EPA-R3-72-012

October 1972
      Ecological Research Series
                Characterization of
           Stream Reaeration  Capacity
                       f
                       I
S322
\
 UJ
 CD
                                   Office of Research and Monitoring


                                   U.S. Environmental Protection Agency

                                   Washington, D.C. 20460

-------
            RESEARCH REPORTING SERIES
Research reports of the  Office  of  Research  and
Monitoring,  Environmental Protection Agency, have
been grouped into five series.  These  five  broad
categories  were established to facilitate further
development  and  application   of   environmental
technology.   Elimination  of traditional grouping
was  consciously  planned  to  foster   technology
transfer   and  a  maximum  interface  in  related
fields.  The five series are:

   1.  Environmental Health Effects Research
   2.  Environmental Protection Technology
   3.  Ecological Research
   4.  Environmental Monitoring
   5.  Socioeconomic Environmental Studies

This report has been assigned  to  the  ECOLOGICAL
RESEARCH  series.   This series describes research
on the effects of pollution on humans,  plant  and
animal   species,  and  materials.   Problems  are
assessed   for   their   long-   and    short-term
influences.    Investigations  include  formation,
transport, and pathway studies  to  determine  the
fate  of  pollutants and their effects.  This work
provides the technical basis for setting standards
to  minimize   undesirable   changes   in   living
organisms   in   the   aquatic,   terrestrial  and
atmospheric environments.

-------
                                                  EPA-R3-72-012
                                                  October 1972
              CHARACTERIZATION

                      OF

         STREAM  REAERATION CAPACITY
                      BY

  E.  C.  Tsivoglou,  Principal  Investigator

    J.  R. Wallace,  Associate Investigator
            Project No. 16050  EOT

                Project Officer

          Dr. Walter M. Sanders III
          Southeast Water Laboratory
           Athens,  Georgia   30601
                 Prepared for
      OFFICE OF  RESEARCH AND  MONITORING
    U.S.  ENVIRONMENTAL PROTECTION AGENCY
           WASHINGTON, D.C.   20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $3 75

-------
            EPA Review Notice
This report has been reviewed by the Environ-
mental Protection Agency and approved for
publication.  Approval does not signify that
the contents necessarily reflect the views
and policies of the Environmental Protection
Agency, nor does mention of trade names or
commercial products constitute endorsement
or recommendation for use.
                   11

-------
                        ABSTRACT
The purposes of this research have been to characterize
stream reaeration capacity in terms of the stream hydraulic
properties and to develop procedures for evaluating the ef-
fects of pollutants on reaeration.  Field studies of the
reaeration capacity and the associated hydraulic properties
of five rivers have been completed, using a gaseous tracer
procedure for field measurement of reaeration.  These
studies have incorporated a wide range of hydraulic features,
such as waterfalls, rapids, shoals and pools, with stream
flows ranging from 5 to 3,000 cfs.  The range of BOD's and
temperatures encountered was also large.  Studies of the ef-
fects of both pure substances and community wastes on the
reaeration capacity have been conducted in a newly designed
test system.

Tests of observed vs. predicted values of K2 have shown that
none of the available models, e.g., O'Connor, Churchill, etc.,
is capable of providing dependable predictions of stream
reaeration capacity, especially under highly turbulent flow
conditions.  A new energy dissipation model has been derived,
by which the reaeration capacity of a stream is explained in
terms of the rate of energy dissipation, measured as the loss
of water surface elevation divided by the time of flow.  Two
distinct forms of the energy dissipation model have been
tested against the observed results, and it has been shown
that both forms provide dependable predictions of stream
reaeration capacity.

The tests of pollutant effects have shown that LAS and com-
munity wastes decrease the reaeration rate coefficient, pure
NTA has no effect, and pure mineral oil increases the
reaeration rate coefficient.

This report was submitted in fulfillment of Project Number
16050 EOT, under the sponsorship of the Environmental Pro-
tection Agency.
                           111

-------
                        CONTENTS


Section                                                 Page

   I     GENERAL SUMMARY AND CONCLUSIONS                    1

  II     RECOMMENDATIONS                                  H

 III     INTRODUCTION                                     15
         Purposes of this Research                        18
         Experimental Plan                                19

  IV     RELEVANT THEORY                                  23
         Molecular Diffusion, Mixing and Gas Transfer     23
         Mathematical Relationships                       31
         Field Tracer Application                         35
         Predictive Models                                38

   V     EXPERIMENTAL PROCEDURES                          47
         Physical and Hydraulic Properties                47
         Tracer Studies                                   63
         DO and BOD Studies                               74
         Pollutant Effects on Reaeration                  74
         Settleability of Tritiated Water                 78
         Sources of Error                                 81

  VI     EXPERIMENTAL RESULTS — FIELD STUDIES            85
         Flint River                                      85
         South River                                     104
         Patuxent River                                  120
         Jackson River                                   128
         Chattahoochee River                             141

 VII     EXPERIMENTAL RESULTS — LABORATORY STUDIES      155
         Pollutant Effects — Pure Substances            156
         Pollutant Effects -- River Waters               170

VIII     HYDRAULIC PROPERTIES RELATED TO REAERATION      175
         Introductory Discussion                         175
         Comparisons with Available Predictive Models    180
         Energy Dissipation Models — Theory             188
         Energy Dissipation Models -- Observed Results   191
              Flint River                                192
              South River                                207
              Patuxent River                             222
              Jackson River                              229
              Chattahoochee River                        234
              Summary Analysis — Five Rivers            241
         Prediction of Stream Reaeration                 245
                            v

-------
Section                                                 Paqe
  IX    ACKNOWLEDGMENTS                                  255




   X    REFERENCES                                       257




  XI    PUBLICATIONS AND PATENTS                         261




 XII    APPENDICES                                       263
                          VI

-------
                         FIGURES                 '


                                                       Page

 1     Molecular Diffusion                               25

 2     Gas Transfer in Stagnant Water                    27

 3     Gas Transfer in Turbulent Water                   29

 4     Typical Cross Section Notes                       49

 5     Rating Curve, Flint River                         50

 6     Average Discharge, Flint River                    51

 7     Average Discharge, South River                    52

 8     Average Discharge, Chattahoochee River            53

 9     Flow Adjustment Procedure                         55

10     Measured Hydraulic Properties, Flint River        57

11     Adjusted Hydraulic Properties, Flint River        58

12     Typical Dye Curves, Flint River                   59

13     Computed and Observed Velocities                  60

14     Summary of Hydraulic Measurement Procedures       61

15     Tracer Release Device (Manual)                    66

16     Field Sampling Arrangement                        69

17     Pressure Pipette                                  71

18     Typical Semilog Plot of Tracer Data               73

19     Pollutant Effects Reactor                         76

20     Typical Semilog Plots of LAS Data                 77

21     Settleability of Tritiated Water—                80
       Experimental Arrangement

22     Flint River Study Locale                          86

23     Channel Profile, Flint River                      91
                           Vll

-------
                                                       Page

24     Observed Krypton Transfer, Flint River            95
       (Upper Study Reaches)

25     Observed Krypton Transfer, Flint River            98
       (Middle Study Reaches)

26     Observed Krypton Transfer, Flint River            99
       (Lower Study Reaches)

27     South River Study Locale                         105

28     Channel Profile, South River                     112

29     Observed Krypton Transfer, South River           117
       (Upper Study Reaches)

30     Observed Krypton Transfer, South River           118
       (Lower Study Reaches)

31     Patuxent River Study Locale                      122

32     Observed Krypton Transfer, Patuxent River        127

33     Jackson River Study Locale                       138

34     Observed Krypton Transfer, Jackson River         139
       (Upper Study Reaches)

35     Observed Krypton Transfer, Jackson River         140
       (Lower Study Reaches)

36     Chattahoochee River Study Locale                 142

37     Water Surface Profile,  Chattahoochee River       148

38     Observed Krypton Transfer, Chattahoochee River   152

39     Typical Pollutant Effects Tests with LAS         157

40     Effect of LAS on K                               160
                         .KIT
41     Effect of LAS on "ALPHA"                         161

42     Typical Pollutant Effects Tests with Mineral Oil 165

43     Effect of Mineral Oil on K,                       167
                                 XlT
44     Effect of Mineral Oil on  "ALPHA"                 168
                          Vlll

-------
                                                        Page

45     Gas Loss vs. Energy Expended, Flint River         193
       (All Results)

46     Gas Remaining vs. Energy Expended, Flint          194
       River (All Results)

47     Gas Loss vs. Energy Expended, Flint River         197
       (Dump XIV)

48     Gas Remaining vs. Energy Expended, Flint          198
       River (Dump XIV)

49     Gas Loss vs. Energy Expended, Flint River         201
       (Mean Values)

50     Gas Remaining vs. Energy Expended, Flint          202
       River (Mean Values)

51     K   vs.  Rate of Energy Expenditure, Flint         203
        OX
       River (All Results)

52     K   vs.  Rate of Energy Expenditure, Flint         205
        OX
       River (Dump XIV)

53     K   vs.  Rate of Energy Expenditure, Flint         206
        ox
       River (Mean Values)

54     Gas Loss vs. Energy Expended, South River         211
       (Lower Reaches)

55     Gas Loss vs. Energy Expended, South River         212
       (Upper Reaches)

56     Gas Loss vs. Energy Expended  , South River        213
       (Dumps VIII and X)

57     Gas Remaining vs. Energy Expended, South          214
       River (Lower Reaches)

58     Gas Remaining vs. Energy Expended, South          215
       River (Upper Reaches)

59     Gas Remaining vs. Energy Expended, South          216
       River (Dumps VIII and X)

60     K   vs.  Rate of Energy Expenditure, South         219
        OX
       River (Lower Reaches)
                           IX

-------
                                                       Page

61     K   vs.  Rate of Energy Expenditure, South        220
        OX
       River (Upper Reaches)

62     K   vs.  Rate of Energy Expenditure, South        221
        OX
       River (Dumps VIII and X)

63     Gas Loss vs. Energy Expended, Patuxent           224
       River

64     Gas Remaining vs. Energy Expended, Patuxent      225
       River

65     K   vs.  Rate of Energy Expenditure, Patuxent     227
        OX
       River

66     Gas Loss vs. Energy Expended, Jackson River      230

67     Gas Remaining vs. Energy Expended, Jackson       231
       River

68     K   vs.  Rate of Energy Expenditure, Jackson      233
        OX
       River

69     Gas Loss vs. Energy Expended, Chattahoochee      237
       River

70     Gas Remaining vs. Energy Expended,               238
       Chattahoochee River

71     K   vs.  Rate of Energy Expenditure,              240
        CJ^k.
       Chattahoochee River
                            x

-------
                         TABLES


No.                                                     Page

 1     Dosing Pattern, Flint River                        87

 2     Typical Hydraulic Properties, Flint River          89

 3     Observed Reaeration Coefficients, Flint River      93

 4     Summary of DO and BOD Results, Flint River        100

 5     DO and BOD at Waterfalls, Flint River             102

 6     Dosing Pattern, South River                       107

 7     Typical Hydraulic Properties, South River         110

 8     Observed Reaeration Coefficients, South River     113

 9     Summary of DO and BOD Results, South River        119

10     DO and BOD at Panola Shoals, South River          120

11     Dosing Pattern, Patuxent River                    123

12     Typical Hydraulic Properties, Patuxent River      124

13     Observed Reaeration Coefficients, Patuxent River  126

14     Jackson River Station Elevations                  130

15     Dosing Pattern, Jackson River                     131

16     Typical Hydraulic Properties, Jackson River       132

17     Observed Reaeration Coefficients, Jackson River   136

18     Dosing Pattern, Chattahoochee River               144

19     Typical Hydraulic Properties, Chattahoochee       146
       River

20     Observed Reaeration Coefficients, Chattahoochee   149
       River

21     Reactor Tests with LAS  (Constant Mixing)          158

22     Reactor Tests with LAS  (Constant LAS)             162
                           XI

-------
No.                                                     Page

23     Reactor Tests with NTA                            163

24     Reactor Tests with Mineral Oil                    166

25     Reaeration Effects of Pollution Loads             171
       (Weekdays)

26     Reaeration Effects of Pollution Loads             172
       (Sundays)

27     Statistical Test of Selected Predictive Models    183

28     Observed Values of the Oxygen Escape Coefficient  248
                           XII

-------
                       SECTION I

            GENERAL SUMMARY AND CONCLUSIONS
The general purpose of the studies reported here has been to
characterize the natural reaeration capacity of nontidal
fresh water streams in terms of the stream hydraulic proper-
ties.  The reaeration capacity controls or limits a stream's
ability to receive and assimilate oxygen-depleting wastes
without serious harm.  This in turn dictates the necessary
degree of waste treatment in any situation, and the asso-
ciated costs.  Hence, the reaeration capacity is a crucial
and valuable natural resource.  In any specific case, it is
a vitally needed waste treatment design parameter, if the
waste treatment facility is to be both adequate and econom-
ical.  Hence, the ability to predict stream reaeration
capacity with accuracy and within acceptable limits of error
has long been an urgent need.

In any stream, the ability to absorb oxygen from the atmo-
sphere, or the reaeration capacity, is a direct function of
the degree of turbulent mixing.  Atmospheric oxygen can be
obtained only at the water surface, and the rate at which
reaeration can take place is therefore directly limited by
the rate of surface water replacement in a flowing stream.
Thus, in a relatively still pool reaeration is a very slow
process, whereas the reaeration capacity of a rapids sec-
tion is very great.  In turn, in any stream the rate of
water surface replacement is controlled by the stream's
physical characteristics, and is related to the associated
hydraulic properties.

Attempts to predict stream reaeration capacity from the hy-
draulic properties date back to 1911, to the work of Black
and Phelps, who proposed a predictive model based upon
stream depth and a "mixing period."  In 1925, Streeter and
Phelps proposed a new model relating the reaeration capacity
to the velocity and depth of flow.  The predictive models
proposed by others since that time have largely followed the
same empirical form.  Until 1966 it was not possible to test
any of the available predictive models against independent
field observations of stream reaeration capacity, as there
was no method of measuring reaeration capacity directly and
independently.  As a result, it was not possible to evaluate
the accuracy or dependability of any of the available pre-
dictive models, or to select the best among them, before
that time.

In 1966 the first direct and independent field measurements
of stream reaeration capacity were made in the Jackson

-------
River, below Covington, West Virginia.  Those measurements
were made by means of a new field tracer procedure developed
by the Federal Water Pollution Control Administration, in
which a radioactive form of the noble gas krypton serves as
a tracer for oxygen.  Limited hydraulic studies were con-
ducted at the same time, and the Jackson River studies
thus also provided the first test of the more commonly used
predictive models.  That test, limited to one stream sec-
tion having a very narrow range of reaeration capacities,
was not conclusive, but did indicate that the two most com-
monly used predictive models provided estimates in the
general range of observed results.

The field tracer procedure for measuring stream reaeration
capacity has since been applied in detailed studies of four
additional streams.  These studies provide the information
presented in subsequent sections of this report.  During the
period 1969-1971 studies were conducted in sections of the
Flint, South, and Chattahoochee Rivers, in Georgia, and the
Patuxent River in Maryland.  In all, including the earlier
Jackson River studies, a total of 323 measurements of the
reaeration rate coefficient, T^2, have been made and are re-
ported here.  They refer to a total of about 70 miles of
stream including rapids, waterfalls, shoals, pools and rel-
atively uniform stream reaches.  In the various field
studies the stream flow has ranged from as little as 5 cfs
to as much as 3,300 cfs, and the river temperature from 10°C
to 35°C; the 5-day BOD of the streams has ranged from about
3 to 30 mg/1, and the observed values of K2 from zero to
15.0/hr.  Thus, the available observed results are regarded
as representative of a great many, probably most, nontidal
fresh water streams.  They have provided the basis for a
new energy dissipation theory of reaeration, as well as the
necessary field data for adequate testing of currently
available predictive models.

It has been known for some time that pollutants such as de-
tergents may affect stream reaeration capacity, and an ad-
ditional purpose of this research has been to develop a pro-
cedure for accurate evaluation of such effects.  A labora-
tory procedure capable of measuring the effect of wastes on
reaeration capacity has been developed, and has been applied
in studies of the effect of pure substances  (LAS, NTA, min-
eral oil) and of the pollution present in river waters.
These studies have proved also to be of considerable assis-
tance in characterizing stream reaeration capacity.

A more detailed, referenced, review of the historical devel-
opment of predictive models for stream reaeration capacity

-------
is provided in Section IV of this report, together with a
detailed discussion of the relationships between turbulence,
mixing and gas transfer.  All of the experimental procedures
are described in Section V, including the gaseous tracer
field method for reaeration measurement, physical (hydrau-
lic) field measurement procedures, and the laboratory system
for testing the effects of pollutants.  All of the observed
field results are presented and discussed in Section VI, and
the pollutant effects results are provided in Section VII.
Section VIII of the report includes the theoretical develop-
ment of the energy dissipation models for predicting reaera-
tion capacity, as well as the detailed results of statisti-
cal testing of these and other available predictive models.
Finally, Section VIII includes recommended models and pro-
cedures for the prediction of stream reaeration capacity.

One additional purpose of this research was to make avail-
able a complete set of data, both reaeration and hydraulic,
for the use of others in conducting further tests of avail-
able models, or refining them, and in developing new pre-
dictive models.  Accordingly, all of the detailed tracer
study and hydraulic study data have been summarized and are
provided in Appendices to this report.

CONCLUSIONS

The following conclusions have been derived from and are
supported by the research studies reported here.

(1)  The reaeration capacity of nontidal fresh water streams
is directly related to the energy expended by the flowing
water.  Reference to the energy equation for flow in open
channels indicates that the reaeration capacity is directly
related to the change of water surface elevation.  The
change of water surface elevation between two stream loca-
tions is just the amount of energy expended in that reach,
in ft-lbs per pound of water; the change in water surface
elevation divided by the time of flow is the average rate
of energy expenditure.

(2)  The reaeration rate coefficient is directly proportion-
al to the rate of energy expenditure in nontidal fresh water
streams.  This basic relationship may be expressed by the
model

                     K  = c(|^.)                          (48)
                      ^     tf


where K2 is the reaeration rate coefficient  (base e) per
hour, tf is the time of flow in hours, Ah is the water sur-
face elevation change in feet, and c is the constant of

-------
proportionality with units per foot.  Other compatible sets
of units may be substituted as desired.

(3)  From equation (48)  and the basic reaeration equation,
the amount or extent of reaeration that takes place in
streams such as those included in these investigations can
be related to the amount of energy expended by the following
model:
                                                        (61)
where Dj is the DO saturation deficit at the upstream end of
a stream section, in mg/1, D2  is the DO saturation deficit,
in mg/1, that would occur at the downstream end if there
were no concurrent sources of oxygen consumption in the
stream section, Ah is the water surface elevation change in
the section, in feet, and c is the same constant of propor-
tionality that occurs in equation (48) .

 (4)  The constant of proportionality, c, that occurs in
equations (48) and (61) is designated the "escape coefficient."
It is the product of two other constants

                      c = (a) x  (b)                       (57)
one of which, a, refers solely to the physical molecular
properties of the diffusing gas and the quality of the water,
and the other, b, refers exclusively to the mixing charac-
teristics and hydraulic properties of the stream.  From
equation  (61) above

                         0.693                           ,c_
                     c =                                 (52)
and the escape coefficient may thus be regarded as a  "half-
height", in feet; the half -height is that amount of water
surface elevation change required for Do to take the  value
0.50 DI.  If the half -height for a particular stream  were
10 feet, and there were no sources of concurrent oxygen  con
sumption, then 50 percent of an initial DO deficit would be
satisfied in 10 feet of fall, 75 percent in 20 feet,  etc.

For any stream section, the escape coefficient is a basic
constant that will remain unchanged unless the degree or
kind of pollution is changed or the hydraulic mixing  regime
is altered significantly.

-------
(5)  The escape coefficient, c, is related to the water
temperature in the same way as K  , namely
            c. (T,«)   K  (T )   C(T)
                  —  ^  "     *^   -L     i/\o«"»>*"in'      /^ro\
            „ /m	T" — r?	7TS	\~ = f^	TS	T" =  1.022   2   1       ID^;
within the range 0°C to 30°C, where T^  and T2  are  the water
temperatures in °C, CS(TI) and C_(T2) are the  respective  DO
saturation limits in mg/1, and the constant  8  =  1.022/°C
has been derived and demonstrated elsewhere  (see Section  IV)
 (6)  The foregoing energy dissipation models  for  stream re-
aeration capacity, equations  (48)  and  (61), have  been  tested
by standard statistical procedures  for each of  the  five
streams studied, with very  satisfactory  results.  In all
cases equation  (61) produced high  correlation (rxy  = 0.90  or
better); equation  (48) produced  comparable results  except  in
streams where the observed  range of K2 was very small.   The
numerical values of the escape coefficient, c,  derived from
separate statistical tests  of equations  (48)  and  (61)  were
in excellent agreement in every  case.

 (7)  The range  of numerical values  taken by the escape co-
efficient, c, is quite small.  For  the five, rivers  studied,
which incorporate a very wide range of stream flow, BOD,
temperature and K2, the range of individual values  of  c was
from 0.0317/ft  to 0.0804/ft at 25°C.  It appears  unlikely
that for any stream that is reasonably well mixed the  es-
cape coefficient will be much less  than  0.030/ft  or much
greater than 0.085/ft at 25°C.

 (8)  From summary analyses  of the  observed results  for all
five rivers taken together, the  single value  c  =  0.054/ft
at 25°C was obtained.  All  of the  observed values of c from
the individual  river studies fall within +_ 50 percent  of
that result.  From those analyses  the models
                 K2 =  0.054  (|ii)                         (59)

and
may be used to estimate  the  reaeration capacity of an "aver
age" or  "typical"  stream that  is  moderately polluted (5-day

-------
BOD in the neighborhood of 15 mg/1) and reasonably well     ^
mixed, at axtemperature of 25°C.

(9)  The observed range of numerical values taken by the
escape coefficient, c, appears to be related primarily to
the degree of pollution of the stream.  Thus, the highest
observed values of c were associated with the least pollu-
tion, measured roughly as 5-day BOD, and vice versa.  How-
ever, the available results do not indicate an exact or ex-
clusive relationship between c and the 5-day BOD, and this
matter requires more extensive investigation.  For the pres-
ent, the predictive equations (58) and (59) may be modified
with caution by substituting a lower value of c  (down to
0.030/ft) for streams that are relatively heavily polluted
(up to about 30 mg/1 of 5-day BOD), or a higher value (up
to c = 0.085/ft) for streams that are relatively lightly
polluted (down to 2 mg/1 or so of 5-day BOD).

 (10)  In sections of stream where the entering flow clearly
does not mix fully with the whole stream volume, as may oc-
cur in stratified pools, reservoirs and estuaries, great
care is necessary in predicting reaeration  capacity at any
time.  Under such hydraulic circumstances,  the entering
stream flow may gain substantially more oxygen than expected
if it tends to remain at the water surface, or it may gain
essentially no oxygen by direct reaeration  if it flows along
the stream bottom.  Such a situation can be unusually sensi-
tive to transient meteorological  and hydrologic  conditions
in terms of their effect on the mixing regime at any time.

 (11)  These studies have indicated that the mixing pools
that occur underneath waterfalls  may add significantly to
the reaeration that would otherwise take place.  Observed
results are for only one such pool, and the available data
are not sufficient to develop an  accurate  relationship, but
indicate that the degree to which such pools enhance reaera-
tion may well be directly related to the ratio of stream
flow  (and energy input) to pool volume, as  well  as  the
specific geometry involved.

 (12)  The available results suggest strongly that stream
flow changes by a factor of two or three do not  usually
significantly modify the stream reaeration capacity.  Flow
changes of that order had no significant effect  upon the
magnitude of K2 in the Flint, South, Patuxent or Chattahoo-
chee Rivers.  A flow increase reduces the  time of flow, but
not usually in the same proportion; this flow increase also
therefore reduces the period in which gas  transfer  may take
place.  Other changes in hydraulic properties, such as the
slope of the water surface, velocity, depth, etc.,  also

-------
occur with changes in discharge.  In essence, changes of
flow by a factor of two or three appear to have little ef-
fect on K2, and, by inference, on the mixing regime.  The
available data are not sufficient to further clarify this
matter.  For the present, however, they indicate clearly
that K2 is not usually significantly affected by flow
changes of the order mentioned above.

(13)  Laboratory tests with the detergent LAS have demon-
strated that LAS reduces the reaeration capacity of water.
In a series of tests at a constant mixing rate, the gas
transfer capacity of distilled water decreased to about 60
percent of the unpolluted magnitude as the LAS concentra-
tion was increased to 12 mg/1.  Additional tests at a con-
stant LAS concentration showed that the gas transfer capac-
ity also decreased as the rate of mixing was increased.
Thus, LAS and similar pollutants appear to bring about the
greatest reduction of reaeration capacity at features such
as rapids and shoals in streams, where reaeration would
otherwise be expected to be at a most beneficial maximum.

(14)  Laboratory tests with mineral oil have demonstrated
that mineral oil enhances the gas transfer capacity of water.
A series of laboratory reactor tests at a constant mixing
rate showed that the gas transfer capacity of distilled
water more than doubled as mineral oil concentration was in-
creased to 400 mg/1.  A small number of tests at a constant
concentration of mineral oil showed that as the mixing rate
was increased the effect on gas transfer capacity diminished.
Concentrations of oil in river waters should be much lower
than these laboratory test concentrations except for major
accidental spills, and these results are therefore not re-
garded as applicable to practical stream reaeration problems
at this time.

(15)  A small number of laboratory tests with pure NTA indi-
cated that NTA itself has little or no effect upon the gas
transfer capacity of water.  In tests with NTA concentra-
tions up to 16 mg/1 no significant change in the gas trans-
fer capacity of distilled water was observed.

(16)  Tests with natural river waters have demonstrated that
the pollution added to streams causes reduction of the
stream reaeration capacity.  For these tests, the gas trans-
fer capacity of "clean" Chattahooch.ee River water  (taken at
the Atlanta Water Works) was compared to that of polluted
river water  (taken below the Clayton STP).  The gas transfer
capacity of South River water samples had to be compared to
that of distilled water, as no relatively "clean" South
River location was available.  The South River water samples

-------
had about a? percent of the gas transfer capability of dis-=>.
tilled water.  On weekdays, the gas transfer capability of
the polluted Chattahoochee water was considerably less than
that of the "clean" Chattahoochee water.  On Sundays, when
pollution of the Chattahoochee is considerably reduced,
there was no significant difference between the gas transfer
capabilities of Chattahoochee River water from the "clean"
and polluted locations.

(17)  Earlier models for predicting stream reaeration capac-
ity from the hydraulic properties have been tested by stan-
dard statistical procedures against the results observed in
this research and reported here.  None of those models
proved capable of predicting stream reaeration capacity
within acceptable limits of error over the range of observed
results.  The differences between predicted and observed
values of K2 were consistently large for highly turbulent
stream reaches and reaches where flow was not uniform.

(18)  The stream hydraulic properties that are related to
reaeration in a primary way are the change in water surface
elevation and the time of flow.  The slope of the water sur-
face, or the change in elevation per unit length, is thus a
primary or causative hydraulic property.  The change in
water surface elevation per unit time, or the rate of energy
dissipation, is a basic dynamic hydraulic property.  It has
been shown by these studies that the range of observation of
the rate of energy dissipation is quite comparable to the
range of associated values of K2, and that the rate of
energy dissipation is a sensitive indicator of stream reaer-
ation over a very wide range of observation.

(19)  Neither the mean forward velocity nor the mean depth
of flow appear to be related to stream reaeration in a pri-
mary way-  The mean velocity results from the slope of the
water surface and the energy gradient, and is a secondary
property in that sense; it has been shown that the range of
observable mean velocities is very much smaller than the
range of associated values of K2, and thus the former can-
not be a highly useful indicator of the latter.  Stream
depth appears also to be a secondary property.  In a turbu-
lent stream there is no apparent reason why depth, per se,
should control or limit reaeration.  As in the case of mean
velocity, the observed range of stream depths has been
shown to be much smaller than the observed range of values
of K2, and depth therefore cannot be a sensitive indicator
of reaeration.

It has been clearly shown in the studies reported here that
most of the real work of gas transfer takes place at abrupt
changes of elevation such as rapids, shoals and falls in

-------
reaches that contain such features.  In such reaches, nei-
ther the mean velocity nor the mean depth of flow can be
regarded as a meaningful or adequate indicator of turbu-
lence or gas transfer.

(20)  Stream bottom roughness is a calculated rather than an
observable or measurable property, and may be calculated
only for quite restrictive conditions of uniform flow.  Many
if not most of the stream sections studied failed to meet
the criteria for uniform flow.  Hence, the bottom roughness
is not regarded as a useful indicator of stream reaeration
capacity.
 r
(21)  It became evident early in the studies reported here
that longitudinal dispersion could not serve as a useful
general indicator of natural stream reaeration capacity.
At hydraulic features such as waterfalls, rapids and shoals,
a great deal of reaeration occurred even though there was
virtually no longitudinal dispersion across those features.
Accordingly, no further analysis seeking a relationship be-
tween reaeration and longitudinal dispersion was conducted.

(22)  The gaseous tracer procedure for direct field measure-
ment of stream reaeration capacity is suitable and effective
for stream flows at least as large as 3,300 cfs, without
significant radiation exposure of field personnel or any
member of the public, and without unduly cumbersome or im-
practical field procedures or equipment.  For the larger
doses required for large stream flows, the field dosing pro-
cedure should be designed to minimize handling the dose and
to minimize the duration of any exposure of the dosing party.
The use of blasting caps for dose release and of downstream
hand sampling procedures for dose assay have proved to be
effective methods for minimizing personnel exposure and ob-
taining the necessary information.  These procedures are
described in more detail in Section V of this report.

 (23)  The reproducibility of results obtained by use of the
gaseous tracer method for stream reaeration capacity has
proved to be generally excellent throughout these studies,
with few exceptions.  In stream reaches where mixing is
very poor and even erratic, the actual reaeration that takes
place at any time will depend upon the momentary mixing sit-
uation, and an individual field measurement of reaeration
therefore may not always reflect "average" conditions.  Such
situations include, for example, pooled reaches that are
subject to thermal stratification wherein the entering
stream flow may either stay at the surface or flow along the
bottom of the pool, depending upon momentary conditions.
Where the magnitude of the reaeration coefficient, K2, is
small and approaches the magnitude of the error associated

-------
with  any individual measurement of K2 , the  accuracy  of  any
individual measurement may be improved by counting larger _
numbers of samples and by extending the counting period; if
necessary, several such measurements of K2  can  also  be  made
to  arrive at a satisfactory mean value.

 (24)   The basic energy dissipation models,  equations (48)
and (61), describe the transfer of many gases,  not just
oxygen, with appropriate modification of the magnitude  of
the escape coefficient.  It has been demonstrated elsewhere
both  theoretically and experimentally that  for  any pair of
different gases
              B      mB      A

 where A  and B denote the two different gases,  K  is  the  gas
 transfer coefficient  (K2) for the specific gas,  Dm  is the
 molecular diffusivity, and d is the molucular  diameter  (see
 Section  IV and related references) .  The escape  coefficient
 is  also  specific for any gas, and is proportional to the
 K2  for that gas.  Hence, equation (6) may be rewritten


              (~) =  (jA =  =  (d^                 (63>
              CB     *B      mB     QA

 where CA and c  are the escape coefficients for  the two
 gases .

 As  an example of the usefulness of equation  (63) , the pre-
 dictive  model for oxygen transfer in a "typical" stream,
.equation (59) , may be readily converted to a predictive
 model for the transfer of carbon dioxide, or nitrogen.
 Using either the ratio of known dif fusivities  for oxygen and
 C02 , or  the ratio of known molecular diameters

                 
-------
                      SECTION II

                    RECOiyiMENDATIONS
The studies reported here and the conclusions derived from
them are subject to certain limitations, as outlined in the
text of this report and in Section I.  The following recom-
mendations, if carried out, should be of considerable as-
sistance in refining and expanding the ability to predict
stream reaeration capacity.

(1)  It is recommended that studies designed to firmly es-
tablish the range of the escape coefficient, c, be under-
taken.  At the one extreme, this will require evaluation
of c for clean natural streams and reaches that receive only
highly treated secondary or tertiary effluents.  At the
other extreme, it will require similar studies in stream
reaches that are heavily polluted by representative commun-
ity and industrial wastes.

(2)  As shown in equation  (57), the escape coefficient, c,
is the product of two other constants, a and b, for any
stream.  The latter is a mixing coefficient, dependent upon
the specific hydraulic properties and mixing characteristics
of the stream.  The former, a, is a molecular coefficient,
related only to the physical molecular properties of oxygen
and the quality of the water.  The coefficient, a, is thus
related to the diffusivity, but is affected by pollutants.
It is recommended that studies be undertaken to establish
the magnitude of the coefficient, a, for clean water.  This
will facilitate evaluation of the effects of pollutants on
a, and will allow separate evaluation of the specific mixing
characteristics of a stream.

(3)  Research studies designed to clarify the basic mecha-
nisms by which pollutants modify the gas transfer capacity
of water are strongly recommended.  The observation reported
here that LAS and mineral oil have opposite effects on K2
demonstrates the importance of such research.  It appears
from these limited studies that entire classes of pollutants
may enhance reaeration capacity, whereas others may reduce
it.  As indicated in Section VII, presently one can only
speculate on the mechanisms involved:  they may include al-
teration of the water surface on a molecular scale, changes
in surface tension, modification of the hydraulic character-
istics of the whole body of fluid, or other factors not
presently envisioned.  Clarification of these mechanisms,
and of the characteristics of pollutants that bring them
into play, should be of particular assistance in refining
the ability to predict stream reaeration capacity.
                           11

-------
(4)   A suitable and accurate test system for measuring the
effect of pollutants on the DO saturation concentration
should be developed.  The reactor test system used here (see
Section VII)  provides for tests of pollutant effect on K2 /
but no studies of the effect on DO saturation have been per-
formed, nor any tests devised for that purpose.  It appears
likely that some pollutants may modify the DO saturation
limit as well as the reaeration rate coefficient, and such
an effect would be of considerable importance in planning
for waste control.

(5)   Detailed investigations of the effects of typical
wastes on stream reaeration capacity are recommended.  Such
information should be of great assistance in planning for
the location and eventual growth of industry, as well as
for assigning economic damages to waste sources and allo-
cating stream self-purification capacity.  The required
studies can be performed in reactor experiments of the type
reported here  (see Section VII), on a somewhat larger scale.
Typical process wastes from paper and textile mills, food
processing plants, chemical production facilities, refin-
eries, etc., should be studied for their effect on K2, with
reference to both waste concentration and degree of turbu-
lent mixing.  Concurrent studies of any effect on the DO
saturation concentration should also be performed.

(6)   Further studies of the specific relationships between
stream flow and reaeration capacity are recommended.
Changes in stream flow modify a number of hydraulic proper-
ties, and in varying degree.  It has been found in the
studies reported here that changes in stream flow by a fac-
tor of two or three had no significant effect on the reaera-
tion rate coefficient, and that implies that flow changes of
that magnitude did not significantly modify the mixing re-
gime.  It is recommended that studies of the relationship
between flow and K2 be undertaken to clarify this matter.
Such investigations should be performed in selected natural
stream channels, and in each case should involve a suffi-
ciently wide range of flow to provide positive results.

(7)   Thorough studies of the relationships between energy
dissipation and reaeration should be conducted for specific
hydraulic features such as waterfalls, mixing pools, hydrau-
lic jumps, spillways, etc.  Such investigations may be
carried out principally and most readily in hydraulic labo-
ratory models, using the tracer method for reaeration capac-
ity, but should also include prototype studies to the ex-
tent necessary to verify scaling factors.  For each specific
feature, energy expenditure and gas transfer capacity should
be interpreted in terms of prominent hydraulic characteris-
tics such as height of free fall, height of hydraulic jump,
                           12

-------
mixing pool retention period, flow  (and energy) to channel
volume ratio, etc.  Such research studies should be of par-
ticular aid in the design of stream hydraulic structures
for maximum reaeration effect, as well as in refining the
ability to predict the reaeration capacity of natural
streams.

(8)  It is recommended that research on the reaeration ca-
pacity of tidal waters and estuaries be performed.  The
studies reported here have been restricted to nontidal
fresh water streams, but do provide some insight regarding
tidal waters.  Tides represent an additional or superimposed
energy input that should, by itself, usually cause addi-
tional mixing, surface water replacement and reaeration.
For example, it appears that at any location the reaeration
capacity may prove to be a direct function of the water sur-
face elevation change associated with high and low tides, as
well as the ratio of fresh water inflow to occupied channel
volume.  The necessary studies may well be carried out
largely in the hydraulics laboratory and in river basin
models such as those located at the Waterways Experiment
Station of the US Army Corps of Engineers, at Vicksburg,
Mississippi, but scaling factors will have to be verified in
the prototypes.  The reaeration capacity of tidal waters and
estuaries is a matter of great economic importance in
coastal regions of the United States, and the proposed re-
search studies should therefore be of considerable practi-
cal value.

(9)  It is recommended that responsible state and federal
agencies develop and undertake long-term programs designed
to obtain and accumulate accurate information regarding
water surface elevation change and time of flow for impor-
tant sections of streams of the United States.  It has been
shown by these studies that water surface elevation change
and time of flow are the basic hydraulic data required for
prediction of natural stream reaeration capacity, and the
recommended program should be of great value for planning
purposes, as well as for solution of immediate waste control
problems.  In the interests of efficiency and economy, the
more critical stream reaches below communities and indus-
tries should be studied early in the program, and the stream
sections studied should be of sufficient length to include
both the critical and recovery zones.  In situations where
secondary waste treatment may not provide adequate protec-
tion of stream oxygen resources, as indicated by the pre-
dicted reaeration capacity, tracer measurement of the re-
aeration capacity at appropriate low stream flows is recom-
mended ,
                            13

-------
(10)   It is recommended that the field data obtained by
Churchill et al5 for TVA streams be restudied in terms of
the energy dissipation models for reaeration capacity,
equations (48)  and (61).   Those data represent the only
other available set of real field observations of recent
times.  Even though they were derived by the indirect oxy-
gen balance method, it appears that the reported values of
K2 should not contain large error for the most part, as
great care was exersized to prevent error due to photosyn-
thesis, BOD, sludge demands,  etc.  Hence, the observed re-
sults may well provide an independent means of studying the
energy dissipation models and of deriving valid estimates of
the escape coefficient for relatively clean natural stream
sections.
                             14

-------
                      SECTION III

                     INTRODUCTION
The ability of a flowing stream to obtain oxygen from the
limitless resources of the atmosphere is the fundamental
process by which the stream is able to "purify itself"
once its dissolved oxygen resources have been depleted.
Without this "reaeration capacity", a stream degraded by
oxygen-depleting wastes could never recover its dissolved
oxygen resources, and the great variety of natural aquatic
life that is taken for granted could not exist.  Thus,
accurate knowledge of the reaeration capacity of a polluted
stream is the necessary basis for determination of waste
treatment requirements, if oxygen resources are to be
adequately protected, and is thereby also the principal
requirement for accurate evaluation of the costs of pollution
control.  For these reasons, attempts to evaluate stream
reaeration capacity date back at least to 1911-'-, and
research on this subject has intensified in recent years as
population has expanded and stream pollution has become a
more widespread and more serious national problem.

The reaeration capacity of a flowing stream depends pri-
marily upon the prevailing degree of turbulence of the
stream.  Oxygen transfer from the atmosphere into the water
can take place only at the air-water interface that exists
at the stream surface, and this interface is constantly
and randomly changing  (being replaced or renewed) due to
turbulent mixing of the flowing water.  Hence, for any
specific degree of oxygen depletion, the rate at which
oxygen can be gained by the flowing water is directly
proportional to the rate at which the water surface is being
replaced from below by turbulent mixing.

Turbulence is a very complex process, and is not as yet
susceptible to independent measurement or evaluation.  As
a result, we have had no independent means of knowing the
rate of water surface renewal in a natural stream, and it
has therefore not been possible to evaluate reaeration
capacity in terms of stream turbulence.  It has thus been
necessary over the years to attempt to evaluate stream
reaeration capacity by the indirect oxygen balance method
of Streeter and Phelps .

Stream self-purification involves two principal processes,
namely:  (a)  the depletion of DO resources by bacterial
degradation of domestic and industrial organic wastes, and
                            15

-------
(b)   the replenishment of the DO resource by absorption of
oxygen from the atmosphere.   Other natural processes modify
the''oxygen balance in a polluted stream or reservoir:  the
anaerobic decomposition of benthal deposits of settleable
organic matter results in a local demand on the DO resources
of the stream; if algae are present in large numbers, they
will add oxygen to the stream by photosynthesis during day-
light hours and will consume DO by respiration during the
dark hours; in some streams, prolific growths of attached
bacterial slimes have a great influence on the oxygen con-
tent of the flowing water; the situation is often further_
complicated by the presence of multiple sources of pollution
and tributary flows. All of these oxygen-influencing pro-
cesses occur simultaneously in a polluted stream, to lesser
or greater degree, and in a specific case any one of them
may dominate the total self-purification process.  Stream
self-purification is thus a very complex process in any real
situation.

In applying the Streeter-Phelps indirect oxygen balance pro-
cedure, an attempt is made to evaluate all of the other pro-
cesses that have influenced an observed stream DO profile,
and then a calculation is made of what the reaeration oxygen
income must have been in order to produce the observed DO
profile.  The approach is much the same as that used in es-
timating the bottom "roughness" of a stream in calculations
related to open channel flow - one cannot obtain a system-
independent direct measure of roughness.

The several oxygen-influencing processes that occur in a
natural stream are not all susceptible to accurate indepen-
dent evaluation, either.  Thus, although the indirect oxygen
balance procedure is entirely logical and valid in concept,
its application incorporates unavoidable errors of assump-
tion, omission and field measurement.  The net result is
that indirectly calculated estimates of reaeration capacity
contain an unknown degree of error, small in some cases and
undoubtedly large in others.  In point of fact, the reaera-
tion rates calculated by the indirect method contain an
error that simply compensates for all of the other errors of
assumption, omission and measurement that have been made.
As a result, it has not been possible to accept such indi-
rect estimates of reaeration as firm or accurate.

Until quite recently, then, it has not been possible to ob-
tain, firm and accurate evaluations of the reaeration capa-
city of a polluted stream - we have not known how to evalu-
ate turbulent mixing, which controls reaeration, and we have
not known how to obtain accurate independent evaluations  of
some of the other oxygen-influencing processes such as photo-
synthesis and bioextraction.
                           16

-------
Faced with the above dilemma, various investigators have at-
tempted over the past 60 years to develop rational mathema-
tical models for the reaeration process itself.  Such models
generally attempt to explain reaeration in terms of turbu-
lence  theory and stream hydraulic properties such as veloc-
ity and depth of flow.  The first such model was provided in
1911 by Black and Phelps, in a report on the pollution of
New York Harbor^.  That model, which attempted to explain
reaeration in terms of molecular diffusion, stream depth and
a "mixing period", is still in use today^.

Since 1911, other attempts have been made to explain reaera-
tion in terms of the hydraulic properties that are associated
with turbulent mixing.  Some of the better known models in-
clude those of Streeter and Phelps (1952)2 , 0'Connor and
Dobbins (1956)4, and Churchill et al (1962)5, all of which
consider reaeration (and turbulence)  to be directly related
to stream velocity and inversely related to stream depth.
Other models include that of Krenkel and Orlob^ who attempted
to explain reaeration in terms of longitudinal dispersion,
and the Thackston model^ which incorporates hydraulic slope
as an additional influencing factor.

All such mathematical models for stream reaeration are re-
ferred to here as predictive models,  rather than indirect,
as their purpose is to predict reaeration independently in
terms of stream hydraulic properties.  In all cases, their
development has been hampered and limited because the only
means of testing the model has been indirect calculation of
the real reaeration income by the questionnable oxygen bal-
ance procedure.  Hence, all of the predictive models must
still be regarded as possible but not proved.

The predictive models for reaeration will be discussed in
greater detail at a later point.  They are regarded as most
important, as they represent the necessary direction of de-
velopment, that is, the explanation of stream reaeration,
and the ability to predict it, in terms of hydraulic prop-
erties.  Thus, although any or all of the predictive models
may prove eventually to be not quite adequate or correct,
all of them provide necessary emphasis and insight into the
important relationships between reaeration, gas transfer and
turbulent mixing in natural streams.

Recognizing the real need for an independent means of evalu-
ating stream reaeration capacity with accuracy and depend-
ability,  in 1964 The Federal Water Pollution Control Admin-
istration began studies to develop such a procedure.  The
result of those studies has been the gaseous tracer proce-
dure that forms the basis of these research studies^.
                           17

-------
The tracer method for reaeration was first demonstrated in
the field in 1966, in studies of self-purification of the
Jackson River below Covington, West Virginia9.  Those field
studies fully demonstrated the techniques and the effective-
ness of the reaeration tracer procedure, and produced the_
first independent observations of stream reaeration capacity -
The gaseous tracer procedure is described in full detail in
a 1967 FWPCA report^-0, which also contains a detailed theo-
retical discussion of the mechanisms of gas transfer in tur-
bulent water systems.  Counting the studies reported here
measurements of reaeration capacity have now been conducted
in six or more inland streams, in a small tidal estuary, and
in a physical model of a tidal stream.
PURPOSES OF THIS RESEARCH

The gaseous tracer procedure for field observation of stream
reaeration capacity now provides the necessary tool for char-
acterizing stream reaeration capacity in terms of the hy-
draulic properties associated with turbulence, and, hence,
for solving the practical problem of predicting reaeration
from field measurements of the relevant stream hydraulic
properties„

Although the tracer method permits highly accurate field eval-
uation of reaeration capacity in specific stream sections,
such observations are not directly extendable to other
streams.  The field method itself is not without certain
limitations and disadvantages related to cost, availability
of special equipment and specially trained personnel, and
radiological safety, and widespread application of the field
tracer procedure by practicing sanitary engineers is regarded
as both unlikely and unnecessary.  In many practical situa-
tions all that is needed is a proved hydraulic model for re-
aeration that provides predictions within acceptable limits
of error.  The general purpose of this research has therefore
been to provide such a predictive model for general use, and
to investigate its limitations and range of error, so that
the field tracer procedure may be reserved for application
in situations where the highest degree of accuracy and de-
pendability is a necessity.

The specific purposes of this research have been as follows:

     (1)  on the basis of direct field measurements of re-
          aeration capacity and hydraulic properties, to
          evaluate the degree of accuracy and the range of
          error normally associated with the various avail-
          able predictive models for reaeration capacity;
                           18

-------
     (2)   by means of direct field tracer and physical
          studies in local streams, to define and evaluate
          the basic relationships between stream reaeration
          capacity and measurable stream hydraulic properties
          such as depth and velocity of flow, slope, dis-
          persion, etc. ;

     (3)   as necessary, to develop modified predictive models,
          or additional models, for predicting stream reaera-
          tion capacity on the basis of measurable hydraulic
          properties;

     (4)   to refine and perfect the field methodology for
          the use of the gaseous tracer procedure in large
          river flows;

     (5)   to develop and demonstrate a standard laboratory
          test procedure for evaluating the actual effects
          of pollutants on stream reaeration capacity, and
          to apply this technique to evaluate the effects of
          various pollutants such as detergents, oils, munic-
          ipal wastes, etc.

As incidental but not negligible purposes, it was planned
that the research program would provide useful and necessary
reaeration data to be of assistance in solving real pollu-
tion problems in the local vicinity, and that it would also
provide a complete set of basic field data on reaeration and
hydraulic properties to be available to other investigators
who might wish to perform independent analysis of the basic
relationships involved.
EXPERIMENTAL PLAN

The general plan of research included concurrent field and
laboratory investigations designed to satisfy the foregoing
purposes to the full extent possible, the field and labora-
tory studies being mutually complementary rather than sepa-
rate project areas.  Field studies in at least two small
(less than 250 cfs) local streams were planned as the initial
project phase, to be followed by similar studies in a larger
stream  (1,000 cfs, or greater) as the second project phase.
It was planned that each field study would include gaseous
tracer evaluation of the reaeration capacity through the
critical pollution zone and into the recovery zone, together
with supporting stream physical and limited oxygen balance
studies.  The streams finally selected for study were the
Flint River below the Atlanta airport, the South River below
the South River Sewage Treatment Plant, and the Chattahoochee
River below the Clayton Sewage Treatment Plant in Atlanta.
                           19

-------
As planned, the gasequs tracer studies employed essentially
the same field procedures as described in earlier reports°/
9'10, with krypton-85 used as the tracer for dissolved oxy-
gen, tritiated water as the dispersion indicator and a fluo-
rescent dye  (rhodamine-WT) for evaluation of time of flow
and longitudinal dispersion.  It was planned that the field
physical studies would be conducted prior to or concurrently
with the tracer studies, and would include flow gaging_ac-
cording to accepted USGS procedures, stream cross-sectioning
 (usually at 500-foot intervals)  and determination of stream
slope.  The resulting data should provide adequate informa-
tion regarding stream flow, hydraulic slope, cross-sectional
area, mean velocity and depth, etc.  Concurrently with the
tracer studies, it was also planned to conduct limited oxy-
gen balance studies for purposes especially of observing any
unusual effects such as might result from the occurrence of
prolific growths of algae or attached bacterial slimes.   It
was planned to conduct all of the field studies during peri-
ods of extended low flow and relatively high stream tempera-
ture, to the extent possible, in order to obtain maximum
relevant information.

The planned laboratory research included studies as necessary
to develop the tracer method further for application to
larger stream flows, and studies to measure the effects of
various pollutants of interest on the reaeration capacity-
The investigations envisioned for extension of the tracer
method to larger flows involved both the possibility of
shielding the necessary tracer doses in lead pigs (and re-
mote dosing, if necessary for personnel protection), and sub-
stitution of a less hazardous radiotracer gas (xenon-133) for
krypton-85 as the tracer for dissolved oxygen.  As the re-
search developed, it became evident that krypton-85 doses as
large as 5.0 or more curies could be used safely without lead
pigs or other such special protection, and that doses of that
magnitude were adequate for measuring gas transfer in flows
up to 3,000 cfs.   Hence, it proved unnecessary to investi-
gate the possible substitution of xenon-133 for krypton-85.

The experimental plans also called for the development of a
standard laboratory reactor test for pollutant effects on re-
aeration,  envisioning a series of reactor tests using stream
receiving water from above the pollution source and, subse-
quently,  water taken from below the source of pollution.  As
the research developed, these plans were modified to include
investigation of the effects of pollutants such as LAS, NTA
and oil,  in order to develop basic information regarding the
effects of such surface active agents and surface contamina-
tion on the gas transfer capacity of water.  Although not
originally planned, other laboratory investigations were also
conducted, including, for example, limited investigations of
                           20

-------
the possibility (suggested by others) that tritiated water
might physically separate from ordinary water because of
settling of the heavier tritiated water molecules.

As is the case with any research project, the initial re-
search plans were modified as research information became
available during the course of the project to the extent
deemed necessary to best satisfy major project purposes.
                            21

-------
                      SECTION IV

                    RELEVANT THEORY
The following summary of theoretical considerations that are
relevant to this research represents the available necessary
theory as of the time when this research was begun.  This
summary has been taken largely from the 1967 FWPCA report on
this subjectlO, and the reader is referred to that report for
specific mathematical derivations and proofs of interest.  No
attempt has been made here to include all of the various the-
ories regarding reaeration that are available in the exten-
sive earlier literature on this subject.  Rather, only di-
rectly relevant and necessary theory is included here.   An
excellent comprehensive 'state-of-the-art' review of all of
the available literature on oxygen transfer in water has been
completed recently by KingH, and the reader is referred to
that report for additional information and references.
MOLECULAR DIFFUSION, MIXING, AND GAS TRANSFER

Reaeration of turbulent water is a purely physical process
that involves entry of the oxygen molecules from the atmo-
sphere into the water at the air-water interface and subse-
quent distribution of this dissolved oxygen throughout the
volume and depth of water.  Reaeration takes place as the
result of the combined effects of molecular diffusion of the
oxygen and physical mixing of the water.  The driving force
for reaeration is the difference between the active partial
pressures of oxygen in the air and in the water.  In a pol-
luted stream, the water is oxygen-deficient, the driving
force is in the air-to-water direction, and this brings about
a net transfer of oxygen from the air to the water.  So long
as such a driving force exists, oxygen transfer will take
place.  When the driving force has disappeared, and there is
no longer any partial pressure difference, the water is said
to be "saturated" with dissolved oxygen.  For example, at
20°C and one atmosphere of air pressure the "saturation lim-
it" for dissolved oxygen in water is 9.17 mg/1, and this is
the maximum attainable DO concentration.  The "saturation
deficit" is the difference between the actual concentration
of the DO in the water and the saturation limit for the ex-
isting temperature and pressure, and is a measure of the
strength of the driving force for reaeration.

Molecular diffusion and physical mixing, or dispersion, are
two quite different processes that complement each other
during reaeration of turbulent water.  As outlined below,
diffusion takes place because of the inherent kinetic energy
                           23

-------
possessed by the oxygen molecules, whereas the dissolved gas
molecules are dispersed, by turbulent mixing, which results
from the application of external forces of one kind or an-
other on the volume elements of water.  The technical liter-
ature is somewhat confusing on this point - for instance, the
commonly-used terms 'eddy diffusion" and 'hydrodiffusion'
really refer to mixing or dispersion of the volume elements
of water, rather than to the molecular diffusion process so
well known in science.

It is also important to bear in mind throughout the following
discussion that the water can obtain additional oxygen only
at the air-water interface, or the water surface.
MOLECULAR DIFFUSION

If a group of dissolved molecules (such as a salt or a gas)
could be placed at some point in a beaker of water, and if
this could be accomplished without disturbing the water, the
dissolved molecules would;  (a) gradually spread out through
the water volume, and (b) eventually achieve a uniform con-
centration throughout the volume of water in the beaker.
They would do this without any movement at all of the water
itself, or in totally quiescent water.  They would do it
because of their own inherent kinetic energy.

Referring to Figure 1, all molecules possess inherent kinetic
energy associated with their surrounding temperature, the
average kinetic energy being just 3/2 kT, where k is the
Boltzmann Constant and T is the absolute temperature.  In
terms of mass and velocity, molecules of a specific mass
will move about with a specific velocity, on the average,
according to the model KE = 1/2 mv .  The dissolved mole-
cules therefore move about more or less as shown by the
arrows in Figure 1, and this motion is entirely random and
takes place in random directions.  It is this movement due
to inherent kinetic energy that allows the dissolved mole-
cules to spread out and eventually achieve uniform concen-
tration in the beaker of water, by molecular diffusion.

Fick's first law of diffusion places molecular diffusion on
a quantitative basis.  Referring to Figure 1, J is the net
flux of molecules  (in mg/cm^/sec) across any plane within

the volume of water; ^£ refers to the concentration gradient
                     dr
across the plane  (dc represents the difference in concentra-
tion of dissolved material on the two sides of the plane, and
dr represents, the infinitesimal distance from one side of the
plane to the other), and is the driving force for diffusion;
Dm is referred to as the coefficient of molecular diffusion,


                             24

-------
          --/:-
                     KE  =   kT
Fick's First Law:
                                         o

        /  /  2 ,   N    ,2        ,mg/cm .
        (mg/cm /sec)  = 'cm /sec)  x '-2/	
                                      cm
        Dm =
RT

N f
 o
(Einstein)
        f = 6rrTlr
                              (Stokes)
Hence,
        Dm - ffT,Tl,r)
and
        dc
        -T— = driving force for diffusion
        dr
                    FIGURE 1



               MOLECULAR DIFFUSION
                   25

-------
and its magnitude depends upon the molecular characteristics
of both the diffusing molecules and the surrounding medium.

In 1905IT Albert Einstein developed an equation for evalua-
tion of the molecular diffusion coefficient, based upon his
studies of the Brownian motion.  Referring to Figure 1, the
diffusion coefficient, D , is seen to be equal to the pro-
duct of the universal gas constant, R, and the absolute tem-
perature, T, divided by Avogadro"s number, No, and a "fric-
tion factor", ft related to the ability of the surrounding
medium to impede the progress of the diffusing molecule.

A little later, Stokes further defined the friction factor,
f, for spherical particles falling freely through water,
and showed the friction factor to be directly proportional
to the viscosity, r\, of the medium and the radius, r, of the
falling sphere.  Hence, the diffusion coefficient, D , is
seen to be a function of the absolute temperature, the vis-
cosity of the fluid and the size of the diffusing particle.

Now a word about gas molecules, and regarding them as spheres.
If we could take a single oxygen molecule and set it down on
a table, and hold it still, it would not look like a sphere.
Presumably, this diatomic molecule might look something like
a dumbbell.  However, one cannot set it down on a table and
hold it still long enough to look at it, because this single
molecule is constantly in motion.  First, it has what we
call "spin", and it spins like a top about an axis; .secondly,
the axis itself "precesses", as though the top were wobbling,
about some other axis; thirdly, the molecule possesses "di~
pole moments" and "quadrupole moments" related to the move-
ment of the atoms with respect to each other.  The combined
effect of all of these motions is to make the molecule behave
like a sphere, even though it wouldn't look like one if it
could sit still on a table.  Nor will the effective diameter
of the operating sphere be the same as the length of the
quiet dumbbell.  The effective diameter of the spherical gas
molecule is of the order of angstrom units (1 angstrom =
10~8cm).

To summarize, then,  molecular diffusion takes place because
of the inherent kinetic energy of the diffusing molecules and
in proportion to the magnitude of the existing concentration
gradient; the diffusion coefficient is a function of the ab-
solute temperature,  the viscosity of the fluid medium and the
size of the diffusing molecules.

Figure 2 illustrates the mechanics of gas transfer in com-
pletely quiescent water.  The water is completely still,
there being no temperature gradients, convection currents, or
other motion of volume elements of water.  (Although such a


                          26

-------
           \\\ mm A
            ++-\
                       *
1
                                Ah
T
en c
re O
0)
> -P
u) oC o
K C 2
•H O
D U
T \
Ac
T _^
.


^s^
Ah



il>— ^-
1
1
h
hl h2
                depth
       (C  - C )  = AC = very small
      .'. J ^   Dm () = very small
               FIGURE 2


      GAS TRANSFER IN STAGNANT WATER
                 27

-------
system might well be impossible to achieve experimentally,
the concept is valid and suitable for our purposes here) .
Initially, there is no dissolved oxygen at all in the water,
so that initially oxygen molecules move only into the water
from the overlying atmosphere.

