WATER POLLUTION CONTROL RESEARCH SERIES 16010 DMG 12/71 FLUSHING OF SMALL SHALLOW LAKES U.S. ENVIRONMENTAL PROTECTION AGENCY ------- WATER POLLUTION CONTROL RESEARCH SERIES The Water Pollution Control Research Series describes the results and progress in the control and abatement-of pollution in our Nation's waters. They provide a central source of information on the research, development and demonstration activities in the Environmental Protection Agency, through inhouse research and grants and contracts with Federal, State, and local agencies, research institutions, and industrial organizations. Inquiries pertaining to Water Pollution Control Research Reports should be directed to the Chief, Publications Branch (Water), Research Information Division, R&M, Environmental Protection Agency, Washington, D,C. 20^60. ------- FLUSHING OF SMALL SHALLOW LAKES by Claud C. Lomax John F. Orsborn The R. L. Albrook Hydraulic Laboratory College of Engineering Research Division Washington State University Pullman, Washington 99163 for the Office of Research and Monitoring ENVIRONMENTAL PROTECTION AGENCY Grant #16010 DMG December 1971 For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, B.C., 20402 - Price 80 cents ------- EPA Review Notice This report has been reviewed by the Environmental Protection Agency and approved for publication. Approval does not signify that the contents necessarily reflect the view and pol- icies of the Environmental Protection Agency, nor does mention of trade names or commercial products consti- tute endorsement or recommendation for use. ii ------- ABSTRACT Restoration of quality to polluted lakes by inflows of clean water with simultaneous outflow of polluted water was investigated. Elliptical basins were used in the laboratory to simulate shallow lakes. The in- vestigation determined the influence of selected geometric and flow parameters on the flushing efficiency. The theoretical analyses are combined with the experimental results to obtain equations for predicting the flushing curves. Application of these equations will give the potential flushing efficiency of a pro- posed flushing scheme. The width of the basin perpendicular to the axis of flow is an impor- tant parameter. In general, narrower basins have better flushing action and less erratic flow patterns than wider basins. The pollution remaining in a basin after any interval of flushing is primarily depend- ent on the time of the flushing interval divided by the detention time of the basin. This report was submitted in fulfillment of Grant Number 16010DMG, under the partial sponsorship of the Water Quality Office, Environmental Pro- tection Agency. iii ------- CONTENTS Conclusions 1 Recommendations 3 Introduction 5 Experimental Apparatus 13 Test Procedure 15 Data Analysis 17 Test Results 21 Discussion of Test Results 23 Acknowledgments 29 References 31 Appendices 33 ------- FIGURES Page 1 NOMENCLATURE AND FLOW ZONE SKETCHES 6 2 RESIDUAL CONCENTRATION AND AVERAGE RATE OF CONCENTRATION REMOVAL 7 3 GENERALIZED BOUNDARY CONDITIONS FOR FLUSHING EFFICIENCY 7 4 SMOOTHED OUTFLOW CONCENTGRAPH 8 5 PREDICTED VALUES, CR/CQ 10 6 FLUSHING PARAMETER, td/t , AS A FUNCTION OF ASPECT RATIO, Y/X, AND INLET WIDTH, Wj_ 11 7 SCHEMATIC OF TEST APPARATUS 13 8 TYPICAL OUTFLOW HYDROGRAPH, C/CQ VERSUS t/td 17 9 CONCENTRATION VERSUS TIME 18 10 FLUSHING CURVES FOR 10-FOOT BY 4-FOOT BASIN 21 11 FLUSHING CURVES FOR 10-FOOT BY 6-FOOT BASIN 21 12 FLUSHING CURVES FOR 10-FOOT BY 10-FOOT BASIN 22 13 FLUSHING CURVES FOR 6-FOOT BY 10-FOOT BASIN 22 14 IDENTICAL RUN COMPARISON. RUNS 60 AND 61 24 15 EXAMPLE OF WELL-DIFFUSED JET 26 16 COMPARISON OF AVERAGE FLUSHING CURVES 27 VI ------- TABLES NO. Pa?e 1 Predicted Values, CR/CQ , from Eqs. 9 and 14 10 2 Summary of Test Conditions 15 3 Data and Computed Values 19 4 Data Summary for 10-Foot by 4-Foot Basin 36 5 Data Summary for 10-Foot by 6-Foot Basin 37 6 Data Summary for 10-Foot by 10-Foot Basin 38 7 Data Summary for 6-Foot by 10-Foot Basin 39 Vll ------- CONCLUSIONS 1. The most important parameter in determining the shape of the flush- ing curve, CR/C0 versus t/td, is the variable behavior of the inflow jet. 2. Inlet velocity and depth of flow have no separable effects on the flushing curves. 