MANAGEMENT AND MEASUREMENT OF "DO" IN IMPOUNDMENTS
James M. Symons, William H. Irwin, Robert M.  Clark,  and Gordon G.  Robeck
                     U. S. Department of the Interior
               Federal Water Pollution Control Administration
                         Cincinnati, Ohio 45226
                             September 1966

-------
           MANAGEMENT AND MEASUREMENT OF "DO" IN IMPOUNDMENTS





James M. Symons, William H. Irwin, Robert M. Clark, and Gordon G.  Robeck








        This paper develops a basis for calculating the hydraulic




 efficiency of any mechanical destratification device by the use of




 the stability concept and a mathematical model, including a solution,




 for studying the DO budget.  Also, the influence of mixing on reduced




 substances and DO resources in stratified impoundments is demonstrated.




                       INFORMATION ABSTRACT




        This paper recommends a basis for the comparison of various




 mechanical devices used for artificially destratifying impoundments by




 calculating their hydraulic efficiency based on impoundment stability.




 Three problems with this calculation are discussed to avoid misinter-




 pretation, influence of size and geometry on stability, natural changes




 in stability with time, and the basic hydraulic inefficiency of spring




 mixing if cold water is raised.  Also, a mathematical model is developed




 for studying the DO budget in impoundments.  A solution to the model is




 presented, using specially designed DO probe equipment.  Finally,




 data is presented on the destratification of a 2930-ac-ft impoundment




 in northern Kentucky by pumping.  The hydraulic efficiency of the pump




 is calculated and the influence of the mixing on the reduced substances




 and the DO resources in the impoundment is shown.  The most favorable




 time for destratification is also discussed.  A bibliography on artificial




 destratification is included.




 KEY WORDS




 Impoundments, Stratification, Destratification, Water Quality, Stability,




 Mechanical Pumping, DO Resources

-------
        MANAGEMENT AND MEASUREMENT OF "DO" IN IMPOUNDMENTS8
               1                          23
James M. Symons ,  M.ASCE, William H.  Irwin ,  Robert M.  Clark ,  A.M.  ASCE
                                       4
                   and Gordon G.  Robeck ,  F.  ASCE
                          INTRODUCTION


       The problem of thermal stratification in storage reservoirs is

widespread throughout the United States.   Considerable attention has

been given to developing engineering methods that will prevent downstream

problems from the discharge of poor quality bottom water from impoundments,

particularly during peaking power operations.  Several possible design

features that might be incorporated into dams and outlet works to

prevent the escape of poor quality water from impoundments have been

extensively reviewed by Kittrell .  One of these techniques, artificial

destratification, has been researched during the past two years by

staff members of the Cincinnati Water Research Laboratory.  The first

a - Presented at the ASCE National Symposium on Quality Standards for
    Natural Waters, Ann Arbor, Michigan, July 19-22, 1966.
Submitted for publication in the Proceedings of the Symposium and in the
Journal Sanitary Engineering Division of the American Society of Civil Engineers,

1.  In Charge, Impoundment Behavior Studies, Engineering Activities,
Research and Development, Cincinnati Water Research Laboratory, FWPCA
Cincinnati, Ohio

2.  Aquatic Biologist, Impoundment Behavior Studies, Engineering Activities,
Research and Development, Cincinnati Water Research Laboratory, FWPCA
Cincinnati, Ohio.

3.  Applied Mathematician, Research and Development, Cincinnati Water
Research Laboratory, FWPCA, Cincinnati, Ohio.

4.  Chief, Engineering Activities, Research and Development, Cincinnati
Water Research Laboratory, FWPCA, Cincinnati, Ohio
This work was performed while the authors were members of the Division
of Water Supply and Pollution Control, Public Health Service, Department
of Health, Education, and Welfare.

-------
                               - 2 -
publication on this subject ,  discussed the theoretical aspects of




artificially breaking the thermal stratification of an impoundment,




and presented early experimental results.   The second paper  dealt with




a larger field trial, and extensive data were presented to demonstrate




the improvement in water quality that could be affected by pumping cold




water from the bottom of a stratified impoundment and discharging it




in the surface, thereby creating an overturn in the body of water.




       Although to date (1966) the method has only been tried on




relatively small bodies of water, 3,000-4,000 ac-ft, artificial




destratification is an effective method for water quality improvement.




This technique is doubly attractive since it improves the water quality




both in the reservoir itself as well as protecting downstream quality.




This can be contrasted to some of the techniques discussed by Kittrell




that only prevent poor quality water from escaping an impoundment, and




do not prevent the formation of poor quality water in the impoundment




itself.




       Since research on this process will continue, one purpose of




this article, the third in the series, is to establish a firm theoretical




basis for comparing the hydraulic efficiencies of various pieces of




equipment that might be used for artificially destratifying a body of




water.  Secondly, although dissolved oxygen  (DO) is not the only




quality parameter in natural waters, it certainly is an important one.




Therefore, the other purpose of  this article is to examine the  influence




that artificial destratification has on the DO resources in an  impoundment

-------
                               - 3 -
from four aspects;  a) the elimination of reduced substances,




b) the addition and redistribution of DO, c) when artificial




destratification should be used for quality control during a summer




season, and d) a description of a method for evaluating DO resources




in impoundments.  In addition, a complete bibliography on the subject




of artificial destratification is included.

-------
                               - 4 -

                    THEORETICAL CONSIDERATIONS




Management of DO in Impoundments



       Hydraulic Efficiency



       There are four major methods currently (1966) available for


                                               67              *
artificial destratification, mechanical pumping ' , the Air-Aqua


      8                 9                            **10
system , compressed air,  and the Aero-Hydraulics Gun    .  If these



four systems are to be compared, their hydraulic performance must be



judged on an equal basis, through the calculation of their hydraulic



efficiency.  The hydraulic efficiency of any destratification device



is evaluated by comparing the actual energy input necessary to cause



overturn in a given body of water,  to the theoretical minimum energy



required to overturn that body  of water.  The theoretical minimum



energy required for artificial  destratification is  defined as the




stability of the body of water.



       The  stability of stratified  impoundments has been discussed  by


               9                  11
Koberg and  Ford  and by Hutchinson   , and can be calculated according



to  the  following definition  - the work  or the energy  required to lift



the weight  of  the  entire body of water  the vertical distance between



the center  of  gravity of the water  mass when  isothermal and  the  mass



center  of gravity  when  the  impoundment  is thermally stratified.   For  the



purpose  of this  paper the  center of gravity  is  taken as a  point  anywhere



 on a horizontal  plane  that  passes  through the  true mass center  of gravity



        Figure  1  schematically  illustrates the  stability concept.  In  the



 upper portion  the  temperature  profile in  a  column of water,  taken as



 a rectangular  slice from a wedge-shaped body of water,  is shown at



 three different  months  of a year.   In the first column, the water is






*Product of Hinde  Engineering Company,  Highland Park, Illinois.

**Product of Aero-Hydraulics Corporation, Montreal, Canada.

-------
                            - 5 -

                            WIND
            HEAT
     HEAT
HEAT
ISO-
THERMAL
 C.E.  G.R.
      x
      H
      Q_
      UJ
      Q
      o
      cc
      UJ
      liJ
              I 82
            FEB
    JUNE
TEMPERATURE
 AUG
                Work of the wind
                     11B"
                             TIME

        Figure 1. Schematic diagram to demonstrate stability
                 and work of the wind.

-------
                                  - 6 -





isothermal,  and the center of gravity along the vertical axis is




approximately 1/3 the depth of the water column.  The isothermal center




of gravity does not change throughout the summer season.  In the




subsequent discussion movement of the position of the center of gravity




is considered along the vertical axis only.




           As the water column is heated from the surface, two different




temperature profiles, labeled "1" and "2", are shown for June and August.




Profile "1" in June, dotted, would occur if there were no wind blowing




across the surface of the water.  In this condition, most of the incident




heat would be stored in the upper layers of the water, making this water




lighter, relative to the remainder of the water column.  This decrease




in weight would lower the center of gravity of the stratified water




column to the position shown by CG.. .  As heating continues into August,




thermal stratification becomes more intense, and the center of gravity




moves down into the water column even farther.  Since the actual center




of gravity continues to lower throughout the heating season, the energy




required to mix the water column, which is evaluated by multiplying the




weight of the water column by the distance between the  isothermal center




of gravity and CG1 at any given time, continually increases with time.




