&EPA
United States
Environmental Protection
Agencv
Industrial Environmental Research EPA-600 280 140
Laboratory June 1980
Cincinnati OH 45268
Research and Development
Development of a
High Current
Oilboom/Skimmer
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution-sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-80-140
June 1980
DEVELOPMENT OF A HIGH CURRENT
OILBOOM/SKIMMER
by
Blair A. Folsom
Ultrasystems, Inc.
Irvine, California 92715
Contract No. 68-03-0403
Project Officer
Stephen J. Dorrler
Oil and Hazardous Material Spills Branch
Industrial Environmental Research Laboratory
Edison, New Jersey 08817
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution con-
trol methods be used. The Industrial Environmental Research Laboratory -
Cincinnati (lERL-Ci) assists in developing and demonstrating new and improved
methodologies that will meet these needs both efficiently and economically.
This report describes the development and testing of a high speed (4 to
6 knot) skimmer for collecting floating oil from calm water. This technique
will be of interest to all those interested in cleaning up oil spills in pro-
tected waters. Further information may be obtained through the Resource
Extraction and Handling Division, Oil and Hazardous Materials Spills Branch,
Edison, New Jersey.
David 6. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
-------
ABSTRACT
A low drag oil spill recovery device has been developed to retain and
recover floating oil slicks in currents up to 3.0 m/sec (6 knots) and low
wave conditions typically encountered in small lakes and harbors, smooth
flowing streams and other inland waterways. The streamlined oil boom/skimmer
utilizes a slightly submerged slotted hydrofoil to skim a thin layer of oil
and water into an attached sump where the kinetic energy of the high speed
flow is dissipated and the oil and water are separated by gravity. The ex-
cess water is discarded and the oil is retained in the sump under conditions
similar to a conventional 16w speed oil boom. The retained oil is recovered
as required by a built-in secondary skimmer. Because of the system's sim-
plicity and low drag, it can be deployed as a high speed skimmer supported
by a catamaran or as a high current boom with several adjacent sections
spanning a high current area.
The development effort focused on bow, sump and stability control system
design and the testing of the complete streamlined oil boom system at
OHMSETT. Thruput efficiencies as high as 99+ percent and 87 percent were
measured under calm surface conditions at 2.0 and 3.0 m/sec (4 and 6 knots)
respectively. Performance degraded in tests with waves'.
This report was submitted in fulfillment of Contract 68-03-0403 by
Ultrasystems, Inc. under the sponsorship of the U.S. Environmental Protection
Agency. This report covers the period March 1, 1974 to March 31, 1980, and
work was completed as of March 28, 1980.
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables .- yii
Abbreviations and Symbols .- * . . . . viii
Acknowledgement > . . . x
1. Introduction 1
2. Conclusions 7
3. Recommendations 9
4. Bow Design 11
5. Sump Design 33
6. Forces and Stability 48
7. System Integration Tests 60
References . . . 80
Appendices
A. Phenomenological Bow Wave Theory 81
B. Detailed Test Procedures - OHMSETT System Integration Tests . 94
C. Oil Handling and Experimental Errors.- OHMSETT Systems
Integration Tests . 99
-------
FIGURES
Number Page
1 Streamlined boom design with rigid sump - Phase I Tests 3
2 Schematic diagram of Phase I oil boom test 4
3 Typical flow over a streamlined shape, shallow submergence 13
4 Small scale models for bow wave tests 16
5 Emperical bow wave correlation verification 17
6 Dimensionless bow wave flow map 18
7 Dimensionless bow wave flow map with data, all solid models.... 20
8 Preliminary slotted model tests 22
9 SI otted models 23
10 Typical flow - NACA 4415 model 25
11 Typical flow - NACA 0015 model 26
12 Correlations for bows of slotted models, NACA 4415 & NACA 0015. 27
13 Comparison of measured and calculated stagnation streamlines... 28
14 Preliminary streamlined oil boom model using NACA 0015 bow 29
15 Preliminary slotted bow design 30
16 Final bow design 32
17 Sump design concept - schematic drawing 34
18 Sump test f 1 ow channel 37
19 Energy dissipation mechanisms 39
20 Final streamlined oil boom/skimmer design... 43
21 Minimum oil droplet size completely removed from final sump.... 45
22 Self-adjusting weir and secondary sump 47
23 Model for Lockheed tests 49
24 Total vertical force vs. difference in free surface levels 51
25 Schematic diagram of streamlined oil boom supported by buoyant
floatation system 52
26 Vertical stability as a function of net vertical force 54
27 Comparison of stability numbers with observed stability-OHMSETT
Tests 1975 56
28 Streamlined oil boom model for System Integration Tests 61
29 Plan view of OHMSETT testing arrangement - 1976 63
30 Streamlined oil boom in system test at OHMSETT 64
31 Oil flow through streamlined oil boom system 68
32 System Tests Results - Calm Surface at 3.0 m/sec (6 knots) 76
33 System Tests with Diversionary Boom - Flow Pattern 77
34 System Tests Results - Calm Water at 2.0 m/sec (4 knots) 79
A-2 Fictitious Bottom Contour For Open Channel Flow Analysis 81
A-3 Theoretical Correlation 63 vs. GB 83
A-4 Theoretical Correlation G-] vs. Go 84
A-5 Three-Dimensional Flow Over Model 86
VI
-------
FIGURES (continued)
Number Page
A-6 Dolphin Nose Model - Correlations 87
A-7 Dolphin Nose Model Data Corrected For Vortices and Data Taken
with SidewalIs 88
A-l Turbulent Bow Wave Flow Model 92
C-l Oil Input - Schematic Diagram 100
C-2 Oi 1 Recovery - Schemati c Di agram 102
C-3 Pear-Shaped Centrifuge Tube, from ASTM Designation D 96-73,API
Standard 2542 - "Standard Method of Test For Water and Sed-
iment in Crude Oils" 103
C-4 Error Associated with Minimum Calibration Divisions On Centri-
fuge Tube ASTM D96-73 104
C-5 Histogram of % Water in Recovered Oil Samples 106
C-6 Error Due to Oil Input Volume Measurement 107
C-7 Histogram of Recovered Oil Volumes 108
C-8 Error Due to Recovered Vol ume Measurement 109
TABLES
Number
1 Test Series Outline 5
2 Energy Dissipation Devices: Important Characteristics and
Resultsof Tests in Sump Test Flow Channel 40
3 Freeboard and Gate Position Adjustments for Optimum Performance 67
4 Test Oil Properties 70
5 Average Wave Heights in OHMSETT'S 1.0 Foot Harbor Chop 71
6 74
C-l Oil Volume Measurement Summary - Errors Due To Minimum Calibra-
tion Divisions ; .110
vn
-------
ABBREVIATIONS AND SYMBOLS
B --freeboard
CD --sump exit slot discharge coefficient
D --sump depth
F --Froude number based on freestream velocity and hydrofoil
character!'c length >
Fj --Local Froude number at Station i
GB --Reciprocal of Froude number squared based on freestream velocity
and freeboard
g --acceleration of gravity
G-, —Reciprocal of Froude number squared based on freestream velocity
and freeboard
GO --Reciprocal of Froude number squared based on freestream velocity
and flow depth over the hydrofoil's highest point.
w1
Gl* "Gl 4
h. --height of channel bottom above arbitrary horizontal datum line
L —characteristic length or effective sump length
Q --flowrate passing over hydrofoil per unit width
QQ --sump flowrate
S --flotation system stiffness
SN --dimensionless stability number
S,, --sump exist slot width
V, --freestream velocity
V. --local velocity at Station i
VD —terminal oil droplet rising velocity in water
K
W --sump width
W, --width of flow passing over hydrofoil measured upstream
hL —hydrofoil width
vm
-------
Y-i --depth of stagnation streamline beneath the free surface
measured far upstream - the thickness of the layer skimmed
Yg —Flow depth over the top of the hydrofoil
Y.. —local flow depth at Station i
YQ —initial position of oil droplets in sump - distance above the
bottom
f —water level difference inside minus outside the sump
y --fluid weight density
-------
ACKNOWLEDGMENTS
The author wishes to thank Messrs. J. S. Dorrler and F. 0. Freestone
of the United States Environmental Protection Agency for their many valuable
suggestions and assistance during this program. Thanks are also due to
Dr. D. C. Wooten of Olson Laboratories and Messrs. G. C. Carver and C.