3. For a predictable jet behavior, the residual concentrations are primarily a function of the time ratio, t/td, and the basin shape (as- pect ratio, Y/X). 4. The theoretical flushing efficiency as measured by ts/td is influ- enced directly by the inlet width, V^, inversely by the aspect ratio, Y/X (Eq. 2), 5. Although the results presented have been derived to apply to flush- ing pollutants from a lake, the same information can be used to evalu- ate the degree of increased pollution versus time where a polluted stream (inflow) is permitted to enter a lake. 6. Perhaps the most significant conclusion reached in this study is that a jet entering a shallow basin with a solid boundary (the bottom) and an air-water interface (the free surface), does not follow predicted patterns for a jet discharging into an infinite medium (1). Considerable work remains to be done to describe the behavior of such a jet includ- ing the influences of density differentials. ------- RECOMMENDATIONS Although this program was conducted under controlled laboratory condi- tions, valuable information has been developed for planning the enhance- ment of water quality in polluted lakes situated within reasonable dis- tances of flushing water sources. Factors such as wind and tidal action were not considered in this program, but their relative effects can be relatively estimated. Wind action would tend to promote more complete mixing in shallow lakes, and tidal action would tend to oppose the flushing action. Temperature differentials between the inflow and the lake were not investigated. Therefore, field pilot studies under natural conditions are an obvious recommendation. Many natural situations ,of various sizes already exist which could be monitored. Care should be taken to avoid systems which are so large that automatic, continuous monitoring becomes too expensive. A system analysis study should be conducted, using the results of this study combined with natural physical and climatological factors, to de- velop an optimum pilot investigation plan. This analysis should be con- ducted prior to the initiation of any field experiments. Based on the hydraulic and quality parameters of the flow system, the system analysis would be used to determine required sampling location and density, data types and probable scaling factors to be evaluated as a result of dif- ferent system sizes. The results of these pilot tests would provide a significant contribu- tion to water resources system design information for the enhancement of quality in polluted lakes as guided by results of this laboratory study. ------- INTRODUCTION Flushing of a lake means reducing the pollution by clean inflow with an equivalent outflow of polluted water. In the process the clean water both displaces and mixes with the polluted water. Many factors influence the effectiveness of the cleansing stream. This study was conducted to investigate those parameters believed to be most important and manageable under laboratory conditions. The parameters selected for study were: 1) the inlet velocity, 2) the inlet width, 3) the depth, and 4) the basin shape. Testing was conducted on two depths, two inlet widths, three inlet velocities, and four elliptical basins. The primary purpose of the project was to evaluate the various parameters to determine their influences on flushing efficiency, and to develop prediction equations based on the geometric and flow charac- teristics of the systems tested. Analyses were completed to: 1) develop a test program, 2) analyze the system for comparison with experimental results, and 3) develop pre- diction equations which incorporate the analytical and experimental re- sults of the study. A dimensional analysis of the various geometric, flow, and fluid parame- ters was completed to isolate the more important variables as a guide for establishing the test program. The average rate of removal of pol- lutant, Qr, was of primary interest and is a varying percent of the in- flow rate, Q.