Hydraulicsof Stratified Flow

Engineering and Research Center

Bureau of Reclamation
Denver, Colorado 80225

d,
i;
 . .
achieved modifications t o the previously presented by US. Bureau of ~'eclamatiorl(USER) designers i n
tentative theory on selective wirhdrawal from stratified designing facilities for selective withdrawal from reser-
reservoirs. The author also attempts t o developand voirs. The contents,should also be of interest t o other
presmt the theory with design curves a n d formulas researchers in ;his field.=Emphasis is pl%d on the
that are of practicable significance. hydraulic engineering aspects of sel,?ctive withdrawal.
.- \\. , .
*
j,
*\
RESULTS lNTROD.UCTlON
: +=:
1. Attempts t o correlate inaccuracies i n withdrawal This third and final report completes a series d e a 6 g
layer thickness prediction with variation i n the with the hydraulics o f stratified flows as applied t o
density gradient from the assumed linear distribu- selective withdrawal from reservoirs.
~
tion proved inconclusive.
'I
'I The studies described by these reports were initiated
2. Dimensionless velocity distribution curves wore on the premise that:many 'water quality parameters
developed for withdrawal layers that were either follow the patterns established by reservoir stratifica. ..
unrestricted o r restricted by the bottom- or the tion. It was also reali5ed that the quality of reservoir
water surface. I n these curves the dimensionless outflow could be controlled through::;selective with-
velocity term (the local velocity divided by the drawal; however, knowledge o f the mechanics of .,
maximum velocity i n the withdrawal layer) is stratified flow and selective withdraw9was limited and
correlated t o the density gradient across the with- more accurate predictive abilities were needed t o
drawal layer and the relative location within the
optimize design and operation o f selective withdrawal ,:
withdrawal layer. structures.
'? . ,,
\:\
3. I t was ger:q-ally observed that for withdrawal
layers not restricted by the water surface or bottom,
'
I n t h e first report i n this series D. L. King presented
a summary of the basic theories and principles dealing
the elevatidh o:'the maximum velocity in the flow
with stratified flows and selective withdrawl. He also
was the same as that o f the center of the withdrawal
discussed hydraulic modeling problems which include
outlet.
similitude and modeling law questions as well as
physical modeling facility and instrumentation diffi-
4. For withdrawal layers that are restricted by
culties. Finally, i n the initial report King evaluated the
either the water surface or the bottom, the location
state of research as of 1966 i n which he not only
of the maximum velocity was found t o shift from
presented a review of literature and a summary, of
the outlnr'centerline towards the restricting bound-
USBR activities, but also an evaluation of areas needing
ary. A curve o evaluate this shift was developed.
additional research and a proposal for research by the
The curve correlates the relative positions of the
Hydraulics Branch, Division of General Research of the
restridti;;< boundary and the outlet centerline (with
USBR.
respect t o the total withdrawal layer thickness) t o
the maximum velocity location.
The second report i n this series." also by D. L; King, ; ,
reviewed past research i n reservoir stratification and'
5. A correlation between uneven discharge distribu-
selective withdrawal. He then presented a tentativ;
tions w ~ t h i nthe withdrawal layer and the pred~cted
theory for aiding i n the solution of design and
half-layer thickness (distances from outlet centerl~ne
operationz<ilective withdrawal problems. I n his analy-
t o layer boundar~es)was developed for half layers
sis King modified the formula for the densimetric
that are n o t restricted by a boundary. It wasfound
Froude number as suggested by Debler.3 This formula
that variations from a uniform d~schargedistribu-
is:
tion could be evaluated. These i n turn could be used
t o develop modified discharsesfor
. use i n evaluation
o i corrected withdrawal layer boundaries. Tliis
F'=- v 0.28% 0.04 (11 i'
correction is only meaningful for unrestricted half
layers. The thickness of the restricted half layers is
established by physical lunits and therefore cannot
be modified. ...
,$!
/+
"Superscript numbers refer t o references listed at the end of this report.
'i
r/

. .
 V = average vClocity i n withdrawal layer
g' = g Aplp It is also noted that, this discharge correction factor i s
po = density at orifice centerline probably not completely accurate. King recommended
&p = density differential across layer in the second r e p o r t ~ h afuture
t work should evaluate
d = thickness o f withdrawal layer the velocity distribution i n the withdrawal layer and
that this information could be used to.determine the
He then developed e~uationsof the general form: discharge distribution i n the layer. This could be done
for cases both w ~ t hand w ~ t h o u twater surface or
Vo
'----=K- Wd bottom interference.
F-&% D' (2)
, C?
.
!r, Because 'of. the nature of the study, it was not possible,
where to consider specific reservoir shapes a@ outlet config-
urations. These factors change from site t o site and
2 = velocity through &ithdrawdorifice
= constant depending upon the sllape,of the
withdrawal orifice and the value of the - thus do n o t lend themselves t o a generalized research
study: The results of this analysis are therefore most
critical densimetric ~~~~d~ number applicable to straight, uniform reservoirs with rela2vely
tive withdrawal c i n be accomplished only symmetrical and unrestricted outlet placemenis. The
for densimetric Froude numbers below the results of this study can be expected to be representa- ,

