the air and promotes diffusion
and mixing.
In order to efficiently design an aeration system for rivers and streams
the oxygen balance for the stream must be understood. The primary
source of oxygen for a stream is natural reaeration at the surface,
which is aided by the velocity and turbulence of the stream. If a
stream is artificially aerated, there will be some loss in the natural
aeration capability, and this must be compensated for by the engineer
in estimating the additional oxygen required to meet a specified water
quality condition. Additional contributions come from photosynthesis,
ground water, drainage, and flow augmentation.
The remainder of the report is organized as follows. In Section III
the scope and purpose, background, and approach methods are intro-
duced. Section IV includes a review of the oxygen dynamics in rivers
and streams and a section on the measurement of parameters in
the oxygen-balance equation. Section V is a review of the use of
surface mechanical aerators, diffusers, downflow contactors (U-Tubes),
and sidestream mixing in river and stream applications. Also included
in Section V is a discussion of the possible benefits of using pure
oxygen instead of air. In Section VI the methodology for designing
the aeration system is developed.
-------
Background
Over the last 50 years an extensive body of literature has accumulated
on the application of aeration in waste treatment. Recently, interest
has developed in the use of aeration technology for destratifying lakes
and reservoirs and in artificially improving the assimilative capacity
of rivers and streams.
Serious consideration of artificial aeration for rivers and streams
appears to have been first initiated by Tyler [l] in the early 1940's.
He proposed that under certain conditions it might be economically
advantageous to treat waste in a stream rather than in concentrated
form as would be the case in a treatment pond. This philosophy has
apparently interested a number of investigators, as there have been
many tests of this concept in the last 5-10 years.
A second philosophy on the use of artificial stream aeration assumes
that all possible waste material entering a stream is first treated in
a waste treatment plant, resulting in the removal of 90-99% of the
BOD. The aeration system would then be used as a supplement where
DO levels on the order of 5 ppm are required to support a viable
aquatic life and its attendant recreational benefits. In this case the
aeration system would probably be operated on a seasonal basis when
DO levels fell below 5 ppm. As will be shown later, much more work
is required to maintain a 5 ppm level than would be needed for a 3 ppm
level; thus relative transfer efficiencies of aeration devices are an
important consideration.
Many of the aeration tests conducted have used site-dependent techniques,
such as turbine venting, weirs, and dams. Although these techniques
have demonstrated reasonable transfer efficiencies, their dependence
on site conditions limits general application. Consequently, the
emphasis in this study has been on more flexible systems, such as
mechanical surface aerators, diffusers, downflow contactors, and
sidestream mixing.
Since 1966, a number of tests have been conducted using mechanical
surface aerators. A fewer number have used diffusers and side-
stream mixing. No tests using U-Tubes in a flowing river were found,
although there is reason to believe from results obtained in lakes and
impoundments that these devices can be used on rivers and streams,
given sufficient water depth. In several tests of diffusers, U-Tubes,
and sidestream mixing, molecular oxygen has been substituted for air.
In many of the test programs the aeration systems were evaluated
under a variety of actual or simulated stream conditions. Performance
data inmost cases was converted to a set of standard conditions, with
the transfer efficiencies stated in pounds of oxygen transferred per
horsepower hour.
-------
Although the economics of river and stream aeration favor a high
oxygen transfer rate, site conditions also restrict the use of
particular aeration devices, for reasons other than cost. Many
rivers are navigable both to shipping and recreational boating; others
are used for recreational boating and water sports. Where public use
is extensive, surface aerators using turbulent mixing would be re-
stricted. In cases where a channel is maintained by dredging, diffuser
systems cannot be located in the channel. Designs of aeration systems
must also consider aesthetic conditions at the site. Large, unsightly
structures with extensive surface agitation may not be acceptable to
the public.
Approach
An extensive survey of the literature was conducted, with particular
emphasis on reports of field tests with mechanical surface aerators,
diffusers, U-Tubes, and sidestrearn mixing. Although necessary in
the development of a total river basin system, reports dealing with
mathematical and simulation models were reviewed only for background
information. Each aeration field test was reviewed for engineering
design, continuity of results, and efficiency.
In addition to a state-of-the-art review, emphasis was placed on the
development of an engineering methodology for river and stream
aeration which could be used by an engineer charged with the develop-
ment of a system for a particular location.
In the course of the study it was evident that some aspects of river
and stream aeration require additional research. Although large-scale
projects may be possible within the present state of the art, additional
research should result in better system effectiveness. Recommendations
for further research are included in this report.
-------
SECTION IV
DISSOLVED OXYGEN DYNAMICS IN RIVERS AND STREAMS
Oxygen Balance
Artificial aeration of a body of water, whether it is a lake or a stream,
will have some effect on the natural oxygen balance. An understanding
of the natural processes controlling oxygen concentration is therefore
necessary before the design of an artificial aeration system can be
undertaken.
There are four main processes controlling oxygen concentrations in
naturally aerated streams:
1. Consumption of oxygen as a result of respiration of
benthic and planktonic organisms, and chemical
oxidation;
2. Exchange of oxygen as a result of atmospheric reaeration;
3. Photosynthetic production of oxygen during the day by
benthic plants and phytoplankton; and
4. Oxygen contribution from ground water, surface
drainage, and storage.
In the first case, consumption of oxygen by respiration is expressed
as a biochemical oxygen demand (BOD) in pounds of oxygen per unit
time. For DO levels below 1 ppm the rate of consumption has been
found to be dependent on the DO concentration [2] ; thus, in highly
polluted streams there may be a diurnal variation in BOD if the
stream becomes oxygen depleted in any particular area. For higher
DO concentrations there does not appear to be the same dependence.
Hence, for rivers or streams in which the minimum DO levels are
on the order of 2 or 3 ppm it is sufficient to assume that the oxygen
demand is a function only of the remaining BOD.
The oxidation of the organic load may occur in the stream or on the
bottom, depending on the physical nature of the material. Because
the rate of demand of the finely dispersed material in the stream
differs from the demand of the larger particles, which settle to the
bottom, the two components are often treated separately.
Also included in the respiration process is the oxygen uptake caused
by nitrification, which is the oxidation of ammonia and nitrites to
nitrates. These compounds are common in the effluents of secondary
-------
waste-treatment plants. Because nitrification is an autotrophic
process carried out by relatively few organisms, an appreciable
time lag may exist between the introduction of ammonia and the
point at which measurable nitrification occurs. Because of the
difference in rate and time, nitrification is often treated separately
from BOD.
The second process is the primary natural mechanism for oxygen-
ating a stream. Oxygen diffuses from the atmosphere directly into
the water, with the exchange rate depending on the rate of renewal
of new water surfaces and on the percentage saturation of oxygen in
the water. Turbulent stream conditions facilitate a higher oxygen
transfer than do quiescent conditions. If the water becomes super-
saturated with oxygen, the direction of transfer may actually be to
the atmosphere. Figure 4. 1 shows typical diffusion rates for a
stream in which there is significant photosynthetic oxygen production.
In this case, oxygen is released to the atmosphere when photosynthesis
during the day results in supersaturation.
The production of oxygen by photosynthesis is the third process in
the oxygen balance. In this process, carbohydrates are synthesized
from carbon dioxide and water, with a subsequent release of oxygen.
This process requires radiant energy from the sun and is consequently
diurnal in nature. The greatest rate of production occurs around noon
and drops to zero at night, as is shown in Figure 4.2. The rate and
total production will depend on the depth of penetration of sunlight,
which is, in turn, dependent on water clarity or turbidity. Maximum
daily production occurs at about 1800 hours, as opposed to the
maximum rate of production, which occurs around noon.
The fourth process is the accrual of oxygen from ground water,
drainage, and storage. This is a site-dependent contribution whereby
incoming waters containing higher (or lower) concentrations of
dissolved oxygen produce changes in the stream DO. This contribution
is usually negligible except when it involves flow augmentation from
hydroelectric storage impoundments.
The dissolved-oxygen level in a stream, therefore, is the net result
of several dynamic processes occurring simultaneously. The
processes interact to produce variations in DO along the length of
a stream, the graphic representation of which is usually called the
DO profile or "oxygen sag curve."
These processes can be quantized on an area basis, e.g., g/mz/hr,
or on a volume basis as g/m3/hr, which is also ppm/hr. The
summation of these processes can be expressed as follows:
q = d + p - r + a (4. 1)
10
-------
o
c
•rH
O
S-i
O
C
o
•«
O
Shaded Area Shows Oxygen
Diffusion into the Stream
1800^ 2400
Time (hours)
Stream Becomes
Supersaturated
Figure 4. 1. Typical Rate-of-Diffusion Plot
c
0)
81
U) .-I
O •>-'
O
-C
Time of Maximum
Production
600
1200 1800
Time (hours)
2400
Figure 4. 2. Typical Plot of Photosynthetic Oxygen Production
11
-------
where
q = rate of change of dissolved oxygen per unit volume (ppm/hr
r = rate of respiration, including oxygen demands by plants,
animals, and aerobic bacteria (ppm/hr)
d = rate of diffusion from the air if the concentration of
dissolved oxygen is below saturation (ppm/hr)
p = rate of production of oxygen by photosynthesis (ppm/hr)
a = rate of accrual from drainage, ground water, or storage
(ppm /hr)
If an oxygen sag curve shows low DO levels at certain stream locatio
or if DO is lower than desired over a particular reach of a stream,
artificial aeration systems may be used as a supplemental source of
oxygen. The fate of the added oxygen will depend on the natural stre
processes.
In designing an aeration system it is necessary that a new oxygen saj
curve be calculated after oxygen is added artifically to raise the DO
level. The shape of the curve will indicate how many aeration devic<
are required and their spacing, in order to keep the DO level above
some specified minimum. In Section VI the specific techniques for
determining the number and spacing of aerators are developed. In
this section it is sufficient to note that the artificial aeration will hav
some effect on the natural balance, and any decrease in the ability oJ
the stream to reaerate naturally must be compensated for by the
aeration system.
In order to obtain a numerical solution to the oxygen balance equatioi
a one-dimensional representation of each of the rate processes in
Equation 4. 1 is developed as expressed in Equation 4.2.
=DT— -U+K(C -C)+P-KL-KN+A-S (4.;
at L 2 3x ax s ' c n v
where
A = rate of accrual of O? from drainage, ground water,
etc. (ppm /day)
C = DO concentration (ppm)
C = saturation value of DO (ppm)
S
12
-------
D = turbulent diffusion (dispersion) coefficient (ft2/day)
JU
K = aeration constant (I/day)
a
K = instream carbonaceous oxidation constant (I/day)
c
K = instream nitrogenous oxidation constant (1/day)
P = photos ynthetic production rate (ppm/day)
S = benthal demand rate (ppm/day)
U = mean stream velocity (ft/day)
t = time (day)
x = distance along the stream (ft)
and
L, and N are given by the solutions to
IT = DT
9t L
and
= D _ u _ KN +N
3t L gx2 3x n a
where
L = carbonaceous BOD (ppm)
L = uniform rate of addition of carbonaceous BOD (ppm/day^
3.
N = nitrogenous BOD (ppm)
N = uniform rate of addition of nitrogenous BOD (ppm/day)
3.
Equations 4.2 through 4.4 form a system of coupled, second-order,
partial differential equations, the solution of which results in DO and
BOD profiles along a stream. The equations are for general non-
steady-state conditions and can be simplified for many stream con-
ditions. Considerable simplification can be achieved, for example,
13
-------
when steady-state conditions are assumed or if diffusion effects are
small. If simplification is not possible, a digital computer greatly
eases the computational burden.
The first step in the solution is to partition the river reach under con-
sideration into a number of sections, each of which is assumed to be
completely mixed or homogeneous. A finite-difference method is ther
used to express the derivatives of the dependent variables C_, L_, and 1
Details of the numerical producer can be found in references 3 and 4.
Although each of the rate terms and constants appearing in Equations
4.2 through 4.4 may be different for each section, they are often
essentially equal and additional simplification can be achieved.
In order to obtain numerical values for terms appearing in Equations
4.2 through 4.4, a number of laboratory and field techniques have
been developed. These techniques are discussed in the next subsectio
along with the practical assumptions that can be made and the relative
orders of magnitude of the measured values.
Evaluation of Stream Parameters in Oxygen-Balance Equation
Saturation Values of C
The solubility of oxygen in water is primarily influenced by temperatu
and salinity. According to SED-ASCE [5] , the saturation value of
oxygen concentration at sea level can be expressed as
where
C = 14. 625 - .41022T + . 0799T2 - . 000077774T:
s
= temperature in C.
In studies conducted on the Passaic River, organic pollution did not
appear to have a discernible effect on the saturation values [6] . It
is reasonable, therefore, to assume that, unless one is dealing with
extremely high organic loadings, the values of Cs given by the above
equation are applicable. Values of Cs for the temperature range of
0°C to 40°C are given in Table 4. ].
Aeration Constant— K_
d
The natural aeration rates that occur in streams depend directly on
the amount of turbulence and consequently are related to hydraulic
parameters, such as velocity and depth of flow. Temperature and
14
-------
TABLE 4. )
Saturation Values of Oxygen in Clean Water
at One-Atmosphere Pressure
Temperature
°F °C
32. 0
35.6
39.2
42.8
46.4
50. 0
53.6
57.2
60. 8
64.4
68.0
71.6
75.2
78.8
82.4
86.0
89.6
93.2
96.8
100.4
104.4
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
C
s
14.6
13.8
13. 1
12.5
11.9
11. 3
10. 8
10.4
10.0
9.5
9.2
8.8
8.5
8.2
7.9
7.6
7.4
7.2
7. 0
6.8
6.6
the type and concentration of pollutants in solution also affect aeration
but to a lesser degree.
The aeration constant for a stream in which there is little biological
and chemical activity can be readily evaluated. The temperature (to
determine Cs) and DO measurements are taken at two stream locations
where relatively steady-state conditions exist. An average value of
Ka can then be calculated using Equation 4.5.
The rate of atmospheric aeration has been demonstrated to be propor-
tional to the oxygen deficit (C -C)[7], i.e.,
Letting C - C = D,
15
-------
-
dt a
which upon integration results in
= K (4.5)
At
where
D = upstream deficit (ppm)
D-, = downstream deficit (ppm)
At = time of travel between stations (days)
As mentioned previously, if the DO concentration in a stream is strongl
influenced by biochemical oxygen demans, photosynthesis, nitrification,
etc. , the above approach is questionable. To minimize these effects,
the time of observation between stations should be minimized; however,
caution must be exercised in order to avoid conditions which may reflec
small differences of small numbers.
Since natural aeration is a function of surface renewal, which, in turn,
depends on hydraulic parameters, a number of empirical and theoret-
ical formulations have been proposed for determining K . Many of
the theoretical studies are based on transport phenomena and consider,
for example, effects of surface tension, molecular diffusivity, and
turbulence. In general, for stream applications these approaches re-
duce to formulations which express Ka in terms of two readily determin^
able stream parameters, viz, depth and velocity. The general form of
the equation is
,
1
(4.6)
Table 4.2 contains values of the various constants in Equation 4.6
determined by different investigators.
Observation of the values given in Table 4.2 indicates that each
formulation will result in a different value for Ka. It should be
realized that each formulation has its limitations. In the case of a
theoretical formulation [9] verification with stream conditions is
needed. And for the remaining empirical equations caution must be
used in applying results to stream conditions other than those from
which the observations were made.
16
-------
TABLE 4.2
Aeration Parameters for Determining K .
cL
Investigator
Churchill et al. [8]
O'Connor & Dobbins [9]
Games on [ 1 0
Lang be in & Durham [11]
Owens et al [12]
C
11.56
1.291
23. 17
7.59
21.62
kl
. 969
. 5
.73
1. 0
.67
k2
1.673
1.5
1.5
1.33
1.85
Usually both the depth H_ and the velocity V_ can be obtained as a
function of stream flow (Equation 4.7a), which would then permit
the aeration constant to be expressed as a direct function of flow,
as shown in Equation 4. 7b.
H = fj(Q)
V = f2(Q)
(4.7a)
From Equation 4. 6
Ka =
k
(4.7b)
H
For the Passaic River [6] this type of analysis resulted in the
following expressions for depth and velocity:
H = 0. 746Q
V = 0.020Q
. 398
. 524
The above expressions were used in Equation 4.6 for the various
values of C_, kl, and k£ given in Table 4.2 The resulting predictions
of Ka are shown in Figure 4. 3, indicating that values of Ka differ
widely according to the various formulations. (The precise reasons
for such a spread are not clear, although the investigators speculated
that perhaps it is because the Passaic is an unusually slow river and
may be out of the range of the formulations . )
17
-------
oo
K
(day
1.00
0.90
0. 80
0.70
0.60
0. 50
0.40
0. 30
0. 20
0. 10
0
Gameson
O'Connor & Dobbins
Owens, Edwards, & Gibbs
Langbein & Durham
Churchill, Elmore, & Buckingham
J I I I I I [_
I
I
100 200 300 400 500 600 700 800 900 1000
Flow (cfs)
Figure 4.3. K vs Flow, as Calculated by Various Prediction Equations (Whipple etal. [6])
3.
-------
Because of the significant spread in the predicted values of the aeration
constant, a serious question arises as to the general applicability of
the formulations. The formulation of O'Connor and Dobbins (see
Table 4.2) seems to provide an adequate description of aeration in
natural waterways, since good agreement was found by the authors
between measured and predicted values for several rivers [9] . Their
formulation was also considered to be the most representative of
aeration of the Passaic River. Other equations, e.g. , that of Churchill
et al. , also provide good results for several rivers [8] .
Since the natural aeration capability is affected by many chemical, bio-
logical, and physical factors, it is quite possible that no one formulation
will be universally applicable. Thus, the above formulations should be
used with discretion, particularly where "white water" turbulence or
vertical stratification is evident. Furthermore, the predictions are
for clean water, and if polluted water is to be aerated, adjustments
on a percentage basis must be made, to account for the effect of the
pollutants .
Since natural aeration is the most significant contributor in maintaining
the DO level in a stream, it is advisable to measure the natural aeration
coefficient before completing the final design of an aeration system.
For preliminary design estimates, either the method of O'Connor and
Dobbins or that of Churchill et al. is probably equally applicable.
Aeration transfer rates are quite sensitive to water conditions. Table
4.3 shows the large range of aeration transfer rates that have been
measured for different water conditions. (Values given are mass
transfer per unit of surface area per unit time. ) For still water,
values are on the order of .034 g/m2/hr; while for water conditions
involving small droplets, rates can be as high as 34 g/m2 /hr. Values
for flowing water lie between these two bounds, which encompass three
orders of magnitude.
The aeration constant is also a function of temperature. Reported
values are usually given in the form
K (T) = K (20) x
3, 2L
where
K (T) = aeration constant at temperature T
a —
K (20) = aeration constant at 20 C
3,
and 9 is a factor ranging from 1.015 to 1.047. A commonly used
value is 1. 024.
19
-------
TABLE 4. 3
Transfer Rate of Oxygen into Water
from Air for Various Flow Conditions
(Odum [13] )
1
• Velocity
(m/sec)
Still water 0. 0
j 0. 0 ;
Moving water , :
Stirred water i
Shallow circulating trough | 0.01
1 0.01
i 0.01
| 0.013
i 0.070
! 0.119
i 0.20
Sewage in circulating !
trough ! 0. 05
! 0. 15
Stream and ponds !
New York Harbor i tidal
Tank with a wave machine '.