Johnson of Ultrasystems, Inc. for their outstanding work on several aspects
of the project.
-------
SECTION 1
INTRODUCTION
Although the current generation of oil spill control equipment can clean
up oil spills over a range of environmental conditions, it is inadequate to
cope with spills in currents greater than 0.5 to 1.0 m/sec (1.0 to 2.0 knots).
The objective of this program was to develop an oil spill control device
capable of retaining and recovering floating oil slicks in currents up to
3.0 m/sec (6 knots) and wave conditions typical of small bays, harbors and
inland waterways with free surface excursions less than about 15 cm (6 inches)
trough to crest. The device was to be deployable as either a boom retaining
oil against a high current or as a high speed skimming craft recovering oil
as it was encountered.
The streamlined oil boom concept investigated here uses a low drag
streamlined hydrofoil to skim a thin layer of floating oil and water into an
attached sump at high speed. In the sump the oil and water are separated
under controlled conditions and the oil is retained against a barrier similar
to a conventional oil boom. A built-in skimmer is operated as required to
recover the retained oil.
This concept has several advantages which make it ideal for this appli-
cation. The overall concept is quite simple and is amenable to fabrication
as a boom spanning wide high current areas. Except for flotation/stability
control and oil transfer systems, no moving parts are required and since the
device is essentially two-dimensional, small easy to handle modules may be
constructed separately and coupled to form a boom of arbitrarily long length.
The streamlined design also reduces hydrodynamic drag to approximately 25
percent of a conventional oil boom's or other bluff body's drag significantly
reducing structural requirements. But perhaps the most important advantage
is that the concept has been successfully developed to retain and recover a
high percentage of oil slicks in currents up to 3.0 m/sec'(6 knots).
The streamlined oil boom development was performed in two phases. Phase
I was a preliminary design and feasibility test of the streamlined oil boom
concept and identified problem areas requiring additional development effort.
Phase II focused on~solving these problems through a series of developmental
tests and integrating the subsystems into an operational preprototype system
for testing at OHMSETT.
The initial boom design investigated in Phase I utilized an airfoil
shaped leading edge section moving slightly beneath the water's surface
1
-------
skimming oil and water into a flexible sump. Early tests indicated that
although the flow over the hydrofoil appeared promising, satisfactory control
of the flexible s,ump shape could not be achieved.
Figure 1 shows an improved design where the sump was formed by the in-
terior of the hydrofoil contours. This model was constructed and tested at
the Lockheed Towing Basin in San Diego, California under a variety of wave
and current conditions. The basic hydrofoil shape is an inverted NACA 4424
airfoil* with a 1.5 m (5 feet) chord and an adjustable slot in the ventral
side. This slot, located in a low pressure region, bleeds excess water from
the sump. End plates were attached to simulate two-dimensional flow and
buoyancy was provided through a trailing cylindrical float and a small
totally submerged control hydrofoil beneath the 1/4 chord point. The model
was mounted to a force balance as shown in Figure 2 and tow tested at speeds
up to 2.0 m/sec (4 knots).
The results of these tests indicated that the strealined oil boom con-
cept was feasible for controlling oil spills in currents greater than 1 m/sec
(2 knots). Thruput efficiencies**in excess of 65 to 75 percent were measur-
ed at 1.5 m/sec (3 knots) with drag coefficients substantially smaller than
bluff bodies. Three design problems associated with further development
were also identified:
• Bow Design - The achievement of a fundamental understanding of the
oil-water flow in the vicinity of the bow wave and the design of
the proper bow shape to give efficient operation,under a wide
variety of conditions.
• Sump Design - The development of a functional oil seperation and
pumping system for handling the flow into and out of the sump.
9 Stability Control System Design - Using flotation and hydrodynamic
control surfaces to design a system with the proper response in
comferined tvave and current Conditions.
The Phase II development solved these problems .through a balanced pro-
gram of analysis to understand the important flow phenomena and iterative
testing to optimize design. Table 1 describes the eight major test series
performed in both phases. Following the Phase I tests at the Lockheed
Towing Basin, bow wave tests were performed at small scale using open chann-
el flow facilities. A phenomenological theory was developed as a parallel
effort to correlate the results of the tests and identify scaling paramet-
ers.
* NACA 4424 has a 4 percent camber with the maximum 40 percent aft of the
leading edge and a maximum thickness of 24 percent of the chord length.
See Summary of Air Foil Data by Abbott, Doenhoff and Stives, National
Advisory Committee for Aeronautics Report #825. 1945.
** Thruput efficiency in the ratio of oil volume collected to oil volume
encountered as a floating oil slick.
-------
Float
End Plate
Oil and Water
Oil Slick
Airfoil-Shaped;
Streamlined Boom
Control Hydrofoil
Figure 1. Streamlined boom design with rigid sump as an
integral part of streamlined shape - Phase I Tests
-------
.Force Measurement
Instrumentation
Camera 1
Towing Carriages
Oil Drum
Viewing Windows
Viewing
Window
Camera 3
Camera 2
Figure 2. Schematic diagram of Phase I oil boom test
-------
TABLE 1. TEST SERIES OUTLINE
Test Series
Tow tests at Lockheed Towing
Basin, San Diego, California
5/72-11/72
Bow wave tests in small flow
channels at University of
California at Irvine and
California State University
at Fullerton 4/74-5/75
Flow channel tests in U.S.
Coast Guard recirculating
flow channel Richardson,
Texas 5/75
Tow tests at Lockheed Towing
Basin, San Diego, California
8/75
Tow tests at OHMSETT** (first
test series) 10/75-11/75
Sump tests in specially
constructed flow channel at
Ultrasystems , Irvine,
California 10/75-7/76
Catamaran supported bay tow
tests in Newport Harbor,
Newport Beach, California
7/76
Tow tests at OHMSETT**
(second test series) 9/76
Scale
(Vertical and
flow Direct-
ion)
Full
1/3
Full
Full
Full
Full
1/3
Full
Model
Width
1.2 m
(4 ft.)
10 cm
(4 in.)
1.2 m
(4 ft.)
0.61 m
(2 ft.)
1.52 m
(5 ft.)
8.9 cm
(3.5 in.)
0.51 m
(20 in.)
1.52 m
(5 ft.)
Goal/Result
Measured forces, verified sys-
tem feasibility in tests with
oil and identified further
development needs.
