tTLEAl
       WATER POLLUTION CONTROL RESEARCH SERIES  I6080DUP12/70
  OXYGEN REGENERATION OF POLLUTED RIVERS
              THE  DELAWARE RIVER
  ENVIRONMENTAL PROTECTION AGENCY
WATER QUALITY OFFICE

-------
 - about our cover
The cover illustration depicts a city in which man's activities coexist in
harmony with the natural environment.  The 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 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
                    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 Improvement
  of Power Releases
16080BWPU/70
Induced Air Mixing of Large Bodies of Polluted Water

-------
           OXYGEN REGENERATION OF POLLUTED RIVERS;
                      THE DELAWARE RIVER

                              by

                      William Whipple, Jr.
                        Joseph  V.  Hunter
                        Frank W. Dittman
                           Shaw L. Yu

                              of

                      Rutgers University

                              and

                      George  E. Mattingly
                              of
                     Princeton  University
                            for the

               ENVIRONMENTAL PROTECTION AGENCY
                    Water Quality Office
                    Program No. 16080 DUP
                       December, 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402  Price $1

-------
                 EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents neces-
sarily 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.

-------
                                 ABSTRACT
Tests of surface instream aerators and of bottom diffuser aerators were
conducted on the Delaware River near Philadelphia in order to determine
the practicability of induced oxygenation of deep navigable rivers.
Techniques used were based mainly upon those developed previously by
Rutgers University upon a smaller river, with special adaptations re-
quired to determine oxygen transfer rates, since the equipment affected
only a small part of the channel.  The diffuser was tested at various
depths up to 38 feet, but its performance in pounds of oxygen per
horsepower hour decreased markedly in the deeper water.  Performance of
the surface aerator appeared to be somewhat improved over results pre-
viously found in a shallower river.  Cost estimates and systems analysis
led to the following conclusions:

       (a)  That induced oxygenation by aerators appears to constitute
an economical alternative to advanced waste treatment on the Delaware
River.

       (b)  That surface aerators can readily be reinforced to operate
economically on large rivers, but can only be used in areas where they
will  not interfere with navigation.

       (c)  That diffuser aerators of the type tested are closely com-
parable in economy to surface aerators, and can be used in port areas
without interference with navigation.

       (d)  That oxygen diffusers developed by others may provide an
even  more economical means of induced oxygenation in waters subject to
navigation, provided certain problems can be solved.

This report was submitted in fulfillment of Grant No. 16080 DUP under
the partial sponsorship of the Environmental Protection Agency.
                               -                        
 KeyWords:    Dissolved  oxygen,   water quality control,   oxygen sag,
              aeration,* stream improvement,*  stream pollution,
              biochemical oxygen demand, dispersion, pollution abate-
              ment,  surface  aerators,* air diffusers,*  induced
              oxygenation,*   instream  aeration.*

-------
                      CONTENTS
Section
    I.
   II.
  III.
   IV.
    V.
   VI.
  VII.
 VIII.
   H.
Conclusions and Recommendations
Introduction
Field Operations
Oxygen Transfer of Aerators
Dispersion Analysis
Delaware Aeration Systems Analysis
Design and Cost Considerations
Acknowledgements
References
Appendices
vii
  1
  7
 23
 lil
 55
 65
 81
 83
                          ii

-------
                               FIGURES
1  General Map                                                      2

2  Dissolved  Oxygen  Criteria and Recent Conditions                  3

3  Camden Aerator Site                                              

k  Aerial View  of Camden Aerator Site                               9

5  Surface Aerator Damaged by Storm                                H

6  Rebuilt Surface Aerator in Operation                            12

7  Student Assistants with Dissolved Oxygen Meter                  13

 8  Student Assistants with Current Meter                           lit

 9   Operation  of Submerged Diffuser Aerator                         17

10  DO and Velocity Traverse  at Upstream Sampling Section          26

11   Typical Flow Pattern in the  Vicinity  of Mechanical Aerator      28

12   Typical Flow Pattern in the  Vicinity  of Diffuser Aerator        29

13   Oxygen Increase Resulting from Mechanical Aerator Operation    32

lU   Cross-Sectional Distribution of Oxygen Increase due to Aeration 3^

15   Compressor Brake  Horsepower  V3 Submergence Depth for Different
    Engine Speed                                                   36

16  Oxygen Transfer Rates VS  Diffuser Submergence                   38

1?  Oxygen Absorption VS Diffuser Submergence:  Comparison of
    Results                                                         39

18  (a)  Sketch of Dye Patterns in a Stream                         U3

    (b)  Plot of One-Dimentional Unsteady Dye Concentrations VS
         Downstream Distance                                        ^3

19  Dye Cloud Frontal Structure VS Time After Injection at Two
    Stream Stations                                                 ^6

20  Differentiated Dye Cloud Frontal Structure VS Time After
    Injection at Two Stream Stations                                ^7
                                   iii

-------
                            FIGURES  (Gont'd)
21  Stream Cross-Sectional Area Affected by  the Diffusion
    Aerator VS Downstream Distance                                 50

22  Graphical Results of Cross-Sectional Area Distribution and
    Reaeration Percentage VS Downstream Direction                   51

23  Estuary Sections, Composite DO Profile                          56

2k  Oxygen Mass Balance on Section i                               58

25  Dissolved Oxygen Dispersion                                    

26  Single and Twin Aerator Facilities                             69

2?  Surface Aerator Mooring                                        71

28  Diffuser Aerator Layouts                                       75

29  Diffuser Aerator at Pier End                                   76
                                    iv

-------
                                 TABLES

                                                                    Page

1  Tabulation of Gas Samples                                         21

2  Summary of Field Test Data for Mechanical Aerator                 31

3  Summary of Oxygen Transfer Results, Mechanical Aerator            31

h  Summary of Field Test Data at Camden, N.J., Diffuser Aerator      31

5  Summary of Field Test Data at Philadelphia, Diffuser Aerator      3U

6  Summary of Oxygen Transfer Results, Diffuser Aerator              3h

7  Computation  of  Oxygen Required to Maintain a h mg/1 DO Level
   in the Critical Region                                            60

 8  Diffuser Aeration System DesignAn Example                       62

 9   Comparison of Oxygen Increase Due to  the  Present Design and
   That Based on DECS DO Response Matrix                            63

10 Cost Estimate,  Surface  Aerator Sites                              70

11  Cost Estimates, Diffusion Aeration  Facilities                     Ik

-------
                                CONCLUSIONS
      1.  Induced oxygenation of deep navigable rivers by river aeration
appears to be an entirely feasible and economical alternative to advanced
waste treatment under appropriate conditions, costing approximately half
as much to achieve a given dissolved oxygen level on the Delaware River.

      2.  Mechanical surface aerators provide an economical means of
adding oxygen to deep rivers; but they may need to be reinforced struc-
turally, and can only be used for areas where they will not interfere
with navigation.

      3.  The submerged diffuser aerators, with coarse holes, have about
the same costs as surface aerators for oxygenation of deep rivers.  They
are relatively most efficient in waters l - 20 feet in depth.  They
should not adversely affect navigation, except perhaps for small, rapidly-
moving craft, such as outboards.  Their relative economy declines rapidly
where long underwater pipelines are required.

      U.  Based upon experiments of others, it appears that diffuser
aerators would have a much higher transfer rate if built with smaller
apertures.  Diffusers with small apertures have been found to plug up
when not in use, due to both physical and biological causes.  However,
means may be found to avoid the plugging.

      .  On the be sis of available estimates, oxygen diffusers developed
by others appear to be less costly than aeration, especially in deep
water.  However, the problem of dispersion would have to be solved before
they could be used in a large navigable river.
                                     Vll

-------
                               RECOMMENDATIONS
      1.  That in planning increased consideration should be given to the
possibilities of induced oxygenation of rivers, either by aeration or
oxygenation, or a combination of both, as appropriate to the circumstances.

      2.  That research and development should be continued upon oxygen
diffusers and fine bubble air diffusers, including means of minimizing
plugging of apertures, and of obtaining dispersion throughout the river
cross section.
                                    ix

-------
                                 SECTION I

                               INTRODUCTION
In 1967 and 1968, the Water Resources Research Institute of New Jersey
conducted tests of instreatti aeration on. the Passaic River, supported by
EPA grants (Phase I, 16080 DUP) and accompanying O'WRR research projects
(B-010-N.J., B-011-N.J., A-017-N.J., and B-002-N .J.).  Financial support
also was provided by the Department of Conservation and Economic De-
velopment, State of New Jersey.  The results, which have been reported
previously (l, 2, 3, k, 5) indicate that the artificial or induced aera-
tion of rivers may provide a more economical method of improving oxygen
levels than advanced waste treatment of effluents and that for small
rivers, surface aerators are more economical than diffusers.  The present
study constitutes Phase II of the original program.  It is based upon
field tests conducted upon the Delaware River in the summers of 1969 and
1970.  The purposes of the study were to test both  surface and diffuser
aerators on a wide, deep, navigable stream; to determine efficiency,
economy, and operating characteristics of these aerators; and to prepare
prototype designs and cost estimates of aerator installations appropriate
to such rivers.

The lower Delaware River, below Trenton, N.J., is designated as the
Delaware Estuary; and it is the sections between Philadelphia and Chester,
Pa., inclusive, which constitute the main problem area from the viewpoint
of water pollution  (see map, Figure 1).  The situation on the Delaware
Estuary is recognized nationally as a major water quality problem.  It
was analyzed in a 1966 report  (6) and has been the  subject of much action
by the Delaware River Basin Commission.  The situation is well summarized
in a recent paper  (7) and a report  (8).  About UO miles of the river suffer
from insufficient dissolved oxygen  at times, as shown in Figure 2.  The
oxygen deficiency interferes with important runs of anadromous fish, be-
sides impairing the quality of the river for recreational and living pur-
poses.  Adopted dissolved oxygen standards will require bringing minimum
daily average dissolved oxygen  (DO) levels up to about 3-5> mg/1.  The
cost of implementing this program is very great, of the order of $UO mil-
lion annually;1 and it will require a high degree of treatment.  In view
of the economic importance of the situation, and the definitive evalua-
tion of the optimum program obtainable by means of  treatment alone, the
Delaware Estuary provides an ideal  situation to study the feasibility of
a program of instream aeration, to be used as a supplement to a  (somewhat
reduced) program of effluent treatment.
 1  Reference 7  indicates annual  cost  estimates  of $46  million and
    million, respectively, for  two  given objectives,  the  adopted  plan ly-
    ing somewhere between them.   Comparison  of these  figures, t iking into
    account differences  in dissolved oxygen  objectives  for  various reaches
    of the river, indicates  that  the annual  cost of adding  the last  1.0 mg/1
    of dissolved oxygen  to the  critical  section  of the  river would be over
    $9 million  annually.

-------
Philadelphia^?  Zone  3
                                  Trenton
             NEW  JERSEY
                 FIG.  1   GENERAL  MAP

-------
.
              8 mg/1 d. o.
                         Zone  2
\
                    \
                                          i
                          Minimun  daily
                              value


                             Zone  3     i     Zone  4   _   I.
                                                          |   stream  criteria
                                           Minimum daily
                                             average
                                           recent summer /
                                                                                      Zone  5
                         FIG.      DISSOLVED OXYGEN CRITERIA  AND  RECENT  CONDITIONS

-------
The Delaware River at the test site is about a half-mile wide.  The nor-
mal tide range is about 6 feet.  Just after high or low tide, the river
flows relatively quickly for a short time; then velocity decreases slowly,
usually ranging between 1 and ls ft/sec, in water depths of 20 to 30 feet.
The main navigation channel is maintained at UO feet of depth, or over.

The demonstration project (Phase II) approved was for a year only, ending
30 November 1969, but notification of the grant was only given in May
1969, leaving little time for preparations.  The tests of surface aerator
and diffusers were carried out during the summer of 1?69 substantially as
planned, at a site near Camden, N.J.  However, upon later analysis of
results and discussions with the Corps of Engineers and Delaware River
Basin Commission, it was found that, as regards diffuser aerators, the
conditions tested did not include water of sufficient depth.  The require-
ments of navigation entirely preclude the use of surface aerators in port
areas and limit diffusers to areas outside the channels and anchorages.
In view of channel configuration and river dispersion characteristics,
diffusers must be utilized in water of 30 or even UO feet in depth along
considerable reaches.  Since performance at such depths had not been
tested, it was decided to postpone submission of the report and to arrange
for additional diffuser tests in deeper water in June 1970.  The Federal
funds had by that time administratively terminated; but tests were conduct-
ed with remaining state funds.

During the final period of analysis, information was received indicating
that under some conditions use of pure oxygen may be more economical than
air diffusers.  Also certain other types of air diffusers may be prefer-
able to those tested.  These possibilities, although beyond the project
scope, are mentioned as a matter of perspective.

The first phase of the project was the field  operations of the aerators
in Gamden, N.J., described in Section II.  As previously indicated, part
of the field operations took place during the second summer.  The second
phase was the calculation of oxygen transfer  of the aerators.  (See
Section III.)  In principle, this analysis resembles similar calculations
for the Passaic River aerators (l); except that subsequent laboratory
work has indicated a change in one of the constants.  In practice, how-
ever, much more elaborate methods were necessary to obtain oxygen uptakes
on the Delaware River, since, in the larger river, it is much more diffi-
cult to determine which portions of the flow are affected by the aerator.
Also, entirely different methods involving a gas chromatograph were re-
quired to obtain oxygen uptake of the deeper diffusers.  Section IV
describes the dispersion field tests, using fluorescent dyes, and the
analysis required to determine dispersion characteristics of the river.

Once the transfer rates, navigation requirements, and dispersion character-
istics had been determined, it was possible to develop a series of proto-
type aeration sites which would be practicable on such a river.  Different
sizes are required, in order to apply to reaches needing more or less
supplementary oxygen.  Cost estimates were prepared for these systems by
design consultants, Hazen and Sawyer.  These  considerations are covered in

-------
Section VI.  Section V is a discussion of a systems analysis of oxygen
characteristics of the Delaware Estuary, which derives an estimate of the
amount of supplemental oxygen required to raise the dissolved oxygen level
by a given amount, and draws inferences as to approximate costs of river
oxygenation by means of mechanical facilities.

This report incorporates mainly the work of the following:

William Whipple, project director, general planning and coordination, basis
for design.

Joseph V. Hunter, Department of Environmental Science, water chemistry and
biochemistry.

Prank W. Dittman, Department of Chemical Engineering, field operations,
power estimates and gas chromatography.

Shaw L. Yu, Department of Civil and Environmental Engineering, oxygen
uptake and systems analysis.

George E.  Mattingly, Department of Civil and  Geological Engineering,
Princeton  University, dispersion  analysis.

Francis P.  Coughlan, Jr., and Melvin Stein, Hazen  and Sawyer,  design con-
sultation  and cost estimates.

Personnel  of U.S. Geological Survey,  Trenton,  N.J.  and Washington, D.C.,
whose contributions are acknowledged elsewhere in  this report.

Dr. Dittman was primarily responsible  for Section  II, Dr. Yu for  Section
III and V,  Dr. Mattingly for Section  IV, and  General Whipple for  Sections
I and VI.   Various students also  assisted in  field work  or  office opera-
tions .

-------
                                SECTION II

                             FIELD OPERATIONS
Operating headquarters for the 1969 Delaware River tests was located in
Camden, New Jersey, at a pier projecting into the river from the east
bank.  The pier is part of the Camden Sewage Treatment Plant, located
about midway between the Ben Franklin and Walt Whitman bridges.  It was
generously made available by the city, free of charge, on recommendation
of the plant manager, Mr. John Frazee.