Sea Surface
Summer j
Winter i
Silver River, Florida |
Subtraction-of-respir - j
ation method ! 0.21
Dye-measured-turnover i
method j 0.21
Green Cove Springs, \
Florida. From •
carbon dioxide by [
respiratory-quotient !
method j 0. 3
Small rivers, diurnal
1 oxygen curve analyses
Ohio River below
Cincinnati 0.05-
0. 09
• Bubbles and drops (K given
per area of drop or bubble) ;
Air bubble :
Air bubbles
Water drops
Depth
(m)
-
-
_
0.1
0. 1
0. 1
0. 1
0. 1
0. 1
0. 1
0.45
0.45
-
-
1.8
-
-
2.77
2.77
0.23
0.5-3.
4.8
:
_
-
Temp.
f°C»
20-25
-
25
0-10
10-20
20-30
12
17
14
13
25-26
25-26
-
-
12-20
2-7
23
23
24
2
15-25
i 37
20-25
1 24
i
K
(g/m2/hr
at 10%
saturation)
.034
0. 03-0. 08
0. 09-0.74
0. 037
0.043
0.47
0. 12
0.52
1. 12
3.8
0.38
1.5 I
0. 08 '
0.23 '
0.31
i
1. 1
5.2
0. 92 :
i
1.00 j
|
)
».
i
0.55
0.6-4. 3
1.5-5. 0
13. 1
2.8-28.
22-34.
20
-------
Biochemical Oxygen Demand (BOD)
Since the various oxygen demands in a stream or river depend on
processes which proceed at different rates, each one should be
described separately. For example, depending on the physical nature
of organic loadings, the fine particulate matter will oxidize in the
stream, while the larger particulates will tend to settle, forming the
benthos, and will oxidize at a different rate. Furthermore, if the
loading contains ammonia or nitrites, oxygen demands again occur at
different rates, and separate rate constants are needed to describe
each of the processes, which are commonly referred to as carbon-
aceous, nitrification, and benthol oxygen demand.
The oxidation of organic matter is essentially a chemical reaction,
initiated either directly by bacteria or indirectly by their enzymes.
As previously stated, the oxygen demand is exerted by two classes
of materials : carbonaceous organic matter and oxidizable nitrogen,
both of which may occur simultaneously [ 14-1 8] . Figure 4.4 illustrates
the two processes and the clear distinction that they proceed at different
rates. It can be seen in this case that a period of at least 19 days was
required before oxidation was essentially completed. The upper curve
represents the total BOD of the water (first and second stage). This is
readily obtained by measuring the oxygen uptake of stream samples as
a function of time (see reference 11). The lower curve represents
only the carbonaceous BOD and can normally be obtained by following
the methods given in Standard Methods [19], which call for the measure-
ment of the oxygen uptake as methylene blue or allylthiourea. The tests
are usually performed in a laboratory at 20°C. By taking the difference
between the two curves, a curve representing the nitrogenous BOD is
obtained. (In both of these tests, time and temperature can be closely
controlled, but it is difficult to simulate the dynamics of the river
environment, including the biological chain. )
The carbonaceous BOD follows a first-order reaction process, viz,
the rate of biochemical oxidation is proportional to the remaining con-
centration of unoxidized substance. Such a process is described
mathematically by
which yields, upon integration,
-K t
L = LQe C (4.9)
where
t = time (days )
L = carbonaceous BOD at time_t (ppm)
21
-------
Q
O
PQ
12 16 20 24
Time (days)
Figure 4.4. Grand River BOD Curve (Courchaine [18])
22
-------
L, = ultimate carbonaceous BOD (ppm)
K = c
arbonaceous deoxygenation constant (-3 )
K can be obtained from the slope of a semilog plot of Equation 4. 9
and is usually on the order of . 1/day at 20 C. In order to evaluate
Kc for a stream it is not necessary to conduct the test for 20 or more
days, since the slope of a semilog plot of Equation 4. 9 can be quite
accurately determined over a time period of 5 days. However, the
ultimate carbonaceous demand Lo must be correctly stated. This
can be clearly illustrated by examining Table 4.4 below. It can be
seen that for Kc - . 1 /day, at the end of 5 days only 68.4% of the
carbonaceous BOD has been consumed; hence, the ultimate car-
bonaceous BOD would be
BOD
BOD
ult
= L =
o
5 day
684
TABLE 4.4.
Percentages of Carbonaceous BOD as a
Function of Time for K = . 1 /day
Time (days )
0
1
2
5
10
20
Remaining BOD (%)
100
79.4
63.0
31.6
10.0
1.0
Consumed BOD (%)
0
20.6
37.0
68.4
90. 0
99.0
Both K and L, depend on temperature. The following expressions
were extracted from a recent study [54 ] .
LQ(T) =
(1 + 0.0113 (T - 20))
~ - O 0 r O .-,
range 20 - 35 C
LQ(T) = LQ(20) (1 + 0.0033 (T - 20))
range 2° - 20 C
23
-------
Kc(T) = Kd(2Q) (0.896) (1.126T-15)
->O ->o°r-
range Z - 32 C
T 20
K (t) = K (1.047 )
Kc(T) = K 2 (1.728) (0.985T-32)
ic° ->->r~
range 15 - 32 C
5T-32)
->->° ,,nO^
range 32 - 40 C
The oxidation of carbonaceous matter is a heterotrophic process
carried out by a great variety of different organisms and having
optimum temperatures ranging from 18° to 25°C. Many of the
bacteria obtain their food and energy requirements from the organic
matter present in the water and have generation times on the order of
20 to 30 minutes [14]. On the other hand, nitrification is an autotrophic
process, i. e. , one which is carried out by specific bacteria which
obtain food and energy from oxidation of ammonia and nitrites. The
nitrogenous oxidation process involves two genus organisms, Nitro-
somonas and Nitrobacter, each of which have an optimum growth
temperature range of 25° to 28°C [20]. Furthermore, generation
times on the order of 31 hours are necessary for nitriting cells to
develop [14]. Often nitrification may not occur for several days, e..g.,
in a study on the Passaic River [ 6] nitrification did not begin until
3 days had elapsed and, in the case of the Grand River, a situation
was found in which nitrogenous BOD did not occur until after 9 days
(see Figures 4.5 and 4.6[18].
Thus, it can be concluded that BOD tests should be conducted over a
long-enough period to determine if and to what extent nitrification
processes will occur. This is especially important if a river or stream
may receive the effluents from a secondary treatment plant.
The ultimate oxygen demand due to nitrification is directly limited by
the amount of oxidizable nitrogen available. The process involves
the oxidation of ammonia to nitrite, which is carried out by the genus
Nitrosomonas, and the second phase, the conversion of nitrite to
nitrite, which is carried out by the genus Nitrobacter. The overall
stoichiometric relation is given below:
NH3 +
(M. W.) —14 64 62
By comparing molecular weights, one finds that 1 part of NH3 will
consume 4. 57 parts of O2 by weight. Thus, 1 mg/1 of ammonia is
equivalent to 4. 57 mg /j. of BOD, and 1 mg/1 of nitrate produced by
24
-------
22
IV
Ul
18
(D
4-)
a
D
0)
Cud
>>
x
O
14
10
Carbonaceous +
Nitrogenous
Carbonaceous
J L
6810
Time (days)
12 14 16
Figure 4. 5. Total Carbonaceous and Nitrogenous BOD
as a Function of Time (Whipple [6])
-------
40
ro
32
24
BOD
(mg/1)
16
Carbonaceous BOD
10
12
14
16
18
20
Figure 4. 6. BOD Curve, Lansing Wastewater Treatment Plant,
24-hr Composite Sample, Taken July 12, I960.
-------
the oxidation of ammonia is equivalent to 1.03 mg/1 of BOD. It is
expected that some variation from the above ratio will exist because
of fixation by carbon dioxide, lowering the ratio of 4. 57 to perhaps
4. 3 or 4.4 [21] .
The rate of oxygen consumption by ammonia and nitrogenous matter
is directly related to the multiplication of the nitrifying bacteria. It
can clearly be seen from Figure 4.6 that the nitrogenous BOD curve
consists of two phases. In the first phase the nitrifiers are lagging;
the bacteria are undergoing multiplication, building up to a maximum
population. Once this maximum is reached, the process enters the
second phase (corresponding to the point of inflection), which is
similar in shape to curves obtained for carbonaceous BOD. Such a
process is clearly not first order. However, for mathematical
expediency in predicting oxygen sag curves for a stream, the process
is frequently assumed to be a first-order reaction [6, 22, 23], i.e.,
one in which
= K N = K (N - N.) (4. 10)
dt n nv o r l '
where
N = nitrification demand (ppm)
t = time (day)
K = nitrification constant (day )
N = ultimate nitrification demand (ppm)
N = nitrification demand at time_t (ppm)
and which integrates to
-K t
N = NQ(1 - e n ) ' (4. 31)
In order to take into consideration the lag time associated with the
nitrification process, Whipple et al. [6] found it advantageous to
include a lag-time term t in the following manner
3,
-K < t - t >
Nt = NQ(1 - e n a ) (4.12)
where the triangular brackets represent the singularity function such
that
< t - t > =0 f o r t < t
a a
< t - t > = t-t for t > t
a 'a a
27
-------
An important point when evaluating the various constants is the fact
that the time lag ta can be a function of where and when a stream
sample is taken. If oxidizable nitrogen was introduced near the point
at which the sample was taken, long lag times can be expected.
However, if the sample was taken a considerable distance downstream
from the introduction of nitrogenous matter, short lag times may
result, since incubation may have initiated during the transit time.
By describing the nitrification process as a first-order process, an
upper bound is obtained on the BOD, since such a representation
results in BOD values larger than those actually occurring (see
Figure 4.7).
As mentioned earlier, the effect of temperature is quite significant.
According to O'Connor [24] , the temperature dependence is given as
Kn = Kn(20) 1-°9 " (4. 13)
More detailed investigations of the twofold nitrification process have
been conducted by several authors [ 25-28] , in which the rate of change
of nitrogenous BOD is expressed in terms of ammonia oxidation and
nitrite oxidation. The expressions are relatively complex, nonlinear
in nature, and, to be of practical use, they often require the assistance
of a digital computer. In view of the many environmental conditions
affecting nitrification (temperature, pH, chemical composition of the
water, the specific genus of nitrifying organism present, etc.) and
the variability of determining other terms in the equations used to
predict the oxygen profile of a stream (Equations 4.2 through 4.4), it
may be concluded that the effort involved in obtaining the precision
offered by the more refined approaches is" questionable. For stream
aeration design purposes, the use of Equation 4. 12 will result in oxygen
demands in excess of what actually is required for nitrification.
However, in view of the many variables affecting stream aeration, its
predictions are perhaps as good as any others.
Benthal Demands S
One of the best methods of determining benthal demands S_ is to measure
the oxygen consumption of the deposit "in situ." This can be effectively
performed by using an opaque respirometer which is only open to the
stream's bottom surface. By monitoring DO levels and subtracting
the oxygen consumed by carbonaceous oxidation and nitrification at
the same time and temperature, the net oxygen demand due to the
benthos can be determined. Given the time period over which the
consumption occurred, one can determine the benthal oxygen demand
on a rate basis, usually expressed as ppm/day, or, when multiplying
by the stream height H_, as g/m2 /day. Having the results in the latter
form and knowing the percentage of the river bottom covered with
28
-------
BOD
(ppm)
First-Order
Approximation
Time (days)
Figure 4. 7. First-Order Representation of a Nitrification Process
29
-------
similar mud deposits permit the determination of the river demand on
a daily basis. This approach gives an oxygen demand rate which is
indicative of the steady-state consumption of the benthos. Making
these measurements during summer months as opposed to other times
of the year eliminates the need for temperature corrections when
determining the needs of an artificial aeration system, even though
such biological uptakes are temperature dependent. Since benthal
demand rates are generally fairly constant and low, the "steady-state"
value associated with the higher consumption rates during the summer
months is a safe value to use for aeration system design.
Net Photo synthetic Oxygen Production P
The production of oxygen by photosynthesis in streams has been con-
sidered by several investigators [6, 13, 29-35] . Depending on the
type of stream and associated climatological conditions, the contribution
can be significant. Relevant parameters include presence of aquatic
plants, especially free-floating and benthic algae; stream clarity;
intensity of sunlight; stream temperature; and stream velocity. A
study conducted on the Passaic River in New Jersey[6] indicated that
the eutrophic zone was found to be limited to 50 centimeters, while
investigations conducted on streams in Oklahoma [31, 35] which have
relatively low turbidity have indicated that photosynthetic productivity
in the benthos is considerable and at times in excess of that which is
free-floating. Other studies, conducted on the Sirsloch River [29]
in the U.S.S. R. and on the Rhine River in Germany [36] , have also
substantiated the finding that photosynthetic production of oxygen can
be significant. In general, it can be concluded that the contribution of
photosynthetic aeration should be considered when designing an aeration
s ys t em .
Photosynthetic oxygen, by its nature, is diurnal in production and conse-
quently is reported as a net production per 24-hour period, usually
expressed in mg / /day. A widely known method for measuring photo-
synthetic oxygen is the light-dark bottle technique. The measurement
procedure requires one clear and one opaque bottle filled with stream
water, which are incubated at several depths in the stream. At the
end of the incubation period, which may run from one to several days
depending on the rate of oxygen production, the difference in oxygen
concentration between the water in the two bottles is assumed to be
equal to the photosynthetic oxygen production.
The light-dark bottle technique requires the assumption that biological
processes in the bottle are the same as those in the stream. This is
not a completely valid assumption because the quiescence imparted to
the samples affects the natural exchange of food and waste products to
and from the organisms, and also because a higher surface-to-volume
ratio is introduced in the sample bottles. In addition, this technique
neglects photosynthetic oxygen from benthic algae or attached plants,
an assumption which can only be made in relatively deep or turbid
30
-------
streams, where sunlight cannot penetrate to the benthos. If measure-
ments are being made on a relatively clear, shallow stream, they
could easily be in error. For this type of stream, Stay et al. [31]
have developed a method which uses three plastic chambers placed
over the stream bottom. Two sets of measurements are required, the
first with plastic bottom plates covering the benthos and the second set
without plates. Each of the three chambers is employed differently.
One is clear and closed to the atmosphere, the second is clear and
open to the atmosphere, and the third is black and closed. The oxygen
concentration in the closed, clear chamber is affected by photo-
synthesis and respiration, while the oxygen in the black chamber is
affected only by respiration. The clear chamber open to the atmos-
phere is a control. The tests are run with and without bottom plates,
in order to account for respiration and photosynthesis of the benthos.
The differences in concentration between the clear and dark chambers
is the photosynthetic contribution.
The oxygen production should be expressed in terms of daily average
rate (ppm/day). A more precise expression can be derived to account
for the variation in sunlight intensity during the day. This precision
is not required, however, when the object is to design an aeration
system.
Longitudinal Dispersion
Longitudinal dispersion is the process of spreading and distributing
a mass of pollutant or oxygenated water along the length of the stream.
The spreading is a result of two diffusion processes, molecular
diffusion and eddy diffusion. In turbulent flow the effect of molecular
diffusion is negligible compared to that of eddy diffusion and conse-
quently can be neglected.
In aeration system design, the rate of spreading and the resulting
distrubition are important. The dispersion characteristic of flowing
water has been found to vary greatly [36-42]. Table 4.5 contains
values of dispersion coefficients Dj_, determined by various investigators
for different rivers.
Examination of the table shows that longitudinal diffusion coefficients
are larger for estuaries than for a conventional stream or river.
This is primarily due to the mixing effects that occur in estuaries.
An interesting point is the considerable increase in the diffusion
coefficient of the Delaware River as a result of using a diffuser aerator
(141,000 ft2 /sec, compared with 1075 ft2 /sec). The investigators
speculated that the large value of the dispersion coefficient was a
consequence of turbulence caused by the large bubbles emitted by
the diffuser, and that a small-bubble diffuser might give smaller
values.
Several approaches have been devised to predict the magnitude of the
longitudinal diffusion coefficient. Some of the schemes have only
31
-------
TABLE 4.5
Typical Longitudinal Dispersion Coefficients for Rivers
Delaware (estuary) [40]
James (estuary) [49 ]
Hudson (estuary) [40 ]
Delaware (estuary) [37]
Delaware (estuary) [44]
Copper Creek, Virginia
[42]
Clinch River, Tenness ee
[42]
Powell, Tennessee [42]
Coachella Canal,
California [42]
Depth
(ft)
1.6
2.8
1.6
1.3
2.8
7. 0
6.9
1.9
2.8
5. ]
Width
(ft)
_ _
--
--
--
52
60
53
61
154
195
175
118
111
80
Discharge
(ft3/sec)
_
54
300
48
483
323
3000
1800
240
140
950
DL(ft2/sec)
195-350
325
4, 850-16, 100
1075
141, 000*
210
230
102
106
150
580
500
87
102
103
*with artificial aerator in operation
been applied to laboratory channel conditions, while others have been
verified with natural stream data. The "routing procedure" as
described by Fischer [42] appears to give the best predictions when
compared to dye-concentration measurements. However, this approach
requires the use of tracer dyes and is mathematically involved to the
extent of requiring the use of a computer. If cost and personnel
limitations do not permit use of the routing technique, the disperson
coefficient can be estimated from field measurements of velocities and
water-surface slope. This technique has also been developed by
Fischer [4]] and will only be briefly described here. (If one is faced
with evaluating the diffusion coefficient of a stream, references 41 and
42 should be consulted.)
The procedure involves measuring the slope of the stream and the cross-
sectional geometry and velocity distribution at one or more typical cross
sections. For many stream situations these field measurements can
usually be completed in a day by a three-or four-man crew.
Equation 4. ]4 is used to calculate DT for streams having a width-to-
depth ratio equal to or greater than o.
32
-------
_!_
A
f
q'(z)d(z)
(4.14)
where
d(z;
q'(z)= J u'(y,z)dy
o
E = .23 dU transverse turbulent mixing coefficient [42]
u'(y,z) = velocity at any point relative to the mean flow
velocity, i.e. , u'(y,z) = u(y,z) - u", where
u(y,z) is the actual velocity
q'(z) = depth of flow at any point on the cross section (ft)
A = area of flow (ft* )
b - width (ft)
y = vertical coordinate
z - transverse coordinate
* 1/2
U = (grSe) function velocity, where g = acceleration
of gravity, r = hydraulic radius, and Se = slope of
the energy gradient
Although the calculation of D^ appears complicated and evaluation by
hand is tedious, the equation can easily be programmed for a computer.
Normally, depth and velocity measurements are taken at a minimum
of 20 vertical sections across the stream. The integrals in Equation
4. 14 are then approximated by summations, viz,
n
k=2
Az
33
-------
where
q'. = i(d. + d. )u'.
i Z % i i+l i
u. = velocity in the i vertical slice (ft/sec)
u1. = u. - u (ft/sec)
u = mean velocity of the flow within the entire cross
section (ft/sec)
d. = depth at the i vertical slice (ft)
Az « width of vertical slice (ft)
n = number of vertical slices (usually more than 20)
# th
E = . 23d.U transfer coefficient between i-1 and the i
z. i . n.
i vertical slice
Table 4.6 has been reproduced from reference 42 and illustrates the
procedure to be followed.
Simplification of the Oxygen Balance Equations (4.2-4.4)
Equations 4.2 through 4.4 describe the temporal and spatial distribution
of BOD and DO. The flow volume and the cross-sectional area are
functions of both space and time. In order to simplify the treatment,
the assumption is usually made that the most severe condition for
design purposes is the one which can be made to approach steady state.