Developed bow and preliminary
sump designs: hydrodynamic
tests, no oil
Refined bow and sump designs,
testing with simulated* and
real oil at up to 2.5 knots
Measured forces and refined
bow and sump design in tests
with simulated* oil at up to
6 knots
System integration, sump design
refinement, testing with oil at
3 and 6 knots, instability
identified
Refined sump design in 6-knot
tests with simulated * and
real oil
Measured flotation requirements
and developed stability control
system; hydrodynamic tests, no
oil
System integration and testing
at 4 and 6 knots
* Simulated oil is floating granular material: polyethelene beads approx. diameter = 0.30
cm (0.125 in.) specific gravity = 0.96.
** OHMSETT is the abbreviation for the U.S. Environmental Protection Agency's Oil and
Hazardous Materials Simulated Environmental Test Tank located in Leonardo, N.J.
-------
Based on the results of these small scale tests, a full-scale 1.2 m
(4 feet) wide model was constructed using the best bow design and a prelim-
inary sump design. Tests in a Coast Guard flow channel in Richardson, Texas
were used to refine bow and sump designs in tests with real and simulated oil
(seenote for Table 1). An improved model was then constructed and tested
at the Lockheed Towing Basin. The model was mounted to a force balance to
determine lift, drag,moments and performance under calm water and period-
ic wave conditions was measured at tow speeds up to 3.0 m/sec (6 knots).
The results of all these tests were then combined into a full-scale 1.5 m
(5 feet) wide model for testing of OHMSETT. The model was attached to a
10.3 m (34 feet) catamaran and tow tested at 1.5 and 3.0 m/sec (3 and 6
knots). Thruput efficiencies as high as 86.5 percent at 1.5 m/sec (3 knots)
and 72.0 percent at 3.0 m/sec (6 knots) were achieved under optimum condit-
ions. .These tests also identified a hydrostatic instability due to the int-
eraction of the weight of water contained in the sump and the catamaran
flotation. For testing purposes, this instability was counteracted by
continuously manually adjusting the flowrate out of the sump. However, this
manual method only achieved a semblance of stability with larae oscillations
in sump depth, freeboard and other operational parmafneters. A stability
control system was clearly required.
The less than perfect thruput efficiency at 3.0 m/sec (6 knots) prompted
a series of full-scale sump development tests. Since the strealined oil
boom is essentially a two-dimentional device, a full-scale (except for width)
sump model was constructed to test several sump design options. The sump
width was chosen as 8.9 cm (3.5 inches), the width of a "2X4" to produce a
channel with flow requirements within the capacity of available pumps.
The stability control problem was attacked both analytically and exper-
imentally. A stability analysis using nonlinear control system analytical
techniques was used to select a promising system which was optimized
through 1/3-scale tow tests in Newport Harbor.
The results of subsystem development in each of these areas were combin-
ed into a full-scale 1.5 m (5 feet) wide streamlined oil boom section for
a final three week test series at OHMSETT. The purpose of these tests was
to determine the overall system performance in retaining and recovering oil
slicks at speeds up to 3.0 m/sec (6 knots) and \yi.th,wave conditions typical
of inland waterways. Overall the system performed well at both 2.0 and 3.0
(4 and 6 knots) testing speeds under .calm water conditions with thruput
efficiencies as high as 99+ and 87 percent at 2.0- m/sec (4 knots) and 3.0
m/sec (6 knots) respectively. In wave tests using a 0.30 m (1.0 foot) harbor
chop, performance degraded significantly. This was expected since in this
wave condition 10 percent of the .waves are higher than Q.-30, m (12 inches)
trough to crest and the streamlined oil boom was designed to accommodate
only 0.15 m (6 inches) 'free surface variation.
This report describes the development and design of the bow, sump and
stability control systems. The details of the final design and testing at
OHMSETT are also discussed.
-------
SECTION 2
CONCLUSIONS
(1) The streamlined oil boom/skimmer was developed as an essent-
ially two-dimentional device so that sections of arbitrary width could be
close coupled to span a wide area.
(2) The final bow design was capable of operating at up to 3.0
m/sec (6 knots) skimming a 3.8 cm (1.5 inch) layer of oil and water into the
sump in a smooth unbroken layer.
(3) The final sump design dissipated 99 percent of the incident
flow's kinetic energy utilizing any of four energy dissipation mechanisms
without breaking the oil slick into small droplets.
(4) An automatic stability control system was developed to adjust
flow out of the sump to equal flow into the sump and reduce flotation
requirements by an order of magnitude.
(5) The OHMSETT tests measured the performance of the device
deployed as a high speed skimming craft with a 1.5 m (5 feet) wide section
suspended between the hulls of a catamaran work platform.
(6) Results of oil slick recovery tests at OHMSETT indicated
that the device is a feasible means of retaining and recovering oil slicks
under calm surface conditions. Thruput efficiencies in the range of 77 to
99+ percent were measured at 2.0 m/sec (4 knots) and 53 to 87 percent were
measured at 3.0 m/sec (6 knots) respectively over a range of slick thick-
nesses and with a 150 centistoke lube oil and a simulated number two fuel
oil.
(7) In similar tests with a one foot harbor chop wave condition,
thruput efficiencies degraded to the range of 0 to 21 percent. This wave
condition with a 30 cm (1.0 foot) average one-tenth highest wave height was
a severe test and the test results indicate that the device as developed is
only effective under calm surface conditions such as those found in small
harbors, lakes, smooth flowing streams and other inland waterways.
-------
(8) The streamlined oil boom was also tested with a conventional
oil boom deployed at slight angles to the tow direction to divert a 3.6 m
(12 feet) wide oil slick into the 1.5 m (5 feet) working section at 3.0 m/sec
(6 knots). Although the boom successfully diverted the slick into the work-
ing section, turbulent bow waves were produced entraining oil beneath the
surface and lowering thruput efficiencies to the range of 18 to 41 percent.
(9) Force measurements confirmed that the streamlined design
lowers drag. The drag coefficient based on frontal area was calculated
to be 0.39 at 3.0 m/sec (6 knots). This compares to a drag coefficient for
conventional oil boom of approximately 1.50.
-------
SECTION 3
RECOMMENDATIONS
Based on the excellent high-speed calm surface results obtained in the
OHMSETT tests it is recommended that the streamlined oil boom be further
developed as a practical deployable system for high current inland applica-
tions. Both deployment options, high speed skimming craft and high curr-
ent boom are feasible and should be pursued. However, less effort will be
required for the high speed skimming craft application.
The following are specific recommendations:
Deployment as a High Speed Skimmer
(1) The 1.5 m (5 feet) wide section tested at OHMSETT combined with
the supporting work platform are essentially a high-speed skimming craft.
However, the work platform is several times larger than actually required
to support the streamlined oil boom section and is unsuitable for transport
to an actual spill site. A flotation/work platform should be designed and
constructed specifically for the streamlined oil boom so that tests on real
world spills may be conducted.
(2) The streamlined oil boom section tested at OHMSETT was designed
and constructed as a research model and is not suitable for long-term service
in the marine environment. Alternative materials and methods of construction
should be investigated to produce a light weight, rigid and corrosion re-
sistant structure.
(3) The overall operation of a high speed skimmer system should also
be investigated. An effective oil recovery pumping and storage system
should be designed to interface with the basic streamlined oil boom unit.
The problem of debris should be addressed and other aspects such as cleaning,
deployment and transportation should also be considered.
Deployment as a High Current Boom
(1) If the device is to be deployed as a boom, each section must be
equipped with its own integral flotation system which maintains the bow
at the proper vertical position to skim a thin layer of oil and water into
the sump. Ideally, the flotation system would be self-adjusting so that
sections spanning a variable current stream could each operate at design
point.