A little later, there will be available dissolved oxygen
molecules in the upper water layer near the water surface;
they also are in constant movement due to their inherent ki-
netic energy, and they move in random directions.  Some of
them escape again to the overlying atmosphere, while others
diffuse to deeper water layers.  However, oxygen molecules
are able to enter the topmost water layer from the overlying
atmosphere more easily than they are able to diffuse down-
ward through the fluid medium.  As a result, the dissolved
gas molecules accumulate fairly rapidly in the uppermost
water layers, and those layers become "saturated."

At any time after the start of the experiment, the net rate
of entry of gas molecules at the air-water interface is just
the rate of entry from above  (constant, because the overlying
atmosphere has constant oxygen concentration) minus the rate
of escape back to the atmosphere (proportional to the dis-
solved oxygen concentration in the uppermost water layer)  «
As a result of the relatively rapid accumulation of gas
molecules in the topmost water layer, the net rate of entry
(or, reaeration) soon becomes very small.

As a result, the deeper water layers soon become "starved"
for oxygen molecules.  Referring to Figure 2, across any  in-
finitesimal distance (depth) Ah, the dissolved oxygen con-
centration difference, AC = (C^ ~ ^2^ ' ^s infinitesimally
small.  Hence, at any depth and at any time, the driving
force for molecular diffusion, the concentration gradient
 AC
    f is very small.  Referring back, then, to Pick's law,
diffusion of oxygen molecules downward is very slow, and re-
aeration of truly stagnant water is a very slow process that
requires days or weeks before the bottom layers of water
approach DO saturation.  The whole process is slow because
of the blocking action of molecular diffusion.
TURBULENT MIXING

Consider now the same beaker of water, but no longer quies-
cent.  Instead, the water is being mixed by some external
force  (perhaps the beaker is being stirred, or is sitting on
a vibrating platform).  We will be concerned now primarily
with volume elements of water, rather than with molecules of
oxygen.  We define a volume element to be infinitesimally


                          28

-------
      (C1  - C  )  = AC  = large
        J_    ^
     /. J ^  - Dm    •=  large
            FIGURE  3
GAS TRANSFER IN TURBULENT WATER

-------
small in the calculus sense, but large enough to contain a
very large number of molecules.

Referring to Figure 3, at the start of our experiment the
water contains no dissolved oxygen.  Volume element No. 1
moves up to the water surface from below and remains there
for a definite, if very small, period of time.  Because it
contained no dissolved oxygen, the net rate of entry of gas
molecules from the overlying atmosphere is very large - at a
maximum - and the volume element gains a relatively large
amount of dissolved oxygen before it leaves the surface to
move downward to a deeper location.  In its downward path it
encounters a second volume element of water, No. 2, that has
never been at the water surface and so contains very little
or no dissolved oxygen.  Thus, the one volume element con-
tains quite a large amount of dissolved oxygen compared to
the other, and at the interface between them there is a
large concentration difference, AC = (Ci - C2) .  Hence, for
that moment across that interface, the driving force for mo-
                    AC
lecular diffusion, (^), is relatively large, and the trans-
fer of dissolved gas molecules from the one volume element to
the other is relatively rapid.

If we now multiply this example by all of the volume elements
of water in the beaker, it is clear that mixing greatly speeds
the reaeration process.  The water surface is constantly re-
placed by volume elements from below, and hence the blocking
action of molecular diffusion is no lonaer present.  The
lower water depths are no more starved  for dissolved oxygen
than the upper locations,  The average concentration of dis-
solved oxygen is at any time the same at all depths and all
locations, including the surface, in a homogeneously mixed
system, and, hence, the net rate of entry of gas molecules at
the water surface remains relatively large until the whole
volume of water approaches the DO saturation concentration.
Note also that the dissolved oxygen concentration gradient
does not now occur in any preferred direction, such as down-
ward.  Instead, there is an average concentration gradient
throughout the whole volume of water, and it is multidirec-
tional.

It is  also clear that the faster the water is mixed, and the
surface replaced, the faster will be the reaeration process.
Instead of days or longer, the water can be saturated with
dissolved oxygen in minutes at high rates of mix.  Thus,
molecular diffusion keeps up with mixing in the turbulent
system, instead of blocking reaeration.  It is also important
to note that in the mixed system the depth of water has no-
thing  to do with the rate of reaeration except insofar as the
depth-to-volume ratio influences the physical rate of water
surface replacement.

                            30

-------
So far as reaeration is concerned, then, the term "turbulence"
has a special meaning relating strictly to the rate of water
surface replacement and to the dispersion of volume elements
of water.  Turbulent mixing of the water and consequent dis-
persion of the dissolved gas molecules takes place due to
the application of external forces, such as the platform
vibration, or a mechanical stirrer, etc.  It enhances mole-
cular diffusion and reaeration as outlined above.

Misconceptions.  The foregoing outline of the fundamental
mechanisms of gas transfer in turbulent water systems indi-
cates that certain widely held concepts of gas transfer are
not, in fact, correct representations of the physical facts.
In the first place, in a well-mixed system, or in a turbulent
natural stream, the surface water layer is not saturated with
dissolved oxygen - constant surface replacement precludes
this.  Also, as indicated above, there is no preferred direc-
tion of oxygen transfer, such as downward, and the physical
depth of a water-course influences reaeration only to the
extent that it influences the rate of water surface replace-
ment in the hydraulic sense.

In particular, in a homogeneously mixed system no stagnant
surface water "film" can exist for any finite period of time,
and, hence, even though it may be an adequate mathematical
convenience in some situations, the "film theory" of gas
transfer is y'rong in concept.  The film theory denies the
obvious fast of physical surface water replacement, and is
based upon the false supposition that a dissolved oxygen
concentration gradient is not present within a well-mixed
system - as has been seen, such a concentration gradient is
the driving force for diffusion, and it exists everywhere
within the unsaturated fluid volume.

A clear distinction must therefore also be made between phys-
ically impossible stagnant surface water films and physically
real hydrodynamic upper layers of water in a system that is
not homogeneously mixed.  For example, in a stratified res-
ervoir the whole volume of water is physically or hydrody-
namically separated into two distinct regions - the lower
region has little opportunity for reaeration because its
volume elements never reach the air-water interface.  In that
case, then, the hydrodynamic situation prevents surface re-
placement and reaeration is very slow.  However, this has to
do with the hydraulic properties of the system, and has no-
thing to do with the film theory.
MATHEJXIATICAL RELATIONSHIPS

Based upon the foregoing considerations of molecular


                           31

-------
diffusion and turbulent mixing, the following mathematical
relationships have been derived and demonstrated experi-
mentally.  The relationships are simply reproduced here, the
reader being referred to the detailed 1967 FWPCA report for
the derivations and the supporting experimental evidence. ^0

Referring to the definitions given earlier, for a turbulent
water system we can write


                  D = (C  - C)                           (1)
                        o


where D is the saturation deficit, C_ is the saturation limit,
                                    O
and C is the momentary average DO, all expressed as mg/1 of
water.

The familiar basic reaeration equation has been derived else-
where from simple first principles^
                    _
                 dt - ~K2D


and states simply that the rate of change of the saturation
deficit  (the driving force) at any time is proportional to
the deficit at that time, or^ the greater the saturation def-
icit the greater the rate of reaeration.  The proportionality
constant, K2, is the "reaeration rate coefficient" for the
specific set of conditions, and its numerical magnitude de-
pends, in particular, upon the degree of turbulent mixing of
the water.

Equation  (2) is solved by


                 D = D e~K2t                              (3)
                      o


where D  is the initial dissolved oxygen deficit  (at t = 0) .

It has also been shown elsewhere-^ that


                 K2 = a n |                               (4)


in a turbulent water system, which states that the reaeration
rate coefficient, K2, is directly proportional to the rate  of
surface  replacement.  In equation  (4) , A is the exposed sur-
face area of water in cm2, v is the whole volume of water in
                            32

-------
cm  and n is the number of new surfaces exposed per unit
time.  Thus the product (n^) is just the area of new surface
exposed in cm2 per unit time and per unit volume.  The pro-
portionality constant, a, is directly related to the coeffi-
cient of molecular diffusion, Dm, and is therefore a function
only of the molecular properties of the dissolved gas and the
physical properties of the water.  The coefficient, a, is
thus a constant for oxygen in clean water at any fixed tem-
perature, but it will be a function of the water temperature
and it may also be modified by the presence of pollutants.
It is not a function of the degree of turbulence.

It should be noted that the quotient (A/V) is properly re-
garded as the reciprocal of the whole depth of water only
under conditions of complete homogeneous mixing.  Thus, for
example, the whole depth of water in a stratified reservoir
is meaningless as a measure of &2 or reaeration capacity.
In point of fact, it is probable that many natural water-
courses, especially large slow-moving rivers, are not homo-
geneously mixed, and in such cases the average depth of flow
is not a measure of the depth that is effective in terms of
surface replacement or reaeration.

With appropriate modification, the foregoing expressions also
describe the absorption or desorption of other gases.  Speci-
fically, consider a dissolved tracer gas, krypton-85 which
has been added to the water.  The amount of krypton-85 pres-
ent  in the atmosphere above the water can be taken to be
zero, for practical purposes.  Hence, the driving force for
gas transfer will be just the partial pressure of the dis-
solved krypton-85 in the water, transfer will be from the
water to the atmosphere, and the tracer gas will be steadily
lost from the water.

Thus, in the case of desorption of the tracer gas we can
write


                     C = CQe~K2t                          (5)


where C is the concentration of the dissolved tracer gas re-
maining in the water at time, t, Co is the concentration at
t = 0, and K2 is the gas transfer coefficient for the tracer
gas.

(At this point, in order to avoid confusion, we will no
longer use the subscript "2" to denote reaeration or gas
transfer.  Instead, the constant K will represent the gas
transfer coefficient for any gas, including oxygen, and the
subscript used will identify the specific gas.  For instance,
                           33

-------
       will  be understood to mean the K2 value for oxygen,
 and "Kkr" will identify the K2 value for krypton, etc.).

 As has been shown  in the earlier worklO, different gases have
 different transfer coefficients because of the different
 molecular characteristics of the gases.  Thus, referring to
 equation (4),  the  numerical value of the constant, a, will
 be different for different gases, under identical hydraulic
 conditions.   In an extended series of experiments involving
 a number of different gases (hydrogen, helium, nitrogen,
 carbon dioxide, oxygen, radon and krypton) , and also on theo-
 retical grounds, it has been demonstrated that for any pair
 of different gases
                  B      mB      A

 Equation  (6) states that the ratio of gas transfer coeffi-
 cients  for  two different gases is just equal to the ratio of
 molecular dif fusivities , and that both ratios are equal to
 the  inverse ratio of their molecular sizes, under identical
 conditions  of turbulent mixing.  In other terms, the larger
 the  gas molecule, the lower its relative gas transfer ability.

 It has  thus been shown, both experimentally and theoretically,
 that for the same conditions of turbulence

                 K,
                 (iT^  = 0.83 - 0.04                       (7)
                  ox

 and  this provides the necessary basis for the use of krypton-
 85 as a tracer for dissolved oxygen in stream studies.

 It should also be noted that the numerical constant, 0.83,
 given in equation (7)  has been demonstrated to be independent
 of the degree of turbulent mixing, independent of the direc-
 tions in which the two gases happen to be moving, and inde-
pendent of temperature within the range 10 to 30°C.

Although the ratio of transfer coefficients for two different
gases is independent of temperature for practical purposes
within the temperature range of interest, it has long been
recognized that the magnitudes of the individual coefficients
 are affected by water temperature.  Specifically, the numeri-
cal magnitude of any value of K increases with increasing
temperature, and vice versa.  Also, for any gas, the satura-
 tion limit decreases with increasing temperature, and vice
versa.   The exact magnitude of the temperature effect on K
has remained a subject of considerable controversy until


                           34

-------
It has been shown in the earlier FWPCA report   that on
quite recently.

It has been she
theoretical grounds


                K(.T2) _ C.S(.TI)
                        Cs(T2)
                                                          (8)
where T-^ and T~ are two different water temperatures, K(T^)
and K(T2) are the respective values of K, and CS(T1) and
CS(T2)  are the respective values of the saturation limit.
Combining equation  (8) with the usual expression for the
temperature effect on K
               K(T2)   C  (T )     /T  _ T )
                  ^  —  °  •*•  _ o \ •*• o    i'                i en
                     -       -- 621                 (9)
where 6 is the commonly applied temperature coefficient.

Using the known dissolved oxygen saturation concentrations
in clean water at standard atmospheric pressure for tempera-
tures from 0°C to 30 °C,12 it has been found that
                6mean = 1'022   0'004


Exactly the same value was found for water containing  large
quantities of chlorides^.

The foregoing predicted mean value for 0 of 1.022 agrees ex-
cellently with the value of 1.024 observed experimentally by
Churchill et al^.  From this combination of theorylO,  experi-
mentally determined values of CS12, and experimentally ob-
served values of 65, the value 6 = 1.022 * 0.004 is taken to
be firmly established.
FIELD TRACER APPLICATION

The mathematical basis of the field tracer technique  for
determining the reaeration rate coefficient,  Kox,  in  natural
streams has been outlined in detail in earlier publications^
10, but is also summarized here for the  sake  of  completeness
and immediate reference.
                           35

-------
Let A and B represent two points along the course of a stream,
with A being the upstream location, and let a quantity of dis*-
solved krypton-85 be introduced at a location upstream from A.
If this tracer dose could be introduced in such a way as to
be uniform in concentration across the stream section, and if
we then collected samples from A and B at the moment of max-
imum tracer concentration at each location and analyzed
these samples for krypton-85, then the numerical value of
Kkr ^or t*le reac-k (AB) could be obtained directly by the use
of equation (5) :


                    ,CB.     -K,  t
                    (TT-)  = e  kr
where C^ and CB are the maximum dissolved krypton-85 concen-
trations at A and B, t is the time of flow between A and B,
and the ratio (Cg/C^) is just the decimal fraction of tracer
gas remaining at point B.  The reaeration rate coefficient,
Kox, for that reach could then be obtained directly from
equation (7) .

In practice, it is not really possible to achieve a truly
uniform cross-sectional tracer dose.  Also, if we only use
the dissolved tracer gas, we have no way of knowing exactly
when the maximum concentration occurs at A and B.  However,
these problems may be solved by the use of additional tracers.

In brief, the field application involves the simultaneous
release of a homogeneous mixture of three tracers.  This
tracer dose is essentially instantaneous, and can be a
point dose in the stream cross-section.  The three tracers
are dissolved krypton-85, tritium in the form of tritiated
water molecules, and a fluorescent dye.  At downstream
sampling locations the dye concentration is monitored contin-
uously by the use of a continuous flow recording fluorometer,
and suitable samples of river water are collected at frequent
intervals as the peak dye concentration approaches and passes
the sampling point.  Samples representative of the maximum
dye concentration are subsequently analyzed for  tritium and
krypton-85.

The fluorescent dye performs two functions:  it indicates
when to sample for the two invisible radioactive tracers, and
it provides an accurate measure of the time of flow between
sampling stations.  The tritiated water provides an accurate
measure of dispersion of the tracer dose:  the concentration
of tritium decreases between sampling locations because the
tracer is dispersed as the result of longitudinal, lateral
and vertical mixing, and, being in the form of water mole-
cules, tritium is not adsorbed on the stream bed or otherwise


                           36

-------
lost in any significant amount.  Because the three tracers
were released simultaneously, the dissolved krypton-85 under-
goes exactly the same dispersion as the tritiated water; in
addition, it is lost to the atmosphere according to the fore-
going models, but, being chemically inert, it suffers no
other losses of any significance.

Under these test conditions, the observed concentrations of
tritium provide an accurate correction for the effects of
dispersion, and hence the decimal fraction of tracer gas re-
maining at point B is just
                                                         (11)
where  (Cjcr/Ctr)A B  are  the  concentration  ratios of krypton-85
and tritium  in  t&e  samples  taken  at the time of the dye peaks
at A and B,  and t is  the  time of  flow between  the two  loca-
tions.  Hence,  an accurate  evaluation of  K-^r can be made  for
any such reach  of stream, and the conversion to a Kox  can
then be made using  equation (7).

The krypton-85  and  tritium  concentrations in the river water
samples are  measured  by a liquid  scintillation counting tech-
nique, for maximum  counting efficiency.   The tracer doses are
quite  small  in  relation to  permissible environmental limits,
and the exposure of personnel handling the tracer dose is
minimal and  controllable.   The  field and  laboratory proce-
dures  have been described in detail in earlier publications^i
10.

As indicated earlier, although  the ratio  of gas transfer  co-
efficients,  (Kkr/Kox) '  -"-s not affected by water temperature
in the range of interest, the individual  coefficients  are
affected.  Hence, for comparing values of Kox  in different
streams, and for seeking  relationships between KQX and stream
hydraulic properties, the field-determined values of Kox  are
converted to a  common temperature by the  use of equations (9)
and  (10).
AVERAGE NATURE OF K

The available procedures  for  estimating  the  reaeration capa-
city of natural  streams generally  require the  assumption of
uniform mixing and turbulence over relatively  long stream

                            37

-------
is
reaches.  in particular,  they treat reaeration as a single
first-order process over  substantial distances, as no other
practical alternative has been available.  Cautiously ap-
plied, this approach does not necessarily lead to large er-
ror in estimates of the reaeration capacity,  However/ it i
decidedly an averaging procedure, rather than an exact one,
as the actual reaeration  rate coefficient, like the actual
rate of water surface replacement, undoubtedly varies sub-
stantially within most such river reaches.  As will be seen
in the subsequent discussion of Experimental Results, in
many streams most of the  real action of gas transfer takes
place in short times of flow.

The tracer method described here and applied in this research
requires no such assumptions as to uniformity of mixing or
turbulence, or constancy  of the reaeration coefficient.  It
provides a direct and quite independent measure of the gas
exchange capacity under existing conditions of mixing and
surface water replacement, whatever they may be.
PREDICTIVE MODELS

A number of theoretical and empirical models for the reaera-
tion coefficient have been proposed.   A recent publication •*•
resulting from a literature search and a state-of-the-art re-
view on the topic of reaeration of streams and reservoirs in-
cludes a discussion of many of these  models.  In addition, in
an earlier report on the use of tracers for the measurement
of reaerationl° several of the models in current use were dis-
cussed.  With these more detailed accounts of previous inves-
tigations readily available, it is not necessary to again
present a complete analysis of the models previously devel-
oped.  However, a comparison between  the values of k2 predic-
ted by several of these models and values determined by the
tracer method is included in one of the following sections of
this report, and for this reason outlines of several of these
models are presented below.

Streeter and Phelps2.  One of the earliest models of stream
self-purification was developed by Streeter and Phelps from
data collected on the Ohio River.  These authors reasoned
that the rate of reaeration is influenced by the hydraulic
characteristics of the stream.  Their model combined the con-
cepts of molecular diffusion and turbulence.  It had been
shown earlier in a model developed by Black and Phelps^ that
when the reaeration process is governed solely by molecular
diffusion the reaeration rate coefficient is inversely pro-
portional to the square of the depth.  To this model Streeter
and Phelps added their concept of the relation between turbu-
lence and stream velocity.  They stated that "under uniform


                           38

-------
physical conditions Cturbulence^ might be expected to be a
power function of the velocity of the form;


                       T = cVn                           (12)
the constants c and n defining the stream type as regards the
fixed physical conditions, such as slope, character of the
bottom, depth, shape, and direction of channel, etc."  By
combining these concepts they formulated the empirical model


                     K? = *£                             (13)
                      ^   ir


where H is the depth, V the velocity of flow, and c and n are
constants that must be evaluated empirically for any specific
stream section.  Streeter and Phelps evaluated these constants
by an indirect method on a number of reaches of the Ohio River,
This process required the estimation of the magnitude of the
effect of each of the stream phenomena that influence the con-
centration of dissolved oxygen, and the subsequent insertion
of these values into the oxygen sag equation from which the
value of &2 was then computed.  Mean values of n determined in
this manner ranged from 0.57 to 5.40, while the values of c
varied from 0.23 to 131.  These investigators suggested that
such wide variations in these coefficients are to be expected
and that further studies would doubtlessly disclose that all
streams would not follow their model for 1^ .
                     4
O ' Connor and Dobbins .   In 1958 O'Connor and Dobbins, working
wTth the idealized oxygen sag equation developed and used by
Streeter and Phelps, developed models for the reaeration co-
efficient which were based on certain theoretical concepts
from the field of fluid turbulence and on assumed relation-
ships defining the rate of surface renewal.  Two flow cate-
gories were specified and an equation was developed for each
category.  The equation

                                 /2
                          480 D
                     k2 =
                              rl
was suggested for use in streams characterized by non-isotro-
pic turbulence.  In this model S is the slope of the stream
channel, DL is the coefficient of molecular diffusivity, and
H is the average depth.  For streams in which turbulence ap-
proaches an isotropic condition, O'Connor and Dobbins
                            39

-------
 approximated  the  reaeration coefficient by the relationship

                          D l/2yl/2
                                 .„„                       (15)
                                 3/2
                           2.31 H
 where  V is  the mean velocity of flow-

 The  verification of these models was based on both field  and
 laboratory  data.  The laboratory data were generated by an
 apparatus consisting of a lattice work moving in simple har-
 monic  motion  in a 5 1/2-inch diameter cylinder containing ap-
 proximately 2,500 ml. of water.  The field data were selected
 from results  collected by other earlier investigators, who
 computed k2 values for the various streams by the indirect
 method.  Much of the field data on hydraulic characteristics
 which  pertained to these earlier investigations was not in
 the  form required by the O'Connor-Dobbins model, and it was
 necessary for O'Connor and Dobbins to make some rather rough
 estimates regarding such variables as mean velocity and depth
 of flow.  These authors reported good agreement between the
 coefficients  calculated by others from earlier river surveys
 and  those calculated by the O'Connor-Dobbins formulae.  In
 addition, O'Connor and Dobbins were of the opinion that their
 laboratory  experiments substantiated their theoretical devel-
 opments .

 Krenkel  and Orlob.   The importance of various factors in the
 reaeration  process were discussed by Krenkel and Orlob.   Con-
 clusions were drawn regarding the effects of oxygen deficit,
 are a-to ^volume ratio, time of exposure of water elements  at
 the  free surface, temperature, partial pressure of solute gas,
 molecular diffusivity, energy dissipation and turbulent dif-
 fusion.  They suggested that the overall mixing characteris-
 tics of a stream section are reflected in a "longitudinal
 mixing coefficient," DL ,defined by the equation
            9t
where c is the concentration of a tracer injected into_a
stream, t is time, x is distance along the stream and u is
the mean stream velocity.

Experiments were conducted in a laboratory flume in order to
examine the relationship between k2 and DL.  Oxygen was first
                           40

-------
        from the water by the addition of sodium sulfite, and
the reaeration rate coefficient was determined by measuring
the concentration profile of a tracer material which affected
the conductivity of the water.

After correcting all test values to a constant temperature of
20°C it was found that the temperature adjusted values of the
reaeration coefficient k21 could be fitted best by a regres-
sion equation of the following form


             k-(20°C) =  (1.138 x 10"5) D,.1*321 H~2'32    (17)
              £                         \ j
in which H is the average depth.  In addition, Krenkel and
Orlob stated that they expected to find the reaeration co-
efficient to be directly proportional to the energy expendi-
ture per unit mass of fluid and that some measure of the size
of turbulent eddies should be effective in the reaeration pro-
cess.  They hypothesized a relation of the form


             k2(20°C) =  (constant) EbHC                  (18)


where E is the energy dissipation per unit mass of fluid, com-
puted from the relation
                     E = u S g                            (19)


in which S is the slope of the energy gradient, g is the grav-
itational constant, and u is the mean stream velocity-  The
constants were evaluated by fitting the flume data, and the
resulting equation was


            k2(20
-------
 techniques were applied to analyze the data.  Variables  in-
 cluded  in the regression analysis were flow/ velocity, mean
 depth,  energy slope, resistance coefficient, fluid_ density,
 fluid viscosity, surface tension, molecular diffusion  co-
 efficient, and a vertical diffusion coefficient.

 Nineteen different combinations of these variables were  ana-
 lyzed,  and the nineteen equations which resulted were  presen-
 ted  and discussed.  The recommended equation had the form


           k9(20°C) = 5.026 V°-969 IT1-673                (21)
             ^4

 where V is the mean stream velocity and H is the mean  depth.
 The  coefficient of multiple correlation for this equation,  as
 determined by the data collected by Churchill et al.,  was
 0.822,  while the correlation coefficients for the other  eigh-
 teen equations ranged from a low of 0.805 to a high of 0.846.
 Thus, any one of the nineteen equations developed by Churchill,
 Elmore, and  Buckingham is essentially as good a predictor as
 any  other.   Since there was no significant improvement in the
 equations when terms other than V and H were used, it  was sug-
 gested  that  the equation


               k2(20°C) = 5 V H~5//3                       (22)


 be used, and that the simplification of the constants  in this
 form, as compared to the values in the regression equation,
 would not significantly affect the predictions in most appli-
 cations .

 Churchill's data have been used by several investigators for
 the development of additional models of k2 .  Isaacs and
 Maag-1  employed these data to evaluate a model of a form simi-
 lar to that recommended by Churchill.  Their equation  is


                k2 = C 00                                (23)
where V - mean stream velocity, H - mean depth, C = constant,
0S = a non-dimensional variable which varies with the channel
geometry, and 0V = a non-dimensional variable which is a mea-
sure of the surface velocity.  The coefficient for channel
shape was added by Isaacs and Maag because they believed the
deviations observed between field and laboratory measurements
                           42

-------
of the reaeration coefficient possibly represented, the depar-
ture in the field from the constant prismatic channel section
used in the laboratory studies.  In addition, they noted that
most theories describing the reaeration process consider con-
ditions at the stream surface to be highly important.  Based
on their use of Churchill's data an equation of the form


                k  = 2.98 00                            (24)
was  recommended.  The average value of 0S was found to be
1.078 and the average for 0V was 1.16.  A correlation coef-
ficient of 0.989 was reported.

Langbein and Durum   used Churchill's field data together with
the field data collected by others  (such as by Streeter and
Phelps) and reported by O'Connor and Dobbins, plus the labora-
tory data of Krenkel and Orlob and Streeter, Wright, and
Kehr1  .  These data, when plotted on the same graph, indicated
that an equation of the form


               k0 = 3.3 V H"1'33                          (25)
would accommodate both the laboratory and field data.  The
failure of Langbein and Durum to include all of Churchill's
data in their analysis was pointed out by Isaacs and Maag   .

Owens, Edwards, and Gibbs   also used data collected by
Churchill as well as the field data of Gameson, Truesdale,  and
Downing^''7 and by regression analysis arrived at the equation


           k2  (20°C) = 9.41V0'67 H~1<85                  (26)


where U is the mean velocity and H is the mean depth.

Thackstori and Krenkel .  The basic premise adopted by Thacks-
ton and Krenkel is one that was originally proposed by Kren-
kel18, i.e., that k2 is proportional to k/H2 where ky is the
vertical mass transfer coefficient and H is the depth of  flow.
In their paper they derive an expression for ky based on  the
mechanics of open channel flow.  An outline of the steps  fol-
lowed by Thackston and Krenkel to estimate the vertical mass
transfer coefficient is presented in the following paragraph.
                            43

-------
 The mean  transfer  coefficient was  assumed to.be proportional
 to the momentum,  transfer  coef f icient e y defined by
                      ^  e
                         y dy
                                                          (27)
 where  T  is  the  fluid  shear  stress,  p  is  the  density,  and u
 the local velocity at  depth  y-  Measurements19  of ky/'ande  in
 the laboratory showed  that neither  cy  nor ky  become zero at
 the free  surface although a  zero value is predicted by the
 use of von Karman's  logarithmic velocity  distribution.  The
 variation of both of these values with depth  was  shown ex-
 perimentally to  be almost identical.   The variation of ev in
 the vertical direction was shown to  follow quite  closely a
 relation  based on Vanoni ' s modification of the  von Karman
 universal velocity distribution.  This relation is


               ey =  K  u* y  (1  - J)                       (28)


 where K is the von Karman coefficient,  UA is  the  shear velo-
 city, and y  is the vertical  coordinate.   Thus,  from equation
 (28) the  average value of e   is  (K/6)  H U* .

 If it is  assumed that  the value of kv  is  proportional  to the
 value of  e  at any value of  the relative  depth  y/H, then it
 follows that the value of ky at the  surface is



           ky = Cl f  H  u* = C2  H u*                        <29>


 Returning to the  basic assumption of Thackston  and Krenkel,
 i.e.,  that
k
                       2   "                               (30)
then the equation for k2 becomes
            ko   =   °4 C2 H **     «  C5  U*               (31)

                     H2                  S~


where the constant €5 must be evaluated experimentally.   Equa-
tion (31) is the basic equation developed by Thackston  and
                            44

-------
Krenkel.  The constant was evaluated from laboratory data,
and when substituted in equation  (31) yields
                    k2 = 0.000215  *-                     (32)


where u,, is computed from
                         u^ =   (HSg)1/2                  (33)


In order to account for the difference between the actual
area of the air-water interface and the projected area,
another term was added to equation  (20) .  Because the Froude
number is an indicator of the roughness of the water surface,
it was incorporated into equation  (31) and the constant was
then reevaluated.  The final form of the equation was then


               k2 = 0.000125 (1 + F1/2)  ^*_              (34)
                                         H

                                    1/2
where F is the Froude number, V/(gH)   . Thackston and Kren-
kel point out that their equation is insensitive to errors in
velocity, and relatively less sensitive to depth than many
of the previously formulated models.

Summary.  With few exceptions, the currently available pre-
dictive models for the reaeration coefficient follow the form
proposed by Streeter and Phelps some 46 years ago2.  Thus,
the models proposed by O'Connor and Dobbins^, Langbein and
Durum-^, Owens et al^-o, and Churchill et al^ are all identi-
cal in form with the original Streeter-Phelps model, and all
of these models differ among themselves only in terms of the
numerical values of empirical coefficients.  As indicated so
clearly by Streeter and Phelps, wide variation of the empiri-
cal coefficients is to be expected.  The Isaacs-Maag model-^
contains an additional empirical factor for channel shape,
but remains identical to the others in form.

Krenkel and Orlob  attempted to explain reaeration in terms
of a longitudinal mixing coefficient, and provided an addi-
tional model in which the slope of the energy gradient was
added to the Streeter-Phelps form.  Later, Thackston and
Krenkel' provided a modified form involving the Froude Num-
ber.
                            45

-------
The main source of difficulty in all of the foregoing efforts
to develop an adequate predictive model for K2 has been the
lack of independently observed accurate values of K9 in nat-
ural streams, for purposes of model testing.  Only Streeter
and Phelps and Churchill et al conducted extensive field
studies to obtain values of K2 by indirect oxygen balance.
Some of the subsequently proposed models, e.g., those of
O'Connor and Dobbins, Langbein and Durum and Isaacs and Maag,
have been based completely upon the values of, K2 obtained by
others.

The studies of Churchill et al are the most extensive and
thorough of recent times.   As noted earlier, these investi-
gators attempted to relate reaeration capacity to hydraulic
properties by means of nineteen different models involving
some eight to ten hydraulic properties.  All of the models
produced very similar correlation coefficients.  These stud-
ies have demonstrated beyond further question that, in such
attempts, the magnitude of a correlation coefficient does not,
in itself, constitute adequate or sufficient evidence of the
real usefulness of any particular empirical model.  Recog-
nizing this, Churchill advised general use of the simplest of
his models, which had about the same correlation coefficient
as the others.
                           46

-------
                        SECTION V

                 EXPERIMENTAL PROCEDURES
This section of the report includes a summary of the experi-
mental procedures that were used in the several research
areas, field and laboratory.  This includes the field stud-
ies of the physical and hydraulic properties of the Flint,
South and Chattahoochee Rivers, the field tracer studies of
the reaeration capacity of those streams, the associated DO
and BOD studies, and the studies of the effects of pollu-
tants on the reaeration capacity.  Details of the various
field and laboratory procedures are provided to the extent
necessary for a full description of the research performed.
Where appropriate, the reader is referred to earlier pub-
lished details of analytical and other procedures.

Descriptions of the streams studied, with maps and relevant
detail such as sampling stations, are provided in Section VI-
Experimental Results.  In general, river sampling stations
were selected both for relevance to specific hydraulic fea-
tures and for accessibility.

PHYSICAL AND HYDRAULIC PROPERTIES

Measurements of hydraulic properties were made at stations
located at 500-foot intervals along the South River, Flint
River and Chattahoochee River.  Data were obtained for com-
puting channel slope, cross sectional area, stream width,
depth, hydraulic radius, velocity, and discharge.  Distances
were measured along the main channel of the stream with a
steel tagline, and a stake was driven into the bank and
marked with a station number at the end of each 500-foot in-
terval.  Levels were run to the top of each stake and to the
water surface opposite the stakes.  The levels were run
using standard leveling procedures and the work was checked
by tying the levels to bench marks located on highway
bridges.  Bench mark elevations were obtained f^om the US
Corps of Engineers, the US Geological Survey, the State
Highway Department of Georgia, and from the Offices of Pub-
lic Works of Fulton and DeKalb Counties.

Standard US Geological Survey stream gaging techniques, with
slight modifications, were employed by the field crew.
Width of cross section was measured with a tagline or steel
tape, and depths and velocities on the Flint and South
Rivers were measured with a top-setting wading rod and Price
current meter.  The procedures followed for the hydraulic
studies on the Chattahoochee River were identical to those
on the smaller rivers with the exception that measurements
                           47

-------
were made  from a boat rather than by wading the stream.   The
field  crew measured an average of thirteen depths and velo-
cities  at  each cross section on the smaller streams and on
the Chattahoochee; all velocities were measured at six-
tenths  of  the depth.

Typical cross section notes are shown in Figure 4.  The dis-
charge,  velocity and geometric properties of each cross sec-
tion were  determined from these notes.  The computation of
cross-sectional area and discharge are also illustrated in
the typical notes of Figure 4.  The values of "distance from
initial point" and "depth" for each cross section were
punched on computer cards for additional processing.  The
wetted  perimeter was computed from the relationship
 where x. is the distance from the initial point on the  stream
 bank to1the itn depth measurement, and dj_ is the depth  of
 the stream at distance x-^.  The summation is taken over all
 the values for a section.  The hydraulic radius was then
 determined as R = A/P, where A is the cross-sectional area
 as shown in Figure 4.  The stream depth was determined  by
 dividing the area of the section by the width of the section.
 The slope of the stream was computed for each 500-foot  sec-
 tion by dividing the change in the elevation of the surface
 of the stream by 500 feet.  The hydraulic properties mea-
 sured at each station are listed in Appendix AIII.

 Additional discharge information was available from US  Geo-
 logical Survey gaging stations on the Flint and Chattahoo-
 chee rivers.   Continuous water-level recordings at Georgia
 Highway 166 on the Chattahoochee and at Upper Riverdale
 Road on the Flint were utilized during the tracer studies.
 Measurements made by the Georgia Tech field team were used
 to establish the lower ends of the ratings.  A typical
 rating curve is shown by Figure 5.  Average discharge curves
 for the three streams are shown in Figures 6, 7 and 8.

Flow Adjustment.  The stream discharge at the time of the
tracer studies was usually somewhat higher or lower than the
discharge at the time of the detailed hydraulic studies.
Therefore it was necessary to adjust the measured hydraulic
properties to the discharge conditions that existed at  the
 time of the tracer studies.  The adjustments were based on
Manning's formula
                   Q -      A R    S1/2                  (36)
                           48

-------
                              Figure  4
              TYPICAL CROSS  SECTION  NOTES
Location
FLINT
                            STA.SI
Cross Section -
Flow -
Air 	
Date
                       S/WD
                          FAST
                   F@
                          /050
                          Water
    65°
           23  /9£8
Surface Water Elevation   —
Meter Number  /75"5
Width    14-.1
                        Party
                       Date Rated    —
Spin Before
1050
                                               After
                 Area  2/.8S
Remarks:
                      sreef.
                     Mean Velocity
                       BANK5 //< TH
           .10
3*
£M













o






Dirt.
from
initiil
point
O
l.o
2-0
3.0
£0
5To
6.0
7-0
8.0
f.o
/o.o
//•






Depth
0
0.54-
0.86
1.17
Z-VS
757
z.&t-
Z.f-B
Z.4&
Z.Z+
/.+8
A/f
aft?
0.48
0






I!





















Rcv-
olu-
tioni
IxiC
?
5"
,5
15
2f>
25
Z5
2S
Zf>
/$
7
^

i






VELOCITY
At
point





















ticil
0
.(56
.308
.665
745
/•'7
/.32
/3S
/.if
/.cs
.s/e
.J^
•117
o
O






AdjuMed
forhor.
ftngte or





















Are*
O
o.S'-f
0-88
1.17
z.*s
Z.S?
2.6-f-
Z-68

i.2^-
/.*§
A/f
ago
0.78
o
Z/-8S





DUchtrgc
0
a of
o.z.7
0.7g
/.87

7.48
J.6Z-
S.-io
Z.fZ
/.2/
O.^f
0./6
0
0
1.
Z0.55





                                                         .92

                                                         .94
                                                         .97
                                                         .96
                              49

-------
                            FIGURE 5
                  RATING CURVE, FLINT RIVER
                  AT UPPER RIVERDALE ROAD
10
          3   4  6 6789 10
100                  1,000
     DISCHARGE, CFS
                                                            10,000
                          50

-------
12
o
UJ
o
a:
<

u
in
FIGURE 6
AVERAGE DISCHARGE
MEASURED IN FLINT RIVER
JUNE 12- JULY 31, 1968

cc
n.
1-
z
U.




PLANT
EATMENT
cc.
t-



v






@










© <








r









UJ
cc
o
Q

(








P


1 	
J




f



_l










U
UJ
cr





JESTERS

T

















?
                    10       20      30      40      50       60       70


                                       DISTANCE  IN  500 FT. STATIONS
80
90
100
10

-------
to
       12
       o
       UJ



       I



       I

       o
IDU
140
120
100
80
40
20
0











r
J
L
^
]
J
i

D
3
0
J-T""

r

CREEK
i-
z
UJ
DC
O


«^M
Jj
ac
z




P ^

UJ
cc
o
CC
ID
U


^=J


T

X*
UJ
UJ
oe
o
_l
irt

•^M



T



'






P




«,«


J — '


J

PL7WT -
CREEK -
ktAI MtN 1
APFINGER
_ ^
u
J
a
z
3.
5
j)

r~"



V

—j 	 1



D ®

•M^MM^^B*



f






FIGURE 7
AVERAGE DISCHARGE
MEASURED IN SOUTH RIVER
AUG. 5 -SEPT. 13,1968
                            20       40      60       80       100      120


                                               DISTANCE IN 500 FT. STATIONS
140
160
180
200
                                  220

-------
               1300
              1200
               1100
u>
            CO
            u.
            o
s  1000
cm

x
o
CO
               900
               800
               700
(j





p
w
w
P5
O
05
0
EH
U
«
ft
o


M
W
W
«
ICKAJACK C
•A
d)
TOY CREEK
D
1
r




i




R CREEK 	
SWEETWATE

i

r~
»
3 CREEK 	
O


ii

















FIGURE 8
AVERAGE DISCHARGE
MEASURED IN
CHATTAHOOCHEE RIVER
JULY-SEPTEMBER 1970
                                       HO        80        120        160


                                               DISTANCE IN 500 FT, STATIONS
                                                                    200
240
280

-------
in which Q = discharge (cfs); n = Manning's roughness co-
efficient; A = area of stream cross section (ft2); R = hy-
draulic radius, A/P (ft); S = slope of the energy grade  line;
and P = wetted perimeter.

In order to adjust the hydraulic parameters at each of the
500-foot stations to values compatible with the discharge
during the tracer studies, the following assumptions were
made:
                  (a)  n2 = n±

                  (b)  S2 = SL
                  (c)  A2 = A1 + B1 Ay
                  (d)  P2 = PI + 2 Ay

where B = width of free surface, P = wetted perimeter and
Ay = change in depth as flow changes from Q-, to Q2, and
where the subscript 1 indicates the value of a variable
measured during the original detailed survey of the rivers,
and the subscript 2 indicates the value of a variable that
existed at the time of the tracer study  (see Figure 9).
The discharge at each sampling station was measured at the
time of the tracer study and hence the value of Q2 at each
of the 500-foot stations was readily estimated.  Changes in
n are negligible when the change in R is of the order of
0.1 foot or less20.   The larger variations in R, i.e.
R2 - R, , were of the order of 0.1 foot or less, therefore,
assumption (a) is justified.  The slope of the energy grade
line may change appreciably if the river control changes.
This can occur when a large increase in flow drowns a control
such as a small fall or rock outcrop.  However, the changes
in discharge were small enough to rule out this occurrence.
Thus, assumption  (b) appears to be valid.  Assumptions  (c)
and (d) , taken together, imply that the sides of the streams
are vertical.  In general, this is a good assumption for the
streams under study although it was not exactly satisfied at
every cross section.  The cross-hatched area on Figure 9
shows the area that is neglected when the assumption is  not
satisfied.

Assumptions (a) through (d)  allow Q2 to be written as
or
                               A  + B  Ay
          Q2 = GI (AI + BI Ay) (FL___I___)            (38)
where
                           54

-------
              . R 2/3  1/2
              A1R1  Sl
          A  = A  + BAY
                 2AY
(1)




(2)



(3)
         FIGURE 9

Flow  Adjustment Procedure
             55

-------
                    c  =
                    C

The constant C^ can be computed from


                    ci •

and B, and P, are known.  Therefore, for a known Q2 , equa-
tion  138) can be solved for Ay and A2 and P2 can be computed
from  assumptions (c) and (d) .  This was the procedure used
in all adjustments of the measured hydraulic properties.

Figure 10 is a typical example of the tabulation of measured
hydraulic properties as obtained directly from the  field
survey data.  Figure 11 is the associated table of  adjusted
hydraulic properties for flows observed at the time of the
reaeration tracer studies, and based upon the foregoing pro-
cedure for adjustment.  Appendix AIII provides the  detailed
data  for measured hydraulic properties for all of the stream
sections studied.

Computed vs. Observed Velocities.  Average values of the hy-
draulic properties were determined for each subreach, i.e.,
the reach of stream between each two sampling stations.  The
average value utilized for each geometric property, such as
depth, area, etc., was the simple arithmatic mean.  The
average computed velocity for a reach was calculated from
the relation


                                                         (39)
where L. is the distance between section i and section  i +  1
and V. is the arithmatic average of the measured velocity at
section i and section i + 1.  The velocities between sampling
stations so determined were later compared with the veloci-
ties determined by the time required for the dye tracer to
pass from one sampling station to another.  The time of
travel was defined as the peak-to-peak time from individual
dye curves for each tracer release (see Figure 12) .

A summary of the results of the velocity and traveltime com-
parisons is presented in Figure 13, which provides veloci-
ties obtained from the time-concentration curves from the
tracer studies as well as those obtained by computation
using equation (39) .   The velocity determined from the  dye
traveltimes did not always equal that determined from the
detailed hydraulic measurements.  In order to illustrate the
difference in velocities obtained, the ratios of "measured"
                           56

-------
Figure 10
Measured Hydraulic Properties
Station
500 ft.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
It
15
16
17
is
19
20
21
2?
23
24
25
Velocity
ft./sec.
1.14
0.64
1, "0
1.25
0.4H
0.23
0.21.
0. 12
0. 34
0.61
0.65
0.60
0.57
0.90
1.13
1. 14
0.98
1.27
0.89
1.09
0.56
0. J«
0.34
0.55
0.55
0.95
Area
ft.*
4.33
19.76
/.ro
9,41
17,46
36.33
43 .99
71.13
24.91
12.30
1 1.21
10.32
19.79
8.23
6.60
5.99
6.44
6.51
9.52
6.ft3
13.85
45.26
32.50
18.61
13.53
.7.45
For Flint River
Width Hyd. Had.
ft. ft.
6.60
14.30
1 3.00
12.60
15.50
34.00
P7.60
flS.bO
9 1 . 5 0
17.20
1 8 . 8 0
19,20
17.20
16.40
1 0,60
16.50
19.60
16.00
14.60
16.50
21.00
23.60
P0.20
10.70
14.10
13.40
0.58
1.15
0.56
0.70
1.03
1 .01
1.38
1.49
1.10
0.69
0.58
0.51
0.50
0.47
0.60
0.35
0.32
0.40
0.64
0.38
0.63
1.75
1.47
1 .38
0.86
0.53
Wet. Per.
ft.
7.51
!7->24
13.68
13,51
17.02
35,87
30.45
47.69
22.74
17,86
19.46
20.08
39,62
17.55
10,98
16.97
19.85
16,17
14,84
17.33
22.13
25.87
22.10
13.50
15.78
14,16
   57

-------
Figure 11
Hydraulic Properties
For Flint River
Adjusted To Average Discharge During Dye
Station Velocity
500 ft. ft./sec.

0 1,32
1 0,52
"2 1.16
3 1,03
4 0,46
5 " 0,22
6 ' 0,20
7 " 0.11
"8 0.32
9 0.60
10 0,^66
11 0,77
i~2 0.46
13 0.92
14 1,15
15 1.22
16 1,06
17 "" 1.19
18" 0.64
~'i9~' 1,13
20 0.55
21 0,16
22 0.29"
23 0.49
24 0.55
25 0.99
Area
ft.2
5.76
14,04
" 6,40
" 6V89
16, 2 2
33.61
36.13
65,63
23,19
11.96
11.59
9,55
15.32
a. 56
6.61
6.65
7.22
5,87
8,64
6.96
13,43
4 1 , 96
25723
14,97
13,81
7799
Ar
ft.
0.22
-0.40
-0.10
-0.20
-0,08
-0.06
-0.14
-0,12
-0.08
-0.02
0.02
-0.04
-0.26
0.02
0.02
0.04
0,04
-0.04
-0.06
0.02
-0.02
-0,14
-0.36
-0.34
0.02
0.04
Studies
Wet. Per.
ft.
7, ¥5
16.44
13.48
13.11
16.86
35.71
30.17
47,65
22.56
17.82
19,50
20,00
39,10
17.59
11.02
17,05
19.93
16.09
14.72
17,37
22,09
25.59 "
21.38
12.62
15.62
1 '
14,24
Hyd. Had.
ft.
0.73
0.85
0,47
0,53
0,96
0,94
1,26
1.38
1.03
0,67
0.59
0,46
0.39
0.49
0,62
0.39
0.36
0.36
0.59
0.40
0,61
1,64
1.18
1.17
0.87
6.56
 58

-------
   20
                                                                      FIGURE  12

                                                                    FLINT RIVER

                                                                 TYPICAL DYE  CURVES
rO
 O
 \-
 UJ
 >-
5
tf=2.60 HOURS
                       ST/V2
tf=2.90  HOURS
                                         STA.3
     5
                                       TINE  (HOURS)

-------
                        Figure 13
            Computed and Observed Velocities
Subreach
FLINT
Plant to Terrell Mill (01)
Computed
Velocity
(Fps)
RIVER
0.285
Terrell Mill to Lee's Mill (12) 0.500
Lee's Mill to Gravel Plant (23
Gravel Plant to U. R'dale(34)
U. R'dale to Valley Hill (45)
Valley Hill to Wavelyn Way (56
Wavelyn Way to End (67)
SOUTH
Plant to Bouldercrest (01)
Bouldercrest to P'ville(12)
P'ville to Waldrup(23)
Waldrup to Snapf inger (34)
Snapfinger to Panola(45)
Panola to End (56)
) 0.774
0.404
0.595
) 0.395
0.451
RIVER
1.41
1.26
1.48
1.20
0.78
1.32
Dye
Velocity
(Fps)

0.279
0.474
0.578
0.421
0.477
0.381
0.396*

1.40
1.13
1.48
1.21
0.86
1.35*
Ratio
Computed/
Dye

1,02
1.05
1.34
0.96
1.25
1.04
1.14

1.01
1.12
1.00
0.99
0.91
0.98
CHATTAHOOCHEE RIVER
0-2
2-3
3-4
4-5
1.64
1.58
1.75
1.57
1.71
1.76
1.85
1.72
0.96
0.90
0.95
0.92
*estimated value
                           60

-------
                        Figure 14

       SUMMARY OF HYDRAULIC MEASUREMENT PROCEDURES
    PROPERTY
    PROCEDURE
1.  Reach Length

2.  Discharge
    Cross Section Area
    Stream Depth
    Wetted Perimeter
    Slope
Taped along stream bank

A.  Current meter at each
    500-ft. station during
    hydraulic studies
B.  Current meter at each
    sampling station during
    tracer studies; gage
    height at each rated
    section

A.  Direct measurement at
    each 500-ft. station
    during hydraulic studies
B.  Direct measurement (for
    discharge) at each sam-
    pling station during
    tracer studies; area at
    each 500-ft. station
    adjusted to discharge
    at time of tracer re-
    lease

A.  Computed from area di-
    vided by width of cross
    section for each 500-ft
    station.
B.  Average over subreach
    computed as mean of
    depths at each 500-ft
    station after adjust-
    ment for discharge at
    time of tracer release

From cross-section measure-
ments for discharge deter-
mination.  Adjusted for dis-
charge for each tracer study

Difference in water surface
elevation divided by length
of reach, Ah/L. Ah was ad-
justed for discharge asso-
ciated with each tracer
                           61

-------
                  Figure  14  (continued)

       SUMMARY OF HYDRAULIC  MEASUREMENT  PROCEDURES
       PROPERTY
     PROCEDURE
6.   Slope (con't)

7.   Hydraulic Radius
    Txrtie of Travel
9.   Velocity
release

From cross-section measure-
ments for discharge deter-
mination.  Adjusted for dis-
charge for each tracer study

Peak-to-peak times from dye
curves for each tracer re-
lease                -

A.  Discharge divided by
    area at each 500-ft
    station
B.  Average for subreach by
    equation (39)
C.  Average for subreach by
    dividing distance by dye
    time of travel
                          62

-------
velocity to "dye" velocity were computed and are included
in the tabulations.  The average ratio for the South River
was 1.01, the ratio for the Chattahooch.ee was 0.93 , while
the average ratio for the Flint River was 1.16.

The higher value for the Flint River can be explained by
referring to the general description of the streams.  The
channel of the flint River is very irregular; it contains
many small riffles and pools, and the channel is littered
with debris.  When the velocity was being measured with
wading rod and current meter, the end of a 500-foot inter-
val occasionally fell in a pool that was too deep to wade
or often among debris such as tree tops that had been left
from timber operations.  The field crew therefore made the
measurement slightly upstream or downstream at a better
cross section.  As a result, the velocities were higher, and
the predicted time of travel less than those Computed from
the dye curves.  The channel of the South is straighter,  with
fewer pools and more uniform depth (average ratio 1-01) ,  and
the channel of the Chattahoochee is the straighest and most
uniform, with no real pools (average ratio 0.93).

Figure 14 summarizes the procedure used for the individual
measurements made at each 500-foot section and the procedure
used to determine an average value of each property over a
subreach.

TRACER STUDIES

The field tracer studies of reaeration capacity were con-
ducted generally in the same manner as those performed in
the initial Jackson River field demonstration of the tracer
technique^, with minor modification.  The analytical pro-
cedure for krypton-85 and tritium was also the same as the
procedure developed earlier, in which the two radioactive
tracers are counted simultaneously for the same sample by a
liquid scintillation technique^f21.  These procedures are
described briefly here for purposes of completeness and con-
tinuity, and modifications are noted, and the reader is re-
ferred to the earlier publications for greater detail.

As  indicated in Section IV of this report, the basic field
procedure consists of releasing an instantaneous point dose
of three tracers and subsequently observing the downstream
concentrations at selected sampling points along the water-
course.  The three tracers are rhodamine-WT fluorescent dye,
tritium in the form of water molecules, and krypton-85 in
the form of a fully dissolved gas.  The three tracers are
contained in a well mixed single dose, so that release of
all three is truly simultaneous.  The fluorescent dye is
monitored continuously at each downstream sampling station
                           63

-------
by the use of a continuous-flow recording fluorometer, and
thus provides accurate information as to time of flow.   It
also indicates when to sample for the other two tracers,
which is accomplished primarily as the dye peak passes the
specific sampling station.  The tracer samples are subse-
quently analyzed in the laboratory for krypton-85 and tri-
tium.  The tritium provides an accurate measure of disper-
sion and dilution.  The krypton-85, dispersed in identical
degree due to the simultaneous dosing technique, provides
the needed additional accurate information as to the gas
transfer that took place between sampling stations.  The
theoretical basis of the analysis for gas transfer and re-
aeration capacity is provided in Section IV of this report,
as well as in earlier publications"'10t

Dose Release and Assay.  In these more recent studies of the
reaeration capacity of the Flint, South, Patuxent and Chat-
tahoochee Rivers, the mixed doses of three tracers (dye, dis-
solved krypton-85 and tritiated water) were procured ready
for immediate use from a vendor, in order to make best use
of project staff time.  (In the earlier Jackson River stud-
ies , the individual mixed tracer doses were prepared by
project staff10).  The required number of  doses  was deliv-
ered by the vendor  by air and express to the Georgia Tech
Radiation Safety Officer and stored for imminent field use.
All of the doses were delivered in glass containers pro-
vided by Georgia Tech, so as to facilitate the field opera-
tion.

For the smaller streams (Flint, South, Patuxent) the indi-
vidual mixed doses were of one quart size.  This dose liquid
volume was essentially all fluorescent dye (20 percent
aqueous solution) , with only a few ml of tritiated water and
the dissolved krypton-85 added.  Initial doses used in the
Chattahoochee studies were about four liters in volume.
However, these proved to be inadequate because of the great
dilution and dispersion, and most of the latter Chattahoo-
chee doses were about nine liters in volume, with propor-
tionately larger quantities of the radioactive tracers (see
Section VI).

The field tracer studies were conducted usually on Saturdays
or Sundays or between school quarters, so as to avoid  con-
flicts _ with the regular coursework of student project aides.
This did not interfere with project objectives or scheduled
progress,  but in the case of the Chattahoochee studies it
was observed that the BOD load from the Clayton STP in
Atlanta was substantially lower on the weekends than during
the usual work week.
                           64

-------
Figure 15 illustrates the manual tracer dosing device used
in the studies of the Flint, South and Patuxent Rivers.  The
base was a steel channel section (1" x 4") with smaller chan-
nels welded underneath it so as to keep the dose off the
stream bed.  The quart dose bottle was first taped to this
base section.  The top, or striker, section was a similar
piece of steel channel with a four foot length of one-inch
steel rod welded to it as shown.  With the tracer dose bot-
tle firmly taped to the base, the striker section was then
also taped to the base, to complete the assembly for dosing
(see Figure 15).  To release the tracer dose, the assembly
was carried by the vertical rod to the main channel of flow
by wading, and placed firmly on a rocky section of the stream
bed.  The submerged dose could then be released by striking
the top of the one-inch steel rod with a hammer, to shatter
the glass bottle.

The dosing device shown in Figure 15 proved to be very de-
pendable and satisfactory in all ways.  Because of its sim-
plicity, very little could go wrong with the dosing proce-
dure.  Its weight was sufficient to facilitate holding it
steady, even in a strong current.  Assembly of the dose was
quick  (three or four minutes), and this was of real assis-
tance in minimizing any external radiation dose received by
the personnel who conducted the dosing operation.  The steel-
to-glass contact resulted in thorough shattering of the glass
bottle, and the actual tracer dose release was thus virtually
instantaneous.  After the release, still holding it by the
vertical rod, the assembly could be readily washed and
cleaned by rinsing it in the stream flow for a few minutes.
Customarily, it was rinsed vigorously for a few minutes and
then left submerged in the stream flow, to be retrieved some
hours later after downstream sampling was complete.

For the doses in the Chattahoochee River, which was much
larger and too deep and swift to wade, the larger glass dose
bottles were suspended, submerged and held in place by guy
ropes from a bridge, and were then shattered by the use of
blasting caps taped to the side of the bottle.  This field
procedure also proved fully satisfactory in terms of sim-
plicity and minimum exposure of project personnel handling
the dose bottle, and resulted in an instantaneous release
for all practical purposes.  These doses were submerged
about two feet below the water surface and were rigged with
a thin metal plate between the bottle and the water surface.
Careful observation demonstrated that no splashing occurred
on detonation of the caps.

The dose bottles were delivered with a rubber septum seal,
so as to facilitate sampling by use of a hypodermic needle.
                           65

-------
             FIGURE 15
         TRACER RELEASE
              DEVICE
              (MANUAQ
BASE
                              STRIKER
                          TRACER  BOTTLE
                66

-------
Initially, the smaller doses, and later the first of the
larger Chattahoochee doses, were thus sampled directly for
dose assay, by pressure pipette, before release.  This pro-
cedure proved, however, -to be somewhat undesirable in terms
of preventing contamination of the external surfaces of the
dose bottle.  In the case of the larger doses, this proce-
dure also required personnel exposure to the dose for longer
periods of time  (15 minutes or more) which, although quite
safe by accepted standards, was nevertheless regarded as
undesirable.  Accordingly, a hand sampling procedure was
devised to eliminate the need for direct dose assay.  This
was a miniaturized  (to hold a one-ounce sampling bottle) DO
sampler^2 attached to a long rod.  The procedure consisted
of sampling the dye patch a hundred yards or so below the
dose point, before it became greatly dispersed.  Several such
samples could be obtained during passage of the dye patch,
and it was found that they provided entirely satisfactory
information as to the krypton:tritium ratio in the tracer
dose.

It is worthy of mention that, no matter how much care is
taken, glass bottles can be damaged in shipment, and this
happened to one nine-liter Chattahoochee River dose during
these studies.  No damage or leakage was apparent until the
drum shipping container was opened in the field for use.
 (Standard project procedure was to monitor the outside of the
shipping container on receipt, and then leave it sealed un-
til actual use in the field).  At that time a cloud of gas-
eous krypton-85 was released and was detected by standard
project safety monitoring procedures, which also were suc-
cessful in preventing any unsafe exposure of project per-
sonnel.  During subsequent safe disposal of that dose, it
was ascertained that the bottle had been cracked in transit.
This occurrence is noted here to emphasize the importance of
following safe procedures in handling and using such materi-
als .

Field Procedures.  Field sampling procedures were essentially
the same as those described earlier in connection with the
Jackson River studies^,10.  The field staff usually included
about six undergraduate civil engineering student aides plus
one or two faculty, the exact number depending upon the de-
tails of the specific study planned.  One field sampling
crew (three aides) set up the first downstream sampling
station and began operation at about the time of dose re-
lease.  Usually, time of flow to the first sampling station
was several hours.  After dose release, the dosing crew
moved to the first sampling station to prove any assistance
that might be necessary, then left once the first station
was in satisfactory operation.  A second field sampling crew
moved in to set the second sampling station in operation
                            67

-------
several hours before tracer was anticipated at that location.
On completion of sampling at the first station, that crew
moved with its equipment to the third downstream sampling
station, to set that in operation well before the appearance
of dye.  This leapfrogging was continued by the two field
crews until the end of the study.