W
critical value)
= channel width
. .. a.3'
.. - tive for many facilities. Highly sinuous reservoirs,$'
reservoirs with severe constrictiop%i,ntake ~tru~;uies q2 .
d = withdrawal layer thickness above - w i t h indirect access to the rese-sow, and otlier similar
: ii
. the orifice centerline physical factors can be expected t o reduce the accu-
D ';= diameter or vertical width of outlet. ., racy o f representation. It may also be desirable, in
some cases, particularly for larger structures or struc-
King then rearra;&d terms t o obtain: tures for which the selective withdrawal ability i s :
critical? to refine this analysis. Model stdies of specific
D4 v02 installations can consider factors that are beyond the
= Ap K' d3 W' (31 scope of this study and therefore can provide accurate
9 predictive capabilities and the most effective design.
which is a convenient form of the equation that may
be aprlied easily i n digital computer solutions. This I n this, the third and final report, an attempt i s made
to develop more accurate predictive methods. This
analysis, however, contains several assumptions and
additional accuracy i s gained through the development
simplifications which limit the flexibility and accuracy
of modifications to the basic formula, equation (3).
of the method. The first of these assumptions is that
~ h e s kmodifications attempt to consider the effect of
the density gradient across each half layer is linear. I n
both deviations from a linear density gradient and
actuality, however, this i s almost never the case. I n
deviations from an squal discharge distribution. This
many cases the deviation from linearity i s extreme. A
report also attempts t o present design curves and
second and equally significant assumption is that the
procedures that are of practicable significance.
total discharge is equally divided between the upper
and lowel portions of the withdrawal layer. This
assumption is probably erroneous i n all cases where the
dens~tyq ~ a d ~ e nIS t not balanced about the withdrawal TEST FACILlTY AND PROCEDURE
centerline. However, this error is generally most severe
i n those cases involving surface or hot!& interference. Figure 1 shows the flume used for the laboratory tests.
As King noted: , . A refrigeration system in the flume was used to create
,,' the density stratification. The strength of the stratifjca-
,
When intcrsectjqn with the reservoir bottom or tion could be controlled easily by adjusting a control
water surface okcurs. this assurnptia_n (equally di- thermistor. The stratification was monitored and re-
vided discharge) i s no longer valid. d is now less than corded by using a series of thermistor probes placedat
the value required t o satisfy equation (31. The rrdesired locations i n the test flume. The thermistors
discharge above or below the orifice is then adjusted \bere connectecl through a scanning device t o a digital
by multiplying the discharge by t h e ratio of the thermometer and a printer where temperatures y e
right side t o the left side of equation (31. The displayed and recorded with an accuracy of U.02 C.
 unit was switched o f f and the reservoir wasallowed t o
stand for 3 to 4 hours. This period o f time allowed
currents to dissipate and the reservoir t o stabilize.
Wher: the stabilization p e ~ i o d was complete, the
withdrawal layer was then given at least 20 minutes to
develop and stabilize, after which data were collected.
Crystals of pstassium pnrmanganate were dropped at a
given station i!?th;-!lume. A t the same time a
stopwatch was started. Then over a period of a few
minutes :he f l o w being withdrawn created a deforma.
tion i n the vertical dye streak created by the falling
crystals. The stopwatch was then stopped, and data
were collected either visually or pliotographically.Tlie
data included: ( 1 ) average water surface elevation for
the run. (2) upper and lower withdrawal layer bound-
ary elevations. (3) elevation ancl magnitude of maxi-
Figure 1. Tcrt flume and observation 01 withdrawal lave^ mum dye streak deflection, ( 4 ) outlet elevation and
Photo P801 -D-74321 discharge, (5) total time interval involved. and (6)
average temperature profile for the run. Where photo-
Two very accurate quartz probes, w i t h a digital graphic data wcre taken, total velocity distribution
thermometer, were used for calibrating ancl checking information resulted.
the thermistor probes. Outflow from the flume oc-
curred through a small orifice whose elevation was The test facility is a three-rlimensional model although
adjustable. The outflow was wasted ancl therefore n o t the reservoir shape has been idealized. The reservoir
returned t o the flume, resulting i n a falling water width is ionsiderecl i n all of the following analyses.
surface in the test flume. When attempts were made to Observations i n the model indicate that tlie withdrawal
maintain a constant watw surface elevation, data layer quickly grows to its full thicknessancl to the full
collection was more difficult because of extraneous width of the reservoir. The layer thickness is nearly
currents ~stablished by the inflow. The withdrawal constant w i t h respect t o time and distance from the
discharge 'was monitored w i t h a differential mercury outlet when the density gradient is constant. There-
manometer across a volumetrically calibrated 318.inch- fore, the analysis may be undertaken for any reservoir
diameter orifice. cross section consirlererl t o be representative.

Temperature was selected instead o f salinity as the

agent for creating stratification for three principal
reasons. First, temperature is a convenient medium for
With N o Bottom or Water
estal~lisliingand altering a stratified reservoir. Second,
Surface Interference
temperature stratification can be monitorecl easily with
one set of probes. Because saline stratifications also
Initial efforts were directed toward improving the
contain temperature stratifications, a dual probe syr.
accuracy of withdrawal layer thickness prediction with
tem with a superimposition of data is required to
the assumption of equal discharge distribution between
obtain the actual density gradient. Finally, temperature
the upper and lower layers accepted, while questioning
stratification creates a hydraulic model that more
*.he linear density gradient assumption. Noted was that
correctly represents the prototype molecular diffusion
f i n the theoretical development something other than
coefficient. As King noted:
.3 linear gradient was assumed, a nonlinear differential
The molecular diffusion coefficient for heat is cm q u a t i o n developed. Solution of this equation would
the order of 5 0 0 times greater than that for sodium LE difficult i f n o t impossible. Therefore, attempts were
chloride. This woulcl tend t o increase the with- made t o develop coefficients based o n the difference
drawal layer thickness i n the thermal heat models between the assumed linear and tlie true density
and thus decrease the apparent critical value of F :' gradients. The coefficients would be used to modify
the results obtained from the conventional analysis so
The test procedure presented herein was followed for that more accurate solutions would be obtained. A l l of
all data shown i n this report. As soon as the test flume the attempts made proved to be futile.
was freshly filled, the refrigeration unit was turned on
and allowed to operate for at least 16 hours. After the Attention was shifted t o the equal discharge distribu.
stratification had been thus created, the refrigeration tion assumption and i t s effect o n the analysis. A
 total v e l o c i t y ~ i s t r i b u t i o ninformation. The areas en-
compassed by the velocity distribution curve,
above and be!pw the orifice centerline, were
h i n i d with a planimeter. From this the discharges both
above and below the orifice centerline were deter-
mined. T!x:se discharges ranged from 25 t o 75 percent ..
of the total flow for cases'where n o bottom or w a h
surface interference occurred. The conclusion w
some method should be found t o &dict th
distribution and i o evaluate its effect on
.. drawal layer thickness calculati0n:'~ffo:ts
again centered on the photographic velocity distribu-
tion data. Using t?? dimensionless parameters devel-'
oped by Bohan and Grace: the,dimensionless
. ; ~ ~ . ~ i s r r i b u t i curve,
on sho+?:in Figure 2, was de
o risr rwuL
LEGo
E Nm
D
These parameters are:^^^

