EPA-R2-73-238
JUNE 1973 Environmental Protection Technology Series
Flow Augmenting Effects of
Additives on Open Channel Flows
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
EPA-R2-73-238
June 1973
FLOW AUGMENTING EFFECTS OF ADDITIVES ON
OPEN CHANNEL FLOWS
By
Charles Derick
Kevin Logie
Contract No. 68-01-0168
Project 11020 GQG
Project Officer
James W. Newsom
Region III
U.S. Environmental Protection Agency
Philadelphia, Pa. 19106
Prepared for '
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20102
Price $1.00 domestic postpaid or 75 cents GPO Bookstore
-------
EPA Review Notice
This report has been reviewed by the Office of Research
and Monitoring, EPA, and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation
for use.
-------
ABSTRACT
Two model open channel configurations (trapezoidal and rectangular)
and three water soluble polymers (Polyox Coagulant, Polyox WSR-301,
and Separan AP-30) were used to experimentally determine the effects
of injecting dilute polymer solutions into open channel water flows.
It was found that for all test cases, injection of the three polymer
additives produced flow characteristic changes reflected as either a
water surface level decrease at constant flow rates or a flow rate
increase at constant static heads. These flow characteristic changes
were found to be dependent, in varying degrees, on channel slope, sur-
face roughness, injection point location, polymer injection method,
flow Reynolds number, and injected polymer concentration.
In addition, two flumes (Parshall and Leopold-Lagco) and two model
side channel spillways (90° V-notch weir and sharp-crested rectangular
weir) were used to determine experimentally the effects of polymer
additives on the flow measuring characteristics of energy dissipators.
It was found that for specific flow systems, the flumes and spillways
could be recalibrated for use with polymer/water flows. However, for
the Leopold-Lagco flume and sharp-crested weir, it was- found that as
the flow increased, the hydraulic jump inherent to these devices was
dissipated by polymer injection, thus preventing further recalibration.
This report is submitted in fulfillment of Contract No. 68-01-0168
under sponsorship of the Office of Research and Monitoring, Environmental
Protection Agency.
ii-i
-------
CONTENTS
SECTION TITLE PAGE
ABSTRACT iii
LIST OF FIGURES vii
I CONCLUSIONS AND RECOMMENDATIONS 1
Conclusions 1
Recommendations 2
II INTRODUCTION 5
Background 5
Purpose 6
Scope 6
III POLYMER SELECTION 9
Literature Search 9
3/4-Inch Pipeline Calibration Tests 9
IV OPEN CHANNEL FLOW EVALUATION 21
Experimental Tests 21
Open Channel Polymer Performance
Results 29
Theoretical Evaluation
(Velocity Profiles) 41
V FLUME AND SPILLWAY TESTS 47
Experimental Test Arrangement
and Procedures 47
Test Results 49
VI DISCUSSION 55
ACKNOWLEDGMENT 59
REFERENCES 61
GLOSSARY 63
v
-------
CONTENTS (cont.)
SECTION TITLE PAGE
APPENDICES 65
A. Fluid Friction Reducing Polymers
Literature Survey 65
B. Formulas Used for Data Analysis 73
VI
-------
FIGURES
PAGE
1. EXPERIMENTAL 3/4-INCH PIPELINE POLYMER
CALIBRATION TEST SETUP 10
2. POLYMER SOLUTION MIXING DEVICE (3/4-INCH
PIPELINE TESTS) 12
3. POLYMER PERFORMANCE CURVES (3/4-INCH
PIPELINE TESTS) 16
4. POLYMER PERFORMANCE CURVES (3/4-INCH
PIPELINE TESTS) 17
5. POLYMER PERFORMANCE CURVES (3/4-INCH
PIPELINE TESTS 18
6. POLYMER EFFECTIVENESS CURVES (3/4-INCH
PIPELINE TESTS) 19
7. EXPERIMENTAL OPEN CHANNEL TEST SETUP 22
8a. SIMPLEX METERING TUBE WITH ORIFICE 24
Sb. PITOT-STATIC TUBE,TEST ASSEMBLY 24
9a. POLYMER MIXING DEVICE (OPEN CHANNEL TESTS) 26
Sb. POLYMER INJECTION SYSTEM (OPEN CHANNEL TESTS) 26
10. SMOOTH-WALL OPEN CHANNEL POLYMER PERFORMANCE
(POLYOX COAGULANT) 31
11. SMOOTH-WALL OPEN CHANNEL POLYMER PERFORMANCE
(POLYOX WSR-301) 32
12. SMOOTH-WALL OPEN CHANNEL POLYMER PERFORMANCE
(SEPARAN AP-30) 33
13. REFERENCED DATA COMPARED TO PRESENT OPEN
CHANNEL DATA (POLYOX WSR-301) 34
Vil
-------
FIGURES (cont.)
PAGE
14a. MAXIMUM POLYMER EFFECTIVENESS (Re=5.0xl04) 35
14b. MAXIMUM POLYMER EFFECTIVENESS (Re=8.5xl04) 35
15. VISUAL RESULTS OF POLYMER PERFORMANCE
(OPEN CHANNEL TESTS) 36
16. ROUGH-WALL OPEN CHANNEL POLYMER PERFORMANCE
(POLYOX COAGULANT) 38
17. INJECTION POINT DEPENDENCE (POLYOX COAGULANT) 40
18. TYPICAL VELOCITY PROFILES FOR TRAPEZOIDAL ,
OPEN CHANNELS 42
19. TYPICAL VELOCITY PROFILES FOR RECTANGULAR
OPEN CHANNELS 43
20. THEORETICAL VELOCITY PROFILES COMPARED TO
THOSE EXPERIMENTALLY DETERMINED IN AN OPEN
CHANNEL AND PIPELINE 44
21. FLUMES AND SIDE CHANNEL SPILLWAYS 48
22. LEOPOLD-LAGCO FLUME CALIBRATION DATA
(WITH/WITHOUT POLYMER) 50
23a. SHARP-CRESTED WEIR CALIBRATION DATA
(WITH/WITHOUT POLYMER) "51
23k. 90° V-NOTCHED WEIR CALIBRATION DATA
(WITH/WITHOUT POLYMER) 51
24. PARSHALL FLUME CALIBRATION DATA
(WITH/WITHOUT POLYMER) 52
25. TYPICAL VISUAL EFFECTS FOUND FOR POLYMER
ADDITIVES WITH FLUMES AND SPILLWAYS 53
-------
SECTION 1
CONCLUSIONS AND RECOMMENDATIONS
CONCLUSIONS
As a result of data obtained from both a literature search
and 3/4-inch pipeline calibration tests, it was determined
that three polymers (Polyox Coagulant, Polyox WSR-301, and
Separan AP-30) exhibited sufficient "friction reducing"
characteristics at low injected concentrations to warrant
testing in the model open channel system. Other additives
such as Guar Gum were deemed unacceptable for open channel
testing, since they produced significantly lower 'friction
reducing" characteristics at injected concentrations con-
siderably higher than the other three polymers.
Experimental data obtained for two open channel configura-
tions (rectangular and trapezoidal) indicate that a flow
augmentation at a constant head, or a water surface level
decrease at a constant flow rate, can be obtained in open
channels by injecting water soluble friction reducing
additives in relatively low concentrations. Maximum flow
augmentation is achieved at an injected concentration
between 25 wppm and 50 wppm (depending upon the polymer
used), and then tends to remain constant or decrease slightly
with increased additive concentrations. It should be noted
that the flow augmentation values for pipeline flow continue
to increase until much larger additive concentrations are
reached before leveling off or decreasing.
Theoretical comparison using Prandtl-von Karman's Universal
Velocity Distribution Law shows that open channel flows
with polymer follows the same trends as pipeline flow and
flat plate flow. However, wall shear stress for open
channels is somewhat lower than that found for pipeline
and flat plate flows at a constant wall shear stress.
/
For smooth-wall channels, percent flow augmentation tends
to decrease as the Reynolds number increases. This may be
caused, however, by inadequate polymer mixing with the
channel flow. In contrast, for rough-wall channels, flow
augmentation is nearly independent of Reynolds number for
the same polymer concentrations.
-------
Both Polyox WSR-301 and PolyoxCoagulant were found to
rapidly degrade when transferred by means of a centrifugal
pump. Separan AP-30, however, was not totally degraded by
the pumping system. Thus, it may be concluded that this
polymer exhibits significant resistance to shear degradation
and may be the best candidate material for long channels.
Polymer degradation along the length of the model channels
was non-existent for the test setup used.
The distance traveled by the polymer additive after injection
prior to measuring flow augmentation was found to be signifi-
cant. Flow augmentation was greater for the polymer
injection locations further upstream and decreased as the
injection point was moved closer to the measuring stations.
The flume and spillway tests produced two distinct results.
The first result shows that if a flume or spillway is used
to "measure flow, calibration tests must be performed using
polymer additives at the specific flow conditions present
in the system. The second indicates that either the
inherent hydraulic jump of the measuring devices remains
within the devices, in which case a calibration curve may
be plotted, or the hydraulic jump is dissipated, thus
negating the flow measuring capabilities of the device.
A definite trend was established for the flumes and spill-
ways used in this program. As the Reynolds number increases
for a specific channel system, the probability of the
dissipation of the inherent hydraulic jump within the
measuring devices increases. Good calibration was obtained
for the Parshall flume, which was actually oversized for
the test setup used. It seems feasible to conclude that
a solution to the problem of hydraulic jump dissipation
for polymer-fluid mediums flowing within flumes or spillways
may be to install oversized devices.
REC OMMENDATIONS
Primary on the list of necessary tasks needed to be accom-
plished before polymers may be used in full-scale open
channel systems is obtaining data for channels varying in
size between those used for this program and those used in
-------
existing full-scale systems. These data are necessary
since no information of this nature has keen reported to
date by other investigators; in order to establish design
criteria to be used for present and future open channel
systems using polymer additives, comparison investigations -
should be performed. Polymer shear degradation along open
channel systems should be established.
Economic feasibility studies should be conducted for full-
scale open channel systems using polymer additives.
Additional flow measuring device (flumes and spillways)
investigations should be conducted to establish whether or -
not a universal calibration can be made for each device
when used for polymer-fluid measurement.
In light of the results of this project, a full-scale open
channel demonstration should be conducted and should include
both a force-fed and gravity-fed system (see Glossary).
The following municipalities have been contacted for said
demonstrations.