The mechanical and diffuser aerators were located on a common north-south
center line about ?5 feet west of the end of the pier, and about 1^0 feet
downstream (see Figures 3 and li).  About 12 pile clusters or individual
pilings had to be driven into the bottom for several purposesto anchor
the aerators, to support the necessary services extending outward from
the end of the pier, and to provide reference points for locating the
boats while instream measurements and samples were being taken.  The Corps
of Engineers granted an official permit for the work, being in navigable
waters.

The services extending out from the pier included: (a) one 12-inch steel
pipe carrying compressed air for the diffuser aerator; (b) one UUO-volt,
3-phase, 200 KVA electrical supply cable for the mechanical aerator;
(c) one 3/4-inch plastic hose for Rhodamine B dye solution used in dis-
persion studies; (d) thermocouple wires and pressure leads of 1/U-inch
plastic tubing to an orifice meter used for air flow measurements.

The entire river installation had to conform to two navigation restric-
tionsthe space at the end of the pier had to be kept clear at all times
for the sewage plant sludge barge, and the navigational anchorage area
could not be infringed upon by any pilings or equipment.

Compressed air for the diffuser aerator was supplied from the pier by the
same diesel-compressor unit used for the Passaic River Study in 1968  (see
Ref. l).  This was a Fuller Company Sutorbilt rotary positive blower with
silencer, driven by a 250 hp Cummins diesel engine, with muffler, through
a clutch and reduction gear assembly.  Pressures up to 21? psi were obtain-
able.  The fuel tank was elevated on a steel framework so that its bottom
was about two feet above the engine fuel intake.  Air from the 12-inch
supply pipe, metered by the orifice plate connected to an inclined manom-
eter, was divided at the outward end into two 8-inch branches leading to
the two 8-inch underwater headers with a total of 160 diffuser nozzles.
lank-Belt 3/h-inch adjust-air diffuser nozzles, each having 12 openings
5/32 inches in diameter, were used.

The 75> horsepower Yeomans mechanical aerator was the same unit used in
the Passaic River study in 196? and 1968 (Ref. 1); electric power for it
was supplied by a portable 200-kilowatt diesel-generator unit on the pier.
Switchgear for the electrical output was mounted in a small existing shed
near the outer end of the pier.

-------
                           N
                                                  0
                                                25
50 ft.
                                                  1.  Diesel blower
                                                  2.  Generator
                                                  3,  4   Trailers
                                                  5.  Shed
                                     Mechanical
                                         aerator
                                                         Diffuser
                                                           aerator
FIG. 3
CAMDEN  AERATOR  SITE

-------
FIG. U    AERIAL VIEW OF CAMDEN AERATCR SITE

-------
Two trailers, one for the office and laboratory and the other for equip-
ment storage, were installed on the pier for the duration of the project.
110-volt power for the office-lab trailer was supplied from a line pre-
viously installed to serve a floodlight at the end of the pier.

The installation of both aerators was completed on Friday, July 11, 1969.
During the following day, a brief but violent storm occurred at about
6 pm; as a result of wave action, the mechanical aerator, designed for
service in treatment plants or quiet waters, collapsed.  It failed to
sink only because its supporting pontoons were undamaged; two of the
three radial pairs of steel channels were badly twisted, and the motor was
damaged by water (see Figure 5>)

A revised design was quickly developed, using fabricated box girders in-
stead of simple channels for the radial members.  Meanwhile, the motor was
dried out and reconditioned.  The installation of the redesigned and re-
built mechanical aerator was completed on Friday, August l.  Initial
operation occurred August 18, and satisfactory operation was achieved on
August 21, after several problems were corrected (see Figure 6).  The
diffuser aerator was not affected by the storm.

Two llt-foot open boats with 10 horsepower outboard motors were used for
direct instream measurements, for taking water samples, and for miscellan-
eous water transportation.  Since the waters around the pier were complete-
ly unprotected against storms, ship wakes, and boat thieves, the boats
were kept at the nearest marina, about 3 miles away, and were brought to
the project daily.  Additional boats and special equipment, used for a
few days during the dye dispersion studies, were furnished by the U.S.
Geological Survey office at Trenton.  Their boats were stored temporarily
at the Gloucester Coast Guard Base, immediately south of Camden.  One
boat each was also borrowed for dispersion tests from the Delaware River
Basin Commission and the Corps of Engineers Philadelphia District.

The data taken from the boats on a daily basis included instream measure-
ments of dissolved oxygen, temperature, velocity, depth, and boat position.
Numerous water samples were also taken to check the dissolved oxygen meter
readings against dissolved oxygen via Winkler titration.  Samples of spent
air from the diffuser were also taken.  When an aerator was operating, two
boats were used to make simultaneous cross-traverses, upstream and down-
stream of the aerators.  Two men were needed in each boat to maneuver and
position the boat, and then to take oxygen and velocity data  (see Figures
7 and 8).  Daily data taken on the pier included atmospheric temperture,
pressure, and oxygen content, air flowmeter readings on the 12-inch supply
line, electrical data on the mechanical aerator drive, and operating data
on the two diesel engines.  Also on the pier, Winkler samples previously
fixed in the boats were titrated in the office-laboratory trailer.

One of the complications of data-taking in a tidal estuary is the diurnal
tidal cycle; water is flowing upstream for about 12 hours of each 2k.
Random eddies, unpredictable mixing zones, and additional turbulence occur
where the incoming tide meets the outgoing river water.  The direction of
                                     10

-------
FIG. 5    SURFACE  AERATOR DAMAGED BY STORM

-------
FIG. 6    REBUILT SURFACE AERATOR IN OPERATION

-------
FIG. 7    STUDENT ASSISTANTS WITH DISSOLVED  OXYGEN METER
                            13

-------

FIG. 8    STUDENT ASSISTANTS WITH CURRENT METER

-------
the flow should not affect the performance of the aerators, but because
of the mixing phenomena, it would affect the accuracy of our oxygen input
calculations.  For consistency, it was decided to operate the aerators
and take data only during the periods of out-going tide.  This decision,
of course, required that data be taken at different time periods each day,
using the tide tables as a guide.  Most of the work could be done in day-
light; on only a few occasions was there need to continue after sundown.

Two Delta Type 85 portable oxygen meters, previously used on the Passaic
River project, were available for the work on the Delaware.  Results with
these portable instruments had been satisfactory, provided that they were
calibrated by the Winkler method at the beginning of each data-taking
period.  These Type 85 meters, however, had cables only 6 to 8 feet long,
which were not adequate for the Delaware.  Therefore, while some data
were taken with the Type 85 meters, portable polographic meters with
cables about U5 feet long were obtained and used during most of the proj-
ecta Delta Type 75 from another OVER project at Rutgers, and a Yellow-
springs meter generously made available by the U.S. Geological Survey at
Trenton.  Meters were calibrated at the beginning of each data-taking run
against the Winkler titration method, azide modification, using the same
sample of river water for both measurements.  The meters were adjusted  to
agree exactly with the Winkler method at the beginning of each run; data
taken during the run usually indicated a maximum difference of - 0.5 mg/1
between the two methods.  During July and August, 1969, the Delaware
River at  Camden usually contained 0.5 to 1.5 mg/1 of oxygen except after
heavy general rains, when it increased to a range of about 1+ to 5 mg/1.

The  measurement of water temperature at or very close to the oxygen probes
is important for two reasons; it affects the meter readings, and it af-
fects the saturation concentration of oxygen in water.  The Delta 85 and
Yellowsprings instruments are each equipped with a built-in thermistor
close to  the oxygen probe, plus a measuring circuit.  The  temperature is
measured  and recorded on the data sheet, and the temperature-compensating
dial of the instrument is set accordingly, before the oxygen measurement
is made.  Since the Delta 75 lacks the temperature-measuring feature, but
still requires the setting of a temperature-compensating dial, it was
used in conjunction with a Model 380 electrical thermometer, made by RFL
Industries of Boonton, N.J.  This thermometer has a cable  of about the
same length as the Delta 75; in operation, the two cables  were taped to-
gether .

Point values of river velocity to define the velocity profile were made
at various depths and locations, using weighted current meters, made by
Gurley Engineering Instruments of Troy, N.Y.  These instruments had cables
about 50  feet long, so that velocity could be measured  at  any desired
depth.  Prior to use on the project, these Gurley meters were recalibrat-
ed,  using a moving boat technique in a swimming pool.  Results agreed well
with the  manufacturer's calibration.  The manufacturer's calibrations were
used in making calculations from the data.  The depth,  at  which each
point  measurement of temperature, velocity, and oxygen  concentration was
made, was obtained by means of red and black  tapes fastened  to  the in-
struments' cables at 5-foot intervals.
                                      15

-------
The position of a given boat for a given measurement was determined by
reference to a coordinate system based on distances from the pier and from
various pilings or pile clusters.  Distances of the boats from the refer-
ence points were obtained using calibrated ropes.  Where possible, the
ropes were stretched between pilings.  When working on the outboard side
of a single piling, the outboard motor had to be kept running to keep the
boat in position at the end of the rope.  This method was required because
of the restriction against driving pilings in the navigational channel.

Ten to fifteen river water samples for Winkler analysis were taken as spot
checks on the meters during each traverse, usually at 1-foot depths only.
Standard 200-ml bottles with ground-glass stoppers were used.  Samples
were "fixed" in the boats immediately after being taken; that is, the oxy-
gen content of the water in the bottle was converted to a chemically
equivalent quantity of iodine (12) in solution.  Back at the office-lab
trailer, the iodine was titrated with standard ^28203 solution to deter-
mine the original oxygen content.

The diffuser manifold, 80 feet long, consisting of two headers 5 feet
apart, was installed between two clumps of piles, supported so as to re-
main in a horizontal position, averaging about 13 feet below the surface.
At this depth the operating characteristics were excellent.  The 5-foot
horizontal distance between the headers brought bubbles up in a fairly
uniform pattern, with an upward velocity which seemed to be perhaps 1.0
to 1.5 ft/sec  (see Figure 9).

In a permanent installation, the clumps of piles actually used would be
objectionable to navigation interests.  One of three alternatives would
be used:

       (a)  Lay the diffuser on a sloping bottom and vary the size of
diffuser apertures to correct (approximately) for the different depths.

       (b)  Use a support structure of some kind to hold the diffuser
horizontal despite a sloping bottom.

       (c)  Lay the diffuser upon a bottom which was either already level,
or was specially leveled prior to the installation.

It had been anticipated that the oxygen input to the Delaware in terms of
pounds or tons per hour might not be calculable accurately if based on a
small observed concentration rise multiplied by a large flow rate.  Also,
the exact quantity of water passing through the aeration zone is diffi-
cult to determine precisely when only part of the stream is being aerated.
In the case of the mechanical aerator, there is no alternate simple,
available way of verifying the oxygen input based on water analysis and
flow rate.  For the diffuser aerator, however, there is an available
method, namely, to take samples of the spent air and analyze them for  oxy-
gen content.  Then, knowing the flow rate and oxygen concentration of  the
inlet air, it should be possible to calculate the oxygen input to the
water and the efficiency of absorption.
                                     16

-------
-


         FIG.  9     OPERATION OF SUBMERGED DIFFUSED AERATOR
                                 17

-------
Samples of spent air were taken from a boat over the aeration zone, using
a large plastic funnel with its stem connected to a rubber hose.  The
hose, in turn, was connected to sample bulbs with inlet and outlet stop-
cocks.  With the stopcocks open, the funnel was inverted and submerged
over the aeration zone, and the spent air allowed to flow through for a
few minutes.  After the bulb had been sufficiently purged, the stopcocks
were closed to trap the required sample.

Initial results, using 5> to 10 ml portions of each sample in the Beckman
OrC-5 gas chromatograph, were disappointing.  The same atmospheric air
sample, for instance, gave 02 contents varying from 19-5 to 22.2 volume
percentgood enough for some purposes, perhaps, but not for calculation
of the percent 02 absorbed.  Obviously, the analyses of the Camden spent
air samples done in this way were not reliable.

The instrument manufacturer, when consulted about the problem and shown
copies of data from several runs of this degree of scatter, attributed
the error to unconscious variations in the operator's technique from one
trial to the next.   The variations occurred in the rate at which the
operator pushed the plunger of the syringe when injecting the 1 ml
sample of gas.  A Beckman Gas Sampling Valve of the manual type was then
obtained.  This valve traps a portion of sample of constant volume in a
loop of capillary tubing; by turning the valve to the alternate position,
the operator then injects the sample at the same rate each time.  By the
use of this device, 02 contents of various portions of the same air sample
ranged from, for example, 20.9 to 21.5 percent 02 j the average for 5
portions of this sample was 21.1/6 02.  Other air  samples gave results as
good or better.

The diffuser aerator at Camden operated at depths averaging about 13
feet.  It was determined after analysis and discussion of aerator results
with agencies concerned that the extensive pier and anchorage areas of
the Philadelphia port and the navigation channels would not permit use
of surface aerators; moreover that diffuser depths of up to UO feet in
some areas would be required.  A search of the literature did not dis-
close any basis upon which the performance of these or any other aerators
at depths of 30 to hO feet could be predicted.  Accordingly, a decision
was made to carry on another set of tests using a diffuser aerator at
various depths in June 1970.

These tests were performed at a deep water pier on the west bank of the
Delaware River, about 3 miles upstream from the Camden test site, made
available through the courtesy of Brig. Gen. Allen F. Clark  (ret'd),
director of the Philadelphia Port Corporation.  At this pier a maximum
water depth of UO feet was available, so that if a variable depth diffu-
ser with controllable air supply could be devised and spent air samples
could be taken, a correlation of percent oxygen absorption with depth
and air flow rate should be obtainable.

A 3-inch diameter steel pipe diffuser header, 6 feet long, was made up
with a 3-inch cap at one end and a 1-inch hose connection at the other
                                      18

-------
end.  At 1-foot intervals there was a Link-Belt 3/U-inch Diffusair nozzle,
with 12 ^/32-inch air outlets.  This was the same kind of nozzles and the
same nozzle spacing as in the large diffuser previously used.  The header
was suspended from an improvised hoist with a manually operated winch and
marked cable, so that the depth in the water could be controlled down to
the maximum of about 1*0 feet.  The axis of the header was horizontal and
it was held roughly perpendicular to the current by the air supply hose.

The header was supplied with air from a Jaeger gasoline-driven recipro-
cating compressor having a maximum output flow of 125 SCFM at 100 psig
pressure.  The air receiver outlet was equipped for the occasion with a
manually variable pressure regulator so that, while the receiver pressure
rose and fell, and the compressor started and stopped automatically, the
pressure and flow rate of the air supplied to the header remained con-
stant.

From the pressure regulator, the air flowed through a Brooks rotameter
with a maximum capacity  of 70.8 SCFM; the meter outlet was equipped with
calibrated temperature and pressure instruments and a sampling  connec-
tion for collecting samples of the compressed air.  The air  then  flowed
through a  standard 1-inch compressed air hose about 0  feet  long  to the
air header in the water.

During the operation,  19 spent air samples and  h  compressed  inlet air
 samples were taken over  a 2-hour  period.  Three different header  depths
were used, and three  different air  flow rates were  used at each depth.
Data taken were  as follows:

       1.   Time of day
       2.   Air rotameter  scale reading
       3.   Air rotameter  gauge pressure
       k.   Air rotameter  temperature,  F
       5.   Atmospheric pressure,  inches of mercury
       6.   Atmospheric dry bulb  temperature,  F
       7.   Atmospheric wet bulb  temperature,  F
       8.   Dissolved  oxygen  in water,  mg/1
       9.   Dissolved  oxygen  sensor depth,  ft.
      10.   Water  temperature, C
      11.   Spent  air  sample  number
      12.   Compressed air sample number
      13.   Air  header depth,  feet.