Two complete sets of field sampling equipment were thus kept
in operation during each tracer study.  A third complete- set
of equipment was kept available in the field vehicles as
standby equipment.

The field sampling arrangement, and the associated equipment,
was essentially the same as in earlier studies10, and is
shown schematically in Figure 16.  The procedure consisted
of placing a submersible pump in the main stream flow and
pumping stream water to a continuous-flow recording fluoro-
meter through a length of flexible hose.  The pump and flu-
orometer were operated with power from a portable gasoline-
powered generator located nearby.  After passing through the
fluorometer for measurement of dye concentration, the pump
flow was split, with most being wasted back to the stream
and a smaller flow being delivered to the bottom of a one-
ounce glass sample bottle.  The sample bottle was allowed
to overflow until a sample was wanted, then removed slowly,
capped with a pressure cap, numbered, and stored for trans-
fer to the laboratory.  A new sample bottle was then in-
stalled for the next sample.  A background sample was taken
at each station well before the arrival of any dye.

The fluorometer dial and recording chart were watched for
arrival of the dye, and the chart was marked occasionally
with the correct time and the corresponding dial reading.
Sampling was begun soon after arrival of.the leading edge of
the dye mass, and continued with increasing frequency as the
peak dye concentration approached.  Each sample was numbered,
and the sample number and time were marked in a field note-
book as well as on the fluorometer recording chart.  All
samples were pressure-capped and sealed with black plastic
tape.   Sampling was usually discontinued after the dye con-
centration had dropped to about three quarters of the peak
concentration.

The three streams (Flint, South, Chattahoochee) were all
near Georgia Tech, where the liquid scintillation counter
was housed,  and the field samples were usually delivered to
the counting laboratory within four or five hours after col-
lection.   The Project Chemist was on duty at the laboratory
during the tracer studies, and the field samples were pre-
pared  for counting and set in the counter at once after re-
ceipt  at the laboratory.  At the worst, field samples waited
                           68

-------
                                    FIGURE 16
                                FIELD SAMPLING
                                  ARRANGEMENT
                        GENERATOR
                                        CONTINUOUS FLOW
                                        RECORDING
                                        FLUOROMETER
INFLOW
 FROM
STREAM
                                 SAMPLE
                                 BOTTLE
                         DISCHARGE
                         TO STREAM
                        69

-------
no more than eight hours or so between collection and being
placed in the laboratory counter.

Other field measurements were also made during the tracer
studies, and other samples taken.  River water temperature
was observed and recorded several times at each sampling
station, and water surface elevations observed relative to
some fixed reference  (e.g., a bridge).  A separate crew
gaged stream flow at key locations by standard procedures.
DO and BOD samples were also taken, the former being eval-
uated by the azide modification of the Winkler Method.

Laboratory Procedures.  The krypton-85 and tritium concen-
trations in the field samples were evaluated by a counting
procedure developed earlier by Cohen et al^1-  In this pro-
cedure the two radioactive tracers are counted simultan-
eously in the same sample by liquid scintillation tech-
niques, the counts being recorded separately by virtue of
the difference in energy of the two different beta emissions.
In brief, a 2-ml aqueous sample is mixed with about 20 ml
of liquid scintillation solution  (Butler's Solution) in a
25-ml glass counting vial, and the scintillations are
counted for a fixed period (e.g., 20 minutes) in a liquid
scintillation counter.  The reader is referred to the
earlier paper for greater detail^l.

Preparation of the individual field samples for counting con-
sisted of transferring 2 ml from the one-ounce sample bot-
tle to a counting vial about half full of Butler's Solution;
the vial was then filled with additional Butler's Solution,
capped, swirled to mix the water and counting solution, and
set aside for subsequent counting.  Three replicate vials
were thus prepared for each field sample to be analyzed, and
usually three samples (corresponding to peak dye concentra-
tion)  were analyzed for each field sampling station.  Thus,
nine separate vials were counted for each sampling station.
The liquid scintillation counter, in addition to automatic
sample changing,  also had provision for recycling samples,
and the laboratory procedure included counting each vial at
least twice.   This procedure, involving considerable repli-
cation and long counting times, provided excellent counting
statistics and accurate results for river samples having a
count rate of as  little as twice the background rate.

The procedure for transferring 2-ml portions of field sample
to the counting vials is of special interest.  Special pre-
cautions against  loss of the dissolved krypton-85 are nec-
essary - for example, the water cannot be poured out of the
sample bottle,  nor can it be subjected to the negative pres-
sure associated with direct transfer by means of a hypoder-
mic syringe.   Figure 17 illustrates the procedure developed


                           70

-------
          PRESSURE  PIPETTE
            (NOT TO SCALE)
 RUBBER
 STOPPER.
                                  DISPOSABLE
                                     cc SYRINGE
2 ml. PIPETTE.
                               DISPOSABLE
                               22g HYPODERMIC
                               NEEDLE
                               FIGURE 17
                  71

-------
earlier  , which involves pressure-pipetting.  In brief,_a
rubber stopper that fits the sample bottle is prepared with
a hypodermic needle and a 2-ml volumetric pipette, as shown.
The needle tip projects barely through the stopper, whereas
the pipette projects well into the sample water.  The full
sample bottle is opened and the prepared stopper firmly in-
serted.  A fully opened syringe is then attached to the
needle, and the plunger is pushed in slowly to force a mea-
sured amount of water up into the pipette.  The stopper is
then removed and the 2-ml portion transferred directly to
the half-full counting vial, the water being allowed to run
down the side of the bottle from a location just above the
Butler's Solution.  The aqueous sample falls directly to the
bottom of the vial, under the counting solution, and thus
has virtually no opportunity to lose krypton-85.  This trans-
fer procedure is repeated to fill the desired number of rep-
licate counting vials.  The vials are then filled to the top
with additional counting solution, capped, and finally
swirled to achieve a uniform mixture of sample and counting
solution.

As noted above, usual practice was to prepare three repli-
cate vials from each field sample for counting.  In the case
of the Chattahoochee River studies, where tracer concentra-
tions proved to be especially low, five or six replicate
counting vials were prepared for counting, to improve ac-
curacy.

The liquid scintillation counter was calibrated against known
standards to provide accurate counting efficiencies for tri-
tium and krypton-85.  Usual efficiencies in these studies
were 25 to 28 percent for tritium and 86 to 90 percent for
krypton-85.  These efficiencies permitted accurate counting
of river water samples having quite low tracer concentra-
tions .

Figure 18 is a typical semilog plot of the results from one
of the tracer studies, showing the usual way of plotting the
krypton:tritium concentration ratios and the range of obser-
vation for each ratio.  For each sampling station, the mean
value for the three samples analyzed (total of nine vials)
is shown by the small circle, and the range of results for
the three samples (three vials each) is shown by the asso-
ciated vertical line with brackets.  Proceeding downstream
in time, the tracer concentrations decrease, especially that
of krypton-85, with corresponding increase in the range of
observation of the ratio at each sampling station.  In most
cases,  results were satisfactorily accurate at all stations,
but in a few, where almost all of the tracer gas had been
lost,  the range of observation at the last downstream sta-
tion was substantial.  In such cases, the results presented
                           72

-------
1.0

o.g


0.8



0.7



0.6
                                           Dump X S
                                                   .ations
                                                                                     SOUTH RIVER

                                                                              KRYPTON TRANSFER COEFFICIENTS


                                                                              Hote;  All values are Kg/hour

                                                                                    (K )   = (Kg)^-
                                                                                    * 2 ox    0.83
                                                                                      FIGURE
                                                                                              18
0.1*
0-3
                                         • Dump X
                  Dump VI-
o.i
       a
                                             Dump
                                                   VI Stations
                                            Time of Flow - Hours
                                           73

-------
later  (Section VI) are enclosed in parentheses to indicate
uncertainty.

DO AND BOD STUDIES

During the studies of the reaeration capacity of the Flint,
South and Chattahoochee Rivers, routine observation of DO
and BOD was carried out as originally planned, according to  •
standard analytical procedures12.  To the extent_possible,
DO and BOD samples were taken !at key sampling points at or
near the time of passing of the dye peak associated with^
the gaseous tracer studies.  DO' s were analyzed in the field
on collection, by the azide modification of"the Winkler
method, and river water temperatures were observed at that
time.  BOD samples were transported to the Georgia Tech
laboratory and set up without delay.  In a number of cases,
the BOD analysis involved a BOD-time series, rather than
only the 5-day observation.

As originally planned, this research incorporated limited
oxygen balance studies in addition to the primary studies
of the relationships between reaeration and hydraulic pro-
perties, for the sake of completeness and for the practical
purpose of providing comprehensive information on real situ-
ations that involved serious pollution.  For example, using
the firm values of K2 to be obtained from the gaseous tracer
studies, it was hoped that highly accurate DO balances might
permit improved evaluations of usually neglected natural
processes such as benthal decomposition and respiration by
attached oxidizing growths.  Unfortunately, in streams as
polluted as the Flint and South proved to be, the perfor-
mance of complete and accurate oxygen balance analyses can
be a major project in itself, and such detailed evaluations
of other natural processes proved for the most part to be
complex beyond the practical limitations of available staff
and funds.

Relevant observations of DO and BOD are presented in the
following report section on Experimental Results, together
with appropriate discussion.

POLLUTANT EFFECTS ON REAERATION

Various pollutants alter the ability of gas molecules to en-
ter and escape water.  This alteration causes the value of
K2 to vary under identical hydraulic and environmental con-
ditions,  depending upon whether clean water or polluted
water is  flowing in the stream.  The pollutant effect on
reaeration is not only related to the pollutant constituents
and concentrations but also the turbulent mixing regime
                           74

-------
within the fluid.  In order to evaluate the effect of vari-
ous pollutants on reaeration, a series of laboratory tests
was conducted on both natural and artificial stream waters.

The Pollutant Effect Reactor.  A constant temperature, open
top reactor, as shown in Figure 19 was constructed for the
purpose of comparing the reaeration rates of varipus test
waters under the same hydrodynamic and environmental condi-
tions.  The test reactor consisted of a support bracket to,
which an open top glass reaction kettle was mounted.  A
variable speed stirrer provided the mixing action.  The
stirrer motor was rigidly attached to the support bracket so
that the reaction kettle and stirrer maintained their re-
spective positions relative to each other.  Constant temper-
ature water was circulated around the reaction kettle for
the purpose of temperature control.

The reactor system was all glass, with the exception of the
stainless steel impeller and shaft.  The reactor could be
operated with a maximum water volume of 4 liters.

Operation of the Pollutant Effect Reactor.  In the typical
test the reactor was initially dismantled, thoroughly
cleaned, and rinsed with distilled water a minimum of five
times before reassembly.  A predetermined volume of distil-
led water  (3,600 ml) was then added to the reactor and al-
lowed to stir until thermal equilibrium was attained.  The
stirrer was stopped and a single homogeneous tracer dose of
tritiated water and dissolved krypton-85 was added below the
test water  surface, in 2 ml of distilled water contained in
a  pipette.  The stirrer was started again and not stopped
until the end of a run.  Radiotracer samples were taken
from the reactor by means of immersion of a 2-ml pipette
until test water filled the pipette to a point above the
fiducial mark.  The pipette was then withdrawn from the
reactor and a 2-ml sample was transferred to a liquid scin-
tillation counting vial that had been previously filled
with 10 to  15 ml of Butler's solution containing the liquid
scintillator.  After the transfer was complete, the vial was
then filled full with Butler's solution, capped, and loaded
into the liquid scintillation counter.  The time at which
each sample was taken was recorded in the notes.

After the completion of a distilled water test, the reactor
was drained by means of a siphon and thoroughly rinsed with-
out dismantling or moving the system.  A predetermined vol-
ume  (equal  to the volume of distilled water used in the pre-
ceding test) of polluted water was then added to the system
and the test was conducted again as described above.  Sev-
eral times  during each test, the speed of the stirrer shaft
was determined and recorded in the notes.
                            75

-------
                                                   variable  speed  motor
support bracket
 stirrer
                                                  open top reaction kettle
                         REACTOR ARRANGEMENT




                         (pollutant studies)










                            FIGURE 19
                                76

-------
   1.0
   0.8
   0.6
   o.k
                                                              TYPICAL REACTOR EXPERIMENT
                                                               Note:   Ti  =

                                                                       2     ^2
                                                                  FIGURE  20
           \
i •= 28.1 mins
     polluted sariple)
   0.2
•8
PS


§
a

-p  0.1
c

-------
Computation of the Pollutant Effect.  For each reactor run,
the krypton: tritium concentration ratios associated with
each sample analyzed were plotted as a logarithmic function
of time.  The line of "best fit" was then obtained by the
method of least squares fitting of the linearized data.
The slope of this best fit line was then recorded as the Kkr
for the particular run.  By comparing the Kkr for each pair
of reactor runs (distilled and polluted) the pollutant ef-
fect could be determined.  Figure 20 shows a typical pair of
laboratory test results for the detergent surfactant LAS
(linear alkyl sulfonate)  in distilled water.  The Kkr as~
sociated with the distilled water run was 0.0412 '/rain, while
the Kkr associated with the LAS polluted water test was
0.0246/min.  The pollutant effect on reaeration during this
particular test reduced the gas transfer rate coefficient by
40 percent.  Much of the available literature describes the
pollutant effect on the reaeration rate coefficient as

                     K0 , polluted
                 
-------
     "     •
field tracer tests.  Even though such a suggestion repre-
sents a clear denial of basic laws of physics and chemistry,
it was. felt that the simplest and clearest response would be
experimental, and the following experiment was therefore
conducted.   (It is worthy of mention also that the nuclear
power industry has for some years been seeking a practical
process for for separating tritiated water from its wastes,
and would hardly have missed the above possibility if it
had any faint semblance of validity.)

Figure 21 shows the experimental column designed and fabri-
cated for this study.  The column was of glass, 36.5 inches
in height.and 2.75 inches in diameter.  Four sampling, ports
were built in at heights as shown, and closed by use of a
rubber serum cap, so that a hypodermic needle could be in-
serted through the serum cap directly into the center of the
column and a 2.0 ml sample withdrawn.

The experimental column was housed in a small storage room
used only by project personnel  (completely unused during the
period of study).  This room was air-conditioned, with ther-
mostatically controlled temperature.  The column was rigidly
supported with clamps to a rigid steel post, which rested on
the concrete floor in the basement of a building.  This ar-
rangement minimized any possible vibration.

The experimental procedure was as follows:  the column was
filled with water containing some tritiated molecules.  A
total volume of  3400 ml was used.  A small amount of food
coloring was added also.  The contents were stirred magnet-
ically until the food coloring appeared uniform throughout.
The stirrer was  then stopped.  A set of samples was taken
at once  (about 2.5 ml from each port) as indicated above,
with careful attention to avoiding any agitation of the liq-
uid in the column.  The set of four samples was then imme-
diately set up for counting and 2.0 ml portions were counted
in the liquid scintillation counter.  Additional sets were
taken in the same manner at intervals through a period of
120 hours, as shown below:

   Depth,          Observed Tritium Count Rate, cpm/ 2 ml
inches below      at     ^      at                at
water surface   t=0 hrs      t=1.6 hrs        t=19.3 hrs

     4.1         11,675        11,747           11,588
    14.1        11,818        11,958           11,740
    24.1        ill,918        11,635           11,765
    33.9       - 11,762        11,968           11,750
                           79

-------
GLASS COLUMN
STIRRING BAR'
                                                 FIGURE  21
                                             SETTLEABILITY  OF
                                             TRITIATED WATER-
                                               EXPERIMENTAL
                                                ARRANGEMENT
                                       •SAMPLE PORTS
                                        (with rubber
                                         serum caps)
                                  MAGNETIC STIRRER
                    80

-------
(continued)

   Depth,  ,        Observed Tritium Count Rate, cpm/2 ml
inches below              at                  at
water surface          t=73.1 hrs        t=118.8 hrs

     4.1                11,547             11,658
    14.1                11,602             11,755
    24.1                11,522             11,793
    33.9                11,527             11,688

The above demonstrate clearly that no settling of the tri-
tiated water molecules occurred over the period of study.
The very small differences in count rate that occurred were
the result of usual counting statistical deviations and the
small errors involved in the volumetric transfers  (two per-
cent or less).  These differences were quite random, as ex-
pected, and there was no statistically significant variation
of count rate with depth for any of the sets of samples.

In contrast to this experiment, in which every reasonable
effort was made to maintain a quiescent volume of water, the
field tracer experiments are conducted in water that is tur-
bulent.  They are completed in a matter of 10 to 20 hours,
rather than several days.  It is clear, from these consider-
ations and the foregoing data, that settling of tritiated
water molecules in the field tracer studies does not occur
to any significant degree.

These results were, of course, as expected.

SOURCES OF ERROR

Although every effort was made during the studies reported
here to eliminate and minimize experimental error, it is
inevitable that some error or discrepancy will occur in any
research, especially when field operations are involved.
The kinds of error usually involved in stream physical
studies, and their likely degrees, are commonly under-
stood and need not be discussed in detail here--reference
to Figure 14 should suffice to call them to mind.  However,
certain other kinds of deviation or error are not so obvious
at first glance.  Some of these became evident especially
during the analysis and interpretation of the field results,
and it was only then that they could be adequately defined
in terms of their degree of importance.  Such errors are
discussed below.

Counting Statistics.  In all of the field tracer studies per-
formed it was attempted to obtain maximum information with
the minimum dose of radioactive tracers.  Doses were
                            81

-------
minimized for reasons both of cost and personnel safety-
Obtaining maximum information often meant that the sampling .
was extended too far downstream from the dose point, so that
at the farthest downstream stations very little tracer gas  •„
remained in the river water.  In such cases, the krypton  85
count available from the liquid scintillation counter_was   ,
too small, compared to the background count, for statistical
accuracy.  Usually this was obvious, and the data for the
sampling station were accordingly rejected.  However, in  a
few cases the statistical accuracy (the variance) appeared
acceptable even though the counts were low, and the data
were retained.  Occasionally, subsequent analysis of the
field results showed such data to be unacceptable after all,
for reasons of such low count rates.  Such cases are iden-
tified in Section VIII of this report, and involve data
specifically from two South River releases.

Dose Assay.  As indicated earlier the smaller (one quart)
tracer doses used in the Flint, South and Patuxent River
studies were sampled directly by pressure pipette before  re-
lease, so as to provide an assay of the dose itself.  This
procedure not only provided verification of the quantities
of tracer received from the vendor, but also provided the
necessary krypton:  tritium concentration ratio for the
dose point in the river.  About midway in these studies the
hand sampling procedure was also initiated during the Patux-
ent River studies, as analysis of prior results occasionally
gave some cause to suspect the direct dose assay by pressure
pipette.  In brief, the results of all of the studies indi-
cate that although most of the direct dose assays were good,
an occasional one yielded a false krypton:  tritium concen-
tration ratio.  Subsequent analysis and interpretation led
to the conclusion that in such cases the krypton-85 and tri-
tium were most probably not uniform in concentration in the
dye within the dose bottle--the triple tracer dose was prob-
ably not wholly uniform in the bottle, even though subse-
quent handling and deliberate mixing in the field before  re-
lease resulted in a satisfactorily uniform released dose.  A
specific example of this discrepancy is identified in Sec-
tion VIII of this report in connection with dose number VIII
in the South River.

Nonuniform^Mixing.  As indicated in Section IV of this re-
port and discussed in greater detail in Sections VI and VIII,
most theoretical analyses of the relationships between gas
transfer on the one hand and turbulent mixing or hydraulic
properties on the other require the assumption of uniform or
homogeneous turbulent mixing.  In contrast, the field tracer
method for gas transfer measures the gas transfer that ac-
tually occurs, whether mixing in the stream is uniform or
                            82

-------
not.  Hence, in stream reaches where turbulent mixing is not
uniform, some discrepancy between observed gas transfer and
that predicted from theory must be expected.  For the most
part in the studies reported here the assumption of uniform
mixing appears to be quite acceptable.  However, in the case
of one reach in particular, namely, the long, deep pool
above Panola Shoals in the South River, mixing is clearly
not uniform or homogeneous.  In that case, the inflowing
water clearly does not usually mix uniformly with the pool
contents, and this leads to more than usual variability in
gas transfer.  This matter is described in greater detail in
Sections VI and VIII of this report.

Pollutant Effect on Reaeration.  As shown in the subsequent
sections of this report,the presence of pollution can
markedly modify the gas transfer capacity of water, both in
the laboratory and in real streams.  In stream situations
where there is only a single major source of pollution, such
as the Flint, Jackson, Patuxent and Chattahoochee River
studies, the gas transfer capacity is modified according to
the degree of pollution, but this does not create signifi-
cant difficulty of interpretation in terms of the basic re-
lationships between gas transfer and the stream hydraulic
properties.  However, where multiple major sources of pol-
lution occur, as in the South River studies reported here,
or where the pollution loads are highly variable from day
to day, the resulting marked differences in pollutant effect
from one reach to another, or from one day to another, can
mask or confuse the analysis of gas transfer in terms of
hydraulic properties.  Such effects are described in detail
in Section VIII of this report in connection with the South
River studies.

Each of the above sources of discrepancy or error is readily
recognizable and identifiable in the observed data if it
occurs to any significant degree, and each of these kinds of
discrepancy thus proved to be uncommon.  Yet each also oc-
curred at least once in recognizable degree and had to be
reckoned with in the final analysis of results.  These spe-
cific situations are clearly shown in the following sections
of this report, especially in connection with the South
River studies.
                            83

-------
                      SECTION VI

         EXPERIMENTAL RESULTS - FIELD STUDIES
This section of the report includes a summary of the rele-
vant observed results from the field studies of the Flint,
South and Chattahoochee Rivers, together with directly rele-
vant discussion.  Although not a part of this research pro-
ject, the results of similar field tracer studies performed
earlier on the Jackson River in West Virginia (1966) and on
the Patuxent River in Maryland (1969) are directly applica-
ble and are also included.  The field survey results sum-
marized here thus represent all such studies performed to
date that have included both independent (tracer) observa-
tion of gas transfer capacity and detailed field measure-
ment of the hydraulic properties.  The detailed data from
which these summarized results have been derived are in-
cluded as Appendices to this report.

FLINT RIVER

The Flint River rises in the southwest region of the City of
Atlanta and flows southward for approximately 350 miles be-
fore joining the Chattahoochee to form the Apalachicola,
which then flows into the Gulf of Mexico.  A headwater reach
extending from the Flint River Sewage Treatment Plant  (STP),
located immediately south of the Atlanta Airport, to a loca-
tion about 1.6 miles south of the intersection of the Flint
River and Georgia Highway 138 was selected for study.  The
length of the river reach studied was 9.9 miles.  Figure 22
is a general map of the Flint River study locale, and shows
relevant features such as the Flint River STP, highways, tri-
butary streams and river sampling stations.

The Flint River study reach includes a wide range of hydrau-
lic features - the widest of any of the streams studied.
The upper two miles are characterized by alternating riffles
and small pools, and there are two larger mill ponds in this
section of the river.  These ponds are shallow, being almost
filled with silt and sludge.  At each mill pond site the
flow spills over a dam and falls a distance of 10 to 12 feet.
Below the second waterfall, the remainder of the study reach
is characterized by a highly variable cross section, with a
variable bottom of stone, clay or sand, and many sections of
the stream contain old treetops resulting from timber opera-
tions along the banks.  In addition, there are two sections,
one about a mile in length and the other about 1.5 miles
long, in which the stream flows in multiple channels.  The
meandering of the flow in these channels is difficult to
                           85

-------
00
                                         N
                                                                                   FIGURE 22
                                                                         FLINT  RIVER STUDY LOCALE

                                                                             VICINITY OF ATLANTA
                                                                                                                          KEY
                                                                                                                                DIRECTION OF FLOW
                                                                                                                            REPRESENTS EVERY
                                                                                                                            FIFTH CROSS SECTION

-------
                                          Table 1
                                Dosing Pattern, Flint River
      Dump 	Flint River Sampling Station Number
     Number    0      IP       1      2P       2      .34
        I ... 4/25,
              204. ,
              485. ,
      II  	5/2 	 Key: 5/2-Dump II released on 5/2/69-Station 1
                              182 	      182-182 me of krypton-85 released
                              463 	      463-463 me of tritium released
     III  ... 5/23

-------
follow, and no single 'typical1  cross section can be used to
describe these stream reaches.   One reach, about 0.3 miles
long, is a marsh in which no meaningful velocity or cross-
sectional measurements can be made.

Average discharge measurements  made during the 1968 summer
field hydraulic studies ranged from 5 cfs above the Flint
River STP to 27 cfs at the lower end of the 9.9 mile study
reach.  Similar flows were encountered during the summer
1969 field tracer studies.  The waste flow from the Flint
River STP, at the upper end of the study reach, was the only
significant source of pollution.  At the time of these
studies, the Flint River STP was a standard rate trickling
filter plant with a design flow of 2.0 mgd, and was somewhat
overloaded.

Dosing Pattern.  Table 1 shows  the dosing pattern that was
used in"the 1969 tracer studies of gas transfer in the Flint
River.  This pattern of seven separate releases was designed
to provide a measure of gas transfer in each of the sub-
reaches at least twice, as a test of reproducibility of re-
sults.  As may be seen, the quantities of tracer were varied,
the krypton-85 dose ranging from 131 me to 419 me and the
tritium dose from 420 to 650 me.  A total of 1.63 curies of
krypton-85 and 3.59 curies of tritium were released in the
seven Flint River tracer studies, the average dose being
233 me. of krypton-85 and 513 me of tritium as tritiated
water.

The study locale map, Figure 22, shows the locations of the
sampling stations listed in Table 1, and Table AI.l; in Ap-
pendix AI, provides more detailed station descriptions.  The
stream sampling stations were located generally both for re-
lationship to hydraulic features of interest and for acces-
sibility.

Hydraulic Properties.  The unusually wide variability of hy-
draulic features in the Flint River is clear from the des-
criptions of stations and reaches:  referring to Figure 22,
Station Number 0 was located about 200 feet downstream from
the entry of the Flint River STP effluent, at a point where
the effluent and the stream flow were apparently well mixed;
proceeding downstream, the reach 0-1P includes a highly tur-
bulent rapids section about 150 feet long emptying into the
pond above the first waterfall - Dump XIV included additional
sampling points immediately above and below the rapids sec-
tions; Station IP was only a few feet above the dam; Station
1 was located about 100 feet below the first waterfall; the
reach 1P-1 includes the first waterfall - the water falls
about 10 feet onto granite surfaces, and the reach below the
                           88

-------
00
                                         Table 2
                              Typical  Hydraulic Properties
                                       Flint River
Dump Reach
III 0-1P
0-1
0-2P
0-2
0-3
III 1P-1
1P-2P
1P-2
1P-3
II 1-2P
1-2
1-3
1-4
II 2P-2
2P-3
2P-4
II 2-3
2-4
IV 3-4
3-5
3-6
Flow
cf s
10
10
10
10
11
10
10
10
11
10
10
11
16
10
13
19
14
19
22
23
24
Time of
Travel Length
hrs ft
2.72
2.79
6.05
6.47
8.00
0.08
3.33
3.75
5.28
3.12
3.47
5.04
10.24
0.35
1.92
7.12
1.57
6.77
4.95
9.45
14.77
3,000
3,250
10,000
10,500
15,500
250
7,000
7,500
12,500
6.750
7,250
12,250
22,350
500
5,500
15,600
5,000
15,100
9,600
19,500
27,600
Velocity
ft /sec
0.31
0.32
0.46
0.45
0.54
0.87
0.58
0.55
0.66
0.60
0.58
0.67
0.61
0.40
0.80
0.61
0.88
0.62
0.54
0.57
0.52
Depth
ft
1.19
1.19
0.93
0.93
0.94
1.52
0.82
0.82
0.88
0.82
0.82
0.88
1.34
1.83
0.96
1.61
0.96
1.61
1.95
1.77
1.70
X-sect
Area
sq.ft.
33
31
22
22
20
12
17
18
17
15
16
16
26
24
16
31
15
31
40
40
47
WS Elev
Change Bottom*
ft Char
9.9
23.8
39.6
51.7
61.1
14.0
29.7
41.8
51.2
15,8
27.8
37.2
57.8
12.0
21.5
42.0
9.4
29.9
20.5
27.4
36.1
Sand,
11
"
"
"
Rock
Sand,
"
"
Sand,
"
"
"
Sand
"
Sand,
Sand
Sand ,
Sand,
"
n
mud
"
"
"
"

sludge
"
"
sludge
n
"
"


sludge

Sludge
Sludge
n
ii

-------
                               Table 2  (continued)
                          Typical Hydraulic Properties
                                   Flint River
Dump Reach
V 4-5
4-6
4-7
V 5-6
5-7
V 6-7
Flow
cfs
21
24
24
25
26
28
Time of
Travel
hrs
4.55
9.90
17.73
5.35
13.18
7.83
Length
ft
8
17
26
8
17
9
,900
,000
,400
,100
,500
,400
Velocity
ft/sec
0.
0.
0.
0.
0.
0.
54
48
41
42
37
33
Depth
ft
1.58
1.57
1.57
1.55
1.55
1.56
X-sect
Area
sq ft
39
49
59
58
69
84
WS Elev
Change Bottom*
ft Char -
8.1
16.8

8.9


Sand ,
ii
ii
Sand,
it
Sand,
Sludge
ii
ii
Sludge
H
Sludge
*Predominant bottom character

-------
  920
LJ
                                                                        FIGURE 23
                                                                   CHANNEL  PROFILE
                                                                     FLINT RIVER
                                                                 JUNE I2-JULY3IJ968
  760
                                30     40     50     60      70
                                   DISTANCE  IN 500 FT. STATIONS

-------
fall remains quite shallow and fast above Station 1; the
reach 1-2P is similar to the reach 0-1P, but contains no
substantial rapids section; Station 2P was located at the
downstream end of the second mill pond, a few feet above the
dam; the reach 2P-2 includes the second waterfall (about 12
feet in height)  and a relatively deep pool (6 to 8 feet deep)
immediately underneath the fall, station 2 being about 100
feet downstream from the pool in the main channel; the reach
2-3 is characterized by alternating riffles and shallow pools,
Station 3 being located just above a gravel plant; the marsh
is located below Station 3, and Mud Creek enters the Flint .
at the lower end of the marsh - Dump IV included additional,
sampling stations above and below the marsh and in Mud
Creek, and demonstrated that some water from the Flint River
above the marsh (Station 3) crosses the marsh to enter Mud
Creek and then returns to the Flint with the Mud Creek flow
below the marsh; the remainder of the study reach, from
Station 4 to Station 7, is relatively uniform, has a shal-
lower slope, and becomes somewhat larger than the upstream
section.

Figure 23 shows the Flint River channel profile for the reach
studied.  Table 2 summarizes the typical hydraulic properties
of the sections studied.  These results represent field mea-
surements of relevant hydraulic features at 500-foot inter-
vals, as described in greater detail in Section V.  The de-
tailed hydraulic properties are provided in Appendix AIII.

Reaeration Coefficients.  Table 3 is a summary of a'..l of the
reaeration coefficients observed in the Flint River
studies,
for the main study reaches.   These studies included iseven
separate tracer releases (Dumps I> II, III, IV, V, XvEI, XIV)
as shown, and 56 separate main reach values of Kox were ob-
served.  In most cases at least two separate values of Kox
were observed for the same reach, both in order to test re-
producibility and to include different flow conditions when
possible.

As indicated earlier, the Flint incorporates a very wide
range of hydraulic features  within the 10 miles or so studied,
from waterfalls to rapids, pools and a swamp.  As a result a
very wide range of values of Kox has been observed, from
0.099 to 15.1 per hour at 25°C.  Referring to Table 3, the
reach 1P-1 includes the first waterfall and the reach 2P-2
the second.  The reach 0-1P  includes a highly turbulent ra-
pids section, and the reach  3-4 includes the swamp.  These
latter two reaches were the  subject of special studies
(Dumps IV and XIV), as noted below.

Reference to Table 3 indicates that in general the results
                            92

-------
                        Table 3
     Observed Reaeration Coefficients, Flint River
                  Main Study Reaches
Reach

0-1P
0-1
0-2P
0-2
0-3
1P-1
1P-2P
1P-2
IP-3

1-2P
1-2
1-3
1-4

2P-2
2P-3
2P-4

2-3
2-4

3-4
3-5
3-6

4-5
4-6
4-7

5-6
5-7
6-7

I II
0.314
0.438
-
0.284
0.540 0.520
0.449
0.361
2.72
0.732
0.397
0.280
0.271
0.268
- -
-
_ _
- -
-
_ _
.-
_ _
K per
ox ^
III
0.166
0.450
0.343
0.461
(0.420)
10.2
0.488
0.678
(0.551)
0.251
0.471
(0.403)
-
2.22
(0.663)
-
(0.236)
-
—
-
-
—
-
-
_
-
_
hour @ 25°C
IV
-
-
—
-
-
-
—
-
-
-
-
0.236
0.185
(0.131)*
0.131
(0.080)*
-
(0.037)*
-
_
V
-
-
—
-
-
-
—
-
—
-
—
—
-
-
0.116
0.107
0.112
0.099
0.110
0.118
XII XIV
0.143
0.377
0.328
- 0.453
15.1
0.541
0.778
-
0.270
0.536
-
- -
2.89
-
- -
_ _
— —
- -
-
— —
0.097
0.100
0.111
0.102
0.116
0.125
Mean
0.155
0.380
0.336
0.451
12.7
0.515
0.728
t-
0.268
0.517
0.426
-
2.61
0.698
—
0.258
—
0.252
-
—
0.115
0.104
0.112
0.101
0.113
0.122
*Questionnable result - not included in mean,
                            93

-------
are highly reproducible from one tracer release to the next,
with few exceptions.   The results that involve Station 6 in
Dump IV are noted as  questionnable because of very poor
counting statistics (very little dissolved krypton-85 re-
mained at Station 6 in that case).  Certain other sources of
deviation from the norm remain in the results—for. example,
Dump XIV took place at a considerably lower flow than oc-
curred during Dumps I and III, and Dump XIV also took place
on a Saturday at a lower than usual degree of pollution.
The significance of these and other factors is discussed at
a later point.

Reference to the data of Table 3 also indicates that, in
terms of the gas transfer capacity, this section of the
Flint can be conveniently regarded as three separate main
reaches, namely, the  reaches 0-2, 2-4, and 4-7.  The reach
0-2 had a mean K   of 0.45 per hour at 25°C, within very
narrow limits (0.44-0.46, three observations); the reach 2-4
is characterized by a Kox of 0.26 per hour at 25°C, about
half that of the reach 0-2, also within quite narrow limits
(0.24-0.28, five observations over the reaches 2-3, 2-4 and
3-4); the lower study reach, 4-7, is characterized by a ^Kox
of 0.11 per hour at 25°C, also within very narrow limits
(0.10-0.13, 13 individual reach observations).

It is evident also from these results that in a stream like
the Flint much of the real work of gas transfer may take
place in very short reaches.  Thus, within the reach 0-2 most
of the gas transfer occurred at the two waterfalls.  This
and other similar effects are more directly demonstrated by
reference to Figures  24 through 26.  Figures 24 through 26
are a graphical presentation of all of the gas transfer re-
sults for the Flint River.  According to the theory, as out-
lined earlier, the observed krypton:tritium ratios have been
plotted for each dump against time of flow on semilbg scales,
and the slope of the  straight line between any two observed
ratios represents the tracer gas transfer rate coefficient.
(It should be recalled here that the logarithmic scale shown
refers to common logs, and therefore to kkr, whereas the
numerical values shown on the graphs are referenced to the
base e, and are values of Kkr) .  It should be noted also that
in preparing these Figures it was desirable for purposes of
clarity to displace certain stations in time—thus, for ex-
ample, in Figure 24 (Upper Study Reaches) the results for
Dump II are displaced to the right by about four hours, in
order to avoid overlapping with Dumps III and XIV.

The numerical values  of Kkr shown on the graphs are refer-
enced to the prevailing river water temperatures at the time
of the specific tracer release  (see Appendix AIV for details)
                            94

-------
o


<


z
o
»—I
(-
<
o
u
a:
   0.02
                             468

                             TIME OF  FLOW/  HOURS
10
12
                                  95

-------
The values of KQX at 25°C in the preceeding Table 3 have
been obtained from these observed values of Kkr by use of
the basic Kkr: K   ratio of 0.83 and the temperature correc-
tion coefficient 0 = 1.022/°C.

Referring now to Figure 24 (Upper Study Reaches), and noting
the parallel nature of the lines, the reproducible pattern
of gas transfer in the specific stream reaches is obvious.
Dumps II and III were at comparable flows (about 10 cfs) ,
whereas Dump XIV was at a lower flow (about 5 cfs) .  River
temperatures were somewhat different (24-27°C for Dumps III
and XIV, and about 19°C for Dump II).  The observed values
of Kkr ranged from a low of 0.014/hr (reach R3-1P,  Dump
XIV) in the pool above the first waterfall to 12.35/hr
(reach 1P-1, Dump XIV)  over the first waterfall.

Times of flow were considerably longer for Dump XIV, reflec-
ting the lower flow, but this did not markedly affect the
observed values of Kkr or Kox at 25°C.   The observed values
of K^j- for specific reaches are highly consistent,  allowing
for the prevailing differences  of river temperature, flow
and degree of pollution.

Further and more direct comparisons can be made in terms of
the actual loss of tracer gas in the specific stream reaches.
Recalling that for any reach the decimal fraction of tracer
gas remaining at the downstream station is just the down-
stream station krypton:tritium concentration ratio divided
by the upstream station ratio,  the fractions of tracer gas
transfer for any specific reach may be compared for different
dumps.  For example, for the reach 1-2, Dump III, the ob-
served krypton:tritium ratios were 0.0664 and 0.0152 at
stations 1 and 2 respectively.   Hence,  at station 2 there
remained 22.9 percent of the tracer gas that was present
earlier at station 1:


                   ^Ml- io°= 22-9

In other terms, 77.1 percent (100-22.9) of the tracer gas
was lost to the atmosphere between stations 1 and 2.

The reach 1-2 was included in all three dumps shown in
Figure 24.  The results of such computations show that for
this reach the average tracer gas loss was 79.3 percent,
with only slight variation:
                            96

-------
              Dump Number      Percent Tracer Gas Lost

                  II                    74.9

                 III                    77.1
                 XIV                    85.9

Similar computations for the other upper study reaches demon-
strate great consistency.  For further example, for the
reach 2P-2, which represents the second waterfall, an aver-
age of 57 percent of the tracer gas was lost, the observed
range of results being 50-66 percent.  All of the observed
data have been reported in terms of percent loss of tracer
gas, as well as in terms of Kkr and KQX, and these results
are presented also in Appendix AIV.  This mode of interpre-
tation of the data, and its significance, is discussed in
greater detail in a subsequent Section of this report.

It is also evident from Figure 24 that for a stream like the
Flint most of the gas transfer takes place in short reaches.
This is obvious, so far as the two waterfalls are concerned,
but is also seen to be true through the rapids section, Rl-
R3, contained within the reach 0-1P.  Referring to Figure
24, it may be seen that the Kkr for the reach 0-1P has the
values 0.117 and 0.142 per hour for Dumps XIV and III, re-
spectively.  However, the more detailed results obtained
from Dump XIV show that, in fact, most of the actual gas
transfer took place in the short rapids reach R1-R3, and
that a very high K^ prevailed there (1.94 per hour); in the
pool R3-1P very little real gas transfer occurred, corre-
sponding to a quite low K^r (0.014 per hour).  More specifi-
cally, in Dump XIV 42.6 percent of the tracer gas present at
Station 0 was lost in the reach 0-1P, during the 4.74 hours
time of flow; of this total gas loss, about 50 percent oc-
curred in the reach 0-Rl (1.75 hours), 36 percent occurred
in the rapids section R1-R3 (0.12 hours) and only 5 percent
occurred in the pool section R3-1P  (2.87 hours).

Figure 25 illustrates the observed data for the Middle Study
Reaches, 2-4.  Considering the reaches 2-3 and 3-4, the ob-
served numerical values of Kk  are quite consistent  (0.184-
0.202 per hour).  Dump IV included intermediate stations
between 3 and 4, for purposes of further evaluating the ef-
fect of the swamp or marsh.  As may be seen, the marsh it-
self, reach BS-AS, yielded a value of K^,^ quite consistent
with the other observed results, indicating that gas trans-
fer through the marsh was not markedly reduced.

Figure 26 provides the observed results for the Lower Study
Reaches, 4-7.  The parallel lines again indicate highly
                           97

-------
KRYPTON ."TRITIUM CONCENTRATION RATIO
P P P CD CD CD CD CD CD CDCDP
2 R p CD 0 CD 0 ~ rO IM It )£}
*— ^ ^ -f ui ^cn oo CD CD CD CD 0



^




>&*
X£*^
x
s

%r~
^%)
T5



N|.r
@^k






s_
^^
\
-DUMP III



-DUMP IV
>^5)
\^ ^ /
*\^?
C^^>^
' oLy '"^
w<7



r DUMP I I
X
\

MIDE
OBSERVE
(F





>D0




S|)
FIGURE 25
FLINT RIVER
LE STUDY REACHES
D KRYPTON TRANSFER
!IVER TEMPS,)




















KEY
@ - STATION 3
0-187 - KKR/HR
0 - OBSERVED RATIO
2
          6          8
TIME OF FLOW,  HOURS
10
12
                 98

-------
o
I—t

<


•z.
o
I-*

<
01
o

o
o
o:
j-
f-
z
o

Q-

o:
                                             FIGURE 26

                                            FLINT RIVER

                                         LOWER STUDY  REACHES

                                     OBSERVED KRYPTON TRANSFER

                                          (RIVER TEMPS,)
                  DUMP XII
                                                    KEY

                                             0 -  STATION
                                             0.080  - KKR/HR
                                            O  -  OBSERVED RATIO
0
0,03
  0,02
                            8         12         16

                          TIME OF FLOW/ HOURS
                                 99

-------
consistent results for the specific reaches, as do the ob-
served values of Kkr.  The observed krypton:tritium ratio
for Station 6, Dump IV, is a dubious result due to poor
counting statistics, as noted earlier.  The loss of tracer
gas for the whole reach 4-7 was remarkably consistent in
Dumps V and XII, being 80.2 and 79.8 percent, respectively,
over the 18-hour flow time.  As indicated earlier, the Flint
becomes somewhat larger below Station 4,  and its general
slope somewhat smaller.  The preceeding Table of Typical
Hydraulic Properties makes clear these and other significant
features.

DO and BOD Results.  The major pollution load received by
the Flint River in the section included in these research
studies came from the Flint River STP, the plant effluent
entering the Flint about 200 feet above Station Number 0.
During the 1969 research study period, the existing Flint
River STP (trickling filter) was overloaded:  the influent
sewage flow averaged about 2.1 mgd, with an average 5-day
BOD of 262 mg/1, and on the average, about 7 to 8 percent
of this influent flow had to be bypassed.  An additional
smaller pollution load entered the Flint via Sullivan Creek
immediately below Station Number 2:  during the 1969 study
period, the Sullivan Creek STP (trickling filter) treated
an average influent flow of 0.82 mgd having a 5-day BOD of
191 mg/1, and accomplished an average BOD removal of 90 per-
cent or a little better.  Some additional pollution entered
the Flint via Mud Creek (between Stations Number 3 and 4).
Table 4 provides a summary of relevant information collected
by the Georgia Water Quality Control Board22.

                        Table 4

                   Flint River, 1969
             Summary of DO and BOD Results

  Location       Temperature        DO            BOD5
  	       	°C           mg/1           mg/1
Above Flint         14-24        5.4-8.8          3-9
  River STP

Below Flint         14-24        2.9-5.8          4-77
  River STP

(Sullivan Cr)       (14-23)      (3.0-6.2)         6-18

Below station 2     15-24        1.8-5.2         16-55

Station 6           16-25        0.8-4.4          3-13

These results for the period April 30 - October 22, 1969,
provide some indication of the highly variable nature of the
                          100

-------
BOD load received by the Flint, but tell only part of the
story.  More recent information from the Georgia WQCB indi-
cates clearly that both the Flint River STP effluent flow
and its 5-day BOD may vary over quite a wide range during
the 24 hours of any one day.  In the upper reaches of the
Flint, this plant effluent constitutes a major portion of the
river flow.  In addition, a substantial portion of the oxy-
gen demand on the river results from the ammonia load pres-
ent in the plant effluent.  During the reaeration tracer
studies available staff and equipment did not permit the ex-
tensive and thorough compositing of BOD samples that would
be necessary for highly accurate oxygen balance analysis.
Nor was the need for thorough analysis of samples for the ni-
trogen series foreseen.  Hence, the BOD results presented
below can be regarded only as a momentary representation of
the quality of the Flint River, and it must be understood
that the quality may very well have been significantly dif-
ferent a few hours earlier or later.

In addition to the foregoing problems associated with con-
ducting an accurate oxygen balance, others are involved.
Zero DO's were observed in the pools above both dams in the
Flint River, and such results must be regarded as indeter-
minate for purposes of oxygen balance.  No doubt, they were
due in part to the presence of organic benthal deposits, es-
pecially in the pool above Station Number IP.  Also, growths
of bacterial slimes and some algae were observed on rock sur-
faces between Stations Number 2P and 2, where a fraction of
the river flowed over those rocks.  A considerable number of
the BOD-time series that were performed also displayed an
erratic BOD process, typical of samples in which there is
substantial nitrification.

As a result of the foregoing difficulties, some of which were
not anticipated, no detailed oxygen balance has been attemp-
ted for the whole Flint River section studied, and, hence,
only very limited evaluations of the DO effects of benthal
decomposition and of the respiration of attached bacterial
growths can be presented here.

Table 5 provides a summary of the DO and BOD data collected
at the two waterfalls on the Flint during the gaseous tracer
study periods, and these results do allow some limited esti-
mates of interest here.  Referring to Table 5, the observed
5-day BOD's at Station Number 1 and 2 were within the range
usually observed by the Georgia WQCB, and were relatively
high.  At both waterfalls, reaches 1P-1 and 2P-2, substantial
net DO increases were observed - an average of 3.4 mg/1 in
reach 1P-1  (typical time of flow 5 minutes) and 4.3 mg/1 in
reach 2P-2  (typical flow time 21 minutes).  Within the
                           101

-------
limitations imposed by certain sources of uncertainty out-
lined below, comparison of the observed net gain of DO with
the total DO gain obtained from the gaseous tracer data pro-
vides information of interest.

                        Table 5
                   Flint River, 1969
           DO and BOD Results at Waterfalls
                       DO, mg/1
1st Waterfall
Date (Dump)
4/25 (I)
5/2 (II)
5/24 (III)
6/24 ( — )
8/30 (XIV)
IP
0.0
0.0
0.0
nd
0.0
1 "
3.3
3.0
3.3
nd
4.0
2nd Waterfall Temp, °C BOD
2P
0.0
0.8
1.6
0.0
0.1
2
3.9
5.0
5.6
4.8
4.5
1P-1
17
22
26
nd
22
2P-2
13
20
25
23
22
1
nd*
41
58
nd
61
                                                       nd
                                                       32
                                                       30
                                                       29
                                                       44
*nd:  not determined.


Reference to equations (3),  (7)  and (11)  of Section IV of
this report indicates that

                lo  (^-)  =   1   lo  A
                     D      0.83      R
                      **                i~i

where A and B refer to upstream and downstream river sampling
stations, respectively, D is the DO saturation deficit, in
mg/1, and R is the krypton:tritium concentration ratio.
Equation (41) may be used to  evaluate the total DO added at
the waterfalls.  Any significant difference between the
total DO added and the observed net DO increase reflects
sources of oxygen consumption in the reach, assuming no
significant errors of measurement.

Certain sources of error are present in the data.  The DO and
BOD samples could not usually be collected exactly at the
time of arrival of the peak  dye concentration at each sta-
tion, as many other tasks had to be performed then, and, con-
sidering the highly variable BOD load in the river, this po-
tential source of discrepancy may not always have been insig-
nificant.  Also, the krypton:tritium concentration ratios at
Stations 1 and 2 are subject to some error due to low tracer
gas concentrations resulting from the large gas losses over
the waterfalls.  DO's were analyzed by use of a Yellow
                           102

-------
Spring? instrument, as well as by the azide modification of
the Winkler method, and some'apparent differences in results
were observed.  Also, the data are not complete in every
case - for example, although DO's were observed at Stations
IP and 1 for Dump II, the tracer release was made at Station
1, so there are no tracer gas transfer data for the reach
1P-1, Dump II.

Referring now to the reach 1P-1, krypton tracer data are
available for Dumps III and XIV.  Referring to the detailed
tracer data in Appendix AIV, the average value of the ratio
(R^/R-^p) was 0.448, and, from equation (41), the ratio of
saturation deficits is just 0.380.  At the mean temperature
of 24°C, and for an elevation of 880 feet above mean sea
level, this indicates a total DO gain between Stations IP
and 1 of 5.1 mg/1.  The observed net gain of DO was 3.7 mg/1,
indicating DO consumption within the reach of about 1.4 mg/1.
This estimate does not allow for any reduction in the effec-
tive value of DO saturation due to the presence of consider-
able pollution  (5-day BOD of 60 mg/1), hence it is likely
that the actual DO consumption between IP and 1 was somewhat
less than 1.4 mg/1, perhaps by 0.2 or 0.3 mg/1.

The reach 1P-1 is very shallow  (a few inches deep) and swift,
and the stream bed is largely rocks and stones.  During these
studies, extensive growths of bacterial slimes attached to
the rocks were observed, together with relatively moderate
to low quantities of blue green algae.  That such bacterial
growths can constitute an important source of oxygen consump-
tion, even in a flow time as short as 5 minutes, is evident
from the BOD data for Dumps III and XIV.  Referring to Table
5, the average 5-day BOD's at Stations 1 and 2 were 60 and
37 mg/1, respectively; for these Dumps, the time of flow for
the reach 1P-2 averaged 4.1 hours.  There is no significant
gain in streamflow between 1 and 2.  The reduction from 60
mg/1 to 37 mg/1 in 4.1 hours leads to an average value of
(K1)r;j_ver for the reach 1P-2 of about 0.12 per hour.  The
associated DO consumption for the reach 1P-1, even in a 5-
minute time of flow, would be about 0.7 or 0.8 mg/1 of DO.
Furthermore, the observed (Kj)river of 0.12 per hour is an
average value for the whole reach 1P-2, and it is likely that
the actual magnitude of this rate coefficient was greater in
short reaches like 1P-1 and lower in other pooled reaches
between Stations 1 and 2, so that the estimated loss of 0.8
mg/1 of DO in the reach 1P-1, due to attached oxidizing
growths, may tend to be a somewhat low estimate.  All such
factors considered, a DO consumption of 1.0 mg/1 between IP
and 1 due to respiration of attached oxidizing growths is re-
garded as a close estimate.
                          103

-------
Similar computations may be made for the second waterfall,
reach. 2P-2.  Krypton tracer data are available for Dumps II,
III and XIV.  The 5-day BOD's were lower for this reach,
averaging about 35 mg/1 for the three Dumps.  More impor-
tantly, the reach 2P-2 is quite different in the hydraulic
sense than the reach 1P-1, the reach below the second water-
fall being pooled immediately below the falls (pool depth
about 6 feet)  and remaining relatively deep (about 2 feet)
and slow above Station 2, with a stream bed of mud and sand.
As a result, attached oxidixing growths do not occur in
large amount between 2P and 2 except on the surfaces of ele-
vated rocks at one side of the waterfall - only a small frac-
tion (perhaps 10 percent) of the flow from Station 2P flows
over these growths.  Instead, there is ample opportunity for
the accumulation of organic bottom deposits between 2P and
2, as well as in the larger pool above Station 2P-  Although
no specific observations were made, it is likely that bottom
deposits of organic material between 2P and 2 were relatively
small,  as most of any settleable BOD load would have been
deposited in the large pool above Station 2P.

The krypton tracer data indicate for the three Dumps an
average value of 0.432 for the ratio (Rj/R2p)•  From equation
(41)   and the data of Table 5, this indicates a total gain of
DO between 2P and 2 of 4.9 mg/1.  The net observed DO gain
between the two stations averaged 4.2 mg/1, indicating
sources of oxygen consumption between the two stations that
utilized about 0.7 mg/1 of DO.  This is regarded as a good
estimate of the combined effects of DO loss from the frac-
tion of the stream flow that was exposed to the attached ox-
idizing growths and the DO loss from minor bottom organic
deposits in the reach.

SOUTH RIVER

The South River originates in southwest Atlanta and forms
part of the headwaters of the Altamaha River system.  Like
the Flint, the South flows through part of metropolitan
Atlanta, and carries both domestic and industrial wastes  as
well as urban storm drainage.  The South flows southeasterly
approximately fifty miles before joining the Ocmulgee.  The
latter subsequently joins the Oconee to form the Altamaha,
which then flows eventually into the Atlantic Ocean near
Brunswick, Georgia.

The portion of the South River  included in  these  studies  ex-
tended from the South River STP to a point  about  one mile up-
stream from the intersection  of the South River and Flat
Bridge Road, a distance  of 18.3 river miles.  Figure  27 is
a general map of the South River study locale, and shows  sa-
lient features such as highways, tributary  streams and  the
                          104

-------
         FIGURE  27

SOUTH RIVER STUDY LOCALE
   VICINITY  OF ATLANTA
                                                                                       REPRESENTS EVERY
                                                                                       TENTH CROSS SECTION

-------
river sampling stations.    Within the 18-mile section stud-
ied, the South receives treated or partially treated domes-
tic and industrial wastes from five treatment plants, namely,
the Old and New South River STP's at the upper end, the
Intrenchment Creek, STP, the Shoal Creek STP< and the Snapfin-
ger STP, as well as from other sundry sources.  All of these
wastes enter the stream above Panola Shoals.  The South was
heavily polluted, as a result, throughout the entire study
section.  Large quantities of floating foam were seen fre-
quently, especially in a pool immediately below Panola Shoals

The South River channel is relatively uniform.  A typical
section in the upper third of the 18-mile study section is
thirty to forty feet wide and one to two feet deep.  The
width and depth increase proceeding further downstream, but
the depth:width ratio remains fairly constant.  The channel
is relatively straight,, with high steep banks.  The bottom
material is primarily sand.  There are occasional pools fol-
lowed by short reaches of rapids, but the predominant char-
acteristic of the channel is its uniformity.  The most un-
usual hydraulic feature occurs at Panola Shoals, where the
water flows in thin sheets over large smooth granitic rocks,
falling about eight feet in a matter of about 30 yards.  The
largest and deepest pool is located immediately above Panola
Shoals.  The river attains a width of approximately 100 feet,
with depths of six to seven feet, in this pool.

Dosing Pattern.  Table 6 shows the dosing pattern that was
used in the 1969 tracer studies of gas transfer in the South
River.  This pattern of eight separate releases was designed
to provide at least two separate measures of gas transfer in
each of the subreaches, to verify reproducibility; in the
case of reach K-L  (Panola Shoals) four separate observations
of Kkj- were made.  As may be seen, the quantities of tracer
were varied somewhat, the krypton-85 dose ranging from 296
me to 597 me and the tritium dose from 349 me to 826 me.  A
total of 2.93 curies of krypton-85 and 3.47 curies of tri-
tium were released in the eight South River tracer studies.

The study locale map, Figure 27, shows the location of the
sampling stations listed in Table 6, and Table AI.2, in
Appendix AI, provides more detailed station descriptions.
The stream sampling stations were selected both for relation-
ship to hydraulic features of interest and for accessibility.

Hydraulic Properties.  Table 7 summarizes the typical hydrau-
lic properties of the sections studied.  These typical re-
sults represent  field measurements of relevant hydraulic
features at 500-foot intervals, as described earlier.  The
detailed hydraulic properties are provided in Appendix AIII.
                          106

-------
                                        Table 6

                               Dosing Pattern, South River
      Dump    	South River Sampling Station Number	
     Number    ABD      E      F      GH      JKLM      N
       VI ...  6/16
              323
              349
      VII	 6/18
                                               330
                                               380
        X ... 6/25
              311
              357
°    VIII 	  6/20
                                                            345
                                                            370
       IX 	 6/23
                                305
                                387
       XI 	 6/27
                                        296
                                        370
     XIII	 8/27
                                                           597
                                                           432

-------
                                     Table  6  (continued)

                                 Dosing  Pattern,  South  River
        Dump     	South  River  Sampling  Station Number	
       Number     ABD       E      F      GH       JKLM      N
         XV	  9/26
                                          421
                                          826
o
CO
      Key:   6/16  - Dump VI  released  6/16/69  at Station A
             323   - 323 me krypton-85 released
             349   - 349 me tritium  released

-------
Figure 28 shows the channel profile for the South River
reach studied.

As indicated above, the South River channel is relatively
uniform, with few exceptional features.  Flow throughout the
study reach varies somewhat depending upon effluent flows
from the several waste treatment plants and from the tribu-
tary streams.  A substantial rapids section occurs in the
upper portion of the reach G-H, and was the subject of spe-
cial investigation in Dump XV.  The reach J-K includes the
large deep pool above Panola Shoals.  This pool is about
800 feet long, with depths up to six or seven feet, and a
typical observed flow-through time of about 2.5 hours.  As
will be shown below, the observed results indicate that the
incoming flow often does not mix completely with the pool
contents - this incomplete mixing phenomenon is apparently
the rule, rather than the exception, in this pool.  The nat-
ural rock barrier at the lower end of the pool (Station K)
develops gradually, rather than being like a vertical dam
wall, and the pool depth decreases gradually toward this
barrier, rather than being at its deepest at Station K.  As
a result, depending upon other factors such as the tempera-
ture differential between the pool contents and the incoming
river flow, the rate of inflow, etc., at some times the flow
over the barrier at Station K may be largely pool surface
water and at other times may be primarily bottom water from
the pool depths.

The reach K-L includes Panola Shoals, where the flow spreads
thinly over wide smooth granitic rock surfaces to fall ap-
proximately eight feet into a small mixing basin or pool.
The time of flow for this reach is about 12 minutes, during
which, as will be seen, considerable gas transfer occurs.

Reaeration Coefficients.  Table 8 provides a summary of all
of the reaeration coefficients observed in the South River
studies, for the main study reaches.  The South River stud-
ies included eight separate tracer releases  (Dumps VI, VII,
VIII, IX, X, XI, XIII, XV) as shown, and 85 separate main
reach values of Kox were obtained.  In most cases at least
two separate values of K   were obtained for the same reach,
and in some cases, notably the reaches G-H, H-J, J-K, and
K-L, four or more tests were conducted.