|_, as a function of time. The following set of dimension- less ratios was developed: V ' Y ' X ' d Terms are defined in Figure 1 and in the Nomenclature. Fluid properties such as density and viscosity were considered constant throughout the system. Grouping the first four dimensionless terms in Eq. 1 yields an expression for the average rate of soluble pollutant removal during passage of some time, t, such that W?X (2) K' is a dimensionless coefficient containing 4/TT for elliptical basins and is a function of not only the geometric and flow parameters, but also time. K1 was found to be independent of the jet momentum force, T?±, and the Reynolds number. For the general case of flushing lakes for various geometric conditions, Qr should approach Qt as Wt approaches Y. Also, as X becomes very ------- large with respect to W^ and Y, Qr should approach Qi- The coefficient K' is used to relate Qr to the actual flushing curve at any rela- tive flushing time, t/td, in Figure 2 for a particular lake geometry. An important time ratio for evaluating flushing efficiency is ts/td, or the relative time at which the real flushing curve departs from the ideal flushing curve. Noting that the volume of the lake is V = 0.25irXYd and hold- ing d constant in Eq. 2, it would be expected that as W^ approaches Y, the average rate of pollutant removal Qj. approaches Q^. At the other extreme, W^ approaches zero, Qi> Qr and the flushing effi- ciency should all tend towards zero. These general boundary conditions are shown graphi- cally in Figure 3. It was found in these tests that for small values of and low inflow discharges, the flushing action can move through the basin as a wave front and be quite efficient. As long as Q^ is finite, the actual flushing curve will follow the ideal flushing curve for a time, ts/td, if external mixing forces such as wind are neglected. 10 feet maximum (a) Plan View of a Longitudinal Basin- Hydraulically Long Lake (b) Elevation, Common to (a) and (c) Developing Developed Jet Zone-, £L2°™ ,-SinkZone (c) Plan View of a Transverse Basin- Hydraulically Short Lake Fig. 1. Nomenclature and Flow Zone Sketches A theoretical analysis of the flushing was developed to relate concen- tration to time. The displacement phase of pollutant removal continues for a time interval, tg/td, with the outflow concentration equal to C/C0 = 1.0. The remaining outflow concentration is assumed to follow some form of an exponential expression. Figure 4 shows the assumed curve. The area under the curves to any time, t/td, is proportional to the pollutant removed from the basin. ------- Qr = QO between 1.0 2.0 3.0 t/td, Relative Length of Flushing Time Fig. 2. Residual Concentration and Average Rate of Concentration Removal 4.0 1.0 0.8 « ^ 0.6 » 5 O-4 o ^ 0.2 •Wj/B / / _/ / Y/X- General parameter being a function of lake shape and equal to K1 at t/tj = 1.0. ASSUMED: Volume of lake constant. Depth of water constant in lake, inlet and outlet. l 0 1.0 •Relative Inlet Width, Wj/Y ^ -Aspect Ratio, Y/X, for Small, Constant Inlet Width, Wj- Fig. 3. Generalized Boundary Conditions for Flushing Efficiency 2.0 ------- Ideal Flushing Curve Smoothed Outflow Concentgraph The initial quantity of dye in the basin equals 1.0 KV, where K is a dye removal coefficient. The area under the curve from t = 0 to t = °° must equal KV. Integrating the equations to find the area and determine "a", f Is- + f L d »*- •-< ;a(t-ts)/td d(_t_; (3) On the right side of Eq. 3, K is multiplied by td to obtain the correct units for new time base. Within the brackets, the first term, tg/tj* accounts for area under the horizontal part of the curve; that is, where C/CO = 1.0. The^area under the exponential decay part of the curve is determined by Jt /t,. Eliminating K and dividing by Q.^ to get V. Qi J_Id a(t-ts)/td a t (4) Divide by td. The lower limit gives one and the upper limit gives zero for "e" raised to its power. Therefore, 1 = ~d+~ Solve for a = -1 1 - (5) (6) ------- Residual in Basin = Total at the start minus that removed. 