 Data items 2-7,  12,  and 13  were taken by Observer No. 1, on shore.  Items
 8-11 were taken by Observer No. 2, in a boat over the zone of rising
 bubbles.  Both Observer 1 and 2 recorded the time (data item l) from syn-
 chronized watches.   Observer No. 3, also in the boat, held the inverted
 funnel in the water  over the rising bubbles and filled the sample bulbs,
 using a collecting period of 30 to 60 seconds for each sample.  The
 weather was cloudy but there was no wind or rain.  The only difficulties
 were caused by the wakes of passing tugs and boats.
                                      19

-------
The chromatographic analysis of these spent air samples and their correla-
tion with depth and flow rate yielded interesting results, which are
summarized in Table 1, and discussed in Section III.

A special feature of the 1969 project was the plotting of the dispersion
pattern of the oxygenated water from the aerators by injecting Rhodamine
B organic dye solution into the water at the point of aeration.  Using
a group of 3 small centrifugal pumps in series, with a capacity of 5 to
20 GPM, river water was pumped continuously up to the working level of
the pier and mixed with dye solution in a jet mixer.  The dye-water mix-
ture was then sent through the 3A-inch hose to a perforated spray hose
within the turbulent mixing zone of whichever aerator was operating dur-
ing the test.

For the mechanical aerator, the perforated hose was extended around a
horizontal triangular pattern, just above the cascade created by the
aerator.  For the diffuser aerator, the perforated hose was underwater,
parallel to the two 8-inch headers, just between and about 1 to 2 feet
above them.

During each test, one of the aerators and the dye solution pumps ran
continuously for several hours.  Boats equipped with samplers and photo-
electric colorimeters moved through the dispersion  zone at various dis-
tances downstream from the aerator, measuring and plotting the dye
concentration vs time at many known points.  Measurements continued for
one  to two hours after the dye pumps  and the aerator were shut down.   In
this way the dispersion pattern  of the  oxygenated water could be plotted
much more accurately  than would  be possible by  means of oxygen analysis.

For  several reasons the dye solution  is better  than oxygen in this type
of study.  Dye  analysis is quicker, more sensitive, and more accurate
than oxygen analysis.  There is  only  one source  of  dye, at a known loca-
tion, and the dye concentration  is not affected by  the atmosphere or by
variations in biochemical oxygen demand.

The  dye dispersion studies were  made  possible by the cooperation of the
U.S. Geological Survey of Trenton, N.J., under the  leadership of Mr.
John McCall.  His staff worked closely with our project staff for sever-
al days to achieve results which constitute an unusually  complete study
of the dispersion pattern of the oxygenated water.
                                      20

-------
                           TABIE 1
TABULATION OF GAS SAMPLES
Taken July 2, 1970 in Philadelphia
Sample
No.
1
2
3
U
6
7
8
9
10
11
12
13
s
16
17
18
19
20
M
il
Depth of
Header
12 '3"
12 13"
12 3"
12 '3"
12  3"
12 '3"
38'3"
38 '3"
38'3"
38'3"
38'3"
38J3"
25'
25'
25'
25'
25'
25'
25'
Air Intake
Corrected
Flow Rate
SCFM Air
93-0
93.0
kk'O
UU.o
16.3
16. U
19.7
19.7
53.7
53.5
QQ J.
QQ Jj
17.9
17.9
U9.6
U9.6
9U.1
92.9
Samples
% 02 in
Spent Air
20.8
19.9
20.06
20. U
19.9
19.91
20.7
19.07
19.6
18.9
19.3
19.7
20.1
19.8
19.6
19.7
19.7
20.3
19. U
21.1
21.0
20.9
20.83
% 02
Absorbed
1
6.5
5.6
3.6
6.5
6.5
1.8
11.3
8.2
s-12.3
10.0
7.7
5.3
7.1
8.2
7.7
7.7
U.2
6.0
mm
N^     purged with N2  15 seconds  ....    *3r0.7
No     Purged with N2  30 seconds	     0.0
                              21

-------
                                SECTION III

                        OXYGEN TRANSFER OF AERATORS
In this section the data on oxygen transfer for both diffuser and surface
aerators are presented and the results discussed.  Also comparisons are
made with oxygen transfer efficiencies obtained from the same equipment
on the Passaic River and other results reported in the literature.

Oxygen Transfer Rates

In the Passaic River aeration report (l) it was demonstrated that in a
natural stream the rate of oxygen transfer for an aerator under steady-
state conditions may be expressed as

                            R  m .22U6 Q (Cd-Cu)                  (>1)
                             1         P
in which

      Rt = oxygen transfer rate in pounds per hour per unit horsepower
       Q = river discharge in cubic feet per second
      Cd = DO concentration downstream of the aerator in milligrams per
           liter
      GU = DO concentration upstream of the aerator in milligrams per
           liter
       P = aerator power consumption in shaft horsepower

Equation 3-1 implies that the amount of oxygen supplied by the aerator
per hour is proportional to the difference between the concentrations of
oxygen downstream and upstream of the aerator and to the river discharge.
Also in this equation the contributions from other oxygen sources or
sinks, such as BOD consumption and atmospheric reaeration, are considered
negligible.

The applicability of Equation 3-1 depends mainly on the following assump-
tions :

      (1)  the dissolved oxygen concentrations within the upstream and
downstream sampling cross-sections are uniform, that is, there is no
significant DO concentration gradient across each of the river sections;
and

      (2)  the capacity of the aerator is such that the entire downstream
cross-section is affected by the aerator at the time DO samples are taken.

During the Passaic River aeration study, the aerators were placed in a
relatively narrow (average width 100 feet) and shallow  (average depth 7
feet near the aeration site during low flows) river.  The above two con-
ditions were generally satisfied (l) and Equation 3-1 was used to compute
oxygen transfer rates for the aerators.  For the Delaware River tests,
                                     23

-------
however, the aerators were placed near the Camden side of the river,
approximately 200 feet from the bank, as shown in Figure 3 in Section II.
The width of the river at the aeration site is about half a mile, and the
depth of water near the aerator about 30 feet.  It is evident that only
a small fraction of the river flow was affected by the aerator when DO
samples were taken; also the DO concentrations were not uniform in the
vicinity of the aerator.  Hence in such cases the conditions for using
Equation 3-1 are not satisfied.  Therefore the following method derived
from a mass balance concept is adopted as an approximation.

It is assumed that a "control section" could be located downstream of the
aerator through which all or most of the aerated water passes.  This con-^
trol section can be determined by observing the flow pattern in the vicini
ty of the aerator in operation.  For a small area,   AAj_, the DO' concentra
tions may be considered uniform and represented by C^.  If the velocity  of
flow for the small area,   AA-^is Vi, then the mass rate of oxygen passing
through the area is CiVi  A AJ..  The total mass rate for the control sec-
tion, M, is then
                           M = Ei CiVi AA                   (3-2)
 Let Ci be the DO concentration before and C'i after  the  aerator was  in
 operation.  The oxygen uptake rate due to the aerator, U,  can be  computed
 by

                        U =  Ii(C'i - Ci)Vi AA               (3-3)

 under the assumption that there is no significant velocity change at
 the control section before and after the aerator is in operation.

 The oxygen transfer rate for the aerator under test conditions, Rt,  can
 then be determined by

                           Rt = U/P                            (3-U)

 Mechanisms of Oxygen Transfer

 The mechanisms of oxygen transfer for the diffuser and the mechanical
 aerators are different.  As shown by Bewtra and Nicholas (9), in the case
 of diffuser aeration the total rate of oxygen transfer constitutes the
 following three components, namely, bubble formation, bubble ascent, and
 bubble breaking at the water surface.  For porous diffusers with small
 capillary openings and a small rate of air flow, the oxygen transfer
 during bubble formation is appreciable.  However, for diffusers with large
 apertures, such as the one tested in this study, bubbles are formed while
 rising in the water and the amount of oxygen transfer during bubble forma-
 tion is negligible.  The aeration during bubble bursting at the water
 surface is related mainly to turbulence conditions.  Although Carver  (10)
 has shown that in conventional aeration tanks the amount of oxygen trans-
 fer in this phase of bubble aeration is small, other investigators, e.g.,
 Barnhart  (11), have observed that a substantial amount of oxygen is
 transferred at highly turbulent aeration tank surfaces.  Therefore for
                                       2U

-------
duffusers with large apertures the total rate of oxygen transfer, N, can
be written as the sum of two individual rates; namely, Na, the rate during
bubble ascent, and N^, the rate during bubble bursting at the surface.

                             N = Na + Nb                      (3-5)

The oxygen transfer during bubble ascent, Na, depends on such factors as
the diffuser submergence, velocity of air bubbles, diameter of air bubbles,
and so forth.  On the other hand, N^ has been expressed as an exponential
function of the water velocity at the surface (12).  It is therefore ex-
pected that for a diffuser the efficiency of oxygen transfer would be
proportional to the depth of submergence, the size of the bubbles, the
air flow rate, and the surface turbulence, as described by Eckenfelder
(13) and Eckenfelder and Ford (lU).

In the case of the mechanical aerator, oxygen transfer to the water occurs
both in the sprayed water and in the turbulent mixing zone (lh).  Carver
(10) has investigated the process of spray aeration and found that oxygen
transfer was correlated to the size of water droplets.  Garland  (l) re-
ported that oxygen transfer efficiency for surface entrainment aerators
generally decreases with increasing aeration basin volume, but increases
with increasing water depth for basins of constant diameter.  In an
earlier study by Kaplovsky, et al  (16), oxygen transfer was found to in-
crease markedly in a certain velocity region.  To date (1970), however,
no information is available concerning the independent evaluation of the
relative importance of each of the aforementioned two phases of oxygen
transfer, i.e., the spray aeration and the turbulent mixing.

Procedures of Computation

As indicated in Section II, it was decided to operate the aerators and
collect data only during the periods of outgoing tide in an attempt to
minimize the mixing effects of the tidal currents on the DO distribution.
For each test run, numerous samples were taken both upstream and down-
stream of the aerator.  Details of the sampling procedures as well as
the locations of the sampling sections are described in Section II.
Figure 10 illustrates the results  of a typical DO and velocity traverse.

To compute the oxygen uptake due to the aerator, the DO readings were
first corrected against the Winkler readings.  The average DO for a
cross-section was then obtained by weighing the DO readings according
to the velocities, and the uptake was determined by Equation 3-3.  It
should be noted that for several tests no DO data were taken at the
downstream "control" section before the aerator was in operation.  In
these cases, the DO concentrations at the upstream sampling section were
taken as approximations of the downstream readings.

As mentioned earlier, the downstream "control" section was located by
examining the flow pattern of the oxygenated water in the vicinity of the
aerator.  To accomplish this, dye tests and floating object tests were
conducted so that the paths of the aerated water could be determined.

-------
                                                          0.8
 
                                                                      O D.O. tng/1
                                                                      + Velocity FPS
                                     125         100         75          50          25
                                       Distance, ft.,  across upstream sampling  section.


                              FIG. 10   DO AND VELOCITY TRAVERSE AT UPSTREAM SAMPLING SECTION

                                                           9/5/69

-------
These tests revealed that in general the aerated flows downstream of the
aerator were divided.  For the mechanical aerator, as shown in Figure 11,
part of the flow passes riverward of pile BW, while the balance seems to
pass landward of pile BE, leaving a section between them without aerated
water.  For the diffuser aerator the flow pattern is generally similar,
with only a small portion of the flow also passing the section between
piles B and BE and some aerated flow going upstream and then turning
back to downstream, as indicated in Figure 12.  More detailed description
of the oxygenated flow patterns is given in Section IV.

For the Philadelphia tests, only air samples were taken for the determina-
tion of percentage of oxygen absorption, as described in Section II.  The
oxygen uptake rates for these tests were computed by knowing the supply
air flow rate, oxygen content in the air, percent absorption, and density
of oxygen.  The results so obtained in the Philadelphia tests for the
diffuser with four nozzles were adjusted to compare with those for the
manifold of 160 nozzles, at both Camden and the Passaic River, on the
assumption that the oxygen transfer per nozzle remained the same.  Equa-
tion 3-U was then used to compute the oxygen transfer rates under test
conditions.  It should be mentioned that for the Philadelphia tests the
power consumptions were obtained by extrapolating the manufacturer's
calibration curves to the greater depth of submergence experienced during
the experiments.

The oxygen transfer rates, R^, were reduced to those under standard con-
ditions, Rs, by the following relationship so that comparisons can be
made with results elsewhere:
                                                                   (3-6)
 in which

       (CS)2Q = sautration DO  concentration at 20C
       (Cs)ij,  = saturation DO  concentration at test  water  temperature,  T
           P = pressure in  inches  of  mercury
           8 = "specific solubility"  of  oxygen
           Cm = DO concentration  at the aerator
           TF = temperature  correction factor =  (
           a = "specific oxygen  transfer rate"
           F = conversion factor

 The  saturation oxygen  concentration,  Cs, was computed by  the SED/ASCE
 Equation  (1).  For  the diffuser  aerator  Cs was  corrected  for pressure,
 which in  turn was evaluated at the mid-point of the diffuser submergence.
 According to laboratory tests not  yet reported  on elsewhere, both a and
  j  values were found  to be 1.0.  The analysis  is as follows:
Rs = Rt x F
F -

" 29.92 '
(Cs )20
)( 3 ) - Cn

J (TF)( a )
                                      27

-------
                           i      J      J      J      I
                                                Effluent

 2.0
                            I      I      I      I
                                1.3  1-3
                            	X	X	XX	X	X	O
                                 1.2  1.1  1.1 AE
                                               //
100
:
100
      1.9  1.9  3-2    I 2.3   2.1  / I
	XX XX X_XO * *XCf 2.3<
         1.8  1.9 |2DBW   2.1  BE '



           x D.O. reading,  mg/1
            ft.
               FIG. 11  TYPICAL FLOW PATTERN  IN THE VICINITY
                               OF MECHANICAL AERATOR
                                  28

-------
                         3-9
                        3.1
                         >
U.I

  X
 3-8
 
3-6
 X
            -;
            ft
 100          0

   i i I l l  I I I I I I
              Ft.
                              IMM
 100
IZI
                                                    Effluent


                                                        oo
                                                           ^x  ^x

            ///1
                                              /
           ,  Ll  la 3-9  3-9 3-93.9U.;
              x	x	i	*-*-Q-*	*

                  I   l   I i  B
          \      V
                                                               BE
   x D.O. reading mg/1
                FIG.12   TYPICAL FLOW PATTERN IN THE VICINITY

                                OF DIFFUSER AERATOR
                                         29

-------
Many environmental factors influence the reaeration characteristics of
surface waters.  One of the most general of these is the quality of the
water itself.  That is, the possible presence of constituents that will
influence reaeration either through effects on the rate constant (trans-
fer characteristics) or the saturation value above and beyond well de-
fined influence of temperature and salinity.

To either substantiate or remove such possibilities, studies were carried
out to determine any influences on either the saturation value or reaera-
tion constants.  Effects on the saturation value (Cs) were evaluated by
deaerating six cylinders with nitrogen and measuring the dissolved oxygen
levels after various time periods until the level remained constant.  For
19C, the DO levels ranged from 9.23 to 9.26 with an average of 9-25-
This places the saturation value only slightly below that noted in
Standard Methods for the Examination of Water and Waste Water.