Reference to Table 8 indicates that reproducibility of the
test results was generally very good.  The South River chan-
nel profile is not so steep as the Flint River profile, and
the channel is considerably more uniform.  It is similar to
the Lower Study Reaches of the Flint.  The observed KQX
values are correspondingly lower, in the general neighborhood
                           109

-------
           Table 7
Typical Hydraulic Properties
         South River
Dump
VI
"
"
11
VI
II
II
VI
II
IX
II
II
II
II
IX
II
II
II
VII
It
II
II
Reach
A-B
A-D
A-E
A-F
B-D
B-E
B-F
D-E
D-F
E-F
E-G
E-H
E-J
E-K
F-G
F-H
F-J
F-K
G-H
G-J
G-K
G-L
Flow
cfs
47
62
73
76
63
74
78
89
92
72
80
83
113
115
82
86
116
118
124
139
141
141
Time of
Travel Length
hrs ft
2.48
3.85
6.28
7.98
1.37
3.80
5.50
2.43
4.13
1.78
6.53
7.17
10.60
13.82
4.75
5.38
8.82
12.03
0.43
3.55
6.82
7.00
10,250
16,250
27,650
36,650
6,000
17,400
26,400
11,400
20,400
9,000
27,400
30,200
40,600
49,200
18,400
21,200
31,600
40,200
2,800
13,200
" 21,800
22,300
Velocity
ft/sec
1.15
1.17
1.22
1.27
1.21
1.27
1.33
1.30
1.37
1.40
1.16
1.17
1.06
0.99
1.07
1.09
1.00
0.93
1.81
1.03
0.89
0.88
X-sect
Depth Area
ft sq ft
1.05
1.30
1.29
1.21
1.51
1.40
1.37
1.34
1.30
1.39
1.64
1.67
1.75
1.84
1.81
1.78
1.85
1.94
1.90
1.97
1.82
2.12
40
53
59
60
52
58
58
68
68
51
69
71
107
116
77
79
116
127
69
135
158
160
WS Elev
Change
ft
12.1
24.0
36.5
44.8
11.9
24.4
32.7
12.5
20.8
8.6
24.2
31.3
45.5
47.7
15.6
22.7
36.9
39.2
7.7
21.9
24.0
36.2
Bottom*
Char
S and , mud
n n
11
"
Sand
"
"
Sand, rock
"
Sand
"
" , rock
n n
n M
Sand, rock
n n
n n
11 II
Rock
" , sand
n n
n n

-------
                              Table 7 (continued)
                         Typical Hydraulic Properties
                                  South River
Dump
VII
"
"
VIII
"
"
"
VIII
II
II
VIII
II
VIII
Reach
H-J
H-K
H-L
J-K
J-L
J-M
J-N
K-L
K-M
K-N
L-M
L-N
M-N
Time of
Flow Travel
cfs hrs
139
141
141
188
188
196
198
190
198
200
198
200
207
3.12
6.38
6.57
2.72
3.00
4.85
7.30
0.28
2.13
4.58
1.85
4.30
2.45
Length
ft
10
19
19
8
9
18
25

9
16
8
16
7
,400
,000
,500
,600
,100
,000
,500
500
,400
,900
,900
,400
,500
Velocity
ft/sec
0.93
0.83
0.82
0.88
0.84
1.03
0.97
0.50
1.22
1.02
1.34
1.05
0.85
Depth
ft
2.10
2.12
2.15
2.40
2.32
2.02
1.86
1.67
1.88
1.58
1.70
1.58
1.43
X-sect
Area
sq ft
149
170
172
214
224
190
204
380
162
196
148
190
243
WS Elev
Change
ft
14.2
16.4
28.5
2.1
14.3
19.0
28.0
12.2
16.9
25-9
4.7
13.7
9.0
Bottom*
Char
Sand,
it

Sand,
11
"
11
n
Sand,
11
Sand,
11
Sand,
rock
mud

mud
"
11
11
«
11
11
mud
11
mud
*Predominant bottom character.

-------
   81
Id

uj 775
5 755
UJ

m  735
                                                                             FIGURE 28

                                                                       CHANNEL PROFILE

                                                                          SOUTH RIVER

                                                                      AUG  5-SEPT.  13,1968
u.
2:
   715
UJ
  695
UJ
  675
  655
                                                    G)  CH)
                   20      40      60      80      100      120

                                    DISTANCE IN 500 FT STATIONS
                                                                140
160
180
200
220

-------
(jj
                                         Table  8
                     Observed Reaeration Coefficients,  South  River
                                  Main  Study Reaches
Reach
A-B
A-D
A-E
A-F
B-D
B-E
B-F
D-E
D-F
E-F
E-G
E-H
E-J
E-K
F-G
F-H
F-J
F-K
G-H
G-J
G-K
G-L
K per hour. @ 25° C
OX
VI VII VIII
0.244
0.285
0.252
0.240
0.358
0.258
0.238
0.204
0.200
0.194
- -
_ _
_ _
_ _ _
_ — _
_ _
_ _ _
_ — _
(0.920)*
0.336
0.173
0.283



0
0
0
0
0
0
0
0
0
0
0
0

IX
-
-
.178
.150
.179
.205
.188
.139
.180
.211
.189
.482
.294
.222
™~
X
0.164
0.208
0.182
0.184
0.289
0.194
0.192
0.144
0.162
0.188
-
-
-
-
0
0
0
0
0
0
0
•"
XI XIII
-
-
— —
-
-
-
- -
.111
.168
.212
.177
.592
.326
.216
"" ™"
XV
-
-
-
-
-
-
-
-
-
-
—
0.528
0.296
0.180
0.263

0
0
0
0
0
0
0
0
0
0




0
0
0
0
0
0
0
0
Mean
.204
.247
.217
.212
.324
.226
.215
.174
.181
.187
-
-
-
-
.125
.174
.212
.183
.534
.313
.198
.273
    *Questionnable  result  -  not  included  in  mean.

-------
                              Table 8  (continued)
                Observed Reaeration Coefficients, South River
                             Main Study Reaches
Reach
VI VII
H-J - 0.255
H-K - 0.122
H-L - 0.241
J-K - (-0.004)*
J-L - 0.228
J-M
J-N
K-L - (4.31)*
K-M
K-N
L-M
L-N
M-N
K
VIII IX X
0.260
0.197
_
(0.119)*(0.131)* -
0.237
0.171
0.154
(1.37)*
0.237
0.175
0.066
0.098
0.122
per hour .<§. 2.5 °.C
XI XIII XV
0.279 - 0.263
0.183 - 0.157
0.246
(0.093)* (0.028)*(0.067)*
0.219 0.231
0.172
0.173
(3.25)* (2.92)*
0.373
0.265
0.096
0.141
0.175

Mean
0.264
0.165
0.244
(0.072)*
0.229
0.172
0.164
(2.96)*
0.305
0.220
0.081
0.120
0.149
*Questionnable result

-------
of 0.2 per hour at 25°C, with a low of perhaps 0.03 or 0.05
per hour in the pool above Panola Shoals to a high of about
3.0 per hour over Panola Shoals.

As noted earlier, the South is heavily polluted from four
separate sources, and this in itself brings about some var-
iability of reaeration capacity.  This is best exemplified
by comparing the observed results for Dumps VI and X, both
of which included the reach A-F.  Both studies were at the
same river flow and essentially the same temperature.  Yet
the observed values of KQX for Dump VI were consistently
significantly higher than the Dump 'X values.  The subject of
pollutant effects on reaeration capacity will be discussed
in greater detail subsequently, but is noted here in terms
of this particular set of observed results.

The reach G-H, containing the rapids section at its upper
end, had a Kox of about 0.53 per hour at 25°C.  Dump XV, a
more detailed study, yielded an observed KQX 1.28/hour at
25°C for the rapids itself, where most of the elevation
change occurred.

The results for the reach J-K, containing the pool above
Panola Shoals, are quite variable compared to any other
reach studied in any of the river studies reported here.  In
six separate studies covering this reach the value observed
for KQX at 25°C varied from 0.00 per hour  (-0.004) to 0.13
per hour.  Without doubt, this high degree of variability is
the result of the erratic and incomplete mixing of the in-
coming flow and the pool contents, as noted earlier, together
with the bottom or top water release at Station K.  The mean
observed Kox of 0.07 per hour at 25°C is judged to be prob-
ably somewhat high.

Because Station K cannot be taken always to be representative
of the whole South River flow, as outlined earlier, the ob-
served values of Kox for Panola Shoals, reach K-L, are also
more variable than usual.  The four observed values ranged
from 1.4 to 4.3 per hour at 25°C.  The mean of 3.0 per hour
is regarded as a close estimate of the reaeration capacity
of the Shoals.

The remarkable consistency of the four observed values of KOx
for the reach J-L, which includes both the pool  (J-K) and the
Shoals  (K-L), provides ample evidence of the validity of the
foregoing outline of hydraulic reasons for the variability of
results in the pool and over the Shoals when taken separately,
For the whole reach J-K the four observed values of Kox
ranged only from 0.219 to 0.237 per hour, and the mean of
0.229 per hour at 25°C is thus seen to be highly accurate.
                           115

-------
Stations J and L were quite representative of the whole flow,
but Station K was very sensitive to the momentary hydraulic
conditions in the pool.

Figures 29 and 30 are a graphical presentation of the results
for the South River studies, and illustrate the foregoing.
The numerical values of Kkr shown on the graphs refer to the
prevailing river water temperatures and flows at the time of
the individual tracer releases, and the values of Kox pre-
sented in the preceeding Table 8 have been obtained by use
of the basic Kkr:Kox ratio of 0.83 and the temperature co-
efficient of 1.022/°C.

Referring now to Figure 29, for the Upper Study Reaches, the
generally parallel nature of the lines is evident.  As noted
earlier, although Dumps  VI and X were performed at essen-
tially the same flow and water temperature, the observed
values of Kkr were consistently higher in Dump VI, indicating
a different pollutant effect.

Reference to Figure 30 (Lower Study Reaches)  illustrates the
behavior of the rapids section in the reach G-H, the long
deep pool in the reach J-K, and the rapid change of eleva-
tion at Panola Shoals, reach K-L. As outlined earlier,
mixing of the inflowing water with the pool contents was er-
ratic and nonhomogeneous in the reach J-K, leading to a wide
range of observed values of Kkr.  In contrast, the four ob-
served values of K,   in the reach H-J, immediately upstream,
were quite consistent (0.255 to 0.279 per hour at 25°C).
The four observed values of Kkr for the reach J-L ranged
only from 0.171 to 0.192 per hour, and led to the equally
consistent set of results for KQX at 25°C  (0.219 to 0.237 per
hour) noted above.

DO and BOD Results.  In the 18.3 mile section of the South
River included in these research studies, the stream re-
ceives wastes from five municipal treatment facilities and
from 16 smaller sources22.  The five municipal plant efflu-
ents are from the Old South River STP (trickling filter, 6.0
mgd) and the New South River STP  (modified activated sludge,
12eO mgd), both located above Station A, the Intrenchment
Creek STP  (high rate trickling filter, 20 mgd) just below
Station B, the Shoal Creek STP  (trickling filter, 3.0 mgd)
just below Station F, and the Snapfinger Creek STP  (trick-
ling filter, 2.0 mgd in 1969) below Station H.  The Snap-
finger Creek STP was greatly overloaded in 1969; it was ex-
panded to 6,0 mgd capacity in 1970, but was still overloaded
at times.

The operation of these waste treatment plants was evaluated
                          116

-------
<
CCL
LU
O
c   o
o
H-
a.
o:
                                                                 FIGURE 29
                                                                SOUTH RIVER
                                                             UPPER STUDY REACHES
                                                          OBSERVED KRYPTON TRANSFER
                                                               (RIVER TEMPS,)
               KEY
           B)- STATION  B
           0.190  -  KKR/HR
           © - OBSERVED  RATIO
    0
    0,10
        0
                                       6         8         10
                                    TIME OF FLOW/ HOURS
12

-------
            2,00
00
o
p
<
        a:
        LiJ
        O
        o
        h-
        Q.
        Ct
                           FIGURE 30
                          SOUTH RIVER
                      LOWER STUDY REACHES
                   OBSERVED KRYPTON TRANSFER
                        (RIVER TEMPS,)
                                                                     DUMP XIII
                                                          DUMP VIII
                         KEY
                     H)~ STATION H
                  0,238 - KKR/HR
                  O-OBSERVED RATIO
           0,04
                                               3          4         5
                                               TIME  OF  FLOW,  HOURS

-------
in 1970 by the Georgia WQCB 3.  During 1970, the total waste
flow from the five plants was 40.1 mgd.  The BODc released
to the South River totalled 13,300 Ibs/day from the five
plants, and the total ammonia nitrogen discharge averaged
3,150 Ibs/day.  These records of the Georgia WQCB indicate
that "The highest organic nitrogen (including ammonia) con-
centrations observed in the river occurred below the con-
fluence with Snapfinger Creek"  (between Stations H and J).
Table 9 provides a summary of relevant information reported
by the Georgia WQCB for the period April 23 - October 9,
196922.

                        Table 9

                   South River, 1969

             Summary o_f DO and BOD Results

                  Temperature       DO          BOD
   Location           °C	      mg/1         ing/

Above South R STP   13-23        6.3-8.8        1-14
  (Above Sta A)

US Hwy 23           14-26        2.6-6.8       15-27
  (Below Sta A)
(Intrenchment Crk)  (16-25)       (2.5-4.9)     (21-40)
Bouldercrest Rd     15-25        2.4-4.8        9-36
  (Sta D)
Panthersville Rd    15-25        2.4-4.8       12-33
  (Sta E)
Ga Hwy 155          15-25        5.2-7.8        6-20
  (Sta L)
Ga Hwy 138          15-26        4.3-8.0        4-10
  (Below Sta N)

It is evident from the foregoing data that the South River
was heavily polluted during the 1969 research study period,
and that there was considerable variability of DO and BOD.
The multiple sources of pollution, the variability of loads
and the variability of flows in the river made accurate
oxygen balance studies for the whole 18.3 mile section a
project beyond the planned scope and the physical capability
of the research staff.  In addition, the presence of large
amounts of organic nitrogen in the river led again to highly
erratic BOD-time series because of the occurrence of nitri-
fication.  As a result, no detailed oxygen balance for the
whole South River study section is presented here.  However,
                          119

-------
the DO and BOD results obtained on separate occasions at
Panola Shoals provide interesting information.  Those re-
sults are summarized in Table 10.

                       Table 10
                   South River, 1969
          D0_ and BOD Results at. Panola Shoals

                  DO, mg/1      Temp        BOD5
   Date (Dump)    K    L          °C        ttig/1

   6/18 (VII)    2.8  5.1        22         7.9

   6/27 (XI)      2.5  6.0        26        12

   8/28 (XIII)   1.9  5.6        23         9.4

Referring to Table 10, the observed 5-day BOD's were within
the range usually found by the Georgia WQCB, and were not
especially high.  For the three Dumps, the observed average
net gain of DO over the Shoals, reach K-L, was 3.2 mg/1, in
an average flow time of 12 minutes.  Krypton tracer data are
available for Dumps VII and XIII (see Appendix A IV) , and
provide an average value for the ratio (RL/RK) of 0.544.  At
the mean river water temperature of 22.5°C and for an eleva-
tion of 700 feet above mean sea level, application of equa-
tion (41)   indicates a total DO gain between Stations K and
L of 3.2 mg/1.  The observed net DO gain for Dumps VII and
XIII averaged 3.0 mg/1, for comparison, as against the aver-
age observed net DO gain of 3.2 mg/1 for all three dates.

The stone surfaces of the Shoals between Stations K and L
were covered by prolific growths of both algae and bacteria
during these research studies, and most of the water flowed
over these growths in relatively thin sheets of flow.  There
can be no doubt that the DO concentration of the flowing
water was affected both by the photosynthetic activity of
the algae and the respiration of the attached bacteria - in
essence, between K and L there was a major source of DO re-
plenishment as well as a major source of DO consumption.
The analysis presented above indicates that these processes
were in near-balance, so far as DO change was concerned -
that is, that about as much DO was added by photosynthesis
as was lost due to respiration.

PATUXENT RIVER

The Patuxent  River below Laurel, Maryland, is a typical rel-
atively small coastal plains stream, nontidal, and
                          120

-------
characterized by alternating small pools and gentle riffles.
Studies of its reaeration capacity were performed in Septem-
ber, 1969, for the Maryland Department of Water Resources as
a part of that Agency's program for protection of the quality
of the stream.

The reach of stream covered by these reaeration studies in-
cluded the critical 7-mile section from the Laurel, Maryland,
Sewage Treatment Plant  (secondary treatment), through the
USDI Patuxent Wildlife Refuge, to a location above Bowie,
Md.  At critical low summer flows the stream depths are rel-
atively shallow, velocities slow for the most part, and the
stream meanders quietly through the Wildlife Refuge.  At
somewhat higher flows, the stream splits into multiple chan-
nels at several locations in the Refuge.

Figure 31 is a map of the section studied, and depicts the
general area, the Wildlife Refuge boundary, location of
stream sampling stations and other relevant features.  The
typical reach between sampling stations was about one mile
in length, or a little more.  At the critical low flows that
prevailed during the tracer study period, the depth of flow
ranged from four to six inches in the riffles to perhaps as
much as two feet in the deeper pools.  The typical riffle
was perhaps 20 to 30 feet in length, with a bottom of small
(three-to five-inch) rounded smooth stones.  Most of the
pools were relatively short and shallow (one foot to 1-1/2
feet deep) and had a mud and silt bottom.  Exceptions were
the longer, deeper pool at Duvall Bridge (Station 4) and the
somewhat deeper flows at the lower end of the study section.

Dosing Pattern.  Table 11 shows the pattern of tracer dos^s
that was used in the 1969 studies of gas transfer and reaera-
tion in the Patuxent River.  The pattern of five separate
tracer releases provided a measure of gas transfer in each of
the subreaches at least twice.  The quantities of tracer used
were fairly uniform, except for the quantity of tritium in
Dump E, and averaged 376 me of krypton-85 and 519 me of tri-
tium.  A total of 1. 879 curies of krypton-85 and a total of
2.595 curies of tritium were released in the five doses.

Hydraulic Properties.  Table 12 provides a summary of typical
hydraulic properties of the Patuxent River study reaches.  As
may be seen, the range of typical velocities and depths is
small compared to similar results for the Flint and South
Rivers.  The range of slopes is also relatively small, and
typical velocities are lower.  No remarkable hydraulic fea-
tures such as violent rapids, falls or shoals occurred in
the Patuxent River study section.  This indicates a higher
degree of hydraulic uniformity than in the Flint or South,
                          121

-------
                                            FORT MEAD
to
        BALTIMORE /
        WASHINGTON
        PARKWAY   /
                                          FIGURE 31
                               PATUXENT RIVER STUDY  LOCALE
                                                                              MILES

-------
NJ
                                         Table  11

                             Dosing Pattern, Patuxent River
Dump
Number
A . . . .


B


C


D . . . .


E ....




1



. 9/6 . .
349 . .
443 . .



. 9/9 . .
362 . .
489 . .




Patuxent River Sampling
2 3 4 5
	 9/5 	
342 	
458 .......



	 9/8 	 	 .
398 	 	
479 	



............... 9/11 	 	
428 	
726 ......

Station Number
6 7
















               Key:   9/5  -  Dump A released  on  9/5/69  at  Station  4
                     342  -  342 me of  kyrpton-85  released
                     458  -  458 me of  tritium released

-------
                                   Table 12
                         Typical Hydraulic Properties
                                Patuxent River
Dump
B
n
n
B
ii
H B
^ A
"
"
A
II
A
Reach
1-2
1-3
1-4
2-3
2-4
3-4
4-5
4-6
4-7
5-6
5-7
6-7
Flow
cf s
9.8
9.8
9.8
9.8
9.8
9.8
19.5
19.5
19.5
19.5
19.5
19.5
Time of
Travel
hrs
3.85
9.08
14.75
5.23
10.90
5.67
6.70
13.92
17.52
7.22
10.82
3.60
Length
ft
5,400
9,600
16,800
4,200
11,400
7,200
8,400
15,000
19,800
6,600
11,400
4,800
Velocity
ft/sec
0.39
0.29
0.32
0.22
0.29
0.35
0.35
0.30
0.31
0.25
0.29
0.37
X-sect
Depth Area
ft sq ft
0.80
0.90
0.90
1.00
1.00
1.00
1.10
1.10
1.10
1.10
1.05
1.00
25
34
31
45
34
28
56
65
63
78
67
53
WS Elev
Change Bottom*
ft Char
7.0 Mud, sand
11.9
21.9
4.8
14.9
10.1
14.7
23.2
29.6
8.5
14.9
6.4
* Predominant bottom character

-------
as a result of which a more narrow range of reaeration rate
coefficients, KQX at 25°C, may well be expected.

Flows were measured by Maryland DWR personnel during the
course of the tracer studies.  They exhibited substantial
short term fluctuations during the study period, and, for
that matter, during certain of the individual tracer release
periods.  For example, observed flows at Duvall Bridge (Sta-
tion 4)  varied from about 7 to 16 cfs during the study peri-
od, and individual flow observations ranged from 7 to 23 cfs
over the whole study section.  Occasional local rains during
tracer studies also brought about some variability of stream
flow.

The stream physical studies were performed by Maryland DWR
personnel in late 1969, after the tracer studies were com-
plete, and included discharge, elevation and cross-sectional
measurements at about 600-foot intervals.  As prevailing
stream flows were significantly higher than those during the
September tracer study period, hydraulic properties that
prevailed during the tracer studies were obtained by adjust-
ing the physical study data accordingly -

In addition to the flow measurements and stream physical
studies, personnel of the Maryland DWR were also primarily
responsible for the conduct of Dump E, the last tracer re-
lease.

Reaeration Coefficients.  All of the reaeration coefficients
obtained in the Patuxent River studies are summarized in
Table 13.  Duplicate values were obtained for the upper
study section  (above Station 4) and triplicate values below
Station 4.  Figure 32 is a graphical presentation of all of
the observed gas transfer results for the Patuxent, the re-
sults for Dumps C, A and E having been displaced successive-
ly to the right on the graph in order to avoid overlapping
or crossing of lines.  The values of Kkr shown in Figure 32
represent the prevailing river temperatures  (see Appendix
AIV for details).  The values of Kox presented in Table 13
have been derived from the basic K^/KOX ratio of 0.83 and
corrected to 25°C by use of the temperature correction co-
efficient 1.022/°C.

Reference to Table 13 and to Figure 32 again indicates good
reproducibility of results, although the data for Dump E ap-
pear to depart more than usual from the results for Dumps A
and C.  Reproducibility of the observed values of Kox at
25°C was excellent between Dumps A and C  (below Station 4),
and very good between Dumps B and D  (above Station 4).  Dump
E took place at a lower flow than Dumps A and C, as indicated
                          125

-------
                      Table 13

  Observed Reaeration Coefficients,  Patuxent River
Reach
1-2
1-3
1-4
2-3
2-4

A
0
0
0
0
0

B
.159
.122
.122
.095
.109
Kox per
C
0.
0.
0.
0.
0.
hour @ 25 °C
D E
167
145
149
131
143

Mean
0.163
0.134
0.136
0.113
0.126
3-4       -   0.123       -      0.155       -         0.139

4-5     0.141   -       0.160      -       0.135       0.145
4-6     0.129   -       0.129      -       0.108       0.122
4-7     0.150   -       0.151      -       0.114       0.138

5-6     0.118   -       0.101      -       0.084       0.101
5-7     0,156   -       0.145      -       0.101       0.134

6-7     0.234   -       0.227      -       0.136       0.199
                         126

-------
o

H
<

z
o
»—I
I-
<
UJ
0
Z
o
o
o
H
D.

CC
v:
                                                                     FIGURE 32
                                                                   PATUXENT RIVER
                                                               OBSERVED  KRYPTON TRANSFER
                                                                    (RIVER TEMPS,)
                 KEY

            5)- STATION 5
            0.0/5- KKR/HR
          G- OBSERVED RATIO
    0,10
                                     15        20         25

                                         TIME OF FLOW/ HOURS

-------
by the somewhat longer times of flow, which may account in
part for the lower values of Kox observed for Dump E.  As
indicated above, substantial fluctuations in river flow
also occurred during some of the studies.

Reference to Table 13 indicates a high degree of uniformity
of results for the whole section of the Patuxent that was
studied.  Neglecting Dump E for the moment, all of the in-
dividual values of' Kox at 25°C fall within the range of 0.10
to 0.23 per hour; except for the reach 6-7, the range of ob-
servation is only 0.10 to 0.17 per hour, and the range of
mean values is only 0.11 to 0.16 per hour.  This consistency
of results clearly shows the effect on Kox of the higher
degree of hydraulic uniformity referred to earlier.  Con-
sidering the flow variations that occurred during these
tracer studies, in the case of this section of the Patuxent
between Stations 1 and 6, no major error would be made by
selecting a single value of Kox for use in oxygen balance
computations in the same range of flow.

JACKSON RIVER

The Jackson River has its source in the mountains of Vir-
ginia and West Virginia, and flows generally east and south-
east through Covington and Clifton Forge, Va.  Below Clifton
Forge it joins the Cowpasture River and becomes the James,
which then flows past Lynchburg to the ocean at Norfolk, Va.
Below Covington the Jackson is relatively wide and shallow,
and is characterized by alternating riffles and natural
pools, with a considerable range of velocities and depths.

The major BOD load to the Jackson River is from a paper mill
in Covington owned by the West Virginia Pulp and Paper Com-
pany.  At times of severe drought flow, essentially all of
the flow of the Jackson is withdrawn, used at the paper mill,
and then returned to the channel, so that at such times the
downstream flow is entirely effluent from the paper mill
waste treatment plant.  Even after 80 to 90 percent BOD re-
moval in this activated sludge treatment plant, the down-
stream natural reaeration capacity is not sufficient to pre-
vent serious or complete depletion of the DO resources of
the river.  Thus, as a further measure to improve the situa-
tion, the West Virginia Pulp and Paper Company operates two
pairs of mechanical aerators in the river itself at loca-
tions below the mill, at times of low stream flow.

Figure 33 is a general map of the Jackson River study locale,
showing the river and such features of interest as stream
sampling stations, highways, towns, the mill location, and
the mechanical aerator locations.  Tracer studies of the
                          128

-------
reaeration capacity of the Jackson between Covington and
Clifton Forge, i.e., the critical reach below the paper
mill, were performed in July, 1966, by personnel of the
FWPCA.  Those studies constituted the first field demonstra-
tion of the gaseous tracer technique for independently eval-
uating stream reaeration capacity^.  During the period im-
mediately prior to the tracer studies, the Jackson was cross'
sectioned at intervals of 1,000 feet between Covington and
Clifton Forge by staff of the Civil Engineering Department
of the Virginia Military Institute.  Slopes and water sur-
face elevations were not obtained at that time.  In 1970,
through the support of the FWPCA, the VMI staff returned to
the Jackson River study section to obtain elevations, as by
that time the results of the research reported here had
clearly indicated that the changes in water surface eleva-
tion were important hydraulic data.  Rather than level the
entire length of the study section from beginning to end,
the VMI staff relied upon available bench marks and levelled
from those to the 1966 river sampling stations.  Some diffi-
culty was encountered in determining the exact locations of
some of the earlier stations, especially Station 10, and at
some locations the sampling station elevation was obtained
by interpolation from elevation measurements at intervening
locations.  The derived 1966 sampling station elevations
were forwarded to the Georgia Tech research staff by Thacks-
ton^4f and are presented in Table 14.  Immediately prior to
the beginning of the 1966 reaeration studies the secondary
treatment units at the Westvaco paper mill failed, and the
paper mill waste treatment plant operated essentially as a
primary plant for several days.  The mechanical aerators
located downstream in the river were idle at the time, be-
cause of the imminent tracer studies.  As a result, zero
DO's occurred in a reach several miles long immediately be-
low the paper mill.

Dosing Pattern.  Table 15 shows the pattern of tracer doses
used during the two-week 1966 reaeration study period.
Eight separate tracer releases were made at two-day inter-
vals, proceeding systematically downstream from Dunlap
Creek, and the releases were designed to provide coverage of
each of the subreaches at least twice.  The quantities of
tracers varied considerably from one release to another,
partly because of the inability of the tracer suppliers to
provide an accurate assay.  The krypton-85 dose therefore
ranged from 231 me to 458 me, and the tritium dose from
273 me to 1,108 me in the individual releases.  The average
dose used was 335 me of krypton-85 and 746 me of tritium.
The total amounts released in the eight studies was 2.68
curies of krypton-85 and 5.97 curies of tritium.
                          129

-------
                        Table 14
           Jackson River Station Elevations
       1966 Sampling        Water Surface Elevation
          Station            Above MSL  (1970), ft.
            0                      1,219.3
            1                      1,211.1
            2                     (1,201.8)*
            3                     (1,187.3)*
            4                      1,179.9
            5                     (1,171.1)*
            6                     (1,163.1)*
            7                     (1,156.9)*
            8                     (1,141.4)*
            9                     (1,132.3)*
           10                     (1,119.1)*
           11                     (1,109.0)*
           12                     (1,093.8)*
           13                      1,083.2
           14                     (1,074.0)*
           15                      1,063.9
      Interpolated value

Hydraulic ^ Properties.   Typical hydraulic properties of the
Jackson River study reaches are summarized in Table 16.  The
18-mile study section was divided into 15 subreaches of av-
erage length 1.2 miles, from Dunlap Creek in Covington al-
most to Clifton Forge.  Stream flow ranged from 90 cfs at
the upper end of the study section to about 130 cfs at the
lower end,  a moderate drought condition.  The cumulative
time of flow for the 18-mile study section was about 2.4
days.
As may be seen from Table 16, average subreach velocities
ranged from 0.27 to 0.67 ft/sec, and average subreach depths
from about 1.7 to 3.1 feet.  From Table 14, the total water
surface elevation change for the 18-mile study section was
about 155 feet.  Although the stream was characterized by a
                          130

-------
          Table 15



Dosing Pattern, Jackson River
Dump
Number 0 1
10 	 7/11 	 ,
350 	 ,
1088 	 ,
11 	 7/13 	 ,
458 	 ,
273 	 ,
,_. 12 	
U)
H
13 	
14 	 	
15 	
16 	
17 	

Jackson River Sampling Station Number
2 3 4 56 789 10 11



231 -231 me of krypton-85 released
554 -554 me of tritium released

	 7/15 	
( 450) 	
273 	 	 	
	 	 	 7/17 . . . 	 	
231 	
542 	
	 	 7/19 	
231 	
554 	
	 	 	 7/21 	 	 	
321 	
1088 	
. . . 	 	 	 	 	 7/23 	
318 	
_ 	 1042 	 	 .
	 	 	 7/25
321
1108

-------
U)
                                           Table  16
                                 Typical Hydraulic Properties
                                         Jackson  River
Dump
10
10
10
10
10
10
10
10
10
Reach
0-1
0-2
0-3
0-4
1-2
1-3
1-4
2-3
2-4
Flow
cf s
90
90
90
93
90
90
94
90
96
Time of
Travel
hrs
1.65
5.97
12.67
17.09
4.32
11.02
15.44
6.70
11.12
Length
ft
4,000
10,800
19,100
26,600
6 ,800
15,100
22,600
8,300
15,800
Velocity
ft/sec
0.673
0.503
0.419
0.432
0.437
0.381
0.407
0.344
0.395
Depth
ft
1.76
1.87
1.97
2.28
1.94
2.03
2.38
2.09
2.56
X-sect
Area
sq ft
134
179
215
215
206
236
231
262
243
WS Elev
Change
ft
8.2
17.5
32.0
39.4
9.3
23.8
31.2
14.5
21.9
     10      3-4      96       4.42       7,500      0.471       3.11      204        7.4
12
12
12
12
12
12
12
12
12
4-5
4-6
4-7
4-8
5-6
5-7
5-8
6-7
6-8
102
102
102
104
102
102
105
102
106
2.
7.
10.
15.
4.
7.
12.
3.
7.
92
37
75
33
45
83
41
38
96
6
11
15
23
5
9
17
3
11
,100
,800
,700
,500
,600
,600
,400
,900
,700
0
0
0
0
0
0
0
0
0
.580
.445
.406
.426
.350
.341
.389
.321
.408
1.67
1.73
1.80
1.95
2.36
2.19
2.22
1.99
2.18
176
229
251
244
292
299
269
318
260
8.8
16.8
23.0
38.5
8.0
14.2
29.7
6.2
21.7
     14      7-8     108       4.58       7,800      0.473      2.26      228       15.5

-------
                                      Table  16  (continued)
                                  Typical  Hydraulic Properties
                                          Jackson River
u>
Dump
14
14
14
15
15
15
15
15
15
15
16
16
16
16
16
16
16
17
17
17
17
Reach
7-9
7-10
7-11
8-9
8-10
8-11
8-12
9-10
9-11
9-12
10-11
10-12
10-13
10-14
11-12
11-13
11-14
11-15
12-13
12-14
12-15
Flow
cfs
108
110
110
108
111
112
112
113
113
115
113
116
117
117
118
122
124
125
128
128
129
Time of
Travel
hrs
9.06
11.78
14.56
4.48
7.20
9.98
15.03
2.72
5.50
10.55
2.78
7.83
10.48
14.10
5.05
7.70
11.32
15.05
2.65
6.27
10.00
Length
ft
12,200
17,600
22,700
4,400
9,800
14,900
22,600
5,500
10,500
18,300
5,000
12,800
18,400
23,900
7,800
13,300
18,800
25,000
5,600
11,000
17,200
Velocity
ft/sec
0.374
0.415
0.433
0.273
0.378
0.415
0.418
0.562
0.530
.482
.500
.454
.488
.471
.429
.480
.461
.461
.587
.487
.478
Depth
ft
2.46
2.42
2.31
2.83
2.54
2.33
2.37
2.30
2.13
2.25
1.95
2.24
2.20
2.17
2.42
2.30
2.23
2.08
2.13
2.10
1.94
X-sect
Area
sq ft
289
265
254
296
294
270
268
201
213
239
226
255
240
248
275
254
269
271
218
263
270
WS Elev
Change
ft
24.7
(37.8)
47.9
9.2
(22.3)
32.4
47.6
(13.1)
23.2
38.4
(10.1)
(25.3)
(35.9)
(45.1)
15.2
24.8
35.0
45.1
10.6
19.8
29.9
     17
13-14
128
3.62
5,500
422
2.06
303
9.2

-------
    Table 16 (continued)
Typical Hydraulic Properties
        Jackson River
Dump
17
17
Reach
13-15
14-15
Flow
cf s
129
130
Time of
Travel
hrs
7.35
3.73
Length
ft
11,600
6,100
Velocity
ft/sec
.438
.454
Depth
ft
1.85
1.66
X-sect
Area
sq ft
294
286
WS Elev
Change
ft
19.3
10.1

-------
considerable range of velocities and depths at the individu-
al 1000-ft cross-sections (depths from 0.6 to 5.3 feet), no
remarkable hydraulic features such as waterfalls, shoals or
violent rapids occurred.

Reaeration Coefficients.  All of the reaeration rate coeffi-
cients obtained during the 1966 studies of the Jackson River
are provided in Table 17.  As may be seen, 80 observations
of the reaeration rate coefficients were obtained.  The
range of these coefficients was from about 0.07/hr to 0.39/hr
at 25°C.  Subreach 9-10 was unusual compared to the rest of
the stream, in that it included a relatively long shallow
riffle section, and was characterized by an associated un-
usually high reaeration capacity.  However, the observed
mean value of KOX (0.36/hr at 25°C) did not compare with the
values observed in the Flint or South at features such as
falls, Panola Shoals or the violent rapids sections in those
streams.

Reference to Table 17 shows that the reproducibility of the
test results was very good.  As noted earlier, the degree of
pollution of the Jackson River during these studies was un-
usually high, due to failure of the secondary treatment units
at the paper mill.  As a result, it is probable that the
values of KQX reported in Table 17 are lower than would have
been found otherwise, especially in the reaches immediately
below the paper mill.  Specifically, the observed results
for Dumps 10 and 11 are probably quite a bit lower than us--
ual, perhaps by 20 to 30 percent, due to that unusual pollu-
tion situation.

Figures 34 and 35 are a graphical presentation of all of the
results for the Jackson River studies.  The numerical values
of Kj,r shown  on the graphs refer to the prevailing river
temperatures and flows at the time of the individual tracer
releases.  The values of Kox shown in Table 17 have been
obtained by use of the Kkr'Kox ratio of 0.83 and the tempera-
ture coefficient of 1.022/°C.

Referring to Figure 34 for the upper study reaches, the gen-
erally parallel nature of the lines is evident, and indicates
a considerable degree of hydraulic uniformity.  This becomes
more evident by reference to Table 17, after correction of
the data to a common temperature.  From Table 17, for the
entire upper study section  (Stations 0 to 8) the range of
observed values of Kox at 25°C was from 0.07 to 0.21 per
hour.  Of the 41 reported values of Kox» 35 fall within the
range 0.10 to 0.15 per hour, and the median result of 0.12
per hour for Kox at 25°C represents a reasonably good single
value for the entire upper study section.
                          135

-------
                                      Table  17
                   Observed Reaeration Coefficients,  Jackson  River
Reach

 0-1
 0-2
 0-3
 0-4

 1-2
 1-3
 1-4

 2-3
 2-4

 3-4

 4-5
 4-6
 4-7
 4-8

 5-6
 5-7
 5-8

 6-7
 6-8

 7-8       -         -       0.132     0.122     0.128      -                       0.127
 7-9                                         0.133      -
 7-10       -                                  0.181      -


0
0
0
0
0
0
0
0
0
0










10
.146
.119
.120
.108
.108
.115
.104
.120
.102
.076
_
-
-
-
_
-
-
_
-


0
0
0
0
0
0
0
0
0
0










11
.134
.139
.117
.104
.104
.114
.100
.094
.081
.064
_
—
-
-
_
-
-
_
-












0
0
0
0
0
0
0
0
0

12
_
-
-
-
_
-
—
_
-
-
.153
.116
.122
.125
.093
.111
.119
.134
.133
K











0
0
0
0
0
0
0
0
0
per hour @ 25 °C
ox
13 14 15
_ 	 	
_ _ _
_
- - -
— _ _
_
— — —
_ _ _
- - -
_
.208
.141
.141
.134
.097
.115
.118
.139
.128


16 17 Mean
- 0
- 0
- 0
- 0
0
0
- 0
- 0
- 0
0
0
0
0
0
- 0
0
0
0
- 0
.140
.129
.119
.106
.124
.115
.102
.107
.092
.070
.181
.129
.132
.130
.095
.113
.119
.137
.131

-------
                                Table 17 (continued)
                   Observed Reaeration Coefficients, Jackson River
Reach

 7-11

 8-9
 8-10
 8-11
 8-12

 9-10
 9-11
 9-12

10-11
10-12
10-13
10-14

11-12
11-13
11-14
11-15
12-13
12-14
12-15

13-14
13-15

14-15      _______      0.102
K per hour @ 25°C
ox
10 11 12 13 14 15
- - - 0.174
- - - - 0.138 0.137
0.217 0.238
- - - - 0.197 0.202
----- 0.181
0.343 0.385
- - - - 0.244 0.248
0.197
- - - - 0.146 0.120 0
0.136 0
------ 0
------ 0
0.144 0
0
------ 0
------
------ o
------ 0
— — — — — —
------ 0
______
16
-
-
.140
.148
.184
.169
.152
.198
.174
-
.283
.192
-
.125
-







0
0
0
0
0
0
0
0
0
17
-
—
_
-
-
—
.184
.195
.191
.168
.213
.196
.161
.181
.140

0
0
0
0
0

0
0


0
0
0

0
0

0

Mean
.137
.228
.200
.364
.246
—
.135
.142
-
—
.160
.197
.183
-
.248
.194
-
.153
-

-------
                                                                      CLIFTON
                                                                      FORGE
                   JACKSON  RIVER STUDY LOCALE
                                                                 .ROSS SECTION
                                                                  SCALE
                                                                  IN MILES
FIGURE  33

-------
        2,00
U)
o
»—«
1-
<


z:
o
I—I

<

H
2
UJ
     o
     u
O
t-
a.

oc
                                                                              FIGURE
                                                                           JACKSON RIVER

                                                                         UPPER STUDY REACHES

                                                                      OBSERVED KRYPTON TRANSFER

                                                                           (RIVER TEMPS,)
                                                                                   KEY

                                                                              2)- STATION 2
                                                                              o.i36   - KKR/HR
                                                                            O- OBSERVED RATIO
        0
        0,04
             0
                            10
15
    20         25        30

TIME OF FLOW, HOURS

-------
 o

 H
 <
UJ
CJ
z
O
O
O
(-
Q.

o:
                                                                      FIGURE 35
                                                                     JACKSON  RIVER
                                                                  LOWER  STUDY REACHES
                                                              OBSERVED  KRYPTON  TRANSFER
                                                                    (RIVER  TEMPS,)
                                                                               KEY
                                                                         12)- STATION 12
                                                                          0.173  - KKR/HR
                                                                        O - OBSERVED RATIO
                                      12         16         20
                                         TIME OF FLOW, HOURS

-------
Figure 35, which depicts the observed results  for  the  lower
study section  (Stations 8 to 15), shows  somewhat more  var-
iability from one subreach to another.   The  subreach 9-10
in particular is clearly different  from  the  others  covered
by Dumps 14 and 15, as noted above,  and  was  characterized by
higher values of KQX  (see Table  17).

CHATTAHOOCHEE RIVER

The Chattahoochee is one of the principal rivers of the
southeast.   It rises in the mountains of northeast Georgia
and flows south and west some 436 miles to the Florida line,
where it joins the Flint River to form the Apalachicola.
The Chattahoochee drains approximately 1,450 square miles of
mountain and piedmont country above Atlanta.  It is the
principal source of public water supply for the Atlanta met-
ropolitan area, where such withdrawals totalled 130 to 140
mgd in 1966.

About 45 miles upstream from Atlanta, the Chattahoochee is
impounded in Lake Sidney Lanier by Buford Dam, constructed
in the 1950's by the Corps of Engineers.  Lake Lanier is a
multiple purpose reservoir, important for flood control,
hydroelectric power and low flow augmentation, as well as
recreation.  The river flow from Buford Dam is reregulated
at Morgan Falls Dam, about 10 miles upstream from Atlanta.
Morgan Falls Dam is owned and operated by the Georgia Power
Company, and a minimum release of 750 cfs is maintained at
all times.

The Chattahoochee falls rather sharply (as much as five feet
per mile) from Morgan Falls to a location near the beginning
of the reaeration tracer study section.  From there, the
fall is more gradual for some 120 miles until the stream
reaches the Fall Line, where it again drops rather abruptly
before flattening out on its way to the Gulf of Mexico.

Figure 36 is a general map showing most of the Chattahoochee
River study locale.  The main reaeration tracer study sec-
tion, about 18.5 miles in length, extended from Station 0 at
Georgia Highway 280 to Station 6 (not shown), located almost
two miles downstream from the Georgia Highway 92 bridge  (Sta-
tion 5).  The main source of pollution,  the Clayton STP, is
located almost a mile above Station 0:  during the reaeration
tracer study period, this plant operated as a primary treat-
ment plant, with frequent bypassing of raw sewage.  In addi-
tion, two steam generating plants,  (Plants Atkinson and Mc-
Donough), located between the Clayton STP and Station 0, use
Chattahoochee River water for cooling, and thus raise the
temperature of the downstream Chattahoochee.  Additional
                          141

-------
*>•
to
               SWEETWATER
                       CR.
                GA I,
                               NICKAJACK CR.'
                                           ATLANTA
                                           CITY
                                           LIMITS
                                                                   PEACHTREE CR,
   KEY
       C SAMPLING STATION
       ^
DIRECTION OF FLOW
                                                            SCALE IN MILES
                                                          012345
                                  FIGURE 36
                       CHATTAHOOCHEE  RIVER LOCALE

-------
pollution entered the main Chattahoochee via the several
tributaries (Nickajack Creek, Sandy Creek, Utoy Creek, Sweet-
water Creek, Camp Creek). During the work week, as a result,
the Chattahoochee was heavily polluted, and at locations in
the upper portion of the reaeration tracer study section the
predominantly mud and sand bottom was the residence of tre-
mendous populations of redworms.

With the exception of Dump XXV, all of the reaeration tracer
studies were performed on Sundays, at quite low, steady flow.
The Chattahoochee is used for power production during the
work week, and the flow in the vicinity of Atlanta therefore
undergoes a sharp fluctuation during each day.  With the co-
operation of the Georgia Power Company and the Corps of Engi-
neers, it was possible to schedule steady critical low flows
on Sundays for reaeration study purposes, without undue in-
terference with power production.  However, as will be seen
in this and subsequent sections of this report, this neces-
sary Sunday tracer study schedule also meant quite low river
BOD's, due to characteristically low weekend pollution loads.
Also, companion studies of the effect of pollution on the
reaeration capacity of the Chattahoochee demonstrated a
sharply different condition during the usual work week.
Hence, the reaeration capacities observed during these tracer
studies represent a pollution condition more of the future
than the present, when pollution of the Chattahoochee in the
vicinity of Atlanta will be considerably reduced.

Tracer Dump XXV, the last study, was conducted at a substan-
tially higher flow than the earlier studies, in order to ob-
tain some information at other flows.  As in the earlier
studies, a steady flow for this tracer study was arranged
with the cooperation of the Georgia Power Company.

Dosing Pattern.  Table 18 shows the dosing pattern that was
used in"th~e tracer studies of the reaeration capacity of the
Chattahoochee River.  Dose XVIII  (not shown, and not reported
further) was similar to Dose XVII and followed it by three
weeks; difficulties associated with the dose assay, together
with unusual and ineffective location of river sampling sta-
tions, resulted in observed reaeration capacities that were
both questionnable and not comparable to the results obtained
at regular sampling points from the other tracer releases.
Accordingly, the results from Dose XVIII are not reported
here.

As indicated in Table 18, a total of nine separate tracer
studies was conducted, beginning in the fall of 1969 and ex-
tending to the fall of 1970.  The Chattahoochee River studies
were begun in the fall of 1969, immediately after completion
of the Flint and South River studies, with the intention of
                           143

-------
                        Table 18

           Dosing Pattern, Chattahoochee River


DumpChattahoochee River Sampling Station Number_
Number   0        2        T~       3        5        6~
XVI....10/12/69
       2,600	
       4,000.. ,
XVII	10/19/69
                1,330	
                4,000	
XIX	7/26/70,
        4,480.. .
        5,000. . ,
XX	8/2/70,
        5,250.,
        5,000.
XXI	8/16/70
                        2,220..
                        5,000. .
XXII	 .8/30/70
                        3,610..
                        5,000. .
XXIII.	9/13/70,
                3,980- . .
                5,000,.
XXIV	9/20/70
                        2,410..
                        5,000. .
XXV	11/24/70,
       5,930	
      10,000	
Key:  7/26/70-Dump XIX released on 7/26/70 - Station  0
      4,480  - 4,480 me of krypton-85 released
      5,000  - 5,000 me of tritium released.
                          144

-------
completing most if not all of the field research by late
1969 or early 1970.  However, it proved to be impractical to
continue the Chattahoochee studies into late November and
December because of considerably colder water temperatures
and frequent rainfall.  Thus, following Dose XVII, the re-
mainder of the field operations were rescheduled for the
summer of 1970.

As with the studies in the Flint and South Rivers, the quan-
tities of tracer varied from one release to another, largely
due to the inability of the supplier to deliver the quantity
of dissolved krypton-85 ordered.  Thus, the amount of kryp-
ton-85 actually released ranged from as little as 1,330 me
to as much as 5,250 me in the series of doses at critical
low flow, whereas the tritium dose was as ordered and ranged
from 4,000 to 5,000 me.  For Dose XXV, at higher flow, the
krypton-85 dose was 5,930 me and the tritium dose 10,000 me,
as requested from the supplier.  A total of 31.8 curies of
krypton-85 were released in the nine studies, and a total
of 48 curies of tritium.  The average dose at critical low
flow  (Doses XVI - XXIV) was 3,240 me of dissolved krypton-
85 and 4 /750 me of tritium.

Hydraulic^ Properties.  The Chattahoochee River is character-
i zed "by a fTigh degree of hydraulic uniformity in the reaera-
tion study section.  In contrast to the Flint and South, the
Chattahoochee below the Georgia Highway 280 bridge has long
straight sections of relatively uniform width and depth,
often with high, steep banks that prevent ready access.
Table 19 shows typical hydraulic properties of the Chatta-
hoochee associated with the flows prevailing during the re-
aeration tracer studies.

Referring to Table 19, at the critical low river flow of
about 1,100 cfs in the tracer study section, the stream
depth, velocity and cross-sectional area were each quite
uniform:  the mean depth of about 4.0 feet did not vary ap-
preciably from reach to reach, or, for that matter, within
any reach; the mean velocity averaged about 1.8 ft/sec, and
the mean cross-sectional area about 600 sq ft.  At the
higher flow associated with Dump XXV  (3,300 cfs), the mean
velocity was higher, about 2.5 ft/sec, the mean depth in-
creased to about 7.6 ft, and the cross-sectional area about
doubled.  No detailed cross-sectional measurements were made
below the Georgia Highway 92 bridge  (Station 5), hence typi-
cal depths and cross-sectional areas are not shown in Table
19 for the reach 5-6.  The bottom character of the study sec
tion was predominantly mud and sand, with occasional short
more rocky areas.
                          145

-------
                                    Table 19
                          Typical Hydraulic Properties
                               Chattahoochee River
Dump
XIX

XXIII


XXIV


XXIV

XXIV
XXV

Reach
0-2
0-3
2-3
2-4
2-5
3-4
3-5
3-6
4-5
4-6
5-6
0-4
0-5
Flow
cf s
1,076
1,076
1,030
1,130
1,130
1,180
1,180
1,180
1,180
1,180
1,180
3,300
3,300
Time of
Travel
hrs .
4.93
8.73
3.27
5.39
9.45
1.70
5.19
7.34
3.49
5.64
2.15
7.53
10.08
Length
ft
31,500
54,000
22,500
35,000
57,500
12,500
35,000
44,000
22,500
31,500
(9,000)*
66,500
89,000
Velocity
ft/sec
1.77
1.72
1.91
1.80
1.69
2.04
1.87
1.67
1.79
1.55
(1.16)*
2.45
2.45
Depth
ft
4.16
4.16
4.02
3.84
3.89
3.62
3.90
—
4.06
—
-
7.65
7.66
X-sect
Area
sq ft
608
625
540
627
668
579
630
—
660
—
-
1,350
1,350
WS Elev
Change
ft
6.7
10.5
3.8
6.6
9.5
3.8
5.6
6.8
2.9
4.2
1.2
13.3
17.8
*Estimated value.

-------
Because of the high steep banks and the general inaccessi-
bility of the Chattahoochee, it was not feasible for the
levelling party to establish a continuous series of refer-
ence elevations directly along the banks of the river.  In-
stead, it was necessary to level along more circuitous
routes, tie into reference elevations and bench marks estab-
lished earlier by others, and thereby establish tapedown lo-
cation elevations for each river sampling point.  Thus, it
was not possible to conduct a single, continuous elevation
survey from Station 0 to Station 6, checking against estab-
lished bench marks along the way, as was done in the Flint
and South River studies.  The procedure used for the Chatta-
hoochee elevation study thus depended, for accuracy, more
heavily than usual on the accuracy of the elevations re-
ported for established bench marks.  The Georgia Tech re-
search party elevation surveys were tied into or checked
against bench marks established by or for the U. S. Army
Corps of Engineers, the Georgia Highway Department, the
U. S. Geological Survey, Six Flags over Georgia, the Georgia
Power Company and the Atlanta Water Works.  Unfortunately,
the results indicate clearly that all of the available bench
marks are not equally valid, and the discrepancies were dis-
covered too late to permit the additional surveys necessary
to resolve conflicts among established bench mark elevations.

Figure 37 shows the water surface elevation profile that has
been developed for the reaeration tracer study section of
the Chattahoochee River, for the steady critical low flow
associated with Dumps XVI through XXIV.  The elevations for
the individual sampling stations were finally established
by a procedure of cross-checking relevant level notes, tape-
downs made during the tracer studies and all available bench
mark elevations.  The resulting water surface profile is
thus believed accurate.  However, some doubt must remain, as
it proved impossible to resolve all bench mark elevation
conflicts.  It is to be hoped that in the future some au-
thority such as the City of Atlanta will undertake a thor-
ough survey designed to firmly establish the elevations of
existing and new bench marks, and to resolve present con-
flicts .

Reaeration Coefficients.  Table 20 provides a summary of all
of~the~reaeration coefficients observed in the Chattahoochee
River studies of 1969-70.  The total of nine separate re-
leases, as detailed in Table 18, yielded 37 observed values
of the reaeration rate coefficient.  Of these, 31 observed
values  (Dumps XVI through XXIV) are associated with the
critical low flow of about 1,100 cfs in the tracer study
reach, and 6 observations of Kox  (Dump XXV) reflect a flow
of about 3,300 cfs.  Of the 31 observed values at low flow,
                           147

-------
00
            750
         LU
         >
         LU
         C/5
         <
         LU
         LU
         >
         o
         PQ
            740
           730
           720
         uj 710
         LU
           700
                      STA. 0
                                   FIGURE  37

                             WATER SURFACE PROFILE

                              CHATTAHOOCHEE RIVER

                               DOSES XVI - XXIV
           •GA. HWY.  92

           U.S.G.S.  GAGE  REF
                         0
20        40         60        80
    DISTANCE  IN  THSDS,  OF FEET
100
120

-------
                                    Table 20
              Observed Reaeration Coefficients, Chattahoochee River

Reach                             Kox per hour @ 25°C
XVI XVII
0.077
XIX
0.100
0.077
XX XXI
0.036
0.036
XXII
-
XXIII
-
XXIV
-
XXV MEAN
0.030 0.061
0.052
0.040
0.042
0-2
0-3
0-4
0-5

2-3       -   0.070   0.049   0.012    -       -      0.040       -         -    0.043
2-4       -   0.042     -                             0.057       -       0.049  0.049
2-5       -                     -              -      0.045       -       0.047  0.046

3-4       - (-0.005)    -       -    (0.081)  (0.002)   (0.084)   (0.049)       -    0.043
3-5       -                     -    0.044   0.026    0.048    0.040        -    0.040
3-6       -                     -            0.019       -     0.038        -    0.029

4-5       -                     -    0.026   0.039    0.029    0.035      0.045  0.035
4-6       -                     -            0.025       -     0.034        -    0.030

5-6       -                     -            (0.005)      -     (0.033)       -   (0.019)
*Parenthetical values questionnable.

-------
four were obtained in the 1969 study period and the remain-
der during the summer and fall of 1970.  The prevailing
river water temperatures ranged from 20°C to 28°C for Dumps
XVI through XXIV, whereas the prevailing river water temper-
ature for Dump XXV, at 3,300 cfs, was 10°C«  The detailed
data for each tracer release are provided in Appendix IV.

As noted earlier, all of the tracer releases associated with
the critical low flow of 1,100 cfs occurred on Sundays, and
pollution loads were quite low.  The significance of this
situation and of other related studies of pollutant effects
on the reaeration capacity will be discussed further below.

Referring to Table 20, the reproducibility of observed re-
sults was quite good in some reaches (e.g., reaches 2-4,
2-5, 3-5, 4-5) and poorer in others (e.g., reach 3-4, 5-6).
The individual observed results for the reaches 3-4 and 5-6
are all contained within parentheses in Table 20, to indi-
cate that they are considered more questionnable than the
other results.  Those reaches were relatively short com-
pared to the others, both in distance and time of flow, and
therefore generally characterized by less gas loss and some-
what larger relative error.  In the case of the reach 3-4,
a sufficient number of observations of Kox was made  (n=5)
to yield a good mean value for that reach; in the case of
reach 5-6, only two observations of Kox were made, one of
which (Kox = 0.005 per hour) was undoubtedly poor.  Hence,
the mean value of Kox for the reach 5-6 is also indicated
as questionnable in Table 20, by the use of parentheses.

That the reproducibility of individual observed results is
generally somewhat poorer for the Chattahoochee River stud-
ies than for others is not surprising in terms of the test
conditions.  Although the tracer doses were the largest used
in any of the studies reported here, the river flow was also
much larger, and the tracer doses were actually the smallest
used on a mc/cfs basis.  In addition, average velocities
(and dispersion) were considerably greater in the Chattahoo-
chee than in the other streams, and reaeration capacity  (and
actual tracer gas losses) considerably smaller.  The net re-
sult of this combination of factors was a somewhat lower
degree of reproducibility of results.

To compensate, the Chattahoochee studies were designed so as
to provide more than the usual two observations of Kox per
basic study reach, so that good estimates of the mean value
of Kox could be derived for each basic reach.  Specifically,
for th- reaches 0-2 and 2-3, four separate values of Kox
were obtained; for the reaches 3-4 and 4-5, five separate
values of K0x were obtained in each case.  As may be seen
from Table 20, this degree of repetition of field testing
                          150

-------
provided a highly consistent set of mean values of Kox for
the whole study section 0-5.

Referring to Table 20, for the basic study reaches 0-2, 2-3,
3-4 and 4-5, the mean value of K_x at 25°C ranged from
0.035 to 0.061 per hour, and tended to decrease proceeding
downstream.  The mean of all 37 observed values of Kox was
0.042 per hour at 25°C; neglecting those reaches that in-
clude station 6 (the reach 5-6 was covered only twice), the
mean of the 31 values of Kox that were observed for the
study section 0-5 was 0.045 per hour at 25°C.

It should be noted also that the observed values of Kox for
Dump XXV (at 3,300 cfs) were completely consistent among
themselves and also were fully consistent with the results
obtained for the other tracer studies conducted at a low
flow of 1,100 cfs.  Thus, an increase in river flow by 300
percent, and an associated near-doubling of stream depth,
made no difference in the prevailing value of Kox.  This
phenomenon was also observed clearly in connection with the
Flint River studies discussed earlier (Dump III vs. Dump
XIV), as well as in connection with the fluctuating flows in
the Patuxent and South River studies.  In brief, in all
such cases observed, changes in river flow by a factor of
two or three did not result in any observable significant
change in the reaeration coefficient, Kox, which remained
constant within relatively narrow limits.