'ts/td Substitute limits and cancel K. Divide by Q^ to get td = V/Q., and divide by tj. 1 (8) Now substitute a = -l/[l-(tg/td)] and simplify. co This equation applies for t/td > tg/td. To obtain CR/CQ when t/td < ts/t, use sd CR The most significant time ratio in the right side of Eq. 9 is t /tH. Expressing this ratio in terms of the geometric and flow parameters, and inverting to give whole numbers leads to The experimental value of tg is equal to the X distance divided by the average velocity of the fastest flow-through element. If v = bVi (12) then for elliptical basins, Eq. 11 becomes Eq. 13, when inverted and simplified, gives tg 4bWt ~d = ~W~ ' In this form, b can be selected to fit any condition. It is feasible to assume a value of b for Eq. 14 and compute .t /tH and the values of CR/CO can then be computed for values of t/t, using Eq. 9 First the limits will be examined. If ts/td = 1.0, then the polluted water is all displaced by the clean inflow and the ideal flushing curve ------- is followed. Obviously, tg/t^ cannot exceed 1.0. Assuming instantane- ous mixing of the inflow with the basin water, ts/tj =0, and the expo- nential curve starts at t/t^ = 0. The assumption of an exponential decay curve implies perfect mixing. This curve and the ideal flushing curve are plotted for Figure 5. It is apparent that any value of tg/t^ be- tween zero and one will give a curve lying between these two curves. 100 _ 80 2 S 60 4- C a o & 1 40 O m O O 20 a J Ideal Mixing t,/td = 0 Ideal Flushing - t,/td - 1.0 1.0 2.0 3.0 t/tj, Relative Length of Flushing Time Fig. 5. Predicted Values, CR/C0 4.0 Table 1 gives the values calculated for predicting CR/CO for various t/tj and ts/td values. Although Table 1 gives valid limits for the assumed conditions, there are other possibilities which are amenable to theoretical analysis. Conditions may change with time so flushing may Table 1. Predicted Values, CR/CO, from Eqs. 9 and 14 t/td £d (1) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 (2) 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1 (3) 0.368 0.331 0.294 0.258 0.221 0.184 0.147 0.110 0.074 0.037 0.000 2 W 0.135 0.109 0.084 0.062 0.042 0.025 0.012 0.004 0.001 0.000 0.000 3 (5) 0.050 0.036 0.024 0.015 0.008 0.003 0.001 0.000 0.000 0.000 0.000 4 (6) 0.018 0.012 0.007 0.004 0.002 0.001 0.000 0.000 0.000 0.000 0.000 10 ------- be either: 1) displacement, 2) near perfect mixing, or 3) very poor mixing to the extent of clean water short circuiting through the basin to the exit. The experimental values cannot be below the ideal flush- ing curve in the lower limit curve as given by values in Table 1. Each curve consists of a short section of the ideal flushing curve plus an ideal mixing curve which starts at tg/td. It is recognized that neither ideal mixing nor a perfect short circuit (upper limit of no flushing) are possible. Variability in the experimental values of t /td can be attributed to: 1) the jet expansion and 2) the type of start, "static" or "running." Examination of the limits of Eq. 14 reveals some of the complexity of establishing b values. td/tg cannot be less than unity. When Y/W£ = 1, td/ts - 1 and b must equal 4/ir. The minimum value of Y/X also occurs when Y = W^. If Y = °°, then Y/W^ = °°, td = °°, and td/ts = <». However, ts is finite. If V^ =0, ts = 0, td = <*>, and td/ts = °o. In this case, Eq. 14 is not valid. The experimental data have been plotted in Figure 6 to show the depend- ence of td/tg on aspect ratio Y/X and inlet width Wj_. These data are averages for the combined data (4 from two depths and three ve- locities. Within each group there is random scatter of ±10 percent except for two readings which were ±25 percent from the average. It should be noted that the running start data gave t,j/ts values which were from two to five times those for the static starts. The curves shown in Figure 6 can be combined with Eq. 14. ~ TJ 10Y = 8 in.) (16) with Wi, X, and Y in feet. In Eqs. 15 and 16, b = 0.29X and 0.24X, respectively, for a typical jet expansion rate. Considerably