Fairfax County, Virginia
Monroe County, Rochester, New York
Onondaga County, Syracuse, New York
City of Cleveland, Cleveland, Ohio
City of Balitmore, Baltimore, Maryland
Dauphin County, Harrisburg, Pennsylvania
A "universal" polymer injection system should be developed.
This device should include a portability characteristic
and an automatic'capability. These features would enable
municipalities to easily eliminate their overflow problems
with a minimum of manpower and initial investment expense.
-------
SECTION II
INTRODUCTION
BACKGROUND
Combined sewer overflows have been recognized throughout the years
as major sources of water pollution. Also, sanitary sewer overflows
are caused by unlawful residential "tie-ins" to a main sewer system,
by unwarranted urban and suburban over-development which results in
insufficient sewer line capacities, .and/or ground water infiltration
through worn or unrepaired older sewer lines. In the past few years,
extensive investigation and development programs have been directed
toward significant elimination of this pollution problem. From
these programs, a number of specific anti-pollution methods have
evolved and are now being closely scrutinized and refined for opera-
tional use.
One of these methods is injecting high molecular weight polymers into
the flow of combined sewer lines at pre-determined concentrations.
It is believed that this injection tends to reduce the frictional
flow resistance of the waste water. More explicitly, the energy re-
quired to drive the waste water can be lowered by polymer injection,
and if one assumes a constant energy level, the flow capacity of a
given sewer line can be increased. As an added feature of such a method
used for increasing older, combined sewer line capacities, it has been
demonstrated that no additional stress due to fluid pressure is trans.-
mitted to these lines when their capacities are increased. Such a
feature has been welcomed, since many of the older lines would need
costly repairs if other flow augmentation methods were employed*
The techniques of polymer injection and the corresponding theory
associated with this phenomenon have been reported by many invest-
igators for completely filled conduits and for flow over flat plates.
Most of those investigations have been performed as laboratory exper-
iments, with a few being performed in full-scale sewer line systems
(Refs. 1, 2, and 3). Unfortunately, only a limited number of invest-
igations have been reported concerning open channel applications,
-------
and these are directed toward partially filled circular
channel configurations.
This report describes and presents the results of an
Environmental Protection Agency sponsored research project
which investigated and demonstrated the effect of polymer
additives on model open channel flows.
PURPOSE
The primary purpose of this project was to experimentally
determine the flow augmentation effects of water-soluble
polymers on the flow of water within different model open
channel configurations, and on energy dissipators such as
flumes, and side-channel spillways used as flow measuring
devices.
In addition, secondary purposes included investigating
polymer insertion methods, evaluating polymer degradation
resistance as a function of insertion point location," and
determining flow augmentation effects of polymer additives
as a function of channel slope, wall roughness, and polymer
type., A theoretical analysis for polymer flow in the open
channels using Prandtl-von Karman's Law of the Wall and
Velocity Defect Law was also performed in order to correlate
the experimentally determined open channel data with data
of previous investigations for flow in completely filled
pipelines and on flat plates.
SCOPE
The work performed for this project was essentially separated
into two tasks which were conducted concurrently. The first
task was composed of both a theoretical analysis and a
comprehensive literature search. The"second task consisted
of performing, in a laboratory atmosphere, the experimental
open channel tests.
Initially, all known references concerning' polymer additive
.effects on fluid flow within pipelines, within open channels,
and over flat plates were gathered and reviewed for
-------
applicability to the investigation. Based upon these data,
a theoretical analysis was conducted using the Prandtl-von
Kantian Universal Velocity Distribution Law. This theory is
utilized to predict velocity profile characteristics of
flow in pipelines, flat plates, and open channels as a
function of the wall shear stress and distance from the
wall.
Experimental tests and evaluations were conducted in
accordance with test plans (Refs. 4 and 5) prepared in the
beginning of the project by Columbia Research Corporation
and approved by the cognizant EPA project officer.
Open channel tests were conducted, using two channel con-
figurations; rectangular and trapezoidal. Data were taken
for each of these configurations for three types of polymers
at two Reynolds numbers, three channel slopes, two channel
roughness levels, and two polymer solution concentrations.
In addition, polymer injection methods and injection point
locations were studied and evaluated for use with the
particular model test system.
Flume and spillway tests were performed, using the open
channel configurations mentioned above. Two flumes
Parshall and Leopold-Lagco and two side-channel spillways
(weirs) 90 V-notch and sharp-crested rectangular were
investigated and calibrated with and without polymer
injection. These tests were conducted using one polymer
additive (found to have the greatest flow augmentation
effect in the open channel configurations) and two channel
slopes.
-------
SECTION III
POLYMER SELECTION
LITERATURE SEARCH
Prior to performing experimental open channel tests, it was
necessary to conduct an extensive literature survey
(Appendix 1). This survey consisted primarily of gathering
theoretical and experimental data pertinent to the friction
reducing capabilities of known water-soluble polymers.
Emphasis was placed on locating correlation data between
pipeline/flat plate flow and on data for open channel flow
systems. The results of this search indicated that little
information has been recorded outlining correlation pro-
cedures for use when comparing the effects of polymer
additives on the water flow within pipelines to that of
the flow over flat plates. In addition, it was found that
polymer data for open channel flow systems have been
scantily reported and pertain exclusively to semicircular
channels or partially filled pipelines.
Based upon this literature search and available information
supplied by polymer manufacturers, it was verified that
the four original polymer additive types specified in the
work statement of this project were superior to all other
polymers as friction reducing additives for water flow.
In light of this, it was decided that all testing efforts
should be directed toward using only these four polymers.
THREE-QUARTER INCH PIPELINE CALIBRATION TESTS
Experimental Test Arrangement
The experimental setup used for the 3/4-inch pipeline
calibration tests is shown in Figure 1. This test arrange-
ment included the following equipment:
,, Component Specification
Centrifugal water circulating pump 20 gpm
Variable displacement polymer feed pump 0 - 0.055 gpm
-------
Rotometer
Pressure Gauge
,Feed Pump
Manometer
Gate Valve
Gentri f ug_a1_
Pump
3/4-inch Pipeline
; : Water Reservoir
Polymer Metering
Pump
Mercury
Manometer
E| B
Rotameter
Polymer
Reservoir
Bypass
Gate Valve
(Flow Rate
Control)
Static Pressure
Gage
Circulation
Pump
Water
eservoir
A \
AD = 150 ft
EC = 130 ft
AB = CD = 10 ft
AE = 5 ft
FIGURE 1. EXPERIMENTAL 3/4-INCH PIPELINE POLYMER
CALIBRATION TEST SET-UP
10
-------
Component Specification
Bourdon-tube pressure gauge 0 - 100 psig
Rotometer 2.4 - 30 gpm
Manometer, mercury ' 100 in.
3/4-inch pipeline 150 ft
Steel tank (water reservoir) 12 gal.
During operation, the water was pumped ~by a Rivett centri-
fugal pump from a reservoir (steel tank) through a Fischer-
Porter rotometer into 150 feet of 3/4-inch nominal I.D.
pipeline back into the reservoir. A gate valve located at
the pump outlet was used to control the flow rates by
diverting excess flow through a bypass circuit.
Preparation of Polymer Solutions
The four different polymer products listed herein were
initially supplied in powder form and had to be prepared as
a solution prior to testing.
Manufacturer Polymer Resin
Union Carbide Corporation Polyox WSR-301 (Polyethylene
Oxide)
Polyox Coagulant (Polyethy-
lene Oxide)
Dow Chemical U.S.A. Separan AP-30 (Acrylamide)
Stein, Hall & Company, Inc. Jaguar Guar Gum (Poly-
saccharide)
The polymer solutions were initially prepared using the
mixing apparatus of Figure 2. Three different techniques
were employed for mixing these solutions, depending on the
polymer type and the concentration of the solution.
The first technique employed a vibrator to feed the polymer
in powdered form into a 1,000-milliliter beaker of tap
water stirred at approximately 300 rpm. As the solution
became more viscous, the rotational speed of the agitator
was reduced to approximately 100 rpm. The stirring was
continued until all of the powder was dissolved and the
solution reached a homogeneous state. This process was
11
-------
Driving Motor
Hot Plate Location
Vibrator
Driving Motor
Graduated
Beaker
Stirrer
_
1
*»
cr
s
^
/// S/SS/SSS
'/Hot Plate /
/ / / / / s / / / //
i
S t
' i
i
i
FIGURE 2. POLYMER SOLUTION MIXING DEVICE
(3/4-INCH PIPELINE TESTS)
12
-------
used in the preparation of solutions of Separan AP-30 and
low concentration (0.1%) solutions of Polyox Coagulant and
Polyox WSR-301.
Guar Gum solutions were prepared using a similar technique.
Since it was found that Guar Gum is readily soluble in
water, the vibrator was not required. The Guar Gum powder
was transferred from a strip of waxed paper into a beaker
of tap water under brisk agitation. As in the first
technique, the agitation was continued until the powder
was completely dissolved.
When mixing the more concentrated (1.0% and 0.5%) solutions
of Polyox Coagulant and Polyox WSR-301, homogeneous
solutions were difficult to obtain using the above techniques,
Thus, an alternate approach was used. A weighed sample of
Polyox was added to a bealcer of boiling water so that a
good polymer dispersion could be obtained. Note that
although both Polyox WSR-301 and Polyox Coagulant are
insoluble in boiling water, the increased temperature has
no effect on their potency as a friction reducing additive.
Room temperature water was then added, the heat source
was removed, and the stirring was continued until the
solution reached a uniform state. At this time, the solu-
tion was placed in a cold water bath and allowed to cool
to room temperature.
Polymer Injection Procedure
All polymer solutions were introduced into the 3/4-inch
pipeline flow downstream of the centrifugal pump by means
of a Meco-0-Matic chemical solution feeder pump. This
pump had the capability of varying the polymer feed rate
to compensate for the flow in the 3/4-inch pipeline in
order to achieve the desired concentration levels.
Experimental Test Procedures
In order to evaluate the effectiveness of each of the four
polymer additives mentioned above, calibration tests were
performed.
13
-------
The flow performance of each additive was rated in terms
of the flow increase that it produced at a constant head
loss (the head loss being the differential pressure
measured across 130 feet of pipeline). The data necessary
for this rating was obtained as follows.
The first step established base data using tap water alone.