Three systems were employed to evaluate effects on the rate of reaeration.
These were slow stirred cylinders, high-speed mixing giving considerable
turbulence and bubble aeration.  In each case, the rate of aeration for
river water was compared to that for New Brunswick tap water.  For the
slow stirred cylinders, the rates of reaeration of tap and river water
were identified.  For both the high turbulence mechanical and the bubble
aerators, the river water values were 99$ of the tap water values.  As
these latter two areas are directly analogous to actual methods employed
in river aeration, it can be concluded that the rate of reaeration is not
influenced by  constituents present in the water at their actual concentra-
tion in the  aquatic environment.

Mechanical Aerator Results

Due to the late start  of tests and the storm accident described earlier
in Section II, only four complete sets of observations were obtained.
The field data, which include dates, water temperature, oxygen concentra-
tions and uptake, and power consumptions are listed in Table 2.  The
results on oxygen transfer rates, both under test and standard conditions,
are given in Table 3.  Also Figure 13 illustrates the oxygen increase
resulting from the aerator operation for the test on September 8, 1969.
The stationing shown on the figure refers to arbitrarily chosen sampling
section.  It can be seen that the maximum oxygen increase at the down-
stream control section was close to 1.0 mg/1, and the net increase in
oxygen content is well indicated.  The upstream variation of DO concen-
tration does not seem to be significant.

The computed oxygen transfer rates show a large variation in the effi-
ciencies, probably partly due to approximations inherent in the methods
of measurerasnt.  The transfer rates varied from 1.18 Ibs 02/hp-hr to
3.78 Ibs 02/hp-hr with an average of 2.56 Ibs 02/hp-hr.  When converted
to standard  conditions, the range became 1.29 Ibs 02/hp-hr to k.%0 02/hp-hr
and the average 3.06 Ibs 02/hp-hr.

The Delaware results indicated substantially higher average transfer
rates for the  mechanical aerator than the Passaic River tests, both for
                                     30

-------
                          TABIE 2
      SUMMARY OF F3EID TEST DATA FOR MECHANICAL AERATOR
Water Temp.
Date C
8/21/69
8/25
9/2
9/8
25.0
27.7
27. U
26.8
d.o. concentrations
Be fere (ttig/1) After
1.23
0.95
1.79
0.78
2. Hi
1.13
2.02
1.02
Uptake
Ibs 02/hr
377-1
172.1;
351.7
118.9
Power
Shaft hp
99.8
98.0
100.0
101. U
                          TABIE 3




    SUMMARY OF OXYGEN TRANSFER RESULTS, MECHANICAL AERATOR
Date
8/21/69
8/25
9/2
9/8

Uptake
Ibs 02/hr
377.1
172 .U
351.7
118.9

Rt
Oxygen transfer rate
under test conditions
Ibs 02/hp hr
3-78
1.76
3-52
1.17

RS
Oxygen transfer rate
under standard condi-
tions, Ibs 02/hp hr

Ave.
U.50
1.9U
U.U9
1.29
3.06
                          TABLE h




SUMMARY OF FIELD TEST DATA AT CAMDEN, N.J. DIFFUSER AERATOR
Water Temp.
Date C
8/1/69
8/7
8/8
8/18
8/22
8/27
9/5
23.1
2U.2
2U.5
26.2
26.0
27.0
27.U
d.o. concentrations
Before (mg/1) After
U.52
3.67
3.13
1.U7
0.75
1.51
0.63
U.6U
U.07
3.35
1.82
0.82
1.92
0.92
Uptake
Ibs 02/hr
103.5
90.5
118.9
105.U
109.2
1U1.9
lU6.it
Brake
hp
108.0
83.5
92.5
76.5
90.0
9U.O
77.0
                             31

-------
               A aerator off
               x aerator on
           200           150100         50          0

             Distance,  ft, across upstream sampling section

                           Upstream Section
   1.50
 .  1.0
o
o
   o.5
          &  aerator  off
          x  aerator  on
          O aerator  off (interpolated)
                                                           X _
        0    50          100         150         200         250

           Distance, ft, across downstream sampling section


                           Downstream Section


              FIG. 13    OXYGEN INCREASE RESULTING TROM

                         MECHANICAL AERATOR OPERATION

                                    9/8/69

-------
field conditions (2.6 Ibs C>2/hp-br compared to l.OU Ibs 02/hp-hr) and
under standard conditions (3-06 Ibs 02/hp-hr compared to 2.12 Ibs 02/hp-
hr).

Compared with most of the transfer rates reported in the literature, even
the results for the Delaware are a little low.  Eckenfelder and Ford (lU)
reported a transfer rate between 3.2 and 3.8 Ibs 02/hp-hr in activated
sludge plants and aerated lagoons5 Susag, et al (17), reported a 1; Ibs
02/hp-hr in laboratory channels; Cleary  (18) an 11.05 Ibs 02/hp-hr and
Kaplovsky  (16) a range of 1.5 to h.% Ibs 02/hp-hr in the Chicago canal;
all referred to standard conditions.  However, the average uptake rate of
210 Ibs 02/hp-hr under standard conditions on the Delaware is higher than
the 180 Ibs 02/hp-hr reported by Imhoff  (19) on the lower Ruhr River in
West Germany.

In summary, the Delaware results seem to fall within the range of values
reported in the literature, the difficulties of measurement and number
of tests were such that the determination is not considered conclusive.

Diffuser Aerator Results
Compared with the Passaic River study, much more data were obtained for
the diffuser aerator on the Delaware River.  As described earlier in
Section II, the diffuser aerator was tested under various depths of sub-
mergence ranging from 11.0 feet to 16.9 feet at Camden, N.J., and a short
replica of the diffuser was tested later at Philadelphia under depths of
12.3 feet, 25 feet, and 38.3 feet.  A summary of the Camden tests data,
including date, water temperature, DO concentrations, power consumption,
and uptake rates is given in Table U, and the Philadelphia data are listed
in Table 5-

The oxygen increase due to diffuser aeration is illustrated in Figure ll;,
using data of September 5, 1969.  It can be seen that the DO increase is
approximately uniform along the vertical direction, but less so horizon-
tally.

The oxygen transfer results for the diffuser aerator are given in Table
6, together with some Passaic River results for comparison purposes.

The observed power consumption data for the Camden  tests and for the
Passaic River tests are plotted against the diffuser submergence for
various engine speeds, as shown in Figure 15.  It is observed that for a
certain engine speed, the brake horsepower varies linearly with the water
depth.  Consequently, power consumptions for greater depth could be esti-
mated by extrapolation, as was carried out in this  study.

Upon examining the results for diffuser aerators given in Table 6, the
following are noted:  (1)  The amount of oxygen absorption increases
substantially as the depth of submergence increases.  Under standard con-
ditions, for an average depth of 7.2 feet, the oxygen absorption averages
3.1 percent, as experienced on the Passaic River.   On the Delaware, the
absorption increases to 5.0 percent for an average  depth of 13-2 feet; to
                                       33

-------
                          TABLE 5

SUMMARY OF FIELD TEST DATA AT PHILADELPHIA* DIFFUSER AERATOR
         July 2, 1970         Water Temp. = 25.5C
Diffuser sub-
mergence, ft.
12.3
12.3
25.0
25.0
38.3
38.3
Air flow rate
SCFM/Nozzle
11.0
11.0
12.U
12. k
13.U
13- U
Percent
Absorption
5.6
3.6
8.2
7.7
8.2
12.3
Uptake
Ibs 02 /hr
102.7
66.0
169-6
159.2
183.2
27U.8
Brake
hp
65
65
11*0
11*0
230
232
  Include only the portion of the data for which the air flow
  rates are comparable with the Passaic and Camden tests.
                          TABLE 6

     SUMMARY  OF  OXYGEN TRANSFER RESULTS, DIFFUSER  AERATOR
Study

Passaic



Del.
at
Camden,
N.J.



Del.
at
Phil.,
Pa.

Depth
of
submer-
gence,
ft.
7.0
7.7
7.9
6.5
11.0
11.2
12.8
Hi. 2
111. 2
lU.il
16.9
12.3
12.3
25.0
25.0
38.3
38.3
Air
flow
rate-
SCFM/
Nozzle
9.1
15.8
17.1
12.6
15.U
13.6
U*. 3
11*. 0
11.7
U*.7
15.9
11.0
11.0
12. U
12. li
13. u
13- U
Percent
absorption
standard
conditions
3.8
2.7
3.2
2.6
U.I
5-U
5-0
6.0
5.0
5.U
5-3
5.5
3.5
6.5
6.1
8.1*
5.6
Rt
Oxygen
transfer rate
test conditions
Ibs 02/hp-hr
1.33
0.92
0.95
0.95
1.35
1.21
1.08
1.29
0.91*
1.23
0.96
1.58
1.02
1.21
1.11*
1.18
0.80
RS
Oxygen transfer
rate, stand-
ard conditions
Ibs 02/hp-hr
1.1*1*
1.16
1.50
1.05
1.38
1-59
1.1*3
1.51
1.08
1.1*1
1.30
1.55
0.99
0.96
0.90
0.81
0.5U
                                 31*

-------
                                                                     DO mg/1
                                                             200   1    2    3
Data of September
FIG. Hi
                                     CROSS -SECTIONAL DISTRIBUTION OF
                                     OXYGEN INCREASE DUE TO AERATION

-------
i
o
                    20
                 .
                 -P
                 ; .
                 JJ
                 i
                 g  10
                                                                        o 1969 Delaware River Data

                                                                        A 1968 Passaic River Data
                                  20
UO          60         80


   Compressor Brake Horsepower,  Hp.
                                                                                             120
1UO
                                        FIG. 1^   COMPRESSOR BRAKE HORSEPOWER VS SUBMERGENCE

                                                     DEPTH FOR DIFFERENT ENGINE SPEED

-------
6.3 percent for a depth of 25 feet; and to 7-0 percent for a depth of
38.3 feet.  (2)  The variation in oxygen uptake with depth of submergence
can be summarized as follows:  68.8 Ibs 02/hr at 1.2 feet; 113.6 Ibs
Oo/hr at 13.2 feet; 130.3 Ibs 02/hr at 25 feet; and 156.k Ibs 02/hr at
38.3 feet, all under standard conditions.  (3)  The power consumption
varies linearly with the depth of submergence, with a range of 52 hp at
a depth of 6.5 feet, to 232 hp at 38.3 feet.  (lj)  The oxygen transfer
rate, Rs, in pounds of oxygen per hp-hr seems to increase somewhat be-
tween depths 7 and 15 feet, but decreases materially with further in-
crease in depth.  The average transfer rates at various depths are:
1.29 Ibs Op/hp-hr at 7.2 feet; 1.36 Ibs 02/hp-hr at 13.2 feet; 0.93
02/hp-hr at 25 feet, and 0.68 Ibs 02/hp-hr at 38.3 feet.  Figure 16
illustrates this variation with the depth of submergence.

The oxygen absorption results for diffusers also were compared with
results of other studies.  Bewtra and Nicholas  (9) reported aeration
data for both the saran tube and the sparger type diffusers operating in
aeration tanks.  Their data, for air flow rates ranging between 10 and
15 SCFM per unit, are plotted on Figure 17 against diffuser submergence.
The relationship between percent absorption and depth appears to be
linear on the log-log paper.  For the saran tubes, which have very fine
bubbles,  the absorption reaches 11$ at a depth  of 12 feet, while for the
spargers, with  larger openings, the absorption  is 7-5$ for the same
depth.  Also Imhoff  (19) used hoses with very fine orifices  (between
0.5 and 0.7  mm  in diameter)  on the lower Ruhr River in West Germany and
obtained  high absorption rates for depths of 8, 16, and 20 feet.  These
results are  also plotted  on  Figure 17.  To compare the above findings
with results of the  present  study, the Passaic  and the Delaware aera-
tion results are plotted  on  the same figure, and an eye-fit line is
drawn to represent  the approximate relationship.  It is noted that the
absorption rates for the  Passaic and the Delaware studies are lower
than both of the aforementioned two sets of results  (see Figure 17).
This may be due partly to the larger bubble sizes as a result of the
larger openings of  the nozzles  (5/32 inch) used in the present study,
and the fact that  most of the tests were conducted in tanks, which
would materially reduce the  upward velocity of  the bubbles.

According to Eckenfelder  (13), and Bewtra and Nicholas  (9), oxygen ab-
sorption  is also correlated  with the supply air flow rate.  However,
no apparent relationship  was found with  the present data.  It may be
possible that, although the  rate of absorption  increases with increas-
ing air flow rate,  the velocity of the rising bubbles may also increase,
resulting in a shorter detention time, and hence less absorption.

Conclusions

     1)  For the mechanical  aerator, an  average oxygen transfer rate of
3.06 Ibs 02/hp-hr was obtained on the Delaware River, while for the
Passaic River a 2.12 Ibs  02/hp-hr was  obtained.  The Delaware results,
however, should not be considered conclusive due to certain difficulties
in taking measurements and small number of experiments.
                                    37

-------
              2.0
CO
                                                                   A   1968  PASSAIC DATA
                                                                   +   1969  DELAWARE DATA
                                                                      1970  DELAWARE DATA
                                                     Depth of Submergence, ft.
                                FIG. 16     OXYGEN  TRANSFER  RATES VS  DIFFUSER SUBMERGENCE

-------
c
3
H
-
:
c
'.
-
C
a
 j
-
-
,0
10
5
c
4
3 .
0


- -0






/
Y
/


 Imhoff
-Q- Saran tubes"! Bewtra &
-Q Spargers J Nicholas
^ Rutgers, Passaic
Q Rutgers, Delaware
(gas samples)
-f Rutgers, Delaware
(mass balance)





J

!>
/






s
/

/

/
A





/

/
/
-
S*
A





!



4
'




y


y
/

X




/


/

4



J
/


/

rf
.



3-


&

3
X
D









'









1




















s







*










T
vT
'






































X





























1









X


















r

















o


o




                      8   9  1O               2O


                       Diffuser Submergence, Ft.
30
40
       FIG. 17   OXYGEN ABSORPTION VS DIFFUSER SUBMERGENCE

                         COMPARISON OF RESULTS
                                39

-------
     2)  For the diffuser aerator, the percent oxygen absorption was
found to increase significantly with increasing depth of submergence,
though not as rapidly as results reported elsewhere for aeration tanks
with fine bubble diffusers.  The average oxygen absorption increases
from 3.8 percent at a depth of 10 feet to 7.0 percent at a depth of
38.3 feet.

     3)  The oxygen transfer rate for the diffuser aerator was found to
decrease generally with increasing water depth, though a slight increase
in efficiency appeared to occur between depths of 7 and 16 feet.  The
transfer rates average about l.U Ibs 02/hp-nr at 10 feet depth and about
0.65 Ibs 02/hp-hr at UO feet depth.

     U)  The diffuser results indicated generally lower absorption rates
than most of those reported in the literature, probably due mainly to
the larger bubble sizes, and the free circulation provided by the river.

-------
                               SECTION IV

                           DISPERSION ANALYSIS
 Introduction

 The purpose of the dispersion test was to determine quantitatively the
 geometry of the aerated plume spreading downstream from the aerator.
 The importance of this plume geometry becomes apparent when several
 aerators are to be installed in the same river.  Optimal placement of
 these  aerators is dependent upon the dispersion characteristics of the
 individual aerators.  Consequently, to avoid repetitious reaeration of
 the same slug of river water, it is appropriate to know the aerator
 dispersion characteristics.