more data, pref- erably under natural condi- tions, are required to extend with confidence Eq. 15 beyond the limits of these tests. td Limit Ideal Flushing — =1.0 Fig. 6. Flushing Parameter, td/ts, as a Function of Aspect Ratio, Y/X, and Inlet Width, WA 11 ------- EXPERIMENTAL APPARATUS Figure 7 is a schematic of the test apparatus. The test basins were constructed on a 12-foot by 12-foot wood platform. The entrance still- ing box, gate, and depth measuring piezometer tap were common for all tests. At the exit, the mixer, Fluorometer tap, control gate, and measuring weir were common. The sides of each basin were fabricated with sheet metal fastened to a plywood template of the desired ellipti- cal shape. The entrance and exit sections were faired to the elliptical basins by a 2-inch radius. Sections were provided for an 8-inch exit width and for both an 8-inch and a 4-inch entrance width for each basin. A constant head tank supplied the clean water. After the desired flow rate was established with regulating valves, a shut-off valve could be closed and reopened without affecting the flow rate. Flow rates were measured with the 90° V-notch weir. Repeated readings on the point gage established the constancy of flow rate during a test run. A tail-water gate at the exit of the discharge channel controlled the depth. Time- lapse, motion pictures in color were taken at intervals. Rates of one picture per second, per two seconds, or per four seconds were used. Samples of the outlet channel flow were pumped at a constant rate through a heat exchanger for temperature stabilization, then through the Fluorom- eter. The Fluorometer readings were recorded with a strip chart recorder. Temperatures were sensed at the outlet of the flow-through door with a thermistor and recorded whenever changes occurred. The inlet and outlet temperatures of the basin were monitored for a few of the final tests. The water in the basin was usually warmer than the inlet supply by 2 to 5 degrees Celsius. The flow was warmed about 1.5 degrees Celsius by heat from the air and lighting during passage through the basin. TEMPERATURE INDICATOR PUMP CHILLER 12" GRID RECORDER FLUOROMETER / CONSTANT TEMPERATURE BATH , 'i i — HEAT P J EXCHANGER TAIL-WATER! CONTROL—' "BAFFLES 1—ENTRANCE BOX PLAN VIEW Fig. 7. Schematic of Test Apparatus 13 ------- TEST PROCEDURE 1. A flow rate was established with the upstream regulating valves. 2. The tailgate was adjusted to establish the desired depth. 3. Flow was shut off without disturbing the regulating valves. 4. The basin was filled to the test depth with water polluted with fluorescent (Rhodamine WT) dye. This dyed water extended to the control gates in the inlet and outlet channels. 5. The entrance box was filled to the necessary depth with clean water. 6. After the basin water became quiescent the upstream control gate was opened. 7. The shutoff valve and the downstream control gate were opened simul- taneously. 8. The Fluororaeter, timing, and camera systems were started with the start of flow into the entrance channel. At the beginning of each test the basin concentration was adjusted to be within the range 85 to 100 percent on the Fluorometer scale being used. The test was terminated when the outflow percentage reached about five percent. Table 2 summarizes the parameters studied in this investigation. Table 2. Summary of Test Conditions Variable (1) Inlet width, / in inches X Inlet velocity, f in feet per < second I Depth, in f inches L Running Start also tested Basin Size, in feet, Y by X 4 by 10 (2) 4.0 8.0 0.1 0.2 0.4 2.0 4.0 No 6 by 10 (3) 4.0 8.0 0.1 0.2 0.4 2,0 4.0 Yes 10 by 10 (4) 8.0 0.1 0.2 0.4 2.0 4.0 No 10 by 6 (5) 4.0 8.0 0.1 0.2 0.4 2.0 4.0 No 15 ------- DATA ANALYSIS The raw data from the strip charts were corrected for temperature vari- ations in the sample temperature by the technique given by Cobb and Bailey (2). Time adjustments were made to correct the raw data for transit time in the inlet, and the delay due to passage through the out- let channel and the tube leading to the Fluorometer. Although