-------
(2) The problem of oil slick diversion should also be addressed. If
the sections are close coupled, the integral flotation systems must divert
oil into the active sections. An alternate method of deployment would
utilize streamlined oil boom sections a substantial distance apart with a
diversionary device or boom directing the oil slick into the active section.
The poor performance demonstrated with diversionary boom at OHMSETT points
to the need for additional work in this area.
(3) Coupling and deploying the sections in a high current environment
are key problems. The section widths should be optimized for good perform-
ance in recovering oil slicks and for easy handling and deployment. Coup-
lings should allow the sections to operate somewhat independently and yet
withstand drag forces. In addition they must be easily field assembled and
disassembled.
(4) An oil recovery system for combining the recovered oil from each
section must also be developed. One possiblity would involve inducing a
cross flow in the sump to sweep the collected oil to one end of the stream-
lined boom for recovery over a weir or other device. Another possibility
would maifold the outputs of each section into a common pumping section.
10
-------
SECTION 4
BOW DESIGN
The primary function of the streamlined oil boom bow is to direct a
layer of oil and water into the sump for further processing. Ideally the
bow shape should skim a thin layer into the sump at high speed with little
or no turbulence so that all of the oil approaching the boom is smoothly
swept into the sump.
The Phase I tests at the Lockheed Towing Basin demonstrated that stream-
lined hydrofoil shapes could skim most of the oil into the sump and operate
with low drag but did not develop a definitive relationship between hydro-
foil shape, tow velocity, freeboard and bow wave turbulence characteristics.
Consequently, the Phase II bow design work focused on determining the
following data:
• The important bow wave scaling parameters
• The relationship between freeboard and flow over
the hydrofoil
• The parameters causing transition from turbulent
to smooth bow waves
• The bow shape producing optimum flow characteristics
for the streamlined oil boom
A reveiw of hydrofoil literature revealed that while substantial work
had been done on deeply submerged high speed and planing applications, little
was known about the "in between" area where velocities and submergence are
small and the bow wave is near transition from turbulent to smooth.
The bow wave on a slightly submerged hydrofoil is a complex phenomenon.
The flow conditions are determined by a balance of inertia!, gravity, viscous
and surface tension forces and are too complicated to calculate exactly
particularly where the bow wave is turbulent or breaking. In order to
obtain design data and a basic understanding of bow wave flow, a simplified
phenonenological theory was developed. The results of this theory identified
scaling parameters and predicted flowrates passing over a hydrofoil as funct-
ions of operating conditions. The transition from turbulent to smooth bow
waves and the influence of specific hydrofoil contours could not be deter-
mined from the theory so experiments were conducted to determine these fact-
11
-------
ors and develop an optimum streamlined oil boom bow.
BOW WAVE HYDRODYNAMICS AND PHENOMENOLOGICAL CORRELATION
As a hydrofoil is lowered into a swiftly moving current, the following
flow regimes occur as shown in Figure 3. At very high freeboard,* no flow
passes over the hydrofoil and a turbulent bow wave forms. This bow wave is
characterized by a turbulent recirculating roller much like a breaking wave
or a hydraulic jump. The stagnation streamline** is identical to the free
surface upstream, intersects the hydrofoil near the bow and is identical to
the hydrofoil's ventral contour downstream. As the hydrofoil is lowered
flow begins to pass over top and the stagnation streamline descends beneath
the free surface upstream and switches to the hydrofoil's dorsal side
downstream. The bow wave is turbulent as before. As the hydrofoil is low-
ered further, the amount of flow passing over the hydrofoil increases and
eventually the turbulent bow wave is swept over the hydrofoil leaving a
smooth bow wave without surface turbulence, air entrainment or recirculation.
Further lowering the hydrofoil increases the flowrate passing over the hydro-
foil and reduces the disturbance to the free surface until at very deep sub-
mergence there is no noticeable free surface disturbance.
Ideally the streamlined oil boom bow should skim a thin layer of fluid
into the sump with a smooth bow wave at all speeds up to 3.0 m/sec (6 knots),
This type of operation can be described hydrodynamically as low speed and
near the free surface. In terms relative to the size of the hydrofoil, low
speeds means that the Froude number is small. The Froude number is a dim-
ensionless parameter expressing the square root of the ratio of inertial
to gravity forces.
F- V
where:
F = Froude number
V]= Free stream velocity
g = Acceleration of gravity
L = A characteristic length
* Freeboard is the hieght of the hydrofoil's uppermost point above the
undisturbed free surface.
** Stagnation streamline is that streamline which divides flow over from
flow under the hydrofoil. Fluid along this streamline is stagnated
(brought to rest) at the hydrofoil bow.
12
-------
NO FLOW OVER HYDROFOIL
TURBULENT ROLLER
FREE SURFACE
(TYPICAL) —
STAGNATION STREAMLINE
ON VENTRAL SIDE
TURBULENT BOW WAVE
TURBULENT ROLLER
STAGNATION STREAMLINE
ON DORSAL SIDE
SMOOTH BOW WAVE
SMOOTH BOW WAVE
DEEPER SUBMERGENCE
REDUCED FREE
SURFACE DISTURBANCE
STAGNATION STREAMLINE
(TYPICAL)
Figure 3. Typical flow over a streamlined shape
at shallow submergence.
13
-------
The characteristic length used here is the maximum thickness of the
hydrofoil and low speed operation means that the Froude number is small,
less than approximately 2.0. Following these definitions, a hydrofoil 30 cm
(1.0 foot) thick would operate at low speed up to 3.5 m/sec (6.7 knots).
In this low speed range both gravity and inertial forces are important but
the pressure effects due to curvature of the streamlines are small.
Operation near the free surface means that the thickness of the water
layer passing over they hydrofoil is the same order of magnitude as the
thickness of the hydrofoil itself. Thus the flow above the hydrofoil is
effectively separated from the flow beneath and the two flows behave inde-
pendently.
A phenomenological theory was developed to identify scaling parameters,
predict the flowrate over the hydrofoil and correlate experimental results
for hydrofoils operating at low Froude numbers and near the free surface.
The analysis applied the classical open channel flow equations to the flow
passing over a friction!ess impermeable,contour consisting of the stagnation
streamline upstream of the hydrofoil and the dorsal surface of the hydrofoil,.
The details of the analysis and results are discussed in Appendix A.
The analysis identified the reciprocal of the Froude number squared
based on freestream velocity (V-|) and free board (B):
r - gB
« - o—
as the important independent scaling parameter and expressed the resulting
flow conditions in similar form:
•..-31
G3 =
where Y-i and Y3 are the flow depths to the stagnation streamline measured
far upstream and over the hydrofoil's highest point respectively. Based on
this analysis the bow wave characteristics are independent'of the hydrofoil
size and shape and the flowrate over the hydrofoil may be calculated given
the freeboard and velocity.
Experiments were conducted over a range of hydrofoil shapes and sizes
to obtain a measure of the correlation's validity. Initially small open
channel flow facilities of the type normally used to model flow in rivers
14
-------
and streams were used to obtain small scale data on hydrofoils ranging in
thickness from 1.0 to 6.0 cm (0.4 to 2.4 inches) and later larger scale
experiments extended the range to 15 cm (6 inch) thickness.