'.6 7 B 9 10 8.l . l;P

V .:
--
VAp
'I yap, and
,
.L-.
.,

where:
Ap = density difference of fluid bewee
. .
elevations maximum and - either directly or indirectly reduced the data collec-
the corresponding local velocity. tion accuracy. The thicknesses of the withdrawal
b m = difference fluid between the layers are so small that inaccuracies i n the thickness
elevationsOf the and measurement (including those due t o parallax) may
either the upper or lower boundary (de- be significant.
pends on which half of the withdrawal
layer is being examined)
2. Secondary currents.-The
-
.. ..7.
-\.;
relatively s m a l l - t i c k -
,
y = the verticaldistance from the
;
i v e l o c i t ~t o a point on the velocity distribu-
ness of the withdrawal layer may be susceptihk t o
\\ errors induced by small secondary currents ifrom
!;,k, tion
several sources. These currents may be the single
\&
Y = the vertical distance from the riaximum
most important source of error in fhe
velocity t o eithdr the upper or lower limit secondary .currents can be caused b ' v w i t h d r a w a l
of the zone of withdrawal from a restricted reservoir. Since the withdrawal
i3
v = the local velocity at y layer has horizontal limits, a .vertical flow must be
= the in the 'One Of established to supply water t o the layer. Secondary :,

withdrawal. currents can also develop when the dye i s dropped

Into the flume. The dye-caused currents are due t o
both the disturbance caused by the falling crystals
Flgure 2 also shows the dimensionless velocity distribu-
and t o density currents caused by the dyed water.
tlon curve develoDed by Bohan and and a few
prototype date points. he prototype data points are
from two sources: a Tennessee Valley Authority study 3. Falling warer surface elevation. -Returning water
of Fontana, WattsEar, and Douglas Reservoirs and a t o the flume i n an attempt t o maintain a constant
USBR study of LaKe Mead.6 Only a limited amount of water surface elevation would induce strong second-
prototype data is available; more would be required t o ary currents; therefore, no water was returned t o
obtain an accurate verification of the model data. the flume during these tests. These currents would
Several sources'of error, when combined, probably no doubt have severely hampered, i f not made
... .. yield the data scatter i n Figure 2. I n the model tests impossible data collection. The reservoir water
these sources of error are: ;.' surface was allowed t o fall as water waswithdrawn.
 density proflles for the total number o f runs. The
withdrawal layer boundaries also were evaluated
approximately for the average conditions. This
2. Eva,uate the : I ~ ve,o=ity ~ ~ ~
disfribution,-By using the dimensionless velocity
changing water surface elevation is a possible source
distribution curve, Figure 2, i n conjunction with
of additional scatter.
the assumed elevation of the maximum velocity (the
- outlet centerline elevation), the known reservoir
For the field tests the greatest source of data scatter, density gradient data, and the withdrawal layer
by far, is error caused by other extraneous currents i n thickness information that was predicted i n Step 1,
the reservoir. These currents result from tributary t h e velocity distribution for the predicted
.~,~flows;
>
outflows through the various spillways, outlet withdrawal layer is evaluated.
,works, alta generating facilities; and atmospheric
xchange (wind, heat, etc.). It would be 3. Integrate curves to determine discharge disfribo-
impossible t o still these currents:' i n a tion.-With this knowledge an integration of the
e reservoir. Possible scatter caused by these velocitydistribution curve i s carriid out, and unit :'
s may be extremely significant. width,, areas representative of the discharges both
..--
, -, above and below the ohfice centerline are evaluated.
"
btaining a dimensionles,~,,v~)ocity.distributipn thij: rebort the integration i s do?i ma
curve, a modified withdrarval layer thickness prediction ever, this also can be computerized.
can be undertaken. The elevation p i the maximum
velocity in the-withdrawal layer must be known Prior 4. Obtain modified discharges.-To consider a shift
t o making the analysis. Boha" and race^ note that. from a uniform discharge distribution, modified
,.* * the maximum velocity within the zone of discharges are developed for the computer program.
withdrawal, in most cases, did ?not occur :at the The ratios of the upper and lower integrations t o
elevation of the orifice center1ine"'and that "Data one-half of the total are first evaluated. These two
analysis indicated that the maximum velocity occurred ratios are then multiplied by the initial total
at the elevation of the orifice centerline only when the discharge to obtain two modified total discharges.
withdrawal zone was vertically symmetrical about the
elevation of the orifice centerline. The maximum ,5. obtain correctedlayer
thicknesspredjct;on.qhe
velocity occurred below the orifice centerline when the modified discharges from step 4 are then used
vertical extent of the lower l i m i t of the withdrawal instead of the initial assumed discharge as data for
zone was less than that of the upper limit. Similarly. the computer program i n Step 1. Thus, two program
the maximum velocity occurred above the orifice . . runs
~. are made, each with exactly the
centerline when the distance from the orifice cenlerline data except for the discharges. The withdrawal la
. t o the lower limit was greater than the distance from boundary limitsfrom the !wo runs are then uni
~.
the orifice centerline t o the upper limit." The auilpr to yield a corrected layer.~h~ p
found a similar tendency i n his data; but it was also upper~boundary from the run using the d
noted that in cases where the bottom and water surface based on the upper ratio (step 4) and t h e
did not interfere with the withdrawal laver. the shift boundary from the run using the discharge ba
from the centerline was small. I n these cases the the lower ratio (Step 4) form the new with
assumption that the maximum velocity occurs at the layerlimits,
elevation of the orifice centerline appears justified.
6. Obtain final layer thickness .p&$icti6n.-~his
The recommendzd method of analysis for obtaining method is convergent with additional applications
h e withdrawal layer thickness consists of six steps: of the steps. A process o f successive approximations
therefore can be applied until change i n the Pre-
1. Basic theoretical prediction.-The initial step dicted layer is negligible. The rate of convergence
consists of predicting the thickness of both the varies with the specific problem, but indications are
upper and lower portions o f the withdrawal layer. that five or less cycles would be satisfactory.
This prediction i s accsmplished by using the digital Observe also that the program used i n Step 1 could
computer program (Appendix) which applies the
 initi*,lly predicted -layer thickAess occu; only for reservoir i n the problem are taken f r o m Figure 3. As
cssei with extreme variations from a balanced? can be scen the full prototype reservoir depth was not'
discharge distribu;ion. I n most cases the t o b .i.,
i
J, modeled. It was realized that the withdrawal layer
nlodification will he only a small percentage o/,,??he , probably would not extend t o either the water surface
initial thickness. 4 c,' bottom. So hypothetical bottom and water surface
l,ounda,cier were used i n the problem to.reducethe
Sample Calculation With N o B o r o m ' amount"irf ,input data. I f the predicted withdrawal
or Surface Interference layer' reaches' these hypothetical Poundaries, then
additional data would be input. ;'As long as the
The sample groblem was obtained from T V proto- predicted w i t h d r a w a l layerdoes not reach the hypo:
type data on Fontana ~ e s e r u o i r .This
~ is done so that thetical boundary, the withdrawal layer will be the
the results can be ,?ompar@(l j o actual P r o t o W e same whether the hypothetical boundaries or the
observations. Figure 3. actual boundaries are used i n the computer p r o g r a m . i.