An initial flow rate was set, and the corresponding head
loss and pump discharge pressure were measured with the
manometer and the pressure gauge, respectively. Once this
information was recorded, the feed pump was brought on
line, and a polymer solution was injected into the pipeline
flow. The injection resulted in a subsequent decrease in
both the head loss and the discharge pressure. In order
to return the head loss to its original value, it was
necessary to increase the flow through the pipeline by
closing the bypass valve. The new flow rate was measured
with the rotometer and recorded. It was then compared to
the initial flow rate using Equation (1) of Appendix 2.
By changing the polymer injection rate and repeating the
above procedure for the same initial pipeline flow, the
flow augmentation was expressed as a function of the polymer
concentration.
The polymer injection rate was varied in two ways by
changing the feed pump setting, and by changing the intial
concentration of the polymer solution. Three different
initial polymer solutions (.1%, .5%, and 1%) and four
different pump settings were used. This made it possible
to obtain at least 12 data points for each polymer additive.
To determine the injected polymer concentration in the
pipeline flow. Equation (2) of Appendix 2 was used. The
volumetric rate of injection "q.^", was determined using a
graduated beaker and a stopwatch, and the time required to
inject 100 milliliters was measured and recorded. All of
the other variables in Equation (2) were either known or
easily computed from the recorded data.
The entire test procedure was repeated for three different
initial pipeline flow rates to insure that the relative
effectiveness of each of the polymers was the same over a
range of Reynolds numbers.
14
-------
Calibration Test Results
In order to illustrate the effectiveness of the polymer
additives in the 3/4-inch pipeline flow, two typical methods
of graphical representation have been used. The first
method plots the experimentally determined flow augmentation
against injected polymer concentration for three different
flow Reynolds numbers. In addition, three different percent
solutions were used for each polymer tested thus permitting
the maximum injected polymer concentrations to be obtained
with the available test setup. The results using this
method are shown in Figures 3, 4, and 5. Comparing the
data on these figures, it is seen that Polyox Coagulant
and Polyox WSR-301 display approximately equal effective-
ness, followed by Separan AP-30 and Guar Gum. For the
polymer concentrations and the minimum and maximum Reynolds
numbers tested, Guar Gum produces a negligible effect on
the pipeline flow, compared to the other three polymers.
Thus, it was decided to discontinue testing with this
polymer. It is also clear that as the Reynolds number
increases, the polymer effectiveness increases and, for
the injected polymer concentrations tested, the effective-
ness continues to increase for constant Reynolds number.
The second method compares the pressure drop versus pipeline
flow rate data of undosed tap water against the pressure
drop at constant injected polymer concentration for water/
polymer flow. Such a method has been used successfully as
a check on the experimentally determined flow augmentation
data previously mentioned. As shown in Figure 6, the
derived flow augmentation for Polyox WSR-301 at a polymer
concentration of 12.4 wppm is approximately 72%. This
value corresponds closely to the experimentally determined
flow augmentation percentage found on Figure 4.
In addition, the results presented above were used to cali-
brate the polymer solutions used for the open channel
testing. This was done simply by repeating the pipeline
tests for each new polymer solution. Thus, the solutions
were mixed until the data on Figures 3, 4, or 5 were
duplicated, thereby insuring a known polymer solution
concentration.
15
-------
80
60
g
H
0)
I
40
20
/
ft
lA
Pori
fc
F"
V
/^A'
r/
/
nc
^
t!^
i-4
r
^
-
^~
!^>
rfja.-^"
ui
.. .
^l
^
a.j
-
. ...
^
^
a
-
-
1
Polyox Coaaulant
A -0. 1% Solution
A -0.5% Solution
A -1.0% Solution
Polyox WSR-301
Q ->.!% Solution
^-0.5% Solution
^-1.0% Solution
Separan AP-30
Q - 0.1% Solution
H- 0.5% Solution
U - 1.0% Solution
Guar Gum
G
«
a
/\
)- 0.1% Solution
>- 0.5% Solution
»- 1.0% Solution
-----
;
A
20
40
60
80
100
120
Injected Polymer Concentration (wppm)
Re = 2.05 x 104
Temp. = 70°F
FIGURE 3. POLYMER PERFORMANCE CURVES
(3/4-INCH PIPELINE TESTS)
16
-------
80
~ 60
a
o
H
40
3
20
«
./
*{/
p
A
/
0
1
y
r
/
^
/A
>
0«
jo
/
\
^
'B
-
.LI
a
-,-
^
^
--- -
,
- a
- -
-
-
Polvox Coaqulant
A-0.1% Solution
A-0.5% Solution
A-l.0% Solution
Polyox WSR-301
^-0.1% Solution
^-0.5% Solution
^-1.0% Solution
Separan AP-30
H-0.1% Solution
B-0.5% Solution
Di -1.0% Solution
...;_.
- .
r
- . -
20
40
60
80
100
Injected Polymer Concentration (wppm)
Re = 3.1 x 104
o
Temp. = 70 F
FIGURE 4. POLYMER PERFORMANCE CURVES
(3/4-INCH PIPELINE TESTS)
17
-------
80
60
0
H
40
H
fe
20
J
4
d
i
A'
I /
IT
;
r
Q
^
t-«
/
2k
/
rm
/"
. .- .
.
-
"^T
..
t
'
"
-.-
_ i
--
...
J
Polyox Coacmlant
A-0.1% Solution
A-0.5% solution
A-l.0% solution
Polvox WSR-301
0-0.1% Solution
^-0.5% Solution.
^-1.0% solution
Separan AP 3O
CD- 0.1% So
B- 0.5% So
Di- 1.0% So
3uar Gum
lution
lution
lution
O- 0.1% Solution
Q- 0.5% Solution
d- 1.0% Solution
':
'
-O
20
40
60
80
100
120
Injected Polymer Concentration (wppm)
4
Re = 4.1 x 10
o
Temp. = 70 F .
FIGURE 5. POLYMER PERFORMANCE CURVES
(3/4-INCH PIPELINE TESTS)
18
-------
a>
4J
a
4J
o
o>
n
Q
a)
n
3
10
0)
n
CM
90 -
80
70 _
60 _
50 _
40
30
20
15 _
10
/
i
/
1
1
5 6 '
/
/
m'
A
/
0
1
'
/
1
Tap WE
/
/;
/ r
/
/
iter
iran AI
'-JU
/Polyox Coagulant
/ '
/_ _ y^
//'
// '
v :\
Polyox
WSR-301
:ted Polyr
ner Cor
icent3
t
ratic
\ \ \ ill
7 8 9 ID 15 20 30 40 . 50 6
)n,
12
.4
wp
i i
0 70 80 90
?m
LOO
Flow Augmentation.
for Polyox WSR-301
Flow Rate (gpm)
FIGURE 6. POLYMER EFFECTIVENESS CURVES (3/4-INCH PIPELINE TESTS)
-------
SECTION IV
OPEN CHANNEL FLOW EVALUATION
EXPERIMENTAL TESTS
Test Arrangement
The general layout of the open channel test facility is
shown in Figure 7. Two different model channel configura-
tions were used for testing r rectangular, and trapezoidal.
Both channels were fabricated from 60 feet of 1/16-inch
sheet steel. The first 20 feet of the channels were used
to ensure that uniform turbulent flow was developed, and
the remaining 40 feet were used for the test section. The
channels were supported by five supports which could be
adjusted to obtain slopes of 0%, 1%, 2%, 3%, and 4%.
Channel surfaces were initially painted with an epoxy paint
to give a smooth finish, and were then roughened by
repainting with a sand and paint mixture. Specifications
for each channel configuration are listed below.
1. Rectangular
Bottom width 6 in.
Side length 6 in.
Channel length 60 in.
Manning factor
(a) Smooth surface 009
(b) Rough surface 013
2. Trapezoidal
Bottom width 5 in.
Side length - 6 in.
Channel length 60 ft.
Interior angle 60°
Manning factor
(a) Smooth surface 009
(b) Rough surface 013
21
-------
Wei
Sight: Glass (typ
Centrifugal Pump
FIGURE 7. EXPERIMENTAL OPEN CHANNEL TEST SETUP
-------
A Bell and Gossett centrifugal pump (1200 gpm capacity) was
used to circulate the water, and flow rates were regulated
with a Bell and Gossett triple duty valve. The water was
pumped from a 300-gallon open reservoir (constructed of
plywood and lined with two 4-mil plastic sheets), through
a 10-foot, 6-inch diameter Simplex metering tube with
orifice, into 50 feet of 6-inch diameter PVC plastic pipe,
through 10 feet of flexible hose, into one of the channel
configurations. The flow then returned to the reservoir
via the channels, to complete the closed loop. Flow rates
Within the open channel system were measured using an
orifice plate and two manometer tubes, located at the
Simplex metering tube (Fig. 8a.). The manometer tubes
were connected to two pressure taps, one on each side of
the orifice plate, and sensed a pressure differential.
Other instrumentation used included a pitot-static tube
and four sight glasses per channel. The pitot-static tube
was used to determine local velocities (for the theoretical
analyses), flow depths, and flow velocity increase. The
tube was adapted to a mechanical positioner (Fig. 8b.)/ so
that it could function at any point in the flow volume.
The sight glasses were constructed of plexiglass and were
etched at 1/8-inch increments, enabling flow depths to be
measured. The glasses were located at the downstream end
of the channel test sections and were spaced at 5-foot
intervals.
i
Preparation of Polymer Solutions
Since larger quantities of polymer solution were required
for the open channel testing, solutions were mixed in a
54-gallon drum fitted with a two-bladed agitator (Fig. 9a).
The agitator was driven with a 1/15 hp, 66 rpm Dayton gear
motor. Three polymer solutions were prepared using this
apparatus: Polyox Coagulant; Polyox WSR-301; and S-eparan
AP-30. The polymers were mixed in 42-gallon batches and
at 1% concentrations.
The preparation of Polyox solutions was accomplished by
first pre-dispersing the powder in anhydrous isopropanol.
The slurry was then added to water which was being stirred
at 66 rpm. The stirring was continued until a uniform
solution was obtained. It was found that the simplest way
23
-------
.Flow
jDirection
Manometer
Leads
FIGURE 8a. SIMPLEX METERING TUBE WITH ORIFICE
Mechanical
Positioner
Pitot-Static,
Tube
FIGURE 8b. PITOT-STATIC TUBE TEST ASSEMBLY
24
-------
to prepare the solutions was to mix them in four stages;
one-quarter of the total quantity was mixed in each stage. .