The majority of the small scale experiments were conducted in an open
flow channel at the California State University at Fullerton. This channel
has a maximum flowrare of 12.5 I/sec (200 gallons/minute), a width of 11 cm
(4.5 inches) and could be operated with subscritical flow at flow velocities
up to 80 cm/sec (2.6 feet/second).* Velocity profiles measured with a pi tot
probe and micromanometer showed uniformity within 5 percent up to within
2.5 cm (1 inch) of the side walls. Tests were conducted by suspending two-
dimensional hydrofoil models in the channel spanning the entire width and
measuring the flow parameters. The free surface locus was measured with a
precision height gauge and the position of the stagnation streamline was
observed with dye injection.
Figure 4 shows the small scale hydrofoil shapes tested. The circular
cylinders of various size were used to measure scale effects and the other
models were used to determine shape effects. The large-scale hydrofoil was
a NACA 4424 shape not shown in Figure 4.
The experimental results are shown in Figure 5, a plot of the flow depth
passing over the highest point on the hydrofoil against freeboard normal-
ized by the acceleration of gravity and freestream velocity expressed in
Froude number form. The theoretical correlations developed from the bow
wave theory (See Appendix A) are also shown. It is remarkable that the
data and theory correlate so well particularly considering the range of sizes
and shapes tested.
The transition from a turbulent to a smooth bow wave as a function of
operating conditions can be described on a flow map. Figure 6 is the?flow-
map where Gg is plotted against the hydrofoil Froude number squared F basec
on the hydrofoil thickness T. Data points representing transition from one
type of bow wave to another can be plotted on this flow map to give a visual
indication of hydrofoil performance. The upper most line represents the
maximum freeboard for which flow passes over the hydrofoil. Along this line,
all of the approaching surface layer's kinetic energy is converted into pot-
ential energy to lift the fluid to the hydrofoil's top. Slightly lowering
the hydrofoil causes flow to proceed with a turbulent bow wave. For Froude
number squared greater than 2.0, the maximum freeboard is limited by the
hydrofoil thickness since the flow has more than enough kinetic energy to
pass over the hydrofoil at any submergence.
At higher velocities the channel Froude number approached 1.0 and the
surface became wavy and unsuitable for hydrofoil experiments.
15
-------
O
1.0cm. DIA.
CYLINDER
o
3.0cm. DIA.
CYLINDER
6.0 cm . DIA.
CYLINDER
DOLPHIN NOSE MODEL
3.0cm. THICK
HEMISPHERICAL NOSE MODEL
3.0cm. THICK
1/4 ROUND MODEL
(QUALITATIVE TESTS ONLY)
STREAMLINED MODEL
2.0 cm. THICK
BLUNT WEDGE LEADING EDGE
3.0 cm. THICK
SHARP LEADING
EDGE THIN MODEL
0.16cm. THICK
SHARP WEDGE LEADING EDGE
EDGE—3.0 cm. THICK
Figure 4. Small scale models for bow wave tests
16
-------
DATA FROM SEVERAL HYDROFOILS
Figure 5. Emperical bow wave correlation verification.
-------
00
1.0
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
-1.8
G0 =-9*
V,
Maximum Freeboard
for Non-Zero Flow
GB=0.5.
B = T
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
V
,2 vi
F --
Turbulent Hydraulic Jump Type
Bow Wave Possible in This Range
Figure 6. Dimensionless bow wave flow map.
-------
The phenomenological theory identified the range -1.0 <% <+0.07 (also
shown in Figure 6) as the range where a turbulent hydraulic jump type bow
wave is possible. The transition line from smooth to turbulent bow waves
should occur somehwere in this range.
Figure 7 is a flowmap showing a compilation of data from all solid
hydrofoils tested. The maximum freeboard for non-zero flow data follows the
theoretically predicted correlation well except for very low values of
Froude number squared. These data points correspond to low velocities where
the dynamic head is of the same order of magnitude as the miniscus height
and surface tension effects are important.
The transition from a turbulent to a smooth bow wave is gradual and was
estimated by observing the flow conditions at which paper chips deposited
upon the surface upstream slow nearly to rest when passing over the hydro-
foil. Dye injection was also used to qualitatively observe the bow wave
turbulence. The apparent data scatter in Figure 7 is partially due to diff-
erences in the performance of various hydrofoils. In the range 0.5< F^< 4
where the streamlined oil boom bow operates, transition occurs in the range
-0.8< GB<0.0.
Data from the flowmap combined with the phenomenological correlation
allow calculation of the minimum flowrate for smooth flow over a hydrofoil.
Minimum flowrate corresponds to maximum Gg and from Figure 7, Gg = 0 appears
to be a conservative value. Applying Gg =0 to the phenomenological correl-
ation equations for smooth flow (Appendix A) gives in the following equation
for minimum flowrate:
Q = _
where Q is the minimum flowrate per unit width. The minimum flowrate is
independent of hydrofoil size and increases as the cube of the velocity
reaching 550 I/sec per meter of hydrofoil width (2,700 gpm per foot of bow
width) at 6 knots.
Similarly, the minimum layer skimmed can be calculated to be:
V
V.2
and at 3.0 m/sec (6 knots) this corresponds to 18.2 cm (7.2 inches).
19
-------
= Ji.
1.0
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
-1.0
-1.2
-1.4
-1.6
-1.8
• o
OB
e
Non-Zero Flow
Max Freeboard
Data Point
Theoretical
Non-Zero Flow
Max Freeboard
« e o
e e e
1.0
i i
-t—• h
e
o
-4 1-
ft
-s—-^ -^
2.0
ro
o
0.2 0.4 0.6s
° o0 e
© o
* 9° 6
e e o
o 0
^_ - o e
0.8
e o
2 K4 T6 T8 27 2T2 2T4
ffi
28 30
e a
Maximum Freeboard
for Smooth Flow
Figure 7. Dimensionless bow wave flow map with data
from all solid models tested.
IT
-------
SLOTTED BOW DESIGNS
The empirical correlation verification in the last section demonstrates
that the minimum flowrate for a smooth bow wave on streamlined hydrofoils
increased as the cube of the velocity reaching quite high values at high
speeds and the thickness of the layer skimmed into the sump increases as the
square of the velocity. Since the thickness of the oil slicks to be recover-
ed by the streamlined oil boom are in the range of 1 to 3 mm, the smooth
flow requirement means that the vast majority of the flow into the sump will
be water and as a result the sump must function as a high flowrate oil/water
separator recovering dilute oil in water mixtures. A reduction of the flow-
rate into the sump is clearly desirable.
Flow into the sump can be reduced with slotted hydrofoils. The hydro-
foil is operated deep enough to achieve a smooth bow wave and some of the
flow is drawn from the hydrofoil's dorsal side to the ventral side through
a slot leaving a much smaller flowrate to enter the sump. The slot width,
location and contours play important roles in determining the amount of flow
sucked through the slot and the characteristics of the secondary bow wave
on the rear edge of the slot. For a very narrow slot width, the slot acts
as a channel with flow drawn through primarily due to the pressure difference
between the dorsal and ventral sides of the hydrofoil. For a very wide slot
width the front and rear sections operate independently each with its own bow
wave. For slot widths between these extremes flow is a complex combination
of the two phenomena.
Before any special models were constructed, an existing model was modi-
fied to allow preliminary testing of the slotted model concept. Figure 8 is
a scale drawing of this model showing one approximate operating condition and
the two stagnation streamlines. The lower one, ending on the bow, separates
flow over from flow under the bow. The upperstagnation streamline, ending at
the rear edge of the slot separates flow through the slot from flow over the
rear of the model. The drawing shows the flow conditions observed when the
model was near the surface. The sharp slot edges caused ventilation (air
entrainment) greatly reducing the effective slot open area. A separated re-
gion on the lower side of the model was also produced. When the slot was
closed by tape, the free surface was only slightly changed (as shown), but
the separated region was greatly reduced. Thus the flow through the slot
was responsible for the majority of the flow separation. As the model was
lowered further beneath the free surface, the flow through the slot diminish-
ed and the two stagnation streamlines merged. This behavior was expected
since the model was symmetric.