The following information i s used at the start of the

Step 1.-The known illformation i s entered into the
analysis:
computer prosram as shown i n t h e ~ ~ . ~ p e n d The
ix.
..:,.
Water surface elevation = 1643 feet r resulting withdrawal layer thicknesses are:
Channel width = 1240 feel
Orifice diameter = 28 feet
. ,
:,!.r From centerline t o jpper limit =;41.1 feet
Bottom:elevation = 1.363 feet : From centerline to lower limit = 32.8 feet
Orifice centerline elevation = 1456 feet
Nithdrawal discharge = 6500 cfs

DOWNSTREAM UPSTREAM /'

VELOCITY I N FT/SEC TEMPERATURE I N *F

F ~ g u r e3 T V A prototype data for sample problem

--
 .
RESERVOIR DESCRIPTION assumed t o be at the ei&atmn o f the outlet centerllne,
'the upper layer thickness equals 4 1 1 feet and the
Elevation :levation lower layer thickness equals 32.8 feet Random local
(feet) (feet) I ("FI elevations across the w ~ t h d r a w a llaver may then be
-- selected. 3
1363 63.2
63.9 - A t these local elevations the densities and.therefore the
1373
64.3,. "1-yAplYAp," t e r m may be evaluated. The velocity
1383
1393 65.1 : distribution curve, Figure 2, is used t o obtain the :

66.2 dimensionless \melocity distribution term (local velocity

1403
67.1 divided by the maximum velocity) at that elevation for
,1413
, j!
1423 68.4 plotting on Figure 4.

Step 3.-Integrate the areas contained between the

1433 ' 69.6
maximum velocity elevation and the upper and lower
1443 70.7
72.0 boundaries of the velocity d k t r i b u t i o n curve, Figure 4.
1453
'.?. 1463 72.9 , These terms are directly proportional t o the discharges
'2 1473 75.6 because the dimensionless velocity term i s directly
1483 76.3 proportional t o the local velocity.
1493

5 1 4 t . 7 4 2 + 9 9 2 + 1 2 3 8 t 1456+ 1 7 1 0 + 1 0 1 t 3 . 0 5 ~ . 1 5 0 1 = 1 9 9 6 2
Lower A r e o = 2.5 ( l o + 1766+1.522 + 1 2 2 4 t 8 7 8 t . 5 3 0 + 0911 + I 4 1.091)=17654

t Averoge A r e a = 18.808.
upper Rotio:; 1.061
Corrected Q ; 6896
,:" , l>. ,

J
Discharge for i n i t i a l &olvot1on=6500CFS
Lower Ratio =0.939
"' Comected Q = 6104

hp ELE

300 497.1
253 491
226 486
194 481
157 476
121 471
91 466
53 461

56 451
116 446 ..
199 441
313 436
429 431
548 426
608 423.

1,::
.'
I
1350
0 .I .2 4 .6 .B 1.0
LOCAL VELOCITY/ M A X I M U M VELOCITY

Figure 4. Sample problem. no boundary effect.

 Step 4.-Divide the area terms evaluated i n Step 3 by predicted thicknesses must therefore be limited to the
the average of their two values. This yields the two half layer that is n o t interfered with.
discharge ratio terms (upper and lower layer) that,
when multiplied by the initial discharge of 6.500 cfs, A dimensionles; velocity distribution curve Was
yield the upper and lower corrected discharges. For developed i n a manner similar t o the no-interference
this problem the corrected upper discharge is 6,896 cfs, case. Photographs of the withdrawal layer were
and the corrected lower disrharge is 6.104 cfs. analyzed and a curve, Figure 5, based on the parameters
. proposed by Bohan and ~ r a c e ~ w developed.
as These
Step 5.-The computer runs as in Step 1 are now made parameters are quite similar to, but not the same as,
using the discharges from Step 4. The 6,896 cfs yields the ones ~ ~ s e
i ndthe no-interference case, Figure 2. The
an upper half-layer depth of 42.3 feet, and t h e 6,104 curve deveQped by Bohan and Grace is also shown i n
cfs yields a lower half-layer depth of 31.9 feet.\+ Figure 5. NO prototype velocity'distrihution data were
available to use i n verifying these curves.
"Step 6.-One more cycle was computed through; the
resulting upper half-layer depth was 42.9feet, and the Once again the probable elevation of the maximum
resulting lower half-layer depth was 31,4.feet. The velocity was needed. Again the maximum velocity
predicted and observed results are shown.gn Figure3. generally was located on the same side of the outlet
--. centerlineas the thinner layer and, Therefore is usually
By dividing the area contained within the entire on the same:sije as the restricting boundary. I n this
dimensionless velocity distribution curve by the case the distance from the maximum velocity to the
discharge for a u n i t width of reservoir. the maximum outlet centerline was significant enough t o evaluate. T o
velocity may be determined. The dimendonless develop a curve that would aid i n predicting the
velocity term and therefore the area contained within maximum velocity location, dimensionless parameters.
' . the dimensionless velocity distribution curve are as proposed by Eohan and Grace: were utilized.These
linearly proportional to the true velocity, and the term parameters were:'
of proportionality is the maxinlum velocity. From thg
maximum velocity, the total velocity distributionmay Y Z
be easily determined. For the sample problem, the - and -
H H
volume defined by the dimensionless velocity '
distl-ibution curve with a unit width was approximately
38 cubic feet and the d~schargefor a unit wldth of zr 1
reservoir was 5.24 cfs. The maximum velo-ity ,,
therefore would be 5.24138 or 0.137 fps.

This computed maximum velocity compares t o an

observed maximum velocity o f 0.09 fps which is '
-
reasonable aqreement i n view of the rather complex
prototype velocity profile. A comparison of the
predicted and observed profiles is shown i n Figure 3. ,
-.'
With Bottom or Water Surface Interference
I ,I ..-/.
/ ,
.$......
-80M.N
'
L ..,N,O
'
.L
1 1-' 4
A similar evaluation wakndertaken for cases c h
either the bottom or the water surface interfe,yed with
the withdrawal layer. Once again, attempts t o modify
the theory so that nonlinear density gradients would be
considered proved futile. So efforts were again shifted
t o an attempt to evaluate the significance o f the .(

discharge distribution assumption. As was noted earlier

in the reoort. Kinq2 - recommended the use of a
d~scharged ~ s t r ~ b u t ~factor
o n developed from the ratio
of the two sides of equation (31 at the boundary layer. , F,gure 5. Velocity dlrtribut,on stratified flow with
The extent of the half layer that i s affected by the boundary effects.
boundary is set by the physical dimens~ons(boundary
and cutlet elevations). Any correction t o the ~ n i t ~ a l l y
-.
 Bottom elevation = 0.00 feet
Orifice centerline elevation = 1.50feet
Withdrawal discharge = 0.00690 cfs
The ratlo of the two is 0.0391. This y~eldsan estimat
The reservoir lnformatton gwen in Table 2 would also discharge of (0.0391! (0.006901 or 0.000135 cfs In t
be known. restr~ctedhalf layer. This would make the unrestricted
discharge (0.00690-.0000135) or 0.006765 cfs.
Table 2
Step 3.-Theelevation of the maximum veloc~tyi s now
RESERVOIR DESCRIPTION predicted. It is known that:

Elevation Temperatures Z = 0.07 feet

!feet) ("CI IGRICCI H = 0.31 feet

I
'i
0 033 09997407 so, 11H = 0 226 I

0.167 .9997334
0.300 .9997249 I n referring this t o Figure 6, it is observed that Y/H =
0.433 .9997154 0.18 and Y is therefore 0.058 feet.This means that the
maximum velocity is 0.012 feet above the outlet
.9997011 centertine at an elevation o f 1 5 1 2 feet.

.9996660 Step 4.-The, problem then i s to-determine the velocity "

.9996424 distribution across the %tire-layer. Following the
calculation procedures as shown anCfipure 7, the
.9996026 dimensir;nless velocity distribution is ob,medt
d for the
.9994788 entire withdrawal layer.

.9982442 Step 5,-The dimensionless curve obtain'kt in Step 4 i s

th:n integrated t o determine not only \;Fe discharge
distribution ,:for flow above and below the outlet
Step i.-The known information is entered intr, the centerline but also the maximum velocity. The calcula-
Program as shown in Appendix The tionprocedure for evaluating the unrestricted discharge
resulting predicted withdrawal layer thicknesses are: is shown on Figure 7. As for the maximum velocity.
once again the ratio of the unit width discharge t o the
-
~. ~rom centerline t o upper limit = 0.07 feet 'integral of the entire dimensionless velocity curve is the
From centerline to lower limit = 0.24 feet maximum velocity. -
: >,

I t should be observed that the withdrawal layer step &-The modified discharge for the unrestricted
extends to, and therefore is restricted by, the water half-layer thickness calculation is.then evaluated. With
surface. the reitricted half-layer discharge of 0.000135 cfs from
7 Step 2 and the unrestricted half-layer discharge of
,.
Step 2.-Evaluation of the left side of 'equation (3) 0.00475 cfs from Step 5, the modified discharge is
yitlds: 0.004886 cfs. This is then inserted into the program of
Step 1.
D~ Po vO2= (0.0417)~( 0 998244) (5.0512
9 32.2 :-L Aftei two applications of this cycle the unrestricted
"Y half-layer thickness is pred~ctedas 0.20 feet. The total
or 0.00000239 withdrawal layer thickness is 0.27 feet. This compares
t o an observed th~ckness o: 0.32 feet during the
and evaluation of the rlght s~deof equation (31 y~elds: laboratory test.
 , , t.ole)(l. + .965)11 / 2. = 0.0987
c :.
Area below center line = n.05)(.965 + .732) + [.0831[.732+ .3781 + t.067) (-378 + ,1341 +
1
1.04~1.134~ / 2. = 0.2165
The lower discharge is t h e r e f o r e (0.2165/0.3152!. Or
0.688 Of the tot01
Lower discharge = (0.6881 10.00691 = 0.00475 Cfs

-
1.0

.3
.,

A
E L E I-* v/V
=
W
ap YAP"

2.' 000 0 1512 1 I

062 3461 1450 929 732
= 145 7964 1.367 615 378
212 10533 1 300 .255 134
252 11899 1260 0 0

0 1 2 3 4 5 6 7 8 9 1 0
LOCbL VELOCITY / M A X I M U M VELOCITY

Ftgure 7. Sample problem wlth boundary effects

REFERENCES 5. Wunderlich, W. 0. and Fan. L. N., "Turbulent

Transfer In Stratified Reservo~rs." ASCE Hydrauhcs
1. King. D. L.. "Hydraulics of Stratified Flow-First Divis~onConference, lowa Clt?, lowa. August 18-20,
Proqress Report-An Analysis of the State o f the A r t Ig7'
a n d a Definition of Research Needs." US. Bureau of
Reclamation Report No. HYD-563, Juae 1966 6. Sartoris, J. J. and Hoffman, D. A., "Meas~lrementof
Currents in Lake Mead With the Deep W% lsoto~~c
2. King, D. L., - , ~ ~ dof stratified
~ ~ ~ ~ l i ~l ~
~ Current
~ - Analyzer
~ ~ ~ U.S.
fDWICA)." ~ Bureau~ of Reclama-
d -
Progress Report-Selective Withdrawal From R ~ tion
~ Report
~ ~ No.. REC-ERC-71-38, October 1971
voirs," U.S. Bureau of Reclamation Report No.
HYD 595. September 1969 7. Elder, R. A and Wunderllch, W. O.."Evaluat~on of
Fontana Reservoir Fleld Measurements." Report No.
3. Debler, W. R., "Stratified Flow Into a ~ ( n esink," 17-90,Tennessee Valley Authority,August 1969
Journal of the Engineering Mechanics Division. ASCE,
July 1959

4. Bohan, J. P and Grace, J. L.. Jr.. "Mechanics o f

Flow From Stratified Reservoirs i~ the Interest of
Water auahty," Technical Report H.69-10. US. Army
Eng~neerWaterways Experiment Station, July 1969
CHAPTER 2. GENERAL I V F O P M P T I O N . ......
CHAPTER 3. 'NPUT. ...............
CHAPTER 4. CURMITTAL

CHAPTEP 5. VUTD'IT .........

CHAPTER 6. b P P E h l D I X

r!-ihpTEP 7. ~FOE~IDIX

CYAPTEP 8. WPEYDIx C ........... .....

CHAPTER 9. PPPEMDIX D .... .........