When this was done, a better dispersion of polymer was
achieved, and the tendency to form polymer "clumps" was
minimized. The Separan AP-30 solutions were mixed using a
technique similar to the one used for the small-scale
tests. A weighed polymer sample was sifted into water
which was being stirred in the 54-gallon drum. When all
of the powder was dissolved, the stirring was stopped. As
was done above, the solution was prepared in four stages.
*
Polymer Injection Procedures
Two basic components of polymer injection systems were
investigated prior to selecting the system (Fig. 9b) used
for the open channel tests. The first component investi-
gated was the forcing device. Two were considered: a
mechanical pump, and a compressed air pump. Since compressed
air at the pressures needed was readily available and the
cost of a mechanical metering pump which would satisfy the
requirements was too high, the compressed air device was
used for the open channel tests. As shown, this device
consisted of a 42-gallon pressure tank and a 120 psig
compressed air supply. The air pressure was varied and
maintained by using a pressure regulator attached to the
tank. In operation, different polymer solutions flow rates
were obtained by changing the pressure settings. To
measure the weight flow of polymer solutions, a Fairbanks
Morris scale and a stop watch were used. The injection was
started and/or stopped with a gate valve connected to the
tank and transfer tubing.
The second component investigated was the device used for
injection of the polymer into the open channel once the
polymer was forced out of the pressure tank. Four such
devices were evaluated, the first two being selected for
the open channel tests (see page 39) . Each was rated
according to its ability to disperse the polymer throughout
"the flow volume. A red dye was mixed with the polymer
solution, and the degree of polymer dispersion in the flow
was visually observed. '
25
-------
. Two-Bladed Agitator
. andMotorAs si_emb ly
55-Gallon Drum
FIGURE 9a.
POLYMER MIXING DEVICE
(OPEN CHAN3STEL TESTS)
Pressure Regulator
7
42r£allon Pressure
Tank
Aluminum Block
Copper Tubing
FIGURE 9b.
POLYMER INJECTION SYSTEM
(OPEN CHANNEL TESTS)
26
-------
The first injection device investigated consisted of an
aluminum block drilled through at a 45° angle, and fitted
with a section of 3/8-inch copper tubing. The block was
attached to the underside of the channel with epoxy glue.
A 20-foot section of 3/8-inch Tygon tubing connected the
pressure tank to the injection device.
The second device was similar in construction to the first;
however, instead of injecting at a point, injection was
accomplished over an area. This was done by taping a
strip of screen (4 inches by 6 inches) across the bottom
of the channel, covering the outlet of the injection
device.
The third device was a plastic injection manifold. It was
constructed of plexiglass and had a 1-inch diameter inlet
which was connected to the pressure tank with a 10-foot
section of 1%-inch flexible tubing. The polymer was
injected into the flow through sixteen 5/16-inch holes.
The holes were evenly spaced across the lower sides and
bottom of the channel. This device was designed to fit
into the rectangular channel only.
The fourth device was the simplest. A 20-foot section of
3/8-inch Tygon tubing was used. One end of the tubing was
connected to the pressure tank and the other end was placed
inside and along the bottom centerline of the channel. The
polymer was introduced in the direction of the water flow.
To minimize the flow obstruction, only 10 inches of the
tubing were placed inside the channel.
.Experimental Test Procedure
Three polymers, selected from the 3/4-inch pipeline cali-
bration tests, were used for testing in the open channel
test facility. The first set of tests was conducted in
the smooth-walled channel. Each polymer was tested in
two channel configurations, at three slopes (1%, 2%, and 4%)
and at two Reynolds numbers (5 x 104 and 8.5 x 104).
The procedure to determine polymer effectiveness commenced
by setting an initial Reynolds number. This was accomplished
by adjusting the flow rate until a combination of the flow
27
-------
velocity and the hydraulic radius produced one of the above
Reynolds numbers. Once this flow rate was obtained, the
following data were recorded:
Dato
Channel slope
Orifice readings
Pitot-static readings
Water level - sight glasses
Initial weight of polymer in pressure tank
Polymer injection pressure.
Polymer was then injected into the flow at a point 49 feet
upstream of the reservoir. The injection point was located
along the bottom centerline of the channel.
As a result of polymer injection, a decrease in the water
level or static head occurred. Thus, the system was given
sufficient time to reach an equilibrium state, and a measure-
ment of the new water level was then made, using both the
sight glasses and pitot-static tube. After recording this
measurement, the flow rate was increased until the level
returned to its original value. The polymer injection was
subsequently stopped, and the data listed below were
recorded:
Orifice readings
Time of polymer injection
Final weight of polymer in pressure tank.
Using a manufacturer-supplied orifice plate calibration
curve and Equation (1) of Appendix 2, the percent flow
augmentation was calculated.
The next step was to reset the flow rate in the channel to
the same initial rate by closing the triple duty valve.
Once this was accomplished, the water levels were checked
to make sure that they returned to their original levels.
If they did not, it was necessary to change the water in
the system to prevent a polymer concentration buildup.
After the initial conditions were re-established, the
pressure tank was set at a different pressure, and polymer
was injected into the flow. The same data as before were
28
-------
recorded. The procedure was repeated at least four times
for each polymer solution, and each time a different polymer
injection rate was used. The percent flow augmentation was
expressed in terms of the bulk polymer concentration in the
flow. The concentrations were calculated using Equation (3)
of Appendix 2.
Tests to determine the effect of channel surface roughness
on flow augmentation were also performed, using the most
effective polymer found for the smooth surface tests. The
same two channel configurations (with a different degree of
wall roughness) were used, as well as the same channel
slopes and initial flow rates. Experimental procedures
were similar to the preceding ones, and the results were
similarly presented.
Three injection point locations were tested to determine
the effect of the point of injection with respect to flow
augmentation. These tests were performed in the smooth-wall
trapezoidal channel at a 2% slope, using Polyox Coagulant.
The first two injection points were located in the channel,
49 feet and 56 feet upstream of the reservoir. The third
point was located in the 6-inch pipeline, 16 feet downstream
of the pump. The methods of injection were the same as
those used in the polymer performance tests; however, for
the third point of injection, the polymer was introduced
through the pipeline into the flow.
Each injection point was evaluated in terms of the maximum
flow augmentation produced at a constant static head. The
polymer concentration in the flow was varied by changing
the injection rate, and the results were presented graphi-
cally. The experimental procedure followed the same steps
as the procedure outlined for the smooth- and rough-wall
open channel tests.
OPEN CHANNEL POLYMER PERFORMANCE RESULTS
Polymer Performance in Smooth Surface Channels
Polymer performance or effectiveness in smooth-wall open
channels has been illustrated by using a number of graphical
methods. The first of these methods uses curves.
29
-------
representing the experimentally determined flow augmentation
versus the injected polymer concentration. These curves
are shown on Figures 10, 11, and 12, and are plotted for
two Reynolds numbers, two channel configurations, and three
channel slopes.
Figure 10 shows the effectiveness of Polyox Coagulant. It
is seen that, in general, this polymer is most effective
in rectangular smooth-wall channels, and at lower Reynolds
numbers. The same results have been obtained for Polyox
WSR-301 and Separan AP-30, as seen in Figures 11 and 12,
respectively. These plotted data also indicate that a
maximum flow increase occurs at a relatively low polymer
concentration (between 20 and 30 wppm) and, unlike the
results found for pipeline flow, the flow augmentation
does not continually increase, but -rather levels off or,
in some cases, begins to decrease with increasing concentra-
tion.
In order to compare these experimental results with those
obtained by other investigators, a portion of Figure 11
has been expanded and is shown on Figure 13. It is seen
that this curve plotted for Polyox WSR-301 in a rectangular
channel flow at a Reynolds number of 5 x 10 and slope of
2% compares well with data obtained from Reference 6 for
the same polymer under similar conditions in a partially
filled pipeline. Note also that, as shown, the. same
trends appear in this reference data when comparing pipeline
flow to open channel flow.
In addition to the above curves, another graphical repre-
sentation has been made. Figures 14a and 14b represent
the maximum percent flow augmentation obtained from
Figures 10, 11, and 12 plotted against channel slopes.
In these figures a general trend is indicated, in which
flow augmentation decreases and then increases with slope
increase. However, this polymer effectiveness dependency
on channel slope is not concrete, since it is not indicated
by all data points.
Visual results of polymer effectiveness are typically shown
in Figure 15. As indicated, a water surface level drop was
produced as the polymers were injected into the flow. The
extent of this drop was directly dependent on the injected
polymer concentration.
30
-------
§
. id
P
G
B
H
50
40
30 -
20
10
50
40
30
20
10 .
GvC
7 ©/d*"
w
&
if
/
^--^
5
J^
0
- - .
Y
$^
r
^^~"
©
E
-H
^^
I
1 1~1
\
-
Reynolds Number
0- Rectangular
Q- Trapezoidal
Reynolds Number
3- Rectangular
3- Trapezoidal
- 5.
Chan
Chan
Chan
Chan
0
4
0 x 10
nel
nel
5 x 104
nel
nel
4
& Chai
r
inel £
Slope
2?
)
4 Chai
inel £
Slope
50
40
30
20
10
. L
-*-©
Channel Slope
10
20 30 40 50 60 70 80 90 100 110 120. 130
Injected Polymer Concentration (wppm)
FIGURE 10. SMOOTH-WALL OPEN CHANNEL POLYMER PERFORMANCE
(POLYOX COAGULANT)
31
-------
a
o
rl
50
40
30
20
10
50
40
30
20
S 10
50
40
30
20
10
/
£
a-H-
f
=© -
-ft
w
(7l
-y
J
JSB
B"n
s
*-
^&=:
~(7) i-
_ j
V
m
clB-iiJ .
O
Reynolds Numbe
O Rectangula
Q- Trapezoida
Reynolds Numbe
0- Rectangula
H- Trapezoida
-
L
/-\
0
]
r - 5
r
'
.0 x 104
r Channel
1 Channel
r = 8.5 x 104
r Channel
1 Channel
25
Yo Cha
T3
-
£ Chai
-
nnel
....
nnel i
-
Slope
Slope
:
A
Ld
Tir
4% Channel Slope
10 20 30 40 50 . 60 70 80 90 100 110 120 130
Injected Polymer Concentration (wppm)
FIGURE 11. SMOOTH-WALL CHANNEL POLYMER PERFORMANCE
(POLYOX WSR-301)
32
-------
50
40
30
20
10
50
40
30
Jj 20
c
i
g1 10
<;
&
H
fc
10
-H-
-G
1% Channel Slope
Reynolds Number = 5.0 x 10
O- Rectangular Channel
9- Trapezoidal Channel
i
Reynolds Number.= 8.5 x 10
D- Rectangular Channel
B - Trapezoidal Channel
-0
2% Channel Slope
.[]
-O
10 20 30 40 50 60 70 80 90 100 110 120 130
Injected Polymer Concentration (wppm)
FIGURE 12. SMOOTH-WALL OPEN CHANNEL POLYMER PERFORMANCE
(SEPARAN AP-30)
33
-------
CO
§
r\
-P
(0
0)
CP
I
100
80
60
40 .