a small error exists due to instrument response lag, no correction was made for it. In Figure 8 the data were reduced to a common comparison base by using the relative concentration; that is, measured concentration divided by initial concentration and multiplied by 100 to return to percentages. A dimensionless time equal to the time from the start of flushing (when clean water enters the basin) divided by the basin detention time has also been used for all comparisons and evaluations. Figure 8 may be thought of as a dimensionless hydrograph of the outflow concentration. i i i i 1 i i ill ii 0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 Fig. 8. Typical Outflow Hydrograph, C/CQ Versus t/td The area under the curve when multiplied by the proper factors gives the quantity of dye removed. This subtracted from the initial quantity gives the residual concentration. Considerable local variation from this aver- age value will occur throughout the basin. 17 ------- Referring to the simplified concentration versus time curve in Figure 9: Initial quantity of dye = KG V dt Dye removed = K/ Q.C t «• Residual quantity of dye = KCRV = K(CDV - f Q.C R / i n dt) Basin volume V = 0.257rXYd Basin detention time = t. = —- = (17) (18) (19) (20) (21) Fig. 9. Concentration Versus Time Eliminate K, substitute Q^d for V, divide by Co and introduce td into the third term of Eq. 19 to obtain: fC0 = k tn/td t0/td (f-) (22) Simplify to: (23) When multiplying by 100 to get the relative concentration on a percentage basis, Eq. 23 is applicable to the curve shown in Figure 8. Because the experimental Cn versus t curve is not mathematically defin- able, discrete steps as shown in Figure 9 must be summed to obtain the area under the curve. Equation 24 shows this summation as it was used in the computer program to obtain the relative residual concentration, CR/C0. 18 ------- CR (7-) 100 uo ={l- o, Wl i 100 (24) The required time and temperature corrections were applied to the data read from the strip chart before they were entered into a program on the GE 235. Table 3 is a printout from this program. The first two columns are the inputs of time and fluorescence concentration in the outflow. The third column shows the relative fluorescence of the outflow, (100C/CO). The fourth column gives the ratio of time to detention time, and the last column shows the mean residual relative concentration in the basin, (100CR/C0). The residual concentration calculated for the terminal time can be com- pared with the residual measured with the Fluorometer after thoroughly mixing the basin contents at the end of the test. The measured value was four percent and the calculated value was one percent for this example. Table 3. Data and Computed Values t, In seconds (1) • 68. 77. 79. 82. 85. 88. 90. 95. 98. 101. 103. 105. 106. 108. 110. 112. 113. 115. 118. 120. 122. 126. 131. 166. 182. 200. 240. 348. 368. 434. 469. 478. C, tn percent (2) 94. 94. 92. 90. 88. 86. 84. 82. 80. 78. 76. 74. 72. 70. 68. 66. 64. 62. 60. 58. 56. 54. 52. 50. 50. 48. 46. 44. 44. 42. 42. 40. 38. c/c0 (3) 100. 100. 98. 96. 94. 91. 89. 87. 85. 83. 81. 79. 77. 74. 72. 70. 68. 66. 64. 62. 60. 57. 55. 53. 53. 51. 49. 47. 47. 45. 45. 43. 40. t/*d (4) .00 .09 .10 .11 .11 .12 .12 .12 .13 .13 .14 .14 .14 .14 .15 .15 .15 .15 .16 .16 .16 .17 .17 .18 .23 .25 .27 .33 .47 .50 .59 .64 .65 CR/CO t, in seconds (5) (1) 100. 648. 91. 660. 90. 691. 89. 725. 89. 743. 89. 767. 88. 814. 88. 855. 87. 910. 87. 960. 87. 980. 86. 1008. 86. 1071. 86. 1160. 86. 1182. 86. 1304. 86. 1312. 85. 1427. 85. 1481. 85. 1555. 85. 1602. 85. 1638. 84. 1713. 84. 1780. 82. 1955. 80. 2004. 79. 2266. 77. 2340. 70. 2677. 68. 2837. 64. 2965. 62. 3000. 62. 3305. C, in percent (2) 38. 36. 34. 32. 30. 30. 28. 28. 26. 26. 24. 24. 22. 22. 20. 20. 18. 16. 16. 14. 14. 12. 12. 10. 10. 8. 8. 6. 6. 4. 4. 3. 3. C/C0 (3) 40. 38. 36. 34. 32. 32. 30. 30. 28. 28. 26. 26. 23. 23. 21. 21. 19. 17. 17. 15. 15. 13. 13. 11. 11. 9. 9. 6. 6. 4. 4. 3. 3. t/'d (4) . .88 .90 .94 .98 1.01 1.04 1.10 1.16 1.23 1.30 1.33 1.37 1.45 1.57 1.60 1.77 1.78 1.94 2.01 2.11 2.17 2.22 2.32 2.42 2.65 2.72 3.07 3.18 3.63 3.85 4.02 4.07 4.48 CR/CO (5) 53. 