Although steady-state conditions are seldom, if ever, attained, the
errors introduced by this assumption are minor compared to variations
occurring in the physical stream and in measurement procedures.
If, in addition to the steady-state conditions, the following assumptions
are applicable:
1. Stream flow is uniform.
2. Carbonaceous and nitrogenous BOD are first-order
reactions .
3. Photosynthetic oxygen, benthal demands, and accrual
processes are uniform along a stream section and are
determined on an average daily basis.
4. Vertical and transverse variations are negligible.
34
-------
TABLE 4.6
Sample Calculation of Dispersion Coefficient from Field Data Test [41]
Data
Vertical
slice
(1)
)
2
3
4
5
6
7
8
9
JO
) 1
12
]3
14
15
16
17
18
19
20
21
22
23
24
25
i
Distance
from left Depth at ]
bank to
left side
start of I of slice
slice (ft) d. (ft)
(2) l(3) i
0.0 : 0.00
2.5 i 1.00
5.0 2.15
7. 5
10.0
12.5
15.0
17.5
20. 0
22.5
25.0
27.5
30.0
32. 5
35.0
37.5
2.61
3.30
3.41
3. 32
3. 14
3.00
3.01
3. 10
2.99
2.78
2.64
2.59
2.66
40.0 2.90
42.5 ! 2.98
45.0 '• 2.84
47.5 2.78
50.0
2.72
52.5 2.39
55.0
57.5
J.82
0.84
60.0 0.34
Computed Quantities
Mean velocity
in s lice p. .
(ft/sec)1
(4)
0. 15
0.25
0.35
0.42
0.65
J. 35
1.80
2.30
2. 35
2.40
2.50
2.55
2.70
2.75
2.65
2.45
2.30
2. 15
1.85
1.50
1. 10
0.70
0. 40
0.20
Discharge through
slice relative
to mean velocity
q! &z (ft/sec)
1 (5)
-2. 00
-5.89
-8. 31
-9.80
-9.20
-3.33
Cumulative
relative
discharge
J
Z q1. A z
i=l. l
(ft3)/sec)
(6)
-2. 00
-5.89
k A7 J-1
•*- "^ •^ ,-,< i\ -
Z E7 d ^ qi
j=2 Z) j i=l
(ft)
(7)
0. 0
-65.7
-16.20 { - J 2 1 . 9
-26.00 -200.2
-35. 19
-278.9
-38.53 -378.5
0.43 -38.09 ' -493.7
4.25
-38.84 -620.9
4.53 -29.31 : -744.8
4.99 ! -24.32 -851.2
5.74 ' -18.58 : -934.2
5.80 -12.79 • -1,003.2
6.46 -6.33 ; -1,057.7
6.56
5.93
4. 90
4. 07
2. 94
0.73
-1.69
-4. 13
-5.5]
-4. 48
-2.28
0.24 ; -1,087.6
6.17 : -1, 086. 4
1 1.05
15. 12
18.06
-1, 057.7
-1 , 014. 4
-958. 3
18.79 : -884.6
17.09 . -804.5
12.96 ! -728.4
7.46 j -653.6
2.98 | -579.5
0.70 i -440. 3
0. 10 -0.70 0. 00 ' -241. 0
1
^ ~Z[ column (5) x column (7)
34,800 ft4 /sec _.„ ft2 ,_
LJ — — — s — u J'-l it / ^ C <_
total area 149 ftz
IjO
Ul
-------
then considerable simplification of Equations 4.2 through 4.4 results:
) ^-^ + K (c - C) - U^ -KL+P-KN+A-S=0 (4.16)
L. j - 2 a s dx c n v ;
dx^
TT
dx
D _ U _ K N+ N = 0 (4.18)
L2 dx
Dobbins [45] has shown that, for freshwater streams, neglecting the
longitudinal diffusion terms results in a prediction which is at most
a few per cent different from a solution which includes the effects of
U2
diffusion. If the value of -^ p— is on the order of 100 or more,
a L
neglecting the diffusion term results in a negligible difference. It
should be noted, however, that the value of DL should include the
effects of the aerator system (see Table 4.5.)
U2
If one is dealing with an estuary, the ratio of ? may be quite small,
a L
meaning that the diffusion is significant. This would be indicative of a
relatively slow-moving body of water which experiences good mixing
(e.g. , tidal effects). If one concludes that the diffusion terms are
negligible, Equations 4. 16 through 4. 18 reduce to
-K(C -C) + KL + KN-P-A+S (4.19)
dK av s ' c n v
+ KL-L =0 (4.20)
x c a
+ KN-N=0 (4.21)
dx n a
Equations 4.20 and 4.21 can be integrated to give
36
-------
KC
~ "zrr X. JL/
T - T 0 U 4- a
JL = _L e t ——
o K
c
K
- ~nx N
N = N e U + T-r-^
o K
n
which can be substituted into 4. 19 to give
K K
c n
K ~ TI x, ~ T T x
dC a,r u + K N e u =P-L-N+A-S
te~ TT(Ca-C) +KcLoe n ° a a
(4.22)
or
K K
K c n
dD , a p. v , "UX+KNe " U . T ._, _ A^c
j— + -7-rD=KLe no +L+N-P-A+S
dx U c o a a
(4.23)
where D represents the oxygen deficit (C - C).
If nitrogenous BOD is small compared to carbonaceous BOD, the two
can be combined in a single expression for BOD. Hence, Equations
4.4, 4.18, and 4.21 may no't be needed in many stream analyses . The
effects of nitrification would be contained in Kc, L, and La and would
not be differentiated from carbonaceous consumption of oxygen. This
has often been done in many of the analyses that appear in the open
literature.
37
-------
SECTION V
AERATION SYSTEMS FOR RIVER
AND STREAM APPLICATIONS
Introduction
Various types of mechanical surface aerators, diffusers, and side-
stream mixing systems have been used in experimental river and
stream aeration projects. Downflow contactors (U-Tubes), although
not used previously for river and stream aeration, have been used
successfully in impoundments and lakes and should be adaptable to
river and stream applications.
In this section the aeration systems which could be used for rivers
and streams are described, and results of recent tests using these
systems are evaluated. The specific application of any of these
systems is extremely site dependent, thus site considerations are
included in the discussion of each system. In most cases, aeration
devices designed specifically for waste treatment have been used
without modification in the stream application. Some improvements,
particularly for surface aerators, can be made to enhance their per-
formance when used in stream environments, since strong pumping
action is not as important there as it is in waste treatment.
Downflow contactors, diffusers, and sidestream mixing systems can
be used with both air and pure oxygen, and the relative merits of the
two approaches are discussed in this section. Since the efficiency of
diffuser systems depends strongly on the diffusion mechanism for
small bubbles, this subject is included in the discussion of diffusers.
Mechanical Surface Aerators
A variety of designs for mechanical surface aerators are available
from a large number of manufacturers. There are no designs of
surface aerators specifically for river and stream aeration, available
systems being used only in waste-treatment applications. A compre-
hensive description of surface aerators can be found in Manual of
Practice No. 5, published by the Water Pollution Control Federation
[47 ] . Some of that information is included in Appendix A.
For river and stream applications, it is desirable that the system be
float-mounted, so that a wide range of flow conditions can be accommo-
dated.
The following types of surface aerators have been developed. (See
Appendix A for more details:
39
-------
Rotating Plate - Creates a peripheral hydraulic jump that
accomplishes oxygen transfer through
entrainment.
Updraft - Pumps large quantities of water at the
surface at relatively low heads. Aeration
efficiency related closely to efficiency
as a pump.
Downdraft - Oxygen is supplied by air self-induced
by negative head produced by rotating
element.
Combination - A rotating element and a sparge ring are
combined to transfer oxygen by dispersing
compressed air fed below the surface to a
rotating agitator or turbine.
Brush - A horizontal revolving shaft with attached
brush-like elements extending slightly
below the surface.
The transfer of oxygen by surface aerators to a body of water is a
direct result of breaking the water into( small droplets and inducing
turbulent mixing on or near the water surface. Each process causes
new unsaturated water surfaces to be exposed to the air, resulting in
their becoming saturated and subsequently mixed with the bulk of the
water.
A significant difference between using surface aerators in streams and
using them in lakes or still bodies of water is the superimposed effect
of the stream current. The current provides the aerator with a supply
of water low in DO compared to water in the mixing zone of influence
of the aerator. Hence, the DO deficit will tend to be greater, enhancing
the transfer of oxygen. Furthermore, since part of the energy con-
sumed by surface aerators is in the pumping of water, aerators which
expend relatively low pumping energy should be considered for stream
applications. The pumpage associated with surface aerators should
be supplied by the manufacturers. For example, the pumping rates
published by two manufacturers for various horsepower units are
listed in Table 5.1.
In a recent study, McKeown [47] measured velocity profiles and DO
concentrations in stabilization basins as affected by surface aerators.
Although the study dealt with basins and not streams, the results are
characteristic of surface aerators and can provide preliminary informa-
tion for stream applications. The investigation included a number of
different sizes of surface aerators. Figure 5. 1 shows the extent of
influence of a 75-hp unit. The profiles shown are typical of surface
aerators .
40
-------
Distance from 75-hp Unit
125 ft
Velocity-Depth-Vector
Figure 5. 1. Example of Velocity Profiles at Various Distances from a 75-hp Aerator
(McKeown [47])
-------
TABLE 5. 1
Surface Aerator Pumping Rates
Horsepower.
5
7.5
10
15
20
25
30
Manufacturer A
"Aqua -Jet"*
3400
3800
5000
6] 00
8300
9800
12,500
Manufacturer B
"AQUARIUS"**
2000
3400
4000
5000
6500
7500
8000
Aqua Aerobic Systems, Inc., Rockford, Illinois
Keene Corporation, Aurora, Illinois
Table 5.2 [47] lists for various size aerators the depth of penetration
of outward flow, the radius at which this depth is reached, and also
the radius at which all discernible outward flow at the surface ceases.
The position of the aerators was such that there was no interference
from adjacent aerators or basin banks. It is interesting to note that
the depth of penetration of the crossover point rarely exceeded 50% of
the bas in depth.
Table 5. 3 [47] contains data on influence of flow and the net DO con-
centration changes at each foot of depth and at varying distances from
the aerator.
It was concluded that the maximum zone of influence for different-size
aerators was 300-400 feet for 100-hp units, 200-250 feet for 75-hp
units, 150-250 feet for 40-to 60-hp units , and out to 150 feet for
1 0- to 25 -hp units.
Attempts to compute the pumpage associated with the various aerators
were not successful, primarily because the flow associated with the
surface aerators is three-dimensional and the data collected did not
consider the vertical dimension.
In another study conducted by Burns et al. [48] , two 15-hp surface
aerators were used to raise the DO in the Jackson River near Covingtoi
Virginia. It was reported that, by placing the aerators at sag points,
transfer efficiencies on the order of 2.2 pounds O?/hp-hr (20°C and
0 upstream DO) were obtained. Average stream velocities were
approximately 3/8 ft/sec.
Kaplovsky et al. [ 49] conducted an investigation on aerating the foreba^
of a canal at Lockport, Illinois, using two model 100 Hi CoWave
42
-------
TABLE 5.2
Vf.
Zone of Outward Aerator Influence (McKeown [47])
Type
NPHP
f
L
L
L
H
JL
L
L
L-
H
H
H
L
L
L
L
10
10
DHP
8
8
10 8
25 2]
40 i 41
40
40
40
50
27
34
26
47
50 47
50 i 50
50 1
50
60 52
60 52
L ; 60
L ' 60
L 75
L 75
L 75
49
49
54
64
55
H 75 64
L i 75 1 59
L 100 j -
L ' 100
L
100
-
-
D
(ft)
11
12
1 1
10
8
8
8
8
13
14
9
12
1 1
11
14
14
12
8
14.
14
10
10
18
18
18
\ '
d
(ft)
5
3
4
2
2
2
3
2
7
6
3
5
6
6
3
3
4
2
5
4
2
5
9
9
7
r}
(ft)
50
50
75
40
r2
(ft)
110
80
100
60
50 , 80
50
50
50
50
50
50
60
60
75
50
75
100
100
75
75
40
60
150
100
100
80
80 i
1
80
100 !
80
60
110
110
110 j
75
110 i
125
110
125
no
60
100
200 !
150 ;
175
i
*
NOTE: (1) Nearest restriction (bank or aerator) at least 2.5r->
distant from aerator reported.
(2) Symbols:
L = low speed aerator
H
NPHP
DHP
D
d
rl
high speed aerator
name plate horsepower
drawn horsepower
basin depth
maximum depth of penetration of outbound flow
radius from aerator shaft to point of maximum
depth (d)
radius from aerator shaft to point where
surface outward flow blends with background
43
-------
TABLE 5. 3
Sample DO Profiles at Various Distances from Aerators (McKeown [47])
(values shown inmg/1)
i
60-hp Unit i
1_
i
!
1
100-hp Unit |
1
25-hpUnit
50-hp Unit
10 -hp Unit
RadialDistanc e i
Depth [
0 !
1 i
2
3 :
4
5
6 :
7 ;
8-10
11-13 :
0 '
1
2
3 ;
4 ;
5-12 i
!
1 !
2 !
3
4
5
6-10
0
1
2
3
4
5
6
7-10
0
1
2
3
4
5
6-10
25 ft
2.4
2 . 4*
1. 9
1.8 !
1.5
1.4
1.4
1.2
1.2
1.2
2.0
1.4
0.2
0.0
0.0
0. 0
2.7
2.3
2.2
2.2
1.9
1.9
1.4
2.5
1.7
1.5
1.4
1. 1
0.8
0.6
1. 0
2.0
2.0
1.9
1.6
1.5
1.5
1 1'5
-.
50 ft i 75 ft
2.1 i 2-0
2.0 ; 1.8
2.0 1.7
1.8 ; 1.7
1.5 ; 2.0
174" 1.6
1.4 1.9
1.4 1.5
1.4 1.4
1.3 1.2
1. 0 0.4
0.4 0. 1
0.2 ; 0.0
6.0 j 0.0
0.0 i 0.0
0.0 : 0.0
2.7 ' 2.2
2.6 ' 2.1
2.4 : 2.0
2.3 ; 1.9
2.4 ' 2.0
2.2 i 1.9
1.9 ;' 1.7
1.5 \ 1.6
1.5 1.4
1.3 1.4
1.3 1.3
1.1 1.3
1.0 1.3
1.0 1.3
1.0 1.3
1.9 ' 1.7
1.7 1.5
1.6 1.6
1.5 1.5
1.5 1.5
1.5 1.5
1.5 • 1.4
100 ft
1.7
1.5
. 4
1.3
L_o
1.2
1.2
1.2
1.2
1.2
0.0
0. 0
0. 0
0. 0
0.0
o.o
1.9
1.4
1.5
1.4
1.4
1.4
1.3
1.3
1.6
1.5
T7T
1.4
1.5
K5
i I-3
Velocity crossover zones are underlined
44
-------
Aerators (73 hp each). Flow rates varied from 800 cfs to 5600 cfs,
and the average depth of the bay was 15 ft.
It was found that at low flow rates (1000-3000 cfs), transfer efficiencies
on the order of 1. 8 Ibs Oz/hp-hr were obtained, while for high flow
rates (4000-6000 range), efficiencies increased to on the order of 4.4 Ibs
G>2/ hp-hr (see Figure 5.2). The values quoted above have been correc-
tec to standard conditions, using the average of upstream and downstream
concentrations.
Figure 5.2 illustrates the differences in oxygen transfer at different
rates of flow. The rapid transition shown is likely to occur if flow
conditions changed, e.g., laminar to turbulent. However, from the
data presented in the report no evidence of such a phenomenon could be
found. It is quite possible that a more realistic curve of increasing
efficiency would show a gradual increase, rather than an inflection.
The authors speculated that the appreciable increase in efficiency might
be attributable to increased shearing of bubbles in the upstream region
or greater retention time of the bubbles in the water. However, it is
more likely that, as the stream velocity increases, more unoxygenated
water is fed to the aerators, resulting in a greater O-> deficit and a
greater rate of transfer.
Whipple et al. [ 6] also found from studies using surface aerators on
the Passaic River (average width 100 ft , average depth 7 ft in low
flow) that an increase in efficiency occurred with flow rate, as shown
in Figure 5.3. The efficiency increased from about 1.2 lb C>2 /hp-hr
at 120 cfs to approximately 2.8 lb/O2/hp-hr at standard conditions.
These values are consistent with those of Kaplovsky. However, the
transition to higher efficiencies is more gradual and perhaps more
representative of the actual situation. Some question remains as to
what happens to the efficiency when the discharge approaches zero.
In contrast to the results of Kaplovsky and Whipple, Susag et al. [50]
found in a series of laboratory tests that very little difference in
oxygen transfer efficiencies occurred between flow and non-flow
conditions. His average results, corrected to standard conditions
(according to the aeration equation) for a 9-in. and a 12-in, unit,
are contained in Table 5.4.
In the authors' opinion the values associated with the 9-inch unit were
more representative, since very little splashing on the sides of their
test tank occurred during that test. The velocity range investigated
was from about . 1 ft/sec to .65 ft/sec for the flow-through tests.
In addition to his tests on the Passaic River [6] , Whipple directed a
series of tests of surface aerators and bottom diffusers on the Delaware
River near Philadelphia in order to determine the practicality of
oxygenating a deep, navigable river [44] . The intent of the study was
45
-------
a
-C
—.
N!
O
-------
4.0
200 400
600
800 1000
Discharge (cfs)
1200 1400 1600 1700
Figure 5. 3. Oxygen Transfer Rate vs Flow for Mechanical Aerator (Whipple [6])
-------
TABLE 5.4
Average Oxygen Transfer Efficiencies,
in Ibs O2/hp-hr (Susag [50] )
Turbine Size
9 inch
12 inch
Flow
3. 73
4. 11
No Flow
3.89
4.86
to test systems on a wide, deep, navigable stream to determine
efficiency, economy, and operating characteristics and to prepare
prototype designs and cost estimates for installations appropriate
to such rivers. This is in contrast to the study on the Passaic River,
which is a relatively small river.
The oxygen deficiency on the Delaware extended over a distance of
about 40 miles, including areas containing heavy industry. The river
is characteristically 2000 to 2500 ft wide above the confluence of
the Schuylkill and somewhat wider below this point. The main channel
is 40 ft deep, with adjacent anchorage areas of about 30 ft. Normal
velocity range is ] to 1-1/2 ft /sec.
Oxygen transfer rates were reduced to standard conditions. The a
and p values were found to be equal to ]. Before being reduced to
standard conditions, the surface aerator transfer rates varied from
1.18 Ibs Oz/hp-hr to 3. 78 Ibs Oz/hp-hr, with an average of 2. 56 Ibs
Oz/hp-hr; after conversion, the range became 1.29 to 4.50, with an
average of 3.06 Ibs Oz/hp-hr.
These values are substantially higher than the average transfer rates
for the mechanical aerator tests conducted on the Passaic River [6].
The average field condition of 2. 56 Ibs Oz/hp-hr on the Delaware
compared to 1.04 Ibs Oz/hp-hr on the Passaic, and under standard
conditions the averages were 3.06 Ibs Oz/hp-hr and 2. 12 Ibs Oz/hp-hr,
respectively. The above results are within the range of values obtained
by other investigators for flow conditions (Table 5. 5).