 Aeration installations may be subdivided into two categories:  (a) aera-
 tion from a line source and (b) from a point source.  In the first
 group  would exist such schemes as bank-to-bank aeration, wherein the
 entire width of a stream is aerated.  Here, at least initially, the dis-
 persion of the plume is in two spatial dimensions, i.e., the vertical
 or with depth and the streamwise direction.  The second aeration cate-
 gory is exemplified by a single, relatively small aerator unit installed
 in a much larger river.  In this case, the initial stages of dispersion
 take place spatially in three dimensions, i.e., in the vertical, in the
 stream direction and in the cross-stream or lateral direction.  When,
 in both of the above instances, the dispersion phenomenon has succeeded
 in spreading throughout one or more spatial directions, the stages that
 follow take place in the remaining dimensions.  That is, in the first
 case of line source aeration when the aerated plume has spread through-
 out the vertical so as to permeate uniformly the entire depth of a
 stream, the remaining stages of dispersion occur only in the stream
 direction.  This then diminishes the concentration with downstream dis-
 tance  only, all extraneous effects neglected.  In the second case, when
 the aerated plume has spread throughout the depth, the dispersion pro-
 cess occurs in both the streamwise and cross-stream directions.  After
 the cross-stream mixing has been completed, the dispersion process then
 takes  place only in the stream direction.

 The dispersion phenomenon is further complicated in the following manner.
 Immediately after the injection of a dispersing quantity into a stream,
 the spreading process is dominated by what is termed convective trans-
 port phenomena.  During this convective period the concentration distri-
butions take on a conspicuous skewness that will be discussed further in
 the following.  This skewness, which initially is quite severe, diminishes
until  at the end of the convective regime the symmetrical distributions
 pertaining to one-dimensional diffusion prevail.  Once the concentration
distributions attain this symmetrical character, their behavior is pre-
dicted by the one-dimensional dispersion dispersion model.  The differ-
ent dispersion regimes are illustrated as follows.  Consider a stream in
which  a vertical cross-sectional plane of water oriented perpendicular
 to the stream direction is instantaneously colored red.  Immediately

-------
after this injection, the one-dimensional concentration distribution
exhibits a decided skewness as shown by Figure 18.  In (a) the view is
that of a vertical plane parallel to the stream direction; in_(b) is
pictured the corresponding concentration distributions where c is the
dye concentration averaged over the cross section of the stream.  Ul-
timately, the concentration distribution exhibits the gaussian shapes
predicted by the simple model.  Until this occurs, however, the dye
cloud spreading phenomenon is convective, and following this interval
it becomes dispersive.  It is therefore important to arrange the data
taking procedures to record the symmetrical gaussian concentration dis-
tributions .

Several schemes (see references) have been devised through which the
longitudinal dispersion coefficient may be determined.  The one employed
in the present study is the change of moment method.  The results of ex-
periment will determine both quantitatively and qualitatively the dis-
persion characteristics downstream from aerators in river conditions
similar to those encountered in the Delaware River.

Theory for One Dimensional Dispersion Model

The three-dimensional, unsteady conservation -of- mass equation for the
dispersion of a conserved quantity will be developed.  This  is  found by
decomposing the velocity field and the concentration  of  our  conserved
quantity  into a time averaged and the fluctuating quantity.  Assuming
conservation of mass for the mean velocity field, we  obtain, in Cartesian
form  (see Daily & Harlemann, 1966)
                                                      + D
                                                          ^   ^   ^

where u, v, w are the usual time averaged velocities.  The quantity c
is the time-averaged concentration and D^B is a molecular dispersion
coefficient that is assumed uniform and isotropic.  Invoking the Fickian
dispersion relations :
                              = - E^
                                   yay                            (2)

-------
                                    -A"*-
fc-t
            velocity
           profile _-
C (x,  t)
    initial
                          (A)
                         t - tr
       Area
                             Area
                                                  Area  A.
      convective  region    >
                                  dispersive  region
                           (B)
FIG. 18   (A)  SKETCH  OF  DYE  PATTERNS  IN  A  STREAM
         (B)  PLOT  OF  ONE- DIMENSIONAL  UNSTEADY  DYE
             CONCENTRATIONS  VS  DOWNSTREAM  DISTANCE

-------
the above equation, dropping the molecular dispersion terms which are
dominated by their turbulent dispersion counterparts, becomes
When v = w = 0 and the turbulent dispersion is characterized solely by
the uniform coefficient, Ex, we obtain our one-dimensional model

                      M + 3 as = E
                      at     sx

Boundary conditions for the above equation depend upon the type of in-
jection of the conserved quantity.  If, as in the case of the dispersion
tests conducted in the present work, the quantity Rhodamine B dye is
steadily injected at a prescribed rate the conditions are
                   ,+co             co
                     c(x,t)dx = [  c(x,o)dx = ~
                  c(,t) = 0; t ^ 0                                   (5)


                  0       = -^
                  ^-inject   dt

Solutions  to the  one-dimensional dispersion  model,  satisfying the  above
boundary conditions  are  given by:
                     a.  .       ,         -[
                                 _ 1__  4Ex(t-T)    .                  (6)
                                        e            dT

 Experimental determination of the  coefficient EX via the method of
 moments (see H.B.  Fischer, 1966) utilizes  the transformation:

                           ? = x - ut                                 (7)

 Substituting this  into equation (1)  and taking  the partial derivative
 with respect to time,  we get
                          2-       3-

                           2    x   ,-2

 Multiplying by ^  and integrating  over  all values of  ^  produces



               -
-------
Defining the variance by
                        x> -

                        4!	                                   do
we obtain the following relationship for EX:


                           - 1 A T^2
                           ~ 2	-


Equation (11) may be transformed
Ex = i^-^-                                (12)
         &
                 9   tJ0   dt       2                                /   %
               a 2 s JL^	^                                (,3)
                t    !-ddt
                     J0dt
where t is the mean time:
                    J0dt

The longitudinal dispersion coefficient, EX will be determined using
the relations (12) through (Hi).

Longitudinal Dispersion

The concentration distributions moving through two stream stations  are
presented in Figure 19.  Differentiating these distributions  produces
the curves shown in Figure 20.  These curves pertain only to  the  diffu-
ser aerator, as  similar data for the mechanical aerator were  found  to
be incomplete due to equipment failures.

Using a computerized moment method, the value of EX for the diffuser
aerator found between the two stream stations in the Delaware River is
8.5 x 10 ft^/min.  While it is true that this value is characteristic
of the aerator in the reach of the river under consideration, it  is
felt that it constitutes a decent first approximation to other reaches
in similar rivers.

It is significant to note from Figure 19 and 20 that the concentration
levels at the front of the dye cloud move downstream ahead of the mean
stream velocity  of the cloud.  When the dye cloud is assumed  to approach
the behavior of  an aerated plume, it is concluded that the dissolved
oxygen level behaves similarly.  Hence, an amount of river greater  than
the mean slug quantity has experienced an increased DO.

-------
    -  50
    - 40
PQ
a

.2
V

8
o
    -  30
- 20
    - 10
                                                          Time  after  injection,  min
                                                                       i
                         40
                                            80
120
160
          FIG. 19  DYE  CLOUD  FRONTAL  STRUCTURE  VS TIME  AFTER  INJECTION  AT



                                    TWO  STREAM  STATIONS

-------
 -  3
  -  1
Time after injection, min.
                        40
80
                                                                     120
            FIG. 20   DIFFERENTIATED  DYE  CLOUD   FRONTAL  STRUCTURE  VS  TIME
                            AFTER  INJECTION  AT   TWO  STREAM  STATIONS
                                                                                            160

-------
Considering travel times, specifically, Figure 19 shows that the leading
edge precedes the mean of the front by U0$ at the upstream station.  At
the station downstream, the travel time of the leading edge is approxi-
mately 2$% of the mean travel time.

The magnitude of the longitudinal dispersion coefficient is significant.
Previous values available in the literature (Yotsukura, et al, 1970)
(32) indicate that naturally a value of 9-6 x 1CP ft2/min., which is the
largest yet determined, has been measured in the Missouri River flowing
at 33,000 cfs.  A value for the Delaware estuary of approximately
160,000 ft2/min has been determined by Paulson  (1969)  (28).

That the longitudinal dispersion coefficient determined in the present
study is an order of magnitude larger than that which  occurs naturally
in similar rivers becomes more apparent from the photographs, shown
elsewhere in this volume, of the turbulent mixing visible at the river
surface over this submerged aerator.  It is to be noted that the dif-
fuser aerator  installed  in the present Delaware test was oriented  at
an angle of about l& with the New Jersey shore.  In this configuration
it presented a projected length of only kO feet normal to the  flow of
the river.  The speculation is therefore put  forth that were it oriented
normal  to  the  river flow, the longitudinal dispersion  phenomena would
have been  even more pronounced.  Such  a violent turbulent  mixing is of
course  associated with the large air bubbles  emitted by the diffuser
nozzles.   A  small bubble diffuser, therefore,  might not give rise  to
dispersion  coefficients of  the size  presently determined.  However, the
smaller bubbles would rise more slowly to  the surface  and  hence would
enable  an  increased DO to be  attained  by  the  affected  river water.  Ob-
viously, both bubble  size,  oxygen  transfer,  and dispersion characteristics
have  to be considered together  in  the  final  analysis  of river  aeration.

Transverse Dispersion

Because it is important  for  the optimal installation of multiple aerators
in the  same river,  an estimate is  obtained for the extent  of  transverse
mixing  downstream of the diffuser aerator.  With the collected data^for
the dye cloud geometry,  a curve-fitting technique is  employed wherein  the
 cross-stream dispersion  is predicted as a function of downstream distance.
However, based upon visual observation of the spreading dye,  the lateral
 dispersion of the two types of aerator may be considered approximately
 the same.

 It is assumed here that  complete vertical mixing has taken place and that
 further mixing occurs only in the  cross-stream direction and longitudin-
 ally.   Furthermore, it is  assumed that cross-stream mixing occurs  mono-
 tonically in the downstream direction.  To insure this monotonicity,  an
 exponential relationship of the following form is assumed:

                       A = a +b(l - e'x) + c(l - e-x)2              (15)

where A and x are the non-dimensionalized, cross-sectional area normal
 to the stream direction, and the downstream direction, respectively.   The

-------
normalizing quantities  are A0  =  12,000 ft2  and x0  =  1000 ft.  The
coefficients a,  b,  and  c  are non-dimensional constants  evaluated using
cross-sectional  areas depicted by the  dye  cloud  at three downstream
stations.  These are shown in  Figure 21.  Numerically,  the  constants
are found to be

                            a  =  0.0666
                            b  =  1.1993
                            c  =  -0.2756

These values enable the calculation of the graphical results presented
in Figure 22. From these results a percentage of reaeration can be
obtained which is hyperbolic in  nature.  Using this  as an estimate one
is able to determine the  amount  of aerated water that is  reaerated
downstream by subsequent  aerators.  Specifically, if two  diffuser  aera-
tors were aligned one  directly upstream of the other at hOO feet  dis-
tance, it may be estimated that  16$ of the aerated water  from the  first
would pass through the  second.  At greater distances the  percentage
would be less.

Conclusions

The results of the dispersion experiment conducted  for the diffuser
aerator in the Delaware River verify the highly dispersive effects that
exist downstream of this type of device.  Quantitatively, the longi-
tudinal dispersion coefficient determined using a one-dimensional moment
method is 8.5 x 10 ft2/min.  This value is  an order of magnitude larger
than the largest "natural" coefficient measured to  date.   It is apparent
that the violent, turbulent,  mixing motions  initiated  in this river by
this type of artificial aeration are responsible for this large value.
In addition, judging by the relatively symmetric concentration distri-
bution shown in Figure 20 at  the upstream  station,  the convective
regime behind this aerator is quite short.   The convective region which
occurs naturally is much longer, and  in Ref.  (32) stated to be of the
order of tens of miles.

In view of this  large  dispersion coefficient found  downstream of this
diffuser it is felt that a smaller bubble  diffuser  might well have a
smaller longitudinal dispersion  coefficient.  However, it is also felt
that the small bubble  diffuser would  inject an  increased amount of
oxygen per hp-hr into  the river.  As  a result,  it is concluded that
future analyses pertaining to varying bubble size include both of these
factors.  Undoubtedly, the violent turbulent mixing motions which exist
downstream of the diffuser aerator are due  to the circulatory fluid
dynamics produced by the rising bubbles interacting with the river
velocities.  Needless  to add, this is an extremely  complicated phenom-
enon to document in detail.   The  present results  pertain specifically
to time and space averaged ramification of these  detailed features.  As
a consequence, the  extrapolation  of the present  to  other situations
should be done with care and  consideration of these factors.

-------
Inj ect ion
 point
         East bank
                                                    East bank

                                             . x /////,
c ross sectional
area 6920 ft. 2
    647
                                6247'   
  |   River width 2600'
                                            cross sectional
                                            area 11, 870 ft.2
                                 r iver width
                                   1900*
  I
  n
                                                                 West bank
        West  bank
        FIG. 21    STREAM  CROSS - SECTIONAL AREA AFFECTED  BY  THE
                    DIFFUSION  AERATOR  VS  DOWNSTREAM  DISTANCE

-------
           - 11
VJT.
           - 10
       to
       3
       O
        o>
        
        bC
        >^
        X
        O
           -  9
           -  8
           -  7
                                Distance Downstream from Diffuser (ft.)
                                                                                             20  -
16 -
                                                                                             12  -
 8  -
                                                                                              4  -
                                                                                                    
-------
Because of the large longitudinal dispersion coefficient, the diffuser
aerator actually supplies dissolved oxygen to a segment of river larger
than the slug described by the mean river velocity.  With regard to the
times of travel at the two stream stations, the leading edge of the dye
cloud exceeds the mean travel time by \&% and 2%%} respectively.

The cross-stream dispersion that occurs downstream of the diffuser aera-
tor is approximated from the available dye data.  Intuitively assuming
a monotonically increasing distribution with downstream direction, the
dependence of aerated cross-sectional area upon distance is

                     A = a + b(l - e-x) + c(l - e-x)2               (16)

where

                     a = 0.0666
                     b = 1.1993
                     c = -0.2756

Using this result the amount of reaeration which  occurs with aerators
placed  in the aerated plumes of upstream aerators  may be predicted using
the results  presented in Figure 22.  The distribution found is hyper-
bolic.   This leads to the conclusion that the time required to increase
the DO  level above that of the ambient at the outer edge of the plume at
the downstream station may now be estimated.  However, this equation
should  not be used to extrapolate dispersion results beyond the reach
actually measured.

To  approximate the distance required for the dye  to spread throughout
the width of the stream, we assume that natural dispersion processes
prevail below the 62U7 foot station.  From reference 23, the downstream
distance estimated for the beginning of the dispersive region  (see
Figure  18)                             9
                          x<  1.8*  
                                   U*Yn

The distance required is x; u is the average velocity; u# is the fric-
tional  velocity; B is the transverse width of spreading; and Yn is_the
average channel depth.  Substituting the values for the Delaware: u" =
1.2? ft/sec; u* = /gYnS  = 1.86 ft/sec where g =  32.2 ft/sec; Yn = hO ft;
S the  slope  determined from discharge, area, and  hydraulic radius; and
B = 1^00 ft, which is transverse distance from the west edge of the plume
to  the  west  shore.  The above equations produce

                     x  <  12.8 miles

This is the  approximate distance from the 62U7 foot station required
for non-tidal natural dispersion to spread the cloud of dye across the
width  of the river.  However, when tidal effects  are incorporated, this
distance decreases.  Through a simple proportion,  the distance required
for the cloud to spread to the center of the river would be x^ = U.5>

-------
miles (see Figure 21).  Both distances, x and xu, are relative to the
62i|? foot stream station.  As a result, the distance downstream  of  the
aerator  required for the dye cloud to reach the west shore is lli.O  miles
assuming that only natural non-tidal dispersion takes place below the
62U? foot station.  The distance required for the dye to reach the
stream center would be 5-7 miles given the same assumptions.