The mean value of Kox of 0.045 per hour at 25°C  (kox, to the
base 10 = 0.47 per day at 25°C) thus constitutes a highly
accurate basic reaeration coefficient for the Chattahoochee
River immediately below the Clayton STP outfall and Plants
Atkinson and McDonough of the Georgia Power Company, for a
distance of about 17 miles  (main study reach 0-5), for
associated steady flows of 1,100 to 3,300 cfs within that
section of the river.  It is emphasized, however, that no
studies were performed during periods of the rapidly fluc-
tuating flows associated with power peaking.  Also, river
BOD's were generally quite low, representative of future
rather than present weekday pollution levels.

Figure 38 is the usual semilog plot of results for the
Chattahoochee River tracer studies.  The observed krypton:
tritium ratios have been plotted vs. time of flow for each
tracer release.  These data reflect river water temperatures
at the time of each study, as provided in Appendix IV, and
have not been corrected to a common temperature.  The asso-
ciated values of K^r at prevailing river temperatures are
shown.  As noted above, the individual results for the short
reaches 3-4 and 5-6 are regarded as more questionnable, and
hence those lines are shown as dashed rather than full lines.
                           151

-------
                                                                           FIGURE  38
                                                                      CHATTAHOOCHEE RIVER
                                                                       OBSERVED  KRYPTON
                                                                           TRANSFER
                                                                        (RIVER TEMPS.)
0.20
                                                    10         12
                                      TIME OF FLOW IN HOURS
                                    152

-------
In connection with Dump XXV, it should be recalled that the
prevailing water temperature was only 10°C, as compared to
temperatures ranging from 20°C to 28°C for the other
studies; also, XXV was at a considerably higher river flow,
as evidenced by the shorter times of flow.

The generally parallel trend of results is evident in Fig-
ure 38, allowing for the individual departures noted above.
Converted to values of Kox at a common temperature, as pro-
vided in Table 20, the results are quite consistent.

BOD Results.  BOD time series were run as a matter of rou-
tine for all of the Chattahoochee River reaeration tracer
studies.  In each case, large BOD samples were collected at
each tracer sampling station and a BOD time series developed
for each such sample.  Samples were also taken for such ana-
lysis at the Atlanta Water Works, well above the Clayton STP
outfall.  The time series results are not reproduced here,
as that degree of detail is not relevant.  However, the 5-
day BOD's are indicative of the general level of pollution
in the study reach at various times, and are discussed
briefly below.

The BOD's associated with the 1969 Sunday tracer studies
were relatively high compared to results observed during the
1970 studies.  Specifically, the 5-day, 20°C Sunday BOD
ranged from 8.0 to 11.4 mg/1 in the reach 0-2 during Dump
XVI, and from 6.7 to 10.7 mg/1 in the reach 0-3 during Dump
XVII.  The BOD was still somewhat high for a Sunday during
Dump XIX, the first of the 1970 series, ranging from 4.5
to 9.2 mg/1 in the reach 0-3.  However, beginning with Dump
XX, on 8/2/70, the 5-day, 20°C, Sunday BOD dropped to new
low levels and remained low.  For example, in the reach 0-3
the observed BOD's ranged from 3.2 to 4.2 mg/1 for Dump XX.
In the succeeding tracer studies, with four stations rep-
resented in each case, the maximum 5-day BOD's observed
were 4.0, 3.4, 4.5, 3.4 and 3.4 mg/1, respectively, for
Dumps XXI through XXV.  The reduction observed in 1970 co-
incided well with improved waste control measures reported
by the Georgia Water Quality Board.

The Sunday BOD's observed and noted above are not indicative
of usual weekday pollution levels, but are quite low in com-
parison.  In addition, as noted earlier, the steady low
flows arranged for the Sunday tracer studies, are in no way
representative of the rapidly changing weekday flow situa-
tion in the Chattahoochee below Atlanta.  Under these cir-
cumstances, detailed oxygen balance studies for the tracer
study section, using the observed reaeration coefficients,
have not been deemed profitable so far as the purposes of
                           153

-------
this research are concerned, and have not been performed.

Pollutant Effects.   Although the research studies of the
effects "of" pollution on stream reaeration capacity have gen-
erally been deferred for separate discussion in Section VII,
some comment regarding the Chattahoochee studies is desir-
able here.  In brief,  as shown in Section VII, a number of
Chattahoochee River water samples were tested for pollutant
effect.  Sunday samples taken during the course of Dumps
XIX through XXIII showed no pollutant effect—that is, the
Kkr(or, the Kox)  was not reduced because of the presence of
pollution added to  the river below the Atlanta Water Works.
In contrast, weekday samples showed a definite reduction due
to such pollution.   Specifically, those tests showed that
during the week Kkr (and Kox)  might be expected to be only
perhaps 80 percent  of  the magnitude that would occur in the
unpolluted condition.   Hence,  the 25°C Kox of 0.045 per hour
observed for the  reach 0-5 during these studies should be
reduced to about  0.036 per hour if it is to be used for oxy-
gen balance studies under weekday conditions at similar
flows.
                          154

-------
                       Section VII

        EXPERIMENTAL RESULTS - LABORATORY STUDIES
This section of the report includes a summary of the ob-
served results of a series of laboratory investigations
dealing with the effects of pollutants on the reaeration co-
efficient, K  .  The main purpose of these studies was to
develop and demonstrate a laboratory test procedure for e-
valuating the effects of a variety of pollutants on the gas
transfer or reaeration capacity.  Although it is suspected
that certain pollutants may also affect the dissolved oxy-
gen saturation limit, Cs, these studies did not include an
investigation of that possibility.

Various investigators have reported that detergent surface
active agents (surfactants) reduce the rate of gas trans-
fer, the reduction being dependent on the concentration and
the hydrodynamic conditions of turbulence and mixing.  In or-
der to study the effect of detergent surfactants on gas
transfer, a series of tests was conducted using various con-
centrations of the detergent surfactant linear alkylate sul-
fonate (LAS) in distilled water.

At the time of the laboratory studies, there was also con-
siderable public attention being devoted to the detergent
"builder" nitrilotriacetic acid (NTA) as a substitute for
high phosphate detergents.  Due to this interest, laboratory
tests were conducted for the purpose of evaluating the pos-
sible effects of pure NTA on reaeration in turbulent water.

The effect of oil on gas transfer in turbulent water was
considered to be of sufficient interest to warrant a sepa-
rate series of tests with various concentrations of reagent
grade mineral oil in distilled water.

The field reaeration measurements reported in Section VI of
this report were often obtained in highly polluted stream
reaches.  As the polluted river waters contain numerous con-
taminants, some of which may effect the reaeration  capacity,
laboratory tests were conducted on actual river water sam-
ples for the purpose of evaluating the effects of mixed pol-
lutants on the reaeration of natural waters.

The experimental results reported in the following portions
of this Section of the report are given in terms of the
krypton transfer coefficient, Kkr, for convenience and con-
sistency, as the principal focus of interest was on the na-
ture and extent of effect rather than upon a particular gas.
The reported results can be converted to values of Kox, if
                          155

-------
that is desired, by use of the 0.83 conversion.  All of  the
tests reported here were conducted at a constant temperature
of 20°C.

POLLUTANT EFFECTS - PURE SUBSTANCES

All of the tests of pollutant effects were conducted in  the
laboratory reactor system described in Section V of this
report  (see Figure 19) .  In brief, one test of the effect of
a pollutant consisted of two reactor runs:  the first was
conducted on clean water and provided a value of Kkr refer-
red to here as Kciean as a baseline evaluation of the hydro-
dynamic conditions; the second run was conducted under the
same hydrodynamic conditions as the first, with the pollu-
tant added, providing a comparable value of Kkr referred to
here as Kponute(j.  In keeping with the current practice of
referring to pollutant effects in terms of an "alpha" fac-
tor, the value of KpOjj_u-t-ed divided by Kciean is denoted
here as "alpha".  For further details of the experiment, the
reader is referred to Section V.

LAS Tests.  Figure 39 is a semilog plot of the krypton :tri-
tiurrT concentration ratios vs. elapsed time for two typical
reactor tests with LAS.  The two distilled water runs (no
LAS present), ISA and 19A, had Kkr values of 2.07/hr and
2.04/hr, respectively.  The test water for run 13B con-
tained 1.0 mg/1 of LAS and the water for run 19B contained
8.3 mg/1 of LAS.  The observed values of Kkr for runs 13B
and 19B were 1.91/hr and 1.40/hr, respectively.  The re-
sulting "alpha" factor for test 13 was
                              ,
                           2.07    '

and for test 19

                       cc = ilM = 0
                           2.04   U
-------
o
I—I

I-
<
<
cc.
g  0.8



I  0.6
0
ce

l-


2^

O
h-
D.
   0,2
   0,1

 0,08


 0,06
K
^
\ \
Xs
X
RUN 1
LAS =
\ =
Kl



X
XV
Xs
X
RUN 19A
LAS = 0
Kkr = 2







TEST 13



= = 1.91 = 0.92
K 2.0:
\\
\
3A — / N
0 MG/L
2.07/hr.




TEST 19

n
Xy-J
X P
\Xi
X
X
X
>

" - i'nS - 0-60
2.04
^D
\>
^\
MG/L N
.04/hr.






\
X
X,
^
x
x
o

FIGURE 39
20°C
EFFECT OF LAS ON Kkr
TYPICAL RESULTS
(LOG SCALE SHIFTED
FOR CLARITY)

*UN 13B
,AS = 0.96
Ckr = !'91/
n
X
\
IX
^ \
°v
\
\


MG/L
'hr.






v
\










/ 	 RUN 19 B
"*n // LAS = 8.3 MG/L
^C K! = 1.40/hr.
X^^ kr
DV
X,


x
\










      0
20         40        60         80

      ELAPSED TIME  IN  MINUTES
100
                              157

-------
       Table 21




Reactor Tests with LAS




   (Constant Mixing)
Test
Number
8

9

10

11

12

13

14

15

16

17

18

19


Run
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
LAS Cone.
mg/1
0.0
3.0
0.0
3.8
0.0
4.8
0.0
6.0
0.0
11.0
0.0
1,0
0.0
2.0
0.0
5.2
0.0
7.0
0.0
9.2
0.0
9.5
0.0
8.3
Kkr/
Water
1.79

1.95

1.81

1.90

1.90

2.07

2.13

2.03

2.13

1.90

2.02

2.04

hr (20°C)
Water 4- LAS

1.43

1.44

1.37

1.29

1.24

1.91

1.83

1.58

1.56

1.43

1.41

1.40

"alpha"

0.80

0.74

0.76

0.68

0.65 .

0.92

0.86

0.78

0.73

0.75

0.70

0.69
         158

-------
the data presented in Table 21, the Run A values of Kkr for
the 12 tests ranged from 1.79/hr to 2.13/hr, and the average
value was 1.97/hr.  All 12 of the Run A values were within
10 percent of the mean for these tests.

Figure 40 is a plot of the results obtained at the essen-
tially constant mixing speed, the observed values of Kkr for
the LAS + Water runs  (Run B data) being plotted against LAS
concentration.  As may be seen, these results appear to plot
as two separate groups.  The main group, for tests 12
through 19, describe a good straight line over the range of
observation, the value of Kkr diminishing as the LAS concen-
tration is increased.  The first four tests, 8 through 11,
plot as a separate group, the results being considerably
lower but adequately fitted with a parallel line (shown as
a dashed line in Figure 40).  The reason for this discrep-
ancy cannot be stated with certainty.  Undoubtedly, between
tests 11 and 12 some change in experimental procedure oc-
curred, possibly an improved procedure for cleaning the
reactor between runs or an improvement in the analytical
method for LAS.  All of the test runs yielded very good
straight line fits of the krypton:tritium concentration
ratios on semilog paper.  Whatever the cause of the discrep-
ancy in runs 8B through HB the results of the entire series
of tests with LAS are clear:  as the concentration of LAS
increases, there is increased effect on the gas transfer co-
efficient, and the relationship involved appears to be lin-
ear for the range of observation indicated in Figure 40.
The line through the main group of data, Runs 12B through
19B, passes through Kkr = 1.97/hr at a concentration of zero
LAS, and that is the experimentally observed value for dis-
tilled water alone, hence the other lower group of data must
be regarded as erroneous in the sense noted above.

Figure 41 is a plot of the data of Table 21 according to the
usual practice of referring to such effects in terms of the
"alpha" factor.  In this case, the "alpha" factors provided
in Table 21 have been plotted against LAS concentration.  As
may be seen, the"alpha" factor becomes smaller as LAS concen-
tration increases, indicating greater reduction of Kkr.  The
smooth curve has been fitted to the data of Figure 41 by eye,
and the data appear to extrapolate back to an "alpha" of
about 1.0 at zero LAS concentration.

Table 22 summarized the results of the six reactor tests
(12 runs)  that were conducted with distilled water in which
the LAS concentration was kept essentially constant while
the rate of mixing was varied.  The LAS concentrations
ranged from 9.8 to 11.8 mg/1 of LAS, with an average of
10.4 mg/1.  The value of Kkr for distilled water only  (Runs
designated A)  is indicative of the actual degree of mixing
or surface water replacement, and varied from 0.078/hr to
                           159

-------
  2,20
  2.00
   1.80
QL
=>
O
  1.60
  1.20
  1.00
            DISTILLED WATER
                TESTS  8-11
        FIGURE 40

          20°C

  EFFECT OF  LAS OH Kkr

    (CONSTANT MIXING)
fESTS  12-19
                                          O

                           4         6         8

                          LAS  CONCENTRATION,  MG/L
                10
12
                            160

-------
1,10
1,00
0,90
          •DISTILLED WATER
      \
       \
V
X
                                                FIGURE 41
                                                   20°C
                                              EFFECT OF LAS
                                               ON "ALPHA"
                                                    (CONSTANT MIXING)
0,80
0,70
0,60
                                   o
     0
                           6         8         10
                    LAS CONCENTRATION/ MG/L
12

-------
4.67/hr.
Test
Number
21

22

23

24

25

26


Run
A
B
A
B
A
B
A
B
A
B
A
B
LAS Cone.
mg/1
0.0
9.8
0.0
10.0
0.0
11.8
0.0
10.3
0.0
10.7
0.0
10.0
                        Table 22

                 Reactor Tests with LAS

              (Constant LAS Concentration)

                                Kkr/hr (20°C)
Water
4

0

0

3

0

0

.67

.439

.078

.58

.180

.895

Water + LAS

2

0

0

2

0

0

.68

.417

.075

.11

.171

.640
                                                    "alpha"


                                                     0.57


                                                     0.95


                                                     0.96


                                                     0.59


                                                     0.95


                                                     0.72
The results of the tests with constant concentration of LAS
are clear from the data provided in Table 22.  If the data
are arrayed in order of increasing value of K^j- for dis-
tilled water, it will be seen that the "alpha" factor de-
creases accordingly.  Thus,  at a constant concentration of
LAS, the greater the degree  of turbulent mixing, the
greater the corresponding reduction of gas transfer, or the
less the gas transfer coefficient.  As shown in Section VI,
in streams like the Flint and the South Rivers, much of the
action of reaeration takes place in short distances  asso-
ciated with rapids, shoals and  waterfalls.  It is at just
such hydraulic features that the damaging effect of deter-
gents will be at a maximum,  in terms of reducing the magni-
tude of K  .
         ox

The reactor tests with LAS have thus provided important in-
sight into the nature and extent of effects on reaeration
capacity of surfactants like LAS and the household deter-
gents that contain such surfactants.  In brief, not only
does LAS or like material reduce the gas transfer and re-
aeration capacity, but this damaging effect of detergents on
stream reaeration capacity is twofold - the resulting reduc-
tion in Kkr or KQX increases as the detergent concentration
                          162

-------
increases,  and for any one concentration this reduction is
at its greatest at hydraulic features such as rapids, shoals
and falls,  where reaeration might otherwise be expected to
be at a most beneficial maximum.  For example, four samples
collected from the South River during the field tracer study
period contained from 1.8 to 2.7 mg/1 of LAS - according to
the results provided in Table 21 and Figure 41, a reduction
of KQX to perhaps 85 percent of the clean water value would
be expected over Panola Shoals, where a great deal of reaer-
ation takes place.

NTA Tests.   Four separate tests of distilled water with NTA
added were conducted, for the purpose of obtaining an ini-
tial evaluation of its effect on gas transfer and reaera-
tion.  The NTA concentrations in these tests ranged from
7.5 to 16.1 mg/1.  The results are tabulated in Table 23.

                        Table 23

                 Reactor Tests with NTA

                                Kkr/hr (20°C)
Test
Number
27

28


Run
A
B
A
B
NTA Cone.
mg/1
0.0
7.8
0.0
16.1
                              Water   Water + NTA   "alpha1
                               0.462
 29


 30
A
B

A
B
0.0
7.5

0.0
7.8
                               0.499
0.194
0.477
                                        0.438
                                        0.475
         0.182
                                        0.477
0.95


0.95


0.94


1.11
The results provided in Table 23 indicate clearly that pure
NTA had no significant effect on the gas transfer capacity
in the range of concentrations and mixing rates tested.
The average "alpha" result was 0.99.

Mineral Oil Tests.  A total of thirteen separate tests were
conducted with mineral oil in distilled water, the mineral
oil concentrations ranging from 16 to 458 mg/1.  Eleven of
these tests were conducted at essentially the same mixing
rate with variable concentration of oil, while four of the
tests were conducted at one concentration but variable mix-
ing rate.  Although such tests with laboratory grade mineral
oil and distilled water cannot be extrapolated directly to
field situations, the tests have provided very interesting
insight into the nature and effects of mineral oil  (and pre-
sumably other oils) on gas transfer in turbulent water sys-
tems.
                          163

-------
Table 24 provides the results obtained from all thirteen
tests with mineral oil.   The first eleven tests, Test 31
through 41, were at essentially one mixing rate, as indi-
cated by the observed values of Kk£ in distilled water only
(no oil).  Referring to  the Run A data for those tests, the
values of Kkr ranged from 0.78/hr to 0.98/hr, with a mean
of 0.90/hr.  Figure 42 is a semilog plot of the Krypton:
tritium concentration ratios for two typical reactor tests
with mineral oil.  The two distilled water runs, 36A and
40A, were characterized  by Kkr values of 0.97/hr and 0.98/
hr, respectively (see Figure 42).  Run 36B contained 360
mg/1 of mineral oil, and an associated Kkr of 1.99/hr was
observed, whereas Run 4OB contained only 16 mg/1 of mineral
oil and had an associated Kkr of 1.14/hr.  The "alpha" fac-
tors were 2.05 for Test  36, and 1.17 for Test 40.  As may
be seen from Figure 42,  the observed data describe excellent
straight lines on the semilog plots, and the individual
derived vlaues of Kkr contain very little possibility of
error.

It is evident from the results presented in Table 24 and
Figure 42 that the presence of mineral oil enhances the gas
transfer capacity of turbulent water, an effect opposite to
that observed for the surfactant LAS.

Figure 43 shows the effect of mineral oil on the magnitude
of Kkr at a constant rate of mixing.  The values of Kk  have
been plotted against the oil concentration, in mg/1, for
Tests 31 through 41.  The trend of results is clear - the
magnitude of Kkr increases sharply with increasing concen-
tration of oil, an effect opposite to that observed with
LAS.  The relationship appears to be curvilinear rather than
straight.  As shown in Figure 43, one result (Test 38) ap-
pears to be quite doubtful, for reasons not known.  Actu-
ally, reference to the results in Table 24 and to Figure 43
indicates that the four  results for the sequence of Tests
37 through 40 are all somewhat high on a relative basis.
Regardless of the reason for such scatter, in no case did
the presence of mineral  oil fail to increase the magnitude
of Kkr'  As snown in Figure 43, the trend of results is al-
so through the observed  mean value for distilled water  (no
oil) at zero oil concentration.

Figure 44 is a plot of the "alpha" factor vs. mineral oil
concentration at essentially constant stirrer speed  (Tests
31 through 41).  In no case was an "alpha" of less than
1.00 observed; in every  case the Kkr for Run B  (water +
oil) was greater than the comparable K,   for Run A.  Refer-
ring to Figure 44, the trend of observed results is quite
clear, there being only  one observation  (Test 38) that is
                          164

-------
o

H
<


z
o
I—I
H

o:
LU
O
z
o
a
     I

  0,8

  0,6


  0,4
o
i-
D.

cc
  0,2
  0,1

 0,08


 0,06
^
fl
V.
\
RUN 4 OB
OIL = 1
Kk = 1
vl
\
\J
^


RUN 361
TEST 40
oc
^>D^
^>s
^
— / ^
6 MG/L
.14/hr.
^""V

<
X
>
//\1
3— / "X
OIL = 360 MG/L
Kkr = 1.99/hr.


TEST 36
1.99
" ~ 0.97
FIGURE 42
20°C
EFFECT OF MINERAL
ON KKR
(LOG SCALE SHIFTE
FOR CLARITY)
_ 1.14 _ ,
0.98 L

^
^,

>x.
"^v^
^^x,^
o^


.17

y 	 Rl
J> OJ
^S- K,
TX^.^^ \
^^



^S"^^O
^

XRUN 36A— '
OIL = 0 MG/
Y - n Q7
\|
= 2.05


nil

D

kr - • - • /
\








IN 40A
:L = o MG/I
= 0.98A
:r
\
^\
"V^U
^x.



^-s
'L ^
/hr

s.
X
\
X
V
HS




J
ir .
k






^






\
      0
                20        40         60        80
                     ELAPSED  TIME  IN MINUTES
100
                           165

-------
                       Table 24



            Reactor Tests with Mineral Oil
Test
Number
31

32

33

34

35

36

37

38

39

40

41

42

43

Run
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
mg/1
0
230
0
458
0
230
0
121
0
45
0
360
0
17
0
41
0
42
0
16
0
123
0
230
0
230

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

0.

5.

1.

894

866

789

849

780

970

904

914

956

978

945

01

43


1.

2.

1.

1.

1.

1.

1.

1.

1.

1.

1.

5.

2.

73

03

70

36

09

99

19

73

41

14

46

01

01

1.

2.

1.

1.

1.

2.

1.

1.

1.

1.

1.

1.

1.

94

34

87

60

40

05

32

89

47

17

54

00

41
                        166

-------
2,20
2,00
1,80
1,60
 1,40
 1,20
 1,00
0,80
0,60


O= — TI
(I
o
/
1
1
1
^-DISTI]



]ST 38
50UBTFUL)
/
/
O

LLED


/
/





V
ox







/^ o













FIGURE 43
20°C
EFFECT OF MINERAL 0 1 _
ON Kkr
(CONSTANT MI\I\^
0        100
                        200        300       400       500
                     MINERAL OIL  CONCENTRATION/ MG/L
600
                           167

-------
2,4
2.2
2,0
1,8
             TEST  38
             (DOUBTFUL)
                                       -0-
1,6
1.4
               O
1.2
1,0
         DISTILLED
         WATER
      FIGURE 44
        20°C
EFFECT OF MINERAL OIL
     ON "ALPHA"
  (CONSTANT MIXING)
             100       200       300       400
              MINERAL OIL CONCENTRATION/ MG/L
                 500
                        168

-------
clearly not compatible with the other results.  Even in that
case, the observed value of "alpha" is too large, rather
than too small, indicating too great a value of Kk  for
Run B.  The reason for this odd result in Test 38 is not
apparent from careful review of the test data, and cannot
be stated here.

Referring again to Table 24, four of the total of thirteen
tests were conducted at the same concentration of mineral
oil  (230 mg/1)  but at different stirrer speeds, as evidenced
by the results for Kkr with distilled water only.  Those
four tests were Numbers 31, 33, 42, and 43.  There are not
sufficient data for plotting and curve-sketching, but again
the results of the tests are quite clear.  In brief, it is
evident that, for constant concentration of mineral oil,
the value of "alpha" diminishes as the rate of mixing is in-
creased, and at relatively high rates of mixing  (see Test
42) the mineral oil was found to have no effect on Kkr
("alpha" = 1.00).  In contrast, at the lowest rates of mix-
ing  (Tests 31 and 33) the highest values of "alpha" were
observed.  Again, these observed results show an effect
just opposite to that observed earlier in the experiments
with LAS - in that case, the greater the turbulent mixing,
the greater the effect of LAS on Kkr and KQX.

The fundamental reasons for the foregoing observed effects
'of mineral oil (and, for that matter, LAS) have not been
explained by these experiments.  In all of the experiments
with mineral oil, the degree of turbulent mixing was suf-
ficiently great to keep the oil apparently uniformly dis-
tributed throughout the test volume of water, at least to
the extent that this could be observed visually.  No oil
film was ever observable at the water surface.  Of course,
the concentrations of oil were relatively large as a rule,
but even at the lowest concentrations, 16 or 17 mg/1, there
was a definite enhancement of gas transfer in the presence
of the oil.  One can speculate as to the reasons for this
observation.  For example, it would appear possible that
the oil altered the quality of the test water in the sense
of making less difficult the escape of the krypton gas
molecules at the water surface.  Alternatively, it appears
quite possible that the oil acted in some way as a lubricant,
modifying the surface tension of the water or increasing the
rate of surface water replacement, especially at the walls
of the vessel, in such a fashion as to facilitate gas trans-
fer.  Neither these nor other speculations can be demonstra-
ted or tested with the available results.  However, the
clear demonstration of the opposing nature of the effects
of mineral oil and LAS emphasizes the importance of further
research designed to examine more closely the basic reasons
                          169

-------
for such effects on the gas transfer capacity of turbulent
water systems.

POLLUTANT EFFECTS - RIVER WATERS

A total of fifteen separate pollutant effects tests were
conducted with water samples taken from the South and
Chattahoochee Rivers in the vicinity of Atlanta.  The pur-
poses of these tests were to demonstrate the test procedure
with natural river water samples and to obtain some estimate
of the actual effects on reaeration capacity of treated or
partially treated wastes routinely released to these streams,

Weekday Tests.  Ten of the tests of pollutant effect on
stream reaeration capacity were conducted on samples col-
lected from the streams during the usual work week  (Monday
through Friday), when pollution loads are relatively high.
Table 25 provides a summary of those test results, and also
includes one other set of results for a test (Number 44)
comparing distilled water with clean Chattahoochee River
water collected at the City of Atlanta Water Works intake.
The comparison of distilled water against clean Chattahoo-
chee River water indicated no significant difference (See
Test 44 - "alpha" = 1.05)  in gas transfer capacity.  In
tests 45 through 49, Chattahoochee River water taken at
Station 0 (See Section VI)  was compared to clean river wa-
ter from the water works intake location.  In each case the
gas transfer capacity for the polluted river water was
lower than that for the clean upstream river water, the
"alpha" values ranging between 0.60 and 0.88.  Although the
number of tests was not sufficient for firm conclusions
regarding the effect of the mixing rate, the results appear
to indicate a decreasing "alpha" factor associated with an
increase in the rate of mixing, as noted earlier in connec-
tion with the pure LAS studies.

Tests 53 and 54 compared polluted Chattahoochee River water
against distilled water.  The river water samples were col-
lected only about 1,000 feet below the Clayton STP outfall,
in order to obtain samples with the highest degree of pol-
lution, and to more definitely ascertain the source of pol-
lution causing the reduction in gas transfer capacity.  The
Run B "alpha" values of 0.80 and 0.83 at two different mix-
ing speeds indicated definitely that the Clayton STP efflu-
ent was responsible for the reduction in gas transfer capa-
city found in the earlier tests (Test Numbers 45 through
49)  of Station 0 river samples.

Tests 50, 51 and 52 compared pollutant effects in the South
River against distilled water, as no really unpolluted
                          170

-------
                       Table 25



   Reaeration Effects of Pollution Loads in Streams



                    (Weekday Samples)
Slea'h"1 -•""•""l" 	 'Polluted '
Test
Number
44
i
45

46

47

48

49

50

51

52

53

54

Run
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
A
B
Distilled or
Water Works
Dist.
WW
WW

WW

WW

WW

WW

Dist.

Dist.

Dist.

Dist.

Dist.
-
KVT./hr(20°c)
River

Chatt

Chatt

Chatt

Chatt

Chatt

Chatt

South

South

South

Chatt

Chatt
Station*

-

0

0

0

0

0

G

A

J

**

**
Clean
2.23

2.35

2.47

0.520

; 1.92

0.49

0.937

0.830

1.04

0.830

2.04

Polluted

2.35

1.48

1.48

0.404

1.17

0.43

0.796

0.680

0.74

0.690

1.63
"alpha"

1.05

0.63

0.60

0.78

0.61

0.88

0.85

0.82

0.71

0.83

0.80
* See Section VI for Station locations.



** About 1,000 ft below Clayton STP outfall.
                          171

-------
location was available as a source of clean test water.  Re-i
ferring to Table 25, the polluted sampling station locations
are the same as those referred to in Section VI, and the ob-
served "alpha" values ranged from 0.71 to 0.85.  The mean
"alpha" value was 0.79 at an average mixing rate associated
with a Kfcr for distilled water of 0.94/hr.  The South is
heavily polluted at the locations sampled, and is known to
have LAS concentrations of 2 mg/1 or more at times.  Refer-
ence to the effects of LAS on gas transfer (see Table 22, es-
pecially)  indicates that LAS was probably not the only pollu-
tant involved in reducing the gas transfer capacity of the
South River water.

Sunday Samples.  All of the Chattahoochee River field tracer
studies of reaeration capacity were conducted on Sundays, be-
cause that was the  only day of the week when steady low river
flows could be obtained (the Chattahoochee is used for power
production, and has a highly variable flow during other days
of the week).  Thus, the river was not as highly polluted
during the tracer studies, compared to usual weekday periods.
Accordingly, a series of five tests was conducted with river
water collected on  Sundays during the field tracer studies.
The results of those pollutant effects tests are provided in
Table 26.   Referring to Table 26, the field study numbers
and the station identification numbers are those referred to
in Section VI, and  all polluted test results were compared
to clean water samples taken at the City of Atlanta Water
Works intake (see Runs designated A in Table 26).

                        Table 26
    Reaeration Effects of Pollution Loads in Streams

                    (Sunday Samples)
Field    Test                  Kkr/hr (20°C)
                                                  "alpha"


                                                   0.94

                                                   1.22


                                                   1.03


                                                   0.96


                                                   1.05
  XIX


   XX


  XXI


 XXII


XXIII
55

56

57

58

59

A
B
A
B
A
B
A
B
A
B
ww
 3

WW
 2

WW
 4

WW
 4

WW
 3
Clean
0

0

0

0

0

.445

.107

.458

.446

.431

Polluted

0

0

0

0

0

.419

.131

.470

.430

.452
                          172

-------
Referring to the results given in Table 26, the observed
values of "alpha" ranged from 0.94 to 1.22, with an average
of 1.04.  Test Number 56 was somewhat different - it was con-
ducted at a considerably lower rate of mixing  (K,   for Run
A = 0.107/hr vs. about 0.44/hr for the other four tests), in
an effort to achieve a degree of turbulence more nearly ap-
proaching that of the river itself.  As may be seen, at that
low mixing rate it appears that it was more difficult to re-
produce the same mixing condition in the test system, and a
high "alpha" value of 1.22 resulted.  Neglecting the result
for Test 56, the observed "alpha" values at essentially con-
stant mixing speed ranged only from 0.94 to 1.05, and aver-
aged 1.00.  These test results indicated clearly that on
Sundays the reaeration capacity of the Chattahoochee River
was not significantly affected by the presence of pollution.
Thus, the reaeration capacities reported earlier in Section
VI of this report probably reflect the pollution situation
of the future  (when the Clayton STP has been upgraded to
secondary treatment and there is no bypass of untreated
sewage) rather than the present weekday polluted condition.
                          173

-------
                     SECTION VIII

      HYDRAULIC PROPERTIES RELATED TO REAERATION
This section of the report contains an introductory discus-
sion regarding the traditional concepts of the relationships
between stream hydraulic properties and reaeration capacity,
together with a critique based upon the theory presented
earlier (Section IV).  This introductory review is followed
by comparisons of the observed reaeration coefficients with
those predicted by the various available predictive models,
together with appropriate comment.  Finally, a new approach
is presented, in which an energy dissipation model for
stream reaeration is developed from simple theoretical con-
siderations and the resulting models for gas transfer and
reaeration are tested with the results obtained from the
field tracer studies.
INTRODUCTORY DISCUSSION

It was noted as early as 1911 by Black and Phelps  that
stream reaeration is directly related to the rate of turbu-
lent mixing, and almost 50 years ago Streeter and Phelps
reasoned that their newly defined reaeration rate coeffi-
cient, K2, could be related to stream turbulence by means of
an empirical model involving the velocity and depth of flow.
Although other hydraulic properties have been considered by
various subsequent investigators, for example, Churchill et
al , Krenkel and Orlob^ and Thackston and Krenkel7, the
predictive models most commonly used today '  still explain
stream reaeration in terms of the mean velocity and depth of
flow.

It has long been recognized that all of the available pre-
dictive models relating ¥-2 to stream hydraulic properties
leave something to be desired.  Most of them incorporate
theoretical assumptions that cannot be verified, none have
been adequately field-tested, and each model appears to
provide adequate predictions of K2 for some streams but not
others.  As indicated earlier, the principal difficulty en-
countered in such attempts to relate stream reaeration capa-
city to the hydraulic properties has been that neither the
reaeration capacity nor the associated degree of turbulence
could be measured directly or independently.

As indicated in Section IV [see equation  (4)] and derived
earlier from simple considerations of the kinetics of
      ^ the reaeration rate coefficient, K2, is directly
                           175

-------
proportional to the rate of water surface replacement in a
turbulent water system.  Hence, "turbulence" refers here
quite specifically to the rate of surface replacement, and
a clear distinction must be made between stream hydraulic
properties that are related to surface replacement  (and
reaeration) in a primary way, or cause it, and those hydrau-
lic properties that may be related to "turbulence"  (and
reaeration) only in an indirect or secondary way.  In any
attempt to relate stream reaeration capacity to hydraulic
properties it will be desirable to express the rate of water
surface replacement in terms of the primary or causative
hydraulic properties.  In addition, it will be necessary to
select or develop ways to measure those primary hydraulic
properties with accuracy and without ambiguity.

Before proceeding to detailed tests of the available pre-
dictive models and, subsequently, to the development and
testing of the new energy dissipation model, some further
consideration of certain of the specific hydraulic properties
appears desirable.  This relates, in particular, to the real
meaning of terms such as "mean velocity" and "mean depth",
whether such properties are of primary or secondary impor-
tance in regard to surface replacement, and whether the
field methods of observing such properties are adequate.

One of the real problems encountered in this research relates
to the basic meaning of certain traditional measures of
hydraulic properties.  Stated another way, sometimes the
very method of measurement of a hydraulic property, or the
method of computing it, modifies its real meaning in the
physical sense.  For example, if we consider a section of a
natural stream, the way in which we obtain the mean water
depth affects its real meaning in terms of its relationship
to the actual rate of water surface replacement.  For in-
stance, the model

            Depth = 0ccuP:'-ed Channel Volume
                        Surface Area

is valid only if there is complete and homogeneous mixing of
all of the water in the channel, especially vertical mixing.
To consider an extreme, in a stratified channel or reservoir
the whole volume is separated into hydrodynamic regions, the
depth that is directly involved in surface replacement  (and
reaeration) is much smaller than the above expression would
imply, and reaeration of the lower region is virtually nil.
In many natural streams of relatively small slope the water
depth that is directly involved in surface replacement and
reaeration is considerably smaller than the measurable whole
depth of flow, due to poor vertical mixing.
                          176

-------
Some of the hydraulic properties of natural watercourses are
considered below, together with methods of observation or
measurement and the resulting implications as regards re-
aeration capacity.

Velocity.  Most of the available models for predicting re-
aeration capacity include the "mean velocity", either di-
rectly or indirectly, and at first glance it seems obvious
that the rate of water surface replacement ought to be a
function of the velocity.  But on closer inspection two
questions present themselves:  first, how shall this "mean
velocity" be measured, and secondly, does this method of
observation affect the usefulness of the result as a measure
of surface replacement?  A third question, namely, is veloc-
ity a basic property that causes surface replacement, is
perhaps the most important question of all.

There are at least two commonly used methods of obtaining
the "mean velocity".  The first involves direct physical
measurement of the velocity at a number of locations in a
stream cross-section by the use of a current meter; if
enough such measurements are made, a reasonably accurate
measure of the average forward velocity through that cross-
section can be obtained.  If, then, this procedure is re-
peated at a sufficient number of cross-sections in a speci-
fied length of stream channel, the results can be combined
to obtain a reasonably accurate estimate of the mean forward
velocity of flow that prevails throughout the length of the
stream section.  This procedure is subject to certain ob-
vious sources of error relating especially to the statisti-
cal adequacy of the number of observations made in any one
cross-section, the statistical adequacy of the number of
cross-sections involved, and the accuracy of the current
meter observations when forward velocities are relatively
small.  But in addition, and even more importantly, there
would appear to be legitimate question as to whether the
forward velocity is that velocity that is most nearly related
to the rate of surface replacement.

The other commonly used method of obtaining the "mean veloc-
ity" involves measurements of the distance travelled and the
time of flow.  The distance travelled can be obtained
readily and with quite adequate accuracy from USGS quad-
rangle sheets or by field survey; the time of flow can be
measured with great accuracy by the use of dye tracers.  The
resulting "mean velocity", the distance divided by the time,
is relatively precise because of the precision of the mea-
sures involved.  Depending upon the degree of homogeneity
and completeness of mixing in the channel, it is not neces-
sarily the same  "mean velocity" as that obtained by the
                           177

-------
first method outlined above, but it clearly reflects the ac-
tual forward velocity that is effective in the channel.
Whether or not such a forward velocity adequately relates to
the rate of surface water replacement is again open to seri-
ous question.

Both of the foregoing procedures for obtaining a representa-
tive mean velocity for a length of stream neglect the real
velocity fluctuations that occur within the stream reach at
changes in slope or cross-section.  Yet, in terms of the
real rate of surface water replacement, the variability of
velocity, or the range of velocities, within a stream reach
may well be of more significance than the magnitude of the
mean velocity itself.  Also, although velocity, or its vari-
ability, may well be related in some way to surface replace-
ment and reaeration, it would not seem to be a primary hy-
draulic  property, or one that is a basic cause of surface
replacement.  Rather, the velocity is the result of some
other property such as channel slope, constrictions, etc.,
and in that sense is more of the nature of a secondary hy-
draulic property.

Thus, although a measure of stream velocity can be obtained
as outlined above, and although such a measure may be a
"mean" in the usual sense, there is real question as to its
usefulness as a representation of surface replacement.  Per-
haps a more meaningful measure would be an estimate of the
mean vertical velocity component in the stream channel, as
this would seem to be more directly relatable to the rate of
surface replacement.  Again, however, the very method of ob-
servation could greatly affect the meaning and the usefulness
of the result.

Depth.  As reaeration is a direct function only of the rate
of surface replacement, stream depth has importance only in
terms of a possible relationship to the rate of surface re-
placement, and only then if mixing is complete and homo-
geneous.  Although it seems unlikely that depth itself is in
any way a direct cause of surface replacement, a brief re-
view of the methods of observation and the meaning of the
derived results is of interest.

One method of obtaining the "mean depth" has already been
outlined, namely, the result of dividing the occupied channel
volume by the whole surface area, and its meaning has been
discussed.  An effective channel volume may also be obtained
as the product of flow and time of passage—this might lead
to a more representative effective depth in some situations
(stratified reaches) but not necessarily in others  (short-
circuiting) .  Another commonly considered method of
                          178

-------
observing the "mean depth" involves measurement of the dimen-
sions of the stream cross-section by field survey, wherein
the mean depth of the cross-section is obtained by dividing
the observed cross-sectional area by the measured stream
width.  If an accurate measure of the "mean depth" of a
length of stream channel is to be obtained by this means, a
substantial number of cross-sections must be included for
purposes of statistical adequacy-  A substantial amount of
field survey work is therefore involved.  However, the re-
sult is subject to much the same criticism as was made for
the volume/surface area method - the "mean depth" obtained
is a measure of the effective depth only if there is complete
and homogeneous mixing in the stream channel, and this re-
quirement is not met in a large number of cases.

It should be noted here that although the term  (A/V) appears
in equation  (4), Section IV, this term is not really to be
construed as the reciprocal of the depth.  Rather, the entire
term  (nA/V) must be considered as an entity, and its inter-
pretation is best restricted to the derived meaning - it is
the rate of surface replacement in cm^/second/unit volume.

The various methods of obtaining an accurate measure of the
mean depth of flow in a length of natural stream channel are
tedious at best, even if there is complete and homogeneous
mixing.  In any event, the real meaning of such measures in
terms of the rate of surface water replacement is not
readily apparent, and it appears unlikely that depth itself
has any causative relationship to surface replacement.

Slope.  The physical slope of a natural stream channel,
namely the decrease in elevation per unit of channel length,
is readily observable by field survey, although the fieldwork
may be somewhat tedious and time-consuming.  Surprisingly,
although the slope would appear to be an important hydraulic
feature, such measurements are not commonly made or available,
Indeed, intuitively the slope of the stream channel would
appear to be more nearly a determining or a causative pro-
perty than most others - it is an independent property ex-
cept where engineering works have modified it, and properties
such as the velocity and depth of flow are functions of the
slope rather than vice versa.  In essence, the steeper the
channel slope, the more violent the tumbling action that
creates water surface replacement, and, hence, it appears
that the channel slope should not only be related to the
rate of surface replacement but should, in fact, be a basic
cause of surface replacement.  As indicated above, it can be
measured with entirely satisfactory accuracy.
                           179

-------
Roughness.  One other property that would seem to be impor-
tant in terms of water surface replacement is the physical
channel roughness, in the sense that a very rough rocky
stream bed should create better vertical mixing than a
smooth sandy stream bed.  Of course, the bottom roughness
cannot be measured directly or independently, and the avail-
able method of obtaining an estimate of bottom roughness,
namely calculation by means of the Manning equation, is cir-
cuitous and subject to substantial error.  In addition, the
character of a stream bed, or its physical roughness, is not
so independent a hydraulic property as might appear at first
glance - in fact, the bottom character results from proper-
ties such as the velocity and the slope of the channel.
Hence, although the bottom roughness may be related in some
way to the degree of vertical mixing and the rate of surface
replacement, it would not appear to be a basic property that
independently causes surface replacement.

Certain of our experimental results have caused us to view
the hydraulic properties in a somewhat different way that
appears to have more promise in terms of developing a basic
relationship between stream reaeration capacity and hydrau-
lic properties.  This point of view involves consideration
of the relationship between surface replacement and energy
dissipation.  However, before proceeding to the development
and testing of this energy dissipation model, the reaeration
rate coefficients observed in these research studies will be
compared to those predicted by use of the several available
predictive models.

Time of Flow.  Reaeration and stream self-purification are
direct functions of time, and the time of flow within a
stream reach is thus an important hydraulic property quite
aside from any other consideration.  In addition, however,
the time of flow within a stream reach is a specific hydrau-
lic characteristic of that reach, and represents the net ef-
fect of other hydraulic properties  (e.g., slope, flow, veloc-
ity, etc.) that are related to turbulence and water surface
replacement.  Thus, even though it is clearly not a property
that independently causes reaeration, the time of flow is
particularly significant in that it ties together both re-
aeration and other stream hydraulic properties that are as-
sociated with turbulent mixing.  With the introduction of the
use of fluorescent dye tracers about a decade ago, accurate
measurement of the time of flow of a stream has become a sim-
ple matter.
COMPARISONS WITH AVAILABLE PREDICTIVE MODELS

In order to determine the predictive capability of  some  of
                          180

-------
the previously available reaeration models, several of
these models were tested by comparing the reaeration coef-
ficients measured during the current study with values pre-
dicted by the models.  The models selected for testing were
the O'Connor-Dobbins model  (4) , the Churchill model (5) ,
the Langbe in-Durum model (14), the model proposed by Owens,
Edwards, and Gibbs (16), and the Thackston-Krenkel model
(7).  Equation (9) was used to adjust each of these models
to a standard temperature of 25 °C.  The forms of the pre-
dictive equations used in this comparison were as follows:
O'Connor-Dobbins
Churchill, et al
                          K_ = 0.573 V°'5/H1>5
                             = 0.543 V° * 969/H1-673
Langbein and Durum
                          K_ = 0.353 V/H
                                        1'33
Owens, Edwards and Gibbs  K  = 1.020 v
                                      0 Q7  1 RR
                                      u ' 3 '
Thackston-Krenkel

     K0 = 1.122 (1+
                               1/2
                                  )  (Sg/H)1/2
The adequacy of each of these five models was tested by
comparing the predicted values of ¥-2 with those measured
during the tracer studies.  The hydraulic parameters used
in the predictive equations were determined by detailed
hydraulic measurements on the various rivers.  The hydrau-
lic measurement techniques have been described in Section
V.  The measurements made on the South, Flint, and Chatta-
hoochee Rivers during the hydraulics study were adjusted by
the procedure described in Section V to make the values of
the hydraulic parameters consistent with the discharge
rates in the streams during the tracer studies.  Such ad-
justments were not possible for the Patuxent and the Jack-
son River studies, and typical values of the hydraulic
parameters were used for these streams.

The five predictive equations taken together include only
three hydraulic terms, velocity (V), depth  (H), and slope
(S).  The velocity was measured directly during each tracer
release by dividing the distance between sampling stations
by the time of flow as measured by the dye tracer.  Hence,
the only two parameters that were adjusted were depth and
slope.  Discharge rates measured during the hydraulic
                          181

-------
studies were, in most cases, not greatly different from
those measured during the tracer studies and, consequently,
the required adjustment in depth at each 500-foot cross
section was generally small (0.1 to 0.2 feet).  As a result,
the values of depth used in the predictive equations for
the South, Flint, and Chattahoochee Rivfers, which were the
average values of the depths at each of the 500-foot cross
sections between sampling stations, are considered to be
quite accurate.  The slope was determined by taking the
difference between the water surface elevations at upstream
and downstream sampling stations, as adjusted for flow
rates, and dividing the difference by the distance between
stations.  The values of slope used in the predictive
equations, like the values of depth, are thus also taken to
be quite accurate.  The unadjusted values of depth and slope
used for the Patuxent and Jackson River studies are believed
to be very close to the values existing at the time of the
tracer studies.

The test of the predictive equations was based on a corre-
lation and regression analysis.  Regression lines of the
form shown below were used to relate predicted and measured
values of K-.



In this regression equation the values of Recalculated from
the predictive equations are represented by X, and Y is the
measured value of I<2.  In addition to determining the slope
(a^) and the intercept (ao), the correlation coefficient
(rxy) was computed for each predictive equation.  The
values of ao/ a.]_r and rxy, as determined by a least squares
fit of the predicted and observed reaeration data, were
used as a measure of the capability Of each of the five
equations tested.  For a good model, that is, an equation
capable of predicting values of the reaeration coefficient
close to the values observed in the field, the value of ao
will be near zero and the value of ai will be near unity.
The possible range of the correlation coefficient is from
-1.0 to +1.0, and a value close to +1.0 indicates that the
variation of the predicted values of K^ about the regres-
sion line is small.

The values of ao, ai, and rxy computed for each model are
showh in Table 27.  Two sets of data were used from both
the Flint River and the South River.  A total of 89 values
of K2 were measured in the Flint River.  Five of these
values were measured at the two waterfalls on the Flint
(two values between stations IP and 1 and three between
2P and 2), and could not be included in the text since the
                          182

-------
                       Table 27

    Statistical Test of Selected Predictive Models
                      Flint River
(84 observations)
Model
O'Connor
Churchill
Langbein
Owen
Thackston
ao .

-0.06
-0.02
-0.09
-0.01
0.06
al

1.14
1.53
2.90
0.73
0.72
rxv
*fcj
0.68
0.75
0.74
0.76
0.34
                                    (43 observations)
                                    0.14
                                    0.14
                                    0.12
                                    0.14
                                    0.06
                                           al
                               0.16
                               0.24
                               0.47
                               0.11
                               0.42
                                      -xy
            0.31
            0.38
            0.41
            0.36
            0.53
                      South River
  Model

O'Connor
Churchill
Langbein
Owen
Thackston
(96 observations)

  ao   al    rxy

 0.34 -0.03  0.00
 0.40 -0.26 -0.06
 0.42 -0.48 -0.07
 0.39 -0.13 -0.05
-0.44  3.03  0.90
                                    (91 observations)
 0.20
 0.20
 0.19
 0.20
-0.06
0.04
0.05
0.10
0.04
1.25
             r
             'xy
0.04
0.06
0.08
0.08
0.67
                    Patuxent River
(30 observations)
Model
0' Connor
Churchill
Langbein
Owen
Thackston
a .

0.08
0.09
0.09
0.10
0.02
al

0.19
0.25
0.46
0.12
0.54
r
XV
•"•-/
0.50
0.53
0.56
0.51
0.44
                          183

-------
     Table 27 (con't)
       Jackson River
             (80 observations)
  Model        ao   al    rxy
O'Connor     0.12  0.27  0.14
Churchill    0.12  0.52  0.21
Langbein     0.10  0.84  0.24
Owen         0.12  0.27  0.19
Thackston    0.11  0.18  0.20
    Chattahoochee River

             (30 Observations)
  Model        a    a.,    r
  	         o	1	xy
O'Connor     0.05 -0.05  -0.06
Churchill    0.05 -0.05  -0.06
Langbein     0.05 -0.04  -0.04
Owen         0.05 -0.03  -0.06
Thackston    0.03  0.25   0.10
            184

-------
available equations are not able to predict K2 values for
such situations.  The second set of data used from the
Flint contains 43 values and is a subset of the 84 values
used in the first test.  All reaches of the Flint which
include waterfalls or rapids were excluded from this sub-
set.  A total of 96 values of K2 were measured on the
South River and all 96 are included in the first test.  In
the second test those values of K2 measured at Panola
Shoals (between stations K and L) and those measured over
a reach containing rapids( G* to T2) were eliminated.

When the data set which includes the 84 values from the
Flint is used to test the models, the O'Connor-Dobbins
model does a better job of predicting the observed values
than do the other four models.  All of the models yield
values of ao close to zero.   (The term "close" is, of
course, a qualitative description.  The value of a0 might
be compared to the average value of the observed K2' s in
order to judge the relative significance of the intercept.
In the present case the average of the 84 observed values of
K2 is 0.38 'and the range is from 0.03 to 2.49.  Thus, the
values of the intercepts shown in Table 27 range from about
3% to about 24% of the average.)  The slope of the O'Connor-
Dobbins model is close to unity, while the Churchill model
and the Langbein model tend to underpredict the observed
values and the Owen model and the Thackston model tend to
overpredict.  An examination of the first four models
listed above shows that they are all of similar form but
that the coefficient in the Owen model is two to three
times that in the other three models.  Thus, the Owen mod-
el can be expected to predict higher values, particularly
when compared to the Churchill and Langbein models which
have similar values of the exponents.  All of the models,
with the exception of the Thackston model, produce correla-
tion coefficients in the range 0.68 to 0.76.  The Thacks-
ton model produces a correlation coefficient of only 0.34
for this set of data.

When all reaches on the Flint which contain waterfalls and
rapids are eliminated, the test shows that all of the mod-
els tend to overpredict the higher observed values and to
underpredict the lower values of K2.  The range of observed
values of K2 in this subset is from 0.03 to 0.53 and the
average value is 0.17.  The Thackston model produces the
highest correlation coefficient for this reduced set of
Flint River K2' s  (^  =0.53).

When the set of K2 values measured on the South is con-
sidered, it is evident that only the Thackston model is
capable of predicting values reasonably close to those
measured.  All of the other four models produce negative
                          185

-------
values of a± and rxy when all 96 values are used in the
computations.  This indicates that the observed values tend
to increase as the predicted values decrease.  The ex-
tremely low values of rxy for the first four models indi-
cate that there is really no correlation between the ob-
served and predicted values and that the negative values of
ai are not significant.  The high correlation coefficient
produced by the Thackston model (rxy = 0.90) is associated
with values of a0 and aj_ significantly different from the
"good" values of zero and unity, respectively.  The value
of -0.44 for a0 and 3.03 for a-j_ indicate that the Thackston
model strongly underpredicted the higher observed values
of K2.  The observed range of K2's predicted by the Thacks-
ton model was from 0.06 to 0.33.

Removing the higher observed values of K2 from the test
data for the South River improved the values of ao, a^,
and rxy only slightly for the first four models.  Some im-
provement is also seen to occur in the slope and intercept
of the Thackston model but the tendency of the model to
underpredict the high observed values and to underpredict
low values is still present.

All of the models tended to overpredict the values of K2
observed on the Patuxent.  The average observed value of
Kj was 0.14, thus, the values of ao (the intercept) for the
first four models represent values almost as large as the
average.  The Thackston model has the largest value of a^,
0.54, and hence, tends to overpredict by a smaller amount
than do the other models.  The correlation coefficient for
the Thackston model was smaller than for the other four
models, but all the values for all five models were similar
(0.44 to 0.56).

The tendency of the models to overpredict observed values
of K£ continues to persist when the values observed and
predicted for the Jackson River are examined.  On the basis
of the Jackson data there is no reason to prefer one model
over any other.  All produce low correlations (rxy ranging
from 0.14 to 0.24).  The values of ao range from 0.10 to
0.12 as compared to an average value of K2 equal to 0.15.
Thus, most of the regression lines for this river are es-
sentially horizontal lines through the mean value of K2
observed.

The Thackston model is the only one of the five that pro-
duced positive values of a-j_ and rxy when the models were
tested with the Chattahoochee data.  However, the regres-
sion line for the Thackston model yields a correlation co-
efficient of only 0.102, and the value of ao  (0.03) is near
the average of the observed values (0.04).


                          186

-------
The test of the five selected models using seven sets of
data from five rivers of varying characteristics indicates
that none of the models are capable of accurately predic-
ting the values of K2 observed during tracer studies on
these rivers.  In most cases, the regression line fitted
to the observed and predicted data had a slope less than
unity, indicating a tendency to overpredict observed values,
and no model was consistently capable of yielding high
correlation coefficients.
                              187

-------
ENERGY DISSIPATION MODELS - THEORY

Consider a length of natural stream channel between  two
points, 1 and 2.  The usual one-dimensional energy equation
indicates that the amount of energy expended in  ft-lbs  per
Ib of water between the two points is
                    2                 2

                      (42)
where V is the velocity in ft/sec, z is the elevation  of
the stream bed above mean sea level in ft, H is the depth
of water in ft, and g is the gravitational constant in
ft/sec^.

Rearranging terms,
                            2     2
                          Vl  - V2
                 - E) * (-^	—)  + Ah
where
                         - (z2 + H2) = Ah              (44)
and Ah is the change in water surface elevation between
points 1 and 2.

With few exceptions, the difference in velocity head,
(Vl  ~ V2 )/29' is negligibly small compared to the change
in elevation head, Ah.  Hence, for most reaches of stream


                    (E.. - E,) * Ah                     (45)
for practical purposes.

The rate of energy expenditure is just the amount of ener-
gy expended per unit time, or
                          188

-------
                                                       (46)
where tf is the time of flow from 1 to 2.

It has also been shown in our earlier work that
                       K2 = al4
where K^ refers to the gas transfer coefficient for any
gas, the constant, a, refers to the molecular properties
of the gas and the quality of the water, n is the number
of surfaces of area A replaced per unit time, and V is the
whole volume of water.  The product  (n=r) is therefore just
the rate of surface replacement in cm^ per second per cm
of volume, if metric units are employed.

It appears logical to suppose that the rate of water sur-
face replacement will be related to the rate of energy
dissipation, probably in a simple and direct way.  Accord-
ingly, the following relationship has been postulated:

POSTULATE:  The rate of water surface replacement is pro-
            portional to the rate of energy dissipation
            in open channel flow.

Using the expressions given in equations  (4) and  (46)
above, the postulate may be expressed as follows:
                      (nf) = b(^)                      (47)
where b is the necessary proportionality constant.

It now follows from equations  (4) and  (47) that


                      K2 = c(~)                       (48)



where c = ab
                          189

-------
Equation (48)  is our basic model relating the reaeration
coefficient, K2, to the stream hydraulic properties.  The
coefficient K? actually refers to any gas, including kryp
ton as well as oxygen, the only difference being the num-
erical magnitude? of the constant, c.  The hydraulic prop-
erties Ah and £ can be me,asure4 ^recti^ an4 independentl
as well as with quite satisfactory accuracy, for any
length of stream channel.  Hence, equation  (48) and its
underlying postulate can be tested directly with field ob
servations.  However, before doing so, one other useful
expression will be derived.

For the length of stream between points 1 and 2, equation
(5) from Section IV indicates that for desorption of the
tracer gas
where C, and C,
are the concentrations of dissolved tracer
                     by its equivalent
gas at points I and 2.  Replacing K
from equation (48), we obtain
                                                       (49)
where y is just the decimal fraction of dissolved tracer
gas remaining at point 2.  It follows also that
               (1 - y)  = z = (1 - e~cAh)
                                       (50)
where z is now the decimal fraction of dissolved gas that
has been lost between points 1 and 2.  Equations  (49) and
(50) refer directly to the tracer gas, krypton, but may
also be used to refer to the decimal fractions of the DO
deficit remaining and satisfied, respectively.

Equation (50) is of very strong interest.  It states, sim-
ply, that gas transfer in a turbulent natural stream is de-
pendent only upon the change in water surface elevation.
In other terms, at a given water temperature the amount of
tracer gas that will be lost to the atmosphere in a spe-
cific length of stream channel, or the amount of DO
                          190

-------
deficit that will be satisfied, can be predicted on the
basis solely of the change in water surface elevation be-
tween the upstream and downstream ends of the length of
stream.  Alternatively, the numerical magnitude of K2 can
be predicted on the basis of the change in water surface ;
elevation and the time of flow, according to equation  (48),

It is emphasized that the foregoing equations have been
developed with the implicit assumption of unrestrained
mixing and surface replacement.  Departures from this con-
dition in real stream situations will cause corresponding
discrepancies between the observed gas transfer character-
istics and those predicted by equations (48) and (50).
Such departures might involve, for example, appreciable
short-circuiting or stratification.  An example of such a
hydraulic situation has already been cited in regard to
the long, deep pool above Panola Shoals in the South River,

Other Models.  Although the two energy dissipation models
represented by equations  (48) and  (50) are regarded as the
basic expressions relating stream hydraulic properties to
reaeration capacity, it should be noted that they also
lead to other models that may prove to be of interest.  As
a single example, equation  (48) suggests that the rate of
energy dissipation can be represented also by the product
of a slope and a velocity.  Specifically,
                               x
                                  tf
in which L is the length of a reach of stream.  The term
 (Ah/L) is thus the slope of the water surface, and  (L/tf)
is the mean velocity determined by the distance travelled
and the time of flow.  The reaeration rate coefficient,
K2, may thus be expected to be proportional to the product
of the slope of the water surface and the mean velocity as
defined in equation  (51).