In a study conducted by Brookhart [5] on the Miami River, floating
mechanical aerators were located at an oxygen sag point occurring
in an impoundment. Four 20-hp high-speed surface aerators were
selected (Welles Products Company, Roscoe, Illinois), and were
placed across the channel so that each aerator would handle equal
amounts of discharge. The channel was approximately 300 ft wide
and the depth about 7 ft. Water temperatures were approximately
25°C. The average transfer rate, computed for tests in which the
incoming water had a DO of approximately 3 ppm, was 1. 0 Ibs
O2/hp-hr (corrected only to standard temperature) in a discharge
range of 400 to 800 cfs (velocity . 1/3 fps).
48
-------
TABLE 5. 5
Average Oxygen Transfer Rate Obtained by
Different Investigators for Surface Aerators
River
Delaware River
Passaic River
Laboratory Channels
Chicago Canal
Miami River
Jackson River
Flambeau River
Investigator
Whipple et al. [ 44]
Whipple et al. [6]
Susag et al. [50 ]
Kaplovsky et al. [49]
Brookhart [51 ]
Burns et al. [48 ]
Lueck et al. [53 ]
Standard Conditions
(Av. Ibs O2/hp-hr)
3. 16
2. 1
-4.0
1.5—4. 5
= 1.0*
2'2
.44 —.90
Velocity Range
(fps)
1 —1-1/2
] /5 —2 /5
1/10-2/3
1/4—1/2
1/3
3/8
•j~
Only corrected to standard temperature
##Not at standard conditions (T= 25°C, Incoming DO = 2 ppm)
*>.
-------
In a surface-aerator study similar to the one conducted by Brookhart,
McKeown installed three 50-hp surface aerators in Gulf Island Pond
on the Androscoggin River in Maine [52] . The pond resembled a
sluggish river, and the BOD^ was generally between 2 and 4 ppm. The
three 50-hp aerators were used with incoming DO levels in the water
of less than 3 ppm. Transfer efficiencies of 1.7 Ibs C>2/hp-hr were
obtained. It was speculated that if the spacing between aerators had
been increased, somewhat higher efficiencies would have resulted.
A comparison of spray aerators and turbine venting aeration was made
in studies conducted on the Flambeau River in Wisconsin [ 53 ] . The
spray aerators were furnished by Welles Products Company and were
known as "Aqua-Lators . " Four Aqua-Lators were installed in the
tailrace of the Pixley Dam powerhouse and used to create a spray by
pumping river water through a slotted disc. Each unit covered an
area approximately 35 ft in diameter and sprayed water in the vicinity
of 8 to 10 ft high. It was reported that droplets in the spray were
rather large.
The efficiencies of the spray aeration device are relatively low and are
compared to results obtained from turbine venting at the same site
(Pixley Dam).
The turbines added 1 . 9 to 2. ] Ibs C>2/hp-hr with oxygen saturation
levels ranging from 3% to 12% and stream flow of 464 to 517 cfs,
while for similar conditions the aerators only added from .44 to .90
Ibs O2/hp-hr.
In the tests only about 7. 5% of the total flow was pumped by the aerators,
and, except for minor leakage, all of the flow passed through the
turbines.
From the limited number of tests, the turbines were 2-1/2 to 5 times
more efficient than the spray aerators.
A summary is presented in Table 5.4 of the transfer efficiencies for
surface aerators found by several of the investigators mentioned in
this section. Most of the tests were conducted with updraft-type
aerators, thus the variations in results do not appear to be attributable
to the type of aerator used. All of the transfer efficiences have been
corrected to standard conditions, except for the results reported by
Brookhart [ 5] ] , which have only been corrected to standard temperature,
and those of Lueck [53 ] , which are reported for a temperature of 25°C
and an incoming DO concentration of 2 ppm.
D iff user Systems
In diffuser systems, air or molecular oxygen is piped to a distribution
system where it is introduced directly into the water through porous
ceramic heads (usually silicon dioxide or alumina), finely perforated
50
-------
tubing, networks of nozzles, orifices, or jets. The diffuser heads
can be installed at various depths below the water surface. Increasing
the height of water above the system provides greater contact time
between the gas bubbles and the water, thus increasing oxygen
absorption. However, by increasing the height of water, one also
increases the hydrostatic head against which the gas must be pumped.
If an air diffuser system is being used, higher operating pressures
are then required, resulting in an overall decrease in transfer
efficiency measured in Ibs O2/hp-hr absorbed. If, on the other hand,
the diffuser system utilizes gaseous oxygen, an increase in the hydro-
static head does not present a problem, since the source would be under
high pressure. Furthermore, when using oxygen instead of air,
absorption rates will increase by up to a factor of five because of the
increase in partial pressure.
In general, the porous-ceramic and perforated-plastic-tubing diffusers
produce smaller bubbles than the other system types and consequently
allow for a greater percentage of oxygen to be transferred to the water,
due to the increased area-to-volume ratio of smaller bubbles. However,
associated with these systems are greater head losses, which for air
diffusers produce lower system efficiencies as measured in Ibs C>2/hp-hr
Furthermore, the fine-pore diffusers are more susceptible to clogging.
The total rate of oxygen transfer is associated with three different
components: bubble formation, bubble ascent, and the breaking of the
bubble at the water surface. For small diffuser openings and low
air-flow rates, oxygen transfer during bubble formation is appreciable.
Conversely, for large aperture diffusers with high air-flow rates,
little transfer occurs during bubble formation, primarily because the
bubbles are formed while rising in the water. The aeration during
bubble bursting at the water surface is primarily due to turbulence.
The oxygen transfer during ascent depends on such factors as size
of the bubble, depth, terminal velocity, etc.
Diffuser System Tests Using Air
Diffuser tests using air have been conducted as early as 1943 in the
Flambeau River at pixley, Wisconsin. Those tests have been de-
scribed by Tyler [54] and Wiley et al. [55] . Noticeable improvement
in DO concentration was obtained for several miles along the river by
introducing air through carborundum plates and porous tubes located
under ]2 ft of water in the headrace and tailrace waters of the Pixley
Dam. Absorption efficiencies were on the order of 7%. (The
efficiencies for diffuser systems are often reported as the percentage
of oxygen absorbed by the water. ) Results of subsequent tests on the
Flambeau River, using drilled pipe (1 /8-in. holes) as diffusers under
4 to 5 ft of water, were reported by Palladino [56]. Considerably lower
absorption efficiencies (1.7%) were obtained for incoming water having
DO levels from 2 to 3 ppm. Operating efficiency was about .38 Ibs
O2/hp-hr.
51
-------
Low absorption efficiencies have also been reported by Bohake for fine
and coarse diffusers using air to aerate the Lippe River at Heil,
Germany[ 57] . Diffuscr depth was about 3 ft, and the incoming water
was nearly depleted of oxygen. The river depth was approximately
]2 ft at the installation. An absorption efficiency of 1.7% was obtained
with the fine-bubble diffuser while the coarse-bubble diffuser ( = 1/4-in,
nozzles) yielded an absorption efficiency of only ) . 0%. Subsequent
tests using diffusers located on the stream bottom increased the
absorption efficiency to J. 5% and the system efficiency to approximately
] . 65 Ibs Q£/hp-hr [73]. The importance of depth was emphasized
in each of the above river applications; the greater the depth of the
diffuser the greater the absorption efficiency, i.e. , the percentage of
oxygen absorbed.
More recent studies using air diffusers have been conducted in the
Passaic [&1 and Delaware [44] Rivers in New Jersey. Similar diffuser
systems were used in each case. The system consisted of two 8-in.
underwater heads with a total of ]60 diffuser nozzles, each having
twelve 5 /32-in. ports . The tests were conducted at various depths in
the Delaware and for a single depth in the Passaic River. Table 5.6
summarizes the results.
TABLE 5.6
Diffuser Results from the Passaic and Delaware Rivers
Mean
Depth
(ft)'
7.2
7.2
7.2
13.2
25.0
38.3
Location
Passaic River
Passaic River
Passaic River
Delaware River
Delaware River
Delaware River
Approximate
Mean Water
Velocity
(ft/sec)
1/4
1
1
1
Ail-
Flow
]&
J V . 6
9
12
12
12
Average
Absorption
of Op
%
2.7
2.0
4.2
5.0
6.3
7.0
Average
Efficiency
(Std. Cond. )
(Ib Oz /hp-hr)
1.20
1.36
.93
.68
In comparing the results, it should be realized that the diffuser systems
are essentially the same, the air flow rates being quite similar, and
the only significant difference other than diffuser depth is the lower
water velocity of the Passaic River. The conclusion can be reached
from the table that oxygen absorption increases substantially with depth.
However, the operating efficiency in Ibs O2/hp-hr increases from the
7.2-ft depth to the 13.2-ft depth, after which it decreases.
The above comparisons for different depths can be made because the
diffuser systems were essentially the same. However, there are a
large number of factors affecting absorption of oxygen from a diffuser
52
-------
system, and caution must be exercised when conclusions on diffuser
performance are made by comparing results of entirely different
systems. For example, one might consider diffuser test results from
two different types of systems in which the air flow rates, water
velocities, and depths of submergence were the same. The oxygen
absorption efficiency and system efficiency in terms of Ibs O2/hp-hr
might differ considerably, since the mean air bubble sizes emerging
from the diffusers could be quite different. The bubble size would,
among other factors, depend on the size and number of diffuser
openings, which would not be the same unless similar equipment was
used.
Additional useful results could have been obtained in the Delaware
River study if the air-flow rate had been varied at the different depths.
This would have provided information for optimizing system efficiency
in terms of Ibs CU/hp-hr and air flow versus the number of nozzles.
The efficiency of diffuser systems should not be a strong function of
stream velocity unless large eddies entrain the air bubbles so as to
increase the air-water contact time. If air bubbles require longer
time to surface because their mean ascent velocity is lowered by eddies,
then higher absorption efficiencies would be expected. However, if
flow conditions significantly affect rise time, the stream would, in all
probability, have adequate natural aeration to correct problem conditions
For the types of diffuser systems used in the above river studies, the
maximum absorption efficiency was 7.0% (Table 5.7). Test results
reported by Imhoff [58] , using finely perforated (.02 in-to .028 in. -
diameter holes) tubing in the Ruhr River (West Germany), indicated
the following absorption efficiencies:
8% @ 8-foot depth
17% @ 16-foot depth
15% @ 20-foot depth
Even higher absorption efficiencies have been obtained from laboratory
tests using 1/2-in, polyethylene tubing having die-formed slits 1-1/2
in- on center. Tests were conducted for low air flows and non-flow
conditions at 23°C. Results have shown that, for 1-cfm air flow with
100 ft of tubing, absorption efficiencies in oxygen-depleted water have
ranged from 14 to 44% at 3- and 10-ft depths, respectively [ 59] .
In view of the relatively high efficiencies obtained by Imhoff for stream
conditions and the even higher values obtained with the 3/2-in.
polyethylene tubing for non-flow situations, it is possible that a poly -
ethelene tubing system might be adaptable to stream applications.
Further study would be required to determine whether or not the system
would have similarly high absorption efficiencies for flow conditions
as it does for non-flow conditions . Specifically, further investigation
53
-------
of its performance as a function of air flow, water velocity, and depth
is needed. In addition, it would be of interest to use this type of
system with molecular oxygen. The following discussion considers
previous tests of diffuser systems with gaseous oxygen.
Diffuser System Tests Using Molecular Oxygen
The results of a series of experiments using pure oxygen in a shallow
diffuser system have been reported by McKeown [60] . Fifteen ceramic
diffusers were evenly spaced along a 4-in* diameter header, 100 ft
long, which was submerged 3 ft below the stream surface. Dissolved
oxygen levels of incoming water varied from 0-1 pprn, and stream
depth in the vicinity of the tests varied from 3 to 5 ft. It was found
that at a feed rate of 300 scfrn absorption was only 1. 1%, while at
10 scfm absorption efficiency was increased to 12%.
In a somewhat limited study conducted by Amberg et al. [6l] on the
Pearl River near Bogalusa, Louisiana, molecular oxygen was diffused
through a multiport feeder placed on the bottom. In this test the
oxygen was produced in a pilot plant on the site.
The multiport diffuser consisted of two 4-in. pipes, 36 in. on center,
and equipped with Walker Process Company's "Sparjets." The
Sparjets on one pipe had 1 /32- in. holes and those on the other had
3/64-in. holes.
During the tests, water flow rate over the diffuser section was
approximately 2440 cfs, with an average depth of 19 ft, and a tempera-
ture of 22 C. The incoming water was moving at an average of 2. 14
ft/sec and had a DO of 7.7 ppm. A summation of five trials is presented
in Table 5.7. Oxygen efficiencies varied from 14. 6% to 2 1. 5%. It was
found that a single header with 1/32-in. orifices was as effective as
the multiple header with two different-size orifices. The observation
was made that, even with the small orifice size and low aeration rates,
a great number of oxygen bubbles were breaking at the water surface,
indicating waste of oxygen.
The efficiency might have been increased by using carborundum or
Saran-wrapped diffusers, commonly used in sewage treatment.
However, such systems may become clogged when immersed in a
river and not used on a continuous basis.
The Effect of Flow on Dispersion of Oxygen from Diffusers
A study of the effects of flow on dispersion has been conducted by
Whipple et al.[ 44 ] , in which Rhodamine-B dye was injected through
a diffuser system into the Delaware River. By measuring the dispersion
of the dye as it travelled downstream and using curve-fitting techniques,
the longitudinal dispersion coefficient was obtained. According to
54
-------
TABJLE 5.7
Reaeration Data for the Pearl River, Louisiana
Using a Double Aeration Header with 1/32-in.
and 3/6]-in. Orifices (Amberg[6]] )
Oxygen
scfm
100
200
300
400
feed rate
lo/day
11, 900
23, 800
35,800
47,700
Oxygen
increas e
in stream
(Ib/day)
2 550
4250
5230
7850
Oxyg en
absorption
efficiency
(%)
21.4
17.8
14.6
16.5
Single Header (1/32-in. Orifices)
100
11, 900
2620
22.0
one-dimensional-dispersion theory, the coefficient was found to be
8.5 x 10 ft^ / min or 1.41 x 10^ ft2/sec. Data on the surface aerator
was found to be incomplete, due to equipment failures. This value
is an order of magnitude larger than the largest "natural" coefficient
for a similar river. It was felt that the large value was primarily
due to turbulent mixing induced by the aeration system. Furthermore,
although this value is characteristic of a given installation, the
authors believed that it constitutes a decent first-order approximation
to other reaches in similar rivers being aerated. It was also concluded
that the DO disperses in a manner similar to the dispersion of the dye.
Such an assumption is reasonable, since changes in the stream charac-
teristics due to such a dye or oxygen are negligible.
A transverse-dispersion coefficient was not reported. However, the
transverse spreading effect was presented in an equation relating
the area affected as a function of distance downstream. The equation
was developed from data on the dye cloud geometry and is shown in
Figure 5.4.
It was noted that, from similar tests using surface aerators, lateral
dispersion was observed to be approximately the same as for the
diffuser s ystem.
Downflow Contactors
The downflow contactor category includes the more widely known U-Tube
System iirst reported in the Netherlands by Bruijn and Tuinzaad[62]
55
-------
11
- 10
o
o
o
-a
O
I
I
I
1000 2000
Distance Downstream from Diffuser (ft)
3000
Figure 5. 4. Graphical Results of Cross-Sectional
Area Distribution (Whipple [44])
56
-------
and more recently in this country by Speece[63] . Aeration by systems
in this category is accomplished by temporarily pressurizing an air-
water mixture as it is forced downward by a slight head over a vertical
tube. More oxygen is transferred near the bottom of the tube by virtue
of increased pressure and lower temperature, which increases the
DO deficit. This relationship is illustrated in Figure 5.5. In a U-Tube
the mixture is released from the tube near the surface, whereas in a
straight downflow contactor it would be released near the bottom.
Several different types of U-Tube systems can be designed[ 65] . In
Figure 5.6 a system is illustrated where oxygen is injected into the
inlet water by a blower. Since only low pressures are required at the
point of injection, centrifugal blowers may be used, which reduce cost
and maintenance requirements. Good control over the air-water ratio,
which ranges between 0 and 20% (by volume), is maintained by adjusting
the air-intake lines.
A second type of U-Tube system is shown in Figure 5.7. in this system
the head is provided by a natural cascade. The system shown has been
used in a fish hatchery where 400 gpm of water at 75% DO saturation
cascades 3 ft into the inlet of a 40-ft-deep U-Tube and emerges at
120% DO saturation [63].
In still another type of U-Tube, air is introduced by a venturi which is
vented to the air as shown in Figure 5. 8. The venturi system has a
somewhat larger head loss than the blower system described above.
The air-water ratio can be controlled by valving the air-intake port.
This system has the advantage of not requiring external power.
Where stratified impoundments are a problem, U-Tubes can effectively
reaerate the water by selective withdrawal of the hypolimnion water.
Such a system has been proposed for the Snake River and is shown in
Figure 5.9. This system is capable of raising the DO from 40% to 95%
saturation. The U-Tube would be a 40-ft-deep trench, 160 ft long and
10 ft wide, on each side of the center baffle.
The transfer of oxygen to water is more effective the deeper the U-Tube;
however, the saturation of dissolved nitrogen (DN) also increases. This
may be a critical factor since fish are adversely affected when nitrogen
super-saturates. For example, it is cited in reference 63 that a
tolerable dissolved nitrogen level for salmon is about 105% saturation
at the water surface. If U-Tubes are restricted to a water depth of
10 ft, the maximum possible super saturation is 106%, and acceptable
nitrogen levels would exist. The mixture of air and water could then
be passed through several ]0-ft stages which would increase the DO
concentration but not the DN.
Any time U-Tubes are considered for an aeration application, attention
must be given in the design to the avoidance of nitrogen supersaturation.
The prime design consideration, of course, is the DO concentration
57
-------
Ul
oo
DO
(mg/1)
Inlet
DO Saturation Value
DO Value in Tube
Outlet
Figure 5. 5. DO Deficit in 40-Ft-Deep U-Tube (Speece [65])
-------
Air Blower
Figure 5. 6. Air Blower Injection Modification of
U-Tube System ( Speece [65])
59
-------
Figure 5.7. Cascade Air Injection (Speece [65])
60
-------
Air
Figure 5.8. Venturi Air Injection (Speece [65])
61
-------
Figure 5.9. Schematic Installation of U-Tube Oxygenation of
Stratified Impoundment Releases (Speece [63])
62
-------
entering and leaving the tube. However, in achieving a desired DO
level the effluent should be checked for possible supersaturation of
nitrogen.
In general, high air-water ratios and deeper depths are required to
obtain DO saturation, as is shown in Table 5.8 [64]. According to
gas transfer theory, the associated air-water ratios that will saturate
the water with nitrogen are those given in Table 5. 9.
TABLE 5.8
Air-Water Ratio Required
for Oxygen Saturation (%)
(Speece [64] )
Inlet DO U-Tube Depth (ft)
(% saturation) 20 30 40 50 60
0
20
40
60
80
_ _
—
—
22
14
23
20
18
14
9
18
16
14
11
8
15
13
12
9
7
13
11
10
8
5
TABLE 5.9
Air-Water Ratio Required
for Nitrogen Saturation (%)
(Speece [64] )
Inlet DO U-Tube Depth (ft)
(% saturation) iQ 30 40 50 60
0
20
40
60
80
_ _
—
—
—
16
— — .
23
21
16
10
21
18
16
13
9
17
15
14
10
8
15
13
11
9
6
63
-------
If the required air-waler ratio to saturate the water with oxygen
exceeds the air-water ratio for nitrogen saturation, then nitrogen
s upersaturation will occur.