20
"
1
Data from Reference 6
Re = 6 x 104
EJ- Pipeline Diameter=2.5 cm
El - Sewerline Diameter=7.5 cm
Data from Figure 11
Re = 5 x 104
Channel Slope = 2%
O Channel Configuration
is Rectanaular
[
E
Vo
f
-T
L
A
A<
y
/
'
i /
JS
'^
^f U
^ [
^^_t
]-
^
.
^
^
'
C
r
1
B ..._
ii
]
)
i
!
©
5 10 15 20
Injected Polymer Concentration (wppm)
25
FIGURE 13.
REFERENCED DATA COMPARED TO PRESENT OPEN CHANNEL DATA
(POLYOX WSR-3.01)
-------
U)
Ul
Max. Flow Augment at ion (%) Max. Flow Augmentation (°/
HtOOJiP'Ul H NJ W *> Ul
OOOOO OOOOO
I
(
i
- L
^\
-^
~~ L
1 2
Slope
FIGURE 14a.
;
i .
s
C
V
^-
/*'
^u
^-^
:.
>
£
i
J
^
',
. i .
^2!
" . t .
J
^~
^ -
..
.. .1.
_
. , i
.
^-^
k
-"^^
Ai.
3 4 1234
(%) Slope (%)
MAXIMUM POLYMER EFFECTIVENESS (Re = 5.0 x 104)
>
I '
-
C
(
J
I
i
. i
i
-- ' ;; -
[Polvox COc
O~ Rectar
Q- Trape2
,
..'
. |:
, -;'. ^
; '..
:.,r' fi
' ; S
, A
. (
L~_
--.
'i :
tqulant pp.]
igular Q .
:oidal . Q-
eparan AP-
^~ Rectang
^- Trapeze
j
; /
-d
;
».
S^
'i
.yox WSR-301
- Rectangular
Trapezoidal
30
ular
idal"
1.234 1234
Slope (%) Slope (%)
FIGURE 14b. MAXIMUM POLYMER EFFECTIVENESS (Re = 8.5 x 10 )
-------
Sight Glass (typ.)
Water Surf acej ;
Water Surface Level j
er Injectioni
I'MIMBMLT-' a?- -.- - - . ~ - . ^ .. -... J
Sight Glass (typ.)
:,<'.-r- f / r i
'' ' (. i I ' '
' ' ('. I I' ( I \
'','' I , I I I
v / ; < - i i k
Initial Water Surface
Level
Water Surface Level
After Polymer Injection
FIGURE 15. VISUAL RESULTS OF POLYMER PERFORMANCE
(OPEN CHANNEL TESTS)
36
-------
Summarizing the above results, it is found that:
1. Generally, Polyox Coagulant is the most effective
polymer under all test conditions. This polymer.
is followed in effectiveness by Polyox WSR-301
and Separan AP-30, respectively.
2. For the same Reynolds numbers and slopes, a
higher maximum flow augmentation was produced in
the rectangular channel configuration.
3. The maximum effectiveness of each polymer additive
occurred at relatively low concentrations
(20 - 30 wppm).
4. Plow augmentation is dependent, in varying degrees,
on flow Reynolds number and channel slope.
Effect of Channel Surface Roughness
Using Polyox Coagulant, the smooth-wall performance tests
were repeated in rough-wall channels. Flow augmentation
was obtained as a function of the additive concentration
at a constant static head. The results are presented in
Figure 16, and are compared to the smooth-wall data as
follows:
1. The concentration at which Polyox Coagulant
reached its maximum effectiveness was increased
from approximately 25 wppm to approximately
50 wppm.
2. At the higher flow Reynolds numbers, greater flow
augmentation was achieved in the rough-wall
channels. For example, in the rectangular channel
at a 2% slope, the maximum flow augmentation was
increased from 19% to 32%.
3. A trend has been established indicating that
polymer effectiveness represented by flow augmen-
tation decreases as slope increases for the high
Reynolds number, and essentially remains the same
for the lower Reynolds number.
37
-------
An -
o n -
JU
*>n -
10 -
1 .
't A
ifr
*/
a/
ft/
1 -
i
,-rf
.
0
_
i
i
i ^
i
i
1% Slope]
>
i i
Reynolds Number =
O- Rectangular
Q- Trapezoidal
Reynolds Number =
D- Rectangular
B- Trapezoidal
5.0 x 104
Channel
Channel
8.5 x 104
Channel
Channel
20
10
O
4% Slope
10 20 30 40 50 60 70 80 90
Injected Polymer Concentration (wppm)
100
FIGURE 16. ROUGH-WALL OPEN CHANNEL POLYMER PERFORMANCE
(POLYOX COAGULANT)
38
-------
Injection Point Location
Polyox Coagulant was injected into the flow at three
different locations. The results (Fig. 17) show that as
the distance traveled by the polymer additive was decreased,
the maximum obtainable flow augmentation was decreased.
When injection occurred approximately 110 feet from the
test section, a flow augmentation of 44% was achieved.
As the distance of injection decreased to 56 feet and 49
feet, the maximum flow augmentation was decreased to
20.5%, respectively.
Polymer Injection System
The injection system previously shown in Figure 9b was
found to be the most efficient for the particular 'test
setup used. As shown, it consisted of a compressed air
pump and an aluminum block fitted with a section of
3/8-inch copper tubing, and utilized both injection over
an area, and injection at a point. It was found that
injection over an area provided better polymer dispersion
at the injection site; however, at a location 10 to 20 feet
downstream, little difference was noted between the area
and point injection. Both resulted in a good dispersion of
pplymer with minimum flow obstruction, and were found to be
equally effective for both channel configurations. The
remaining two injection devices were found to be unsatis^-
factory for the following reasons:
1. The plexiglass device obstructed the flow.
2. The 3/8-inch flexible tubing failed to produce
sufficient polymer dispersion.
General Observations
Besides noting a water surface level decrease with polymer
addition, other observations were made concerning polymer
additives in open channel flows. These are presented below.
1. As the polymer additives were injected, the
characteristic turbulence and wave motion were
seen to diminish and the flow became more laminar.
39
-------
50
c
o
rl
4J
g
§
H
PM
40
30
20
10
/
i
//__
(
/
/
/
/-\
"fcr
-------
2. There was no polymer shear degradation noticeable
within the channels tested. However/ both Polyox
WSR-301 and Polyox Coagulant were found to rapidly
degrade when transferred through the centrifugal
pump. In contrast, Separan AP-30 never totally
degraded through the pump, consequently requiring
complete drainage of the test system after a few
test runs with this polymer.
THEORETICAL EVALUATION (VELOCITY PROFILES)
Test Procedure
Using the pitot-static tube, centerline and cross-sectional
velocity profiles were initially measured for tap water
flowing in each of the trapezoidal and rectangular channels.
The results are shown in Figures 18 and 19, and indicate
the type of flow within the channels. Centerline velocity
profiles were then measured in the same two channels at
two shear velocities for both water and water/polymer flows.
Local velocities and corresponding depths were non-dimension-
alized and graphically presented. The resulting experimental
points were compared with theoretically derived curves. The
theoretical curves were based on an equation derived from
the Prandtl-von Karman Universal Velocity Distribution Law
applied to smooth-wall open channel flow (see Ref. 7).
This equation is presented in Appendix 2 as Equation (4).
Results
Comparing the theoretical and experimental results for tap
water (Fig. 20), it is seen that good' agreement exists.
This, consequently, establishes the validity of Equation (4)
with respect to the flow in the two channel configurations.
Polyox Coagulant injection into the flow results in new
non-dimensionalized profiles as shown in Figure 20.
When compared with the curves for undosed tap water, an
upward shift is noted in each case. This shift, account-
ing for the velocity increase, is represented by a constant,
" AB", which when inserted into the Universal Velocity
Distribution Law, yields Equation (5) in Appendix 2. In
41
-------
Water Surface
3
4J
«J
o
m
3 2
S
6-5
I",
Distance
r~ -
, . -< j | [ j j -«-.- [-.
later Surfe
Maximum- Flc
ijeasurina £
ice Height i in.
>w Rate =2-23 gpm
tation S- = 2
_
_
--
-_
f
...
....
;
.
--
..-
:._
-
~i
~*
...
.
,r
TT-
X
9
^
^
<
(T1
"/
-
-
Cei
ite
rline profile.
FIGURE 18
local Centerlinfc Velocity
vCft/sac>
TYPICAL VELOCITY PROFILES FOR TRAPEZOIDAL
OPEN CHANNEL
-------
+J
+J
s
Water Syrfaqe
J -*r '""""" ~~"
7/////////S
4 *
§
o
1
15
-L^4^-l }_v.J-=.j^'-Ll- I-.-!-, I j-^j
.-':: ! . :}}" :!. i". I--!'/ ;:! :
2 ..,..:._
water Surtace Height = J.s in
Maximum Flow Rate = 260 gpm
Measuring Station =0= = 3
1 . _L_.U_.
Centerline Profile
Local Centerline Velocity
v (ft/sec)
O
H
43
-------
40r-
V/Vf
v/v,
I0 (wall shear stress) "=57.4 dynes/cm
;-H Predicted from data in Reference 8
'H
Polymer concentration = 13 wppm
(Polyox Coagulan
Theoretical Curve
© Tap Water - Rectangular Channel
0 Polymer Additive - Rectangular
Channel
6 Tap Water - Trapezoidal Channel
Polymer Additive - Trapezoidal
Channel
80.4 dynes/cm
Polymer concentration = 9 wppm
(Polyox Coagulant)
Predicted from data in Reference 8
.. Theoretical Curve i
10
102
103
FIGURE 20. THEORETICAL VELOCITY PROFILES COMPARED TO THOSE EXPERIMENTALLY
DETERMINED IN AN OPEN CHANNEL AND PIPELINE
-------
theory, this constant is a function of the polymer additive
properties and the wall shear stress. The same phenomenon
is encountered in pipeline flow and flow over flat plates.