This is shown in the lower portion of Figure 1, by the  curve labeled




"condition I."




           In reality, wind is always present  in any natural situation.




The influence of wind blowing across the surface of  the water column




changes the temperature profile from the condition labeled "1"  to the




condition labeled "2".  The surface of the water column does not become as




warm, and the warm water extends  farther down  into the water column.




At any given time, therefore, the  center of gravity  of  the stratified

-------
                               - 7 -
water column does not move down to the level of CG.. , but only descends


to the level shown In Figure 1 as CG2-  Since wind has the effect of


preventing stratification from being as intense as without wind, the


actual energy necessary to mix the water column is considerably


reduced because of wind action.  This is shown in  the lower portion


of Figure 1, where the "condition 2" curve is considerably below that


of the "condition 1," curve.  The distance between these two curves
                                         •11
has been definied as "work of the wind."


       In summary, at any given time, the energy  that is represented


by the distance "G" in Figure 1, is the energy that would have  been


required to mix the stratified water column  in the absence of any


wind.  Since the wind has put into the water the  quantity of work


indicated by dimension "B", only the remainder, the stability,  "S",


must be supplied by any mechanical device, to make the water column


isothermal.  In any natural situation, the stability  can be arithmetically

          a
calculated  .  This theoretical minimum energy requirement for


destratification, when compared with the energy input of a given


mechanical device that caused destratification, defines the hydraulic


efficiency of the mechanical equipment.


       The stability  curve, "S" in the lower portion  of Figure  1 also


shows  that stability  reaches a maximum in  an impoundment during the


middle of the heating season, usually in  late July, and then begins


to decline.  The maximum stability in a  body of water occurs when  the


warm water has penetrated to the  depth of  the  isothermal  center of


gravity along the vertical  axis.  As  the water warms  below  this depth,


and  the water body approaches  an  isothermally warm condition,  the

-------
                              - 8 -






distance decreases between the isothermal and the stratified centers




of gravity, and the stability declines, reaching zero when the water




body is again isothermal, but warm.




       Problems with the Stability Concept




       While stability is a useful method of calculating the hydraulic




efficiency of an artificial destratification device, it must be used




carefully to avoid misinterpretation.  Three problems will be discussed.




The first is the effect of size and geometry on the calculation of




stability.  Table 1 summarizes data from three lakes near Cincinnati,




Ohio, which are under study.  Column 2 through 6 lists the pertinent




physical data for these lakes.  The stability of each lake, column  7,




was calculated assuming the same temperature profile in each  lake,




20°C for the upper 5 feet of water, and remainder of the water body at




4°C.  As seen in column 7, the stability of these three lakes is different




        Hutchinson   suggested that stability should be calculated per




unit of lake surface, to eliminate differances of size and  geometry.




Column  8 shows, however, this  is not a reliable method of calculating




a comparable parameter.  At  least  for  these three lakes, the  stability




per unit volume,  column  9,  is  a better method of obtaining  a  parameter




that can be compared  from  lake  to  lake.  Table  1 demonstrates that




the stability  calculated  for  one body  of water  cannot be assumed  to




exist  in other  lakes  and  reservoirs  in the  same  general geographic




location.  To  properly  evaluate the hydraulic efficiency of a mixing

-------
                            Table 1




           COMPARISON OF STABILITY IN LAKES WITH




THE SAME ASSUMED TEMPERATURE PROFILE BUT DIFFERENT SIZE AND GEOMETRY
Lake
(1)
Falmouth
Bullock Pen
Boltz
Surface
ac
(2)
225
142
93
Greatest
Depth
ft
(3)
42
44
62
Avg.
Depth
ft
(4)
20.5
22.8
31.5
Vertical
Isothermal
Center
of Gravity
ft
(5)
14.232
15.059
20.867
Volume
ac-ft
(6)
4605
3237
2930
Stability
kw-hr
(7)
21.8
15.5
15.0
Stability
area
w-hr
ac
(8)
96.9
109.2
161.3
Stability
vol
w-hr
ac-ft
(9)
4.7
4.8
5.2
i f

-------
                                 - 10 -
operation in a certain impoundment, the stability for that particular




impoundment must be calculated.  Table 1 also illustrates the difficulty




in using a second nearby impoundment as a control for comparison purposes




in an artificial destratification study.  Differences in size and geometry




between the two impoundments may prevent the second from being used as




an "exact" control.




       Second, the continuing natural change that the stability of a given




body of water undergoes throughout a summer season is another difficulty




with its use for the calculation of hydraulic efficiency.  As shown in the




lower portion of Figure 1, as stratification begins, the stability rises




from zero, reaches a maximum in late July, and declines towards zero




again, because of continued wind-mixing and nighttime cooling in August




and September.  Therefore, if the stability in a given impoundment were




calculated when a mixing operating began, the natural change in stability




occurring during the time of the mixing operation, must be included in




any proper evaluation of hydraulic efficiency.




       If a destratification operation should take place in the spring,




and should require a month to complete, the actual energy  input during




mixing should be compared with  the calculated stability at the start of




the operation, plus the natural increase in stability that would have




taken place during the month.   During a fall mixing operation, the natural




decline  in stability with  time  should be subtracted from the initial




stability, in order to properly evaluate the hydraulic efficiency of




the overturn operation.

-------
                              - 11 -



       Third, the hydraulic inefficiency of a spring destratification


operation, in which cold vater is pumped from the bottom to the surface,


must be recognized.  This is illustrated by the data in Table 2.


Here, temperature profiles were assumed for four different months


during the year and the stability calculated for each of the profiles,


(see Column  3).  The energy that must be added to the lake for mixing is


low in the spring and fall, and high in mid-summer.


       If destratification is  to be carried out by bringing cold water


from the bottom as a heat sink to absorb heat  from the upper layers


and thereby  make the lake isothermal, the volume that must be moved


is the volume of cold water that  is present at the start of the operation.


For this  example, these volumes are shown in Column  4, Table 2.


       Although the volume of  cold water that must be pumped in the


spring is high, the work requirements are not, since the elevation

                                                                         12
head through which the water must be lifted, based on density  difference  ,


is much  smaller in the spring  than  in the summer or  fall.  Practically,


however,  mixing equipment cannot  take advantage  of  this  low elevation


head in  the  spring and so the  actual energy  input  at any time  is


directly  proportional  to the volume of  water  that  must  be  moved,


Column 4,  Table  2.  Therefore, although theoretical  energy requirements

                       an
are  low  in the  spring,/hydraulically  inefficient operation results


because  of practical  considerations.  The  improvement  in hydraulic


efficiency as  the summer  season progresses  is  again shown  in  Column 5,


Table  2  where  the volume  required to  mix the lake  is calculated per


unit  of original stability.

-------
                                  -  12 -




                                  Table 2




         VARIATION IN PUMPING REQUIREMENTS TO DESTRATIFY BULLOCK PEN LAKE AT




                             DIFFERENT SEASONS


Month
(1)
April


Assumed
Temperature
Profile
(2)
0-5 ft - 20°C


Stability
kw-hr
(3)
13.8

Cold Water
Volume
Required
to mix
Lake ac-ft
(4)
2560
Cold Water
Volume Required
to mix the
Lake per Unit
Stability
ac-ft /kw-hr
(5)
180
June
August
September
5-45 ft - 9°C




0-15 ft - 23°C




15-45 ft - 9°C




0-25 ft - 25°C




25-45 ft - 10°C




0-40 ft - 25°C




40-45 ft - 10°C
                                 32.9
                                 28.2
2.5
            1440
             640
                                                 35
43
23
15
Note:  Data calculated  for Bullock Pen Lake,  see Table 1. for physical data.

-------
                              -  13 -



       The practical inefficiency of destratification in the spring could


be overcome by pumping the smaller volume of warm water down into the


colder layers.  The theoretical  energy requirements would be the same,


since pumping cold water up involves moving a large volume through a


small elevation head while pumping warm water down would result in


moving a small volume through a  larger elevation head.  As noted before,


however, most mechanical equipment cannot take advantage of the elevation


head differences, therefore the  actual input energy is directly proportional


to the water volume moved.  This would make pumping warm water down in


the spring a hydraulically more  effecient operation.