Another possibility for avoiding nitrogen s upersaturation is to use
pure oxygen in place of air. With such a system, initial DN would not
be altered, but the DO would increase. Another advantage of an
oxygen injection system is that the change in DO is approximately five
times that of air (Figure 5. 10).
Tests conducted bySpeece[65] using a variety of parameters for
4-in. diameter U-Tubes resulted in the following significant trends:
1. Increasing the air-water ratio increases the change in
DO at a diminishing rate.
2. Increasing the depth at which air is introduced results
in a reduction of change in DO for a given air-water
ratio.
3. Increasing the depth at which air is introduced reduces
the head loss due to air injection in the system for a
given air-water ratio.
4. There is an equilibrium air-injection depth at which the
air injection head loss is zero.
5. Higher water velocities reduce the head loss due to air
injeclion and decrease the change in DO through the
U-Tube. These changes are a consequence of the more
nearly equal bubble residence times in both legs of the
tube and the reduced time for oxygen transfer to occur.
6. The head loss due to air injection is proportional to
the U-Tube depth.
7. Minimum diffuser submergence gave maximum transfer
economy.
The authors extrapolated their findings to a 60 in.-diameter U-Tube
and found that transfer economies in excess of 3 Ibs O^/hp-hr could
be obtained (Figure 5.11). Furthermore, it was determined that
minimum air injection submergence resulted in maximum transfer
economy. Figure 5. 11 shows that associated with each U-Tube depth
and velocity there is an optimum outlet DO. The optimum outlet DO
was found to be dependent on the velocity, as indicated in Figure 5. 12.
Figure 5. 12 also indicates that the lower velocity of 3.5 ft/sec gives
rise to more efficient oxygen transfer. However, if one is interested
in transferring a given amount of oxygen into water, It may be more
economical to use smaller U-Tubes with higher velocities, in which
lower operating efficiencies are offset by lower capital expenditure.
64
-------
GO
o
Q
10 h
Diffused Gas - Oxygen
O Diffused Gas - Air
6 8 10 12 14 16 18 20
% A/W
Figure 5. 10. Increase in DO vs Per Cent A/W
for Diffusion of Oxygen and Air
65
-------
U-Tube Depth (ft)
40
30
20
10
Nominal Velocity = 5.5 ft/sec
Inlet DO = 0. 2 mg/1
ro
O
en
X!
I
I
4 6
Outlet DO (mg/1)
10
Figure 5.11. Effect of Depth on Oxygen Transfer
Economy in 60-in. U-Tube
66
-------
From the work of Speece and Adams [65] it has been concluded that
initial bubble size has no noticeable effect on the change in DO in a
U-Tube. The diffusers used in the tests consisted of a nylon cloth
and perforated units with 1/32-in. and ]/4-in. holes.
An exhaustive U-Tube testing program has recently been completed
by Rocketdyne, in which a 2-in.-diameter U-Tube system was used [66]
Variables investigated included depth, water velocity, air-water ratio,
and aspirator configuration:
Depth
Velocity
Air-Water
Aspirator
9 to 45 ft
1 .4 to 3. 4 ft/sec
0 to . 2 (volume ratio)
center-plug and venturi type
Quantitative analyses for rate of nitrogen transfer relative to that of
oxygen were conducted by a combination of vacuum degassing and mass
spectrometry. The results are interesting in that the ratios of DN to
that of DO were found to have an average value of 2. 4 . Table 5.10
contains the results from the samples investigated.
TABLE 5.10
Results of Chemical Analysis for
Dissolved Nitrogen (Rocketdyne [66 ]
Superficial water velocity (ft/sec)
Air/water at 1 atm (68°F)
DO concentration (mg/1):
Entrance
Exit
DN concentration (mg / 1):
Entrance
Exit
DN change /DO change
Run Number
293 375
1.9
5.8
2.2
7 .4
3.8
16.8
2.5
1.4
7.0
1. 1
5.6
2.6
12.9
2.3
This result is in contrast to the findings of Speece [ 65] which, for the
conditions investigated, indicated that the nitrogen gas transfer out of
the bubbles was insignificant.
68
-------
The discrepancy between Rockefcdyne's and Speece's investigations lies
in the fact that the test water used by the former was subjected to
vacuum degassing and, in so doing, nitrogen as well as oxygen was
stripped from the water, which resulted in a large nitrogen deficit and
consequently a large nitrogen transfer. In the case of Speece's work,
well water was used, which is characteristically high in DN. He-nce,
only a small amount can be transferred at a relatively low rate.
Usually, in a stream or river the DN level is expected to be near
saturation, since there are virtually no nitrogen sinks and the water
surface quickly establishes equilibrium with the atmosphere. Hence,
the nitrogen-oxygen transfer results in the Rocketdyne report can be
misleading when applied to a stream or river situation.
These tests also indicated that pressure losses in a venturi aspirator
were considerably lower than losses from center-plug aspirators
(Figure 5. 13), and also that the pinimum flow passage for the venturi
remains higher than for the center-plug aspirator, thus reducing the
chance for plugging.
A modification of the U-Tube into a straight doxyaiflow contactor has
been investigated recently by McKeown in the Androscoggin River in
Maine [ 52], The modification consisted essentially of removing the
return leg of the U-Tube, which then permitted the oxygenated water
and remaining air bubble to rise freely.
Extensive testing involved collecting DO and velocity profiles in the
vicinity of the contactor, which defined the zone of influence. The
site was a deep pond, formed by a power dam, with a span of about
1300 ft. Water depth was over 60 ft, and the general character
resembled a sluggish river.
The aeration system consisted of an axial flow pump with intake 4 ft
below the surface and whose turbulent discharge, containing large
quantities of entrained air, flowed through a header box into an 3 8-in.
corrugated pipe. Nominal liquid velocities in the pipe were on the
order of 6 ft/sec, and depths up to 40 ft were investigated. Since
there was no direct method of calculating how much air became en-
trained, the discharge from an auxiliary air blower was also fed into
the head box. Sparge rings or diffusers were not used to form small
bubbles, since sufficient turbulence existed in the header box, and
the associated sheer force was adequate to produce small bubbles.
The velocity profile associated with a 10-ft downdraft bubble contactor
is shown in Figure 5.14.
Mixing was observed to occur some 15 ft below the discharge, and
the representation of the rising bubbles (Figure 5. 14) is based on
visual observation. The vertical area zone of influence was estimated
to be approximately 800 ft2 (40 ft wide by 20 ft deep). The major
69
-------
-------
20'
10'
10'
15'
20'
25 !
30'
35!
40' u-
0.08
o;o?
/\
0. 16
0.^10 ^
\ X ' <'
0.05 \ ' 0. 15
\
0. 12
V
0. 05
0. 06
horizontal velocity component
given in knots (kn)
0. 06
0. 04
Figure 5. 14. Zone of Influence of the Downdraft Bubble
Contactor on the Surrounding Water
(McKeown [52])
71
-------
difference in the area of influence for different depth tubes was that
of depth.
The effects are shown in Table 5.11.
TABLE 5.11.
Vertical Area Affected by Bubble Contactor
Depth of Tube
(ft)
10
20
40
Vertical Area Affected
Width Depth
40 ft x 20 ft ( 800 ft2 )
45 ft x 35 ft (1575 ft2 )
50 ft x 60 ft (3000 ft2 )
In general, McKeown's findings can be summarized as follows:
1. The blower produced no increase in DO or economy.
2. The average efficiency for all the tubes is above 1 . 4 Ib
O? /hp-hr (converted to standard conditions).
Sidestream Pres s urization
Sidestream pressurization is a technique for oxygenating river water,
in which a small percentage of the flow volume is drawn off, mixed
with oxygen under pressure, and the resulting supersaturated mixture
diffused back into the river. There have been very few tests of this
technique, although the concept is promising, particularly for large
rivers where the oxygen requirement is great enough to justify the
cost of constructing on site a gaseous-oxygen generating plant. In
the preceding subsections, cost has not been considered in the dis-
cussions, primarily because site conditions have a significant affect
011 cost. For Sidestream mixing, however, the cost of oxygen is a
major factor and is discussed in'this subsection.
In one of the Sidestream pressurization studies conducted by Amberg
et al. [61], water entering the Pearl River was pumped through an
oxygen-diffusion system operating at a pressure of 68 psig. The
system was designed to pump 10, 000 gal/min under a 204-ft head
through a 150-ft-long, 14-in. -diameter pipe where oxygen was added
through spargers. In this process the water became supersaturated
with oxygen and was returned to the river through a diffuser header
placed across the bottom of the river. The diffuser header was
equipped with twelve 2-in. nozzles and fifteen 1-1/2-in. nozzles, and
was tapered from a 14-in.-to an 8-in. diameter.
72
-------
Initially, considerable trouble was experienced from clogging of
pumps and diffuser nozzles with debris carried in the river. However,
these difficulties were overcome with appropriate modifications
(screens and increasing nozzle sizes to 1. 5 in. ). Table 5. 12 illus-
trates the effects of adding 30, 000 Ibs of oxygen. It can be seen that
about 2 ppm or 16,400 Ibs of oxygen was added to the Pearl River at
the first station, where mixing was considered to be complete. The
oxygen absorption efficiency for the system when aerating 25 cfs of
the total stream flow (1. 64% of the total) was 54. 6%.
The oxygen was released in very small, discrete bubbles through the
return header. The small bubbles permit considerable oxygen
absorption in the water as they rise. The average head of water over
the diffuser was 9 ft.
The daily cost of adding 16, 400 Ibs of oxygen (54. 6% of 30, 000 Ibs)
based on $30 /ton (delivered) was:
$450 for oxygen
$ 59 for power ($. 005 /kw-hr)
This resulted in. adding 1.5 to 1.6 ppm oxygen over a 9 -mi. stretch.
The above costs do not include capital, maintenance, etc., and thus do
not provide a true picture. In addition, the system study was only
conducted for a limited set of conditions which were probably not
optimum.
The above sides tream oxygenation system was designed by the Linde
Division of Union Carbide Corporation and is referred to as the
"LINDOX" System.
A smaller system was tested in Brewton, Alabama. The system was
designed to inject 3000 Ibs of oxygen per day into the effluent of a paper
mill of the Container Corporation of America. Union Carbide claims
that injection efficiencies of from 55 to 75% were achieved [67] .
According to Union Carbide Corporation, preliminary capital cost
estimates for sidestream aeration can be made on the basis of $4000
per daily ton of oxygen injected. This price includes pump installation,
concrete pad, c ontrol panel, injection thimble, dispersion header , etc.
Oxygen costs will vary according to the installation and depend on such
factors as quantity used, transportation costs , etc. Typical prices ,
according to Union Carbide Corporation (Spring 1971), vary from
$35 to $50 per ton. They also suggest that power costs for pumping
can be estimated to be approximately 15% of the total oxygen cost.
In other investigations conducted by Linde, parameters have been
established for the following process variables:
73
-------
TABLE 5. 12
Dissolved Oxygen Added to the Pearl River
at an Oxygen Addition Rate of 30, 000 Ib /day
by Sidestream Oxygenation (Amberg et al. [6 l]
(Water temperature was 25. 5 C)
Station
Lakeview (above
mill)
1 . 5 miles (below
mill)
3. 0 miles (below
mill)
6. 0 miles (below
mill)
9. 0 miles (below
mill)
Flow
(cfs)
1528
1528
1528
1528
1528
Before oxygenation
DO DO
(ppm) (Ib/day)
7.5
5.0
4.2
3. 1
2.7
61, 600
41, 100
34, 600
25, 500
22,200
After oxygenation
Flow DO DO
(cfs) (ppm) (Ib/day)
1524
1524
1524
1524
1524
7.5
7.0
6.0
5.0
4.3
61,600
57, 500
49,250
41, 100
35, 300
DO increase
(ppm) (Ib/day)
...
2.0
1.8
1.9
1.6
• • •
16, 400
14,650
15, 600
13, 100
-------
1. Equilibrium oxygen concentration in water versus C>2
partial pressure.
2. DO at various pressures in the pumped stream versus
efficiency.
3. Bypass line pressure and input oxygen concentration versus
overall efficiency.
4. Velocity (Reynolds Number) in bypass versus overall
efficiency.
5. Contact time and oxygen input concentration versus DO
in pumped stream.
6. Fraction bypassed and Input oxygen concentration versus
overall efficiency.
7. Cost of pumping.
Use of Pure Oxygen
Pure (molecular) oxygen can be used to replace air in several aeration
systems, such as U-Tubes, diff user systems, venting of turbines, and
sidestream pressurization. The form of the oxygen can either be
gaseous or liquid, although oxygen in the gaseous form is more easily
injected. The following discussion deals with various applications and
the relevant findings .
The only liquid oxygen (LOX) test reviewed in this study was conducted
by Midwest Research Institute and reported in July, 1970 [68]. In that
study, the investigators dealt with injecting LOX into both static and
flowing water which was vacuum-stripped of oxygen. The water was
then stored under a nitrogen blanket.
The rationale behind using LOX was the following:
1. LOX is the most economical form for transportation of
oxygen.
2. LOX is more dense than water, hence it sinks and prevents
loss es.
3. Evaporation of LOX in water could be at rates that would
produce high-pressure bubbles.
4. The cooling is localized, thus increasing the driving force.
5. LOX introduction imparts turbulence.
75
-------
The program encountered considerable difficulty, and consequently a
definitive assessment of the aeration process was not able to be given.
However, based on the work carried out, certain trends were observed,
and a limited number of conclusions were drawn by the investigators.
Among those conclusions were the following:
1. Based on mass transfer, LOX is at least as attractive
as gaseous oxygen if the quantities are less than those
necessary to require construction of a production plant
on s ite.
2. If the oxygen consumption is such that on-site production
must be considered, then gaseous oxygen would be more
economical.
3. When the flow was turbulent, absorption appeared to be
more effective, and DO concentrations up to 30 pprn were
obtained.
4. Increasing the contact time of the LOX with water increased
the efficiency of absorption but not the mass-transfer
coefficient.
5. Varying the water temperature from 7 to 30°C had no
apparent effect on the absorption efficiency.
6. The initial DO content of the water did not affect the
mass-transfer coefficient significantly.
These results would tend to indicate that the injection of LOX into water
is a difficult problem, and a better alternative would be to transport
oxygen as LOX but deliver it to the water in gaseous form.
Since the cost of pure oxygen is based on rate of consumption and
distance from the source, the prediction was made that, when the
consumption rate exceeded 25 tons/day, a separation plant should be
erected at or near the site. If the demand exceeds 25 tons/day, the
oxygen should be piped as gaseous oxygen, and if the rates were smaller,
the oxygen should be transported as LOX. This conclusion is based on
pricing schedules from four major industrial gas suppliers for supplying
oxygen to ten different locations. Figure 5. 15 illustrates the results,
indicating that the four companies quoted similar prices. The prices
include delivery within a 100-mi. radius of the separation plant.
The minimum price of pure oxygen is directly a function of the cost
of power. To produce one ton of oxygen, 350 kwh are required. It
was pointed out that a lower bound would be about $7/ton of oxygen.
This is based on an industrial rate of slightly under $. 02/kwh. Con-
sumption rates of 1000 tons /day would be required to drop the cost to
$10/ton. Such a consumption rate is atypical, since this rate would
be indicative of the domestic waste requirements of metropolitan
New York City.
76
-------
10,000,000
1, 000,000
100,000
50, 000
1^
n!
-c
U]
,£>
— 10,000
c
o
•r-l
•«->
ex
c
o
o
X
o
1, 000
100
10
On-Site Plant
LOX Transport
0
t t t t t Econom:
r ~
Breakpoint
O Producer No. 1
D Producer No. 2
V Producer No. 3
A Producer No. 4
A Minimum Cost Based on Thermodynamics 1 KWH =
J L
J L
2 3 4567 89 10 11
Cost of Pure Oxygen (cents per pound)
Figure 5. 15. Cost of Pure Oxygen (Both [60])
77
-------
An interesting and valuable theoretical study has been conducted by
Speece [69] to predict the transfer of oxygen out of a bubble of oxygen
and the transfer of nitrogen from water into an oxygen bubble. The
results of this study are important, since a knowledge of conditions
which enable efficient oxygen absorption is necessary in order to
design an economically competitive oxygen injection system. The
significant findings of his study can be summarized as follows:
1. When small bubbles of approximately 2-mm diameter are
released from depths in excess of 100 ft, essentially all
of the oxygen is transferred.
2. The initial concentration of DO in the water has very
little effect on the absorption of pure oxygen.
3. The absorption of oxygen is greater for 2-mrn bubbles
than it is for 4-mm bubbles. At 40 ft there is approxi-
mately a 100% difference.
Findings 1 and 3 above are illustrated in Figure 5. 16.
In another test concerned with the feasibility of using pure oxygen,
Amberg et al. [70] added substantial quantities of oxygen to water
passing through a power turbine at Willamette Falls on the Willamette
River in Oregon. The turbine was vented with pure oxygen and also
with air. In this test the goal was to achieve a 5 ppm concentration of
DO to meet a state standard. It was found that aeration with air did
not offer a practical solution. Through the use of a sparge ring and
pure-oxygen absorption, efficiency on the order of 40% was obtainable.
The relative cost of air and oxygen for this test are shown in Table 5. 13,
where it can be seen that at high DO levels of the intake water the use
of pure oxygen is more economical.
In a study conducted by Pfeffer and McKinney [71] using oxygen-enriched
air to aerate industrial wastes, it was indicated that the rate of oxygen
transfer is considerably increased as the oxygen content of the gas
increases. Upon examination of Figure 5. 17 it can be seen that at a
given DO level the rate of oxygen transfer (slopes of the curves)
increases significantly with increase in oxygen content of the gas.
In oxygen-absorption tests conducted by Carver [72] it was found that,
when pure oxygen was used for aeration, the rate of oxygen transfer
was independent of the DO content in the liquid between 0 to 12 mg/1.
Hybrid or Mixed Systems
A "mixed" system would be two different types of systems in combination.
Such a system might be economically feasible when large changes in DO
are needed, when considerable shifting of the DO sag point occurs with
78
-------
100
G
01
o
o
in
-------
TABLE 5.13
Comparison of Reaeration Studies with Air and Oxygen
Location
Willamette Falls
Willamette Falls (a)
Willamette Falls (b)
Gas
Used
Air
Oxygen
Oxygen
DO to
Turbine
(ppm)
7 . 8-8.0
7.2
7.8
Oxygen
Absorption
Efficiency (%)
6.0
33.9
39.0
Oxygen
Exchange
(Ib/kw-hr)
0.53
13.9
9.5
Power Cost
per 1000 Ib
Oxygen
Dissolved
(dollars)
9.52
0. 36
0. 53
(a) Trial No. 1, Single -Opening Vent to Turbine
(b) Trial No. 2, Sparge Ring to Turbine
00
o
-------
o
Q
M
o
3
0"
T)
HI
X
44. 3% 02
10 15 20
Aeration Time (min)
25
Figure 5. 17. Effect of Percentage Oxygen in Aerating Gas on the
Rate of Oxygen Transfer (Pfeffer & McKinney [71])
-------
different seasons, or if unusual conditions are associated with a
stream. For example, if the water level in an impoundment where
a U-Tube is normally used drops significantly, surface aerators
might be used to temporarily continue oxygen delivery until water
depth again increases .
If a mixed system is to be considered, there are several natural com-
binations which can be ma'de. Reasonable combinations would be units
which work efficiently at low DO levels followed by units which operate
well in surface aerators or air diffusers acting as the primary system,
and the U-Tube, sidestream pressurization, or diffusers using pure
oxygen as a secondary fixed system.