These brief tests confirmed that flow over the bow could be reduced
with a slotted hydrofoil and demonstrated that the slot should be carefully
designed to eliminate sharp edges. This would increase the effective slot
area and flow through the slot and reduce separation under the model. A
cambered profile is another desirable modification to provide an increased
pressure differential across the slot.
The two new slotted hydrofoils shown in Figure 9 were constructed
incorporating these improvements. Both were based on standard NACA profiles
with slots added to divert a portion of the flow above the hydrofoil to the
21
-------
Free Surface
With Slot Open
Upper Stagnation
Streamline
Center Section
(Rectangular)
Free Surface
With Slot Closed
Rear Section
(Triangular)
ro
ro
Lower Stagnation
Streamline
Front Section
(Cylindrical)
Separated Flow
With Slot Open
Note: Approximate only
Velocity - 90 on/sec
(Scales to 5 ft/sec on
3 in. thick hydrofoil)
Continuation of
Lower Stagnation
Streamline
Figure 8. Preliminary slotted model tests
-------
Full Scale
Chord • 13.3 cm
Thickness » 2.0 cm
NACA 4415 Profile Model With 1 Slot
(Slot Width is Adjustable)
CO
NACA 0015 Profile Model With 2 Slots
(Slot Widths are Adjustable)
Figure 9. Slotted models.
-------
ventral side. The NACA 4415 model was operated with the cambered side down
and the slot was located with the exit end in the region of lowest pressure.
The slot could be adjusted from closed to more than 5 cm wide by sliding the
two sections. The NACA 0015 model was noncambered but was equipped with two
adjustable slots with more streamlined contours.
Figure 10 shows typical flows observed with the NACA 4415 model at
slots widths of 1.0 and 3.0 cm (at zero angle of attack). These experiments
were conducted at nearly the same velocity and freeboard and demonstrate the
effects of variable slot width. = When the slot width was reduced the amount
of flow through the slot was also reduced and the upper stagnation stream-
line approached and, at zero slot opening, merged with the lower stagnation
streamline. Figure 10 shows that, although the flowrates over the bow were
approximately the same for each slot width, a larger portion of the flow
passed through the wider slot. The regions of separated flow were caused
by the small radius of curvature of the slot walls and viscous effects.
Typical flow over the NACA 0015 model is shown in Figure 11. Since
there are three model sections, there are three stagnation streamlines. In
these experiments the velocity, freeboard and sum of the slot widths were
maintained constant and the widths of the front and rear slots were varied.
Although the relative magnitudes of the flowrates through the slots varied,
the total was essentially constant leaving nearly the same flowrate over
the rear portion. However, when the front slot was closed, a turbulent bow
wave formed.
Although the phenomenological theory developed for solid bodies as no
provision for slot effects, the data for the front portion of the model can
still be expressed on the same coordinates. Figure 12 shows the data for
the experiments on the NACA 4415 and NACA 0015 slotted models. Both the
depth of the stagnation streamline data and the depth of the flow over the
model data follow the correlation lines closely.
Although the phenonenological theory cannot predict the flowrates
through the slots, it can be extended to predict the locus of the stagnation
streamlines on a point by point basis from measured free surface data.
Figure 13 shows the results of applying this "no energy loss analysis" to
calculate the stagnation streamlines on the NACA 0015 model in one exper-
iment. The agreement between measured and calculated streamlines is excell-
ent except in the vicinity of the bow of each section where the analysis
assumptions are invalid because of pressure effects.
Based on the favorable results obtained with the NACA 0015 model, the
front two sections of the model were incorporated into a complete streamlined
oil boom model including a preliminary sump as shown in Figure 14. This de-
sign also performed well but produced a turbulent bow wave on the rear edge
of the slot under some conditions.
The next bow design iteration is shown in Figure 15. Since the slot
width used in previous designs was the same order of magnitude as the hydro-
foil thickness, the forward portion was designed as a separate NACA 4424
24
-------
en
Free Surface
Upper Stagnation
Streamline
Lower Stagnation
Streamline
Slot Width - 1.0 cm.
Slot Width • 3.0 cm.
Figure 10. Typical flow - NACA 4415 model.
(Same velocity, freeboard, and angle of attack - variable slot width)
-------
ro
CTl
Turbulent
Bow Wave
Smooth
Bow Wave
Separated Flow
Free Surface
Smooth
Bow Wave
Stagnation Streamline
on Center Portion
Stagnation Streamline
on Rear Portion
Stagnation Streamline
on Front Portion
Figure 11. Typical flow - NACA 0015 model.
(same Velocity, Freeboard and angle of attack - variable slotwidths)
-------
Smooth Flow,
No Losses
gB
0.4 V
:TT
Theoretical
Maximum
Smooth Flow
Hydraulic
Jump
-TT2 ^170 ^078 -0.6 -0.4 -0.2 o 0.2 0.4 V,
Figure 12. Correlations for bows of slotted models
NACA 4415 and NACA 0015.
27
-------
Free Surf act
Model
Lower Stagnation Streamline
Middle Stagnation
Streamline
Separated Flow-
Upper
. Stagnation
Streamline
ro
Co
Model = NACA 0015
Slot Widths = 1.5cm (Each)
Velocity = 74 cm/sec
Freeboard = -1.22 cm
Solid =« Measured
Dotted = Calculated
Figure 13. Comparison of measured and calculated stagnation streamlines.
-------
LOWER
STAGNATION
STREAMLINE
UPPER
STAGNATION
STREAMLINE
SURFACE JET
HYDRAULIC JUMP
FREE
SURFACE
OUTLET SLOT
r\>
10
FRONT TWO SECTIONS
OF NACA 0015 MODEL
FLOW CONDITIONS MEASURED IN
SMALL SCALE OPEN CHANNEL
FLOW FACILITY
Figure 14. Preliminary streamlined oil boom model using NACA 0015 bow sections.
-------
SECONDARY SUMP
STAGNATION
STREAMLINE
(Calculated from
flowrate-0.11 cm
beneath surface)
FREE
SURFACE
UPPER
STAGNATION
STREAMLINE
SURFACE JET
HYDRAULIC JUMP
SECONDARY SUMP
FLOW IS PUMPED OUT
CO
o
LOWER
STAGNATION
STREAMLINE
RECIRCULATION
ZONE
NACA 4424 PROFILE WITH
2.B cm THICKNESS
NOTES:
VELOCITY = 65.81 cm/Sec
FREEBOARD = -2.30 CM
FLOWRATE OUT OF
SECONDARY SUMP = 0.054 L'Sec
Figure 15. Preliminary slotted bow design,
-------
airfoil and operated beneath the transition from smooth to turbulent flow.
The rear portion was designed with a relatively sharp leading edge to skim
a thinner layer into the sump and the slot contours were arranged to produce
gradually accelerating flow. This design performed very well allowing thin
layers to be skimmed at high as well as low speeds without aeration or a
turbulent bow wave. As a result of these tests an identical large-scale
model was constructed and tested at the Coast Guard Flow Channel in Texas
(see Table 1). In these tests, the bow design performed well and only minor
modifications were made in subsequent revisions.