16
COMPuTFR PROCRAMS D F v E L O P F n BY THE BUREAU RF
SUBJECT T O TPE rOLLOWIN& CONDITIOMS. CrNSUL

PSSISTANCE WITH CONVERSION TO OTHFR COMPU

PROvICED. T H F PRQGPAMS H A V F R E E N P E v F L O P E F FOR

P N n tiU WARRAVTY 0 5 T O ~ C C U R A C Y I U S E F U L N E S S OR

EXP&SSED O R IMPLIED
S A T I S F I E S THE E Q U A L I T Y . T H F S E TWO n E P T H S A P F T H E U P P E R A N D LOWER

L I M I T S OF THC WITHPRAW4L LAYER

THE COMPUTATTON TS C A R R I F C FORWARC i N A S f R I F S O F 5 T E P S 4 s SHOWN

IN THE A C C O V ~ A N Y T N CF L O W W A R T , PFGIMKING WITH THE CO~RECTION OF

T V E T E M P E R A T I I R E P E P D T N ~ SF O R T H E V A R I C U S LFVELS (THIS C4N BE

O M I T T E n F Y RFMOVTNG T H R F E ST4TEMEWTS F P O H T K E P R O S F A M ) . WITH THF

C O R R F C T T E M P F R A T U R F S T H E ' ~ F N S I T I E S A R E T Y F N COYPIJTFD. THE LEFT

HAND TERM TS THEN EV4LllATEP FOR THF PARTICULAR OUTLET

C0NDIT:IONS. T H E PROGRAM TWEN E V A L U A T E S T H F R!GYT WAND TERM AT

FACH TFMPERATuRE L E v F L STARTING FFOK THE HIGHEST. THESE VALUFS

A R E T H E N COMPARED TO T Y F L E F T H4h1n TFRM UNTTL THE P ~ I ~ TO F

EQUALITY I S PASSFD. THAT INTERVAL I S THEN BROKEN INTO 100

T N C R F M E N T S A N D A c A I N Tl4E RIGHT HAN" TEOMS ARE COMPVTED AND

r O M P A R E D TO T H E L E F T H A N q T E R M U N T I L T H E POINT Or FQuALITY I S

A G A I N 9ASSED. ~ i l FP O S I T I O N O F T H E UPPER L T M I T OF uITHn44WAL IS

.~
..,.

T H U S OF!TATNEn. A SIMTLPP P R O C E n U R F 15 ' l k ! ~ h ! EXECUTfg TO D B T A I N

i'
T H E L O W E R 90\1NIJ4Qy. T H E P F C G R A M WILL COMPENSATE FOR CPSES I N

I~HICI-I EITHER THE K A T E ~ R S S R F A C E 0 9 - THE R O T T P M I -/ iS , L O ~ A T E D IN WHAT

W O U L D : 3 T H E R W T S E RE T H E C W P U T E O WITH~RAWAL LAYER. ALSO I T WILL

S O L V E C A S E S TN r l H I C H T H E U P P C R A N D . L C W E R hOUND4RTES 4QE BOTH

PETWEEY THE CAME TEMPERATURE LFVELS.

IN P T
-----1.

THE FIRST $1 DATA CARnS CONTAIN COPFCcPOYDTNF VALUES OF

19
 SUBYITTAL INSTRUCTIONS
......................
T H E D E C K S w I L n R E S T A C K F P ACCORDING T O THE D I ~ G R P ~ ' J N APPENDIX

P.

OllTPL!T
------
P R I N T E F r)UTP18T C I ? N S I S T S 3F: 1. LISTIEIC- OF THE ELFvA T I O N A B O V E

T H E ROTTOM. T 4 E rORQFCTE9 TEMPERATURE AT lHAT LEVEL , AND THE

PENSTTY AT TH4T LEVEL 2. SHAPF OF THE OIJTLET (SQUARE OR

CIRCULAP) 3.THi OUTLET SI7E 4.THF OUTLFT ELEVATI O N 5.THE

DISCHARGE THE I I P P E P L I P ? T OF W I T H D R A W A L THE LOWER LIMIT OF

kITHPRAb44L. BN E X A V P L E O F T H I S P 9 I N T F D OUTPUT ? S SHOWN I N THE

SAMPLE PROBLFP ( h P P E N I X R ) . WHEN P R O I G T Y P E D A T A AQE \!SED* THF

F G R M P T F O R P D I N T T N F 'HE D I S C H A R G F MUST B E MOPIFIED (FOR EXAMPLEI

FROM F a . 5 TO F R . O ) .

/,
*
,A
9
+
TNPUT D A T A
Elevation Temperature Elevation
(feet) (feet)

5.67 202
5.83 212
 7.1150 1351)
a
.......... -.,...I

CONVERSION FACTORS-BRITISH TO METRIC UNITS OF MEASUREMEPIT

the ASTM Metrtc Pmctlce Gwde.

The metric uom and mnverrlon factors ado~redby the ASTM are baned an the "inrwnanonaiSyrrem of Unm"
~ ~ & q w : e 51
j for S~:rcmc 1nlernal.onil b'Un7ell. f k e d b, tnc lntcrnarmna Covmlttce for Wriqnn and
Mear~rer t h n w t r a ir &o k M n n as m e Cora; or MKSA tmetcrrl ogrlm Imarrlrecond arnwrel wrlem Tnlr
i,rien h r &,plea oy !,e nwrn.t on01 D ~ g a ~ : r t l i ofor
n S t d w d J ral on 1 SO R v c a r r m c r o ~ l ' o nR 31.

that force which, when applied to a body havinga mars c.51 kg, giver it an acceleration of 1 mheclxc. There unitr
must bc dirtinguirhed from the iinconrtantl local weight of s budy having a m a s of 1 kg, that is. theweight of a
bDdy in that force with which a body is attracted t o the earth and is equal to the marsof a body multiplied hv the
acceleration due t o gravity. Howeuer. becaur~i t is general practice ro ue "pound" rather than the technically
correct term "pound-force." the term "kilogram" lor derived mars unit1 h a been u a d in this guide instcad o t
"kilogramforce" i n exprerring the c a v e r r i ~ nfactors for forcer The newton unit of force will find increasing ur:$
and is ~rrentialin SI units.

Where approx~mateor nommal Engi8rh unrtr are ured to expres a value or range of valua. the cmveited metrts
u n t i l in parenthere are also approximate or nommal Where prmse Engllrh unttr are used. the converled mewc
u n m are e x p r e r d ar equally slgnlficant valuer

Table i

OUANTITiES AND UNITS OF SPACE

Multiply 8V T J obtain

LENGTH

Mii ................. 25.4 (exastlyl . . . . . . . . . . . . . . . . . . . . . . Micron

Inches ............... 25.4 lexactlyl . . . . . . . . . . . . . . . . . . . Millimeters
Inches ............... 2.54 lexacrlyl' . . . . . . . . . . . . . . . . . . Centimeters
Fcef . . . . . . . . . . . . . . . . 30.48lexactlyl . . . . . . . . . . . . . . . . . . Centimeters
Feel . . . . . . . . . . . . . . . . 0.3048 iexacrlyl. . . . . . . . . . . . . . . . . . . . Meters
Feet . . . . . . . . . . . . . . . . 0.0003048 lexactlyl* . . . . . . . . . . . . . . Kilometers
Yards . . . . . . . . . . . . . . . 0.9144 lexactiyl . . . . . . . . . . . . . . . . . . . . Meters
Miles lrtarute) . . . . . . . . . . 1,609,344Iexactlyl. . . . . . . . . . . . . . . . . . . . . Mcterr
Miles . . . . . . . . . . . . . . . . 1,609344lexactlyl . . . . . . . . . . . . . . . Kilometers