52. 50. 49. 48. 47. 45. 43. 41. 39. 39. 38. 35. 33. 32. 28. 28." 25. 24. 23. 22. 21. 20. 19. 16. 15. 12. 12. 9. 8. 7. 7. 5. 19 ------- TEST RESULTS The computed values for each run were plotted and the smooth curves shown in Figures 10 through 13 were drawn. The solid line was drawn through the visual average of the data. The dashed lines show the range for two identical runs having the greatest spread, and the dotted lines border the inclusive field for all data taken on the basin. 100 DESCRIPTION Visual average of all data Inclusive field for all data Identical runs with, widest spread Idea! "| Flushing^ Curve J o 1.0 2.0 3.0 3.8 Relative Length of Flushing Time Flushing Curves for 10-Foot by 4-Foot Basin (X-Y) 100 DESCRIPTION Visual average of all data Inclusive field for all data Identical runs with widest spread Ideal 1 Flushing}- Curve J 1.0 2.0 3.0 3.8 , Relative Length of Flushing Time 11. Flushing Curves for 10-Foot by 6-Foot Basin (X-Y) 21 ------- 100 80 60 i § 40 O ------- With a data spread for identical conditions nearly as great as the in- clusive field for all tests of a basin, the effect of the test parame- ters is either completely masked or barely discernible. Motion pictures of identical conditions show completely different jet behavior. Selected frames from the motion pictures do not show the com- plete jet action and the dispersion with time but they do indicate stages of the dispersion and jet movement and permit comparisons to be made. Figure 14 shows the persistence of "hot" islands of pollutant for Run I that do not occur in identical Run II. In the first run of another test, the inflow jet spread laterally almost across the basin, mixed with the dye, and swept the pollution toward the outlets in a "front." In an identical run the inflow jet remained iso- lated and mixed only at the edges so that a clean or slightly mixed stream penetrated to the outlet. Whether such jet behavior is a random phenomena or is triggered by subtle differences in the test conditions remains to be determined. DISCUSSION OF TEST RESULTS Two distinct yet related flushing actions take place. Initially, the clean water, displaces polluted water and this is the most efficient flushing possible. In fact, if a barrier could isolate the clean water from the polluted water and prevent mixing as the jet spread laterally, the basin would be cleansed in a time, t/tj = 1.0. Figures 10 through 13 show this ideal flushing curve as a straight line between Cn/Co = 100, t/td = 0.0; and CR/CO - 0.0, t/td = 1.0. The efficiency of the flushing can be judged by how close the experimental curve comes to this ideal curve. However, mixing does occur and this modifies the flushing curve. Ideally, the incoming water should spread across the basin and mix uniformly at the front. This would accomplish two desirable objectives: 1) the time of the displacement phase would be long, and 2) there would be no iso- lated islands of highly polluted water left in the basin. Figure 15 shows the strong demarcation that occurs between the original basin water and the inflow when the jet diffuses across the basin later- ally. This run had a long initial displacement time, t Figure 16 shows a reproduction of the average curves for the basins studied. The wider the basin, the greater the relative time, t/t^, re- quired to reach a given relative residual concentration. The time dur- ing which the flushing curve coincides with the ideal flushing curve is inversely proportional to the width of the basin. 23 ------- Run I Run II CS ^ ..__ 360 SECS. Fig. 14. Identical Run Comparison, Runs 60 and 61 X by Y = 6 ft by 10 ft, Wi =8 in., d ••-- 2 in., v - 0.1 fps. Note dye islands in Run I. Arrows indicate circulation 24 ------- Run I Run II . 6 1300 SECS. DYE ISLANDS FORMED i H • L DYE i: i • * Hm 2420 SECS. DYE ISLAND PERSISTS 1670 SECS. NO ISLAN. Fig. 14. (continued) 25 ------- 50 SECS. 3 SECS. 193 SECS. 260 SECS. Fig. 15. Example of Well-Diffused Jet 26 ------- 100 _ 80 o ZJ •o £ 240 c |