The constant, "AB", has been evaluated for pipeline flow
by several investigators, and is plotted in terms of the
wall shear stress of Fabula (Ref. 8). Thus, for purposes
of comparison, the "AB" values obtained for pipeline flow
at given wall shear stresses and polymer concentrations
were used in Equation (5) and a predicted curve is also
plotted in Figure 20.
The experimental points found for the open channels are
seen to lie below the predicted curve, indicating that the
local velocity increase in open channel flow is not as great
as the increase experienced in pipeline and flat plate flow.
45
-------
SECTION V
FLUME AND SPILLWAY TESTS
EXPERIMENTAL TEST ARRANGEMENT AND PROCEDURES
Arrangement
A series of tests was conducted in the open channel facility
to determine the effects of a polymer additive on the
following four types of flow measurement devices (Fig. 21):
LeopoldLagco flume
Parshall flume
Sharp-Crested weir
V-notch weir.
The Parshall flume and the sharp-crested weir were adapted
to the smooth-wall rectangular channel and tested at a 1%
slope, while the V-notch weir and the Leopold-Lagco flume
were tested in the smooth-wall trapezoidal channel at a
slope of 2%. The additive used for this testing was
Polyox Coagulant.
'Procedures
The tests were conducted using the following experimental
procedure. The first step was to establish an experimental
calibration curve using tap water for each of above devices.
The curve was obtained by measuring a static head with a
pointer gauge at a specified point in the flow (usually at
the inlet to the device) and recording it along with the
corresponding flow rate. Since the orifice plate was only
accurate for flows greater than 150 gpm, lower flow rates
were determined using a stop watch, a 5-gallon bucket, and
a scale. The time and weight of water required to fill the
bucket were recorded. Once a sufficient number of data
points were obtained, a curve was plotted expressing the
head as a function of the flow rate.
47
-------
00
L Sharp-Crested
Rectangular Weir-
90° V-Notched Weirs
Leopold-Lagco Flume
Channel Section Cyp.
Parshall Plume
Adtptar Section
Cxiitir.9
Ks TEST SETUP
Sh»rp Cr«»ted
Weir
v-Koich weir
FIGURE 21. FLUMES AND SIDE CHANNEL SPILLWAYS
-------
The next step was to introduce polymer into the flow. The
polymer was injected at a rate required to yield a pre-
determined concentration. The same procedure as above was
repeated, and a new calibration curve was developed for
each device at a given polymer concentration.
TEST RESULTS
Results from the flume and spillway tests are presented in
Figures 22, 23, and 24. Comparison of the undosed tap
water data (experimentally determined versus manufacturer
supplied or calculated) shows that for the Leopold-Lagco
flume (Fig. 22), the 90° V-notch weir (Fig. 23a), and the
sharp-crested weir (Fig. 23b), the experimental data lie
below the manufacturer supplied or calculated data. For
the Leopold-Lagco flume, the discrepancy is due to the
fact that the experimental data was plotted for a 2% channel
slope whereas the manufacturer's data was plotted for a
0-1% slope. In the case of the weirs, the discrepancies
occurred because the calculated calibration data did not
take into account the finite stream velocity immediately
upstream of the weirs. As shown in Figure 23a, the experi-
mentally determined data agree well with an adjusted
calibration curve when the upstream velocity is considered.
Figures 22 and 23a show that two of the flow measurement
devices the Leopold-Lagco flume and the sharp-crested
weir are ineffective above a transition flow rate when
polymer is added to the flow. At this flow, the inherent
hydraulic jump created by the flume or spillway is dissi-
pated, as illustrated in Figure 25. This transition flow
rate occurs at approximately 29 gpm for the Leopold-Lagco
flume, and above 90 gpm for the sharp-crested weir.
For the V-notch weir calibration (Fig. 23b), the hydraulic
jump was maintained, but the polymer had an effect on the
calibration at the higher flow rates, resulting in a sub-
sequent deviation from the water calibration curve.
49
-------
01
o
»*
M
(D
s
3 4 5 6 8 10
Flow (gpm)
20
30 40 50 60 80 100
FIGURE 22. LEOPOLDrLAGCO FLUME CALIBRATION DATA
(WITH/WITHOUT POLYMER)
-------
/-
Calibration Curve (Upstream Velocity
Assumed Negligible) -
Adjusted Calibration Curve (Upstream-"";
Velocity Taken into Account) j
is
O
£
o
.2
y-
© - Undosed Tap Water Data
D- Polyox Coagulant Data
(Hydraulic Jump Dissipated)
y- polyox Coagulant Data
(Hydraulic Jump Not Dissipated)
50
100
150
200
250
FIGURE 23a.
Plow (gpm)
SHARP-CRESTED WEIR CALIBRATION DATA
(WITH/WITHOUT POLYMER)
.4
.3
n
o
I
s
a .1
- ' , -
1*
i
l_
Calit
/^
--
ratic
/
y8
n Cur
0 '
ve-^,
0
HI
.. - . -
..:...
*^~
------
^
. .
^
9
T|..c
^^
i ~
~*~
- '
Q- Polyox Coagulant Data
Undosed Tap Water Data
0 - Rectangular Channel
Q - Trapezoidal Channel
20 40
&0 86
100
FIGURE 23b. 90
Flow (gpm)
V-NOTCHED WEIR CALIBRATION DATA
(WITH/WITHOUT POLYMER)
51
-------
1.0
.8
0)
.6
-
/
-
x
<"
C
xf
o
alcul
^
- - '
ated
X
Q
-
-
Calib
s*
ratio
A
- - -
n Cur
^
'
ve-*,
-"'
^
'
- -
^
- -
Q-Undosed Tap Water Data
B-Polyox Coagulant Data (21.5 wppm)
-
-..: ...
- ,
50 100 150 200 250 3
Flow (gpm)
FIGURE 24.
PARSHALL FLUME CALIBRATION DATA
(WITH/WITHOUT POLYMER)
.52
-------
Euro Vatoy
(a)
Chaiuwl Scctioru (Typ.) ^ **^* l*opoia-Lagco
Trapazoidil Flurao-
Leopold-Laqco Flume
Sharp-Crested
Weir
-b-
with
polymer
FIGURE 25. TYPICAL VISUAL EFFECTS FOUND FOR POLYMER
ADDITIVES WITHIN FLUMES AND SPILLWAYS
5?
-------
SECTION VI
DISCUSSION
Although much pipeline information for the polymers tested
is found in the literature, additional data were assimilated,
using a 3/4-inch nominal I.D. pipeline to provide a means
of calibrating the polymer solution concentration used for
the open channel tests. Initially, the experimental data
obtained in the 3/4-inch pipeline were compared to other
data found in the literature, and were found to be compatible.
These data were then used to establish calibration curves
to test each polymer solution used for open channel testing.
In addition, from the calibration tests it was decided to
eliminate Guar Gum from further testing, since relatively
high concentrations of this polymer were found necessary
to produce a minor flow augmentation effect in the calibra-
tion pipeline loop.
The open channel tests were performed primarily to establish
what type of effect,' if any, the addition of polymers would
have on water flow within open channel configurations.
Qualitatively, when polymer was injected into the flow,
three characteristic effects on the flow were noticed. The
first and most dramatic of these was that the level of the
flow decreased at constant flow rate. This effect was
observed ,for both channel configurations tested.
The second characteristic effect necessarily follows from
the first, and was clearly visible. Since the flow rate
before and after polymer addition was maintained at a
constant value, arid since the level of flow decreased upon
polymer addition (decreasing the flow area)/ continuity of
flow was satisfied only if the water velocity increased.
Such a flow velocity increase was evident when the polymer
was first introduced into the water flow. At this time/ a
mini-hydraulic jump was instantly established at the outlet
of the test section. This jump indicated that the upstream
flow (with polymer) was moving more quickly than the down-
stream flow (without polymer). This jump rapidly traversed
downstream, after which the flow level decreased.
55
-------
The third characteristic effect was less noticeable. During
polymer injection, a decrease in surface turbidity and wave
motion was noticed. The flow seemed to approach a tranquil
condition upon polymer addition. This effect was only quali-
tatively noted, and was not investigated thoroughly enough
to establish quantitative results.
The most significant discovery during this investigation
was that Polyox Coagulant, Polyox WSR-301, and Separan AP-30
exhibited a pronounced "flow augmenting" effect on the flow
when injected into both rectangular and trapezoidal open
channel flows. This type of effectiveness represented by
friction reduction, together with flow augmentation, had
been repeatedly demonstrated for completely filled pipeline
flows and partially filled pipeline flows, but had never
been exhibited for true open channel configurations.
Flow performance within conduits is generally expressed
as percent friction reduction (at a ""constant flow rate),
percent flow increase (at a constant energy or static head
level), or a combination of both flow increase and friction
reduction (mainly for gravity flows). In the case of the
open channel test performed for this project, "percent flow
increase" or flow augmentation was used to analyze the test
data. Note also that in this project, flow augmentation
percentage was experimentally determined, rather than
derived from friction reduction data.
Flow increases were achieved for all test cases; however,
Polyox Coagulant caused the. maximum increase. Maximum flow
augmentation is dependent on channel slope,-but concrete
relationships have not been established and this dependency
is felt to be relatively minor. In the case of smooth-
walled channels, it was found that a higher Reynolds number
flow exhibited a lower flow increase capability. This may
indicate either improper mixing of the injected polymer;
that the channels used for testing were not long enough
to permit adequate performance of the polymer additives;
or that, at the high Reynolds number used for these tests,
the specific polymers used degraded rapidly due to the
shear forces. Since the rough-walled channel tests exhibited
a less pronounced difference between flow augmentation for
high and low Reynolds numbers, it is believed that there
existed in the smooth-walled channel tests at high Reynolds
number an improper polymer mixing action.
56
-------
Several polymer injection systems were considered for use
with this specific open channel system. It should be noted
that all of the systems investigated have merit, depending
upon the channel system in question. For larger scale
systems, the one used for these tests is not necessarily
the most adequate.
An injection point location investigation was also conducted.
It was found that injection of polymer at two different
locations along the open channel did not appreciably affect
the flow increase results. However, injecting the polymer
into the 6-inch pipeline approximately 50 feet from the
entrance of the open channels did permit a greater flow
increase capability. This may be due to a more thorough
mixture of the polymer within the pipeline flow before it
is introduced into the open channel.