AAEA -
Square inchsr . . . . . . . . . . . 6.4516 1 e x d i y l . . . . . . . . . . . . . . Squarecentimcterr
Square feet ............ '929.03 . . . . . . . . . . . . . , . . . . . . Square centimeters
Square feet ............ 0.092903 . . . . . . . . . . . . . . . . . . . . Square meterr
Square yards . . . . . . . . . . . 0.835127 . . . . . . . . . . . . . . . . . . . . . Square meterr
Asrer . . . . . . . . . . . . . . . . '0.40469 . . . . . . . . . . . . . . . . . . . . . . . . Hectares
Acres ................ '4.046.9 . . . . :.. . . . . . . . . . . . . . . . . . . Square meterr
'0.0040469 . . . . . . . . . . . . . . . . Square kilometerr
2.58999 . . . . . . . . . . . . . . . . . . "Square kilometerr

VOLUME

16.3871 . . . . . . . . . . . . . . . . . . . Cubic centimefcrs

.......... ....................
I.
CAPACITY
<A'

. . . . . . . . . . . . . . .. . .. ..... .. . C. ~ b i ~ c e n t i m e t e r r
79 ~ 7 2 7
28.5729.. ,.:
. . . . . . . . . . . . . . . . . . . . Milliliters
0.473179 . . . . . . . . . . . . . . . . . . Cubicdecimelers
0.473166 . . . . . . . . . . . . . . . . . . . . . . . . Liters
Ouans 1US.l . . . . . . . . :. . '945.358 . . . . . . . . . . . . . . . . . . . Cubic centimeters
Ouanr (U.S.1 ........... '0.946331 . . . . . . . . . . . . . . . . . . . . . . . . Liters
Gallanr(U.S.l . . . . . . . .,. . . '3.785.43 . . . . . . . . . . . . . . . . . . . . . Cubic centimeters
GallonrlUS.1 ......-... 3.78543 . . . . . . . . . . . . . . . . . . . . Cubic d e e i m e ~ r r
Gailoor [US.) ........... 3.78533 .. .. .. .. .. .. .. .. .. .. .. . . . . . . . . . . . . . Liters
Galianr (US.) ........... '0.00378!543 . . . . . . . . . . . . . . . . . . . Cubic meters
Gallonr (U.K.1 .......... 4.546W . . . . . . . . . . . . . . . . . . . Cubicdecimeters
Gallonr(U.K.I .......... 4.54596 . . . . . . . . . . . . . . . . . . . . . . . . . Liter*
Cubic feet ............. 26.3160.. . . . . . . . . . .... .. .. . . .. .. . . . . . . Liters
Cubic yards ............ '754.55 . . .. . . . .. .. ..... . . . . . . . . . . . Liters
krefeet ............. '1.233.5 ...
. . . . . . . . . . . . . . . . . . . . Cubic meterr
Acre-feet ............ -1,233,500 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Literr
 Table Il-conimum

QUANTITIES AND UNITS OF MECHANICS ~vIuI~;~;~ ay ~~~b~~~

Multlpl" 8, Toabral" WORK AND ENERGY*

1 8 ~ u l. . . . .
Brit;* Vlerntal ~ n i o '0.252 . . . . . . . . . . . . . . . . . . . . . . . . . . . Kilogram caloriet
Bliti* mmmal units i.td ..... 1.055.06 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . JOuleS
Grains l l l 7 . O M L I . . . . . . . . . 6479891 Ieramlvl . . . . . . . . . . . . . . . . . . . . . . . . Miliivaml atu p r pound ............ . . . . . . . . . . . . . . . . . . . . . . . l ~ u l pergram
2.326 ( e ~ a n l y l c
~~o~ommr l a w gainrl . . . . i:31.1035 ................................ Grams FwrpOundr . . . . . . . . . . . . . -1,35582 . . . . . . . . . . . . . . . .< . . . . . . . . . . . . . . . h d e x
Ouncerl&dpl ............ 28.3495 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Grams
P m d s lavdpl ............ ......................
0.45358237 l ~ x a ~ t l y l Kilogramr POWER
Shon~ons(2,OMlb) ........ 907.185 .............................. Kilogiamr
Shon rons(2PMib) . . . . . . . . 0.907185 . . . . . . . . . . . . . . . . . . . . . . . . . . . . M c l r i c r m l Horxpawn.. . . . . . . . . . . . . 745.700 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . W"ir
. Conglons 12.240lbl . . . . . . . . 1,018.05
- ................................ Kilograms Blu p r hour . . . . . . . . . . . . . 0.293071 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .W'".
...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . WdN
FORCEIAREA -~
Foot-pound$ per remnd 1.35582

HEAT TRANSFER
P ~ a~ ~ . .
r ~ u inch . ~. . . . ~ 0.070307
~ . . . ~. . . . . .P. . . . . .~ . Kilraramr p r vlvareemtimrt~r
Pounds per rquoie inch ....... 0.689476 ................. Nwdronr per q v a r ~ c e n I i m i t e r 8tu in.1hr t c 2 d g e e F Ik.
Po~ndrpersquarefmt . . . . 4.88243 .................... KilogrrmsparquaremeIcr Vlorml conducliv8lyI . . . . . . . 1.442 . . . . . . . . . . . . . . . . . . . . . . . M i i l i w s ~ l c ~ d e gCr e ~
d . tool
~ o ~ ~ square ~ ~ r . . . . 47.8803 ..................... ~nutonlprquaremeur Blu i n i h r ff2degre F (k.
thermal mnducsivityi ....... 0.1240 . . . . . . . . . . . . . . . . . . . . . . . Kgcallhi m degrsec
-1.4880 . . . . . . . . . . . . . . . . . . . . . Kg -1 mlhr m2 degree c
- MASSIVOLUME (DENSITY1 am nibr tc2 degree F . . . . . . . .
i BNhr n2d w e F IC.
thermal mnductanml ....... 0.568 . . . . . . . . . . . . . . . . . . . . . . . ~ i l l ~ a r d r m ~ d c g r e e ~
Blvlhr tt2dcgce F IC.
merm.l mnducronce) . . . . . . . 4.882 . . . . . . . . . . . . . . . . . . . . . . . . KQwllhr m 2 w r e e c
Degrr F hr 1t218tv IR.
rerirmcc~
,llcrmai ........ 1.751 . . . . . . . . . . . . . . . . . . . . . . . ~ ~ r r c r m ~ l m i l l i w r
Bfu4bdnlree F lc. heat capacity! . 4.1868 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jlgdegrr,
B n l l b &qrw F . . . . . . . . . . . . 'l.MO . . . . . . . . . . . . . . . . . . . . . . . . . . Ca1tgramdegresC
F12lnr lthemsl ditlulivity! . . . . 0.2581 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . cm2l*c
.
F ? I ~ lthormal dilBsiuity) .... -0.09290 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ ' l h r