When considering a mixed system, it should be kept in mind that a
system which operates well in the 3 to 5 ppm DO range will also work
well in the 1 to 5 ppm DO range. Hence, although in a particular
application it may be found that a mixed system is more economical,
in general, a non-hybrid system offers fewer complications and will,
inmost cases, be less expensive.
82
-------
SECTION VI
ENGINEERING METHODOLOGY
FOR RIVER AND STREAM AERATION
In the previous sections the present state of the art in river and stream
aeration was reviewed. The oxygen balance in a river or stream has
been shown to be relatively complex. In this section a methodology is
developed for the design of an aeration system. It will be shown that
from an engineering viewpoint much of the rigor required in the under-
standing of aeration and of oxygen balance can be simplified considerably
in the system design.
An outline of the steps required in the design process is shown in
Figure 6. 1. In the following paragraphs each one of the steps will be
discussed, with supporting calculations and examples given where
necessary. It should be noted that additional refinement can be added
to the procedure, but in this section it is only intended to present the
framework of the methodology.
Problem Recognition
The first step is a very obvious one, but one which may be difficult, in
that artificial aeration of rivers and streams is not a generally accepted
practice. This step requires the recognition that a problem in river
water quality exists and that the problem can be solved by artificial
aeration.
In some cases all known sources of industrial and municipal waste may
have been treated to the degree that 90 to 99% of the BOD is removed.
During some periods of the year, however, there may still be times
when fish die off or there is a noticeable decrease in some of the more
desirable forms of aquatic life. This is most likely to occur in the
summer months, when flow volume decreases and temperatures rise.
The condition may also develop during the winter, when ice cover on
a river prevents natural reaeration through the surface. Extensive
eutrophication is an indication that oxygen-depletion conditions may
develop. In these cases the assimilative capacity of the river or stream
may be increased by artificial aeration, and recognition of this fact-
constitutes the first step in the problem solution.
Preliminary Assessment
For a preliminary assessment of the problem, a profile along the
critical reach in the river should be obtained. If profiles are available
from past measurements and no significant changes in BOD loading or
flow volume have occurred, these profiles can be used. The profiles
should be for the worst case, i.e. , low flow and high temperature. If
83
-------
PROBLEM RECOGNITION
Dissolved oxygen levels in river or stream
are below acceptable levels.
PRELIMINARY ASSESSMENT
Dissolved oxygen profile is obtained along critical reach.
Profile should be obtained for worst conditions, i.e., low
flow during summer months or possibly under ice in winter.
Temperature, flow velocity, and depth are measured along
critical reach. From these measurements flow rates are
calculated, and reaeration coefficients are estimated from
those empirical formulations most suited to the particular
river or stream.
_L
OXYGEN REQUIREMENTS
Select firstpoint where DO decreases below the accepted mini-
mim level (5 ppm).
Determine step increase in DO necessary to bring DO above
minimum requirements.
Calculate daily loss of natural reaeration.
Plot new DO profile from step increase data, old profile, and
calculated loss of natural reaeration.
Examine new profile for points where DO may again decrease
below minimum.
Trade off costs of making larger initial step increase of DO at
one location versus smaller increases at several locations.
T
EQUIPMENT SELECTION
Convert transfer efficiencies of several
possible aeration systems from standard
conditions to conditions present in the
particular application at hand.
Determine size and number of units
necessary to supply required amount
of oxygen.
Determine cost per year for each
system.
Compare costs.
SITE CONSIDERATIONS
Consider site factors which
may eliminate any particular
aeration systems from
further consideration.
Figure 6. 1
Outline of Engineering
Methodology
FINAL SYSTEM DESIGN
Compare cost and site factors and select system.
Refine system design.
84
-------
no profiles are available, or if the ones that are available are
questionable, a survey of DO should be made over the critical reach.
In either case, measurements should definitely be made of the flow
velocity, temperature, and depth. Thes e measurements are required
for flow rate calculations and for estimates of reaeration coefficients
which are used in the determination, of oxygen requirements and in
compensation for loss in natural aeration.
The flow rate is simply the velocity times the cross-sectional area.
For the preliminary assessment the cross-sectional area can be
estimated from the width and average depth. Whenever significant
variations occur in c ross-s ectional area, a new flow rate should be
calculated for that section. The aeration coefficient can be measured
as described in Section IV, or it can be calculated using whichever of
the empirical methods best fits the river or stream. The methods of
Churchill [8] and O'Connor [9] appear to have wide application.
Determination of Oxygen Requirements
Once it has been determined that the DO levels in a stream or river
must be increased by artificial aeration and a preliminary assessment
of the problem has been made, the next task is to determine how much
oxygen must be added and where to add it. To answer these questions it
is first assumed that DO profiles of the river in question have been
obtained for the worst set of conditions. The profile used to determine
the critical reach of a river may not be a recently measured one but
may be one associated with a one-in-twenty or one-in-thirty year
low-flow situation.
In order to illustrate the methodology, we assume that the DO profile,
upon which an artificial system will be designed, is the one given in
Figure 6.2. For the purpose of illustrating a case of a "polishing"
action, this profile shows a low point of only about 3 ppm.
Assuming it is required that the DO level should not drop below 5 ppm
in the stream, it is then necessary to begin aeration artificially no
further downstream than the one-mile station. Assuming then that
artificial aeration would begin at this station, it is necessary to
calculate fche amount of oxygen that should be added. Since it has been
assumed that the minimum DO level would be 5 ppm, it might seem
logical to raise the DO level at this point by 2 or 3 ppm. However,
the higher the DO level is raised, the less efficient the transfer of
oxygen becomes, since the rate of transfer of oxygen is proportional
to the oxygen deficit, i. e. ,
= KLa(Cs - C) =(KLa)D (6.]}
85
-------
7
6
5
DO
(ppm)
3
2
1
0 123456789 10 11 12
Distance (miles)
Figure 6. 2. Example DO Profile
86
-------
where
C = the Oj concentration of the water (ppm)
C = the saturation level (ppm)
K, a = the overall transfer coefficient of an aerator (1/hr)
t = the time (hrs)
D = the oxygen deficit (ppm)
Equation 6. 1 shows that the rate of change in oxygen concentration at
any time is proportional to the magnitude of the deficit at that time.
Hence, at low initial DO levels the operating efficiencies of a system
are greater than for conditions where DO levels in the water are
initially high. Figure 6. 3 illustrates the work expenditure for step
oxygen increases of 1 ppm, in water where the saturation value is
9.2 ppm. The work required is proportional to
C
s
C - C
s m
where C = mean DO concentration at aerator
m
0< C < C
m s
This behavior is indicative of artificial aeration systems operating at
a nominal pressure of 1 atmosphere, viz, surface aerators and shallow
diffuser systems. When initial DO is on the order of 6 ppm or higher,
systems which transfer oxygen from air into water become relatively
inefficient. For instance, increasing the DO level from 8 to 9 ppm
requires approximately 13 times as much work as raising the DO from
0 to 1 ppm.
In order to illustrate the methodology for determining oxygen require-
ments and the spacing of aeration systems, the following case has been
selected:
87
-------
10. 0
9.2
7.
6.
DO 5.
3.
2.
1.
10
15
20
Relative Work Requirements
Figure 6.3. Relative Work Requirements for a Unit Increase in
DO in Water Initially Having Different DO Levels
88
-------
DO Profile - Figure 6.2
Flow Rate - 1000 cfs
Velocity - 0.5 mph
T - 24°C
K - 0.4/day
cL
The values listed above are assumed to be constant over the 12-mile
reach. This assumption is made for the purpose of illustrating the
methodology rather than for indicating exact stream conditions within
the section.
As discussed above, unless pure oxygen is being dissolved directly, the
work required to transfer oxygen from air increases significantly for
DO levels above 6 ppm. Since the objective is to prevent DO from
decreasing below 5 ppm, the first aeration devices must be placed in
the stream at mile 1. For a first cut, a step increase of 1 ppm is
made at this point, raising the DO to 6 ppm. The loss in natural
aeration due to the artificial addition must then be calculated.
As a consequence of increasing the DO level of the stream from 5 ppm
to 6 ppm, the natural aeration rate will decrease. The decrease can
be computed as follows;
~ = K (C - 5. 0)
dt .. av s '
5 ppm
= K (C - 6. 0)
dt , av s '
6 ppm
Subtracting the above two equations,
. dC dC dC _ . n T,.
AdT = dT, ~ dTV ~ 1-u a
6 ppm 5 ppm
Since the loss in natural aeration depends only on the increase of DO,
an upper bound on the loss of oxygen from natural aeration can be
established by assuming that the maximum loss occurred over the entire
length. Hence, the upper bound for the loss in natural aeration at the
12-mile station is
12 mi.x (I/. 5 mi. /hr) x (1/24 hr/day) x 1. 0 ppm x .4/day = . 4 ppm
89
-------
In Figure 6.4 the effect of the loss in natural aeration is shown at
mile 12, where it is seen that instead of a ].0 ppm step increase
over the old DO profile there is only a 0.6 ppm increase. Now it is
also noted that the new DO profile crosses the 5 ppm minimum level
at mile 4. At this point a second set of aerators must be added and
the loss in natural aeration again calculated. At mile 7 there is another
crossover, but at this point.it appears that a .5 ppm increase will be
sufficient. The new profile shows that this is true. Thus a total of
2.5 ppm of DO has baen added at three separate locations. The oxygen
requirement for this case is;
Ibs O2 = ]000 ft3 /sec x 62.4 lbs/ft3 x 3600 sec/hr
x 24 hr/day x 2. 5 x 10"
= 13, 350 Ibs O2/day
By making a larger initial increase in DO at mile 1, it may be possible
to eliminate the need for additional sites downstream. It is also known
that the work required to raise DO from 5 to 7 ppm is greater than twice
that required to raise it from 5 to 6.
To pursue this further, two more sets of calculations are made. In
Figure 6. 5 the initial DO is raised from 5 to 7 ppm, and it is shown
that a second set of aerators will be required at mile 5.7. The step
increase required at this point is greater than .5, thus, for convenience,
1 . 0 is selected as the value.
The total oxygen needed in this case will then be 16,020 Ibs/day, but
only two locations are required. It is not known at this point which
alternative is better, since the cost of constructing and maintaining three
sites must be weighed against the transfer efficiency of the aerators
employed in the two-site case illustrated in Figure 6. 5.
There is still another alternative, which is to raise the DO from 5 to
8 ppm at mile ] , as shown in Figure 6. 6. The only feasible way of
doing this would be to use pure oxygen, since conventional surface
aerators and diffusers become quite inefficient in this DO range. The
costs associated with this alternative are discussed in Section V and
also later in this section.
The above example illustrates the basic procedure for determining
oxygen requirements and the spacing of aeration systems. There are
several important factors which will affect the calculations. These
include:
3 . Value of aeration constant. A high value will lead to the
requirement for a high initial step increase or more
aerator sites downstream.
90
-------
DO 5
(ppm)
4
Mean Values of Parameters
Flow Rate = 1000 cfs
Velocity = . 5 mph
T = 24°C
Ka = 0. 4/day
Initial DO
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Distance (miles)
Figure 6.4. DO Profile Before and After Oxygen Addition at Three Locations
-------
vO
7
6
DO 5
(ppm)
4
3
2
1
Mean Values of Parameters
Flow Rate = 1000 cfs
Velocity = . 5 mph
T
K
= 24°C
= 0. 4/day
Initial DO
56789
Distance (miles)
10 11 12 13
Figure 6. 5. DO Profile Before and After Oxygen Addition at Two Locations
-------
sO
DO
(ppm)
A-lean Values of Parameters
Flow Rate = 1000 cfs
Velocity = .5 mph
T
K,
= 24°C
= 0. 4/day
Initial DO
0123456? 8 9 10 11 12 13
Distance (miles)
Figure 6.6. DO Profile Before and After Oxygen Addition at One Location
-------
2. Flow rate. Large flows require a high oxygen input to
increase the DO initially. Loss in natural aeration is
less significant for large flow rates.
3. Pure oxygen versus air. If pure oxygen is available on
the site, it is possible to make large initial increases in
DO without suffering a significant loss in transfer efficiency.
As can be seen from the above discussion, the transfer efficiencies of
the aeration devices play a significant role in system design. In the
following step in the design methodology, the characteristics of specific
aerators affecting the equipment selection are discussed. This is
followed by a discussion of site factors, which may eliminate certain
types of systems from consideration, depending on specific conditions
at the site.
Selection of Aeration Units
As discussed previously in Section V, the currently available aeration
systems are surface aerators, diffusers, downflow contactors (including
U-Tubes), and sidestream mixing devices using molecular oxygen.
There are also several other types which may be used with one of the
above systems to form a "hybrid" system. The hybrid approach has
not been explored in great detail but offers some promise with additional
development.
The selection of a particular aeration system depends on a number of
factors, including transfer efficiency, cost, maintainability, and site
suitability. Site factors are very important, and it is quite likely that
even if a system is suitable by virtue of its transfer efficiency and
cost, it might be aesthetically displeasing within a particular area.
Engineering Considerations Affecting the Selection of an Artificial
Aeration System
According to the laws governing the transfer of oxygen to water, it is
desirable to locate the aeration, system at the point of maximum oxygen
deficit. This is particularly true when oxygen is being transferred from
air, since the partial pressure of oxygen is then much less than it would
be if pure oxygen were being transferred.
Unfortunately, it is not possible to locate the aeration system at this
point if some minimum DO level must be maintained. If, for instance,
the lowest DO concentration is 4 ppm and a level of 5 ppm must be
maintained, the aeration system must be located where DO first drops
below 5 ppm. This type of consideration tends to favor a system using
pure oxygen, since the system using air cannot be located at its most
efficient operating point.
94
-------
In order to compare transfer efficiencies of the various aeration systems
for the application at hand, they must first be converted from standard
conditions to field conditions. For systems using air the field transfer
rate can be calculated using Equation 6.2.
TR[PC C ]9
TRf= - S7^ - ^ - (6.2)
C t
st
where
TR = field transfer rate (Ibs/hp-hr)
TR = transfer rate at standard conditions (Ib/hp-hr)
C = DO level seen by aerator (ppm)
C = saturation DO level at stream water temperature (ppm)
s
C = saturation concentration at standard conditions = 9. 17 ppm
P = pressure at which the system operations (mm of Hg)
rate of transfer of Oz in stream water
00 ~ rate of transfer of O^ in clean water
a _ saturation concentration of O;? in stream water
saturation concentration of O? in clean water
6 = temperature-dependent coefficient (average value is 1.024)
For a given river or stream cc > P > 6. and C would be fixed, and
S
TR =k. [k, ™ - C ] (6.3)
1 L 2 760 mJ v '
where k^ and k£ would be calculated constants. Thus, the transfer rate
for a system will depend on the DO in the water and on the operation
pressure.
The exact evaluation of C^ is difficult to obtain, consequently, the trans-
fer rate cannot be precisely determined. Susag et al. [ 50] have proposed
three methods for estimating Cm. The first method assumes Cm = Cu,
the upstream DO value; the second considers Cm = 1 /2 (Cu + C^), i. e. ,
the average of upstream and downstream DO levels; and the third uses
a logarithmic average based on the aeration equation. Based on
95
-------
laboratory tests using surface aerators, the proposers of these methods
found the third technique to give the best results. However, this area
warrants further investigation. The application of these methods for
diffusers, U-Tubes, and sidestream pressurization systems should be
verified.
For a surface aeration unit at or near sea level, the pressure is
essentially 760 mm, and the transfer rate will depend inversely on Cm
(the upper limit of Cm = (P Cs) = k£). Hence, if it is desirable to raise
the DO level C_ of the stream to a high value, the overall efficiency
would drop off rapidly. However, if a pressurized system or a pure-
oxygen system were used, high transfer efficiencies could be achieved.
For example, if an air pressurized system were used at 70 psig, transfer
efficiencies would increase significantly. If such a system used molecular
oxygen in lieu of air, the transfer efficiency should increase by a factor
of five compared to that of air, and perhaps on the order of twenty com-
pared to the surface aerator system.
Each of the aeration systems has particular advantages and disadvantages
relating to the river and stream application. A summary of these con-
siderations is presented in Table 6. 1
Economic Considerations Associated with an Aeration System
One of the most significant factors associated with an aeration system is
transfer efficiency (Ibs O2/hp-hr). It has been shown in this section how
to calculate oxygen requirements for a particular application. From
these calculations it was considered that there are several possible
alternatives and that the final selection will depend on transfer efficiencies
of the available devices and on site factors at the point where oxygen
should be added.
The cost of adding oxygen at one point, such as the case shown in Figure
6.6, will depend on the transfer efficiency of a device attempting to
increase DO from 5 to 8 ppm. It will be shown in a design example that
this can be achieved using pure oxygen but that the cost of using surface
aerators would be prohibitive at this location.
Other alternatives are shown in Figures 6.4 and 6.5. In these cases,
the cost of constructing additional systems at more than one location must
be considered. These costs will depend on specific conditions at the site,
such as:
3. Type of soil for supporting structures
2. Availability of power
3. Accessibility for maintenance
96
-------
TABLE 6. 1
Characteristic Features of Aeration Systems
Type of
Aerator
Features
Disadvantage s
Mechan-
ical
Surface
Aerator
Wide choice of commercially
available units.
High-speed, lightweight, electrical
units can provide portability.
Most units do not require direct
vertical support, i. e. , they float.
However, provision must be made
for sudden increases in stream
depth due to sudden storms.
Ideal operating DO level is under
6 ppm.
Range of instream transfer rates:
1 — 4 Ibs O2/hp-hr
Range of cost: 10-hpunit, $ 3,400
75-hp unit, $18, 500
Total cost increases as the number
of sites increases, because of
electrical service and connections.
Mooring cables may re-
strict boating.
Northern climates may
cause freezing problems.
Their presence may pre-
sent aesthetic and/or
noise problems.
Being exposed, they are
vulnerable to vandalism.
Internal clogging from
dust particles in the air.
Except for plastic tubing,
installations are gener-
ally fixed.
Cannot be used in a
channel maintained by
dredging.
Long distance between
compressor and diffuse?
results in large head
los ses.
Diffuser
Can be used with air or molecular
oxygen.
Does not provide any surface
obstructions.
Not vulnerable to vandalism.
Operates with a minimum of noise
and aesthetic upset.
Diffuser heads are often porous
ceramic, however, perforated
plastic tubing is also available.
Range of absorption efficiencies
(with air) :
>orous ceramic: 3 — 10%
perforated tube : as high as 40%
(when flow rate is low)
Range of instream transfer rates
(porous ceramic only):
. 7 — 1. 4 Ibs O2/hp-hr
Ideal operating range is usually
under 6 ppm when used with air.
Ideal operating depth: air, 10-15 ft
pure oxygen, stream depth
97
-------
TABLE 6. 1
(continued)
Type of
Aerator
Features
Disadvantages
Side-
stream
Mixer
Works efficiently with high initial
DO in water.
Does not interfere with boating.
Oxygen-absorption efficiencies are
normally over 50%.
Is not seriously affected by cold
weather climates.
Requires a supply of pure oxygen.
Requires sophisticated
equipment and continual
maintenance.
Relatively high initial cost
Small installations are
expensive to operate be-
cause of high cost of
oxygen.
Not portable.
Limited number of com-
mercial units available.
Down-
flow
Con-
tactor
(U-Tube)
If natural head is available and
sufficient stream depth, this sys-
tem offers maintenance-free
operation.
Can be located to the side of a main
channel; does not restrict boating.
Can be used with air or molecular
oxygen.
Ideally suited for high DO levels
(near saturation).
Not portable.