-------
                                SECTION V

                    DELAWARE AERATION SYSTEM ANALYSIS
Objectives

The feasibility of the application of instream aeration to meet  dissolved
oxygen standards on the Delaware Estuary is examined in Sections V and
VI.  In this section a systems analysis is carried out to estimate the
amount of oxygen input required to raise the existing DO level to a spe-
cific, desired level.

Various desired DO levels have been suggested for the Delaware Estuary
(7).  In this analysis it is assumed that the present DO level will be
maintained at no less than 3 mg/1 by treatment, and that a further im-
provement to a minimum level of h mg/1 is to be achieved by instream
aeration.

The Delaware Estuary has been intensively studied by the Environmental
Protection Agency, the Delaware River Basin Commission and others; and
reference has been made to various published and unpublished results in
order to relate the present analysis to what has been done previously.

The Critical Region

The estuary between Trenton and Listen Point is divided into 30 sections,
the length of which being either 10,000 feet or 20,000 feet, except
section one, which is assumed to be 21,000 feet long, as described by
Pence et al  (35).

The critical region is designated as the region in which the present DO
levels are below 3 mg/1.  The region was determined by using the iso-
variate plot of dissolved oxygen for 1961; loading conditions.  Two
critical fresh water flows were investigated;  namely, a l-in-25 year low
flow and a 6000 cfs minimum flow assumed to be provided by an upstream
reservoir system.  For details, see Pence e_t al  (35).

Composite DO profiles for the months June through September, derived
from these figures are plotted  in Figure 23.   For the l-in-25 year low
flow, the critical region includes Sections 10 through 20, while  for the
6000 cfs minimum flow it includes Sections 12  through 21.  To obtain a
conservative estimate of the oxygen input required,  it was decided to
consider Sections 10 through 21 as the critical region that required
instream aeration.  As can be seen in  Figure 23, the present DO levels
in Sections  9, 22, and 23 are lower than the desired k mg/1, and  thus
require aeration.  However, the actual DO profile to be raised by ad-
vanced treatment will undoubtedly extend beyond  the  critical region
making it unnecessary to add oxygen to those sections.
                                     55

-------
v-n.
                   I   I	!	1
I  I  I  I  I  I  I  I	\	1	1	1	1	1 I  I  I  I I I  I  I
                                6000 cfs flow
                   l-in-25 yr.  flow
                   Scale, miles

                 i	  I     i     i
                                                                  j   GO
                                                                           1'
                                                                           xo
                                      IV) i\3rorororor\jr\}f\}
                                      O I' i\> Vjj r- vn. ON ^J CD
t\3

VQ
                                FIG.  23   ESTUARY  SECTIONS, COMPOSITE D.O.  PROFILE

                                      June Through September,  Delaware River

                           Based on  l-in-25 Year and 6,OOO CFS Minimum Flow at Trenton

-------
 Oxygen Mass  Balance  in a Section

 Following procedures of Pence et  al  (3%}, any section i can be represent-
 ed by a free body, as  shown in Figure 2k.  Mathematically, the oxygen
 mass balance in  the  section can be expressed as  (35)
                          oxygen entering section
                                                                      Si
 natural aeration                               oxygen leaving section

                  BOD  consumption


 in which

     i = 1, 2, 3.-..n  and i-1 = upstream, i+1 = downstream
     L = ultimate carbonaceous BOD demand
     Q = net flow from section to section
     IF = volume of section
     5 = an advection  coefficient due to tidal motion
     E = an eddy exchange coefficient
     d = the decay rate of BOD
    Cs = saturation DO concentration
    Ci = DO concentration in section i
     r = atmospheric reaeration coefficient; and
     S = other sources or sinks of oxygen.

 In this analysis the following assumptions were made to obtain approximate
 estimates of oxygen required:

     1)  Qi-1 i = Qi   i+1  i.e. constant flow

     2)  Ci_i = Ci+l = Ci^ i.e. uniform DO level

     3)   C i-1, i =   ?i  i+1  i.e. constant advective coefficient

Equation 1 is then simplified to:

                   riVi(Cs - Ci) = diLi - Si                 (2)

Now if the DO level is to be raised artificially from 3 mg/1 to h mg/1 in
the critical region, the reduction in natural reaeration must be supple-
mented by instream aeration in order to maintain the higher DO level.


                                     57

-------
              -diLi
             1th  Section
            Net  Flow
FIG. 2h   OXYGEN MASS  BALANCE ON SECTION i
               (After Pence et al.)

-------
But Sj[ remains unchanged.

Therefore

                      rjViCCs - 3) - rjViCCs - U) = N               (3)

where

     NJ_ = oxygen required from instream aeration.

Thus                                _
                              % = BVj^lVi                          (U)

where

     B = a units conversion constant.

In order to allow a DO sag after the oxygen level is raised to k mg/1,
and also to account for  the effect of dispersion, the oxygen required,
was computed for a DO raise from 3 mg/1 to Lfe mg/1, by the following
equation:
where

     NJ_ is in Ibs 02/day

     ^ in liters, and

     B = 2.20U6 x KT6

Data for the hydraulic  geometry and  atmospheric reaeration coefficient
were obtained from the  EPA  office  in Metuchen, N.J.   The results as to
oxygen required are  listed  in Table  7-   The  computation gave a total re-
quired oxygen input  of  30U,869  Ibs/day  for sections  10 through 21.
Since the Delaware Estuary  is tidal, with tidal flows far in excess of
flows due to net discharge,  additional  net discharges up to 6000 cfs do
not correspond to materially increased  BOD loads  or  oxygen requirements,
and, from Figure 23,  it appears that the critical period is the period
of extreme low flow.  For purposes of computation the total area affect-
ed by either period  was included.

As described earlier in Section III of  this  report,  Table 2, the actual
oxygen transfer rates of the diffuser aerator  for raising DO from 3-0 to
li.5 mg/1 are as follows:

                        At  20 ft        32UO Ibs  02/day
                            25 ft        3530 Ibs  02/day
                            kO ft        Ii580 Ibs  02/day

-------
                     TABLE 7

COMPUTATION OF OXYGEN REQUIRED TO MAINTAIN A 4 MG/L
           D.O. LEVEL IN THS CRITICAL REGION
i
Section
10
11
12
13
14
15
16
17
18
19
20
21
r
I/day
.14
.14
.10
.10
.16
.16
.23
.16
.15
.15
.11
.16
Volume
liters x 106
20,70O
19,794
19,057
17,642
21,606
56,209
58 , 446
70,566
84,866
9O,728
108,992
56,039
N*
Ibs O^/day
9,581
9,161
6,300
5,832
11,429
29,732
44,441
37,327
42,085
44,992
39,636
24,353
 N = 2.2046 x 10"6(V) (1.5)r          Total     304,869
                             60

-------
This diffuser is 80 feet long, with 160 nozzles.  Based on the above ef-
ficiencies, the number of the diffusers required to supply the oxygen
demand was computed for each section for the three depths of submergence.
The results are given in Table 8.

Comparison with DECS Report

In a tidal estuary, the effect of adding oxygen at one section will not
be confined to the section volume alone, but will spread to both up-
stream and downstream sections due to tidal mixing and dispersion.  The
DECS Report (6) has presented a "response matrix," which gives the
responses in DO increase in all the sections due to an input of 10,000
Ibs/day of oxygen at any section.  Since the DECS model was based on the
general equation of the form of Equation 1, it was intended to compare
the results of this study with those computed by the DECS model.  If
placed at 25> foot depth, a total of 91 such diffusers would be required.
However, as is brought out in the next section, design considerations
will vary the type of aerator used, so that this result is only illus-
trative.

To compare the above results with those computed by the DECS model  (36),
the actual oxygen input at each section was first calculated, based on
the number of aerators placed at 20 foot depth as indicated in Table 8.
The oxygen input vector was then multiplied with the DECS response
matrix, to give the actual DO increase in each section.  The results
are listed in Table 9.

The system results based upon the present approach (treating each sec-
tion separately) are not much different from those of the DECS model.
The latter shows a considerable dispersion of oxygenation effects to
adjacent sections, which is undoubtedly realistic, about 16% less aggre-
gate dissolved oxygen response, which is presumably due to differences
in the approach used.  The total additional oxygen demand obtained by
the methods of this study should be conservative, since it provides
enough oxygen to maintain a mean level of k*5 mg/1 throughout the entire
critical regions, which would correspond to local levels of over 5  mg/1
at aerators and minimums of li mg/1 in the middle of the stream.  There
seems no reason to believe that the aeration process itself would accen-
tuate either the normal or benthic BOD demands, under situations given.
If natural aeration is affected at all, other than by the reduction of
the oxygen deficit, it would be increased.

As will be discussed  in the next section, it is unlikely that such  a
large difference as 1.0 mg/1 will be required.  Small indicated deficien-
cies of the DO level below the required norm towards the ends of  the
critical region are not of significance, because a treatment program
which establishes minimum levels of 3 mg/1 in that part  of the river
would not result in P quantum jump at the ends  of the critical regxon
but in a  gradual slope or sag curve.  The results of the aeration program
could more easily be  adjusted to such a sag curve than to the assumed
step profile.
                                     61

-------
                                               TABLE 8
                              DIFFUSER AERATION SYSTEM DESIGN  - AN EXAMPLE




Section
10
11
12
13
14
15
16
17
18
19
20
21





Oxygen
required
to maintain
4 mg/1
Ibs/day
9,581
9,161
6,3OO
5,832
11,429
29,732
44,441
37,327
42,085
44,992
39,636
24,353


Total



Number of
If placed
at 20 
depth
3
3
2
2
4
10
14
12
13
14
13
8


98



8Of diff users
If placed If
at 25
depth
3
3
2
2
4
9
13
11
12
13
12
7


91



needed
placed
at 40'
depth
3
2
2
2
3
7
10
9
10
10
9
6


73






Remarks
(1) At 20 * the diff user trans-
fers 1.09 Ibs O2/hp hr at a
brake hp of 124 hp; this gives
3240 Ibs O2/day.

(2) At 25 the diffuser trans-
fers 0.98 Ibs O-2/hp hr at a
brake hp of 15O hpj this gives
3530 Ibs O2/day.

(3) At 4Or the diffuser trans-
fers 0.83 Ibs O2/hp hr at a
brake hp of 230 hp; this gives
4580 Ibs O2/day.

(4) All for a temp0 of 25C
and average d.o 375 mg/le
ro

-------
TABLE  9
COMPARISON OF THE OXYGEN INCREASE
PRESENT DESIGN AND THAT BASED ON DECS D.O



Section
1-6
7
8
9
1O
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

Number
of
dif fuser s
...
___
	
	
3
3
2
2
4
10
14
12
13
14
13
8
...
	
...
...
__ _
....
. . .
. _ .
...

Oxygen
input
Ibs O2/day
___
	
	
	
9,72O
9,720
6,48O
6,480
12,960
32,4OO
45,360
38,880
42 , 120
45,360
42 , 120
25,920
	
	
	
_ 
	
...
___
___
	

Oxygen
increase
mg/1
 W M
	
	
	
1.52
1.59
1.54
1.67
1.70
1.63
1.53
1.56
1.5O
1.51
1.59
1.60
	
	
	
	
	
	
_ 
___
	
DUE TO THE
. RESPONSE MATRIX
Oxygen increase
based on DECS
response matrix
mg/1
am 	 
0.01
O.07
0.23
0.56
0.77
0.90
1.00
1.10
1.23
1.34
1.43
1.48
1.41
1.21
0.87
0.65
0.49
0.37
0.28
0.21
0.15
0.11
O.06
0.02
         63

-------
                               SECTION VI

                     DESIGN AND COST CONSIDERATIONS
The area of oxygen deficiency on the Delaware River, extends over a
distance of about UO miles, including extensive port area developments,
heavy industrial sites, the Philadelphia Navy Yard and other developed
water front.  The river is generally from 2000 feet to 2500 feet wide
above the confluence with the Schuylkill, and wider below this point;
and it has been extensively dredged, with a IjO-foot main channel and
30-foot anchorage areas.

Since passage of anadromous fish is an important consideration on the
Delaware, consideration was given initially to the possibility of pro-
viding adequate oxygen only along the lesser developed New Jersey shore,
in order to provide for assured fish passage at minimum cost.  However,
experiments with the young of shad and striped bass have now shown that
these fish do not immediately perceive water of deficient oxygen; and
when suffocation begins they react randomly, being as likely as not to
swim deeper into the oxygen-deficient area (37).  Moreover, legally
adopted water quality standards apply to the river as a whole.  Accord-
ingly, it is accepted as basic that supplemental oxygenation must be
effective throughout the entire river cross-section.

Spacing of Aerators

Figure 25 shows schematically a system of aerator sites evenly spaced on
each side of a river.  Five-day BOD values during summer periods are
currently about 5 mgA> and dissolved oxygen falls at times to below 1.0
mg/1.  However, it is not intended that oxygenation be applied to pre-
sent conditions, but be used in combination with added effluent treatment,
which will reduce prevailing levels of BOD from the present values of
about 5.0 mg/1 (5 day) to perhaps 1|.0 mg/1.  If the influence of various
sources and sinks of aeration is assumed unchanged by river aeration or
degree of effluent treatment, the steady-state condition to be satisfied
in a section where DO level remains level, based upon the usual first
order relationship, is the following

                       K2  (CS-C) = K,L


where K2 is aeration coefficient, Cs is  oxygen saturation of water,  C is
actual DO level, K, is BOD removal  coefficien^and L is BOD  (ultimate).
If aeration raises DO level C temporarily to C , mg/1, a corresponding
expression will be

              dc/dt = K2 (Cs - c')  - K,L                         (2)

                 or   K2 (Cs - G')  = K,L - dc/dt                 (3)

-------
River  bank
                                 Aerator
River  bank
                  (A)  PLAN,  AERATOR SPACING
                 (B) DISSOLVED OXYGEN PROFILE
            FIG. 25    DISSOLVED OXYGEN DISPERSION

-------
Dividing Eq.  (3) by Eq.  (l) we obtain

                     Cs  - C!  _ K,L - dc/dt

                     Cs  - C         K,L
(U)
Substituting values of Cg = 8.17,  C'  = U-5,  C  =  3.0, K, = 0.37, and
L = 1.5 x 5-day BOD = 6.0

Eq. (i|) may be reduced to give



                      dc/dt = 0.6U5 mg/1  1 day
Therefore the rate  at which DO will  decline  after  the  aeration-to U.5
mg/1 will be 0.6U5  mg/1  per day.   The  aeration of  water  at  the  aerator
site to a DO of U-5 mg/1,  at  a time  when the river as  a  whole was slight-
ly  above U.O, would be immediately subject to dilution effects  during
the dispersion process.   The  actual  reduction in DO concentration would
be  a complex variable depending  upon dilution as well  as natural aera-
tion and biochemical action.   As an  approximation  it may be assumed that
the process is represented by a  dilution reducing  the  DO immediately to
U.25 mg/1, followed by reduction by  aeration and biochemical processes
only,  at the rate  of 0.61*5 mg/l/day.
 If spacing of aerator  sites  is assumed to be not over 2.0 miles,  the
 basic  design condition will  be an oxygenated plume from one site  dis-
 persing to the center  of the river, a distance of 5-U5 miles,  at  a rate
 of 1.27 ft/sec, and thereafter flowing another 2.0 miles until the next
 oxygenated plume arrived. Application of Eq. (U) indicates that  DO
 would  decline by 0.2h  mg/1 during this movement.  An alternative  assump-
 tion is that flow near the bank, during periods of low tide, might be
 only one half of that  actually observed, or 0.63 ft/sec, and that it
 might  circulate under  tidal  influence from one site almost to the other,
 and then return, a distance  approaching U.O miles.  The computed  reduc-
 tion of DO during this cycle would be 0.25 mg/1.  Therefore it may be
 seen that, under the assumptions made, a spacing of aerator stations  ol
 2.0 miles will be satisfactory, and this spacing is assumed for design
 purposes.  Where particular  circumstances warranted it, spacing closer
 than two miles would be operationally acceptable, down to as little as
 UOO feet for 25-foot diffusers.  Below this interval, as shown in
 Section IV, such an aerator  would be likely to rehandle a substantial
 proportion of aerated water  from a site immediately upstream, and thus
 lose efficiency.  Larger aerators would require further spacing.
                                     67

-------
For all types of aerators considered, an increase in number of aerator
sites would add to total cost, primarily on account of cost of electri-
cal connections and electric service. Accordingly, based upon general
design considerations, aerators may be tentatively considered to be
spaced about two miles apart, down each side of the river.  The sites
would handle unequal quantities of oxygenation, depending upon a BOD
systems analysis to determine the requirements of each section.