Theoretically, using the Prandtl-von Karman Universal
Velocity Distribution Law and experimental data obtained
using a pitot-static tube within the open channels, it has
been found that a flow increase shift is noted but is less
than that found for pipeline and flat plate flow at the
same wall shear stress. This is to be expected since, in
general, the polymer effectiveness within open channels is
less than that for pipeline flow.
The flume and side channel data predict that if the inherent
hydraulic jump associated with these devices is dissipated
by the injection of polymer, the devices- should not be used
for measuring the open channel flow. However, they may be
calibrated for polymer flows if the hydraulic jump remains.
Of all the devices tested, the Parshall flume appeared to
be the most accurate.
57
-------
ACKNOWLEDGMENTS
The support of this work by the Environmental Protection Agency
and the interest and contributions of the following men are
acknowledged with appreciation.
Mr. George Kirkpatrick, former Staff Engineer and Messrs. Harry
Torno and Frank Condon, Staff Engineers, Municipal Pollution
Control Section, for their technical guidance and sponsorship.
Mr. Richard Field, Chief, Storm and Combined Sewer Technology
Branch, for his many useful insights.
Mr. James Newsom, Contract Project Officer, for his support,
guidance, and assistance.
59
-------
REFERENCES
1. Cahan, J.I.; Miller, L.V.; Tatum, D.R., "Polymers
Can Relieve Surcharged Sewers," Edward H. Richardson
Associates, Inc., Newark, New Jersey, Reprinted from
The American City, September 1971.
2. Leary, R.D., et al.; "The Use of Water-Soluble Polymers
in Reducing Sanitary Sewer Overflows: A Feasibility
Study," Hercules, Inc., Wilmington, Delaware, Final
Report prepared for Sewerage Commission of the City
of Milwaukee, April 16, 1971.
3. The Western Company, Research Division; "Polymers for
Sewer Flow Control," U.S. Department of the Interior,
FWPCA, Report No. WP-20-22, 1969.
4. Derick, C.T., and Sieracki, L.M., "Test Program to
Determine the Effects of Polymer Additives on Open
Channel Flow," Report No. 115-1, Columbia Research
Corporation, Gaithersburg, Maryland, February 23, 1972.
5. Derick, C.T., and Sieracki, L.M., "Test Program to
Determine the Effects of Polymer Additives on Flumes
and Spillways," Report No. 115-2, Columbia Research
Corporation, Gaithersburg, Maryland, March 23, 1972.
6. Sellin, R.H.J., and Barnard, B.J.S., "Open Channel
Applications for Dilute Polymer Solutions," Journal
of Hydraulic Research, Vol. 8, 1970.
7. Chow, Ven Te, Open Channel Hydraulics, McGraw-Hill,
New York, New York, 1959.
8. Fabula, Andrew G., "Attainable Friction Reduction in
Large, Fast Vessels," Naval Undersea Warfare Center,
Report No. NUWC TP123, p. 8, February 1969.
61
-------
GLOSSARY
Force-Fed System - A flow system using mechanical means
(i.e., a pump) to move the flow through the system.
Gravity-Fed System - A flow system using the force of
gravity to move the flow through the system.
Hydraulic Jump - A fluid flow phenomena indicated by a
sudden rise in the elevation of the liquid surface when a
rapidly flowing liquid stream in an open channel suddenly
changes to a slowly flowing liquid stream with a larger
cross-sectional area.
Injected Polymer Concentration - The concentration of
polymer expressed in weight parts per million (wppm)
within the flow system being investigated.
Percent Solution - The term'used to specify the amount of
polymer by weight in a polymer/water or polymer/water/
solvent solution used for injection into the flow system
being investigated.
63
-------
APPENDIX 1
FLUID FRICTION REDUCING POLYMERS
LITERATURE SURVEY
Theoretical and Experimental Reports (Pipeline and Flat
Plate Flow)
1. Arranaga, A.B., "Friction Reduction. Characteristics
of Colloidal and Fibrous Substances," U.S. Naval
Ordnance Test Station, Memorandum to W.D. White,
26 June 1967.
2. Bilgen, E., "Effect of Dilute Polymer Solutions on
Discharge Coefficients of Fluid Meters," The American
Society of Mechanical Engineers, Paper No. 70-FE-38,
1970.
3. Buechley, T.C., Derby, J.A., Melton, L.L., "A Device
to Measure Frictional Pressure Reduction of Fluids in
Turbulent Pipe Flow," Paper Prepared for British
Society of Rheology and presented at meeting in
Swindon, Wilts., United Kingdom, September 19, 1967.
4. Crowley, P.A., and Witbeck, N.C., "Evaluation of
Friction Reducing Additives for Pipeline Use,"
Technical Report No. 101-F, Columbia Research Corpora-
tion, October 1970.
5. Ellis, A.T., Ting, R.Y., Nadolink, R.H., "Some Storage
and Shear History Effects on Polymeric Friction
Reduction," Journal of Hydronautics, Vol. 6, No. 2,
pp. 66-68, July 1972.
6. Evans, A.P., "Formulation of Highly Concentrated Drag
Reducing Fluids for Naval Ship Applications," NSRDC,
Annapolis Report No. 8-669, February 1972.
7. Fabula, A.G., "Attainable Friction Reduction on Large
Fast Vessels," Naval Undersea Warfare Center, NUWC TP 123,
February 1969.
-------
8. Fabula, A.G., and Burns, T.J., "Dilution in a Turbulent
Boundary Layer with Polymeric Friction Reduction,"
presented to AIAA 2nd Advanced Marine Vehicles and
Propulsion Meeting, May 1969.
9. Felsen, I.M., "Turbulent Flow Drag-Reduction by Dilute
Poly (Ethylene Oxide) Solutions in Capillary Tubes,"
NSRDC Annapolis Report No. 3240, March 1971.
10. Forester, R.H., Rozelle, L.T., Larson, R.E., "Develop-
ment of a Fluid Concentrated Dispersion of a Water-
Soluble Polymer Capable of Reducing the Friction of
Water under Turbulent Conditions," North Star Research
and Development Institute, February 26, 1968.
11. Gadd, G.D., "Turbulence Damping and Drag Reduction
Produced by Certain Additives in Water," Nature,
Vol. 206, pp. 463-467, 1965.
12. Gadd, G.E., "Reduction of Turbulent Friction in Liquids
by Dissolved Additives," Nature, Vol. 212, Nov. 26, 1966.
13. Giles, W.B., "Similarity Laws of Friction-Reduced
Flows," Journal of Hydronautics, Vol. 2, No. 1, 1968,
p. 34.
14. Goodman, Alvin S., et al., "Use of Mathematical Models
in Water Quality Control Studies," Northeastern Univ.,
Boston, Mass. Dept. of Civil Engineering, July 1969.
15. Granville, P.S., "The Frictional Resistance and
Velocity Similarity Laws of Drag-Reducing Dilute
Polymer Solutions," DTMB Hydromechanics Technical
Note No. 61, 1966.
16. Granville, P.S., "The Frictional Resistance and
Velocity Similarity Laws of Drag-Reducing Dilute
Polymer Solutions," NSRDC Report 2502, Sept. 1967;
also Journal of Ship Research, Vol. 12, No. 3,
Sept. 1968.
17. Granville, P.S., "Progress in Prictional Drag Reduction
Summer 1968 to Summer 1969, " NSRDC, CaderocTc, Tech.
Note 143, August 1969.
66
-------
18. Granville, P.S., "Drag Reduction of Flat Plates with
Slot Ejection of Polymer Solution," Naval Ship R&D
Center, Hydromechanics Laboratory Technical Note 140,
July 1969.
19. Howard, R.B., and McCrory, D.M., "The Correlation
between Heat and Momentum Transfer for Solution of
Drag Reducing Agents," NSRDC Annapolis, Report No. 3232,
Jan. 1971.
20. Hoyt, J.W., "Hydrodynamics of Dilute Polymer Solutions
and Suspensions," Naval Undersea Warfare Center,
Draft Copy, January 1969.
21. Hoyt, J.W., and Fabula, A.G., "The Effect of Additives
on Fluid Friction," Fifth Symposium on Naval Hydro-
dynamics, September 10-12, 1964, Bergen, Norway,
ACR-112, Office of Naval Research, Department of the
Navy, Washington, D.C.
22. Johnson, B., and Barchi, R., "The Effect of Drag
Reducing Additives on Boundary Layer Turbulence,"
AlAA 3rd Propulsion Joint Specialist Conference,
Paper No. 67-459, July 1967.
23. Kline, S.J., Reynolds, W.C., Schraub, F.B., Runstadler,
P.W., "The Structure of Turbulent Boundary Layers,"
Journal of Fluid Mechanics, Vol. 30, Part 4, pp. 741-773,
1967.
24. Kuo, Y., and Tanner, R.I., "A Burgers-Type Model of
Turbulent Decay in a Non-Newtonian Fluid," The American
Society of Mechanical Engineers, Paper No. 71-WA/APM-ll,
1971.
25. Lescarboura, J.A., "Rheology Notes, Research Report
No. 105-1-1-1-70," Continental Oil Company, September
1970.
26. Lescarboura, J.A., Culter, J.D., Wahl, H.A., "Drag
Reduction with a Polymeric Additive in Crude Oil
Pipelines," Society of Petroleum Engineers of AIME,
Paper No. SPE 3087, 1970.
67
-------
27. Melton, L.L., and Malone, W.T., "Fluid Mechanics
Research and Engineering Application in Non-Newtonian
Fluid Systems," Society of Petroleum Engineers Journal,
Vol. 4, 56-66, March 1964.
28. Merrill, E.W., "Drag Reduction in Turbulent Liquid
Flow," Summary of Talk, 1 June 1965.
29. Miloh, T., and Poreh, M., "The Resistance to Rotation
of Free and Enclosed Disks," The American Society of
Mechanical Engineers, Paper No. 71-APM-25, 1971.
30. Nash, J.M., "A General Concept for Modeling Analytic
Requirements for Dilute Polymer Drag Reduction
Phenomena," Paper presented to Fuels Handling Equip-
ment Division, Mechanical Technology Department, MERDC,
Fort Belvoir, Virginia, 1972.
31. Nece, Ronald E., "Hydraulic and Mixing Characteristics
of Suction Manifolds," Washington Univ., Seattle,
Charles W. Harris Hydraulics Lab.., TR 26, June 1969.