WATER VAPOR TRANSMISSION --

BENDING MOMENT OR TOROUE

VELOCITY

Felmrx m d ........... 30.48 lerrotlyl . . . . . . . . . . . . . . . . . . Centimeters per sand

Frfpaxond ........... D3M8lerartlyI' . . . . . . . . . . . . . . . . . . . M~!eirperrecond
F e l per year ............. -0.985873 x 10-8 . . . . . . . . . . . . . . . Centimeters par recond
M i l e r p o hour............ 1.W9344 (eiacllyl . . . . . . . . . . . . . . . . .Kilomeierr per hour
Mil- per hour .............. 0.44704 lexmdy! . . . . . . . . . . . . . . . . . . . Mnerr per wsond; Table III
ACCELERATION.

Fcerperrerand2 ........... .0.3048 . . . . . . . . . . . . . . . . . . . . . . . . . Meler5perr~mnd2 , . ' M"lli.1" BY TO obtain

FLOW
C u b r feet perrquars 0 0 1 per &y beepsgel .... '304.8 . . . . . . . . . . . ~ i l e r r w r r q v r r em-ter pa day
.. ~o~ndmo~dr foot l v i ~ ~ i t .y.l . ... '4.8824 . . . . . . . ~ i l ~ g a mm n d per r p v a r ~msicr
Cubic l e t per I R m d
llecuno-tae!l ............ -0.028317 . . . . .: . . . . . . . . . . . . . . .
..........
S u a w f e ~ pt r x m n d i v i ~ o s i w l '0.092903
. . . . . . . . . . . Square m e t e ~ p e r % o n d
Cubic 1e1 perrn~nvzc . . . . . . . . 0.4719 . . . . . . . . . . . . . . . . . . . . . . . . . .~ i r ~ ~ . ~ ~ F~d m~n h.n i l~d s g
d r e ~lrhmgc).
............. 519 eiarfly
....
...... OB8309 . . . . . . . . . . . . . . . . . . . . . . . . . . LRer. per xrond Vol'l per lnil ...................... . . . . . . .Celriusor
. . . . . Kelvin degrcesIrhmgal*
10.764 . . . . . . . . . . . . .Lvmcnr per square meter
Gallonr IUB.1 per minute 0.03937 Kl!ovoi!sper millimmer
~ u m per m square foot i f o n t - m d w ........
FORCEf ohm-circuiar mils cer loot . . . . . . . . . . . . . . 0.001662 . . . . . . ~ h ~ r q u rmillimsrarr
re per meter
foot . . . . . . . . . . . . . . .
MillhUrie~p e r e u b , ~ '35.3147 . . . . . . . . . . . MiIlimrLr permbir mPrlr
Pards ................ .0.453592 .............................. Milliampsperquarefoot . . . . . . . . . . . . . . .
. . . . . . . . . . .
................ '1.4482 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kilogramr
.. Ga!lonl per square yard . . . . . . . . . . . . . . . .
'10.7639 Milll~mp5~er5~uarem~ter
'4.527219 . . . . . . . . . . . Litersper 5 W W e metor
Pounds
Pounds ................ '4.4482 x 105 . . . . . . . . . . . . . . . . . . . . . . . . . . .Nmvmnr
.Dyner Plrundr per inch . . . . . . . . . . . . . . . . . . . . . -0,17858
. . . . . . . . . . . ~~~~~~~~~~~~~~~CN
G P O 854.215
Selective outlet works provide an important means b y which the quality of water Selective outlet works provide an importaiit means by which the quality of water
withdrawn from reservoirs may be controlled. This i s the thirdand final report in aserier withdrawn from reservoirs may be controlled. This is the third and final report in a series
and ir part of a continuing effort t o develop accurate pracllcable derign and operating and ir part of a continuing effort to develop accurate practicable design and operating
criteria for ruch outlets. The rtudier discurred here refine previourly developed analyses. criteria for ruch outlets. The studies discussed here refine'previaurly developed analyses.
including evaluation o f previous simplifying arrumptions, such as a linear density gradient including evaluation of previous simplifying assumption:. w c h ar a linear density gradient
and equal half-layer discharger. A method is presented for predicting velocity distributions and equal half-layer discharger. A method is presented for predicting velocity dirtributionr
within a withdrawal layer. Layers restricted by either the water surface or reservoir within a withdraw1 layer. Layers restricted by either the water rurface or reservoir
b o i f o m and unrertricted layers are considered. The method is compared with experimental bottom and unresrricted layers are considered. The merhod is compared with experimental
and prototype data. Step-by-step design procedures are included. and proratype data. Srepbvrtep design procedures are included.

........................................................................

ABSTRACT ABSTRACT

Selective outlet works. provide an important means by which the quality of water Selective outlet workr provide an important means by by>ieh the qvality of water
w i t h d r a w from reservoirs may be controlled. Thir is the third and final report i n a series withdrawn from rcrcrvoirr may bc controlled. Thir is the t h i i d a n d final ieporr in a rerier
and ir part of a continuing effort to develop accurate practicable derign and operating and ir par? of a continuing effort l o develop accurate pr<&icable derign and operating
criteria for such outlets. The studies discurred here refine previourly developed analvrer. criteria for ruch outlets. The studies discussed here refine previously developed analyses.
including evaluation of previous simplifying assumptions, such ar a linear density gradient inc;uding evaluation ofpreviour simplifying assumptions, such as a linear density gradient . ,_- ..----
and equal half-layer dircharger. A method is presented for predicting velocity distributions and equal half-layer d;&harges. A method is presenred for predicting velocity dirtribut;onr
wilhin a withdrawal layer. Layerr restricted by either the water surface or reservoir within a withdradilayer. Layers rerrricted by either the water surface-a.?.'rerervoir
bottom and unrestricred layerr are considered. The method i r compared with experimental bottom and unrestricted layers are considered. The method ir eornpared&ith experimental
and prototype data. Stepby-step derign procedures are included. and prototype data. Step-by-step design pro