Other such relationships between reaeration and usual hy-
draulic properties may also be developed similarly from
the basic models given by equations  (48) and  (50), but such
relationships need not be provided here.  Instead, the ex-
perimental evidence in support of the basic models is pre-
sented below.
ENERGY DISSIPATION MODELS - OBSERVED RESULTS

The detailed field study results reported in Appendix AIV
                         191

-------
are used here for the purpose of testing the validity and
accuracy of the energy dissipation models, equations  (48),
(49) and (50).   The five rivers involved will be taken in
the order of their appearance in Section VI, namely, the
Flint, South, Patuxent, Jackson and Chattahoochee.  In each
case, the basic energy dissipation models for reaeration
will be tested according to standard statistical procedures,
and the results also presented in graphical form for ease
of interpretation.

Flint River.  Reference to Appendix AIV indicates that a
total of 89 observations of the reaeration capacity of
reaches of the Flint River were made during the field tra-
cer studies.  Of these, six observations cannot be used
here because the elevation of Station 7 was not accurately
measured during the field physical studies—the levelling
party proceeded only part of the distance between Stations
6 and 7 before discontinuing operations on the Flint.
Hence, the magnitude of Ah for the reaches 4-7, 5-7 and
6-7 is not accurately known for Dumps V and XII, and those
six observations cannot be used here.  This does not de-
tract from their use earlier, however, in tests of the
magnitude of K?, as that depends only upon the observed
gas loss and time of flow, both of which are accurately
known.

Figuza 45 presents all of the individual observed results
(see Appendix IV) for the Flint River in a form suitable
for interpretation in terms of equation  (50).  The percent
loss of tracer gas has been plotted on natural scales
against the water surface elevation change, Ah, in feet.
As may be seen, the data describe an exponential curve of
the form predicted by equation  (50).  The water surface
elevation change is a measure of the energy expended in the
reach of stream in each case, as outlined earlier, and the
observed curve thus provides powerful visual support of
the energy dissipation model given by equation  (50).

For purposes of statistical testing, the data shown in
Figure 45 have been plotted in the form required by equa-
tion  (49).  If the percent tracer gas remaining is plotted
against Ah on semilog scales, equation  (49) predicts that
a straight line should result, and the degree of correla-
tion between the logarithm of the percent gas remaining
and the water surface elevation change can thus be readily
obtained by usual methods.  For the 83 observed results
shown in Figure 46, a correlation coefficient of 0.957 was
found, indicating that 92 percent of the variation in the
logarithm of percent tracer gas remaining has been explain-
                          192

-------
V£>
                                                                        • DUMP XIV
                                                                        O OTHER DUMPS
                                                                     FIGURE 45
                                                                    FLINT RIVER
                                                                       25°C
                                                                   GAS LOSS VS,
                                                                  ENERGY EXPENDED
                 •  *STATION 7 OMITTED—SEE TEXT
                                                                   ALL RESULTS
                                  20         30        40         50
                              WATER SURFACE  ELEVATION  CHANGE/  Ah,  FEET

-------
  100
  80
  60
< 20
UJ
CO
<
CO
LU
O
   10
\ °
V& 0
• Sv
9f^j
01









3
3
^

0\jl
GO^

.'*^








r, O
O

^P °
*°\\
A^
(Ah)1/2 = 12.8 FT—/, ^K
n - fn \

r =0 .957
xy o











KEY: •
C











DUMP XIV
OTHER DUM
t

•









PS










0

N
\ rt
\
\\o c
•\^
\ >
* «\










FIGURE 46
pi T MT R T VPR
.'.'.". 25°c

GAS REMAINING VS ,
FNF'RfiY FXPFNOFD

ALL RESULTS
























. 	 (Ah.) , /0 = 14. 3 FT
_/^ /
^» n = 55
v r 0.981
ON xy
\\
°\ N,
\
\








%
N.
\ \

\







o

\
\


              10        20        30        40         50
                    WATER SURFACE ELEVATION CHANGE,  Ah, FEET
60
70
                                 194

-------
ed in terms of the relationship with  Ah predicted by equa-
tion (49).  When the regression  line  was forced through the
origin  (100 percent gas  remaining at  Ah" = 0),  as required
by the theory, the magnitude  of  the correlation coefficient
did not change, but remained  0.957,
                            .,i_v.,i, .„,„..„ _.T ..  ir	^   ,  the1
                            SKa"* nitegat'lve' slope of 6. 0543.
_,	,-,--,-.-	^-,	.-,--0,-_!iM«'"Of- ;c' ift equation (49); it
is referred tb he'fe as CQ,  to  indicate  that it is associ-
ated with the regression  line  forced through the origin.'
This observed value of c   is more readily interpreted and
understood in terms of a°half-height.   The half-height is
that numerical value of Ah that is  associated with  exactly
50 percent tracer gas remaining.  It is readily  shown that
the half-height,  (Ah),,-,  can  be found  in any case  from
the relation          '
                     (Ah)1/2  = ~~                   (52)
                                  o
Referring to the data at  hand
              (Ah)l/2 -T       =  12'8  feet
Hence, in the case of  these  Flint River data, 50 percent of
the tracer gas remained  dissolved in the water after the
river had dropped 12.8 feet,  25  percent remained after a
total fall of 25.6 feet,  12.5 percent remained after a
fall of 38.4 feet, etc.   Conversely, 50 percent of the tra-
cer gas was lost to  the  atmosphere in the first 12.8 feet
of fall, a total of  75 percent was lost in 25.6 feet of
fall, etc.

Referring to equation  (48),  and  recalling that Kkr = 0.83
K  , it is readily shown that
 ox7             2


             [(Ah)l/2]ox  = 0.83[
-------
case, 50 percent of the existing DO deficit would be satis-
fied by reaeration in 10.6 feet of fall, 75 percent would
be satisfied in 21.2 feet of fall, etc., if there were no
other sources of dissolved oxygen consumption.

Figures 45 and 46 also illustrate the variations that may
occur from one time to another in a stream like the Flint.
Referring to these figures, the data obtained from Dump
XIV have been blacked in so as to make them stand out on
the graphs.  In each case, the data for Dump XIV are clear-
ly different from the rest of the results in regard to the
curve or line of best fit.  Referring to Figure 45, the
Dump XIV results are clearly high, showing greater tracer
gas loss than usual; in Figure 46, the slope of the line
of best fit is clearly steeper for the Dump XIV data taken
alone.

Dump XIV was the only Flint River study conducted on a week-
end  (a Saturday), all of the other tracer releases having
been conducted on weekdays.  As a result, the degree of
pollution from the single source  (the Flint River STP)  was
lower during Dump XIV than on weekdays, and so was the
river flow, and the times of flow were substantially longer
for Dump XIV.  The data for Dump XIV alone will be discuss-
ed below, but the effect of the lower degree of pollution
is clear from Figures 45 and 46.

Referring to Figure 46, as indicated above, for all 83 re-
sults a correlation coefficient of 0.957 and a half-height
of 12.8 feet were found.  If the 28 results obtained in
Dump XIV are excluded, the remaining 55 observations have
a half-height of 14.3 feet and a correlation coefficient
of 0.981 for a line forced through the origin, and the
spread of data is clearly reduced.  Hence, the usual (week-
day) half-height is taken to be 14.3 feet, rather than 12.8
feet, and the half-height for satisfaction of the DO defi-
cit under weekday pollution conditions is 11.9 feet, accord-
ing to equation  (53), rather than the 10.6 feet given ear-
lier.

As noted earlier in Section VI, Dump XIV was conducted es-
pecially for study of gas transfer in the rapids section
above Station IP, as well as to obtain additional results
for the upper reaches of the Flint.  Hence, a relatively
large number of sampling stations was involved, and a lar-
ger than usual number  (n = 28) of observed results was
obtained.  Figures 47 and 48 show the 28 results for Dump
XIV taken above.  Figure 47 shows the relationship between
percent loss of tracer gas and water surface elevation
change, as predicted by equation  (50).  The data for this
                          196

-------
            100
VD
             80
          CNJ

          "60
          O
          _J
          in
          o
          LU
          O
          C£
          LU
          Q_
             20
              0
                                                          coo
                                             FIGURE 47
                                            FLINT RIVER
                                               25°C
                                           GAS LOSS VS,
                                          ENERGY EXPENDED
                                             DUMP XIV
               0
10        20        30        40         50        60
      WATER SURFACE ELEVATION CHANGE,  Ah,  FEET
70

-------
  100

  80


  60
  20
ce

CO
LU
Q-
   10
N&
\-k ,
o\V
°V
\



ALL
THA
DEE










- THE SHALLOW
POOL ABOVE
STATION

\
O X
\\

REACHES
T INCLUDE
P POOL, 2P








IP


<~\
O\J>
V \
THE \o








FIGURE 48
FLINT RIVER
25°C
GAS REMAINING VS ,
ENERGY EXPENDED
DUMP XIV


-(Ah)1/2
\ \
\ N
\ ^
\
\






= 11.7 FT

s.
\
\
\
o\
\












\
o\
               10         20        30        40         50
                 WATER SURFACE ELEVATION CHANGE,  Ah,  FEET
60
                              198

-------
single tracer study are highly consistent, again providing
powerful support of the energy dissipation models.

Figure 48 shows the same 28 observations for Dump XIV plot-
ted in the form required by equation  (49), namely, the
logarithm of the percent gas remaining vs. Ah, or energy
expended.  A regression line forced through the origin has
a high degree of correlation (0.991) and a half-height for
krypton of 10.4 feet.  However, upon closer examination of
the data certain finer details stand out and are of inter-
est.  As may be seen, the data array themselves in a set
of parallel lines all having the same slope associated with
a half-height of about 11.7 feet, still substantially lower
than the 14.3 feet corresponding to the usual weekday pol-
lution level.  The effect of the pools and falls in modi-
fying the reaeration capacity is clarified by Figure 48.

Referring to Figure 48, the main set of results describes
a straight line that passes through the origin, as expect-
ed from equation (49).  However, a separate set describes
a parallel line that is well apart, and below, the first
line.  This separate set of seven results represents all
of the study reaches that include the deep pool, reach
2P-2, below the second waterfall.  This same pattern is re-
produced in earlier Dumps  (II and III).  Specifically, the
data shown in Figure 48 indicate that for the reach 2P-2,
which has a Ah of 11.85 feet, the observed gas remaining
(33.4 percent, from the lower linelis considerably less
than the expected gas remaining  (50 percent, from the main
line, which passes through the origin).  In other terms,
the observed gas loss in the reach 2P-2 was 66.6 percent
instead of the expected 50 percent, and this is reflected
in all reaches that include the fall and pool.  In essence,
these results indicate clearly that gas transfer in the
deep pool below the second waterfall is enhanced by the
action that takes place as the result of the water falling
into the pool.  Whether the additional gas transfer is due
to air entrainment or simply to repetitive circulatory
vertical mixing in the pool cannot be determined from these
results, but the added effect on gas transfer is evident.

Although it is less obvious from Figure 48, another pool
effect appears to be identifiable in terms of tracer gas
losses.  Referring to Figure 48, the dashed line at the
top left represents as nearly as possible the long shallow
pool immediately above the first waterfall.  The three
points represent the reaches between the rapids section
immediately above this pool and the lower end of the pool,
Station IP.  These data appear to indicate that gas loss
in this pool was somewhat lower than it should have been--
                          199

-------
possibly due to poor mixing and some underflow in this
pool.  However, that conclusion must remain in doubt, as
the observable difference is not large and this particular
pool (reach R-1P)  was not studied specifically in other
tracer releases.

Figures 49 and 50 show similar curves for the main study
reaches, with a mean value shown for reaches having two or
more observed results.  All results for the main study
reaches are included, so that Figures 49 and 50 are compar-
able to Figures 45 and 46.  As may be seen, the fits are
excellent, and the spread of results has been reduced con-
siderably by the use of average values for each main study
reach.

The foregoing results and analyses provide a clear and pos-
itive demonstration of the validity and basic nature of the
energy dissipation models, equations (49) and (50), as re-
gards the Flint River studies.  The relationship predicted
by equation (48) between K   and the rate of energy expen-
diture, (Ah/tf) has also been subjected to statistical
analysis,  using the individual observed values of K    (cor-
rected to 25° C) and of (Ah/tf) provided in AppendijPAIV.
The results and their interpretation in terms of equation
 (48) are shown in Figures 51, 52 and 53.  These figures
show the results for values of  (Ah/tf)  less than 12.0, in
order to avoid the scale distortion that would result from
inclusion of the very high values (25.9 to 184)  associated
with the rapids section and the two waterfalls.   Instead,
for reaches having (Ah/tf) greater than 12.0, the observed
and predicted values of K   are shown in tabular form.
                         ox
Figure 51 shows the 74 results associated with a  (Ah/tf)
less than 12.0.  For these 74 observations, which do not
include the rapids or the two waterfalls, a correlation
coefficient of 0.931 was obtained, and a slope of 0.0635/
ft, for a regression line forced through the origin as re-
quired by equation (48).  The correlation coefficient for
a regression line with a nonzero intercept was also 0.931,
and the slope only slightly different  (0.0629/ft).

For all 83 individual observed results, which include the
rapids section above Station IP and the two waterfalls,
1P-1 and 2P-2, a correlation coefficient of 0.982 was ob-
tained, and the regression line slope was 0.0702/ft.  A
regression line forced through the origin had the same
correlation coefficient (0.982) and a slope of 0.0699/ft.

Referring to Figure 51, the data obtained from Dump XIV
have again been blacked in so as to make them stand out
                          200

-------
                100
                 80
NJ
O
M
          CO
          O
          CO
          <
          CD
          UJ
          O
          or
          UJ
          a.
60
                 20
                  0
                               o
                   o
                                             FIGURE 49
                                            FLINT RIVER
                                   GAS LOSS VS, ENERGY EXPENDED
                                         MAIN STUDY REACHES
                                     (MEAN VALUES BY REACHES)
                                               25°C
                   0
            10         20         30         40        50
              WATER SURFACE ELEVATION CHANGE,  Ah,  FEET
60

-------
CD


^
LU
CD
LO
O
LU
Qu
100
80
60.
40
20
10
8
6
4
^
^—_
\
\0
X















)
\
•\
















i O
\
0







FIGURE 50
FLINT RIVER
GAS REMAINING VS , ENERGY
EXPENDED
MAIN STUDY REACHES
(MEAN VALUES BY REACHES)
25°C

(Ah)1/2 =

\








12.8 FT

\
°\
0 \
\











\
\
\

       0
10        20        30         40         50

 WATER SURFACE ELEVATION  CHANGE/ Ah, FEET
60
                               202

-------
                    1,00
to
o
U)
                   0,80
                O

                LO
                CM
                o;
                ZD
                O
cc
Ul
CL

 X
 O
                   0,60
                    0,^10
                    0,20
                       0
                        0
4          6          8

 (r1)/ FEET PER HOUR
                                                              10
12

-------
on the graph.  Again, they are clearly different from the
rest of the results in regard to the slope of a line of
best fit, and demonstrate the effect on K   of reduced pol-
lution during Dump XIV.  When the 28 results obtained from
Dump XIV are eliminated from the total of 83, the remain-
ing 55 results have a correlation coefficient of 0.995, and
the slope of the regression line forced through the origin
is 0.0585/ft.  Four of these 55 results have a (Ah/tf)
greater than 12.0, and when they are eliminated,  the re-
maining 51 observations yield a correlation coefficient of
0.962 and a slope for the regression line forced through
the origin of 0.0583/ft.  This latter line is shown as a
dashed line in Figure 51.  As outlined earlier, it repre-
sents the usual (weekday) pollution condition of the Flint
River during these studies.

Figure 52 shows the relationship between K   and the rate
of energy dissipation for the data of Dump XIV taken alone,
as required by equation  (48).  For the 23 observed results
associated with a (Ah/tf) less than 12.0, the correlation
coefficient is 0.978 and the slope 0.0778/ft for a regres-
sion line forced through the origin.  For the regression
line with a nonzero intercept the correlation coefficient
of 0.982 is only very slightly different.  For values of
(Ah/t,-) greater than 12.0, the five additional results are
tabulated in Figure 52.  As may be seen, the correspond-
ence between observed and predicted values of K   is very
good.  The reaches 1P-1 and 2P-2 are the two wa?lrfalls,
and the reaches R1-R2, R2-R3 and R1-R3 represent the rapids
section above Station IP.  In agreement with the results
shown earlier in Figure 48, for the reach 2P-2 the predict-
ed value of K   is considerably lower than that observed,
indicating the additional effect of the waterfall into the
deep pool.  Also, the dashed line at the lower left of
Figure 52 shows the same three observations thus designated
in Figure 48, representing the shallow pool above Station
IP.  It should be noted also that the slope 0.0778/ft is
considerably greater, for this less polluted condition,
than that presented in Figure 51.

For all 28 results obtained from Dump XIV, and a line of
best fit forced through the origin, the correlation coeffi-
cient for the relationship between K   and  (Ah/t,.) , equa-
tion  (48), is 0.995, and the slope o?xthe line ii 0.0803/
ft.  For the regression line with a nonzero intercept, the
corresponding values are 0.995 and 0.0809/ft, respectively.

Figure 53 shows the relationship between K   and the rate
of energy expenditure for values of  (Ah/tf?Xless than 12.0
with all of the data averaged by main study reaches, and is
                          204

-------
              1,00
to
o
Ul
              0,80
              0,60
ID
O
n:

DC


S   0,40
 X
 o
              0,20
FIGURE 52
FLINT RIVER
25°C
KQX VS, RATE OF
ENERGY EXPENDITURE
DUMP XIV








t
^
>
>^
jrrf^
/£f






q/C
/
. ^r
S^
£
fS
0 	 SHALLOW
ABOVE £










0
J)S
j*





1 POOL
TATION IP






>
/
O ^
r&S
O>r
S o
o












o >
/
/
^\_


y
jS
'


0.978








xy
SLOPE = 0.0778/FT





^h,


K
ox
REACH vt^' OBS. PRED.

2P-2
25.9 2.89 2.01 "
R2-R3 35.4 2.31 2.75
R1-R3 37.9 2.39 2.94
R1-R2 41.4 2.50 3.22
1P-1
184. 15.1
14.3
                   0
                                                   8
10
12
                                              TT-)/  FEET PER HOUR

-------
              1,00
NJ

O
CSl

 X

cc


o
o;
LU
0.

 x
 o
              0,80
             0,60
             0,40
             0,20
                0
FIGURE 53
FLINT RIVER
25°C
KQX VS, RATE OF
ENERGY EXPENDITURE
MAIN STUDY REACHES
(MEAN VALUES BY REACH)



SL

JP
/
/
,


OPE = 0.06
S
/%










.,
f\S
25/FT If
^









o
^ o

D









^



REACH





O .S
Cj



, Ah,









K
ox
tf DBS. PRED.
2P-2 29.7 2.61 1.86
R1-R3 38.2 2.40 2.38
1P-1 1
32. 12.7 11.4
                  0
                                                    8
10
12
                                              ^r-)/  FEET PER HOUR
                                               f

-------
thus comparable to Figure 51.  The three study reaches
having larger values of  (Ah/tf)  are again shown in tabular
form to avoid scale distortion.   The agreement between the
observed and predicted values of K   for the first water-
fall (1P-1) and for the rapids secHfon  (R1-R3) is excel-
lent; as before, the observed K   for the second waterfall
and deep pool is considerably larger than the predicted
result, the predicted result being only 71 percent of the
observed value.  The slope of the fitted line, 0.0625/ft,
is quite comparable to the slope of 0.0635/ft shown in
Figure 51.

The foregoing results from the Flint River studies by them-
selves provide clear and conclusive evidence of the valid-
ity and the fundamental nature of the energy dissipation
models for reaeration capacity, equations (48),  (49) and
(50).  As indicated by the statistical tests, the relation-
ships involved are very strong, and the correlation coef-
ficients are fully as high as could be expected in terms
of the range of experimental error that must be anticipated
in such field operations.  As indicated in Section VI, the
Flint contains the widest range of hydraulic features, and
the most turbulent sections, of any of the rivers studied,
and, therefore, the widest range of observed reaeration
results.  Under these circumstances, the foregoing results
and statistical tests provide ample support of the energy
dissipation theory.

South River.  As reported earlier in Section VI of this re-
port, the South River was the most polluted stream studied
during these investigations.  Four separate waste treat-
ment facilities discharge into the South within the tracer
study section.  The two  larger plants, the South River STP
and the Intrenchment Creek STP, each with a usual effluent
flow of about 14 mgd, release their effluents just above
Station A and below Station B, respectively.  The Shoal
Creek STP usually releases about 3 mgd just below Station
F, and the Snapfinger Creek STP usual effluent flow of
about 2.5 mgd is released in the immediate vicinity of Sta-
tion H.

In addition to this multiple loading pattern, the records
of the Georgia Water Quality Control Board  (WQCB) show that
during the 1969 tracer study period the South River STP by-
passed substantial quantities of sewage flow directly to
the South River several times; the Intrenchment Creek STP
bypassed only a small part of its flow for brief periods
during Dumps VIII and IX; the Shoal Creek STP bypassed
about half of its flow fairly frequently during the tracer
study period while flooding filters; there are no records
                         207

-------
of bypasses at the Snapfinger Creek STP.

The Georgia. WQCB records also show, from occasional samples
taken during the tracer study period, that the 5-day BOD of
the South River STP effluent ranged from 20 to 43 mg/1  (6
observations); the Intrenchment Creek STP effluent had  a 5-
day BOD of 26  to 48 mg/1 (3 observations); the 5-day BOD of
the Shoal Creek STP effluent ranged from 13 to 16 mg/1  (3
observations); the effluent from the Snapfinger Creek STP
contained 15 and 37 mg/1 of 5-day BOD on two observations.

Initial analysis of the gas transfer data obtained in the
South River tracer studies of reaeration indicated that
there were marked differences in reaeration capacity from
one dump to another and from one section to another, even
though the results from each individual dump were consis-
tent among themselves.  For example, as noted earlier in
Section VI, Dumps VI and X covered the same section of
stream (Stations A to F),  and both dumps took place at the
same river flow and temperature.  Yet the reaeration results
(Kox) were consistently lower in Dump X than in Dump VI.
Also, the initial analysis showed that gas loss in the upper
reaches (above Station G or H)  tended to be consistently
lower per foot of elevation change than in the lower reaches,
even though the semilog plot of gas remaining vs. Ah ap-
peared to be good for the  data from each individual dump.
Accordingly, the records of the Georgia WQCB were examined,
as noted above, and the available results of analysis of
river samples  for 5-day BOD were also studied.

The foregoing  analysis of  the pollution situation during
each tracer dump indicated clearly that for reasonable con-
sistency as regarded the degree of pollution the South River
tracer study results had to be broken down into three groups.
The first and  largest group of results is that for the lower
study reaches—all reaches below Station H; these reaches
are the farthest removed from the two larger waste treatment
plants (South  River and Intrenchment Creek), and the avail-
able records showed relatively low river BOD's (6.0 to  10.0
mg/1, 5-day).   Thus, the lower study reach reflects the
lowest degree  of pollution.  The second data group is that
for the upper  study reaches—all reaches above Station H.
This group, representing the reaches just below the two
larger waste treatment plants, reflects a somewhat higher
usual degree of pollution, with river 5-day BOD's in the
range 10 or 12 to 15 or 18 mg/1.  A third group of results
also stands out, consisting of all of the results for Dumps
VIII and X.  This group is associated with the highest de-
gree of pollution, namely river 5-day BOD's of 20 to 30
mg/1.  Specifically, Dump X was associated with river 5-day
BOD's of 20 and 31 mg/1 at Panola Shoals.  No river BOD's
                           208

-------
are available for the exact date of Dump VIII, but the
Georgia WQCB records show that on the preceeding day a large
bypass (almost 7 hours long) occurred at the South River
STP;  Dump VIII was made at Station J, some 18 hours time of
flow below the South River STP, hence was directly affected.
In addition, a smaller bypass occurred at the Intrenchment
Creek STP.

The following figures illustrate the observed results for
the South River tracer studies, and the effects of the degree
of pollution associated with each release.  In reviewing all
of the data contained in Appendix AIV, it was also concluded
that the results for Station K in Dumps IX and XI should be
rejected because of the very small amounts of krypton-85 re-
maining—even though the counting results appeared initially
to be acceptably consistent, the actual counts were not
greatly above the background count rate.  It is also noted
that, referring to Table 6 in Section VI, these releases con-
tained the lowest quantities of tracer gas initially.

Although single graphs containing all of the 87 observed re-
sults from the South River studies are not provided here,
statistical testing was conducted.  Referring to equation
(49), which relates the logarithm of the percent gas remain-
ing to the amount of energy expended, Ah, a correlation co-
efficient of 0.909 was obtained, indicating a strong rela-
tionship for the 87 observations; the slope obtained, Co/
was -0.0383, indicating a half-height,  (Ah)]y2' of 18.1 feet
for krypton.  These results were for a regression line
forced through the origin, and again were not significantly
lower than the results for the unconstrained line of best
fit.

Regarding equation (48), the model that relates Kox to the
rate of energy dissipation,  (Ah/tf), the correlation coeffi-
cient found for the 87 observations was 0.971, and the slope
of the regression line forced through the origin was
0.0538/ft.  For the regression line with a nonzero intercept
the corresponding results were 0.974 and 0.0559/ft, respec-
tively.

The foregoing results taken alone provide quite adequate
support of the energy dissipation models developed earlier.
However, as noted above, the data are more suitably separa-
ted into three groups that reflect the actual conditions of
pollution.  When this is done, the relationships predicted
by the energy dissipation models are better clarified, and
the effects of pollution on the reaeration capacity of a
stream become evident.

                           209

-------
Figures 54, 55, and 56 show the relationship between percent
loss of tracer gas and energy expended (Ah), as predicted by
equation (50).  Figure 54 represents the least polluted con-
dition, the lower study reaches; Figure 55 reflects the us-
ual pollution condition in the upper study reaches; Figure
56 includes those data reflecting the most polluted condi-
tion, namely,  Dumps VIII and X.  The strength of the rela-
tionship is obvious in each Figure.  (In Figure 56, four
points that involve Station J stand out;  they have been ig-
nored in fitting a smooth curve to the remainder of the ob-
served results, for reasons that will be explained below).

It is clear from Figures 54, 55, and 56 that although the
basic energy dissipation model, equation (50) , provides an
excellent fit of the observed results in each case, with in-
creasing pollution there is less gas transfer.  For example,
for a water surface elevation change of 30 feet, under the
least polluted condition (Figure 54) 75 percent gas loss
occurred; with more usual levels of pollution (Figure 55) 65
percent gas loss occurred;  with the heaviest pollution  (Fig-
ure 56) only 54 percent gas loss occurred.

The same observed results have been replotted in Figures 57,
58 and 59 in the form required by equation (49)  and to clar-
ify the effects of pollution.  Figure 57 shows the semilog
plot of percent gas remaining vs. Ah for the lower study
reaches.  These are the same 44 observations shown earlier
in Figure 54.   The straight line fit of the data is excel-
lent, and a correlation coefficient of 0.979 has been cal-
culated for these results with the regression line forced
through the origin.  The slope of this regression line was
C0 = -0.0452/ft, and, from equation  (51), the resulting
half-height was (Ah)jy2 =15.3 feet for krypton.  From equa-
tion  (53), the comparable half-height for satisfying the DO
saturation deficit is 12.7 feet.

Figure 58 includes the same 23 observations shown earlier in
Figure 55 for the upper study reaches of the South, and rep-
resents a more usual upstream pollution condition.  As may
be seen, again  the fit of data by a straight line is excel-
lent and provides powerful experimental support of the ener^
gy dissipation model, equation  (49).  The correlation coeffi-
cient for these results, with a regression line forced
through the origin, was 0.992, and the slope was Co =
-0.0346/ft.  From equation  (51), this leads to a half-height,
(Ah)l/2<- of 20.0 feet for krypton, or 16.6 feet for DO satu-
ration deficit.

Figure 59 shows a similar plot of the 20 observations asso-
ciated with the heaviest pollution condition  (see also
                              210

-------
NJ
\->
H
                o
                LT|
                CSJ
                CO
                CO
                O
CO
<
CD
                UJ
                O
                ct:
                LU
                Q.
                                                   FIGURE

                                                  SOUTH  RIVER

                                                     25°C

                                                 GAS LOSS  VS,

                                                ENERGY  EXPENDED
                                                              LOWER  STUDY  REACHES
                                                                (LEAST  POLLUTION
                     0
                                  WATER SURFACE ELEVATION  CHANGE/  Ah, FEET

-------
to
h-1
KJ
                 o
                 i_n
                 CNl
                  CO
                  CO
                  o
CO
<
<£>
                  CJ
                  o:
                  LU
                  Q_
1UU
80
60
40
20
0
(



v
L


/
"/
/°



#



S&





^






FIGURE 55
SOUTH RIVER
25°C
GAS LOSS VS,
ENERGY EXPENDED
UPPER STUDY REACHES
(USUAL POLLUTION)
} 10 20 30 40 50 6(
                                    WATER SURFACE  ELEVATION CHANGE/  Ah, FEET

-------
K)
M
CO
             CNl
              CO
              CO
              O
              CO
              O
              o:
              UJ
              Q_
   DUMP VIII
  ALL REACHES
WITH STATION J
  (SEE TEXT)
                                          FIGURE 56
                                         SOUTH RIVER
                                            25°C
                                        GAS LOSS VS,
                                       ENERGY EXPENDED
                                       DUMPS VIII S X
                                    (HEAVIEST POLLUTION)
                              10        20        30         40         50
                               WATER SURFACE ELEVATION  CHANGE/  Ah,  FEET

-------
NJ
                 QJ
                 ct:

                 CO
                 LU
                 o
                 o:
                 LLT
'JU
80
60
40
20
^
ooN?
\
>x



(t




V
O fckft
°HU
oo^






\
^Jyp
00\
^h)1/2 = 15.3 FT O>
r = 0.979 CP
xy








FIGURE 57
SOUTH RIVER
GAS REMAINING VS ,
ENERGY EXPENDED
LOWER STUDY REACHES
(LEAST POLLUTION)
25°C


\
"\



X




                       0
10        20         30         /IO         50

 WATER SURFACE ELEVATION  CHANGE,  Ah, FEET
60

-------
                  100


                   80



                   60
N)

M
U1
LU
o:

CO

to

I—
•z.
UJ

cc

m   on
o-   20
                   10
\
>>
m









T<2L o
O^^k_
N


(Ah)
x"






>JD
r\ ^^k.
" \
1/2 - 2°-°
= 0.992
f







FIGURE 58
SOUTH RIVER
GAS REMAINING VS ,
ENERGY EXPENDED
UPPER STUDY REACHES
(USUAL POLLUTION)
25°C

X
FT ^«v
oV



X
\



s
                     0
                 10         20         30        40         50

                  WATER SURFACE  ELEVATION CHANGE/  Ah,  FEET
60

-------
                   100
                    80
                    60
to
H-1
LU
cr

to
<
CD
                LU
                   20
                    10
\^
V

X


I
AI
WITI
( c
v>-


^^QS/I

\

DUMP VIII -
jL REACHES
I STATION :
3EE TEXT)


-(Ah)
rx]
L/2 =
= 0
"N^
^^
u/ °-J
r

FIGURE 59
SOUTH RIVER
GAS REMAINING VS ,
ENERGY EXPENDED
DUMPS VIII & X
(HEAVIEST POLLUTION)
25°C




26.3 FT
.998


s ^S^^








x^
"\










                      0
                10         20         30        40         50

                 WATER SURFACE ELEVATION CHANGE/  Ah,  FEET
60

-------
Figure 56).   Here, the discrepancy involving Station J be-
comes evident.  The main set of data, excluding reaches that
involve J, again is fitted by a straight line through the
origin, and the fit is excellent.  The four observations in-
volving J have been fitted by a parallel line (dashed), but
it is displaced substantially below the other.  There are two
possibilities regarding the cause of this discrepancy.  The
reach J-K includes the long deep pool above Panola Shoals,
and it will be recalled  (Section VI) that mixing through
this pool was found to be very erratic—sometimes the flow
entering the pool plunges and stays underneath the main pool
volume, whereas at other times the entering flow stays at
the pool surface and does not mix with the main pool volume.
Hence, one possible explanation of the discrepancy shown in
Figure 59 is that the flow from J stayed at the surface
through the pool, resulting in greater gas loss than would
have occurred if the flow had mixed fully with the pool vol-
ume.

The second possible explanation of the discrepancy involving
Station J is that the dose assay was in error, and this is
regarded as the most likely explanation.  This source of
error was discussed briefly in Section V.  In the case of
Dump VIII, the dose point was at Station J.  Also, the
specific discrepancy noted in Figure 59 did not occur in
other tracer studies across these reaches.  Hence, in brief,
it appears most probable that in this case the discrepancy
represents an error in dose assay, possibly involving loss
of tracer gas during the assay procedure.  The four results
involving Station J have therefore been excluded from the
subsequent analysis.

For the remaining 16 results shown in Figure 59, a correla-
tion coefficient of 0.998 was obtained for a regression line
forced through the origin.  The slope of the line was
-0.0263/ft, yielding a half-height of 26.3 feet for krypton
and 21.8 feet for DO saturation deficit.

The foregoing analysis of the South River data not only pro-
vides the most powerful of support for the energy dissipa-
tion models, equations  (49) and  (50), but also provides con-
siderable insight regarding the effects of pollution on gas
transfer and reaeration capacity.  The clearly observable
decrease in the half-height for gas transfer with increased
pollution shows that the reaeration capacity of a stream can
be sharply reduced by the presence of untreated or partially
treated domestic sewage.  It should be noted in this regard
also that this reduction in reaeration capacity cannot be
laid exclusively at the door of the detergent LAS—the few
available results obtained later, taken together with the
                           217

-------
laboratory studies referred to in Section VII, indicate
strongly that observed river concentrations of LAS  (in the
neighborhood of 2.0 mg/1)  should not cause so sharp a change
in reaeration capacity.

Figures 60, 61 and 62 show the relationship between the ob-
served values of Kox (25°C) and (Ah/tf) for the three groups
of data, as required for testing equation (48).  Figure 60
includes the 44 observations for the lower study reaches,
three of these results being shown in tabular form to avoid
scale distortion.   As may be seen, the relationship between
KQX an<3 tne rate of energy expenditure is very strong and
provides excellent support of the energy dissipation model,
equation (48).  The agreement between observed and predicted
values of Kox for Panola Shoals, reach K-L,  as shown in the
table, is excellent—the mean observed value of Kox was
3.49/hour vs. a mean predicted value of 3.29/hr.  For a re-
gression line forced through the origin the correlation co-
efficient is 0.995, and the slope 0.0595/ft for the 44 ob-
servations.  If the analysis is limited to the 41 observa-
tions having (Ah/tf) less than 8.0, the corresponding re-
sults are 0.949 and 0.0556/ft, as shown in Figure 60.

As may also be seen from Figure 60, the observed results are
somewhat more erratic at quite low values of  (Ah/tf), espe-
cially for values less than 1.0.  No special effort was made
in the South River studies to achieve the greatest possible
accuracy for reaches where gas loss was likely to be very
low, such as the reach J-K (the pool above Panola Shaols).
As a result, for such occasional reaches the experimental
error involved in laboratory analysis for tracer concentrar-
tions was somewhat larger than usual, and the observed re-
sults more erratic.  For example, Figure 60 shows two values
of K0x to be zero  (at Ah/tf = 0.65 and 1.61); in both cases,
this is because the observed gas loss was essentially zero
(see also Figure 54) within the reach J-K.  In later studies
of the Chattahoochee River reaeration capacity, a higher de-
gree of accuracy was obtained by counting larger numbers of
samples for longer counting periods.

Figure 61 is a similar plot of the observed data for the
upper study reaches, reflecting the intermediate level of
pollution.   Again, the relationship between Kox and  (Ah/tf)
is very strong, and the results for higher values of
(Ah/tf) shown in the table also reflect good agreement be-
tween observed and predicted values of KQX.   For the 17 ob-
servations having  (Ah/tf)  less than 10.0, the correlation
coefficient for the line forced through the origin was 0.959,
and the slope 0.0414/ft.  It may be noted that, as expected,
the slope has been reduced, compared to Figure 59, for the
                           218

-------
              0.60
              0,50
NJ
         cxi
         cr
         LU
         Q_
          X
          o
              0,30
               0.20
               0.10
FIGURE 60
SOUTH RIVER
25°C
Kox VS RATE OF
ENERGY EXPENDITURE
LOWER STUDY REACHES
(LEAST POLLUTION)











o
o
./
































SLOPE = 0.0556/FT
r = 0.949 0



^

/^


AY

^.S
/^°
u



S6
./" O
&!& o















o
o ^
o 0$*^°

XJ^
09
















o /
jy*u
<*&
















*T








(








S"

-------
              0,60
to
NJ
O
              0,50
              0,40
           un
           Csl
           cc
           ID
           o
           LU
           a.
0,30
            x
            o
              0,20
              0,10
FIGURE 61
SOUTH RIVER
25°C
KQX VS, RATE OF
ENERGY EXPENDITURE
UPPER STUDY REACHES
(USUAL POLLUTION)




/



/
/




a
?(°




y
r



y
,/SLO
/

REA
fi-H
G-H
G*-
G-H
G*-

O/
PE = 0.041
= 0.959
xy

CH DUMP


4/FT

<^>




K
OX
f OBS . PRED .
IX 11.3 0.48 0.47
XI 12.1 0.59 0.50
H XV 12.4 0.53 0.52
VII 17.7 0.92 0.73
T2 XV 23.9 1.28 0.99
i i i
   8          10

FEET PER HOUR
                                                                                  12
                                                                                          16

-------
to
NJ
            C_5
            LO
            CNI
             cc
             Z3
             O
ce
U)
Q.
                0.30
                0.25
                0.20
                0.15
              X
              o
                0.10
                0.05
FIGURE 62
SOUTH RIVER
25°C
KQX VS. RATE OF
ENERGY EXPENDITURE
DUMPS VIII* & X
(HEAVIEST POLLUTION)




y^ *S1




AT I ON J OM





S 0
ITTED-SEE

ST.

^

TEXT

DPE - 0.03
r =0.99
xy
0
;/



1 R/FT
XJ O


REAC
B-D
K-L

/
s'

H DUMP
X 9
VIII 43

s' o


Ah^ K
fcf OBS .
.00 0.29
.0 1.37
1




OX
PREP.
0.28
1.34
                                                                4          5

                                                           (f1),  FEET PER HOUR

-------
increased level of pollution.   For all results, the corres-
ponding values were 0.983 and 0.0479/ft, respectively.

Figure 62 shows the same relationship for the 16 observed
results representing the heaviest pollution condition  (Sta-
tion J, Dump VIII, has been deleted, as explained earlier).
The correlation coefficient for the regression line forced
through the origin was 0.999,  and the slope 0.0318/ft.  This
slope reflects a further reduction in gas transfer and reae-
ration capacity associated with additional pollution.

The South River data thus also provide the most powerful of
support of the energy dissipation models for reaeration that
were developed earlier.  In addition, because of the multiple
pollution sources and variable pollution load in the stream,
these studies have provided considerable insight regarding
the actual effects of pollution on stream reaeration capac-
ity.

Patuxent River.  As indicated earlier in Section VI, the tra-
cer studies of reaeration capacity of the Patuxent River were
conducted in the early fall of 1969 for the Maryland Depart-
ment of Water Resources  (DWR).  The gaseous tracer studies
were conducted by the Georgia Tech research staff, with the
exception of Dump E, whereas the Maryland DWR personnel per-
formed the field physical studies, including measurement of
stream flow, etc.  The surveys of water surface elevation,
cross-sectional area, etc., were conducted considerably
later than the field tracer studies, and at a substantially
higher flow.  The available data permit adequate adjustment
to average flows generally prevailing during the earlier
tracer studies, but are not sufficiently detailed to allow
adjustments as accurate as those that can be made for the
Flint, South and Chattahoochee studies by the procedures
outlined in Section V.

Also, as noted in Section VI, stream flow fluctuated sub-
stantially from one dump to the next during the Patuxent
River tracer studies, and, in some cases, during the course
of a single dump.  As a result, the physical study data,
while adequate, cannot be taken to be as accurate as those
for the other stream tracer studies.  This refers especially
to the water surface elevation change between sampling sta-
tions during the actual tracer studies.  Time of flow  is
measured routinely during the tracer studies, and those data
are as accurate as usual, as are the gas loss results  and
reaeration coefficients.  Although the Patuxent River  stud-
ies were not a part of the research  sponsored under  this
project, they provide additional insight into the relation-
ships between energy dissipation and reaeration, and into
the effects of pollution.
                          222

-------
Figure 63 shows the percent loss o,f tracer gas corresponding
to energy expenditure  (Ah) for all of the Patuxent River re-
sults (n = 30).  The observed results are somewhat more
variable than in the other stream studies, but the trend of
results is quite evident and fits the energy dissipation
model, equation (50) very well.

The same data have been replotted in Figure 64 in the form
required by equation  (49), namely the logarithm of the per-
cent tracer gas remaining vs. energy expended, Ah.  As may
be seen, the data describe the expected straight line very
well.  For a regression line forced through the origin,
these results have a correlation coefficient of 0.942, in-
dicating a very strong relationship.  It should be noted
also that the slope of the regression line, as represented
by (Ah)1/2, is steeper than the slopes found for the Flint
or the South River data.  The Patuxent had only one source
of pollution, a highly treated domestic waste, and the
(Ah)1/2 of 10.5 feet for krypton clearly indicates the low
degree of pollution of the Patuxent as it passes through
the Wildlife Refuge.  In contrast, the lowest values of
(Ah)1/2 observed for the Flint and South Rivers were 11.7
feet  (Dump XIV, Flint River) and 15.3 feet  (South River,
Lower Study Reaches), and a result as high as 26.3 feet was
found for the heaviest pollution  (Dumps VIII and X) in the
South.

In terms of overcoming an existing DO saturation deficit,
use of equation (53) indicates that the Patuxent below
Laurel, Maryland, has a half-height of 8.7 feet, the lowest
result observed in any of these studies.  Reference to
Table 12, Section VI, indicates that the water surface ele-
vation of the Patuxent falls about 7.4 feet per mile, or
8.7 feet in 1.18 miles, at low flows.  Hence, under the con-
ditions of flow and pollution load encountered during the
1969 studies, at 25°C reaeration will provide DO income to
the extent of 50 percent of the prevailing DO deficit about
every 1.18 miles.

BOD analyses performed by the Maryland DWR on samples taken
during the reaeration tracer studies clearly showed that
most of the BOD present below the Laurel STP was second
stage demand, rather than carbonaceous, and that pollution
levels were low.  BOD time series were obtained for 12 sam-
ples collected during the survey period and represented
each of the tracer study sampling stations at least once.
In each case a 10-day BOD curve was developed.  Only two of
the 12 samples displayed a distinct first stage lasting 3
or 4 days before the onset of nitrification.  The remaining
10 samples had virtually no first stage BOD, but only sec-
ond stage.  The 5-day BOD for all samples ranged from about


                          223

-------
NJ
to
       LP\
       CNl
       in
       o;
       LU
       Q_
             0
10         15        20         25         30

WATER SURFACE ELEVATION  CHANGE,  Ah, FEET

-------
              100


               80



               60
to
tvj
Ul
LU
cc

CO

CD

I-
z
LU

cc
LlJ
Q.
               20
                10
\
\
\
c
c


(




•v
^v
o°^X(
{J ^J
Ah)1/2 = 1
r - 0.94
xy





*
V
>v U
g\S
8 ^
0.5 FT ^
2
C







\







FIGURE 64
PATUXENT RIVER
25°C
GAS REMAINING VS.
ENERGY EXPENDED
ALL RESULTS


NO
IN^
^*



O



\
                  0
                            10         15         20         25         30

                      WATER SURFACE ELEVATION  CHANGE/ Ah,  FEET
35

-------
4.0 to 9.0 mg/1,  and averaged 6.8 mg/1.  Thus, compared to
the other streams studied, the Patuxent was clearly the
cleanest, and, hence, the half-height for krypton transfer
was the lowest value observed in the five streams".

Figure 65 shows the plot of Kox at 25°C vs. (Ah/tf), as re-
quired by equation (48).  As may be seen, the range of ob-
servation of both KQX and (Ah/tf) is very small, and the re-
sulting correlation coefficient for a line forced through
the origin is only 0.245.  The Patuxent was characterized by
a relatively high degree of hydraulic uniformity, as noted
earlier, leading  to such a small range of observation of Kox
and (Ah/tf)  that  this particular statistical test is not as
meaningful as in  the cases cited earlier.  In essence, the
range of total experimental error is relatively large in
this case compared to the range of observation.  For example,
for the regression line with a nonzero intercept the corre-
lation coefficient was substantially greater  (0.559), the
intercept was at  Kox = 0.066/hr, and the slope was only
0.0430/ft.

The theory,  equation (48), requires that the regression line
go through the origin,  and, hence, that is the line shown in
Figure 65, with the corresponding low degree of correlation.
The resulting slope of 0.0804/ft is then steeper than that
found for any other set of results except those for Dump
XIV in the Flint  (0.0803), which is just as expected as a
result of the low degree of pollution.

Reference to the  results provided in Section VI and Appen-
dix AIVprovides some further insight into the range of ex-
perimental error.  In brief, the data show that in terms of
Kox and percent gas loss the observed results were satisfac-
torily reproducible for most reaches from one dump to an-
other.  The times of flow showed more variability, as one
indication of the fluctuating river flows that occurred
during the tracer studies.  As indicated earlier, due to
short-term fluctuations in stream flow during the tracer
studies, as well  as water surface elevation measurements
made at another time and a considerably higher flow, the Ah
values reported here are also not as stable or dependable
as usual.  This combination of short-term flow fluctuation
and error of measurement leads to a variability of perhaps
as much as 0.5 feet in any reported value of Ah; the times
of flow in any reach vary by as much as 1.0 hour or more
due.to.short~term flow changes.  The net effect of such
variability is still minimal as regards the results presen-
ted in Figures 63 and 64, because of the wide range of ob-
servation of percent gas transfer and Ah, and so a high
correlation still results.  However, the net effect on the
                           226

-------
to
NJ
-J
              0,25
              0,20
           o
           LO

           0X1
              0,15
cr
LU
            X
            o
              0,10
              0,05
                 0







/
/
/





0
/
/
/


SLOPE
r =
xy
O
O
j/°£
/




0
c
- 0.0804/1
= 0.245
y/
Q S
o JD
O
)






,T >x

o 0
o
o







/

O
O









FIGURE 155
PATUXENT RIVER
25°C
KQX VS, RATE OF
ENERGY EXPENDITURE
ALL RESULTS
                   0
                 0,5
1,0
1,5
2,0
2,5
3,0
3,5
                                              t~)/ FEET PER  HOUR
                                              rf

-------
small quotient (Ah/tf)  is large, and, coupled with the  small
range of observation of Kox and (Ah/tf),results in the  low
degree of correlation.

Thus, the half-height of 10.5 feet for krypton, and the as-
sociated high degree of correlation are as expected, and
provide excellent support of the energy dissipation models,
equations (49) and (50).  The slope of 0.0804/ft obtained
for equation  (48) from Figure 65 is also regarded as a  firm
and accurate result, despite the low apparent correlation—
in this case, the slope is determined on the basis of two
firm points, namely, the origin and what amounts to a mean
of the observed results.  These results for a stream that
has very little first stage BOD provide excellent additional
evidence, compared to the Flint and South, of the effects of
the presence  (and absence) of pollution on reaeration capac-
ity.

The results presented thus far now lead to two more general
observations regarding the relationship between reaeration
capacity and stream hydraulic properties.  In the first
place, it is clear that, of the three energy dissipation
models developed earlier, the relationship between gas trans-
fer and Ah is both simpler and more accurate and dependable
than the relationship between Kox and  (Ah/tf).  The former,
represented by equations  (49) and (50) tends to minimize
errors of measurement,  whereas the latter, represented by
equation (48), tends to magnify such errors.  Thus, the de-
gree of correlation found in tests of equation (49) is al-
ways high,  whereas the combination of experimental error and
a small range of observation can lead to a larger error of
prediction when using equation  (48).  Equation (49) involves
the total amount of energy expenditure, while equation  (48)
involves the rate of energy expenditure, and, hence, just as
with any other kinetic analysis, the former is much less
sensitive to errors of measurement than the latter.  Thus,
equation (49) is the preferred energy dissipation model for
testing and subsequent application.

The results presented thus far also now lead to what appears
to be a quite good estimate of the reaeration capability of
a clean stream, measured in terms of the fundamental half-
height for gas transfer.  From these results, the range of
half-heights for krypton transfer is quite small  (10.5  feet
in the Patuxent to 26.3 feet in the polluted South).  Al-
though no observations have been made in a stream that  has
never been polluted, the evidence now available indicates
that a half-height less than, say, 8.0 feet for krypton
transfer is quite unlikely.  The matter of predicting reaer-
ation from water surface elevation change, and application
                          228

-------
in oxygen sag computations, will be discussed in greater de-
tail at a later point in this report.

Jackson River.  The tracer studies of the reaeration capac-
ity of the Jackson River below Covington, West Virginia,
were conducted in 1966, and were not a part of this research
project.  As indicated in Section VI, water surface eleva-
tions were not measured during the field tracer study peri-
od, but were obtained later  (1970) by others.  Also, although
an average flow above Covington was obtained during the tra-
cer study period, stream flows were not measured at key lo-
cations in the study section during each individual tracer
release.  Thus, for these data it is not now possible to
adjust water surface elevation changes to individual flows
for each tracer release by the procedure outlined in Sec-
tion V.  Hence, the values of Ah provided in Appendix IV are
not quite as accurate as those obtained for the Flint and
South, and only one value of Ah is given for each stream
reach.

Also, as noted in Section VI, immediately before the begin-
ning of the tracer studies in 1966 the secondary treatment
units failed at the Westvaco paper mill, and the treatment
plant operated only as a primary plant during the first two
tracer studies.  As a result, the pollution load in the
Jackson was unusually great, especially in the reaches 0-1
and 1-2, and zero DO's were observed in that vicinity.

A total of 80 observations of reaeration capacity were made
during the 1966 tracer studies, and the data are provided in
Appendix AIV. Figure 66 shows all of the individual results,
plotted in the form required by equation (50) , the energy
dissipation model relating percent gas loss to water surface
elevation change, Ah.  As may be seen, the results describe
a curve of the predicted form, and the fit of data is excel-
lent.  These same 80 results have been plotted in Figure 67
in the form required by equation  (49) , namely the logarithm
of the percent gas remaining vs. water surface elevation
change.  The straight line fit of the data, as predicted by
equation (49) , is excellent.  With a regression line forced
through the origin the correlation coefficient is 0.963, as
shown in Figure 67.  The slope of this regression line is
C0 = -0.0444/ft, resulting in a half-height for krypton of
15.6 ft.  This is essentially identical with the results for
the lower reaches of the South River.

Thus, the gas transfer results obtained in 1966, together
with the water surface elevation changes obtained by others
in 1970, provide strong support of the energy dissipation
models, equations  (49) and  (50).
                           229

-------
to
U)
o
                 LO
                 CsJ
                 CO
                 CO
                 o
                 CO
                 LU
                 C_)
                 o:
                 L±J
                 Q_
                                                                    FIGURE 66

                                                                  JACKSON RIVER

                                                                      25°C

                                                                  GAS LOSS VS,

                                                                 ENERGY EXPENDED

                                                                   ALL RESULTS
                                 10        20         30        40         50

                                  WATER SURFACE  ELEVATION CHANGE, Ah,  FEET

-------
NJ
CO
IUU
80
60
0 40
^
z
<
2:
LU
ce
CO
CD
£ 20
LU
Cd
LU
Q-
10
8
6
V

>.
N^t
Sr^y Cft



(





*V
s80"^
C
C
Ah)1/2 =
r =0.9









O
L O Q
X
i
15.6 FT
63











\ 8
0 o










FIGURE 67
JACKSON RIVER
25°C
GAS REMAINING VS ,
ENERGY EXPENDED
ALL RESULTS



\




0 10 20 30 40
WATER SURFACE ELEVATION CHANGE



\












50 60 7(
, Ah, FEET

-------
Figure 68 shows the results for the Jackson River in the
form required by equation  (48) , which is the model relating
Kox to the rate of energy dissipation.  The range of obser-
vation is larger than for the Patuxent, but still much
smaller than for the Flint or South, and as in those earlier
cases, it is evident that this relationship is more sensi-
tive to transitory flow fluctuations and errors of field
measurement.  Certain of the observed results have been
identified in Figure 68 to illustrate this sensitivity.

Referring to Figure 68, the two results for the reach 0-1,
immediately below the paper mill, have been blacked in.
Quite low observed values of Kox were observed, as a result
of the unusual pollution situation noted above.  These two
results are thus regarded as reflecting a different, or sep-
arate pollution situation.  Also, the five observations rep-
resenting the reaches 9-10 and 10-11 have been partially
blacked in for the purpose of identifying another kind of
problem, and will be discussed below.

Considering all 80 observations, the correlation coefficient
for the relationship shown in Figure 68 was 0.739, both for
the regression line forced through the origin and for the
regression line with a nonzero intercept.  The slope of the
line forced through the origin was 0.0533/ft.

When the two points representing the reach 0-1 are elimina-
ted because of the unusually great pollution, for the re-
maining 78 results the correlation coefficient of the re^
gression line forced through the origin is 0.811, a distinct
improvement, and the slope is 0.0549/ft.  The correlation
might be further improved somewhat by deleting additional
results (e.g., for the reach 1-2) that were certainly also
affected by the large upstream pollution load, but the
greatest effect is undoubtedly in the reach 0-1, and further
refinement is not regarded as either necessary or desirable.
The results for the 78 observations yield entirely adequate
estimates of correlation and slope, and those estimates are
shown in Figure 68.  They provide excellent support of the
energy dissipation model, equation  (48), especially when it
is recalled that there was a substantial time lapse  (4 years)
between the tracer studies and the observation of water sur-
face elevations, that the latter were quite independently
observed by others, and that refined estimates of Ah for
each reach during each tracer release cannot be made.

One other source of error, or deviation, is illustrated by
Figure 68.  Referring to that Figure, the two partially
blacked in points that are above the line of best fit repre-
sent all of the results for the reach 9-10, and the three
                          232

-------
              0,50
to
CO
to
                            FIGURE 68
                          JACKSON  RIVER
                              25°C
                         KQX VS,  RATE OF
                       ENERGY EXPENDITURE
                           ALL RESULTS
                                                                      SLOPE = 0.0549/FT
                                                                 • REACH 0-1
                                                                   REACHES 9-10 AND 10-11
                                                 ,  FEET  PER  HOUR
                                              tf

-------
similar points that fall below the line are all results for
the reach 10-11.   During the 1970 field observation of ele-
vation changes,- considerable difficulty was encountered in
finding the exact location of Station 10 from the notes of
the 1966 tracer studies.  Between Stations 9 and 11 consid-
erable elevation change occurs (23.3 feet), especially in
the upper portion.  The exact location of Station 10 deter-
mines the amount of elevation change in each of the two seg-
ments, and, therefore, the magnitude of (Ah/tf) in each seg-
ment.  The observed values of percent gas transfer, time of
flow and KOx were all quite consistent from one tracer re-
lease to the next, and the results therefore suggest strong-
ly that the reported elevation of Station 10 is in error be-
cause of the difficulty in finding the exact station loca-
tion.  For example, if the Ah for the reach 9-10 were two
feet larger, and the Ah for reach 10-11 were correspondingly
two feet smaller, all five affected points in Figure 68
would fit the straight line much better.

As with the Patuxent, the Jackson River study section con-
tained no unusual or remarkable hydraulic features such as
dams or waterfalls or shoals or violent rapids, but consis-
ted mainly of alternating gentle riffles and shallow pools.
In that sense, the Jackson possessed a considerably higher
degree of hydraulic uniformity than the Flint or South
Rivers, and the range of observed values  of KQX (25°C) and
of  (Ah/tf) was therefore relatively small.  Specifically,
the range of observation of both KQX (25°C) and (Ah/tf) was
only about twice that of the Patuxent.   The degree of corre-
lation associated with equation  (48), rxy = 0.811, although
still quite adequate, was thus relatively sensitive to prob-
lems such as those outlined above in connection with the
reach 0-1 and Station 10.  In contrast, the correlation
associated with equation  (49), rxy = 0.963, is again excel-
lent, depending only upon the amounts of total elevation
change and total gas transfer, rather than upon rates.

The Jackson River studies, representing the first field tra-
cer studies of stream reaeration capacity, thus also provide
the strongest of experimental evidence supporting the energy
dissipation models, equations  (48),  (49) and (50).

Chattahoochee jliver.  As reported in Section VI, the tracer
studies of the" reaeration capacity of the Chattahoochee
River were conducted in the fall of 1969 and the summer of
1970 as a part of this research.  In order to obtain steady
flows, it was necessary to conduct these studies on Sundays,
when demands for power are low, as the river is used for the
production of electric power.  As a result, the degree of
pollution, in terms of observed BOD's, was quite low.  The
                          234

-------
Chattahoochee studies included a total of nine tracer re-
leases; of these, eight were conducted at .a critical low
flow of about 1,100 cfs in the study section, while one
(the last) was at a flow of about 3,300 cfs.  The observed
reaeration capacity, measured in terms of Kox at 25°C, was
low relative to the other streams studied, being only about
one-third that of the Patuxent.

A total of 37 individual observations of reaeration capacity
were made in the 18-mile study section below the Clayton
STP, and the detailed results are provided in Appendix AIV
as well as in Section VI.  Six of these observations were
associated with the higher flow of 3,300 cfs, the remainder
being for a steady flow of 1,100 cfs.  As noted in Section
VI, the BOD's were somewhat higher during the earlier
studies, but were quite low in 1970, when most of the ob-
served results were obtained.  In none of the tracer studies
did the BOD's approach usual magnitudes for weekday pollu-
tion loads.

The water surface elevation changes reported here were ob-
tained and adjusted in accordance with the procedures out-
lined in Section V to the extent possible.  As noted in Sec-
tion VI, the field measurements of water surface elevation
change were more difficult than usual because of the inac-
cessibility of the river at many locations where the banks
were high and steep.  Coupled with that problem, discrepan-
cies were found among the reported elevations of the several
available bench marks.  The water surface elevation changes
reported here were obtained by thorough analysis of all of
the available bench mark information, together with the re-
search staff level records and routine tapedowns, and are
believed to be quite accurate.  However, it was not possible
to resolve every last discrepancy among all of the reported
bench mark elevations, and hence some question must remain
until such discrepancies are finally corrected by a compre-
hensive survey of the area.  In the meanwhile, the extent
of any error that might thus be incorporated in the water
surface elevation changes reported here is small—at a max-
imum, the total water surface elevation change for the en-
tire 18-mile study section would be increased by no more
than two feet.  Of course, the reaeration capacity is ob-
served quite independently, and the values of Kox are not
affected.