32. Patel, V.C., "Calibration of the Preston Tube and
Limitations of its Use in Pressure Gradients," Journal
of Fluid Mechanics, Vol. 23, Part 1, September 1965.
33. Pomeroy, Richard D., "Flow Velocities in Small Sewers,"
Journal WPCF, Vol. 39, No. 9, pp. 1525-1547, September .
1967.
34. Pruitt, G.T., Rosen, B., Crawford, H.R., "Effect of
Polymer Coiling on Drag Reduction," The Western Company,
Report No. ETMB-2, August 1966.
35. Ram, A., Finkelstein, E., Elata, C., "Reduction of
Friction in Oil Pipelines by Polymer Additives,"
I and EC Process Design and Development, Vol. 6,
No. 3, pp. 309-313, July-1967.
36. Seyer, F.A., and Metzner, A.B., "The Turbulence
Structure of Viscoelastic Drag Reducing System,"
Unpublished.
68
-------
37. Sieracki, L.M., Smith, S.P., Whitney, R.D., "Operational
Evaluation, of Hydrocarbon-Soluble Friction Reducing
Additives," Columbia Research Corporation Report No.
101-G, October 1971.
38. Spangler, J.G., "Studies of Viscous Drag Reduction
with Polymers Including Turbulence Measurements and
Roughness Effects," LTV Research Center Report No.
0-71000/8R-12, October 1968; also in "Viscous Drag
Reduction," C.S. Wells, Ed., Plenum Press, New York,
1969.
39. Treiber, K.L., Crowley, P.A., Witbeck, N.W., "Evaluation
of Friction Reduction Additives for Pipeline Use,"
Paper presented at the 67th National Meeting of A.I.Ch.E,,
Feb. 15-18, 1970, Atlanta, Georgia.
40. Treiber,'K.L., and Sieracki, L.M., "Test Report - The
Effect of Non-Newtonian Friction Reducing Additives in
a Diesel Fuel Pipeline," Columbia Research Corporation
Report No. 101-2, 14 December 1970.
41. Walters, R.R., and Wells, Jr., C.S., "Effects of
Distributed Injection of Polymer Solutions on Turbulent
Diffusion," Journal of Hydronautics, Vol. 6, No. 2,
pp. 69-76, July 1972.
42. Wells, C.S., "An Analysis of Uniform Injection of a
Drag-Reducing Fluid into a Turbulent Boundary Layer,"
LTV Research Center Report No. 0-71100/8R-14,
November 1968; also in "Viscous Drag Reduction,"
C.S. Wells, Ed., Plenum Press, New York, 1969.
43. White, A.L., "Pilot Plant Evaluation of Polyisobutylene
(Oppanols) Friction Reduction Agent," U.S. Army
Mobility Equipment Research and Development Center,
Report No. 12, 468-7, December 1968.
44. White, A.L., "Friction Reduction Agents," U.S. Army
Mobility Equipment Research and Development Center,
Report No. 12, 468-4, July 1968.
45. Wu, J., "Drag Reduction in External Flows of Additive
Solutions," in "Viscous Drag Reduction," C.S. Wells, Ed.,
Plenum Press, New York, 1969.
69
-------
46. Wu, J., and Tulin, M.P.,- "Drag Reduction t>y Ejecting
Additive Solutions into Pure-Water Boundary Layer,"
The American Society of Mechanical Engineers, Paper
No. 72~Fe-12, 1972.
Theoretical and Experimental Reports (Open Channel Flow)
47. Sellin, R.H.J., and Barnard, B.J.S., "Oppn Channel
Applications for Dilute Polymer Solutions," Journal
of Hydraulic Research, Vol. 8, 1970.
Books
48. Chow, Ven Te, Open Channel Hydraulics, McGraw-Hill
Book Company, Inc., New York, New York, 1959.
49. Daily, J.W., and Harleman, D.R.F., Fluid Dynamics,
Addison-Wesley Publishing Company, Inc., Reading,
Massachusetts, pp. 293-302, 1966.
50. Schlichting, H., "Boundary Layer Theory," McGraw-Hill,
New York, 4th Ed., 1960, p. 542.
Full Scale Reports
51. Cahalan, J.I., Miller, L.V., Tatman, D.R., "Polymers
Can Relieve Surcharged Sewers," Edward H. Richardson
Associates, Inc., Newark, New Jersey, Reprinted from
The American City, September 1971.
52. Field, R., "Management and Control of Combined Sewer
Overflows," Environmental Protection Agency, NERC,
Office of Research and Monitoring, Edison, New Jersey,
May 1972.
53. Hayes, Seay, Mattern, and Mattern Architects-Engineers,
"Engineering Investigation of Sewer Overflow Problem,"
U.S. Department of the Interior, FWQA, Report No.
11024DMS05/70, May 1970.
70
-------
54. Leary, R.D., et al., "The Use of Water-Soluble Polymers
in Reducing Sanitary Sewer Overflow: A Feasibility
Study," Hercules, Inc., Wilmington, Delware, Final
Report prepared for Sewerage Commission of the City of
, Milwaukee, April 16, 1971.
55. Science Information Services Department, Franklin
Institute Research Laboratories, "Selected Urban Storm
Water Runoff Abstracts," Environmental Protection
Agency, Report No. 11024FJE04/71, Third Quarterly
Issue, April 1971.
56. Storm and Combined Sewer Pollution Control Branch,
"Combined Sewer Overflow Seminar Papers," Compilation
of technical papers and discussions presented at a
seminar at Hudson-Delaware B.asins Office, Edison,
New Jersey, November 4-5, 1969.
57. The Western Company, Research Division, "Polymers for
Sewer Flow Control," U.S. Department of the Interior,
FWPCA, Report No. WP-20-22, 1969.
Supplier Bulletins
58. Chemical and Physical Properties; Cellulose Gum,
Hercules, Inc., Specialty Products Division, Wilmington,
Delaware.
59. Powdered Gum Guar Type M, MEER Corporation, North
Bergen, New Jersey.
60. Jaguar, Stein, Hall and Co., Inc., Water Soluble
Polymer Department, New York, N.Y.
61. Liquid Polymers, Nalco Chemical Company, Chicago,
Illinois, Bulletin 21.
62. Separan AP-30, Dow Chemical Company, Moorestown,
New Jersey.
63. Polyox Friction Reducing Agent, FRA, Union Carbide
Corporation, Moorestown, New Jersey, Bulletin No.
F-41350, March 1966.
71
-------
64. Polyox Water-Soluble Resins, Union Carbide Corporation,
New York, New York, Bulletin No. F-40246E, 1968.
72
-------
APPENDIX 2
FORMULAS USED FOR DATA ANALYSIS
THREE-QUARTER-INCH PIPELINE CALIBRATION TESTS
Percent flow augmentation was calculated using the following
formulas:
%(FA) = Q2 " Ql x 100% (1)
Q2
where: Q]_ = initial flow rate without polymer
Q2 = final flow rate with polymer
Polymer concentration was found by using:
?gl
c = -=-^- x C (2)
P mt p
where: C = average bulk polymer injected
concentration
^ = mass density of polymer solution
g^ = volumetric rate of injection
m. = total mass flow rate in pipeline
Cp = initial concentration of the polymer
solution
OPEN CHANNEL TESTS
Polymer concentration was found by using:
mi
C = x C (3)
p mt p
where: C = average bulk polymer injected
concentration
m. = injection mass flow rate
73
-------
fh. = mass flow rate of water in the channel
C = initial solution concentration
P
THEORETICAL ANALYSIS (VELOCITY PROFILES)
The basic Prandtl-von Karman Universal Velocity Distribution
Law can be expressed for open channel flow as follows:
v (9yVf)
v r; »7|- 1 __ Y *-
= 5.75 log £
where: v = local velocity
y = perpendicular distance from channel
bottom
^ = kinematic viscosity
V.r = shear velocity = yg
for open channels
RTT = hydraulic radius
S = channel slope
g = 32.2 = gravitational constant
sec
The above equation is found to change by a constant factor
when polymer is injected into the flow. Thus, it is seen
that Equation (4) becomes:
v c _c , 9yvf + A B
= 5.75 log ~^
where: A B is a function of polymer additive
properties and the wall shear"stress.
74
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
1.- Repf, ft No.
3. Accession No,
J. Title
Flow Augmenting Effects of Additives on Open
Channel Flows
Derick, C.
Logie, K.
COLUMBIA RESEARCH CORPORATION
Clopper Road
Gaithersburg, Maryland 20760
12. Sponsoring Organization
10. Project No.
1102OGQG
11, Contract/Grant No.
68-01-0168
Environmental Protection Agency Report No, EPA-R2-73-238, June 1973
l-j. A osifsd
Two model open channel configurations (trapezoidal and rectangular) and
three water soluble polymers (Polyox Coagulant, Polyox WSR-301, and
Separan AP-30) were used to experimentally determine the effects of
injecting dilute polymer solutions into open channel water flows.
It was found that for all test cases, injection of the three polymer
additives produced flow characteristic changes reflected as either a
water surface level decrease at constant flow rates or a flow rate
increase at constant static heads. These flow characteristic changes
were found to be dependent, in varying degrees, on channel slope, surface
roughness, injection point location, polymer injection method,__flow
Reynolds number, and injected polymer concentration.
In addition, two flumes (Parshall and Leopold-Lagco) and two model side
channel spillways (90 V-notch weir and sharp-crested rectangular weir)
were used to determine experimentally the effects of polymer additives ~on
the flow measuring characteristics of energy dissipators.
17a. Descriptors
*Polymers, *Open Channels, *Flumes, *Fluid Friction, *Water Polution
Control, *Weirs, Spillways, Combined Sewers, Non-Newtonian Flow,
Turbulent Flow, Flow Augmentation, Flow Measurement, Flow Profiles,
Roughness, Slopes, Hydraulic Jump, Reynolds Number, Solutions
17b. Identifiers
*Parshall Flume, *Leopold-Lagco Flume, Injected Concentrations
17f.. COWRR Field & Group 92E 05D 05G 08B
IS. Availability ' 19s Si> sarity *ass> '
; (Report)
20. Security C/ass,
21.
22.
JW>, of
Pages
.Price
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 2O24O
C. T. Der ick
| institution Columbia Research- Corporation
,'SSfC !O2 i'^SV JUNE !97S)
------- |