The final bow design tested in both OHMSETT test series is shown in
Figure 16. Modifications include small changes to the sump inlet contours
to achieve thinner layers skimmed into the sump and changes to the aft port-
ion to accommodate a different sump design.
The OHMSETT tests demonstrated that this final design can operate with
a smooth bow wave at 3.0 m/sec skimming 77 I/sec per meter or a 2.5 cm layer
(6 knots skimming 374 gpm per foot width or a 1.0 inch layer). However, a
3.8 cm (1.5 inch) layer skimmed was chosen as the nominal operating point
to better accommodate waves. Tests at this operating point under calm
conditions demonstrated that all surface material passes into the sump
smoothly. ..Under low wave conditions, some wave energy is reflected so that
waves up to 10 cm (4.0 inches) through to crest can be tolerated with the
entire free surface still entering the sump smoothly. As wave heights in-
crease further, portions of the wave troughs pass under the bow and some
surface material is lost. A buoyant flotation system moving in response to
the waves or an operating point skimming a thicker layer would improve
performance in larger waves.
31
-------
CO
ro
15
(6
22.9 cm
(9 in.)
1
I
.2 cm
in.)
1
-^ 48 cm (19 in.) ^-
' (
15.2 cm (6 in.)
v* ---x
NACA 4424 AIRFOIL
CHORD = 63.5 cm (25 in.)
THICKNESS = 15.2 cm (6 in.)
0.32 cm (0.125 in.) NOSE RADIUS
CHORD LINE
Figure 16. Final bow design.
-------
SECTION 5
SUMP DESIGN
During the Phase I development, little work was done on sump design.
For the full-scale tests at the Lockheed Towing Basin a flexible plastic
membrane was installed in the sump and all material exiting the sump was
recovered in this plastic "bag." The original plan was to either separate
the oil from the excess water in the sump or to recover the entire flowrate
passing over the bow for subsequent shore based separation. However, the
large flowrate required for a smooth bow wave makes subsequent shore based
separation impractical. Based on the 3.0 m/sec (6 knots) final bow design
operating point skimming a 3.8 cm (1.5 inch) layer into the sump, a 30 m
(100 feet) wide streamlined oil boom would collect 212,400 1/min (56,100 gpm)
enough to fill a 30 m (100 feet) diameter tank at 29 cm (0.95 feet) per min-
ute. This gigantic flowrate is clearly impractical for a portable or even
a permanent oil/water separation system. Consequently the Phase II sump
development work focused on developing a design which incorporated oil/
water separation within the sump discarding the excess water (the bulk of
the flow) and retaining and recovering the oil.
Figure 17 shows the basic sump design concept operating at 3.0 m/sec
(6 knots) retaining and recovering an oil slick. A 3.8 cm (1.5 inch) layer
of'oil and water enters the sump at essentially freestream velocity and
flows down a 1/5 incline. Since the flow is moving at high speed it's Froude
number is greater than 1.0 and a hydraulic jump forms as shown. The posit-
ion of the jump is stabilized by the tailwater level which in turn is
regulated by the width of the exit slot in the sump's aft portion. The
hydraulic jump, one of several possible energy dissipation mechanisms, slows
the flow to roughly 0.3 m/sec (0.6 knots). Some of the oil slick remains
on the surface downstream of the jump while the remainder is broken into
droplets in the turbulent eddies and mixed with the water. Since the flow
downstream of the hydraulic jump is slow moving and relatively quiescent,
many of these droplets rise to the surface to form a reconstituted oil slick
in the aft portion of the sump. The excess water and any droplets not ris-
ing to the surface exit through the slot.
The aft end of the sump forms a barrier retaining the reconstituted oil
slick much as a conventional oil boom since the sump velocity is slow enough
to preclude the usual boom failure mechanisms. The aft end of the sump is
adjustable and acts as a weir to removing the oil once enough has accumulated
in a surface slick. The amount of excess water recovered with the oil de-
pends upon the recovery pumping rate, the thickness of the oil slick in the
33
-------
INCLINED
SUMP INLET
•HYDRAULIC JUMP
ENERGY DISSIPATION
RISING OIL
DROPLETS
RECONSTITUTED
OIL SLICK
WEIR
SECONDARY
SUMP
CO
SUMP INLET FLOW 3.8cm
(1.5 in.) LAYER"OF
OIL AND WATER ENTER-
ING AT 3.Dm/sec (6 knots)
SUMP BOTTOM
SMALL OIL
DROPLETS
ARE LOST
ADJUSTABLE
EXIT SLOT
Figure 17. Sump design concept - schematic drawing,
-------
sump, and specific weir geometry.
In the special case where the fluid entering the sump contains a high
percentage of oil, oil/water separation in the sump is not necessary. The
exit slot is closed and all the flow entering the sump is recovered over the
weir. However, such conditions are improbable for inland applications. If
the fluid entering the sump were 50 percent oil, the oil slick would be 1.9
cm (0.75 inches) thick corresponding to an oil flowrate of 3,500 1/min per
meter (280 gpm per foot) at 3.0 m/sec (6 knots). More probably, oil slicks
in the range, of 1.0 to 3.0 mm would be encountered corresponding to 3 to 8
percent of the total sump flow.
The inclusion of oil/water separation into the sump design makes the
characteristics of the energy dissipation process critically important. If
the oil is broken into fine droplets or emulsified, the sump residence time
will not be long enough to allow the droplets to rise to the surface. The
oil/water separation process can also be hampered by high turbulence levels
and nonuniform velocity profiles produced by energy dissipation. Since
inertia!, gravitational, viscous and surface tension forces are all import-
ant and the process is sensitive to oil properties, experiments must be
conducted at full-scale.
The sump must perform several interrelated functions simultaneously:
Receive all flow passing over the bow
Dissipate kinetic energy
Separate oil from excess water
Discard excess water
Retain oil in sump
Recover oil from sump
Interface with the stability control system
The requirements for each of these functions and the sump's basic operation-
al parameters can be deduced from the bow design and normal operating point.
Based on a 3.8 cm (1.5 inch) layer skimmed into the sump at 3.0 m/sec (6
knots), the flowrate into the sump is 116 I/sec per meter width (561 gpm per
foot width). A 3 mm oil slick would enter the sump at 9.2 I/sec per meter
width (44 gpm per foot width) representing 7.9 percent of the total. If
the oil is to be retained in the sump under conditions similar to a convent-
ional oil boom, the sump velocity must be no more than 0.30 m/sec (1.0 feet/
second). For this sump velocity continuity requires that the flow depth must
increase to 38 cm (15.0 inches), assuming no width change, and since the
kinetic energy is proportional to the square of the velocity, 99 percent of
the total flow's incident kinetic energy and 100 percent of the oil's kin-
etic energy must be dissipated. Assuming an effective sump length of 3.0
m (10 feet), if the oil is not recovered it will build up in the sump at a
rate of 18.0 cm/min (7.1 inches/minute). If the oil is recovered as it is
encountered, the equilibrium sump oil layer thickness will be 3.0 cm (1.2
inches.)
35
-------
ENERGY DISSIPATION
Early in the sump design process it was recognized that energy dissipa-
tion was the most critical function. If in the process of energy dissipa-
tion, the oil slick is broken into very fine droplets, there is little hope
of recovering a substantial portion in a compact sump. A gravity separator
with a capacity of 2.100 1/min (560 gpm ) (equivalent to a 0.3 m_(l foot)
width of streamlined oil boom) would require approximately 5.3 nr (1,400
of space and weighs 7.0 metric tons (6.9 tons) dry.* This is clearly impra-
ctical for a streamlined oil boom section even a few feet wide.