       This variation in pumping requirements with changing seasons


does not eliminate the usefulness of the stability concept  for calculating


hydraulic efficiency, but the discussion above should  indicate that


a  low calculated hydraulic efficiency may occur during a spring destratification


operation and  should  also indicate  that different  season comparisons


are unfair.

       Another practical consideration  for  any mixing  operation  is  the


change in hydraulic efficiency as the mixing nears completion.   In

                                         9
Figure 2  from the work  of Koberg  and Ford  ,  the  rate of  change  in


stability per  unit  of time decreases as  the stability  approaches  zero.


There are  at  least  three  reasons  for  this.   First,  since so much of the


water in most  lakes  is  in  the upper portions,  a  very  slight temperature


gradient  can  produce  some  stability.  Practically,  complete elimination


of stability,  particularly  in the spring may be  impossible.  Second,


as mixing nears completion the  density  of  the body of water is  nearly


uniform.   Under these conditions some of the water that has been raised


to the  surface may  sink to the  bottom and  be repumped.  This  would lead


to excessive  energy inputs to complete the job.

-------
                              - 14 -
       Finally, as a body of water is cooled to near its final



temperature, there is a smaller decrease in stability per unit volume



of water pumped.  This is caused by the non-linear change in water


                                  12
density with changing temperature.    In the example shown in Table 2,



for the April data, the first 155 ac-ft of bottom water pumped up



decreases the stability 2.3 kw-hrs, while the last 155 ac-ft of water



mixed only eliminates 0.8 kw-hrs of stability.  The higher the



average (final) temperature of the water, the less pronounced this



effect would be, as the density change of water tends to be linear



over small temperature ranges.  Keeping a continuous record of



stability during a destratification operation is recommended  to avoid



excessive energy inputs in the late stages of the mixing.

-------
                         - 15 -
                       MIXING TIME, hours

Figure 2.  Decrease in stability with application of compressed
          a ir (Reference 9).

-------
                                    - 16 -





        Measurement of DO Budget in Impoundments



               Mathematical Model



               One  common mathematical model used  to describe  the change



        of DO  concentrations with  time in  the  euphotic  zone of impoundments,



        where  sedimentation  is  not an  important  factor,  is shown in Equation (1) ,


                                  13
         (see O'Connell and Thomas   .)






                 *- k (L -y) -  k  (D) + (R-P)                           (1)
                         cl
                where  D =  DO deficit  at any time



                      t =  time



                      k- = Deoxygenation rate constant



                      L = Total organic biochemical oxygen demand
                       a


                      y =  Biochemical oxygen demand satisfied at any time



                          (y-L  [1-exp flc.t)]}
                              a.          J-


                      k0 = Atmospheric rearation rate constant



                      R = Rate of oxygen demand by the algal population



                      P = Rate of oxygen produced by the algal population



         Integration yields Equation (2).
l. [expC-^t) - exp(-kzt)]+      - [l-exp(-k2t)] + DjexpC-k^) ]     (2)
                where D  = Initial DO deficit.
                       a

-------
                             - 17 -




                      14
       Although Hull's   suggestion of measuring all the DO sources



and sinks in a given impoundment to evaluate the DO budget is valid,



this is often time-consuming and evaluates the budget only under one



set of environmental conditions.  If a reliable model can be developed,



the changes in the constants, as related to changing sunlight, wind,



algal counts, and nutrient level, will permit general conclusions to



be drawn, permit predictions to be made, and permit comparisons of



one impoundment to another.


       Through the years, research efforts have attempted to devise



methods to evaluate and understand the physical meaning  of the five



constants (k-, k_, L , R, and P) in Equation (2).  While excessive
            J.   f-   Q-


concern over  the precise meaning of these constants has, with some


                                      14             15
justification, been criticized by Hull   and Ettinger   , this does  not



negate the utility of the overall model. The important  feature of



any mathematical model is whether or not its results accurately



reproduce what is occurring  in nature.  The model  shown in Equation (2)



may not be an exact representation of the complete DO budget  in  an



impoundment,  but, as will be shown in Figure 10,  the change  in  "D"



with  time derived from  the solution of  the model,  and the  change in



the dissolved oxygen  deficit with  time  in the  natural environment  are



similar.  Therefore,  the model  is  acceptable  for  further use.



       Evaluation of  the Model  by  the DO Probe Method



       There  are many  different  methods of  studying the DO budget in



natural  waters,  for example, Odum   -  diurnal  cycle method,  Hull


                                                         19
- "light"  and "dark"  -  bottle method,  Goldman  and Carter   - carbon-14


                    20
method,  and  Verduin   - changes in C02 concentration in

-------
                            - 18 -




unconfined water method.  Each method has advantages and drawbacks.


The method developed in this study, called the DO probe method, using


special equipment has  three distinct advantages; (a) the inaccuracies


in bottle experiments  are minimized, (b) DO  is measured directly


and  (c) the model,  Equation  (2), may be  solved from data gathered.


Future  study will undoubtedly  reveal further improvements  in  this


approach, but  this  is  a significant advance.


        Details of the  design and  operation  of this equipment,  and  a


discussion of  its improvements over the  bottle technique are  contained


 in reference  21, and shall be  noted only briefly here.  The apparatus


 consists  of  three DO probes  that  will  record the changes  in DO


 concentration with  time under  three different conditions.   One probe


 is placed in a black container that prevents the admission of either


 sunlight  or  air.  Under these  conditions the constants  "k2",  and "P"


 are zero  in  Equation  (1).  Integrating Equation  (1)  under these


 conditions yields  Equation (3).


                D = Lfl  [l-exp(-k1t)] +  R(t)  + Dfl             (3)


        A second DO probe  is  in a clear plastic container that admits



 sunlight, but not air.  Under these conditions,  Equation (1)  can be


 integrated with "k"  = 0.   This integration yields Equation (4).




               D = L   [l-exp(-k t)] + (R-P)(t) + D           (4)
                    a          1                  a

        The third DO probe is unconfined, therefore the change in DO


 deficit with  time  is  represented by Equation (1), which when



 integrated yields Equation (2).

-------
                              - 19 -




       Evaluation of the five constants in the model is accomplished



by solving Equation (3) for "In " t "R", and "La" using DO vs. time



data obtained from the DO probe in  the dark container.  These constants



are then entered in Equation  (4) and  "P" is evaluated through the use



of the DO vs. time data taken from  the light  container DO probe.


Finally, Equation (2)  is  solved from  the DO data  taken with  the probe



that  is unconfined, and the previously developed  constants.


       For  those interested in the  details of the solution,  an example



is  included in  the Appendix.   The application of  this method, using

                                                   21
data  from  the equipment developed for this purpose  , will  be demonstrated



in  the results  section of this paper.

-------
                             - 20 -




                       EQUIPMENT AND METHODS









Management of DO in Impoundments




       The equipment and methods used  in this study have been explained




In detail  in references 7 and  21 and shall be briefly summarized




here.  Artificial  destratification was studied  from August 6, to




September  10, 1965 in  Boltz Lake, a 2,930 ac-ft  impoundment near




Cincinnati, Ohio.   Destratification was accomplished by pumping




cold water from within 2 feet  of the bottom  and  discharging it in




the  surface with a 13  ac-ft/day capacity, raft-mounted pump, driven by




a 15.5 kw  gasoline engine.  Analysis of temperature, and DO, sulfide,




manganese, and  total COD concentrations at various depths, weekly




from May 14  to  October 29, 1965 demonstrated the changes in stability




and  DO resources that  occurred in this impoundment before, during,




and  after  the pumping  operation.  A nearby,  unmixed  lake, Bullock




Pen, served  as  the control.   Physical  details  of both  lakes are




summarized in Table 1.