It was pointed out in Section VI that the individual ob-
served values of Kox in certain reaches varied more, on a
percentage basis, than in the studies of other rivers, al-
though the absolute variation was not larger than usual.
Increasing the number of observations of KQX per reach
                          235

-------
compensated for this,  and the set of mean values of Kox for
all reaches was quite  consistent.  The higher percentage
variability within some reaches occurred, in essence, be-
cause the reaeration capacity of the Chattahoochee is rela-
tively small—in any one tracer study, the total gas loss
was no more than 25 percent or so, compared to 80 or 90 per-
cent or more in the studies of other rivers.  In addition,
the Chattahoochee tracer doses were relatively small (mc/cfs)
and velocities and dispersion were large compared to the
other river studies.  The following analysis of results is
therefore based primarily upon the mean values of Kox ob-
served for each reach, as shown in Table 20 / Section VI.
As will be seen, the statistical tests of the energy dissi-
pation models are not  satisfactory when all 37 individual
results are considered, but, in contrast, yield very satis-
factory results in terms of the mean values by study reaches.

Figure 69 is a plot of the data in terms of the energy dis-
sipation model, equation (50), wherein the tracer gas loss
is related to water surface elevation change.  The mean
values by reaches have been shown (see Table 20 f section
VI) , and, as may be seen, fit that model very well to the
extent permitted by the range of observation.  The dashed
line extrapolation of  the observed data has been fitted
from the companion plot of results shown in Figure 70.

Figure 70 shows the mean values by reaches plotted in the
form required by equation (49), namely, the logarithm of
the percent tracer gas remaining vs. the water surface ele-
vation change.  As may be seen, the data describe a straight
line as predicted by the theory, and the fit is excellent.
For the eleven mean values, a correlation coefficient of
0.986 resulted for the line forced through the origin,  and
the associated slope (-0.0322/ft) yielded a half-height of
21.5 feet.  These results thus provide additional powerful
evidence in support of the energy dissipation models, equa-
tions (49) and (50).  For the unconstrained regression line,
both the correlation coefficient and the slope were essen-
tially the same.

The 37 individual results have also been tested statisti-
cally in terms of equation  (49).  Considering all 37 re-
sults, regardless of differences in flow and BOD, a correla-
tion coefficient of 0.703 resulted, and a slope of
-0.0239/ft, for the unconstrained regression line.  For the
line forced through the origin, the correlation coefficient
was reduced to 0.684 and the slope increased to -0.0284/ft.
Comparing these results to those shown in Figure 70, it is
evident that taking the mean values by study reach has vir-
tually eliminated the  variability of individual observations
                          236

-------
OJ
                100
                 80
              o
              LP\
              OJ
               CO
               00
               O
               CO
               <
               CD
               LU
               CJ
               ct:
               LU
               Q_
                 60
20
                   0


                                                                  FIGURE 69
                                                            CHATTAHOOCHEE RIVER
                                                                    25°C
                                                                 GAS LOSS VS
                                                              ENERGY EXPENDED
                                                           (MEAN  VALUES  BY REACH)
                    0
             10         20        30        10         50
             WATER  SURFACE ELEVATION CHANGE, Ah,  FEET
60

-------
100
90
80
70
60
I 50
LU
Cd
C/5
*^C
i-
y Zj°
o:
Q_
30




?n
XX
^>o
V








X










u*\>.
^s,^




FIGURE 70
CHATTAHOOCHEE RIVER
25°C
GAS REMAINING VS
ENERGY EXPENDED
(MEAN VALUES BY REACH)















/ 	 (Ah)l/2 = 21'5 FT
rx 0.986
>









vk^
Xx-









^x








0
5         10         15        20        25
WATER SURFACE ELEVATION CHANGE, Ah, FEET
30
                     238

-------
and resulted in a highly consistent fit of equation  (49) , in
keeping with earlier comment.

Figure 71 shows the data plotted in the form required by
equation (48) , the energy dissipation model relating Kox to
the rate of energy expenditure.  Although the range of ob-
servation is very small, and despite the variability of in-
dividual observations of KQX within certain study reaches,
the fit of the eleven mean values by study reach is excel-
lent.  For the regression line forced through the origin, a
correlation coefficient of 0.927 resulted, and a slope of
0.0374/ft.   For the unconstrained regression line, the cor-
relation coefficient was 0.939 and the slope 0.0324/ft.
Thus, the Chattahoochee data, taken as mean values by study
reach, also provide powerful support of the energy dissipa-
tion theory in the form of equation (48).

The statistical analysis of the 37 individual observations
of reaeration capacity  (Kox) and energy dissipation rate
(Ah/tf) leads to quite different results.  As noted earlier,
the small range of observation, coupled with the fact that
the magnitude of the absolute error of measurement is com-
parable to the magnitude of KOX itself, leads to poor corre-
lation, just as these factors also affected the fit of sim-
ilar data for the Patuxent.  For the unconstrained regres-
sion line, a correlation coefficient of only 0.298 was cal-
culated for the 37 individual results, with a slope of
0.0169/ft and an intercept of 0.022/hr.  For a regression
line forced through the origin, as required by the theory,
the slope improved markedly  (0.0339/ft, as compared to the
0.0374/ft shown in Figure 71), but the correlation went es-
sentially to zero.  Such results are typical where the mag-
nitude of error or deviation is comparable to the magnitude
of the observations themselves, as pointed out also earlier
in connection with the statistical tests of the predictive
models of O'Connor, Churchill, and others.

These particular results for the Chattahoochee River empha-
size the importance of obtaining an adequate number of ob-
servations of KOX for rivers wherfe the magnitude of KOx is
low.  In this case, the provision for four or five obser-
vations of Kox in each study reach proved adequate to yield
a highly consistent set of mean values, and a very satisfac-
tory test of equation  (48), even though the magnitude of Kox
was not much larger than the absolute error of observation.

Thus, the Chattahoochee River data also provide firm support
of the energy dissipation models, equations  (48) and  (49),
notwithstanding the several problems of measurement noted
above,,
                          239

-------
£>.
O
                 0,10
                0,08
             o
             LT\
             CNJ
             QC.
             CC
             LU
              X
              o
                0,06
                0,04
                0,02
                   0
                                            SLOPE =  0.0374
                                    FIGURE 71

                               CHATTAHOOCHEE RIVER
                                      25°C
                                                               KQX VS, RATE OF
                                                             ENERGY EXPENDITURE
                                                           (MEAN VALUES BY REACH)
                     0
0,5
1,0
1,5
2,0
2,5
                                            ,Ah.
                                            —)  / FEET PER HOUR
3,0

-------
Before closing this discussion of the Chattahoochee River
reaeration studies, it should also be noted that the half-
height of 21.5 feet reported in Figure 70 does not appear
to be fully consistent with the half-heights observed for
the other rivers studied.  The BOD's observed for river sam-
ples taken during each Chattahoochee tracer study tended to
be low, at least in 1970, although 5-day BOD's in the range
of 10 mg/1 were found for Dumps XVI, XVII and XIX.  Although
the observed half-height of 21.5 feet is well within the
range of observation for the four other rivers (10 feet to
26 feet), it appears to be somewhat greater than might be
expected on the basis of BOD results only; if BOD as a mea-
sure of the degree of pollution were the only criterion, a
half-height of about 15 feet for the Chattahoochee might be
more nearly in keeping with the results for the other
streams.

Of course, BOD is an indication of only one kind of pollu-
tion, and other pollutants that affect reaeration capacity
may well have been present but not detected in a BOD test.
Also, the Sunday pollution situation is unusual for the
Chattahoochee, and it is not at all impossible that some re-
sidual effect of greater weekday pollution was still opera-
tive on Sundays--for example, cumulative contamination of
the stream bed with material such as grease could quite pos-
sibly influence water surface replacement rates.   Whatever
the cause, the half-height of 21.5 feet is firmly estab-
lished by the observed results.

Summary Analysis—Five Rivers.  As a final and general test
of the  energy dissipation models, and of their usefulness
for extrapolation to other streams, summary statistical
tests of equations  (48) and  (49), using all of the data from
the five river studies, have been performed, and the results
are provided below.  To recapitulate, the streams involved
are the Flint, South, Patuxent, Jackson and Chattahoochee
Rivers, and the reaeration tracer studies took place over a
four year period, 1966-1970.  The studies involved several
different research staffs during that time, and took place
in three separate States.  They covered a total of about 70
miles of stream, in each case involving a critical section
below a pollution source.

The wastes discharged to the five streams were quite varied
in character and strength, and generally typical of communi-
ty wastes of mixed domestic and industrial origin.  They
ranged  from untreated or bypassed raw waste  (e.g., South
River)  to a highly treated secondary effluent  (Patuxent
River), and in one stream  (Jackson River) included the ef-
fluents from a large paper mill.  The five-day BOD's of the
                          241

-------
river water ranged from as little as 3.0 mg/1 to as much as
30 mg/1 or more during individual tracer studies.  Stream
flows during the tracer studies ranged from as little as 5
cfs (Flint River, Dump XIV)  to as much as 3,300 cfs  (Chat-
tahoochee, Dump XXV).   Stream temperatures were primarily
within the range 18°C to 28°C, although in one or two stud-
ies they were as low as 10°C and as high as 35°C.  Thus, the
data provided in Appendix AIV, and analyzed below, incorpo-
rate a quite wide range of pollution loading, flow and
stream temperature.  The corresponding observed values of
K2 at 25°C ranged from essentially zero in pools to a magni-
tude of 15.0 per hour at a waterfall.

A total of 323 individual observations of gas transfer and
K2 were made in the five rivers, the individual results
being reported in Appendix AIV and Section VI.  In most
cases, each specific stream reach was covered by two or more
tracer releases, in order to verify reproducibility and to
provide mean values of K2 for each stream reach that mini-
mize the inevitable errors of field measurement.  Reproduc-
ibility was generally excellent.

For purposes of these summary tests of all of the observed
results, the mean values by stream reach have been used as
representative of the best available results.  These are the
means reported in the individual tables of Observed Reaera-
tion Coefficients, Section VI, and drawn from the data pro-
vided in Appendix AIV.  A total of 101 mean values of K2
are thus available, and the corresponding mean values of gas
loss, Ah and (Ah/tf)  have been obtained directly from the
results given in Appendix AIV-  For the sake of consistency,
only those reaches having available two or more observations
have been used—this has resulted mainly in deletion of
those few results from special studies of the rapids sec-
tions in the Flint (reaches R1-R3) and South  (reaches in-
volving Station T2), wherein only one observation was made.

As indicated earlier,  it has been necessary to delete Sta-
tion 7, Flint River, because the water surface elevations
at that location were not measured and are not available.,
Also, as outlined in the preceeding section regarding the
South River studies, Station J, Dump VIII, has been elimi-
nated because of a bad dose assay, and Station K, Dumps IX
and XI, is excluded because of extremely low krypton-85
count rates.

Referring now to Equation (49) , the energy dissipation model
that relates the logarithm of the percent tracer gas remain-
ing to the amount of energy expenditure, Ah, a correlation
coefficient of 0.915 was obtained for the regression line
                          242

-------
forced through the origin, for the 101 available results.
The slope of this line of best fit was -0.0456/ft, which
corresponds to a half-height of 15.2 feet for the tracer
gas.  For the unconstrained regression line the correlation
coefficient was the same, the slope was -0.0450/ft, and the
intercept -0.0133.  Elimination of the two waterfalls
(Flint River, reaches 1P-1 and 2P-2) and of Panola Shoals
(South River, reach K-L) changed neither the correlation
coefficient nor the slope of either regression line.

Thus, the summary analysis of all of the observed results
for the five rivers studied indicates that at 25 °C equation
(49) takes the form

                                  Ah                    (54)
where y is the percent tracer gas remaining at Station 2,
GI and C2 are the respective concentrations of dissolved  tra-
cer gas at Stations i  and  2, and Ah is the water surface el-
evation change, in feet, between stations 1 and 2.  Inter-
preted according to equation  (52) , on the average for these
five streams 50 percent of the tracer gas was lost to the
atmosphere in every 15.2 feet of fall.

The foregoing result may be converted to DO deficit terms by
use of equation (53) .  This leads to
                       . = e~0.0549Ah
                     Dl

wherein D]_ and D2 are the respective DO deficits at Stations
1 and 2, in mg/1, and Ah is as before.  Thus, the half-
height for the DO deficit is 12.6 feet, whence 50 percent of
a prevailing DO deficit would be satisfied in 12.6 feet of
fall, 75 percent in 25.2 feet of fall, etc., if there were
no sources of concurrent DO consumption.

Statistical tests of equation  (48) , the model relating K2
to the rate of energy dissipation,  (Ah/tf ) , were equally
satisfactory.  For these tests, the two waterfalls  (Flint
River, reaches 1P-1 and 2P-2) and Panola Shoals  (South
River, reach K-L) were handled separately  in order to avoid
the undue statistical influence of the very high associated
values of K2 and (Ah/tf).  Thus, the 98 results having a
(Ah/tf) of less than 12.0 were tested first, and following
that the three results associated with the waterfalls and
Panola Shoals were added for the test of all 101 available
results .


                          243

-------
For the 98 results having (Ah/tf)  < 12.0, the coefficient of
correlation obtained was 0.905 for the regression line
forced through the origin, and the slope of the line was
0.0524/ft.  For the unconstrained regression line, the cor-
relation coefficient was identical, the slope was 0.0509/ft,
and the intercept was 0.0075.  Accordingly, on the average
for the five rivers studied, equation  (48) may be written as

                    K0 = 0.0524 (~)                     (56)
                     I           tf

where K2 is the DO reaeration coefficient (to the base e) at
25°C, per hour, Ah is the water surface elevation change in
the stream reach, in feet, and tf is the time of flow within
the reach, in hours.

Inclusion of the two waterfalls and Panola Shoals expands
the scale of observation considerably and improves the cor-
relation.  However, these three additional results then ap-
pear to have an unduly great effect on the line of best fit.
For example, the inclusion of Panola Shoals alone (n=99)
changes the correlation coefficient to 0.98, while the slope
stays much the same (0.0534/ft), for the regression line
forced through the origin.  Adding the results for the two
waterfalls  (n=101) yields an even higher correlation coef-
ficient, but a significantly changed slope  (0.0681/ft).

Rather than accept these last results, equation (56) is pre-
ferred, with its associated correlation coefficient of
0.905, as most representative of the main set of results.
The values of K2 for Panola Shoals and the two waterfalls
may then be predicted from equation (56) and compared to the
observed results:
                                 K_/hr, 25°C
         Reach       f      Observe'd     Predicted
         K-L       54.9       2.96         2.88
         2P-2      29.7       2.61         1.55
         1P-1     182        12.7          9.5

As may be seen, the prediction for Panola Shoals is excel-
lent, whereas the predictions for the two waterfalls are
somewhat low.  As noted earlier, the reach 2P-2 includes a
deep mixing pool below the waterfall, and the data indicated
an associated enhancement of gas transfer; the prediction
for the reach 1P-1 is better, and agrees quite well with
one of the two available observed results  (9.5/hr vs.  10.2
and 15.1) .
                          244

-------
The foregoing tests involving the available results for the
five rivers provide ample verification and support of the
energy dissipation models for stream reaeration capacity,
and of their basic nature and general applicability.  Those
models now permit rational prediction of stream reaeration,
with a degree of accuracy and dependability not heretofore
possible.

PREDICTION OF STREAM REAERATION

The primary objective of this research has been to improve
the ability to predict the reaeration capacity of a stream
section in terms of the hydraulic properties of that sec-
tion.  Reaeration is a direct function of turbulent mixing
in the specific sense of water surface replacement, and
hence, to be successful, the predictive model must explain
reaeration in terms of the hydraulic properties that cause
surface replacement.

The predictive models that have been available prior to this
research attempt for the most part to relate the reaeration
rate coefficient, K2, to the velocity and depth of stream
flow  (see Section IV).  The original model of this form was
proposed in 1925 by Streeter and Phelps^, and those authors
indicated clearly then that they did not regard their model
as sufficiently accurate or general for widespread applica-
tion.  The more recent and currently available predictive
models largely follow the same empirical form.  They dis-
agree among themselves, mainly regarding the magnitude of
empirical coefficients.  As has been seen, they have proved
to be quite unsuccessful when tested against accurate ob-
served values of K2, especially for stream sections that are
relatively turbulent.

The energy dissipation models, equations  (48),  (49) and  (50),
have been derived from simple considerations of energy expen-
diture in open channel flow and the kinetics of gas transfer,
and their theoretical basis is thus sound and complete.  They
have been tested against the magnitudes of gas transfer and
K2 observed in tracer studies of the reaeration capacities
of five rivers.  As has been shown, in each separate case
the observed data provide powerful support of the energy
dissipation models.  Those models have also been tested, in
a summary analysis, against all of the observed data from
the five rivers taken together, and these summary tests also
resulted in unquestionnable verification of the models.  The
energy dissipation models are thus regarded as thoroughly
demonstrated in both the theoretical and experimental senses.
They may now therefore serve as basic models for the predic-
tion of stream reaeration capacity.
                          245

-------
Equations (48)  and (50)  may be regarded as the two basic
energy dissipation models.  They are quite different, yet
intimately related.  The one, equation (48), refers to the
rate of energy dissipation, and relates that to the rate of
gas transfer.  The other, equation  (50) ,  refers to the
amount of energy dissipated, and relates that to the amount
of gas transferred.  Equation (49) is merely a mirror form
of equation  (50),  referring to the amount of gas that re-
mains, or has not been transferred.  As has been noted,
amounts are easier to measure accurately than rates, as a
rule, and equation (50)  is therefore the simpler and more
dependable form.

The Escape Coefficient.   In order to use equations (48),
(49) and (50) to predict stream reaeration capacity, the
only term that needs to be evaluated is the exponential co-
efficient,  c.  Because of its nature, as  outlined below, c
is designated the "escape coefficient", and will be referred
to in that manner for the remainder of this report.

Although it has been mostly ignored in the discussions pre-
sented thus far, it is now emphasized that the exponential
coefficient c is common to all three models, and has the
same numerical value in all three models  when restricted to
a single gas and a single stream.  In the material presented
thus far, c has been discussed primarily as a half-height
for krypton in connection with equations  (49) and  (50), and
as the slope of a line that refers to oxygen in connection
with equation  (48).  Actually, although arrived at in dif-
ferent ways and by different statistical tests, these are
only different manifestations of the same number, and may
be expressed as the same number through use of relationships
presented earlier.  The following example serves to illus-
trate these conversions.

Referring to the observed results for the Patuxent River, as
reported earlier, Figures 64 and 65 provide the required in-
formation.   From the statistical analysis according to equa-
tion  (49), the observed half-height for krypton was 10.5
feet  (see Figure 64).  This is converted to a half-height
for oxygen of  8.7 feet by use of the basic Kkr:Kox ratio of
0.83, as expressed in equation  (53).  The latter is in turn
converted to the required slope,  c, by equation  (52).  The
value of c thus derived is 0.0797/ft.  This result could
also have been obtained more directly from the slope of the
regression line forced through the origin, which also is
converted to the gas oxygen by use of the basic ratio 0.83.

Reference now to Figure 65 provides the results obtained
from separate  statistical testing according to equation  (48),
and the desired comparison.  From Figure 65, for the
                          246

-------
regression line forced through the origin the slope, c, was
0.0804/ft.  This is in excellent agreement with the value
c = 0.0797/ft derived from equation  (49).  The single value,
c = 0.080/ft, is thus firmly established for the Patuxent,
and may be used in any of the models, equations  (48),  (49)
and (50), for oxygen transfer predictions for that stream.

The foregoing example was chosen also to illustrate certain
other matters.  As indicated in the discussion of the Pa-
tuxent River results, the quality of the observed data was
not regarded as quite as good as that for the other streams.
This is evident, for instance, in the unusually large
spread of results in Figure 64.  Also, the range of obser-
vation of Kox was very small, as shown in Figure 65, and
this resulted in poor correlation in connection with the
testing of equation  (48)—the correlation coefficient was
only 0.559 for the unconstrained line of best fit, and was
reduced to a very poor 0.245 for the regression line forced
through the origin.  Nevertheless, the agreement between the
separately obtained values of c, namely 0.0797/ft and
0.0804/ft, was excellent.

The reason for such excellent agreement between the two
values of c is eminently clear.  The energy dissipation
theory requires that the regression lines go through the
origin in both Figures 64 and 65.  This is simply to state
that if no energy is expended essentially no gas transfer
will occur.   (The transfer that takes place as the result
solely of molecular diffusion of the gas is trivial and en-
tirely negligible on a comparative basis).  The correlation
coefficient is not the most important matter here, but
rather the slope.  Hence, the larger correlation coefficient
 (0.559) was rejected in favor of the smaller (0.245) with
its corresponding theoretically valid slope.

As has been shown, the correlation coefficients associated
with equation  (49) have been very good or excellent in all
cases, and those associated with equation  (48) were equally
good for the Flint, South and Jackson Rivers, and for the
summary analysis of results for all rivers taken together.
Excellent correlation was also found for equation  (48) using
the mean values by reaches for the Cahttahoochee, but not
for the individual values of Kox because, just as with the
Patuxent, the range of observation was very small, and in
the case of the Chattahoochee the magnitude of Kox was of
the same order as the magnitude of experimental error.

Table 28 summarizes all of the values of the escape coeffi-
cient,  c/ found in these studies, as reported in the earlier
part of Section VIII and converted to the gas oxygen as
above«  All results refer to 25°C, and all refer to the line
                          247

-------
00
                                          Table 28

                      Observed Values of the Oxygen Escape  Coefficient


                                     Escape Coefficient,/hr,  25°C
River
Flint


South



Patuxent
Jackson
Chattahoochee
All Rivers
Results
All
Dump XIV
Other1
Lower Reaches
Upper Reaches
VIII and X
All
All
All
All2
All2
Eq(49)
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0654
0796
0584
0545
0417
0317
0461
0797
0535
0388
0549
EgJ48)
0
0
0
0
0
0
0
0
0
0
0
.0635
.0778
.0583
.0556
.0414
.0318
.0538
.0804
.0549
.0374
.0524
Mean ^aa;l/2' rt
0
0
0
0
0
0
0
0
0
0
0
.0645
.0787
.0583
.0550
.0416
.0318
.0500
.0800
.0542
.0381
.0537
10
8
11
12
16
21
13
8
12
18
12
.8
.8
.9
.6
.7
.8
.9
.7
.8
.2
.9
     1.  All results exclusive  of  Dump  XIV
     2.  Mean values by reaches

-------
of best fit through the origin as required by the theory.
As may be seen, the coefficients shown for the summary anal-
ysis of data for all five rivers are the same coefficients
shown in equations  (55) and  (56) , and have the same inter-
pretation.  The last column of Table 28 shows the half-
height for oxygen deficit, obtained from the mean values of
c by equation  (52).  The half-height is a more convenient
and readily understood means of  expressing the magnitude of
the escape coefficient and its meaning as regards the re-
aeration capacity.

Referring to Table 28, the agreement between the values of
c obtained from separate testing of equations (48) and (49)
is excellent in each case.  The  only disagreement worthy of
mention relates to testing of the South River data all to-
gether, in which case the derived values of c (0.0461 and
0.0538) are within + 8 percent of their mean.  This dis-
agreement is undoubtedly due to  the multiple and variable
pollution load situation in the  South, and no such dis-
agreement occurs when the same data are grouped according to
the degree of pollution.  The range of all observed values
of the escape coefficient, c, will be discussed further be-
low.

Nature and Range of the Escape Coefficient.  Before pro-
ceeding to the matter~of predicting stream reaeration by use
of the energy dissipation models, a brief recapitulation of
the nature of the escape coefficient, c, is desirable.  Re-
ferring back to the derivation of equation  (48), it was
shown that
                       c =  (a) x  (b)                    (57)

where the coefficient, a, was taken from equation  (4) and
the coefficient, b, from equation  (47).  The two coeffi-
cients, a and b, are quite  distinct.   From its original
derivationlO, the coefficient, a,  refers specifically to the
physical molecular properties of the diffusing gas and to
the quality of the water medium.   It thus refers to  such
properties as the diffusivity of oxygen and its molecular
diameter, and to the kind and degree of pollution present,
in the sense that pollutants can influence the ability of
the gas to diffuse or transfer in  water.  The coefficient,
a, thus has nothing to do with turbulence or mixing, and is
not affected by those or other hydraulic properties  of the
water.

Referring now to equation  (47) and its accompanying  postu-
late, it is evident that the coefficient, b, refers  exclu-
sively to the hydraulic properties of  the water, and has
                          249

-------
nothing to do with the inherent molecular nature of the gas.
Thus, b refers to those properties associated with the rates
of water surface replacement and energy expenditure.  It is
a mixing coefficient,  and has to do with both the degree
and quality of mixing.  In that sense, it may be influenced
by the "uniformity" or nonuniformity of flow, or by flow
itself to the extent that changes in stream flow could af-
fect the mixing regime.  As has been seen, however, changes
in stream flow to the  extent of a factor of two or three
have had no discernible effect upon Kox, so that it appears
unlikely that changes  of stream flow within that range have
much real effect in any specific stream.

Having distinguished between the two components of the es-
cape coefficient, c, it should also be noted that the pres-
ence of a pollutant could affect both.  For instance, it
was shown in Section VII that LAS and mineral oil had op-
posite effects on the  gas transfer coefficient, K^j-, one
(LAS) depressing it and the other (oil)  raising it.  Those
results indicate the possibility that a specific pollutant,
oil for example, may affect both the ability of the gas
molecules to diffuse within the water medium, or to escape
from it, and the freedom with which the water surface can
be replaced (by altering the surface tension, for instance).
This has not been proved by the results reported here, but
appears to be a definite possibility.

In the original derivation of the basic reaeration equa-
tion10 it was shown that the net rate of gain of oxygen in
reaeration may be regarded as limited by the ability of the
gas molecules to escape from the water medium, and that gas
transfer is controlled by both the molecular characteristics
of the gas and the rate of water surface replacement.  Ac-
cordingly, the coefficient c, defined as in equation  (57),
has been designated the "escape coefficient."

Referring now to the observed values of the escape coeffi-
cient shown in Table 28, it is at once evident that the
range of results is very small.  For the five streams stud-
ied, the range of c is only from 0.0317/ft to 0.0804/ft, the
comparable half-heights ranging from 8.6 feet to 21.9 feet.
Thus, the largest value is only 2.5 times the smaller. Re-
ferring to the mean value for c of 0.0537/ft for all rivers,
all of the results for individual streams fall within the
range described by the mean plus or minus 50 percent.

It has already been noted that, when all five streams are
considered, the studies reported here incorporate a very
wide range of stream flow (5 to 3,300 cfs), pollution load
(at least 3 to 30 mg/1 of 5-day BOD), river water tempera-
ture (10°C to 35°C) and reaeration capacity  (Kox at 25°C
                          250

-------
from zero to 15.0/hr).  The derived values of the escape co-
efficient are therefore regarded as being representative of
a great many, probably most, streams.  The corresponding
very small range of c thus, by itself, indicates the funda-
mental nature of the escape coefficient.  It also indicates
that the reaeration capacity of a stream should now be pre-
dictable within relatively narrow limits of error from the
stream hydraulic properties.

Both the field tracer studies of reaeration capacity and the
laboratory studies of the effects of pollutants on gas
transfer also show clearly that the escape coefficient is
markedly affected by the kind and degree of pollution pre-
sent in the water.  In brief, the greater the degree of pol-
lution, the lower the escape coefficient.  This was clearly
shown in connection with the laboratory scale LAS studies
(Section VII) , and was particularly evident in the results
from the South River field studies , as well as from all of
the field results taken together. (It should be noted here
that, although mineral oil did enhance gas transfer, rather
than depress it, the quantities of oil required were large,
and this is therefore not taken to be a common significant
effect in streams) .  In essence, pollutants reduce the mag-
nitude of the molecular coefficient, a, and thereby also
reduce c.  The data indicate, in fact, that the presence of
pollution is a major factor, if not the principal one, in-
volved in the observed range of the escape coefficient.
This will be of great assistance in narrowing the range of
predicted values of stream reaeration capacity.

Predicting Stream Reaeration.  The reaeration capacity of a
section of stream may now be predicted on the basis of the
foregoing information.  Two modes of prediction are avail-
able, as represented by equations (48) and  (50) , depending
upon the information available.  The one, equation  (48), re-
quires an estimate of the escape coefficient, c, plus the
water surface elevation change and the time of flow in the
stream section.  The other, equation  (50), does not require
time of flow information.

Equations  (55) and (56) provide a sound initial basis for
predicting stream reaeration.  They are given here again,
for convenience, and a single value of c is shown, namely
the mean of the two separately derived results  (0.0524/ft
and 0.0549/ft), as shown in Table 28.
and
                       2 = e-0.0537Ah                    (58)

                     Dl
                          251

-------
                     K. = 0.0537 ()                   (59)
                      2           tf

where, for the stream section between locations 1 and 2:
D]_ is the DO saturation deficit at the upstream location, in
mg/1; D2 is the DO saturation deficit that would occur at
the downstream location if there were no sources of oxygen
consumption within the reach, in mg/1; Ah is the water sur-
face elevation change between locations 1 and 2, in feet;
tf is the time of flow between locations 1 and 2, in hours;
and K2 is the reaeration rate coefficient (Kox) per hour and
to the base e.  Both equations apply at 25 °C.  For correc-
tion to other temperatures,  the escape coefficient,
0.0537/ft in this case, is corrected according to the rela-
tionships given earlier in equations  (9) and (10) , namely
where T is the other temperature, in "Centigrade.

Specifically, according to the results presented earlier,
equations (58) and (59) are regarded as capable of yielding
quite good predictions of the reaeration capacity of a
"typical" or "average" stream section that is moderately
polluted and reasonably well mixed.

The range of values of the escape coefficient shown in
Table 28 now provides the basis for modifying equations  (58)
and (59) to the extent warranted by specific stream condi-
tions.  Although no field tracer studies were conducted in
a stream that had never been polluted, the Patuxent River
was quite clean, receiving only a highly treated secondary
effluent.  It seems unlikely that the magnitude of c for a
never-polluted stream would be much larger than the mean of
0.080/ft observed in the Patuxent studies.  On the other
end of the scale, the mean of c = 0.0318 ft observed for
Dumps VIII and X in the South River represents a quite heavy
pollution load, in the neighborhood of 30 mg/1 of 5-day BOD.
Depending, then, upon the actual degree of pollution, as in-
dicated at least roughly by the BOD of the stream, the mean
value c = 0.0537/ft used in equations  (58) and  (59) may be
replaced by a more appropriate value within the range, say,
0.030/ft to 0.085/ft, for a temperature of 25°C.

Expressed as half-heights for the DO saturation deficit at
25 °C, the foregoing mean value of c for a moderately pollu-
ted stream represents a half -height of about 13 feet, and
the suggested range of prediction is from a half-height of
                          252

-------
about 8 feet for a never-polluted stream to 23 feet for a
heavily polluted stream.

The numerical values of c suggested above are for streams
that are reasonably well mixed in the longitudinal, vertical
and lateral senses.  When mixing is evidently poor or erra-
tic in one direction or another, great care is necessary in
predicting reaeration capacity.  Such a situation is well
illustrated by the experience in attempting to measure the
reaeration capacity of the long, deep pool above Panola
Shoals in the South River.  As shown earlier, at times the
entering flow proceeded through the pool at the surface, and
at other times it dropped to flow through along the bottom
to the pool outlet.  As a result, a wide range of reaeration
coefficients was observed, from zero to relatively high
values, and no single value could be regarded as truly rep-
resentative for the whole pool volume.  In other streams ex-
tensive lateral short-circuiting occurs.  In such situations
it is most important to exercise judgment regarding the por-
tion of apparent stream volume that is adequately mixed
with the flow, and the value of c to assign.

The BOD is, of course, also only a general indicator of pol-
lution, and specific kinds of industrial waste will no doubt
affect the reaeration capacity in somewhat different degree.
So that again, judgment is necessary in selecting the best
value of the escape coefficient, c, in specific situations.
Hopefully, additional research will in the future provide
firm information categorizing specific wastes in terms of
their effect upon the escape coefficient.

In brief, the quality and accuracy of any prediction of the
reaeration capability of a stream will depend principally
upon the quality and accuracy of the information available.
The prediction can be no better than the information upon
which it is based.  If it is merely an office exercise, or
sensitivity analysis, based upon crude estimates of Ah, then
the prediction will be crude to the same extent.  On the
other hand, if the prediction can be based upon accurate
field observations of Ah and time of flow, a considerably
better and more accurate estimate of reaeration capacity
will result.  If a specific waste of unknown impact is in-
volved, tests of the sort described in Section VII can be
of great aid in refining and improving the prediction.

Thus, the predictive models provided here are not intended
to serve as substitutes for accurate work.  Used judiciously,
and provided with accurate information, they can provide
very good estimates of stream reaeration capacity.  In the
final analysis, however, every stream must be regarded as
individual in terms of its own peculiar mixing
                          253

-------
characteristics,  the character of its waste load, etc.
Thus, where the economic or ecological situation requires it,
there can be no substitute for direct accurate field evalu-
ation of stream reaeration capacity by the tracer procedures
that form the basis of this report.
                         254

-------
                       SECTION IX

                     ACKNOWLEDGMENTS


The field and laboratory research  studies  reported here
could not have been accomplished without the generous assis-
tance and cooperation of many  individuals  and organizations.
Those who contributed and participated  in  a major degree are
acknowledged below.  There were numerous other persons who
also helped at one point or another, perhaps to lesser de-
gree, but nevertheless in an important  way at the time, and
their help is neither forgotten nor does it go unappreciated.

The Georgia Tech student assistants deserve unqualified ap-
preciation for their efforts.  They gave unstintingly of
their sincere interest as well as  their time and hard work,
and conducted themselves and their work in all ways as true
professionals.  Those who contributed in a major way include:
L. A. Neal, D. E. Hicks, R. F. Holbrook and H. F. Reheis, all
of whom were project leaders at one time or another; G. T.
Smith, L. A. Kendrick, R. W. Reineke, W. C. Carr, T. E.
Hopkins, K. J. Hatcher, L. C. Hicks and J. R. Rist all con-
tributed in many ways.

Linda Porter, the Project Chemist, worked  long and at times
most unusual hours,  was exceptionally  conscientious and de-
pendable, and displayed a great deal of understanding and
forebearance at times with engineers and other circumstances.

The Georgia Water Quality Control  Board provided help and
encouragement throughout the course of  the project.  The
Water Resources Division, Georgia  District, of the U.S.
Geological Survey was most helpful in lending the project
stream gaging and other equipment  at the outset and providing
other aid; the Georgia Power Company was most cooperative in
arranging for steady low flows for the  Chattahoochee River
studies; the Region IV Office of the Environmental Protection
Agency  (then, the Federal Water Pollution  Control Administra-
tion) helped in many ways, especially at the start of the
project; the Maryland Department of Water  Resources was in-
strumental in arranging for the Patuxent River studies, and
participated fully in that portion of the  research.

Field research of the type reported here is difficult at
best:  it must be scheduled in appropriate season, the
weather and stream conditions must be suitable at the right
moment, a goodly number of individuals  must be available on
short notice, vital supplies such  as tracer doses must be on
hand when needed, etc.  Under such circumstances, the sympa-
thetic understanding and unhesitating assistance of
                           255

-------
administrative officials and staff is vital, otherwise field
opportunities are lost.   The School of Civil Engineering
administrative staff provided much vital assistance at cru-
cial times, and,  under somewhat trying circumstances at
times, always managed to provide the needed help.  The Pro-
curement Office at Georgia Tech handled a great many "RUSH"
requisitions, both efficiently and pleasantly.  The Health
Physics staff at  Georgia Tech provided much assistance, some
at odd hours.  The Project Officer, Dr. Walter M. Sanders,III,
was especially helpful throughout the course of the project,
and displayed a good deal of understanding regarding the
delays that finally prove to be inevitable in field opera-
tions.

Margaret Smith, of the School of Civil Engineering staff,
deserves special  appreciation for typing this entire report,
and for carrying  out that demanding task with great patience
and pleasantness.
                          256

-------
                        SECTION X

                        REFERENCES

 1.   Black,  W.  M.f and Phelps, E.  B. "The Discharge of Sewage
     into New York Harbor."  Report made to the Board of Es-
     timate  and Apportionment, New York City (March, 1911).

 2.   Streeter,  H. W.,  and Phelps,  E. B. "A Study of the Pol-
     lution  and Natural Purification of the Ohio River. III.
     Factors Concerned in the Phenomena of Oxidation and
     Reaeration."  Public Health Bulletin No. 146, U.S. Public
     Health  Service, Washington,  D. C.  (1925).

 3.   Velz, C. J. "Applied Stream Sanitation."  John Wiley and
     Sons, New York, New York, 619 pp.  (1970).

 4.   O'Connor,  D. J.,  and Dobbins, W. E.   "The  Mechanism of
     Reaeration in Natural Streams."  Journal of Sanitary
     Engineering Division, Proceedings  of American Society of
     Cxvil Engineers,  82,  No. SA6, 1115 (December, 1956).

 5.   Churchill, M. A., Elmore, H.  L., and Buckingham, R. A.
     "Prediction of Stream Reaeration Rates."  Journal of
     Sanitary Engineering Division, Proceedings of American
     Society of Civil Engineers,  88, No.  SA4, 1 (July, 1962).

 6.   Krenkel, P. A., and Orlob, G. T. "Turbulent Diffusion and
     the Reaeration Coefficient."   Journal of Sanitary Engi-
     neering Division, Proceedings of American  Society of
     Civil Engineers,  88,  No. SA2, 53  (March, 1962).

 7.   Thackston, E. L., and Krenkel, P.  A.  "Reaeration Predic-
     tion in Natural Streams."  Journal of Sanitary Engineer-
     ing Division, Proceedings of  American Society of Civil
     Engineers, 95_, SA1, 65  (1969) .

 8.   Tsivoglou, E. C., O'Connell,  R. L.,  Walter, C. M., God-
     sil, P. J., and Logsdon, G.  S.  "Tracer Measurements of
     Atmospheric Reaeration.  I.  Laboratory Studies."  Journal
     of Water Pollution Control Federation, 37, No. 10, 1343-
     62  (October, 1965).

 9.   Tsivoglou, E. C., Cohen, J.  B., Shearer, S. D., and God-
     sil, P. J.  "Tracer Measurement of Stream Reaeration. II.
     Field Studies."  Journal of Water Pollution Control Fed-
     eration., 40, No.  2, Part 1,  ^85-305  (February, 1969).

10.   Tsivoglou, E. C.  "Tracer Measurement of Stream Reaera-
     tion."   USDI, Federal Water Pollution Control Administra-
     tion, Washington, D.  C., 86 pp. (June, 1967);  (Available
     US Government Printing Office, June, 1969).


                            257

-------
11.   King,  D.  L.  "Reaeration of Streams and -Reservoirs,,
     Analysis  and Bibliography."  USDI, Bureau of Reclama-
     tion,  Report no.  REC-OCE-70-55 (December, 1970);
     Available National Technical Information Service,
     Operations Division,  Springfield, Virginia, 22151.

12.   "Standard Methods for the Examination of Water and
     Wastewater."  Eleventh Edition, APHA, AWWA, WPCF  (1960).

13.   Isaacs, W. P., and Maag, J. A. "Investigation of the
     Effects of Channel Geometry and Surface Velocity on the
     Reaeration Rate Coefficient."  Engineering Bulletin,
     Purdue University, Volume LIII (1969) .

14.   Langbein, W. B.,  and Durum, W. H.  "The Aeration Capac-
     ity of Streams."  USDI, Geological Survey, Bulletin 542
     (1967).

15.   Streeter, H. W.,  Wright, C. T., and Kehr, R. W. "Mea-
     sures  of  Natural Oxidation in Polluted Streams. III. An
     Experimental Study of Atmospheric Reaeration under
     Stream-Flow Conditions."  Sewage Works Journal, VIII,
     No. 2,  282-316 (March, 1936).

16.   Owens,  M., Edwards, R. W., and Gibbs, J. W.  "Some
     Reaeration Studies in Streams."  International Journal
     of Air and Water Pollution, 8_, 469-486  (1964) .

17.   Gameson,  A.  L. H., Truesdale, G. A., and Downing, A. L.,
     "Reaeration Studies in Lakeland Beck," Journal of
     Institution of Water Engineers, Vol. 9, p. 571 (1955).

18.   Krenkel,  P.  A., "Turbulent Diffusion and the Kinetics
     of Oxygen Absorption," Ph.D. Thesis, University of
     California  (1960).

19.   Al-Saffar, A. M., "Eddy Diffusion and Mass Transfer in
     Open-Channel Flow," Ph.D. Thesis, University of Cali-
     fornia at Berkeley (1964).

20.   Daugherty, R. L., and Franzini, J. B., Fluid Mechanics,
     McGraw-Hill, New York, p. 283  (1965).      ~~"''

21.   Cohen, J. B., Setser, J. L., Kelly, W. D. , and Shearer,
     S. D.   "Analytical Method for the Determination of
     3H85Kr in Aqueous Samples." Talanta, 15,  233  (1968).

22.   Georgia Water Quality Control Board. "Water Quality
     Data—Atlanta Area.  Chattahoochee, Flint and South
     Rivers,  1968 and 1969 Data."  Georgia Water Quality
     Control Board, Atlanta, Georgia  (1970).
                           258

-------
23.   Georgia Water Quality Control Board. "Upper Ocmulgee
     River Basin Water Quality Survey."  Georgia Water
     Quality Control Board, Atlanta, Georgia (1971).

24.   Thackston, E. L. Personal Communication (July, 1971),
                           259

-------
                       SECTION XI

                PUBLICATIONS AND PATENTS
Publications

1.  Tsivoglou, E. C. , and Wallace, J. R. "Hydraulic Proper-
    ties Related to Stream Reaeration."  Proceedings, Iso-
    tope Hydrology, 1970.  International Atomic Energy
    Agency, Vienna, 509-522  (1970).

2.  Tsivoglou/ E. C., McClanahan, M. A., and Sanders, W. M.
    "Direct Tracer Measurement of the Reaeration Capacity of
    Streams and Estuaries."  Proceedings, July 7-8, 1970,
    Symposium, U.S. Environmental Protection Agency Publ.
    No. 16050 FOR 01/72.  Supt. Doc., U.S. Gov't. Pr. Office
    (1972).
Patents

None.
                           261

-------
                      SECTION XII

                      APPENDICES

                                                  Page No.

AI       River Sampling Locations	265

         Flint River	266
         South River	268
         Chattahoochee River	270

All      River Discharge at Time of Tracer
         Studies	271

         Flint River	272
         South River	273
         Chattahoochee River	274

AIII     Measured Hydraulic Properties	275

         Flint River	276
         South River	282
         Chattahoochee River	292

AIV      Detailed Tracer Study Results and
         Reaeration Coefficients	301

         Flint River	302
         South River	306
         Patuxent River	310
         Jackson River	312
         Chattahoochee River	316
                          263

-------
       APPENDIX AI
RIVER SAMPLING LOCATIONS
           265

-------
               RIVER SAMPLING LOCATIONS
River

Flint



Flint


Flint


Flint


Flint


Flint



Flint


Flint



Flint
Flint


Flint



Flint


Flint
Station

  0



  Rl


  R2


  R3


  IP


  1



  2P


  2
            Location
  3*
  BS
  AS
1500 ft. upstream from centerline
of Interstate Highway 285 between
sections 0 and 1.

1500 ft. upstream from Terrell
Mill dam, above rapids.

1200 ft. upstream from Terrell
Mill dam, mid-point of rapids.

1000 ft. upstream from Terrell
Mill dam, below rapids.

5 ft. upstream from Terrell Mill
dam between sections 7 and 8.

20 ft. downstream from center
line of Terrell Mill Road bridge,
150 ft. below dam.

5 ft. upstream from Lee's Mill
dam between sections 22 and 23.

75 ft. upstream from center line
of Lee's Mill Road bridge, 100
ft. below dam.

9600 ft. upstream from upper
Riverdale Road bridge, adjacent to
gravel quary between sections 30
and 31.

600 ft. downstream from station 3,
adjacent to gravel quarry.

1 mi. upstream from Upper River-
dale Road, before swamp between
sections 41 and 42.

4000 ft. upstream from Upper
Riverdale Road, after swamp.

10 ft. downstream from confluence
with Mud Creek, 2400 ft. upstream
from Upper Riverdale Road between
sections 47 and 48.
                           266

-------
River          Station                 Location
Flint            4         50 ft. upstream  from  center  line
                           of Upper Riverdale  Road bridge
                           between sections 52 and 53.

Flint            5         Upstream side  of Valley Hill Road
                           between sections 70 and 71.

Flint            6         50 ft. upstream  from  confluence
                           with Vester's  Creek between  sec-
                           tions  86 and 87.

Flint            7         100 ft. upstream from abandoned
                           Bethel Church  Road  bridge.
                            267

-------
River
Station
                                       Location
South



South



South



South



South



South



South


South


South
  A
  B
  D
  G*
  T2
  H
South
South
100 ft. downstream from center
line of Forrest Park Road bridge
between stations 4 and 5.

150 ft. upstream from confluence
with Intrenchment Creek between
sections 24 and 25.

50 ft. upstream from center line
of Bouldercrest Road bridge be-
tween sections 36 and 37.

20 ft. downstream from center line
of Panthersville Road bridge be-
tween sections 59 and 60.

50 ft. upstream from confluence
with Shoal Creek between sections
77 and 78.

50 ft. upstream from centerline of
Flakes Mill Road bridge between
sections 114 and 115.

50 ft. downstream from Flakes
Mill Road bridge, above rapids.

1000 ft. downstream from Flakes
Mill Road bridge, below rapids.

2800 ft. downstream from Flakes
Mill Road bridge, above treated
sewage discharge from Snapfinger
Creek Facility between sections
120 and 121.

500 ft. downstream from con-
fluence with Corn Creek between
stations 140 and 141.

4600 ft. upstream from McDonough
Road bridge.
                          268

-------
River         Station                  Location
South            K         75 ft. upstream from McDonough
                           Road bridge, above shoals between
                           stations 158 and 159.

South            L         425 ft. downstream from McDonough
                           Road bridge, below shoals between
                           sections 159 and 160.

South            M         9400 ft. downstream from McDonough
                           Road bridge off Lyons Farm Road
                           in pasture between sections 177
                           and 178.

South            N         Downstream side of Flat Bridge.
                           269

-------
River
Station
Location
Chattahoochee    0
Chattahoochee
Chattahoochee
Chattahoochee    4
Chattahoochee    5
Chattahoochee    6
            500 ft.  upstream from Georgia
            Highway 280 at section 17.

            7000 ft. downstream from Inter-
            state Highway 20 near section 80,

            12,500 ft.  upstream from Georgia
            Highway 166 bridge, at abandoned
            ferry crossing between sections
            124 and 125.

            50 ft. downstream from center
            line of Georgia Highway 166
            bridge near section 150.

            50 ft. downstream from center
            line of  Georgia Highway 92
            bridge near section 195.

            9000 ft. downstream from Georgia
            Highway  92  bridge,  at Brown's
            Lake.
                          270

-------
              APPENDIX All






RIVER DISCHARGE AT TIME OF TRACER STUDIES
                   271

-------
RIVER DISCHARGE AT TIME OF TRACER STUDIES

               Flint River

                 Station         Discharge(cfs)
                    0                10
                    1                10
                    2                10
    II              1                 9.2
                    2                 9.2
                    3                19
                    4                25
    III             0                10
                    1                10
                    2                10
                    3                22
    IV              3                24
                    4                24
                    5                25
                    6                27
    V               4                21
                    5                21
                    6                22
                    7                28
    XII             4                21
                    5                21
                    6                22
                    7                28
    XIV             0                 5.5
                    1                 5.5
                    2                 5.5
                  272

-------
RIVER DISCHARGE AT TIME OF TRACER STUDIES

                South River

                  Station         Discharge(cfs)

                     A                48
                     B                48
                     D                75
                     E               100
                     F               107
                     G               146
     VII             G               123
                     H               138
                     J               155
                     K               158
                     L               158
     VIII            J               186
                     K               190
                     L               190
                     M               206
     IX              E                69
                     F                85
                     G                90
                     H               135
                     J               158
                     K               158
     X               A                48
                     B                48
                     D                75
                     E               100
                     F               107
     XI              F               100
                     G               105
                     H               H2
                     J               130
                     K               130
     XIII             J               110
                      K
                      L
                      M              121
     XV              G              HO
                      H               122
                      J              140
                      K              142
                      L              142
                     273

-------
RIVER DISCHARGE AT TIME OF TRACER STUDIES
           Chattahoochee River
Dump
XVI

XVII


XIX


XX


XXI


XXII



XXIII



XXIV



XXV





Station
0
2
2
3
4
0
2
3
0
2
3
3
4
5
3
4
5
6
2
3
4
5
3
4
5
6
0
2
3
4
5
6
                               Discharge(cfs)

                                  1000
                                  1000
                                   925
                                   950
                                   910
                                  1030
                                  1080
                                  1100
                                   910
                                   920
                                   950
                                  1200
                                  1210
                                  1170
                                  1200
                                  1210
                                  1170
                                  1170
                                  1000
                                  1130
                                  1170
                                  1120
                                  1200
                                  1230
                                  1175
                                 (1175)
                                  3000
                                  3000
                                  3300
                                  3300
                                  3300
                                 (3300)
                  274

-------
        APPENDIX AllI
MEASURED HYDRAULIC PROPERTIES
             275

-------
HYDRAULIC MEASUREMENTS





     FLINT RIVER
;TA.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
FLOW
(CFS)
4.94
12.65
10.01
11.76
8.38
8*36
8.82
8.54
8.47
7.50
7.29
8.26
11.28
7.41
7.46
6.83
6.31
8.27
8.47
7.23
AREA
(SQ FT)
4.33
19,76
7,7o
9.41
17.46
36.33
41.99
71.13
24,91
12.30
11.21
10.32
19,79
8.23
6.60
5.99
6,4i*
6.5l
9,52
6.63
VEL,
(FPS)
1.14
.64
1.30
1,25
.48
,23
.21
.12
.34
.61
,65
,80
,57
.90
1.13
1.14
,98
1.27
.89
1.09
DEPTH
(FT)
.66
1.38
,59
,75
1,13
1,07
1*52
1.55
1,16
.72
,60
,54
1,15
.50
,62
,36
,33
.41
.65
,40
WET.PE
(FT)
6.86
17.24
13.18
12.81
17.02
34,37
27.74
46.31
21.79
17.54
18.94
19,27
17.50
17,55
10.78
16.55
19.64
16.17
14.84
16.53
           276

-------
HYDRAULIC MEASUREMENTS





     FllNT RIVER
STA.
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
FlOW
(CFS)
7t76
8.15
11.05
i0.2«*
7.44
7.08
14.74
14.74
15.76
15*63
15.68
11.15
16.1*9
11.58
14.12
2o.43
20.03
11.91
15.96
17.72
AREA
(SQ Fj)
13,85
15.26
32.5Q
18,61
13,53
7.45
14.89
12,39
15.60
15,79
15,68
8,32
15,86
8,2i
12, 5o
20.43
32,30
13.38
19, 70
30,55
VEL,
(FPS)
.56
,18
.34
,55
,55
.95
,99
1.19
1.01
.99
1.00
1.3*
1.04
1,41
1,13
1.00
.62
.89
.81
.58
DEPTH
(FT)
,66
1,92
1,61
1.74
.96
,56
.58
.53
,85
1,09
,95
,35
,63
1.26
1.47
1,56
2.78
1,91
1,06
1.70
WET.PER,
(FT)
21.13
24.67
20.98
11.9.0
14. <*8
13.51
26.10
23.57
18.36
80.31
16.61
23.84
25.42
6.94
8.74
14.85
13.98
7.28
18.85
18,78
          277

-------
HYDRAULIC MEASUREMENTS





     FilMT RIVER
STA.

4*0
4i
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
FLOW
(CFS)
15.12
13.47
17.21
12.68
12.96
81.74
2o,63
23.95
17.14
20.54
2Q.86
18.36
18*10
2Q.95
17.90
17.68
17.48
19.95
17.42
16.00
AREA
(SQ FT)
27.50
24.50
11.7g
18.38
16.83
34.5i
33.82
57.02
3lt7if
21.85
39.35
44.78
51.72
47.62
35,lo
35.36
19.00
30,70
31.10
25,40
VEL.
(FPS)
.55
.55
1.46
.69
.77
,63
.61
.42
.54
,94
.53
.41
.35
.44
.51
.50
,92
.65
,56
.63
DEPTH
(FT)
1.90
1,38
1,01
2.14
1,36
1.75
1,83
2,55
1.63
1,53
1.76
2,52
2.72
1,72
1.78
2,18
1,06
1.41
1,63
,94
WEEPER,
(FT)
U.91
18.67
12.40
8.92
13.66
20.83
18.96
23.20
19.99
15.76
23.10
19.43
20.09
28.91
20.83
17.65
18.51
23.18
19,66
27,34
         278

-------
HYDRAULIC MEASUREMENTS





     FLINT RIVER
STA.
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
FLOW

-------
HYDRAULIC MEASUREMENTS





     FLINT RIVER
5TA,

80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
9?
98
99
FLOW
(CFS)
21.53
2t*,89
21,06
2^,00
29.19
2Qti4i4
32.99
29.54
21.21
20.82
3Q.97
29.67
23.59
27,54
27. tO
26.H
26.18
32.38
29.55
23.71
AREA
(SQ FT)
52.5i
59.27
45,78
33,80
4Q,5tf
76.10
54.09
57,92
40.78
29,32
119,12
61.81
63.77
59,88
28.8(f
43.52
40.28
40.4?
40.48
48.38
VEL.
(FPS)
.41
.42
.46
.71
.72
.40
.61
,51
.52
.71
.26
.48
,37
.46
,95
,60
,65
.80
,73
.49
DEPTH
(FT)
1,94
2,02
1.64
1,21
1.84
1,99
1,52
2,02
1.27
1,20
1,97
1.14
2,82
1,39
1.34
.93
1.61
1.40
1,42
1.56
WET.PER,
(FT)
27,63
32,13
29,51
28.68
23,05
39,46
37,31
32.41
32.49
25.67
62,85
56.36
24.69
*7.18
24,20
47.93
26,47
31.03
29.41
32,02
          280

-------
                  HYDRAULIC  MEASUREMENTS


                       FLINT RIVER
STA.      FlOW     AREA       VEL,    DEPTH      WET  prb
         
-------
HYDRAULIC MEASUREMENTS




     SOUTH RIVER
} i A «
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
FLOW

-------
HroRAULIC MEASUREMENTS






     SOUTH RIVER
TA.
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
FLOW
(CFS)
56.35
55.66
50.89
42.00
39.19
39.61
77.30
75.52
78. 4o
94.44
86.39
I00.8ti
86. 8b
63.16
68.98
83.86
83.22
55.97
57.73
69.13
A^EA

-------
HYDRAULIC MEASUREMENTS






     SOUTH RIVER
TA.
4U
'41
42
43
44
45
46
47
•48
49
50
51
52
53
54
55
56
57
58
59
FLOW
(CFS)
70.313
75.77
75.89
79. 5c
57.6,}
56.00
61.4o
84. 7o
80.95
95.65
87.81
98.12
70. 2b
71.57
73.91
86. 36
75.3o
88.37
98.62
89.45
A"tA
(SQ FT)
48.52
79. 7 o
b3.24
74.32
88.66
53.33
46.93
t>9.25
57.40
71. 58
61.84
62.90
56.2D
36.80
48.51
58.75
54.96
69.58
59.41
63.68
VEL.
(FpS)
1.45
.95
1.20
1.07
.65
1.05
1,31
1.43
1.41
1.34
1.42
1.56
1.25
1.26
1.53
1.17
1.37
1.27
1.66
1.40
DEPTH
(FT)
1.20
1.99
1.56
1.73
2.14
1.29
1.25
1.54
1.71
1.39
1.03
1.03
1.10
1.14
.85
1.20
.98
1.18
1.06
1.10
WET. PER.
(FT)
41 .08
40.68
42.36
46.07
42.29
41.75
30.00
39.45
34.96
51.78
57.12
61.36
51.64
50.31
57.22
49.39
56.74
59.78
56.33
58.53
             284

-------
HYDRAULIC MEASUREMENTS






     SOUTH RIVER
;TA.
60
61
o2
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79

FLOw
ICFS)
78.21
67.45
70.00
68.75
93.42
98. Ob
93.46
92.34
64.32
64.60
64.18
71.41
70.90
66. 8b
65.6o
61.74
61.57
65,46
89.90
107.04

A^tA

-------
HYDRAULIC M




     SOUTH RIVER
TA.
,iO
.a
152
83
rt 4
d"5
cib
b7
88
69
90
91
92
93
94
95
9b
97
98
99
FLOW
(CFS)
77.04
89.27
100.55
114.30
107. Do
111.39
99.71
100.94
81.66
88.87
88. 4i
81. 4u
80.79
78. Ib
88. 8u
90.15
98.51
88.05
85.2o
88.3i
A^ta
(SO FT)
52.41
81.90
08.87
77.23
63. bM
128.05
63.11
107.38
71.63
56.97
t>4.53
61.67
74.12
52.11
59.60
52.41
63.97
59.90
o5.5b
73.15
Vfc.L,
(FVS)
1.47
1.09
1.46
1.48
1.28
.87
1.58
.94
1.14
1.56
1.37
1.32
1.09
1.5Q
1.49
1.72
1.54
1.47
1.30
1.13
uEPTH
(FT)
1.46
1.64
1.68
1..17
1.78
2. no
1.66
2.15
1.43
1.54
2.08
1.50
1.48
1.58
1.P.4
1.81
1.12
1.13
2.05
1.50
WET.PEK.
(FT)
37.33
51 .05
41.68
66.34
47.69
66.25
3R.49
50.90
50.40
37.91
33.95
41.89
51.37
33.91
48.56
30.23
57.55
53.41
32.41
53.54
          286

-------
HYDRAULIC MEASUREMENTS




     SOUTH RIVER
STA.
100
101
102
103
J -u4
105
106
Iu7
Iu8
109
110
111
112
113
114
115
116
117
118
119
FLOW
(CFSJ
81.31
32.15
88.74
84. 7b
106.80
100.46
88. 61)
91.6tJ
102. 7b
90.99
85. Ib
87.09
85.6e
120.01
106.11
107.34
101.30
102.02
97. 8b
103.91
A^LA

-------
HYDRAULIC MEASUREMENTS




     SOUTH RIVER
STA.
120
121
122
123
124
125
126
127
128
129
1,50
131
162
133
134
135
136
137
138
139
FLOw
(CFS)
114.99
128,91
97. 3b
95.90
98. 10
100.8U
93.9o
91.12
87. 2b
95.36
88.52
87.69
94,63
100.72
95.24
111.27
115.69
107.51
10b.22
100.90
A^LA
(SQ FT)
8.1.55
121.61
120.20
141. U3
150.93
120. OU
06. 65
78.55
77.90
113.52
95.18
130.88
181,93
167.87
170.07
191.84
93.30
182.22
96. bo
76.44
Vt£L.
(FpS)
1.41
1.06
.81
.68
.65
.84
i.41
1.16
1.12
.84
.93
.67
.52
.60
.56
.58
1.24
,59
1.10
1.32
UFPTH
(FT)
1.36
2.25
2.50
2.04
2,52
2.00
1.39
1.32
1.59
2.06
1.80
2.06
2.33
2.27
2.21
2.52
1.37
2.85
1.34
.93
WET. PER
(FT)
60.47
54.86
48.23
70.17
60.67
60.86
48.12
59.57
49.41
55.58
53.28
63.82
78.60
74.28
77.34
76.49
68.39
64.41
72.18
82.21
          288

-------
HYDRAULIC MEASUREMENTS




     SOUTH RIVER
STA.
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
FLOvv
(CFS)
116.84
111.31
101. 9o
109. 5b
109,59
104.73
122.1*
124.62
119.54
127.30
183.73
126.31
136. Ob
131.52
116.32
123.10
121.44
134.04
149.95
124.95
A^EA

-------
HYDRAULIC MEASUREMFNTS




     SOUTH RIVER
STA.