Aerator Type

Unlike the case for small rivers such as the Passaic, oxygen transfer
efficiency is not the determining factor in selection of aerator type.
In this important port area  the requirements of navigation greatly
limit use of mechanical  aerators, since their pile moorings would
greatly interfere with navigation in the main port and anchorage areas.
They can be used outside of  port activity  and anchorage areas; but the
site must have water  of  sufficient depth and current to carry away the
aerated mass and disperse it into the main flow of the stream.

Surface Aerator Sites

The layout  of  a suitable surface aerator site should be governed as far
as practicable by the following criteria.

      (a)  Not  more  than  two  aerators along a stream line
      (b)  As close  to the main channel as  possible
      (c)  Layout as compact  as possible
      (d)  Layout such that any one aerator can be removed individually.

Two suggested  layouts are outlined in Figure 26,  one with one aerator
and the other with  two.  The twin aerator  layout  could easily be expand-
ed to take  a third  aerator shoreward of the two shown.  Such sites
would require submarine  cable connections  to shore, and a source of
electric power.  Cost estimates are  indicated in  Table 10.  These esti-
mates are on a similar basis to estimates  previously reported  (1),
modified to provide for  creosoted piling,  submarine cable, and naviga-
tion lights.  They  are based upon electric drive, 75 hp surface aerators,
with frames specially reinforced, and upon installation of a number of
facilities in one contract.   Figure  27 shows the  mooring  of a  surface
aerator between pile  clusters.

As shown in Section III, the surface aerators have a transfer rate of
3.06 Ibs/hp-hr at standard conditions.  This amounts to 1.70 IbsAp-hr
field output at 25C  and a mean DO level of 3.75  mg/1.  Therefore, the
field output of the three-aerator site in  pounds per day would be
3 x 75 x 2k x 1.70  =  9150 Ibs/day.

Annual costs computed as indicated in the  paragraph following are
$1;9,200.  With a 135-day operating period  annually, this would amount
to U.O cents per pound of oxygen utilized.  Corresponding figures for
                                     68

-------
                          Channel
        (A)  SINGLE
            AERATOR
        To
       shore
            Channel ^^
	 	 	   	  JL.   	
(B)  TWIN
    AERATORS
                                      Symbols
                                   9  Pile  cluster
                                 sn :rz   Submarine cable
   FIG. 26    SINGLE  AND  TWIN AERATOR FACILITIES

-------
                                    Table 1O



                                 COST ESTIMATE


                            SURFACE AERATOR SITES
                                  Single          Two            Three
                                  Aerator       Aerators       Aerators


Installed horsepower                 75           150             225

Equipment and Construction
   Items                          $57,000       $92,700        $130,500

Engineering and Contingencies      11,^00         18,500        	26,100

   Total Construction Cost        $68,UOO       $111,200        $156,600



Operation and Maintenance         .$20,000       $  26,500        $ 33,000


Interest and Amortization
   at 15 years 6%  basis  (10.356)       7,000         11,$00          16,200

     Total Annual  Costs            $27,000       $38,000         $1+9,200
                                       70

-------
SURFACE AERATOR MOORING

-------
the two aerator sites are U.6 cents and for the  single  aerator site  6.6
cents per pound of oxygen utilized.

Equivalent Annual Costs

The costs in Table 10 are reduced to equivalent total annual costs per
unit of oxygen transferred, for comparison with other types of aerator,
in the following manner.

The social discount rate for economic comparison of one government in-
vestment with another may be approximated as the long-term U.S. bond
yield rate, less an allowance for the inflation increment in the current
rates, and plus an increment for risk not borne by the bond holder (38).
In November 1970, U.S. bonds of maturity exceeding ten years had an
average interest rate of 6.1%.  Allowing for a depreciation component
of 1.7% and a risk component of !.(#, the costs of alternatives may be
compared on the basis of an economic interest/discount rate of 6%
annually.  With an economic life of 15 years, this corresponds to a
combined rate of 10.3$  annually for both interest and  amortization of
the original investment.

Diffuser Aerator Facilities

The costs  of providing  a diffuser aeration facility  depends very largely
upon  distance  of the diffusers from the  shore,  since underwater pipe-
lines are  very expensive.   The distance  between the  blower and the
manifold  is assumed to  be 1000 feet; and the resulting estimated pipe-
line  cost  is more than  half the  entire construction  cost  in each case.
Installation nearer shore would  be less  expensive, although on-shore
pipelines would be required in most such cases.

Another very important  design condition  is the  question of whether  the
diffusers  are  capable of operation for long periods  of time without
clogging  or obstruction by sediment.   One  unpublished  study of river
aeration potential (preliminary  by others)  eliminated  diffuser aerators
from  consideration because of this question.  However, the diffuser
aerators  tested have holes of 5/32-inch  diameter, with a  ball valve
which at  least partly seals off  the manifold when not  in  use. It is
assumed that these diffusers would remain  operable over a period  of
four  years,  on the average,  and would  then require to  be  removed  and
serviced.   Much greater efficiency could be obtained by fine  bubble
aerators,  but  because of this question of  clogging,  the coarse bubble
aerator has been retained as a basis for comparison.  If  necessary,
devices could  be provided to seal off  the  orifices of  the diffuser
when  not  in use.

A third important consideration  in design  is the  relationship of  the
equipment to navigation requirements,  in the areas in  which the pier-
 head line and channel line are closely adjacent.  According to dis-
 cussions  with the Philadelphia District, Corps of Engineers,  it must
                                     72

-------
be accepted as a constraint that there can be no diffusers laid in the
channel and anchorage areas on account of potential interference with
mooring or with maintenance dredging.  On the other hand, placing of
diffusers back of the pierhead line might leave them in slack water or
eddy conditions during portions of the tidal cycle, such as to detract
seriously from their effectiveness.  Accordingly, in areas where navi-
gation requirements preclude the use of surface aerators, there are
three possible solutions, as follows.  First, in open water, use single
long diffusers, as illustrated in Figure 28B and cases a, b, and c in
Table 11.  Second, where the pierhead line is close to the channel line,
the necessary diffuser capacity could be provided by one or more short
(25) diffusers perpendicular to the line of stream flow, successive
diffusers being separated by at least UOO feet along the stream line,
in order to avoid reprocessing the aerated water a second time.  These
short multiple diffusers illustrated in Figure 28 (A) and cases d and
e of Table 11, would be comparatively expensive, on account of the under-
water pipelines.  Third, where piers exist, the outer ends of which are
not used for mooring purposes, a good solution would be a diffuser laid
parallel to the line of stream flow, close to the pier, as illustrated
in Figure 29 and with costs as shown in case f, Table 11.  Judging
from the flow characteristics shown by diffusers in the tests, the
surface water would flow away from the pier at a fairly rapid rate,
as shown by solid lines on the figure, while water from lower levels
would flow in towards the diffuser as shown by dotted lines.  It is
believed that such a design, if not over 80 feet long, would function
satisfactorily, except briefly at slack tide.  At such periods all types
of aerators would probably recirculate some of the aerated water.  An
examination of the map shows that there are a great number of finger
piers in the Philadelphia area, many of which would undoubtedly be
adaptable to such an installation.

The following design assumptions were made for the diffuser aerators
by consulting engineers Hazen and Sawyer, in preparing the cost
estimates.

     1.  In relating air flows to estimated shaft horsepower, average
         blower efficiency was assumed at 80$ and a discharge pres-
         sure was estimated to overcome the depth plus 2%% for losses.
         These values were checked against the maximum allowable pre-
         sure drop in the diffuser nozzles when located one foot on
         center, with 12 orifices "open" and a discharge air flow of
         15 scfm per diffuser nozzle.

     2.  An optimum air velocity of 50 to 60 feet per second was used
         to size the main and diffuser header pipes.

     3.  No excavation of the river bottom is included in these costs.
         The diffuser would lie directly on the bottom.

     ii.  Current market prices (material only) were used for the air
                                    73

-------
                                                     Table 11




                                                  COST ESTIMATES




                                           DIFFUSION AERATION FACILITIES
Case Designation
Manifold Systems
River Depth, ft.
SCFM
Equipment and
Construction
Contingencies and
Engineering at 2Q%
Total
Annual Operation and
Maintenance Costs
a
1-80'
20
2620
$101,000
20,000
$121,000
$ 18,000
b
1-160'
20
52liO
$128,000
26,000
$15^000
$ 26,000
c
1-80'
30
2790
$110,000
22,000
$132,000
$ 22,000
d
1-25 '
1*0
1000
$ 8U,ooo
17,000
$101,000
$ 15,000
e
3- 25'
ho
3000
$176,000
35,000
$211^000
$ 3^,000
f
1-80'*
30
2790
$ 77,500
15,500
$ 93,000
$ 19,000
Interest and Amortization
at 15 years 6% basis
(10.356)
12.500
Total Annual Cost $30,500


15,900
$11,900

13,700
$35,700

io,Uoo
$25,i|00

21,800
$55,800

9,600
$28,600

-;;-  Installed at pier end

-------
          (A)  RESTRICTED PORT AREAS
                                 Channel
IE            "U  .                :_IT
                I  /


                  y
                     Pier
                              Shore
            (B) NORMAL CHANNEL AREAS
  C_^.                         Channel - 
                                    Symbols
                                       Aerator
                                       Air supply
                                       Blower
      FIG.28    DIFFUSER AERATOR LAYOUTS
                           75

-------
  ELEVATION
  PLAN
             Pier
          j
                                                 V
ii   n
h   H



n   ii

h   n
                                                    \
                                             /      ^
                                             /  <&
                                _ii
                                 >
FIG. 29    DIFFUSER AERATOR AT PIER END
                         76

-------
    blower with its accompanying electric drive motor, with an
    allowance of 30$ for installation.  The estimates of horse-
    power were generally increased to account for economic equip-
    ment sizes.

.  The blower assembly was assumed to include the following:

    a.  Purchase of blower unit and accessoriesone or two
        stage blower, V-belt drive or reduction gear, electric
        drive motor, starter, base, coupling guard, relief
        and unloading valve, inter-stage piping, inlet filter
        silencer, discharge silencer and wiring.

    b.  Installation of blower with electric drive motor complete
        with timber or concrete foundation, and

    c.  Assistance in intial start-up operation.

6.  The air piping system includes the following:

    a.  Preliminary river soundings.

    b.  Purchase, erection and installation of 1000 ft. of Schedule
        iiO steel pipe (main header) and various lengths of diffuser
        piping.

    c.  Valves, fittings and flexible couplings.

    d.  Purchase and installation of proper anchorage and  supports.

    e.  Purchase and installation of diffuser  nozzles similar to
        previously purchased.

    f.  Assistance in initial  start-up operation.

 7.  Electrical costs were estimated based on a source of electric
    power approximately 1,000  ft. from the site.   Among the princi-
    pal items included  are  the following:   1,000  ft.  conduit,
    3,000 ft.  of cable  sized for motor load, lighting transformer
    feeder breaker, lighting transformer,  lighting panel,  two
    floodlights, and six incandescent fixtures.   In addition,
    20 ft.  of service  cable from the service  pole is included.
    The cost of the poles was  assumed backcharged by the utility.

 8.  Each piping is planned  to  be removed  for cleaning and  repair
    and reinstalled once every four years.

 9.  The following  schedule  of  equipment operation was used:
                                77

-------
                 Number of
                Months/Year      Hours of Operation (all cases)

                    6            No operation
                    3            Full (2h hr/day)  continuous  operation
                    3            Half-day (12 hr/day)  operation

    10.  Annual costs of removal from river were estimated as $0% of
         the original installation cost,  prorated over four years.

    11.  During a daily visit to each of  approximately ten sites, one
         serviceman can inspect, repair and maintain equipment.

    12.  Electric power costs based on Public Service  Gas and Electric
         Company of New Jersey rate schedule for large power  and light
         users.

    13.  Annual maintenance costs estimated as 3% of total equipment
         costs.

The transfer efficiency of diffusers varies with depth, and must be
adjusted to field conditons.  In this case, field conditions  are cal-
culated as a mean DO value of 3-75 mg/1 and temperature of 25C, and
also pressure corresponding to the mean depth of water during rise of
bubbles.  A change in oxygen saturation value appropriate to  that
depth must be allowed for.  Values for the diffuser tested are as
follows:

               Depth             Percent Absorption
                                 Standard      Field
                                Conditions   Conditions

                 20                5.6          U.8
                 30                6.8          6.9
                 hO                7.0          8.1

Applying these percentages to the various cases outlined in Table 11,
the costs per pound of oxygen used during an assumed 135 day annual
operational period may be computed.

The diffuser aerator with lowest unit cost is case f,  a large pier-end
installation (801 unit) which would provide oxygen at I|.li0/lb.02 in 30'
water.  Case c, a similar installation in open river without  a pier,
assuming a 1000' underwater pipeline, would have costs of 5'50/lb.02
If the water were only 20 ft. deep (case a) costs would rise  to 7.2.
Smaller installations would be more expensive.  For example,  a single
25-foot diffuser in kO feet of water, in the open river  (case d),
would have costs of 9.3#

When a  comparison is made as to costs between diffuser and mechanical
                                    78

-------
aerators, account must be taken of volume of output, since, for both
classes of aerator, unit costs for the smaller units rise rapidly.
When compared on the basis of equivalent outputs, the diffuser aerators
appear somewhat more costly than the surface aerators, except for case f,
the pier-end aerator, which is less costly.

Based upon these results, the surface aerators appear to be the more
economical for reaches of the river where they do not interfere with
navigation.  In port areas and other developed waterfronts where
surface installations would not be acceptable, the diffusers would be
used, installed preferably upon pier ends or other favorable points
where underwater pipeline construction would be minimized.