Measurement  of  DO  Budget  in Impoundments




       The ability of  the  special DO probe  equipment  to  produce  data




necessary  for  the  solution of Equations (2), (3), and  (4) was  demonstrated




by making  studies  in a 750-gal.  tank  filled  with artificial  impoundment




water.  The  simulated  impoundment water was  created  by mixing  some




settled  sewage,  nitrogen,  phosphorus,  sodium bicarbonate, and  algal




seed with  Cincinnati,  Ohio  tap water.   This  mixture  was  stirred  15




minutes  each hour  and  illuminated  14  hours  each day  until  an observable




algal  growth developed.   The  day of  the test,  some sewage was  added




as  a fresh source  of organic  material.  The three DO probes,  with

-------
                            - 21 -







open trap doors on the containers were lowered into the tank, the




triggers were tripped, and the  test begun.  The DO concentration




data was obtained  for  an  8-hour period, under illumination, using




the recording and  timing  circuitry  detailed  in reference  21.  Before




the  test began,  the  DO probes  were  calibrated and adjusted against




 the  Winkter method for DO analysis.  The  temperature  of  the  test




 was  fairly constant  at 20°C.

-------
                             - 22 -




                             RESULTS




Management of DO in Impoundments




       Hydraulic Efficiency




       The approximate magnitude  of  the work of  the wind  in the control




 lake,  Bullock Pen,  during  the  1965 season  is shown  in  Figure 3.  The




 "G"  energy was  approximated  by assuming that all of the heat incident




 to the lake  was concentrated in the  upper  5  feet of the water.  The




 "G"  energy  is much  higher  than the stability,  the lower curve,




 indicating  that,  at least  for  this impoundment,  the wind  has applied




 much more energy for  mixing  than would mechanical equipment.









        Figure 4 compares the stability pattern for  the test and the




 control lakes.   The control  lake, Bullock Pen, shows  the  typical




 stability pattern,  reaching  a  maximum stability of 45  kw-hrs in




 mid-July.  At this  time the  calculated vertical distance  between the




 isothermal  and actual centers  of gravity  was 0.17 inches. The test




 lake, Boltz Lake, also reaches its maximum stability  about the same




 time, but the calculated stability for Boltz Lake is  somewhat  above




 that for the control throughout the  unmixed period because of




 differences  in size and geometry.  In spite of this slight




 inconsistency,  Bullock Pen Lake was  assumed to be a proper control




 for  this operation.




        Artificial destratification of Boltz Lake was  started on




 August 6, 1965  when the pumping equipment became available.  As  shown




 in Figure 4, through  the 5 weeks of  pumping, the stability in




 Boltz Lake dropped  more sharply than the  natural decline  of  the  control.




 After the 5  weeks of  pumping,  the calculated stability in the  test

-------
200
                            Work of the wind- B
                           STABILITY- "S"
       MAY
JUNE
JULY       AUG
     1965
SEPT
OCT
                                                                Isi

                                                                I
   Figure 3.  Estimation of the work of the wind in Bullock Pen
             Lake (Control lake).

-------
   50
   40
 .  30
CD
<
   20
                                               TEST  LAKE
                                                      CONTROL LAKE
                                                              NET A "S"
                                                              18 kw-hrs
                          i
                          N>
                                   JULY       AUG.
                                        1965
    APR.     MAY      JUNE


Figure 4. Stability in test and control lakes.
SEPT
OCT
NOV.

-------
                              - 25 -

 lake was 18 kw-hr lower than the control at the same time.  Therefore,

 the hydraulic efficiency of this particular piece of mixing equipment,


 calculated according to Equation (5) was a rather low 18/14,300 = 0.13%.

       Net change in stability   „ .   ...   ... J
       Total energy input	  = «ydraulic efficiency             (5)




       Elimination of Reduced Substances


       In Figure 5, the total quantity of sulfide, in the control lake,

 calculated as the equivalent oxygen demand rises throughout the summer,

 reaching a peak in mid-September.  After this time, the natural

 overturn of the lake provides DO throughout the depth, thus oxidizing


 and eliminating the sulfides.  The test lake rises similarly to the

 control prior to the artificial destratification operation.  As

 pumping provided DO throughout the depth of the test lake, the sulfides

were oxidized and their quantity lowered, as shown in the dotted portion

 of the test lake curve.  Assuming that the test lake would have risen

 to a level approximately equal to that of the control lake on September 10,

 the net change in the sulfide content caused by pumping was 13,000

pounds, as the oxygen equivalent.  The oxygen equivalent of the sulfide

was calculated as twice its concentration, since two moles of oxygen

are required to oxidize one mole of sulfide.

       A similar, but smaller, change is shown in the manganese oxygen


demand in Figure 6.   For this analysis, all of the manganese measured

 in both lakes was taken to be soluble, divalent manganese and the oxygen

equivalent of the manganese was calculated as 0.29 of its concentration

since one-half mole of oxygen is required to oxidize one mole of

manganous manganese.   The net change in the manganese oxygen equivalent

-------
                             - 26 -




between the control and test lakes at the end of the pumping operation,




September 10, was 1,300 pounds of oxygen equivalent.  As with the




sulfides, the oxidation of the manganese in the test lake was




accomplished by the destratification providing DO to the lower




level of the lake.




       The total COD content of the test and control lake was also




calculated for each week during the study period.  These data were




more scattered than those shown in Figures 5 and 6, however, and no




significant reduction in the total COD content in the test lake could




be noted during the pumping period.  This is probably because the




degradable portion of the total COD was small, so even though the




addition of DO to the lake enhanced the biodegradation of this portion




of the organic material, the total COD measurement  probably was  too




insensitive  to detect any biodegradation.




       In  summary, the  artificial  destratification  operations on




Boltz Lake,  from August 6 through  September  10,  1965  supplied a  total




of 14,300  pounds of DO, which  satisfied  an  equivalent quantity of




inorganic  oxygen demand.




       Addition and Redistribution of DO




       The decline in  the total  quantity of DO in both  lakes through




 the  early  part of  the  summer  season,  because of continuing  oxygen




 demand  in  the water,  is shown in Figure  7.   When pumping began  in




 the  test lake,  the quantity of DO rose,  but at this same time,  because




 of continued wind-mixing,  additional DO was appearing in the control  lake.

-------
  20
3
O
  16
UJ
Q

Z
LU
O
>-
X


UJ
Q
L
_J
 I
12
 8
                                                               o
                                                            CONTROL LAKE
                                                                               i
                                                                               !
        APR
               MAY
JUNE
JULY       AUG.

     1965
SEPT.
OCT.
NOV.
    Figure 5.  Total sulfides as oxygen equivalent in test and control lakes,

-------
-Q

in
3
O
e
Ill

•>

§2
                 CONTROL LAKE
                               o


                             PUMPING
                                                                               ' -'
                                                                               00
       APR.
MAY
JUNE
JULY      AUG.

   1965
SEPT.
OCT.
NOV.
    Figure 6. Total manganese as oxygen equivalent in test and control lakes.

-------
                             - 29 -




Based on these data, although 14,300 pounds of DO were supplied to the




test lake (Figures 5 and 6) the pumping capacity of the equipment was




not sufficient to add the desired quantity of DO above the inorganic




oxygen demand.  The oxygen demand caused by the natural fall overturn




in late September is also shown in Figure 7.  This occurred even in




the test lake, which restratified slightly after the pumping operation




stopped.




       Although the desired quantities of DO beyond the inorganic oxygen




demand were not added to Boltz Lake, Figure 8 shows the redistribution




of the existing DO resources caused by the pumping operation.  Here,




DO, absent below 20 feet before pumping began, was present, although




in low concentrations, at all depths after pumping.  Since the goal




of any destratification operation is to provide reasonably high levels




of DO at all depths, Figure 8 indicates further the undercapacity of




this particular piece of equipment.  It does show in general, however,




that a favorable redistribution of DO resources is accomplished by




artificial destratification.




       Optimum Time for Destratification




       Figures 5, 6, and 7 show a general deterioration in water quality




from May to August.  The total quantity of DO is lowered in the lakes,




and the quantity of sulfide and manganese rises.  While the mixing




operation improved the water quality markedly in August, the goal of




managing the DO resources in an impoundment through artificial destratification




is to maintain quality throughout the summer season.  The Boltz Lake




studies were planned to begin in spring, in an effort to prevent




any  water quality deterioration, but the lake was actually mixed in




August because the mixing equipment was not ready for operation.




Early season mixing is currently (1966) under study.