High initial cost compared
to other systems.
Requires depth of at least
10 ft, 30 to 40 ft for high
efficiency.
Possible problems with
nitrogen saturation when
used at depths in excess
of 10 ft.
Usually requires a cus-
tom installation.
98
-------
After the sizes and number of units required to meet a given oxygen
requirement have been determined, the capital, construction, and
engineering costs can be computed. This cost is converted to an annual
amortization rate to which are added operating and maintenance costs.
Where there are several possible types of systems, a cost analysis
should be performed for each one. The final selection may be based
on factors other than cost, but it is important to have the cost infor-
mation available.
Design and Cost Examples
The following two examples are offered as illustrations of the procedure
for estimating costs and also to show some of the differences in approach
using surface aerators and sidestream mixing with pure oxygen.
Figures 6.4 and 6. 6 are graphic illustrations of the two examples. In
Figure 6.4 oxygen is injected at three different locations, and in
Figure 6. 6 the same DO profile is treated at only one location. It is
fairly obvious from the previous discussions in this section that pure
oxygen can be injected efficiently at a point where the DO level in the
water is fairly high. The transfer efficiency of a surface aerator,
however, is best at a low DO concentration.
For the two cases, surface aerators will be used at each of the three
locations in Figure 6.4 while a sidestream mixing system will be used
at the single location shown in Figure 6.6.
The following design conditions apply to both examples :
Velocity . 5 mph
Discharge rate 1000 cfs
Length of critical reach 12 mi.
Temperature 2 4°C
Pressure 1 atmosphere
oc = 6 .95
Aeration constant, K .4/day
3,
In the first example, surface aerators are used at two locations to
raise DO from 5 to 6 ppm and at a third to raise DO from 5 to 5. 5 ppm.
The oxygen requirements at each location are
= 1000 ft3 /sec x 624 lbs/ft3 x 3600 sec/hr x 24 hr/day
l-2 A
x 1 . 0 x 10 = 5, 380 Ibs O2/day = 222 Ibs
Ibs O2/day
Ibs O2/day = 1000 ft3 /sec x 624 lbs/ft3 x 3600 sec/hr x 24 hr/day
x 0.5 x 10"6 = 2,695 Ibs O2/day = 112 Ibs O2/hr
99
-------
To determine how much horsepower is required at each location, the
standard transfer efficiency for a surface aerator is converted to a
field transfer rate using Equation 6.Z.
- C
TR = — m
C f
st
For this example, surface aerators with a transfer efficiency of 2 . 2 Ibs
O2/hp-hr are selected. There will be two field transfer rates, one for
the case where DO is increased from 5 to 6 ppm and the other when DO
is raised from 5 to 5.5 ppm.
Tofoc or 760 5+6i i A ~O - in nc
2.2 .95 x 8.5 x =-7-77 T— I x 10 x *. 10 x . 95
L 760 ^
-j 2 = 9. 17 x 10-6
= . 64 Ibs O2/hp-hr
TR
oornc oc 760 5 + 5 . 5 i ,n-6 i i o nc
2. Z [ . 95 x 8.5 x yr-pr - - s - j x 10 x 1. 10 x . 95
3 9. 17 x ]0
= .71 Ibs O2/hp-hr
The hp requirement at locations 1 and 2 will be
, 222
KPl,2 =
At location 3
hp3 =777 = 158
Cost estimates for 75-hp electrical surface aerators have been reported
in a recent study[6] . The estimates include all costs necessary to
install and operate the systems (cables, piling, lights, specially
reinforced frames, etc.) and are given for single-mounted units and
100
-------
aerators mounted in clusters of two and three. (Prices are based on
335 clays of annual use.)
Arrangement Single Double Triple
Total Horsepower 75 150 2Z5
Equipment and Construction 57,000 92,700 130,500
Engineering and Contingencies 11. 400 18,500 26, 100
$68, 400 $]11,200 $156,600
Operation and Maintenance 20,000 26,500 33,000
Amortization & Interest 7. OOP 1 1, 500 ]6. 200
(15 years @ 6%) $27, 000 $38, 000 $49, 200
Assuming that the above costs are representative of the costs associated
with the given example, the yearly expense of a system for the three
locations can be estimated as follows;
At the two sites requiring 347 hp, one triple cluster and one
double cluster could be used at each location, while at the
third site one double cluster could be installed. This results
in a slight excess at the first and second locations and a slight
deficit at the third location. The total provision is on the
conservative side.
The total annual costs would then be
2 x $49,200 + 3 x $38,000 = $212,400
In the second example, illustrated by Figure 6.6, the design conditions
given for the previous example also apply. The oxygen in this case will
be injected by sidestream pressurixation at one location, raising DO
from 5 to 8 ppm .
Based on correspondence with the manufacturer [ 67] , it is estimated
that installation costs will be approximately $4000 per daily ton of
oxygen injected. This includes pump installation, concrete pad for
oxygen supply system, fabrication of oxygen control panel, injection
thimble, and dispersion header.
The operating costs for supplying oxygen at the site would be about
$35 to $50 per ton. The power required to pump water at 100 psig is
about 50 kw per 1000 gpm. (The oxygen and water are mixed at
approximately 1 00 psig.) The injection efficiencies for the system
101
-------
should be on the order of 55-75% when 1 to 3% of the total flow is
pumped to the sidestream unit.
The oxygen requirement to raise DO by 3 ppm at the one location is;
lbsO2/day= 1000 ft 3-/sec x 62. 4 Ibs /ft3 x 3600 s ec /hr x 24hr/day
x 3 x 10~6
- 16,020
Assuming an injection efficiency of 65%, the injection system should
deliver
16, 020 Ibs !•> 7 <. ^ /j
—TT—-> AAA iu—77 = 12.3 tons O,/day
. 65 x 2000 Ibs /ton 2 *
The installation costs would then be
J2. 3 tons x $4000/ton = $49,200
This cost can be amortized over 15 years at a 6% interest rate, yielding
an annual cost of $4,700/year.
The operating costs are calculated as follows:
Assuming the cost of O;? is $45/ton and that 3% of the river flow is
pumped to the mixer:
O? costs (135 days ) = 12. 3 tons x $45/ton x 135 day/yr
= $74,800/yr
The cost of pumping water to the mixer over a 135-day period is:
Power required = 1 000 ft3 /sec x 7. 48 gal/ft3 x 60 min/sec x . 03
x 50 kw/1000 gal/rnin = 675 kw
Power cost = 675 kw x 24 hr/day x 135 day/yr x $.015 /kw-hr
= $32,800
Maintenance costs are estimated as follows:
Personnel - 2 men @ $8, 000/yr = $16, 000
Equipment 5, OOP
$21,000
102
-------
The total annual cost is: $ 4,700
74, 800'
32,800
21,OOP
$ 133, 300
Summary
The cost of using surface aerators in this application is more than
1-1/2 times as high as that of using the sidestream system. The basic
reason for this is the low transfer efficiency of the surface aerator
when working with an incoming DO level of 5 ppm. The oxygen-injection
system is not affected by the relatively high DO and thus performs the
job at a much lower cost.
These two examples serve only to illustrate the methodology. For each
case, the cost calculations should be made, but site factors will also
influence the final result. These factors are considered next.
Site Factors
Depending on the uses of a stream or river, the application of certain
aeration systems may be precluded or severely limited. For example,
if the waterway is navigable and is used for shipping and pleasure
boating, the placement of obstructions such as surface mechanical
aerators or U-Tubes in the river would present an obvious problem. In
this case, the tendency would be to select a d iff user or sidestream
pressurization system. This situation does not completely eliminate
the possible use of surface aerators or U-Tubes, since they can be
installed outside of a main channel; although, this would not permit
optimum utilization of the equipment. On the other hand, if a river
channel is dredged periodically, the positioning of diffuser units or
outflow lines from a sidestream pressurization system would also be
constrained.
Aesthetic considerations and noise levels may also carry considerable
weight when selecting a system. If the installation is to be near a
town or city, opposition may arise if the units produce considerable
foaming and frothing or a continuous whine or roar.
In some cases there may be shifting of the low points in the DO profile,
and consideration must be given to a system which affords portability.
Such a situation would favor the selection of high-speed surface aerators,
which are relatively small and light compared to low-speed units.
U-Tubes, sidestream pressurization, and diffuser systems are generally
not portable or easily moved.
If artificial aeration is necessary during winter months, freezing may
be a problem with surface aerators. When the surface of a river
103
-------
freezes, the aerator may be tilted or lifted and consequently may
become ineffective. In this case, a diffuser or sidestream pressuri-
zation system would be favored, since they would not be affected by
conditions on the surface.
Another factor that must be considered from a practical viewpoint
is the vulnerability of the system to vandalism. For example, with
a diffuser system, accessibility to any of the components can be
minimized; the diffusers are under water, the feed lines and mani-
folds are submerged or buried, and the compressors can be located
in a blockhouse. If sufficient natural head and water depth exist in a
stream, a U-Tube system is also relatively impervious to vandals.
The system in this case would essentially be a submerged concrete
structure whose cross-sectional shape corresponded to that of a
U-Tube.
Final Selection of an Aeration System
As discussed above, site considerations play a major role in the final
selection of an aeration system. In the design examples, several
possible solutions were shown for a selected DO profile, each having
a different cost. The final choice must involve a trade-off between
the costs of a system, the cost of construction and maintenance, and
the aesthetic considerations at the possible sites.
For each of the possible technical solutions, the corresponding site
problems should be listed. The site problems can be characterized as
those affecting cost and those affecting aesthetic qualities. From each
combination there will emerge a top candidate system. If several
combinations appear to be satisfactory, the final selection will
obviously be the one with the lowest estimated cost. In the examples
discussed previously, the use of pure oxygen in a sidestream mixing
system emerged as a top choice on a cost basis and would probably
also be a better choice aesthetically.
However, further examination might have indicated that diffuser
systems using plastic pipe were both less expensive and more acceptable
aesthetically. Unless there are very obvious reasons not to do so,
calculations and a trade-off should be performed for each possible
alternative.
In this report it has been, assumed that the minimum acceptable DO
level is 5 ppm. The results of the design examples may differ con-
siderably if this standard is lowered, for example, to 3 or 4 ppm.
The methodology, however, would not be affected.
104
-------
SECTION VII
ACKNOWLEDGEMENTS
The support and assistance of the Project Officer,
Dr. Curtis C. Harlin, Jr., of the Robert S. Kerr
Water Research Center, EPA, is acknowledged
with sincere thanks. In addition, the assistance of
Mr. Lowell Leach, of the Robert S. Kerr Water
Research Center.and of Mr. James Basilico and
Mr. Charles Myers, of the EPA, has been appreciated.
105
-------
SECTION VIII
REFERENCES
[1] Tyler, R. G. , "Accelerated Reaeration, " Sew. Works J. , _14,
4, 834, July, 1942.
[2] ZoBell, C. E. , "The Effect of Oxygen Tension on the Rate of
Oxidation of Organic Matter in the Sea, " J. Mar. Res. , 3_,
211-223, 1940. ~
[3] Dresnack, R. , and Metzger, I., "Oxygen Response and Aeration
in Streams, " Proc. 23rd Industrial Waste Conference, Purdue
Univ., Lafayette, Ind. , May, 1968.
[4] Dresnack, R., Transient Dissolved Oxygen Profiles in Streams,
Doctoral Thesis, New York Univ. , 19667
[5] American Society of Civil Engineers, SED, Twenty-Ninth Progress
Report of the Committee on Sanitary Engineering Research, 86,
SA 4, 41, I960.
[6] Whipple, W. , et al. , Instream Aeration of Polluted Rivers, Water
Resources Research Institute, Rutgers Univ., August, 1969.
[7] Littleton Research and Engineering Co. , An Engineering Economic
Study for the Development of an Optimum Mechanical Aeration Sys-
tem for Quiet Rivers and Ponds, Report #G-182, July, 1970.
[8] Churchill, M. A. , et al. , "The Prediction of Stream Reaeration
Rates, " J. San. Eng. Div. , Proc. American Society of Civil Eng. ,
July, 1962.
[9] O'Connor, D. J. , and Dobbins, W. E. , American Society of Civil
Eng. , 123, 641, 1958.
[10] Gameson, A. L. H. , and Truesdale, G. A. , J. Institute of Water
Eng. , _1_3, 175-87, 1959.
[11] Langbein, W. B. , and Durrem, W. H. , Geological Survey Circular
542, U. S. Dept. of the Interior, Washington, D. C.
[12] Ownes, M. , Edwards, R. W. , and Gibbs, J. W. , International J.
of Air and Water Pollution, _8, 469, 1964.
[13] Odum, H. T. , "Primary Production in Flowing Waters, " Limnol.
Oceanog. , !_, 102-117, 1956.
[14] Buswell, A.M., VanMeter, I. , and Gerke, J. R., "Study of Nitri-
fication Phase of the BPD Test, " Sewage and Industrial Wastes,
22, 4, 508, February, 1950.
107
-------
[15] Gaffney, P. E. , and Heukelekian, H. , "Oxygen Demand Measure-
ment Errors in Pure Organic Compounds -Nitrification Studies, "
Sewage and Industrial Wastes, 30, 4, 503, April, 1958.
[16] Sawyer, C. N. , and Bradney, L. , "Modernization of the BOD
Test for Determining the Efficacy of Sewage Treatment Processes, "
Sew. Works J. , 1_8_, 6, 1113, November, 1946.
[17] Ruchhoft, C. C. , Placak, O. R., and Ettinger, M. B. , "Correction
of BOD Velocity Constants for Nitrification, " Sew. Works J. , 20,
5, 832, September, 1948.
[18] Courchaine, R. J. , "Significance of Nitrification in Stream
Analysis-Effects on the Oxygen Balance, " 30th Annual Conference,
Michigan Water Pollution Control Assn. , May, 1962.
[19] American Public Health Assn. , Standard Methods for the Examina-
tion of Water, Sewage and Industrial Wastes, 13th Edition, New
York, 1971.
[20] Salle, A. J. , Fundamental Principles of Bacteriology, 3rd Edition,
McGraw-Hill Book Co., New York, 1948.
[21] Velz, C. J. , Applied Stream Sanitation, John Wiley & Sons, New
York, 1970.
[22] O'Connor, D. J. , "The Temporal and Spatial Distribution of Dis-
solved Oxygen in Streams, " Water Resources Research, _3> 1 • 65,
1st Quarter, 1967.
[23] Gameson, A. L. H. , "Some Aspects of the Carbon, Nitrogen, and
Sulphur Cycles in the Thames Estuary, Part II, " in Effects of
Pollution on Living Material, pp 47-54, Institute of Biology, London,
1959.
[24] O'Connor, D. J. , "Stream and Estuarine Analysis," Manhattan
College Summer Institute in Water Pollution Control, New York, N. Y.
[25] Buswell, A. M. , Shiota, T. , Lawrence, N. , Van Meter, I. , Applied
Microbiology, 2^, 21-5, 1954.
[26] Garrett, M. T. , Proc. Industrial Water and Waste Conference,
Rice Univ., Houston, Texas, 1961.
[27] Knowles, G. , Downing, A. L. , Barrett, M. J. , J. General Micro-
biology, 3J3, 263-78, 1965.
[28] Stratton, F. E. , and McCarty, P. O. , "Prediction of Nitrification
Effects on the Dissolved Oxygen Balance of Streams," Environmental
Science and Technology, J_, 5, May, 1967.
108
-------
[29] Winberg, G. G. , and Sivko, T. N. , "The Significance of Photo-
synthetic Aeration in the Oxygen Balance of Polluted Waters, "
International J. of Air/Water Pollution, 6, 267-75, 1962.
[30] Innjatovic, L. R. , "Effect of Photosynthesis on Oxygen Saturation, "
J. Water Pollution Control Fed. , May, 1968.
[31] Stay, F. S. , et al. , The Components of Oxygenation in Flowing
Streams, U. S. Dept. of the Interior, FWPCA, April, 1967.
[32] Hull, C. H. J. , "Photosynthesis as a Factor in the Oxygen Balance
of Reservoirs, " Symposium on Streamflow Regulation for Quality
Control. , PHS Pub. No. 999-WP-30, June, 1965.
[33] Hull, C. H. J. , "Discussion of 'Effects of Impoundments on Oxygen
Resources' by Churchill, " Oxygen Relationships in Streams, W58-2,
pp 124-29, 1958.
[34] Copeland, B. J. , and Duffer, W. R. , "Use of a Clear Plastic Dome
to Measure Gaseous Diffusion Rates in Natural Waters, " Lira no 1.
Oceanog. . _9_, 4, 494-499, October, 1964.
[35] Hornuff, L. E. , A Survey of Four Oklahoma Streams with Reference
to Production, Oklahoma Fishery Research Laboratory Reference to
Production Report 62, June, 1957.
[36] Kn'opp, H. , "Investigation of the Oxygen Production Potential of
River Plankton, " Hydrology, 22, 152-66, I960.
[37] Paulson, R. W. , "The Longitudinal Diffusion Coefficient in the
Delaware River Estuary as Determined from a Steady-State Model, "
Water Resources Research, _5, 1, 59, 1969.
[38] Harleman, D. R. F. , and Ippen, A. T., "The Turbulent Diffusion
and Convection of Saline Water in an Idealized Estuary, International
Assn. Sci. Hydro. Publ. , _5_1, I960.
[39] Orlob, G. T., "Eddy Diffusion in Homogeneous Turbulence, Proc.
American Society Civil Eng. , 85, No. HY 9, September, 1959.
[40] O'Connor, D. J. , "Oxygen Balance of an Estuary, " Proc. American
Society Civil Eng. , _86, SA 3, May, I960.
[41] Fischer, H., "The Mechanics of Dispersion in Natural Streams, "
j. Hydraulics Div. , Proc. American Society Civil Eng., p 187, 1967.
[42] Fischer, H. , "Dispersion Predictions in Natural Streams," J. San-
itary Eng. Div. , Proc. American Society Civil Eng. , p 927, 1968.
[43] Elder, J. W. , "The Dispersion of Marked Fluid in Turbulent Shear
Flow," J. Fluid Mechanics, 5, 544-60, 1959.
109
-------
[44] Whipple, W. , etal. , Oxygen Regeneration of Polluted Rivers:
The Delaware River, Environmental Protection Agency Program
No. 16080 DUP, December, 1970.
[45] Dobbins, W. E. , "BOD and Oxygen Relationships in Streams, "
J. Sanitary Eng. Div. , Proc. American Society Civil Eng. ,
pp 53-78, June, 1964.
[46] Aeration in Waste-water Treatment, Water Pollution Control
Federation Manual of Practice No. 5, 1971.
[47] McKeown, J. J. , and Buckley, D. B. , "Mixing Characteristics of
Aerated Stabilization Basins," TAPPI 8th Water and Air Conference,
1971.
[48] Burns, O. B. , etal. , "Pilot Mechanical Aeration Studies of the
Jackson River in Covington, Virginia, " Proc. 21st Industrial Waste
Conference, Purdue "Univ. , Lafayette, Ind. , 799, 1966.
[49] Kaplovsky, A. J. , etal. , 'Artificial Aeration of Canals in Chicago, "
36th Annual Meeting of the Water Pollution Control Federation,
Seattle, Washington, October, 1963.
[50] Susag, R. H. , Polta, R.C., and Schroepfer, G. J. , "Mechanical
Surface Aeration of Receiving Waters, " J. Water Pollution Control
Fed. , 38, 1, January, 1966.
[51] Brookhart, N. M. , Mechanical Aeration Project Performance and
Feasibility Study, Miami Conservancy District Report, November,
1969.