160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
FLOW
(CFS)
114.68
134.09
134.09
122.02
149.85
138.01
134.54
164.40
127.85
122.58
126.00
126.98
129.15
120.11
122.75
120.96
112.01
122.45
114.81
121.25
ARE. A

-------
HYDRAULIC MEASUREMENTS






     SOUTH RIVER
STA.
180
181
182
183
184
185
186
187
188
189
190
191
192
193
FLOW
(CFS)
.00
110.58
117.65
129.55
105.73
113.78
110.84
126.42
127.45
103.90
122.28
124.26
112.14
140.98
AREA
ISO FT)
.00
97.00
74.45
107.07
89.60
90.30
95.55
86.59
88.51
89.26
78. b9
82.84
88.30
114.62
VEL.
(FPS)
.00
1.14
1.58
1.21
1.18
1.26
1.16
1.46
1.44
1.22
1.55
1.50
1.27
1.23
DEPTH
(FT)
.00
1.67
1.26
1.78
1.57
1.29
1.47
1.40
1.05
1.21
1.29
1.34
1.40
1.85
WET. PER
(FT)
5.12
59.01
59.43
61 ,18
57.56
70.41
66.04
62.71
84.87
74.34
61.13
62. 9R
63.59
62.82
           291

-------
ST'A.

15
ib
17
18
19
20
21
22
23
24
25

910.05
908. 2b
914.2.0
915.50
941. 7u

A^L/,

109^.00
c39°.50
855. hO
701 .uO
726.30
695. uO
674.10
78^.00
653. uO
60°. 00
645.uO

Vi_L.
(FpS)
1.43
1.68
1.07
1.63
1.71
2.09
2.11
2-10
2-15
1.66
1.40
1.70
1.79
2.18
2.11
1.27
1.35
1.16
1.40
1.50
i.46
292
L/F.PTH
(FT)
7.23
5,66
5.30
6,43
5.69
4.81
4.57
4.76
4.35
5.45
5.75
5.20
4.58
3.89
4.54
4.34
4.29
4.89
3.63
3.95
4.03

WFT.PFK
(FT)
187. Ob
1^0.42
181 .34
17?. 54
185.5(4
16^.09
167.7/
161 .79
171.96
178.56
IQP.29
175. Qi
189.5^
181.31
161 .79
162.37
15R.3u
161 ,44
181.12
155.35
16 J.Ob


-------
                 HYDRAULIC
                  CH* i TAHOOcHEc
5T
   .
-— • * n *
36
37
08
59
40
41
42
43
44
45
46
47
+8
49
bli
51
52
53
54
55
r i_w»v
(CFS)
960.81
961.20
962.9,.
956.8^
981.25
986. '49
992.16
991.29
986.65
1002.75
877.82
881. Go
681.79
1052.55
881.79
861. 9u
859. 7u
863.25
8b2.99
844. 7b
A!^LA

-------
II M T^HOOrHLc RIV^P
TA. FL^rt A°c-,
(CFS) (iO hi)
Mb 858. 3o al'i.i.'O
•-;/ ' 858. 9 u 734. JO
58 855. bj 583.70
::-9 " 857.CH 405.40
|>U 866. ?U 71°.uO
,,,i 8o7.77 65~7.'tl
o2 871.0u 435.50
05 ' 869. Co 496. uO
OH 870.6/ 56^.50
o5 869.44 o08.;;,0
bb 1165.25 93°.oO
o7 1163. Id 66 q. i, 0
L-.8 1162.39 021 . o'l
•59 116'j.3^ />8".-.. 0
7 1.1 1155.90 oB'-'.vO
^'1 116'). CD 76^.^.0
I'd Ilfi2.3u 59^.110
7o 115'). 8u 41 i . jO
/4 H52.2i 57°. uO
75 Il4o.°9 75'!-. 60
VLL.
(FpS)
i
i
d
i
i
i
2
X
i
1
L
1
i
i
1
i
i
d
1
1
.67
.17
.23
.73
.22
.32
.00
.75
.54
.43
.25
.34
.87
.48
.69
.52
.97
.80
.99
.52
L/EPTH wr
(FT)
3
.06
1
T.PEK.
(FT)
6°
4.43 3ft?
2.
3
4
4
2
3
3
3
5
5
3
5
4
4
3
2
3
5
.35
.20
.67
.47
.59
.10
.43
.38
.57
.17
.84
.23
.38
.41
.69
.74
.62
.13
,0i
. ftu
163. 9j>
1
1
}
1
1
1
1
1.
1
1
1
1
56
53
i*R
68
60
66
81
69
71
63
51
57
175
.0/4
.61
.72
.9j
-7,,
.17
.64
.3,;
.78
.On
.85
.37
.07
161 .45
1
50
.98
lftl.2o
1
49
.35
         294

-------
HYDRAULIC MEA






 an n !\HOOcHLi£  RIV^P
^ 1 A .
76
77
78
79
dO
61
02
o3
o4
rJ5
66
ay
88
o9
9 0
91
92
93
9 4
95
hLU« AnLA
(CFS) (SQ KT)
1095.16 720.50
1095. 40 001-90
1100.2o o36..,0
1099. lu 506.50
1043.92 o5q.7n
1051.34 735.20
1051.54 o4°.in
1049.79 55°-. 40
1051.3J o53.oO
944.67 o51.5H
947.63 o4^.20
946.99 o5^.10
1059. 2^ 657.9ri
1058.79 710. DO
1058.95 67^.^0
1077.25 o95.uO
109^.24 747.10
1121. 9j 587.40
1133. 6«i 54'. 40
1005.60 o81.iO
Vt_L.
(FpS)
1.52
1.82
1.73
2-17
1.59
1.43
1.62
1.88
1.61
1.45
1.46
1.45
1.61
1.49
1.56
1.55
1.47
1.91
2-09
1.73
UFPTH
(FT)
4.26
3.52
3.83
3.47
4.12
4.90
3.84
3.10
4.08
4.34
4.48
4.50
3.94
4.33
4.65
4.06
4.39
4.45
3.31.
3.68
V'ET.PEu
(FT)
170.77
172,86
168.7u
147.26
161.95
15?.3t>
172. 4(j
181.24
162.31
152.2-.J
147.35
147.27
168.5.3
166.3.£
148.22
172.67
171.62
134.0v
165.52
162. 4b
         295

-------
HYDRAULIC MEAS
STA.
96
97
98
99
lOu
lul
M2
1 , 3
104
1-5
Iu6
107
1-6
109
llu
111
1 12
113
114
115
FLO*,
(CFS)
1007.51
1008.40
10Q6.45
842.34
841.15
840.67
841.55
. Ou
1900.31
1902.81
1884.6?
1785.03
1683.14
1570.6o
1471.65
1367.57
1296.00
1613. 6b
1553.2o
1492.40
A^tA
(SO FT)
451 .80
491.90
52^.70
478.60
486, 2 H
509. 50
56*. 60
.00
81^.10
361 . uO
981 .60
915.40
940.30
87?. 60
67?.i.,-0
94q. 70
8 1 o . g n
996.10
958.80
910.00
^fc.L.
(EpS)
2.23
2-05
1.90
1.76
1.73
1.65
1.48
.00
2.32
2.21
1^92
1.95
1.79
1.80
2.19
1.44
1.60
1.62
1.62
1.64
uEPTH
(FT)
3.09
3.22
3.56
3.23
3.29
3.29
3.95
.00
4.93
5.09
5.58
5.06
5.40
5.70
5.33
5.55
4.63
5.89
5.07
4.84
WET. PER
(ET)
148.94
156.09
152.74
148.93
149.33
156.69
145.51
.Ou
168. On
172.J.5
178.54
184.01
176.24
155.67
1^8.03
174,87
177. J 9
172.26
192.26
192.02
         296

-------
HYDRAULIC MEASUREMENTS
  H*ITAHOOcHbE RIVrR
STA.
116
117
1 13
119
120
121
122
123
124
125
.126
127
128
129
130
151
132
133
1-J4
135
FLOW
(CFS)
1426.11
1370.45
1313. 60
1247.32
1183.21
1121.3?
10bl.6u
1000,24
1101.15
1533.01
919.35
919.10
921.12
921.5o
916, 3u
9l2.6o
9l3.5u
913.25
913.42
1124.37
ADLA
(SQ FT)
93°. 10
736.80
746.40
67^. oO
71 ^.IT
75'. oO
717.50
d 1 3 . 2 0
754.20
92^.50
557.20
5 0 5 . u 0
606. f,-0
57°. 60
615. uO
57"-. UO
55n.50
543.60
54^.70
576.60
VtiL.
(FpS)
1.53
1.86
1.76
1.86
1.65
1.49
1.48
1.23
1.46
1.66
1.65
1.82
1.52
1.59
1.49
1.59
1.66
1.68
1.68
1.95
UEPTH
(FT)
5.75
4.09
4.22
4.66
4.54
4.62
4.43
4.78
4.01
4.55
2.81
2.30
3,52
2.97
3.27
2.76
2.79
2.63
2.44
2.93
Vi'FT.PEK
(FT)
164.31
182.10
179.86
146.9Q
160.59
165.32
162.2.7
172.45
190. 16
?04.3ii
19^,29
221.44
173.76
196.49
189. So
209. R9
198.75
20P.5Q
224. 9^
1.98.16
         29?

-------
HYDRAULIC MEASUREMENTS
SfA» FLO;*
(CFS)
1
1
1
1
1
1
1
1
56
57
38
59
40
Ml
42
43
144
1
1
1
1
1
1
1
1
1
1
1
45
<+6
47
nb
49
50
51
52
53
54
55
1
1
1
097
099
098
1070
1
1
1
1
1
1
1
1

1

1

1

1
057
043
047
050
047
0 5 1
048
117

12U

119

125

Iij3
.8*
.4b
.37
.3.5
.95
.24
.**
.70
.39
,6b
.9b
.lu
.00
.87
.00
.9u
.00
.02
.00
.15
A^LA
( SQ HI)
oOQ.90
591 -10
603.50
733.10
705.30
63^.30
60°. uO
644. D0
56^.10
653.20
49°. 50
09^.20
. 00
64^.90
.00
o7r|.bO
.uO
65°. 00
. 00
o4^.90
V£L.
(FpS)
1
1
i
1
1
1
1
1
1
1
2
i

i

1

1

1
.80
.86
.82
.46
.50
.65
.72
.63
.87
.61
.10
.88
.00
.73
.00
.67
.00
.73
.00
.70
DEPTH
(FT)
3
3
3
4
4
3
3
3
2
3
3
3

3

3

3

3
.08
.46
.66
.31
.30
.33
.01
.60
.81
.61
.03
.30
.00
.29
.00
.39
.00
.83
.00
,55
WET. PER
(FT)
2on.7o
173.00
167.22
172.0i
167.64
191. 80
2n3.3h
180. 7b
199.97
183.63
166.26
l«?.5o
.Oo
19P.39
.no
199.12
.00
170.91
.no
1*4. 2«

-------
HYDRAULIC M






 CHMTAHOOcHLE
STA.
156
157
358
159
160
161
162
163
164
165
166
167
168
169
1.70
171
J 72
173
174
175
FL.OIN
(CFS)
.00
1077.15
.00
1058.99
.Ou
1042.84
.00
1030. 8u
.Ou
1058.00
.00
1092.01
.00
1130.62
.Ou
1169.32
. Ou
1214.21
.Ou
1210.47
APLA
,0
VEL.
(FpS)
.00
i.64
.00
i.47
.00
i.41
.00
1.50
.00
1.78
.00
^.04
.00
J..76
.00
1 . 55
.00
1.53
.00
1.57
uEPTH
(FT)
.00
3.49
.00
5.4?
.00
4.57
.00
3.54
.00
2.99
,00
3.48
.00
3.61
.00
4.31
.On
4.22
.00
4.28
t-'FT.PEn.
(FT)
.00
189.82
.Oo
135.69
.Oo
J6".?5
.Ou
19'4.qrt
.Ou
200.19
.Oo
150.9/4
.Oy
363. 8 u
.Or,
177.Fu
.Ou
1«9.8i
.Ou
3«2.3ii
         299

-------
HYDDAuLIC '.'EMbUREMfNJ fS






 CHVf TAHOOcHhc  RIVFR
STA.

176
177
178
179
1dO
181
.1 82
183
1<34
185
186
187
1 8b
.189
.190
191
192
193
194
195
F|_0rt A^LA
(CFSJ (SQ FT)
.00 . .J 0
12o9.4o 817.20
.00 . (., 0
1303.01 886.40
.Go .GO
1302. 8b 81^.40
.00 . 0 0
1303.54 835.60
.00 .00
1296.17 815.20
.00 . o 0
1301.30 77^.00
.00 .tjO
.00 . u 0
13o3.6u 874.90
.00 . u o
.00 .oO
1260.54 62^.00
.00 .00
1227. 80 870.80
VLL.

-------
        APPENDIX AIV
DETAILED TRACER STUDY RESULTS
 AND REAERATION COEFFICIENTS
             301

-------
                                            FIELD TRACER DATA

                                               FLINT RIVER
      Dump
       No.
        II
LO
O
NJ
        III
Observed

Reach
0-1
0-2
1-2
1-2P
1-2
1-3
1-4
2P-2
2P-3
2P-4
2-3
2-4
3-4
0-1P
0-1
0-2P
0-2
0-3
1P-1
1P-2P
1P-2
1P-3
1-2P
1-2
1-3
Kr:H
Upstr
.2399
.2399
.1199
.2223
.2223
.2223
.2223
.1124
.1124
.1124
.0558
.0558
.0406
.1955
.1955
.1955
.1955
.1955
.1328
.1328
.1328
.1328
.0664
.0664
.0664
Ratio
Dnstr
.1199
.0290
.0290
.1124
.0558
.0406
.0153
.0558
.0406
.0153
.0406
.0153
.0153
.1328
.0664
.0331
.0152
.0112
.0664
.0331
.0152
.0112
.0331
.0152
.0112
Flow
Time
hrs
3.030
6.960
3.930
3.120
3.470
5.040
10.240
.350
1.920
7.120
1.570
6.770
5.200
2.720
2.798
6.050
6.470
7.998
.078
3.328
3.745
5.278
3.250
3.667
5.200
River
Temp.
°C
19.0
16.7
15.0
21.5
21.3
20.4
18.7
19.5
18.7
17.5
18.5
17.3
17.0
26.5
26.5
26.4
26.4
26.1
27.0
26.4
26.3
25.9
26.4
26.3
25.9
Tcr
River
Temp
.2287
.3031
.3605
.2183
.3983
.3371
.2609
2.0026
.5300
.2796
.2017
.1905
.1871
.1421
.3855
.2935
.3946
.3568
8.8729
.4173
.5785
.4674
.2144
.4020
.3413

K
ox
25°C
.3140
.4375
.5399
.2839
.5201
.4489
.3606
2.7195
.7324
.3966
.2800
.2714
.2683
.1658
.4496
.3430
.4612
.4197
10.2350
.4877
.6775
.5513
.2506
.4709
.4028
% Tracer Gas

lost
.5460
.9201
.8281
.5206
.7764
.8470
.9533
.5461
.6887
.9040
.3057
.7824
.6859
.3122
.6480
.8213
.9159
.9383
.4844
.7400
.8782
.9106
.4913
.7614
.8242
25°C
rem
.4539
.0798
.1718
.4793
.2235
.1529
.0466
.4538
.3112
.0959
.6942
.2175
.3140
.6877
.3519
.1786
.0840
.0616
.5155
.2599
.1217
.0893
.5086
.2385
.1757
WS Elev Chnge
ft
23.83
51.65
27.82
15.80
27.83
37.25
57.76
12.03
21.45
41.96
9.42
29.93
20.51
9.88
23.83
39.58
51.65
61.07
13.95
29.70
41.77
51.19
15.75
27.82
37.24
ft/hr
7.86
7.42
7.07
5.06
8.02
7.39
5.64
34.37
11.17
5.89
6.00
4.42
3.94
3.63
8.51
6.54
7.98
7.63
178.84
8.92
11.15
9.69
4.84
7.58
7.16

-------
                                            FIELD TRACER DATA


                                               FLINT RIVER
      Dump

       No.



        Ill
        IV
o
u>
Observed
Kr:H Ratio
Reach
2P-2
2P-3
2-3
3*-BS
3*-AS
3*-F
3*-4
3*-5
3*-6
BS-AS
BS-F
BS-4
BS-5
BS-6
AS-F
AS-4
AS-5
AS-6
F-4
F-5
F-6
4-5
4-6
5-6
Upstr
.0331
.0331
.0152
.1598
.1598
.1598
.1598
.1598
.1598
.1173
.1173
.1173
.1173
.1173
.0936
.0936
.0936
.0936
.0755
.0755
.0755
.0643
.0643
.0398
Dnstr
.0152
.0112
.0112
.1173
.0936
.0755
.0643
.0398
.0339
.0936
.0755
. 0643
.0398
.0339
.0755
.0643
.0398
.0339
.0643
.0398
.0339
.0398
.0339
.0339
Flow
Time
hrs
.417
1.950
1.533
1.430
2.900
3.420
4.950
9.450
14.770
1.470
1.990
3.520
8.020
13.340
.520
2.050
6.550
11.870
1.533
6.030
11.350
4.500
9.820
5.320
River
Temp.
°C
25.6
25.2
25.1
21.4
21.4
21.5
21.9
22.8
23.3
21.4
21.6
22.2
23.2
23.5
22.0
22.8
23.5
23.7
23.0
23.7
23.8
23.9
23.9
23.9
'"kr
River
Temp
1.8645
.5528
.1959
.2160
.1843
.2192
.1837
.1468
.1048
.1534
.2215
.1705
.1345
.0928
.4139
.1828
.1303
.0854
.1040
.1058
.0703
.1064
.0650
.0300

K
ox
25°C
2.2173
.6631
.2356
.2815
.2401
.2850
.2364
.1852
.1310
.1999
.2873
.2184
.1686
.1156
.5323
.2310
.1622
.1058
.1309
.1312
.0870
.1313
.0802
.0371
% Tracer Gas
25°C
lost
.5358
.6581
.2590
.2840
.4390
.5547
.6214
.7662
.7994
.2164
.3778
.4717
.6744
.7220
.2052
.3251
.5860
.6475
.1535
.4814
.5593
.3876
.4802
.1512
rem
.4641
.3418
.7409
.7159
.5609
.4452
.3785
.2337
.2005
.7835
.6221
.5282
.3255
.2779
.7947
.6748
.4139
.3524
.8464
.5185
.4406
.6123
.5197
.8487
WS Elev Chnge
ft
12.07
21.49
9.42
5.30
9.08
15.64
19.31
27.39
36.11
3.78
10.34
14.01
22.09
30.81
6.56
10.23
18.31
27.03
3.67
11.75
20.47
8.08
16.80
8.72
ft/hr
28.94
11.02
6.14
3.70
3.13
4.57
3.90
2.89
2.44
2.57
5.19
3.98
2.75
2.30
12.61
4.99
2.79
2.27
2.39
1.94
1.80
1.79
1.71
1.63

-------
                                             FIELD TRACER DATA

                                                FLINT RIVER
      Dump
       No.

        V
UJ
O
        XII
        XIV
Observed

Reach
4-5
4-6
4-7
5-6
5-7
6-7
4-5
4-6
4-7
5-6
5-7
6-7
0-R1
0-R2
0-R3
0-1P
0-1
0-2P
0-2
R1-R2
R1-R3
R1-1P
Rl-1
R1-2P
Rl-2
Kr:H
Upstr
.5729
.5729
.5729
.3728
.3728
.2419
.2952
.2952
.2952
.2075
.2075
.1329
.5392
.5392
.5392
.5392
.5392
.5392
.5392
.4044
.4044
.4044
.4044
.4044
.4044
Ratio
Dnstr
.3728
.2419
.1134
.2419
.1134
.1134
.2075
.1329
.0594
.1329
.0594
.0594
.4044
.3654
.3222
.3092
.1224
.0503
.0172
.3654
.3222
.3092
.1224
.0503
.0172
Flow
Time
hrs
4.550
9.900
17.730
5.350
13.180
7.830
4.670
10.240
18.320
5.570
13.650
8.080
1.750
1.800
1.870
4.740
4.810
8.880
9.340
.050
.117
2.990
3.060
7.130
7.590
River
Temp.
°C
24.0
24.1
24.2
24.2
24.3
24.3
22.0
22.3
22.6
22.5
22.8
23.0
24.0
24.0
24.0
24.3
24.3
24.1
24.1
24.0
24.0
24.5
24.5
24.1
24.1
Kr
River
Temp
.0944
.0870
.0913
.0808
.0902
.0967
.0754
.0779
.0875
.0800
.0916
.0996
.1642
.2160
.2752
.1172
.3081
.2670
.3688
2.0283
1.9425
.0897
.3904
.2922
.4159

K
ox
25°C
.1162
.1069
.1119
.0990
.1104
.1183
.0970
.0995
.1110
.1018
.1158
.1253
.2023
.2660
.3389
.1434
.3769
.3281
.4531
2.4975
2.3918
.1093
.4755
.3591
.5110
% Tracer Gas
25
lost
.3553
.5848
.8075
.3559
.7011
.5364
.3134
.5710
.8153
.3754
.7307
.5686
.2546
.3279
.4091
.4312
.7780
.9109
.9701
.0984
.2072
.2375
.7011
.8805
.9600
°C
rem
.6446
.4151
.1924
.6440
.2988
.4635
.6865
.4289
.1846
.6245
.2692
.4313
.7453
.6720
.5908
.5687
.2219
.0890
.0298
.9015
.7927
.7624
.2988
.1194
.0399
WS Elev
ft
7.83
16.71

8.88


7.92
16.00

8.08


3.49
5.56
7.93
10.04
23.86
39.89
51.74
2.07
4.44
6.55
20.37
36.40
48.25
Chnge
ft/hr
1.72
1.68

1.65


1.69
1.56

1.45


1.99
3.08
4.24
2.11
4.96
4.49
5.53
41.40
37.94
2.19
6.65
5.10
6.35

-------
                                            FIELD TRACER DATA


                                               FLINT RIVER
     Dump

      No.


       XIV
o
en
Observed

Reach
R2-R3
R2-1P
R2-1
R2-2P
R2-2
R3-1P
R3-1
R3-2P
R3-2
1P-1
1P-2P
1P-2
1-2P
1-2
2P-2
Kr:H
Upstr
.3654
.3654
.3654
.3654
.3654
.3222
.3222
.3222
.3222
.3092
.3092
.3092
.1224
.1224
.0503
Ratio
Dnstr
.3222
.3092
.1224
.0503
.0172
.3092
.1224
.0503
.0172
.1224
.0503
.0172
.0503
.0172
.0172
Flow
Time
hrs
.067
2.940
3.010
7.080
7.540
2.870
2.950
7.020
7.470
.075
4.150
4.600
4.070
4.530
.458
River
Temp.
°C
24.0
24.5
24.5
24.1
24.1
24.5
24.5
24.1
24.1
24.3
23.8
23.8
23.8
23.8
24.0
±xkr
River
Temp
1.8784
.0567
.3632
.2800
.4052
.0143
.3279
.2644
.3922
12.3532
.4375
.6279
.2184
.4331
2.3430

K
ox
25°C
2.3129
.0691
.4424
.3440
.4979
.0174
.3994
.3249
.4818
15.1119
.5410
.7766
.2701
.5357
2.8850
% Tracer Gas
25
lost
.1206
.1552
.6689
.8675
.9556
.0406
.6239
.8494
.9496
.6096
.8449
.9484
.5985
.8665
.6660
°C
rem
.8793
.8447
.3310
.1324
.0443
.9593
.3760
.1505
.0503
.3903
.1550
.0515
.4014
.1334
.3339
WS Elev Chnge
ft
2.37
4.48
18.30
34.33
46.18
2.11
15.93
31.96
43.81
13.82
29.85
41.70
16.03
27.88
11.85
ft/hr
35.37
1.52
6.07
4.84
6.12
.73
5.40
4.55
5.86
184.26
7.19
9.06
3.93
6.15
25.87

-------
                                           FIELD TRACER DATA


                                              SOUTH RIVER
     Dump

      No.

       VI
o
cr\
       VII
       VIII
Observed
Kr:H Ratio
Reach
A-B
A-D
A-E
A-F
B-D
B-E
B-F
D-E
D-F
E-F
G-H
G-J
G-K
G-L
H-J
H-K
H-L
J-K
J-L
K-L
J-K
J-L
J-M
J-N
K-L
Upstr
.5359
.5359
.5359
.5359
.3345
.3345
.3345
.2280
.2280
.1532
.5014
.5014
.5014
.5014
.3641
.3641
.3641
.1922
.1922
.1941
.5341
.5341
.5341
.5341
.4104
Dnstr
.3345
.2280
.1532
.1164
.2280
.1532
.1164
.1532
.1164
.1164
.3641
.1922
.1941
.1018
.1922
.1941
.1018
.1941
.1018
.1018
.4104
.2998
.2709
.2108
.2998
Flow
Time
hrs
2.480
3.850
6.280
7.980
1.370
3.800
5.500
2.430
4.130
1.700
.433
3.550
6.820
7.000
3.120
6.380
6.570
3.270
3.450
.184
2.720
3.000
4.850
7.300
.283
River
Temp.
°C
22.1
22.1
22.7
23.2
22.2
23.0
23.7
23.5
24.2
25.3
23.5
23.5
23.6
23.6
23.5
23.6
23.6
23.7
23.7
24.1
23.9
23.9
24.3
24.7
23.9
*kr
River
Temp
.1900
.2219
.1993
.1913
.2796
.2053
.1918
.1635
.1627
.1616
.7391
.2700
.1391
.2277
.2046
.0985
.1939
-.0029
.1843
3.5082
.0968
.1923
.1399
.1273
1.1088

K
ox
25°C
.2438
.2847
.2524
.2397
.3580
.2584
.2378
.2035
.1995
.1935
.9200
.3361
.1728
.2829
.2547
.1224
.2409
-.0036
.2284
4.3104
.1194
.2374
.1712
.1543
1.3668
% Tracer Gas
25°C
lost
.3946
.5974
.7318
.7955
.3344
.5574
.6623
.3367
.4954
.2389
.2815
.6285
.6240
.8067
.4830
.4771
.7312
-.0098
.4801
.4822
.2364
.4463
.4980
.6075
.2746
rem
.6053
.4025
.2681
.2044
.6655
.4425
.3376
.6632
.5045
.7610
.7184
.3714
.3759
.1932
.5169
.5228
.2687
1.0098
.5198
.5177
.7635
.5536
.5019
.3924
.7253
WS Elev Chnge
ft
12.07
24.00
36.47
44.81
11.93
24.40
32.74
12.47
20.81
8.34
7.67
21.88
24.03
36.15
14.21
16.36
28.48
2.15
14.27
12.12
2.14
14.31
19.02
28.02
12.17
ft/hr
4.86
6.23
5.80
5.61
8.70
6.42
5.95
5.13
5.03
4.90
17.71
6.16
3.52
5.16
4.55
2.56
4.33
.65
4.13
65.86
.78
4.77
3.92
3.84
43.00

-------
                                      FIELD TRACER DATA

                                         SOUTH RIVER
Dump
 No.
  VIII
   IX
Observed
Kr;H Ratio
Reach
K-M
K-N
L-M
L-N
M-N
E-F
E-G
E-H
E-J
E-K
F-G
F-H
F-J
F-K
G-H
G-J
G-K
H-J
H-K
J-K
A-B
A-D
A-E
A-F
B-D
B-E
Upstr
.4104
.4104
.2998
.2998
.2709
.4484
.4484
.4484
.4484
.4484
.3484
.3484
.3484
.3484
.2037
.2037
.2037
.1587
.1587
.0767
.4997
.4997
.4997
.4997
.3561
.3561
Dnstr
.2709
.2108
.2709
.2108
.2108
.3484
.2037
.1587
.0767
.0543
.2037
.1587
.0767
.0543
.1587
.0767
.0543
.0767
.0543
.0543
.3561
.2582
.1899
.1425
.2582
.1899
Flow
Time
hrs.
2.130
4.580
1.850
4.300
2.450
1.780
6.530
7.170
10.600
13.820
4.750
5.380
8.820
12.030
.633
4.070
7.280
3.430
6.650
3.220
2.420
3.720
6.130
7.830
1.300
3.720
River
Temp.
°C
24.7
25.1
24.8
25.2
25.5
23.0
23.7
23.7
23.9
24.0
23.9
23.9
24.0
24.2
24.3
24.2
24.4
24.2
24.3
24.5
26.2
26.3
26.9,
27.3
26.4
27.2
K.r
River
Temp
.1950
.1454
.0549
.0818
.1022
.1416
.1207
.1447
.1664
.1526
.1129
.1461
.1715
.1544
.3941
.2398
.1814
.2118
.1611
.1071
. 1399
.1775
.1578
.1601
.2474
.1689

K
ox
25°C
.2365
.1748
.0664
.0982
.1218
.1782
.1496
.1794
.2054
.1879
.1393
.1800
.2111
.1893
.4822
.2940
.2215
.2597
.1971
.1305
.1642
.2079
.1824
.1835
.2891
.1940
% Tracer Gas
25°C
lost
.3417
.4854
.0970
.2957
.2195
.2315
.5557
.6562
.8359
.8842
.4227
.5524
,7869
.8489
.2238
.6296
.7377
.5226
.6631
.2944
.2810
.4737
.6047
.6966
.2680
.4507
rem
.6582
.5145
.9029
.7042
.7804
.7684
.4442
.3437
.1640
.1157
.5772
.4475
.2130
.1510
.7761
.3703
.2622
.4773
.3368
.7055
.7189
.5262
.3952
.3033
.7319
.5492
WS Elev Chnge
ft
16.88
25.88
4.71
13.71
9.00
8.56
24.15
31.29
45.49
47.74
15.59
22.73
36.93
39.18
7.14
21.34
23.59
14.20
16.45
2.25
11.93
23.63
35.53
44.68
11.70
23.60
ft/hr->
7.92
5.65
2.54
3.19
3.67
4.80
3.69
4.36
4.29
3.45
3.28
4.22
4.18
3.25
11.27
5.24
3.24
4.13
2.47
.69
4.92
6.35
5.79
5.70
9.00
6.34

-------
                                            FIELD TRACER DATA

                                               SOUTH RIVER
      Dump
       No.

        X
        XI
U)
o
oo
        XIII
Observed

Reach
B-F
D-E
D-F
E-F
F-G
F-H
F-J
F-K
G-H
G-J
G-K
H-J
H-K
J-K
J-K
J-L
J-M
J-N
K-L
K-M
K-N
L-M
L-N
M-N
Kr:H
Upstr
.3561
.2582
.2582
.1899
.4591
.4591
.4591
.4591
.2969
.2969
.2969
.2181
.2181
.0964
1.3811
1.3811
1.3811
1.3811
1.2755
1.2755
1.2755
.7188
.7188
.6027
Ratio
Dnstr
.1425
.1899
.1425
.1425
.2969
.2181
.0964
.0721
.2181
.0964
.0721
.0964
.0721
.0721
1.2755
.7188
.6027
.3958
.7188
.6027
.3958
.6027
.3958
.3958
Flow
Time
hrs
5.410
2.420
4.110
1.690
4.630
5.250
8.680
12.250
.617
4.050
7.620
3.430
7.000
3.570
3.500
3.720
5.950
8.870
.217
2.450
5.370
2.230
5.150
2.920
River
Temp.
°C
27.8
27.7
28.2
28.9
25.9
25.9
26.0
26.4
25.8
26.2
26.7
26.3
26.8
27.2
23.5
23.5
23.9
24.1
24.0
24.4
24.5
24.5
24.5
24.5
*kr
River
m
.1692
.1268
.1444
.1697
.0941
.1417
.1797
.1510
.5000
.2775
.1856
.2378
.1580
.0813
.0227
.1755
.1393
.1408
2.6431
.3059
.2179
.0789
.1158
.1440

K
ox
25°C
.1918
.1441
.1623
.1878
.1112
.1675
.2118
.1765
.5921
.3258
.2155
.2785
.1830
.0934
.0282
.2185
.1719
.1731
3.2545
.3734
.2654
.0961
.1411
.1754
% Tracer Gas
25
lost
.5774
.2513
.4253
.2316
.3477
.5180
.7827
.8338
.2615
.6655
.7442
.5474
.6548
.2419
.0788
.4907
.5722
.7204
.4435
.5320
.6936
.1630
.4529
.3463
°C
rem
.4225
.7486
.5746
.7683
.6522
.4819
.2172
.1661
.7384
.3344
.2557
.4525
.3451
.7580
.9211
.5092
.4277
.2795
.5564
.4679
.3063
.8369
.5470
.6536
WS Elev
ft
32.75
11.90
21.05
9.15
15.67
23.16
37.35
39.58
7.49
21.68
23.91
14.19
16.42
2.23
2.17
14.17
19.26
28.26
12.00
17.09
26.09
5.09
14.09
9.00
Chnge
ft/hr
6.05
4.91
5.12
5.41
3.38
4.41
4.30
3.23
12.13
5.35
3.13
4.13
2.34
.62
.62
3.80T
3.23
3.19
55.29
6.97
4.86
2.28
2.74
3.08

-------
                                          FIELD TRACER DATA

                                             SOUTH RIVER
      XV
OJ
o
vo
Observed
Kr:H Ratio
Reach
G*-T2
G*-H
G*-J
G*-S
G*-K
G*-L
T2-H
T2-J
T2-S
T2-K
T2-L
H-J
H-S
H-K
H-L
J-S
J-K
J-L
S-K
S-L
K-L
Upstr
.5097
.5097
.5097
.5097
.5097
.5097
.4242
.4242
.4242
.4242
.4242
.4298
.4298
.4298
.4298
.2367
.2367
.2367
.2368
.2368
.1981
Dnstr
.4242
.4298
.2367
.2368
.1981
.1233
.4298
.2367
.2368
.1981
.1233
.2367
.2368
.1981
.1233
.2368
.1981
.1233
.1981
.1233
.1233
Flow
Time
hrs
.203
.462
3.670
4.470
7.270
7.480
.259
3.460
4.260
7.060
7.280
3.210
4.010
6.810
7.020
.800
3.600
3.820
2.800
3.020
.217
River
Temp.
°C
17.5
17.1
17.5
17.8
18.6
18.6
16.8
17.6
17.9
18.6
18.7
17.6
17.9
IS. 7
18.7
19.0
19.6
19.6
19.8
19.8
20.3
tcr
River
Temp
.9037
.3690
.2089
.1714
.1299
.1897
-.0500
.1686
.1368
.1078
.1697
.1858
.1435
.1136
.1778
-.0007
.0493
.1707
.0636
.2161
2.1868

K
ox
25°C
1.2819
.5280
.2964
.2416
.1799
.2627
-.0721
.2386
.1924
.1493
.2345
.2629
.2089
.1571
.2458
-.0010
.0669
.2313
.0859
.2916
2.9185
% Tracer Gas
25°C
lost
.1942
.1832
.5946
.5919
.6623
.8043
-.0156
.4961
.4935
.5831
.7576
.5037
.5011
.5885
.7612
-.0006
.1811
.5198
.1810
.5186
.4088
rem
.8057
.8167
.4053
.4080
.3376
.1956
1.0156
.5038
.5064
.4168
.2423
.4962
.4988
.4114
.2387
1.0006
.8188
.4801
.8189
.4813
.5911
WS Elev Chnge
ft
4.86
5.75
20.02
21.31
22.20
34.27
.89
15.16
16.45
17.34
29.41
14.27
15.56
16.45
28.52
1.29
2.18
14.25
.89
12.96
12.07
ft/hr-
23.94
12.45
5.46
4.77
3.05
4.58
3.43
4.38
3.86
2.45
4.03
4.44
3.88
2.41
4.06
1.61
.60
3.73
.31
4.29
55.62

-------
                                              FIELD TRACER DATA

                                               PATUXENT RIVER
       Dump
        No.
U)
M
O
Observed
Kr:H Ratio
Reach
4-5
4-6
4-7
5-6
5-7
6-7
1-2
1-3
1-4
2-3
2-4
3-4
4-5
4-6
4-7
5-6
5-7
6-7
1-2
1-3
1-4
2-3
2-4
3-4
Upstr
.7960
.7960
.7960
.3850
.3850
.1990
.8700
.8700
.8700
.5390
.5390
.3650
.8240
.8240
.8240
.4350
.4350
.2720
.7430
.7430
.7430
.5230
.5230
.3350
Dnstr
.3850
.1990
.1040
.1990
.1040
.1040
.5390
.3650
.2110
.3650
.2110
.2110
.4350
.2720
.1550
.2720
.1550
.1550
.5230
.3350
.1970
.3350
.1970
.1970
Flow
Time
hrs
6.700
13.920
17.520
7.220
10.820
3.600
3.850
9.080
14.750
5.230
10.900
5.670
5.280
11.260
14.490
5.980
9.210
3.230
2.750
7.270
11.940
4.520
9.190
4.670
River
Temp.
°C
21.6
21.8
21.7
22.0
21.8
21.5
22.2
22.4
22.5
22.5
22.6
22.6
20.8
21.3
21.3
21.8
21.7
21.4
21.3
20.8
20.2
20.5
19.9
19.3
Kr
River
Temp
.1084
.0995
.1161
.0914
.1209
.1802
.1243
.0956
.0960
.0745
.0860
.0966
.1209
.0984
.1153
.0785
.1120
.1741
.1276
.1095
.1111
.0985
.1062
.1136

K
ox
25°C
.1406
.1286
.1503
.1175
.1562
.2343
.1592
.1219
.1221
.0948
.1092
.1226
.1597
.1285.
.1505
.1014
.1450
.2268
.1667
.1446
.1487
.1309
.1430
.1550
% Tracer Gas
25
lost
.5425
.7737
.8877
.5056
.7542
.5035
.3988
.6011
.7759
.3374
.6277
.4386
.5033
.6991
.8364
.3955
.6700
.4556
.3165
.5822
.7709
.3881
.6641
.4517
°C
rem
.4574
.2262
.1122
.4943
.2457
.4964
.6011
.3988
.2240
.6625
.3722
.5613
.4966
.3008
.1635
.6044
.3299
.5443
.6834
.4177
.2290
.6118
.3358
.5482
WS Elev Chnge
ft
14.70
23.20
29.60
8.50
14.90
6.40
7.00
11.80
21.90
4.80
14.90
10.10
14.70
23.20
29.60
8.50
14.90
6.40
7.00
11.80
21.90
4.80
14.90
10.10
ft/hr
2.19
1.66
1.68
1.17
1.37
1.77
1.81
1.29
1.48
.91
1.36
1.78
2.78
2.06
2.04
1.42
1.61
1.98
2.54
1.62
1.83
1.06
1.62
2.16

-------
                                           FIELD TRACER DATA

                                            PATUXENT RIVER
      Dump
       No.
Observed
Kr:H Ratio
Reach
4-5
4-6
4-7
5-6
5-7
6-7
Upstr
.6370
.6370
.6370
.3100
.3100
.1920
Dnstr
.3100
.1920
.1320
.1920
.1320
.1320
Flow
Time
hrs
7.620
15.770
19.730
8.150
12.110
3.960
River
Temp.
°C
17.3
17.4
17.2
17.4
17.2
16.8
Kr
River
Temp
.0945
.0760
.0797
.0587
.0705
.0946

K
ox
25°C
.1346
.1081
.1138
.0835
.1006
.1362
% Tracer Gas
25
lost
.5732
.7570
.8451
.4317
.6364
.3610
°C
rem
.4267
.2429
.1548
.5682
.3635
.6389
WS Elev Chnge
ft
14.70
23.20
29.60
8.50
14.90
6.40
ft/hr
1.92
1.47
1.50
1.04
1.23
1.61
U)

-------
                                             FIELD TRACER DATA

                                               JACKSON RIVER
      Dump
       No.

        10
co
M
NJ
        11
Observed

Reach
0-1
0-2
0-3
0-4
1-2
1-3
1-4
2-3
2-4
3-4
0-1
0-2
0-3
0-4
1-2
1-3
1-4
2-3
2-4
3-4
Kr:H
Upstr
.3220
.3220
.3220
.3220
.2510
.2510
.2510
.1600
.1600
.0760
1.6780
1.6780
1.6780
1.6780
1.3000
1.3000
1.3000
.6840
.6840
.4000
Ratio
Dnstr
.2510
.1600
.0760
.0558
.1600
.0760
.0558
.0760
.0558
.0558
1.3000
.6840
.4000
.3040
.6840
.4000
.3040
.4000
.3040
.3040
Flow
Time
hrs
1.650
5.970
12.670
17.090
4.320
11.020
15.440
6.700
11.120
4.420
1.850
6.580
12.780
17.380
4.730
10.930
15.530
6.200
10.800
4.600
River
Temp.
°C
35.0
32.8
31.3
31.0
32.0
30.8
30.6
30.0
30.0
30.0
35.0
32.8
31.5
31.1
32.0
30.9
30.6
30.0
30.0
30.0
Kr
River
Temp
.1509
.1171
.1139
.1025
.1042
.1084
.0973
.1111
.0947
.0699
.1379
.1363
.1121
.0982
.1357
.1078
.0935
.0865
.0750
.0596

K
ox
25°C
.1463
.1191
.1197
.1084
.1078
.1151
.1038
.1200
.1023
.0755
.1337
.1386
.1173
.1037
.1404
.1142
.0997
.0935
.0811
.0644
% Tracer Gas
25
lost
.1815
.4457
.7160
.7852
.3206
.6511
.7358
.4871
.6112
.2420
.1856
.5310
.7119
.7759
.4238
.6453
.7237
.3819
.5168
.2181
°C
rem
.8184
.5542
.2839
.2147
.6793
.3488
.2641
.5128
.3887
.7579
.8143
.4689
.2880
.2240
.5761
.3546
.2762
.6180
.4831
.7818
WS Elev
ft
8.20
17.50
32.00
39.40
9.30
23.80
31.20
14.50
21.90
7.40
8.20
17.50
32.00
39.40
9.30
23.80
31.20
14.50
21.90
7.40
Chnge
ft/hr
4.96
2.93
2.52
2.30
2.15
2.15
2.02
2.16
1.96
1.67
4.43
2.65
2.50
2.26
1.96
2.17
2.00
2.33
2.02
1.60

-------
                                           FIELD TRACER  DATA
                                             JACKSON RIVER
     Dump
      No.
       12
CO
I—'
co
       13
Observed

Reach
4-5
4-6
4-7
4-8
5-6
5-7
5-8
6-7
6-8
7-8
4-5
4-6
4-7
4-8
5-6
5-7
5-8
6-7
6-8
7-8
Kr:H
Upstr
1.6680
1.6680
1.6680
1.6680
1.1230
1.1230
1.1230
.7720
.7720
.5120
.4260
.4260
.4260
.4260
.2440
.2440
.2440
.1640
.1640
.1100
Ratio
Dnstr
1.1230
.7720
.5120
.2960
.7720
.5120
.2960
.5120
.2960
.2960
.2440
.1640
.1100
.0563
.1640
.1100
.0563
.1100
.0563
.0563
Flow
Time
hrs
2.920
7.370
10.750
15.330
4.450
7.830
12.410
3.380
7.960
4.580
3.030
7.630
10.800
16.880
4.600
7.770
13.850
3.170
9.250
6.080
River
Temp.
°C
28.0
28.6
28.7
28.8
29.0 -
29.0
29.0
29.0
29.0
29.0
28.0
28.0
28.3
28.5
28.0
28.4
28.7
29.0
29.0
29.0
"kr
River
Temp
.1354
.1045
.1098
.1127
.0842
.1003
.1074
.1214
.1204
.1196
.1839
.1251
.1253
.1198
.0863
.1025
.1058
.1259
,1155
.1101

ox
25°C
.1529
.1164
.1221
.1251
.0930
.1107
.1186
.1341
.1330
.1321
.2075
.1412
.1405
.1338
.0974
.1147
.1177
.1391
.1276
.1216
% Tracer
25°C
lost
.3096
.5095
.6636
.7964
.2907
.5132
.7054
.3136
.5846
.3948
.4066
.5910
.7163
.8466
.3107
.5228
.7415
.3065
.6247
.4587
Gas


WS Elev Chnge
rem
6903
4904
3363
2035
7092
4867
2945
6863
4153
6051
5933
4089
2836
1533
6892
4771
2584
6934
3752
5412
ft
8.80
16.80
23.00
38.50
8.00
14.20
29.70
6.20
21.70
15.50
8.80
16.80
23.00
38.50
8.00
14.20
29.70
6.20
21.70
15.50
ft/hr
3.01
2.27
2.13
2.51
1.79
1.81
2.39
1.83
2.72
3.38
2.90
2.20
2.12
2.28
1.73
1.82
2.14
1.95
2.34
2.54

-------
                                              FIELD TRACER DATA
                                                JACKSON RIVER
       Dump
        No.
         14
U)
         15
Observed
Kr:H Ratio
Reach
7-8
7-9
7-10
7-11
8-9
8-10
8-11
9-10
9-11
10-11
8-9
8-10
8-11
8-12
9-10
9-11
9-12
10-11
10-12
11-12
Upstr
.4180
.4180
.4180
.4180
.2480
.2480
.2480
.1500
.1500
.0686
.2950
.2950
.2950
.2950
.1910
.1910
.1910
.0825
.0825
.0621
Dnstr
.2480
.1500
.0686
.0490
.1500
.0686
.0490
.0686
.0490
.0490
.1910
.0825
.0621
.0330
.0825
.0621
.0330
.0621
.0330
.0330
Flow
Time
hrs
4.520
8.900
11.650
14.430
4.380
7.130
9.910
2.750
5.530
2.780
3.900
6.530
9.310
14.360
2.630
5.410
10.460
2.780
7.830
5.050
River
Temp.
°C
29.0
27.0
26.6
26.3
25.0
25.0
25.0
25.0
25.0
25.0
24.0
24.4
24.9
25.6
25.0
25.5
26.2
26.0
26.6
27.0
Tcr
River
Temp
.1154
.1151
.1551
.1485
.1147
.1802
.1636
.2844
.2023
.1210
.1114
.1951
.1673
.1525
.3191
.2076
.1678
.1021
.1170
.1251

K
ox
25°C
.1275
.1328
.1805
.1739
.1383
.2171
.1971
.3427
.2437
.1458
.1372
.2381
.2020
.1813
.3845
.2475
.1970
.1204
.1361
.1444
% Tracer Gas
25
lost
.3803
.6251
.8254
.8755
.3951
.7233
.8024
.5426
.6733
.2857
.3587
.7249
.7902
.8849
.5680
.6708
.8192
.2426
.5872
.4540
°C
rem
.6196
.3748
.1745
.1244
.6048
.2766
.1975
.4573
.3266
.7142
.6412
.2750
.2097
.1150
.4319
.3291
.1807
.7573
.4127
.5459
WS Elev Chnge
ft
15.50
24.70
37.80
47.90
9.20
22.30
32.40
13.10
23.20
10.10
9.20
22.30
32.40
47.60
13.10
23.20
38.40
10.10
25.30
15.20
ft/hr
3.42
2.77
3.24
3.31
2.10
3.13
3.26
4.76
4.19
3.63
2.35
3.42
3.48
3.31
4.98
4.28
3.67
3.63
3.23
3.00

-------
                                           FIELD  TRACER DATA

                                             JACKSON RIVER
      Dump
       No.

       16
to
M
Ul
       17
Observed
Kr:H Ratio
Reach
10-11
10-12
10-13
10-14
11-12
11-13
11-14
12-13
12-14
13-14
11-12
11-13
11-14
11-15
12-13
12-14
12-15
13-14
13-15
14-15-
Upstr
.3060
.3060
.3060
.3060
.2310
.2310
.2310
.1250
.1250
.0661
.2900
.2900
.2900
.2900
.1420
.1420
.1420
.0828
.0828
.0481
Dnstr
.2310
.1250
.0661
.0451
.1250
.0661
.0451
.0661
.0451
.0451
.1420
.0828
.0481
.0346
.0828
.0481
.0346
.0481
.0346
.0346
Flow
Time
hrs
2.420
7.300
9.950
13.480
4.880
7.530
11.060
2.650
6.180
3.530
4.780
7.900
11.520
15.250
3.120
6.740
10.470
3.620
7.350
3.730
River
Temp.
°C
25.0
25.0
25.3
25.7
25.0
25.4
25.9
26.0
26.6
27.0
24.0
24.0
24.3
25.0
24.0
24.5
25.4
25.0
26.0
27.0
^r
River
Temp
.1161
.1226
.1540
.1420
.1258
.1661
.1476
.2404
.1649
.1082
.1493
.1586
.1559
.1394
.1728
. 1606
.1348
.1500-
.1187
.0883

ox
25°C
.1399
.1477
.1843
.1685
.1516
.1984
.1744
.2834
.1919
.1249
.1839
.1953
.1907
. 1679
.2128
.1956
.1610
.1807
.1399
.1018
% Tracer Gas
25°C
lost
.2450
.5915
.7818
.8482
.4588
.7107
.7984
.4638
.6263
.3065
.5179
.7222
.8386
.8806
.4237
.6652
.7533
.-4190
.5742
.27 as
rem
.7549
.4084
.2181
.1517
.5411
.2892
.2015
.5361
.3736
.6934
.4820
.2777
.1613
.1193
.5762
.3347
.2466
.5809-
.4257
.7294
WS Elev Chnge
ft
10.10
25.30
35.90
45.10
15.20
25.80
35.00
10.60
19.80
9.20
15.20
25.80
35.00
45.10
10. &0
19.80
29^.90
9 . 20
Ir-9.30
im 10
ft/hr
4.17
3.47
3.61
3.35
3.11
3.42
3.16
4.00
3.20
2.60
3.17
3.26
3.03
2.95
3.39
2.93
2.85
2.54
2'. 62
2.70

-------
                                           FIELD TRACER DATA

                                          CHATTAHOOCHEE RIVER
LO
M
(T\
      Dump
       No.

      XVI

      XVII
      XIX
      XX
      XXI
      XXII
Reach
0-2
2-3
2-4
3-4
0-2
0-3
2-3
0-2
0-3
2-3
3-4
3-5
4-5
3-4
3-5
3-6
4-5
4-6
5-6
Upstr
.6507
.3336
.3336
.2804
.8958
.8958
.5956
1.0495
1.0495
.8867
.4446
.4446
.3975
.7214
.7214
,7214
.7193
.7193
.6490
Dnstr
.4714
.2804
.2825
.2825
.5956
.5083
.5083
.8867
.8515
.8515
.3975
.3699
.3699
.7193
.6490
.6432
.6490
.6432
.6432
5.060

3.300
5.330
2.030

4.930
8.730
3.800

5.440
9.320
3.880
 ,650
 ,010
 ,360

 ,760
 ,180
 ,580
 ,420
 ,820
                                           2.400
River
Temp.
°C
25.1
20.2
20.3
20.5
25.0
25.6
26.3
27.0
27.4
28.0
25.2
25.4
25.5
21.5
21.7
22.0
21.8
22.1
22.5
Ttr
River
Temp
.0637
.0526
.0311
-.0036
.0827
,0649
.0417
.0309
.0224
.0104
.0678
.0367
.0214
.0016
.0204
.0151
.0300
.0192
.0037

K
ox
25°C
.0765
.0704
.0416
.0048
.0997
.0771
.0488
.0357
.0256
.0117
.0814
.0438
.0255
.0021
.0264
.0194
.0388
.0246
.0047
% Tracer Gas

lost
.2750
.1753
.1682
-.0082
.3351
.4283
.1427
.1490
.1799
.0372
.1055
.1666
.0687
.0031
.1074
.1152
.1044
.1122
.0094
25°C
rem
.7249
.8246
.8317
1.0082
.6648
.5716
.8572
.8509
.8200
.9627
.8944
.8333
.9312
.9968
.8925
.8847
.8955
.8877
.9905
WS Elev
ft
6.53
3.80
6.53
2.73
6.69
10.46
3.77
7.24
10.63
3.39
2.73
5.57
2.84
2.70
5.59
6.28
2.89
3.58
.69
Chnge
ft/hr
1.29
1.15
1.22
1.34
1.35
1.19
.99
1.33
1,14
.87
1.65
1.11
.84
1.53
1.07
.82
.84
.61
.28

-------
                                            FIELD TRACER DATA

                                           CHATTAHOOCHEE RIVER
        Dump
         No.

       XXIII
oo
       XXIV
       XXV
Observed

Reach
-2-3
2-4
2-5
3-4
3-5
4-5
3-4
3-5
3-6
4-5
4-6
5-6
0-2
0-4
0-5
2-4
2-5
4-5
Kr:H
Upstr
.7954
.7954
.7954
.7173
.7173
.6230
.4813
.4813
.4813
.4509
,4509
.4089
.5930
.5930
.5930
.5578
.5578
.4947
Ratio
Dnstr
.7173
.6230
.5663
.6230
.5663
.5663
.4509
.4089
.3867
.4089
.3867
.3867
.5578
.4947
.4617
.4947
.4617
.4617
Flow
Time
hrs
3.270
5.390
9.450
2.120
6.180
4.060
1.700
5.190
7.340
3.490
5.640
2.150
3.430
7.530
10.080
4.100
6.650
2.550
River
Temp.
°C
22.7
22.7
23.1
22.8
23.3
23.5
22.0
22.5
22.6
22.7
22.8
23.0
9.7
9.9
9.9
10.0
10.0
10.0
*kr
River
Temp
.0316
.0453
.0359
.0664
.0382
.0235
.0383
.0314
.0298
.0280
.0272
.0259
.0178
.0240
.0248
.0292
.0284
.0270
K % Tracer Gas
ox
25°C
.0400
.0574
.0451
.0840
.0478
.0292
.0493
.0399
.0378
.0354
.0344
.0326
.0299
.0402
.0415
.0488
.0474
.0452
25
lost
.1029
.2265
.2981
.1374
.2175
.0938
.0672
.1581
.2059
.0976
.1488
.0566
.0818
.2225
.2936
.1532
.2305
.0912
°C
rem
.8970
.7734
.7018
.8625
.7824
.9061
.9327
.8418
.7940
.9023
.8511
.9433
.9181
.7774
.7063
.8467
.7694
.9087
WS
ft
3.80
6.62
9.47
2.82
5.67
2.85
2.62
5.56
6.77
2.94
4.15
1.21
7.73
13.25
17.79
5.52
10.06
4.54
Elev Chnge
ft/hr
1.16
1.22
1.00
1.33
.91
.70
1.54
1.07
.92
.84
.73
.56
2.25
1.75
1.76
1.34
1.51
1.78

-------
                               Subject Field & Group
                            05F, 05C
                                           SELECTED WATER RESOURCES ABSTRACTS
                                                  INPUT TRANSACTION  FORM
    c   Organization
   	1     Georgia Institute of Technology
           Atlanta, Georgia
           CHARACTERIZATION OF STREAM REAERATION CAPACITY
   ] Q  Authors)
  	Tsivoglou, E. C.
       Wallace, J. R.
                                lit  Project Designation
                                       Project 16050 EDT
   22
               Environmental Protection Agency  report
               number EPA-R3-72-012, October, 1972.
       Descriptors (Starred First)
        *Reaeration, *Streams, *Stream Pollution, *Streamflow Hydraulics, *Tracers,
        *Radioisotopes, *Dissolved Oxygen
   25
Identifiers (Starred First)
 *Stream Self-Purification, *Flint  River,  *South River,  *Patuxent River,  *Jackson
 River, *Chattahoochee River, *Turbulence,  *Mixing,  *Gas Transfer,  *Stream  Hydraulic
 Properties
   27
      Abstract  The purposes of this research have been to characterize stream reaeration
       capacity in terms of the stream hydraulic properties and to develop procedures for
evaluating  the effects of pollutants on reaeration.  Field studies of the reaeration capacity
and the associated hydraulic properties of five rivers have been completed, using a gaseous
tracer procedure  for field measurement of reaeration.  These studies have incorporated a wide
range  of hydraulic features, such as waterfalls, rapids, shoals and pools, with stream flows
ranging from  5 to 3,000 cfs.  The range of BOD's and temper'atures encountered was also large.
Studies of  the effects of both pure substances and community wastes on the reaeration capacity
have been conducted in a newly designed test system.  Tests of observed vs. predicted values
of K2  have  shown  that none of the available models, e.g., O'Connor, Churchill, etc., is
capable of  providing dependable predictions of stream reaeration capacity, especially under
highly turbulent  flow conditions.  A new energy dissipation model has been derived, by which
the reaeration capacity of a stream is explained in terms of the rate of energy dissipation,
measured as the loss of water surface elevation divided by the time of flow.  Two distinct
forms  of the  energy dissipation model have been tested against the observed results, and it
has been shown that both forms provide dependable predictions of stream reaeration capacity.
The tests of  pollutant effects have shown that LAS and community wastes decrease the reaera-
tion rate coefficient,  pure NTA has no effect, and pure mineral oil increases the reaeration
rate coefficient.   (Tsivoglou-Georgia Tech)
  Abstractor
          E.  C.  Tsivoglou
                                 Georgia Institute  of  Technology
   WR-102  (REV JULY 1969)
   WRSI C

-------