-------
                               CONTROL LAKE
                                                                             ' .
                                                                             I
   APR.
MAY
JUNE
JULY       AUG.
   1965
                                                  SEPT.
                                                OCT.
NOV.
Figure 7. Total quantity of dissolved oxygen in test and control lakes.

-------
                     - 31 -
      BEFORE PUMPING
   10
   20
JE 30
o.
LL)
Q
  40
  501
  60
AFTER 38 PUMPING DAYS
                I
            1
I
          I     2345678
               DISSOLVED OXYGEN, mg/l

     Figure 8. Redistribution of dissolved oxygen

              resources in Boltz Lake during

              artificial destratification.

-------
                                 - 32 -




       Spring mixing has some hydraulic drawbacks:  (a) the practical



inefficiency of raising cold water when the volume of cold water is high



(see Table 2) and (b) a naturally increasing stability with time at this



season of the year, (see the lower portion of Figure 1 and Figure 4).



Quality protection, not hydraulic efficiency, is the goal of this



technique, however, and so some hydraulic sacrifices should be made



in an attempt to prevent the quality deterioration that naturally occurs



from the beginning of the heating season through the time of maximum



stability in late July.  Some of this inefficiency could be overcome by



pumping the warm water layer down into the water mass in the spring.



Measurement of DO Budget by the DO Probe Method



       The data plotted in Figure 9 was obtained with the special equipment

                     21
previously described.    Some important water quality parameters of the



artificial impoundment water used for the test are noted on Figure 9.



This figure shows the excellent data that may be obtained using this



equipment, note the extremely expanded ordinate scale.  Approximate



curves showing the trends have been placed by eye through the data points.



       The solutions to Equations (2), (3), and (4) are taken from the



deficit data taken from these curves.  Table 3 summarizes the data used for



the solution of the equations.



       Using these data and the method in Appendix I the solution for



Equation (2) is:
D = WQ.Q^5  V {exp[(-0.15)(t)] - exp[(-0)(t) ]} + ^~^ {l-exp[ (-0) (t) ]



     Da{exp[(-0)(t)]}                                                 (6)

-------
3.4O —
             I
TEST CONDITIONS
 pH - 7.9
 ALKALINITY-132 mg/l AS CaC03
 COD-l5mg/l
 BODS - 0.8 mg/l
 VSS-5mg/l
 ALGAL COUNT-3xlOVml
 TEMPERATURE-2C°C
 ILLUMINATED
         DARK DISSOLVED OXYGEN
         PROBE  DATA-USED FOR
         EQUATION (3)
                                     V.LIGHT DISSOLVED OXYGEN PROBE
                                       DATA-USED FOR EQUATION (4)
                                    UNCONFINED DISSOLVED OXYGEN
                                    PROBE DATA-USED FOR EQUATION (2)
                                                                      i •
                                                                       •
                                                                    B
                                TIME,hours

    Figure 9. Data from special dissolved oxygen probe equipment.

-------
                        - 34 -



                     TABLE 3




DATA FROM FIGURE 9 FOR THE SOLUTION OF THE MATHEMATICAL MODEL




                    DO Deficit - mg/1
Time
Hours
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
Dark Probe
2.82
2.88
2.94
2.99
3.03
3.07
3.11
3.14
3.18
3.21
3.24
3.27
3.30
3.33
3.36
3.39
3.41
Light Probe
2.77
2.82
2.86
2.90
2.94
2.98
3.01
3.04
3.08
3.10
3.13
3.16
3.19
3.21
3.23
3.25
3.27
Unconfined Probe
2.74
2.78
2.81
2.84
2.87
2.90
2.92
2.95
2.97
2.99
3.00
3.01
3.01
3.01
3.00
2.99
2.98

-------
                               - 35 -





Figure 10 compares the line of best fit from Equation (6) with the




deficit data obtained with the unconfined DO probe.  The fairly




close comparison of the data generated from the mathematical model




and the data taken with the DO probe shows the general reliability of
the model.  The low values of k 2 and (R-p) are in agreement with the




low turbulence and low algal population in the test media.  These data




show that the "DO probe method," using this special equipment, and




the mathematical model represented by Equation (2), is a useful method




for studying the DO budget in impoundments.

-------
U   3
LL
LU
Q
* E
o c

Q
LU
              I

          SOLUTION FROM EQUATION (6)-
              \
                   UNCONFINED PROBE DATA FROM FIGURE 9
;
I
I
1
                                      A   A   A
\
     0        I        234567
                                   TIME, hours

      Figure 10. Comparison of  Equation (6) with the unconfined  probe data.
                                                                           U)
                                                                           ••'
                                                  8

-------
                            - 37 -




                           FUTURE PLANS






       Continued study and research are necessary to completely




define the potential of artificial destratification for water quality




control.  Comparison of different destratification methods, spring




mixing, and the mixing of larger bodies of water, must be accomplished.




The first two items are currently (1966) under study, a diffused air




system was used, starting in the spring of 1966, for the destratification




of Boltz Lake.  This system has the disadvantage that it cannot be




adapted to mixing warm water down in the spring.  Eventually, other




destratification methods must be tried, and water quality  control by




this  technique must be demonstrated in a large storage reservoir.




Finally, the  DO probe equipment must be used  in  the  field  to gain




experience with the influence of various natural environmental  factors




on the  constants  in the mathematical model, and  to determine whether




good  correlations  between the model  and data  taken may be  obtained




under a variety of conditions.

-------
                            - 38 -




                   SUMMARY AND CONCLUSIONS








Management of DO in_ Impoundments




       Hydraulic Efficiency




       The data presented here, and in reference 6 and 7 have indicated,




that artificial destratification is an effective method of impoundment




water quality improvement, although to date (1966) it has only been




researched on a relatively small scale.  Any piece of mechanical




artificial destratification equipment should be as hydraulically




efficient as possible, and this can be evaluated by calculating the




stability of the impoundment to be mixed.  The calculated hydraulic




efficiency, "the net change in stability/the total energy input,"




is a good method of comparing one piece of destratification equipment




with another, or evaluating various designs of a given piece of




equipment.  The equipment used in this study had a relatively poor




hydraulic efficiency,  0.1370, probably because of poor hydraulic design.




       Oxygenation Efficiency




       Effective water quality  improvement  is even more  important




than a high hydraulic  efficiency, and  although  the hydraulic efficiency




of this  pumping equipment was  relatively  low, the pump  did  supply




14,300 pounds of oxygen to  the  test  impoundment with  an expenditure




of 14,300 kw-hr of  input energy.  The  calculated  oxygenation efficiency




of this  equipment was  1 pound  of  oxygen/kw-hr.  This  is an  acceptable




value, but  this equipment was  still  undersized  for Boltz Lake,  as




only enough oxygen  was provided to  satisfy  the  inorganic oxygen




demand,  and very  little DO  was  actually  added to  the  water.  In




spite  of this,  the  DO  resources in  the impoundment were favorably




redistributed,  and  at  least some  DO  was  present at all  levels  of the

-------
                                - 39  -




    lake  after  the mixing  operation.


    Best  Time for Destratification


           The  major  advantage  of artificial  destratification over the


    devices for preventing escape of  poor  quality water from impoundments  ,


    is that mixing creates a  two-fold benefit,  protecting  downstream users


    as well as  improving the  quality  of the water in  the impoundment


    itself.  In order to obtain these benefits,  an impoundment  should


    be destratified  in the spring,  to prevent the formation of  bad


    quality water, rather  than  in mid-summer  to improve poor quality water.


    Spring mixing  is  necessary, in  spite of the practical  hydraulic


    inefficiency of  raising cold water  at this season of the year, and


    the necessity  to occasionally  remix during the summer  season.


    These hydraulic  sacrifices  must be made  in order to maintain good


    quality water throughout the heating season.


    Measurement of DO Budget


           To properly evaluate the DO budget in impoundments,  an adequate


    mathematical model and a method  of  solving the model are needed.


    The model presented in Equation  (2),



Mk.)
 2  1
                                      [l-exp(-k2t)] + DjexpC-k^) ]   (2)
                                                      21
    is adequate, and the special  three DO probe system   is effective for


    determining the five constants.  Use of  this tool will permit further


    study of the DO budget  and will  permit comparison of the DO budget


    from one impoundment to another.