[52] National Council of the Paper Industry for Air and Stream Improve-
ment, Results of a Cooperative Field Study of a Downflow Bubble
Contactor and a Conventional Surface Aerator, Techn. Bull. #237,
June, 1970.
[53] JLueck, B. F. , etal., Evaluation of the Spray Type "Aqua-Lator"
for River Aeration, State of Wisconsin Commission on Water Pol-
lution, Bulletin No. WP-109, March, 1964.
[54] Tyler, R. G. , "Polluted Streams Cleared up by Aeration, " Civil
Eng. , 16, 348, August, 1946.
[55] Wiley, A. J. , et al. , "River Reaeration, " Paper Trade J., 124,
12, 123, 1947.
[56] Palladino, A. J. , "Investigation of Methods of Stream Improve-
ment, " Industrial Water and Wastes, _6, 3, 87, 1961.
[57] Bohnke, B. , "Effect of Organic Wastewater and Cooling Water on
Self-Purification of Waters, " Proc. 22nd Industrial Waste Conference,
Purdue Univ. , Lafayette, Ind., 752, 1967.
110
-------
[58] Imhoff, K. R. , "Oxygen Management and Artificial Reaeration in
theArea of Baldeney Lake and the .Lower Ruhr River, " Das Gras
und Wasserfoch, 109, Germany, 1968.
[59] Sellner, E. P. , "Am-Aqua Aerated Lagoons, " presented at the
38th Annual Conference of the Arizona Water and Pollution Control
Assn. , April, 1966.
[60] National Council of the Paper Industry for Air and Stream Improve-
ment, Artificial Reaeration of Receiving Waters, Tech. Bull. #229,
New York, May, 1969.
[61] Amberg, H. R. , et al. , "Aeration of Streams with Air and Molecular
Oxygen, "
[62] Bruijn, J. , and Tuinzaad, H. , "The Relationship Between Depth of
U-Tubes and the Aeration Process, " American Water Works Assn.
_,L, _50, 879-885, July, 1958.
[63] Speece, R. E. , "U-Tube Stream Reaeration, " presented at the 7th
Annual Sanitary and Water Resources Engineering Conference, Van-
derbilt Univ. , May, 1968.
[64] Speece, R. E. , "U-Tube Oxygenation for Economical Saturation of
Fish Hatchery Water, " presented at the American Fisheries Society
Meetings, September 1968.
[65] Speece, R. E. , and Adams, J. L. , "U-Tube Oxygenation Operation
Characteristics, " Proc. of the 23rd Industrial Waste Conference,
Purdue Univ. , Lafayette, Ind. , May, 1968.
[66] Rocketdyne (Division of North American Rockwell Corp. ), The U-Tube
for Water Aeration, Final Report, Project R - 8043, Canoga Park,
California, March, 1970.
[67] Union Carbide Corporation, private communication dated March
1971.
[68] Both, T. D. , et al. , Oxygenation of Aqueous Bodies Using Liquid
Oxygen - Loxination, Midwest Research Institute, Kansas City,
Missouri, March, 1970.
[69] Speece, R. E. , "The Use of Pure Oxygen in River and Impoundment
Aeration, " Proc. of the 24th Industrial Waste Conference, Purdue
Univ., Lafayette, Ind., May, 1969.
[70] Amberg, H. R. , etal. , "Reaeration of Streams with Molecular
Oxygen, " Industrial Water Engineering, pp 15-20, February, 1967.
Ill
-------
[7l]Pfeffer, J. T. , and McKinney, R. E. , "Oxygen-Enriched Air
for Biological Waste Treatment, " Water and Sewage Works,
p 381, October, 1965.
[72] Carver, C. E. , "Absorption of Oxygen in Bubble Aeration, "
Biological Treatment of Sewage and Industrial Wastes, Volume I:
Aerobic Oxidation, Reinhold Publishing Corp., New York, 1956.
[73] Bohnke, B. , Possibilities of Artificial Aeration of Streams and
Waters, Illustrated by the Lippe as an Example, Lippe Associa-
tion, Essen, Germany.
112
-------
SECTION IX
APPENDICES
A. Diffusers and Mechanical Aerators [46]
Diffusers
Diffusers are devices which introduce air into liquids. They are
installed in various locations below the liquid surface on either fixed
or retractable mountings. Air is furnished by blowers, which are
usually in a central location. The blowers operate at a pressure
sufficient to overcome the static head of liquid above the diffusers and
the distribution losses. In many large plants, gas engines operated
from the digester gas are used to drive the blowers.
Diffusers may be classified as porous or nonporous (Table A. 1).
Porous diffusers in the form of plates or tubes are either of the
ceramic type, constructed of silicon dioxide or aluminum oxide grains
held in a porous mass with a ceramic binder, or of the non-ceramic
type, consisting of plastic-wrapped tubes or plastic-cloth tubes.
Nonporous diffusers may be of the nozzle, orifice, valve, or shear
type. Nozzle and orifice-type diffusers are constructed of metal or
plastic, have larger openings, and release larger bubbles than the
porous-type diffusers. Valve-type diffusers have a disc or valve
which closes when the air supply is shut off. They release larger
bubbles than do porous diffusers.
Shear-type diffusers provide for the reduction of the bubble size by
the shearing force of the water entering the diffuser at the open top in a
counter-flow direction to the upflowing air. These diffusers are square
in shape.
Other diffuser arrangements include water jets, which consist of com-
bined air units, and perforated or slotted pipes, which are used
occasionally as a temporary expedient or for unusual conditions. The
perforated or slotted-pipe diffusers are not offered for deep-submergence
diffusion, show a low oxygen absorption efficiency, and are readily
clogged or corroded. Many of the diffuser applications in the past have
been in treatment tanks, and much of the following discussion is in
relation to that type of usage.
Diffuser mountings may be either fixed or retractable. Portable hoists
may be used for raising the headers out of the tank for servicing. Some-
times porous plates are mounted in plate holders installed on the floor
of the tank. The location of the diffuser in an aerator tank has a con-
siderable impact on the efficiency of the device. Data are available from
studies on the oxygen transfer of diffusers at various tank locations,
113
-------
TABLE A-l
List of Diffuser Manufacturers
Manufacturer
Diffuser Type
Description
Aer-O-Flo Div.
Clow Corp.
Carborundum Co.
Chicago Pump
FMC Corp.
Dorr-Oliver, Inc.
Eimco Corp.
An Envirotech Div.
Filtros Plant
Ferro Corp.
FMC Corp. Link-Belt
Division
Fuller Co. , Infilco
Products
Hinde Engr. Co.
Keene Corp. (formerly
American Well Works
Norton Co.
Ray Products Co.
Rex Chainbelt Inc. ,
Pacific Flush
Tank Div.
Walker Process
Equipment Div.
Chicago Bridge &c
Iron Co.
Nonporous
Porous
Porous plastic
Porous plastic
flexible media
Nonporous
Porous
Nonporous
Nonporous
Nonporous
Nonporous
Porous
Nonporous
Porous plastic
Nonporous
Nonporous
Porous
Porous
Nonporous
Nonporous
Stainles s-steel-nozzle type with
check valve
Ceramic plates and tubes
Saran wrapped media, metal core
with integral end caps and
control orifice
Saran cloth flexible media, metal
frame with integral end cap
and orifice
Plastic nozzle type with integral
control orifice
Ceramic tubes with cast-iron
end caps and control orifice
Plastic and metal valve type with
integral control orifice
Nonporous metal shear type with
control orifice
Tubular metal grid, nozzle-type
diffusion, 0. 5 to 1 m below
water surface
Plastic base with elastomer cover
Ceramic plates and ceramic tubes
with and without integral end
caps and control orifice
Metal nozzle type, adjustable,
with 4 to 12 openings; ball
check valve
Porous plastic media, plastic
pan-type holder
Plastic aeration tubing
Variable-flow, multiple-orifice type,
made of cast bronze or aluminum
magnesium
Ceramic plates and tubes
Plastic media, various mountings
Nonporous-metal-valve type with
plastic ball valve
Plastic-nozzle type
114
-------
water depths, numbers of diffusers, tank width, diffuser spacings, and
air rates. Table A. 2 reports the results of these studies. Similar
studies have also been made of nonporous diffusers.
Mechanical Aerators
Over the past decade mechanical aerators have been widely used to
supply oxygen to treatment plants treating a broad range of flows and
organic loads, providing high removal efficiencies at comparable
power cos ts .
Mechanical aerators can transfer atmospheric oxygen to liquid by surface
renewal and interchange; and, when properly designed, mechanical aer-
ators meet the mixing requirements at various single and multiple-unit
installations for a broad range of tank sizes and configurations. Mechan-
ical aerators can also transfer atmospheric oxygen by dispersing
compressed air fed below the surface to a rotating agitator or turbine.
In the former case for an updraft-type aerator, oxygen is introduced
to the tank contents by lifting large volumes of liquid above the "water
surface and exposing it in thin films to the atmosphere. With the
plate-type aerator a high degree of turbulence is generated. Both
types of aerators transfer oxygen through incorporation and dispersion
of air into the liquid because of high surface agitation.
In the latter case, air bubbles are discharged from a pipe or sparge
ring beneath the turbine and are broken up by the hydraulic shearing
action created by the high-speed rotating blades of the turbine moving
through the liquid. The sparge ring is fed by compressed air; hence,
this system is actually a combined mechanical diffused-air system.
In the downdraft system oxygen may be supplied by air, self-induced
from the negative head produced by the rotor. With this system
external blowers or compressors are not required.
Brush-type mechanical aerators consist of a horizontal revolving shaft
with combs, blades, or circular discs with T-shaped bars attached,
extending below the water surface.
Table A. 3 lists some of the manufacturers of mechanical aerators.
The current trend is toward the use of mechanical surface aerators in
larger basins. Today there are many installations throughout the world
utilizing multiple units in large tanks. Mechanical aerators have been
used in both small and large activated sludge plants and aerated lagoons
treating domestic and industrial wastes.
Efficiencies up to 7 Ib O2/hp-hr have been attained for various aerators.
Actually it has been found that very high efficiencies can be achieved
on small pilot aerators but that for practical application, under normal
conditions, the larger aerators have an efficiency range of 2 to 4 lb/O2
115
-------
TABLE A, 2
Diffuser Efficiencies at Various
Locations in Tank Under Standard Conditions
Location
Diffusers mounted on a
header on the wall side
Diffuser s mounted on both side s
of a header near tank well
Diff users mounted on both
sides of a header located
near both tank walls
Diffuser s mounted on both
sides of multiple
headers
Type
Porous
Porous
Porous
Porous
Depth
(ft)
12. 75
12.4
12.4
12.4
Air Rate
4 cfm/ft
8 cfm/dif.
12 cfm/ft
8 cfm/dif.
8 cfm/ft
8 cfm/dif.
16 cfm/ft
8 cfm/dif.
20 cfm/ft
10 cfm/dif.
16 cfm/ft
8 cfm/dif.
72 cfm/ft
8 cfm/dif.
Effi-
ciency
(%)
9.7
11.5
12
12.8
12.5
14
16
Tank
Width
(ft)
24
24
24
24
24
116
-------
TABLE A-3
List of Mechanical Aerator Manufacturers
Manufacturer
Type
Chicago Pump
FMC Corp.
Dorr-Oliver Inc.
Eimco Corporation,
An Envirotech
Division
Fuller Co. , Infilco
Products
Keene Corp. (formerly
American Well
Works)
Lakeside Engineering
Corp.
Mixing Equip. Co.
Inc.
Permutit Co. , Div.
Ritter Pfaudler
Corp.
Vogt Manufacturing Co.
Walker Process
Equipment,
Div. Chicago Bridge
& Iron Co.
Welles Products Corp.
Yeomans -Clow
Updraft
Combination
Updraft
Plate
Plate
Updraft
Combination
Downdraft
Updraft
Updraft
Updraft
Aerator Characteristics
"Chicago"- propeller-driven flow
discharged against diffuser cone
at top
"D-O Aerator" -induced by sub-
surface rotor and compressed
air
"Simcar "- induced by rotating
impeller at top
"Vortair"-induced by horizontal,
radially vaned impeller at top
"Aer o-Accelator "-induced by
sub-surface rotor and com-
pressed air
"American"-induced by horizontal
vaned impeller at top
Horizontal rotating
"Lightnin1 "-induced by rotating
impeller at top with rotor below
surface
"Permaerator "-induced by surface,
sub-surface, and compressed
air
"Aer-O-Mix" - Produced by impeller
in tube; with radia] inlet troughs.
Induced by rotating updraft impeller
"Aqua-Later" -either submersible
or non-submersible pump dis-
charge through vertical tube
"YeoCone"-induced by spiral
vaned revolving cone at top and
draft tube
Sigma-induced by scoop-type
revolving blades at top
117
-------
transferred under standard conditions per horsepower hour. This
efficiency range normally covers the entire range of aerators
marketed today.
Plate Types
Typical plate types are the "Vortair," manufactured by Infilco/Fuller,
and the "American Aerator," manufactured by American Well Pump.
The "Vortair" consists of a circular flat plate with vertical blades
attached at the periphery of the plate. The "Vortair" uses a standard
motor and gear-drive unit. The plate rotates in a horizontal plane a
short distance below the normal water surface. When the aerator is
in operation, the top of the plate is clear of water. The performance
of the aerator depends on establishing proper design relationships to
determine the effects of diameter, blade sizes, number of blades,
speed of rotation, submergence, and other variables that affect horse-
power requirements and oxygen transfer.
Updraft Types
There are a number of updraft types available, including the Yeomans-
Clow "Sigma," the Eimco "Simcar," the Mixco "Lightnin," the Welles
Products " Aqua-Lator, " the Walker "Intens -Aer, " and one manufactured
by Chicago Pump, all of which are described below.
The Yeomans-Clow "Sigma" aerator operates on the updraft principle
with an impeller located at the surface of the liquid, designed to pump
large quantities of liquid at a low head. Individual blades, attached to
the rotating drive ring, are designed specifically to insure maximum
hydraulic efficiency. The use of individually mounted blades allows
for great flexibility in changing the capacity of the unit on the job site
at any desired future date. The unit is driven by a standard motor
and gear drive and is adaptable to the use of a variable-speed drive.
The Eimco "Simcar" aerator consists of a standard motor and gear-drive
unit, supported by walkway beams spanning the aeration tank. The
impeller is a cone-shaped disc with square-bar blades radiating outward
from the center. The blades are at or just below the surface of the
liquid. Liquid is drawn upward and outward by the rotating impeller
into a center cone and then is propelled outward in a low trajectory.
The Mixco "Lightnin'' aerator is a pitched blade, open-style turbine
utilizing a motor and gear-drive unit. The aerator is located at the
surface of the liquid and operates on the updraft principle. The aerator
usually consists of four blades pitched at a 45-degree angle. Operating
speeds are generally between 30 and 60 rpm. At times, a smaller rotor
is installed on an extended shaft near the bottom of the tank to increase
turbulence and maintain solids in suspension.
118
-------
The "Aqua-Lator" (Welles Products Corporation) is a relatively simple
aerator, consisting of a special submersible or nonsubmersible pump,
a riser tube, a fiberglass-covered float, and an orifice-diffuser
assembly. Power is supplied from an on-shore control station by
means of a submersible power cable. The "Aqua-Lator" is supported
in the liquid by its own integral float, which automatically adjusts to
water-level changes.
The Walker "Intens-Aer" aerator consists of an updraft unit pumping
large volumes of liquid at low head by means of a slinging-blade
impeller located near the surface of the liquid. Oxygen transfer ossurs
from the large interfacial area created by the impeller discharge
pattern, from high turbulence and surface wave action, and from en-
trained air bubbles carried down with the circulating tank liquid by
the high tank-turnover rates established.
A mechanical aerator manufactured by Chicago Pump, FMC Corporation,
consists of an impeller on a vertical shaft located near the top of a
vertical draft tube, which is in the center of the tank. This impeller
lifts the liquid up the tube and discharges it against a diffuser cone,
which deflects it horizontally and down to the surface.
Downdraft Types
The Vogt Manufacturing Company makes the "Aer-O-Mix," in which
an impeller in a vertical tube forces liquid from the top down through
the tube to the bottom of the tank. The liquid entrains the air from
above. The mixture expelled from the bottom of the tube rises through
the tank. This system, in principle, is a "downflow contactor" as well
as a surface aerator.
Combination Types
The "D-O" turbine (Dorr-Oliver, Inc. ) is powered by a standard,
direct-connected motor and gear drive that is secured to a beam
structure spanning the tank. Two or more turbine impellers can be
used, one usually located near the bottom of the tank above a sparge
ring, and another located about 30 in. below the surface. The upper
impellers, or any intermediate impellers located below the surface,
serve as shearing devices for the compressed air released through
the sparge ring.
The "D-O" aerator transfers oxygen through the emission of air at the
sparge ring and the shearing action on the bubbles by the rotating
impellers located above the sparge ring.
The Permutit " Permaerator" uses two motor-driven impellers and a
sparge ring at the lower impeller. The Permutit aerator sparge ring
is designed to release air within the diameter of the adjacent impeller.
The upper impeller on the Permutit unit is located near the surface and
functions as a surface aerator.
119 olI.S. GOVERNMENT PRINTING OFFICE: 19" 484-482/201-3
-------
1
5
2
Sub/iM'1 Fii-lriSr. Group
O5G
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
JBF Scientific CorDoration
2 Ray Avenue, Burlington, Massachusetts 01803
Title
ENGINEERING METHODOLOGY FOR RIVER AND STREAM REAERATION
1Q Authors)
Murro, Ronald P.
Yeaple, Donald S.
16
Project Designation
16080 FSN 10/71
91 Note
22
Citation
23
Descriptors (Starred First)
*Water quality control, *Aeration, Dissolved oxygen, Oxygenation,
River Basins.
25
Identifiers (Starred Firsl)
*Induced aeration, ^Oxygenation efficiency, Air distribution,
Molecular oxygen, Surface aerators, Diffusers, Downflow contactors,
U-Tubes, Sidestream mixing, Oxygen balance.
27
Abstract
Results of recent activities in river and stream aeration by artificial techniques
are reviewed, and a rational engineering methodology is developed for future
river and stream aeration projects.
The development of the methodology follows from a thorough review of the oxy-
gen dynamics in rivers and streams and the capabilities of aeration systems
within the present state of the art. The report shows how the theoretical work
can be simplified considerably and applied to the solution of river and stream
water quality problems. It is assumed that aeration would only be used as a
"polishing" action after all identifiable waste sources have received at least
secondary treatment.
The results indicate that,with careful consideration of site factors,artificial
aeration can be applied successfully to raise dissolved oxygen to 5 pprn, using
mechanical surface aerators, diffusers, downflow contactors, and sidestream
mixing. However, since the transfer of oxygen from air into water is relatively
inefficient above 5 pprn DO, the introduction of molecular oxygen through side-
stream mixing, U-Tubes, and pos sibly diffusers should be considered, depending
on the volume of water to be aerated. In cases where DO may be maintained at levels
lower than 5 ppm, systems using air are competitive with molecular oxygen, depend-
ing on site conditions.
A bstrnctor Institution
Donald S. Yeaple, Sr StffEngJTRF firip.ntifir. Corporation
WR:t02
-------
ENVIRONMENTAL PROTECTION AGENCY
Publications Distribution Section
Route 8, Box 116, Hwy. 70, West
Raleigh, North Carolina 27607
Official Business
POSTAGE AND FEES PAID
ENVIRONMENTAL PROTECTION AGENCY
------- |