Possibilities of Oxygenation

There are also possibilities of raising DO level by diffusion of pure
oxygen, which have been reported on both for rivers (39,1*0) and in much
more detail, for waste treatment plants (1*1).  Although investigation
of  such possiblilities was not within the scope of the demonstration
project, other studies underway indicate that oxygen diffusion offers
possibilities of providing an economical alternative to air diffusion
under conditions assumed.  This is particularly true for water of 30-1*0
feet deep.  There are two reasons for this:  (a) the oxygen supply lines,
being much  smaller, can utilize flexible plastic tubing rather than
steel pipe, and (b) the oxygen diffusers function at 1*0 feet depth as
well or better than at lesser depths, with added cost only for initial
installation.  The cost of the oxygen is a considerable item, of  course,
but for installations of the larger sizes in particular, oxygen dif-
fusion appears considerably more economical.  These conclusions are
based upon  estimates of cost of diffusers provided by the  Martin Marietta
Corporation, and provision for construction cost, provision of oxygen,
and annual  removal and cleaning of diffusers.  The annual  removal of
diffusers would be much simpler for the small oxygen diffusers and
flexible  supply lines than with the heavy  steel aeration pipe lines and
diffusers.  However, further work will be  required  to determine various
design  aspects since there are several alternative methods of oxygenation.
Also it will be necessary to make an  entirely different dispersion
analysis  since the dispersion analysis of  the present report will not
be applicable.

Overall Economy  of Induced Oxygenation

An accurate cost estimate for adding  oxygen to  a major river  cannot be
be made  without  further research  and  design  studies,  particularly to
 determine which  areas  in  the river  can be  oxygenated with  surface
 aerators and which of  the remainder can be served  by  pierhead installa-
 tions.   Also,  possibilities  of further economies by use  of pure  oxygen
 should be developed.   However,  some  very  rough  approximations  can be
made.   It is shown in  Section V  that  the  total  additional  oxygen re-
 quirement on the Delaware  amounts to  about 30^,000 Ibs/day (Table 7).
                                   79

-------
Allowing extra for irregularities in spacing and sizing, this might
amount to about U8 million pounds of oxygen annually provided by a mini-
mum of perhaps 50 sites, an average of 960,000 Ibs 02 annually per site.
Such an input would require two of the 8 foot pier-end diffusers (case
f) or intermediate between two and three surface aerators.  Since the
distribution of added oxygen to obtain the desired oxygen distribution
would be somewhat irregular (as shown in Section IV), and the size of
diffuser facilities will usually be limited by practical considerations,
there will undoubtedly be a considerably larger number of sites required.
Some sites will undoubtedly be required of size much less than the mean,
with corresponding increase in unit price.  It appears likely that the
mean unit cost of raising the minimum average DO level from 3-0 to U.O
mg/1 by means of induced aeration would cost about 50 per pound of oxy-
gen added during the critical period.  This would correspond to total
cost, including amortization of $1^,800,000 annually.  Since this is much
less than the costs of achieving the same results by treatment alone, it
appears that the instream aeration should be adopted in planning^as an
alternative to advanced degrees of waste treatment, where the main
objective is the maintenance of satisfactory dissolved oxygen levels.
                                      80

-------
                               SECTION VII

                            ACKNOWLEDGEMENTS
The cooperation of numerous agencies and individuals was received in
carrying on this investigation, in addition to the co-authors of this
report.

The New Jersey Department of Conservation and Economic Development
cooperated in furnishing non-Federal matching funds for the project,
and in granting the necessary state permit for the installation.

The U.S. Geological Survey, Trenton Office, headed by Mr. John E.
McCall, district engineer, and Mr. Peter W. Anderson, chief of the
Water Quality Branch, assisted by organizing the field dispersion tests
and providing boats and special equipment for field work, under direct
supervision of Mr. John Murphy.  Mr. F. A. Kilpatrick and Mr. N. Yotsukura
of the Washington Office, U.S.G.S., were consulted regarding interpreta-
tion of results.

Colonel James A. Johnson, district engineer, Philadelphia District, Corps
of Engineers, and Mr. Phillips and Mr. Cable of that office assisted
materially by providing hydrographic data, granting a Federal permit for
the test installation, loaning a boat for the dispersion tests, and by
giving advice regarding navigation constraints required for prototype
installations.

The Metuchen Office, Environmental Protection Agency, headed by Mr.
Kenneth Walker, made available tentative conclusions of previous office
studies concerning this matter.

The staff of the Delaware River Basin Commission, especially Mr. James
Wright and Mr. Herbert Hewlett, gave moral support for the study as a
whole.

Mr. John Frazee, plant manager of the Camden Sewage Treatment Plant,
gave invaluable local support in setting up the test site, and helped
obtain approval of the Camden City government to provide the site with-
out charge.  The Camden police sent patrols with both car and boat
which were essential to maintenance of security.

Brig. General Allen F. Clark, director of the Philadelphia Port Corpora-
tion, made a pier, an air compressor and technical assistants available
for the second set of tests, run in June 1970.

Finally acknowledgement should be made of the students who participated,
in various capacities, including Messrs. K. Hilsen, F. Woll, R. Carver,
R. Lynch, R. Gessner, D. Chen, L. Michna and P. Whipple.
                                     81

-------
                              SECTION VIII

                               REFERENCES
 1.  Whipple, W., Jr., Hunter, J.V., Davidson, B., Dittman, F., Yu, S.,
     "Instream Aeration of Polluted Rivers," Water Resources Research
     Institute  (1969).

 2.   Hunter, J.V., and Whipple, W. , Jr.,  "Evaluating Instream Aerators
     of  Polluted Rivers,"  Journal WPCF,  1;2, No.  8, pp 2^9-262, Pt. II
     (1970).

 3.   Whipple, W., Jr., Coughlan, F.P.,  and Yu, S.L., "Instream Aerators
     for Polluted Rivers,"  Journal Sanitary Engineering Division, ASCE,
     96, SA5, PP 1153-1165  (1970).

 li.   Yu, S.L.,  "Aerator Performance in  Natural Streams,"  Journal Sani-
     tary Engineering Division, ASCE, 96, No. SA5, Proceedings Paper
     7601,  pp 1099-111U  (1970;.

 5.   Tarassov,  V.J., Perlis,  H.J., and  Davidson,  B., "Optimization of a
     Class  of River Aeration  Problems by the Use  of Multivariable
     Distributed Parameter  Control Theory,"  Water Resources Research,
     5,  No.  3,  PP 563-573  (1969).

 6.   Federal Water Quality  Administration, "Delaware Estuary Comprehen-
     sive Study, Preliminary  Report and Findings,"  Philadelphia, Pa.
     (1966).

 7.   Wright, J.F., and Porges, R., irater Quality Planning and Manage-
     ment Experiences  of the  Delaware River Basin Commission,"  5th
     International Water Pollution Research Conference, July - Aug.,
     pp  1-3/1 - I-3A7  (1970).

 8.   Delaware River Basin Commission,  "Final  Progress Report, Delaware
     Estuary, and Bay Water Quality Sampling  and  Mathematical Modelling
     Project,"   105 PP  (1970).

 9.   Bewtra, J.K., and Nicholas, W.R.,  "Oxygenation from Diffused Air
     in  Aeration Tanks,"  Journal  WPCF, 36, No. 10, pp 1195- 1221).  (l?6U).

10.   Carver, C.E.,  "Oxygen  Transfer  from Falling  Water Droplets,"
     Journal Sanitary Engineering  Division, ASCE, 95, No. SA2, pp 239-
         (1969).
11.  Barnhart, E.L. , "Transfer of Oxygen in Aqueous Solutions,"   Journal
     Sanitary Engineering Division, ASCE, 95 ,  No.  SA3,  PP 6U5-661 (19697 -

12.  Streeter, H.W., Wright, C.T., and Kehr, R.W., Sewage Works  Journal,
     8, p 181 (1936).

-------
                           REFERENCES (Cont'd)
13.  Eckenfelder, W.W., "Absorption of Oxygen From Air  Bubbles  in
     Water,"  Journal Sanitary Engineering Division,  ASCE,  85,  No. SAl;,
     PP 89-99 (1959;.

lU.  Eckenfelder, W.W., and Ford, D.L., "New Concepts in Oxygen Transfer
     and Aeration,"  in Advances in Water  Quality Improvement,  edited by
     E.F. Gloyna and W.W.  Eckenfelder, University of  Texas  Press, Austin,
     Texas (1968).

15.  Garland, C.F., "Research on the Influence of Basin Volume  and
     Geometry on Performance of Surf ace -Entraintnent Aerators,"  paper
     presented at 1969 ASME/AIChE Stream Pollution Abatement Conference,
     Rutgers - The State University, New Brunswick, N.J. (1969).

16.  Kaplovsky, A.J., Walters, W.R., and Sosewitz, B.,  "Artificial Aera-
     tion of Canals in Chicago,"  Journal  WPCF,  36, No. k,  PP k63-klk
     (19610.

17.  Susag, R.H., Polta, R.C., and Schroepfer, G.J.,  "Mechanical Surface
     Aeration of Receiving Waters,"  Journal WPCF, 38,  No.  1 (1966).

18.  deary, E.J., "Potentialities for Induced Oxygenation of Rivers,"
     Resources for the Future, Inc., Washington, D. C.  (196U).

19.  Imhoff, K.R., "Oxygen Management and Artificial Reaeration in the
     Area of Baldeney Lake and the Lower Ruhr River,"  Das Gas  und
     Wasserfach  (Germany), 109, p 936 (1968).

20.  Elder, J.W., "The Dispersion of Marked Fluid in Turbulent  Shear
     Flow,"  Journal Fluid Mechanics, 5, pp 5;-5"60 (1959).

21.  Fischer, H., "A Note  on the One Dimensional Dispersion Model,
     Air and Water Pollution,"  International Journal,  10,  pp Ui3-
     (1966).

22.  Fischer, H., "The Mechanics of Dispersion in Natural Streams,"
     Journal Hydraulics Division, ASCE, p  18? (196?).

23.  Fischer, H., "Dispersion Predictions  in Natural  Streams,"   Journal
     Sanitary Engineering  Division, ASCE,  p 927 (1968).

2U  Fischer, H., "Methods for Predicting  Dispersion  Coefficients in
     Natural Streams, with Applications to Lower Reaches of the Green
     and Duwarnish Rivers, Washington," U.S. Geological Survey,  Pro-
     fessional Paper 582-A (1968).

25-  Glover, R., "Dispersion of Dissolved  or Suspended Materials  in
     Flowing Streams,"  U.S. Geological Survey,  Professional Paper U33-B
     (19610.

-------
                           REFERENCES (Cont'd)
26.  Holley, E.R., "Some Data on Diffusion and Turbulence in Relation
     to Reaeration,"  University of Illinois,  Water  Resources Report
     No. 21 (1969).

27.  Jabson, H., and Sayre, ., "Predicting Concentration Profiles  in
     Open Channels,"  ASCE National Water Resources  Engr. Mtg.,  Memphis,
     Tenn., paper No. Ilii7.

28.  Paulson, R.W., "The Longitudinal Diffusion Coefficient in  the
     Delaware River Estuary as Determined from a Steady State Model,"
     Water Resources Research, 5, No. 1, p 59 (1969).

29.  Taylor, G.I., "Dispersion of Soluble Matter in Solvent Flowing
     Slowly Through a Tube,"  Proc. Roy. Soc.  Series A, 219, p  186  (1953).

30.  Wilson, J.F., "Time-of-Travel Measurements and Other Applications
     of Dye Tracing,"  International Assoc. Sci. Hydrol., Pub.  No.  76
     (1968).

31.  Wilson, J.,  "An Empirical Formula for Determining the Amount of
     Dye Needed for Time-of-Travel Measurements,"  U.S. Geological
     Survey, Prof. Paper 600-D, pp D5h-D56 (1968).

32.  Yotsukura, N., Fischer, H., and Sayre, W., "Measurements of Mixing
     Characteristics of the Missouri River Between Sioux City,  Iowa and
     Plattsmouth, Nebraska,"  U.S. Geological Survey, Water-Supply
     Paper  1899-G (1970).

33.  Bischoff, K., and Levenspiel, 0., "Fluid Dispersion-Generalization
     and Comparison of Mathematical Models - I. Generalization of Models
     and -II. Comparison of Models,"  Chem. Engr. Sci., 17, PP 2U5-255
     and pp 257-26U  (1962).

3U.  Daily, J., and Harleman, D.R.F., "Fluid Dynamics,"  Addison-Wesley
     Publishing Co.  (1966).

35.  Pence, G.D.,  Jr., Jeglie, J.M., and Thomann, R.V.,  "Time-Varying
     Dissolved Oxygen Model,"  Journal Sanitary Engineering Division,
     ASCE,  9h, SA2, pp 381-UOO (196).Especially Figs. 16 and 1Y.

36.  Information  informally furnished by the Metuchen Office, EPA,
     courtesy Mr.  Kenneth  Walker, representing pending studies.

37.  Dorfman, D.,  and Westman, J.,  "Responses of Some Anadromous Fishes
     to Varied  Oxygen Concentrations and Increased Temperatures,"  Part
     II, final  completion  report,  Project  B-012-N.J., Water Resources
     Research Institute,  New Brunswick,  N.J.  (1970).

-------
                           REFERENCES (Cont'd)


38.  Whipple, W., Jr., "Economic Basis for Water Resources Analysis,"
     Water Resources Research Institute, Rutgers, New Brunswick, N.J.,
     116 pp (1968).

39.  Anon, "For Rivers Breathing Room,"  Chemical Week, pp 131-2 (1969).

liO.  Amberg, H.R., Wise, D.W., and Aspitarte, T.R., "Aeration of Streams
     with Air and Molecular Oxygen,"  TAPPI, 52, No. 10, pp 1866-1871
     (1969).

lil.  Union Carbide Corp., "Investigation of the Use of High Purity Oxygen
     Aeration in the Conventional Activated Sludge Process,"  EPA,
     Water Pollution Control Research Series, 170^0 DNW05/70, 186 pp.
                                    86

-------
1
Accession Number
w
5
n j SHftJec* Field&. Group
&
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
    Title
             Oxygen Regeneration of Polluted Rivers:  The Delaware River
To]
      Whipple, William Jr.
      Hunter, Joseph V.
      Yu, Shaw L.
      Bit t man, Prank .
      Mattingly,  George W.
                                    j 1
                                    Project Dea{0nafion
                                                          16080 DUP  12/70
                                21 |
    Citation
23
Descriptors (Starred First)
   * Dissolved oxygen, * water quality  control,  * oxygen sag,  # aeration, # stream
     improvement, stream pollution, biochemical  oxygen demand, dispersion, pollu-
     tion abatement.
25
Identifiers (Starred First)

   *  Surface aerators, -* air diffusers,  * induced oxygenation, * instream aeration
27
    Abstract
        Tests of surface instream aerators  and  of bottom diffuser aerators were con-
        ducted on the Delaware River near Philadelphia in order to determine the
        practicability of induced oxygenation of deep navigable rivers.  The diffuser
        was tested at various depths up  to  38 feet,  but its performance in pounds of
        oxygen per horsepower hour decreased markedly in the deeper water.  Perfor-
        mance of the surface aerator appeared to be  somewhat improved over results
        previously found in a shallower  river.   Cost estimates and systems analysis
        led to the conclusion that induced  oxygenation by aerators appears to con-
        stitute an economical alternative to advanced waste treatment on the Delaware
        River.  This would require structurally reinforced surface aerators in some
        areas, and bottom diffuser aerators where the surface aerators would inter-
        fere with navigation.  However,  oxygen  diffusers developed by others may
        provide an even more economical  means of induced oxygenation for such
        rivers.
Abstractor
         William Whipple, Jr.
                               Institution
                                      Rutgers - The State University of N.J,
  WR:I02 (REV. JUUY 1969)
  WRSIC
                         SEND WITH COPY Of DOCUMENT, TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                         j^,,^,                      U.S. DEPARTMENT OF THE INTERIOR
                                                   WASHINGTON, D. C. 20240
                                                                               GPO: I97O - 407 -891

-------