-------
                              - 40 -
Conclusions




       Although research on this project is continuing,  the conclusions




that may be drawn at this time are:




       1.  Engineering methods are available that will prevent the




       escape of poor quality bottom water from impoundments.




       2.  Artificial destratification is a better method of impoundment




       water quality control since it has the double benefit of




       preventing downstream damage and improving water quality in




       the impoundment itself.




       3.  The proper use of the stability concept for the calculation




       of the hydraulic efficiency of a mechanical mixing device is




       a useful method of comparing various pieces of equipment and




       improving equipment design.




       4.  In this study, reduced materials, sulfides and manganese,




       were eliminated from the test lake by the artificial destratification




       operation.




       5.  The equipment used in this study was insufficient in capacity




       to add the desired dissolved oxygen beyond the inorganic oxygen




       demand, but the available DO resources in the test lake were




       favorably redistributed by  the pumping so that DO was present at




       all levels.




       6.  In spite of the practical hydraulic inefficiency of raising




       large volumes of cold water, spring mixing is better than mid-




       summer destratification because  early mixing will prevent any




       bottom water quality deterioration.

-------
                                 - 41  -
            7.  Equation  (2)


	:	  iexp^-knt) - exp(-k  t)]+  (~1 [  l-exp(-k t)]  +D  [exp(-k t) ]    (2)
K._"'*•-,         L           *-       K2



            is a  satisfactory  mathematical model for  representing the DO

            budget  in  impounded  waters.

            8.  The special three DO  probe equipment  designed for this

            study is capable of  supplying the data necessary for solving

            the DO  model.

-------
                               - 42 -




                            ACKNOWLEDGMENTS









       The authors wish to thank Lavrence Kamphake,  George Holtzer,  and




Richard Shibiya for the laboratory chemical analyses,  Glenn R.  Gruber




for help with the field work, and Thomas A. Entzminger of the Ohio




River Basin Project who supplied the computer program for the solution




of the Table 3 data, yielding Equation (6).  In addition, acknowledgment




is made of the cooperation of the personnel of the State of Kentucky




Department of Fish and Wildlife Resources.

-------
                            -  43  -

                            REFERENCES
        5.   Kittrell,  F.W.,  "Effects of Impoundments on Dissolved
 Oxygen Resources,"  Sewage  and Industrial Wastes. Vol. 31, 1959
 p.  1065-1078.

        6.   Irwin, W.H.,  Symons, J.M., and Robeck, G.G., "Impoundment
 Destratrlfication by  Mechanical Pumping," Presented at ASCE National
 Symposium  on Sanitary Engineering Research, Development, and Design
 at  University Park, Pa., July 27-30, 1965.  In Press.

        7.   Symons, J.M., Irwin, W.H., and Robeck, G.G., "Impoundment
 Water  Quality Changes Changes Caused by Mixing,"  Presented at the
 ASCE Water Resources  Engineering Conference, Denver, Colorado,
 May 16-20,  1966.  In  Press.

        8.   Ogborn, C.M., "Aeration System Keeps Water Tasting Fresh,"
 Public Works Magazine. Vol. 97, April 1966, p. 84-86.

        9.   Koberg, G.E.  and Ford, M.E., Jr., "Elimination of Thermal
 Stratification in Reservoirs and the Resulting Benefits," Geological
 Survey Water Supply Paper 1809-M, U.S. Government Printing Office,
 Washington, D.C., 1965,  28 pp.

       10.   Bryan, J.G.,  "Physical Control of Water Quality,"  The
 Journal. British Waterworks Association. Vol. XLVI, No. 395, August 1964,
 p.  546.

        11.  Hutchinson,  G.E.,  A Treatise on Limnology. John Wiley &
 Sons,  Inc., New York, New York, 1957, Vol. 1.

        12.  Symons, J.M., Weibel, S.R., and Robeck, G.G., "Impoundment
 Influences  on Water Quality," Journal American Water Works Association,
 Vol. 57, 1, Jan. 1965, pp. 51-75.

        13.  O'Connell, R.L., and Thomas, N.A., "Effect of Benthic Algae
 on  Stream  Dissolved Oxygen," Journal of the Sanitary Engineering
 Division,  ASCE, Vol.  91, SA3, Proc. Paper 4345, June 1965, pp. 1-16.

        14.  Hull, C.H.J., Discussion of "Effect of Benthic Algae on
 Stream Dissolved Oxygen" by Richard L. O'Connell and Nelson A. Thomas,
 Journal of  the Sanitary  Engineering Division, ASCE, Vol. 92, SA1,
 Proc.  Paper 4637, Feb. 1966, pp. 306-313.

        15.  Ettinger, M.B., "How to Plan an Inconsequential Research
 Project," Journal of  the Sanitary Engineering Division. ASCE, Vol. 91,
 SA4, Proc. Paper 4437, August, 1965, pp. 19-22.

        16.  Odum, H.T.,  "Primary Production in Flowing Waters,"
 Limnology  and Oceanography, Vol. 1, 1956, pp. 102-117.

        17.  Hull, C.H.J., "Oxygenation of Baltimore Harbor by Planktonic
Algae," Journal Water Pollution Control Federation, Vol. 35, 1963,
pp.  587-606.

-------
                           -  44 -


       18.  Hull, C.H.J., "Photosynthetic Oxygenation of a  Polluted
Estuary," Advances in Water Pollution Research,  Proceedings of  the 1st
International Conference, held in London, United Kingdom, in
September, 1962, Pergamon Press, London, U.K.,  1964,  Vol. 3, pp.  347-374.

       19.  Goldman, C.R, and Carter, R.A.,  "An Investigation by
Rapid Carbon-14 Bioassay of Factors Affecting the Cultural  Eutrophication
of Lake Tahoe, California-Nevada," Journal Water Pollution Control
Federation. Vol. 37, 1965, pp. 1044-1059.

       20.  Verduin, J., "Energy Fixation and Utilization by Natural
Communities in Western Lake Erie," Ecology,  Vol. 37,  1956,  pp.  40-50.

       21.  Symons, J.M., Discussion of "Effect of Benthic Algae  on
Stream Dissolved Oxygen," By Richard L. O'Connell and Nelson A. Thomas,
Journal of the Sanitary Engineering Division, ASCE, Vol. 92, SA1,
Proc. Paper 4637, Feb. 1966, pp. 301-306.

       22.  Willers, F.A., Practical Analysis. Dover Publications,
New York, New York  (1948).

       23.  Whittaker, E.T. and Robinson, G., The Calculus of Observations.
D. Van Nostrand  Company, Princeton, New Jersey  (1926).

-------
                            - 45 -
                           APPENDIX I
Solution of Mathematical Model with Assumed Data
       The assumed constants for this example are:  k.. = 0.10/hr,
k0 = 0.05/hr, R = 0.10/hr, P = 0.30/hr, L  = 3.00 mg/1, D  = 0.47 mg/1.
 2                                       a               a
Using the appropriate constants and solving Equations (2), (3), and
(4) yields the deficit data shown in Table 4, and plotted in Figure 11.
For simplicity, the units of the constants will not be used.
       Several approaches to the solution of this problem were
attempted.  Various forms of exponential approximation or Prony's
      22                                               23
method  , and a modification of Newton Raphson's method   were
tried without success.  Much of the difficulty in using these techniques
was caused by an insufficient number of significant figures and error
caused by rounding off.  The most productive technique was simple
iteration, which will be developed using the data in  Table 4.  The
solution presented was developed using a Honeywell 400 computer.

       Considering the simulated dark probe data and solving Equation (3)
at t = 7.5 and t = 8.0 yields,
       2.81 = L { 1-exp fck.O.S)]] + R (7.5) + 0.47            (7)
               3.          L
       2.92 = L  (1-exp [-k  (8.0) ]} + R (8.0)+ 0.47           (8)
               ^           4-
Solving Equations (7) and (8) for R and L  yields,
                                         fl
           2.45 (1-exp [-^(7.5)]} - 2.34 fl-exp  [-^(8.0)]}     (g)
       R * 8.0 {1-exp [-1^(7.5)1} - 7.5 {1-exp [-^(7.5)]}

            8.0 (2.34) - 7.5 (2.45J .	^_
        a   8.0 {1-exp [-1^(7.5) ]} -7.5 {1-exp [-^(8.0)]}     (10)

-------
                    - 46 -
                   TABLE 4
      DO DEFICIT DATA FROM ASSUMED CONSTANTS
            D6 Deficit - mg/1


Time
hrs
(1)
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
kl
D
a
Eq. (3)
Simulated
Dark Probe
(2)
0.47
0.67
0.87
1.04
1.21
1.38
1.55
1.72
1.86
2.00
2.14
2.28
2.42
2.56
2.67
2.81
2.92
= 0.10, k =0
« 0.47 mg/1
Eq. (4)
Simulated
Light Probe
(3)
0.47
0.52
0.57
0.59
0.61
0.63
0.65
0.67
0.66
0.65
0.64
0,63
0.62
0.61
0.57
0.56
0.52
k =0.10, k =0

Eq. (2)
Simulated
Unconfined Probe
(4)
0.47
0.52
0.56
0.59
0.61
0.63
0.65
0.66
0.66
0.66
0.65
0.64
0.62
0.60
0.58
0.55
0.52
k^O.10, k2=0.05
P * 0, R » 0.10  P=0.30, R=0.10  P=0.30, R=0.10
L  - 3.00        L =3.00         L =3.00
 a          .      a               a

-------
   3.0
CJ>
E
   2.0
O
u,
LU
Q
z
LU
X
O
Q
UJ
O
GO
CO
Q
    1.0
                       I                 I
          ASSUMED CONSTANTS IN THE MODEL

           K, - 0.10
           K2 - 0.05
           La - 3.00 mg/l
           R - 0.10
           P * 0.30
           Da» 0.47 mg/l
        SIMULATED DARK DISSOLVED
        OXYGEN PROBE DATA
        (EQUATION 3)
        SIMULATED UNCONFINED DISSOLVED
        OXYGEN PROBE DATA
        (EQUATION 2)
                                                                             i
                                                                             i
                                        SIMULATED LIGHT DISSOLVED
                                        OXYGEN PROBE DATA
                                        (EQUATION 4)

                       I
I
I
I
I
I
                                                                         •
                                   TIME, hours

      Figure 11.  Assumed dissolved oxygen data to verify solution of dissolved
                oxygen model.

-------
                             - 48 -
      To solve for k , Equation (3) can be reformulated as follows:


           D - L  [1-exp (-k.t)] + R(t) + D                  (3)
                a           I              a
           D
-D -R(t) = L  [1-exp (-k.t)]
  a         a           i
                                                             (11)
           D-D -R(t)

           	1	1 -exp  (-ktt)                        (12)
              D-D -R(t)

            l-[	1	1 -  exp  (-kxt)                        (13)

                  a


                 D-D  -R(t)
                 D-D  -R(t)


                -t   L(t)    »  -  '"I
                      a


                   D-D -R(t)

            Using [	!L	 ]   as K,  Equation (15) may be written
                      Jj
                       a




            In  a value may be obtained  for  k. at  each
                       3                                       i.


set of values fer D and  t using the dark probe data of  Table  4 and

                                                            2
Equation (16).  The average value  for k- is the new value, k  .  Then


                                         2      2
Equations (9) and  (10) are resolved for L  and R .  The process  is
                                         3


repeated until kn - k    and the solution has converged.   A tolerance  of



0.005 is placed on k-  for convergenge, that is when k     is within



0.005 of k. the iteration stops.

-------
                               - 49




Rearranging Equation (3) into     a)
,  ,  _ r

L  + R L
                                                    _

                                             a      l-exP(-klt)



a least squares technique gives the best value of L  and R, using the
                                                   3


final value of k. .



       To illustrate, starting with k. = 0,50 and L  = 0.76799 and
                                     JL             9

 1                                                          2
R  = 0.21201 from Equations (9) and (10), the solution for k^ would



be as shown in Table 5.  Discarding any values for which 1-K is zero



or negative, k1 = 0.33426.  Continuing in this manner until k  = k^

 1 o

kl"  = 0.11.  The least squares solution of Equation(16$with k.^ = Q.l\



yields R - 0.10 and L  = 2.72.



       In the next step, using k, = 0.11 and L  = 2.72 from the
                                X             3


previous solution, Equation (4) is solved for (R-P) at each time



except t=0 using the light probe data in Table 4.  This yields  16



values for (R-P) that are averaged to yield the final value of  (R-P) .



For these data (R-P) = -0.19.  P can then be evaluated by  difference,



if required.


       In solving Equation (2) a slightly different approach is used.



Here, equally spaced abscissa values are used, then Equation (2)



is reformulated as follows,



            D - C1[exp(-k1t)] + C2[exp (-k^) ] + Q           (17)
where
             2    Vkl      k2       a
                                                               (19)
                   -P)                                          (20)

                   k2

-------
              - 50 -
             TABLE 5
DATA FOR SOLUTION OF EQUATION (16)
t
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
R(t)
0.0
0.105
0.212
0.315
0.424
0.525
0.630
0.735
0.840
0.945
1.050
1.155
1.260
1.365
1.470
1.575
1.680
D-D
a
0.00
0.20
0.40
0.57
0.74
0.91
1.08
1.25
1.39
1.53
1.67
1.81
1.95
2.09
2.20
2.34
2.45
D-D -R(t)
r a i . ,,
1 L J K
a
0.0000
0.12239
0.24478
0.32811
0.41144
0.49477
0.57810
0.66142
0.70569
0.74995
0.79422
0.83848
0.88275
0.92701
0.93222
0.97648
0.98168
ln,l-K . k
t 1
-
-0.26111
-0.28075
-0.26511
-0.26504
-0.27309
-0.28766
-0.30943
-0.30578
-0.30802
-0.31619
-0.33148
-0.35724
-0.40269
-0.38449
-0.50000
-0.50000

-------
                              . 51 -
Solving Equation (17) at (t , D ) for C [exp(-k t )] yields,
                           J_   JL       £»       ff i




            C2[exp (-k2t1^ * Di"Ci ^exp ("k1t1)J ~Q          (21)




The solution for C-[exp(-k0t )] at (t0,D0) is
            C2[exp (-k2t2)] = E>2-C  [exp(-k1t2)] - Q           (22)




Dividing Equation (21) by Equation (22) yields,





            exp [-k.(t -t )] - -~j;
Repeating this process at(t2>D2)and(t3,D3)yields,
                               D -C  [exp(-k t )] -Q

            exp [-kCt-t)] -           .^t ,] .-,             <»)
Since equally spaced abscissa values are used,  (t- -t2) =  ^t2~t



the right hand side of Equations  (23) and  (24) may be equated,



yielding
       D1-C1[exp(-k t )]-Q      D  -C  [exp(-k  t  ) ]-Q
        1111           ]                                         (27)




                                                                  (28)




                                                                  (29)



and    ZtDD-CD)]                                          (30)

-------
                              - 52 -







Using values of k  , L  , and (R-P) derived from the solution of
                 J-   3


Equations (3) and  (4), the quadratic Equation (26) may be solved



for its positive root, as follows.




       The coefficients for k~ and the constant in Equation (26)




may be determined  from each three successive data points, for




example (t=0.5, 1.0, 1.5), (t=1.0, 1.5, 2.0) and so on.  This




yields 14 polynomial equations in the form of Equation (26).  The



fourteen sets of coefficients and constants are then averaged to




yield an "average" Equation (26) which is then solved for k_ as the



only unknown, which in this example = 0.09.  A summary of all of the




solutions for the constants in Equation (2) is shown in Table 6.






                             TABLE 6




         COMPARISON OF SOLUTIONS FOR  CONSTANTS IN EQUATION (2)
                                      R
Assumed 0.10 0.05
Dark Probe 0.11
Light Probe
Unconfined - 0.09
Probe
0.10
0.10
(R-P)
                                                 -0.20
                                                 -0.19
            3.00



            2.72

-------
                               - 53 -
       To demonstrate the adequacy of the model,  Equation (2),  and



the method of solution, deficit data was generated using the



constants            of Table 6, in Equation (2).  This curve is



plotted in Figure 12 and is compared to the simulated unconfined DO



data from Column 4 in Table 4.  The fit curve is low as time progresses,



probably because k« is too large.  Fit curves calculated for the dark-



and light-probe systems are much better, as ky *-s not included.



       The solution is very sensitive to the final value of k-.



The computational procedure was repeated using a tolerance of 0.001



as a measure of convergence for k1.  Under these circumstances k



is reduced 0.03 to 0.08.  This slight reduction has a large impact



on the remaining calculations.  One method of reducing this instability



might be to evaluate Equations (9) and  (10) at each group of two



times (0.5, 1.0; 1,0,  1.5; and so  forth)  instead of just at



t = 7.5 and 8.0 to obtain an  "average"  L   and R  before calculating
                                        a

 2
k  .  In spite of these minor  difficulties  the usefulness of this



equipment and method for studying  the DO  budget  in  impoundments has



been demonstrated.

-------
 M

o
u.
UJ
Q

z
UJ
CD


§

Q
UJ


O
CO
    1.5
1.0
0.5
   0.0
       SIMULATED DATA-COL. 4-TABLE 4
                                         I
                                        SOLUTION  TO EQUATION (2)
                                        USING CONSTANTS
                                        TABLE 6
                                              I
                                        4        5
                                       TIME, hours
                                                                       8
                                                                              ! 'I

                                                                              I
       Figure 12.  Comparison of the solution of Equation (2) with the simulated data.

-------
                              -  55 -

                           APPENDIX II

                           BIBLIOGRAPHY
       1.   Hooper,  F.F.,  Ball,  R.C.  and  Tanner, H.A.,  "An Experiment  in
the Artificial Circulation of a Small Michigan Lake,"  Transactions  of
the American Fisheries Society. Vol. 82, 1952, p.  222-240.

       2.   Grim, J.,  "Bin See Wird Umgepflugt," Allg.  Fisch-Ztg., Jahrg..
Vol. 77, 1952, p. 281-283.

       3.   Mercier, P. and Gay, S.,  "Effects de L'aeration  artificielle
sous-lacustre au lac de Bret," Revue Suisse D'Hydrologic, Vol.  XVI,
Fasc. 2, 1954.

       4.   Cooley,  P. and Harris, S.L.,  "Prevention of Stratification
in Reservoirs," Journal of the Institution of Water Engineers,  Vol. 8,
1954, p. 517-537.

       5.   Streiff, A., "Compressed Air  vs. Drought,"  Compressed Air
Magazine,  Vol. 60,  August 1955, p. 232.

       6.   Derby, R.L., "Chlorination of Deep Reservoirs for Taste  and
Odor Control," Journal American Water Works Association. Vol. 48,
1956, p. 775-780.

       7.   Mercier, P., "L'aeration Naturelle et  Artificielle des Lacs,"
Rev, suisse d"hydrol.. Vol. 19, 1957, p. 613.

       8.   Schmitz, W.R. and Hasler, A.D., "Artificially Induced
Circulation of Lakes by Means of Compressed Air," Science,  Vol. 128,
October 1958, p. 1088-1089.

       9.   Riddick, T.M., "Forced Circulation of Reservoir Waters,"
Water and Sewage Works. Vol. 104, 1957,  p. 231-237.  "Forced Circulation
of Large Bodies of Water," Proceedings of the American Society of Civil
Engineers. Journal Sanitary Engineering Division, Vol. 84,  Paper 1703,
SA-4, July 1958, 21 pp.

      10.  Heath, W.A. ,  "Compressed Air Revives Polluted Swedish Lakes,"
Water and Sewage Works. Vol. 108, 1961,  p. 200.

      11.  Nickerson,  H.D., "Gloucester-Forced Circulation of Babson
Reservoir," Sanitalk,  Vol. 9,  Summer  1961, p. 1-10 &  27.

      12.  Patriarche, M.H., "Air-Induced Winter Circulation of Two
Shallow Michigan Lakes,"  Journal of  WildlifeManagement, Vol. 25,
1961, p. 282-289.

      13.  Laurie, A.H.,  "The  Application  of the  'Bubble-Gun1 Low  Lift
Pump," Water  and Waste Treatment, Vol.  8,  1961, p. 363.

      14.  Anonymous,  "'Bubble-Gun'  for Destratification," Pneumatics
Breakwaters Ltd., Water  and Water Engineering,. Vol. 65, 1961, p. 7071.

-------
                               - 56 -


      15.  Yount, J.L., Biologist,  Florida State Board of Health,
Mixed 300 Acre Lake in 1962 with Compressed Air, Personal Communication.

      16.  Anonymous, "Improvement  in the Water Quality of Reservoir
Discharge Through Reservoir Mixing  and Aeration," Water and Water
Engineering, Vol. 66, 1962, p. 112.

      17.  Ford, M.E., Jr., "Air Injection for Control of Reservoir
Limnology," Journal American Water  Works Association. Vol. 55, 1963,
p. 267.

      18.  Karlgren, L. and Lindgren, 0., "Aeration Studies in Lake
Trask," Vattenhygien. Vol. 19, 1963, p. 67.
                                            11
      19.  Bernhardt, H., "Erste Ergbnisse  uber die Belttftungsversuche
an der Wahnbachtalsperre." Vom Wasser. Vol. XXX, 1963, p. 11.

      20.  Bryan, J.G., "Physical Control of Water Quality," The Journal,
British Waterworks Association, Vol. XLVI, No. 395, August 1964, p. 546.

      21.  Ridley, J.E., "Thermal Stratification and Thermocline Control
in Storage Reservoirs," Proceedings of the Society for Water Treatment
and Examination. Vol.  13, 1964, p.  275.

      22.  Forty-first Report on the Results of the Bacterial, Chemical,
and Biological Examination of the London Waters for the Years 1963-1964,
"Thermal  Stratification and the Possibility of Thermocline Control of
Storage Reservoirs," p. 104.

      23.  Robert, G.E., and Ford, M.E., Jr., "Elimination of Thermal
Stratification  in Reservoirs and the Resulting Benefits," Geological
Survey Water  Supply Paper 1809-M, U.S. Government Printing Office,
Washington, D.C.,  1965, 28 pp.

      24.  Bryan, J.G., "Improvement in  the Quality of Reservoir
Discharges Through Reservoir Mixing and  Aeration," Symposium on Streamflow
Regulation  for  Quality Control, PHS Publ. No.  999-WP-30,  Cincinnati,
Ohio, June  1965, 420 pp.

      25.  Anonymous,  New Zealand  Outdoors. May  1965.

      26.   Irwin, W.H., Symons, J.M. and Robeck,  G.G., "Impoundment
Destratification by Mechanical  Pumping," Presented at ASCE National
Symposium on  Sanitary  Engineering  Research, Development,  and Design
at University Park,  Pa., July 27-30,  1965,  In Press.

       27.   Ogborn, C.M.."Aeration  System Keeps Water  Tasting Fresh,"
Public Works  Magazine, Vol.  97, April  1966, p.  84-86.

       28.   Symons,  J.M.,  Irwin, W.H.,  and Robeck, G.G.,  "Impoundment
Water Quality Changes  Caused by Mixing," Presented  at the ASCE  Water
Resources Engineering Conference,  Denver, Colorado,  May  16-20,  1966.
In Press.

-------
      29.  Burns, J.M., Jr., "Reservoir 'Turn-over'  Improves Water
Quality," Water and Wastes Engineering, Vol. 3, No.  5, May 1966, p.  81.

      30.  Anonymous, "Water Treatment with Compressed Air," Industrial
Wa t_er Engineer ing, Vol. 3, No. 6, June 1966, p. 40-41.

      31.  Symons, J.M., DeMarco, J., Irwin, W.H. and Robeck, G.G.,
"Enhancing Biodegradation of Synthetic Organics in Stratified Impoundments
by Artificial Modification of the Thermal Profile,"  To be Presented
at the International Association of Scientific Hydrology's meeting
"Symposium on Lakes and Reservoirs" at Lake Garda, Italy, October 10-15,  1966,
In Press.

      32.  Fast, A., reported by Roach, R., "Reservoir Aeration Reduces
Evaporation, Improves Water," Palo Alto Times, August 4, 1966, p. 1.

-------