Ideally, the energy dissipation mechanism should be simple, nonclogging,
insensitive to flowrate and oil properties, and avoid breaking the oil slick
into fine droplets. Since energy dissipation through turbulence generation
and decay meets all of these requirements except for its unknown performance
on oil slicks, a majority of the sump development focused on iteratively
testing with oil, various sump designs employing this priciple and measur-
ing the subsequent separation of oil and water.
Experiments were conducted in the sump test flow channel, shown schem-
atically in Figure 18. This flow channel was constructed to be a full-
scale internal model of the sump in all dimensions except width. Full-scale
allowed meaningful experiments to.be conducted with oil without scaling
problems and the high costs of tow testing a complete streamlined oil boom
model merely to obtain sump design data. Since the streamlined oil boom
is essentially two-dimentional, a specific width is not important for accur-
ate testing. A width of 8.9 cm (3.5 inches) was chosen to produce a 3.8
cm (1.5 inch) layer of water at 3.0 m/sec (6 knots) entering the sump
with the available pumps and equipment. The channel bottom had a series of
1.9 cm (0.75 inch) holes joining a plenum beneath to simulate a porous sump
bottom. The holes could be individually plugged or left open to model any
flow exit configuration. Water passed through the porous bottom into the
plenum, a manifold system and into two pumps which supplied water to the
head tank. Water from the head tank entered the channel (or sump) in a thin
layer at high speed (supercritical flow) and flowed down the incline. The
flow's kinetic energy was dissipated through a hydraulic jump or other diss-
ipation mechanism, flowed through the porous bottom and the cycle repeated.
The channel was designed to model sumps up to 7.3 m (24 feet) long and the
aft portion could be blocked off to simulate shorter sumps.
A light lubricating oil was used for the majority of the tests. It was
deposited on the surface of the high speed flow immediately downstream from
the head tank and recovered from the surface in the aft portion of the
simulated sump with a small self-priming recovery pump. The channel was
thoroughly cleaned after each test so that the volumes of oil deposited and
recovered would provide and accurate measure of sump performance.
* Dimensions are based on a model 14.6 gravity separator with a capacity
of 500 gpm manufactured by Hyde Products, Inc., Cleveland, Ohio.
36
-------
CLEAR
PLASTIC
'SIDEWALL
ADJUSTABLE
HEAD GATE
HEAD
'TANK
46 cm (18 in)
CO
POROUS
BOTTOM
(HOLES MAY BE
CLOSED OR .LEFT
OPEN)
INCLINE
MATCHES
FULL-SCAL
DESIGN
7.3 m (24.0 ft
61 cm (24 in)
FLANGED OUTLETS
ON BACK SIDE
OF PLENUM
PLENUM COLLECTS
WATER FROM
POROUS BOTTOM
SUCTION
TUBING
BLOCK SUPPORTS
FOR CHANNEL
TWO PUMPS
80 GPM EACH
303 1/min. (80 gpm)
EACH
OIL DRUM
SUPPORT FOR
HEAD TANK
Figure 18. Sump test flow channel.
-------
Several energy dissipation mechanisms were investigated analytically
and those found suitable for the streamlined oil boom were iteratively test-
ed in the sump test flow channel with oil. Out of the many designs tested
the four shown in Figure 19 performed the best. The important characteris-
tics of each device along with the results of the oil tests in the sump
test flow channel are shown in Table 2.
The normal hydraulic jump is the simplest design and has no obstruct-
ions to cause clogging. Flow enters the sump at essentially freestream
velocity and accelerates down the incline where a turbulent roller forms
dissipating energy. The hydraulic jump automatically positions itself in
response to the tailwater level producing a relatively uniform downstream
velocity profile. The scale of the turbulent eddies produced is of the
same order as the roller dimensions and these eddies decay downstream. The
normal hydraulic jump sump design was tested full-scale in the first OHMSETT
test series (see Table 1) at 1.5 and 3.0 m/sec (3 and 6 knots). Good res-
ults were obrained at 1.5 m/sec (3 knots) with the majority of the oil
slick appearing on the surface immediately downstream of the roller and
thruput efficiencies as high as 86.5 percent were achieved. At 3.0 m/sec
(6 knots) however, the amount of energy dissipated in the jump was four
times as great producing large violent eddies and breaking the oil slick
into much smaller droplets. Efficient oil/water separation was impossible
with these small droplets and the high sump turbulence level and as a result
thruput efficiency dropped to near zero. The experiments in the sump test
flow channel confirmed this performance and a maximum thruput efficiency of
25 percent was obtained with a sump length twice as long as that used for
the other options. Because of the desire to develop a system with good 3.0
m/sec (6 knot) performance most of the development work focused on the other
energy dissipation mechanisms. However, the freedom from clogging, simplic-
ity and good 1.5 m/sec (3.0 knot) performance make the hydradulic jump an
excellent candidate for a lower speed design.
The array of surface piercing rods dissipates energy through the hydro-
dynamic drag on each rod. Since the rods are surface piercing energy is dis-
sipated uniformly'throughout the flow depth without the formation of a roller
or other flow separation. The equations of motion for flow through an array
of surface piercing rods can be developed by assurming a drag coefficient for
each rod and modifying the usual equations for open channel flow to account
for the energy loss. The final equations after some manipulation are identi-
cal to the equations of two dimensional open channel flow with the surface
friction terms replaced by terms involving the rod drag coefficient.
The hydrodynamic drag coefficient for cylindrical rods normal to the
flow (based on frontal area) is approximately 100 times as great as the
drag coefficient for a flat plate such as the sump bottom (based on the
plate area). Consequently a relatively close packed array where the front-
al area of the rods is approximately the same as the sump bottom area can
provide energy dissipation in approximately 1 percent of the space required
for flat plate friction. Furthermore since the energy dissipation is propor-
tional to the square of the velocity, the array of surface piercing rods
38
-------
NORMAL HYDRAULIC JUMP
.HIGHSPEED FLOW ROLLER
LOW SPEED FLOW
ARRAY OF VERTICAL
SURFACE PIERCING RODS
90"
ARRAY OF INCLINED
SURFACE PIERCING RODS
FLOW DETAILS POROUS
WCLINED PLATE
RODS
POROUS INCLINED PLATE
RODS
PIVOTED FLOW
STRAIGMTENERSI6I
FREE SURFACE
\ \ \
FLOW STRAIGHTENING PLATE
Figure 19. Energy dissipation mechanisms.
39
-------
TABLE 2. ENERGY DISSIPATION DEVICES; IMPORTANT CHARACTERISTICS
AND RESULTS OF TESTS IN SUMP TEST FLOW CHANNEL
Energy
Dissipation Device
Construction Details
Thuuput*
Efficiency
Clogging
Rating**
(l=Best)
Hydraulic Jump
Incline Slope = 1/5
25
Array of vertical surface piercing
rods
Rod diameter 0.64 cm (1/4 in)
Rod density - 0.088/cn2(0.57/in2)
Array geometry staggered rows
spaced 2.5 cm (1.0 in) in
flow direction
Array of inclined surface piercing
rods
Inclined expanded metal
Incline angle - 45° in flow
direction
Rod diameter = 0.64 cm (1/4 in)
Rod density 0.12/cm2 (0.76/in2)
Array geometry = staggered rows
spaced 1.3 cm (0.5 in) in flow
direction with every third row
empty
Double layer of expanded metal
louvers inclining in direction
of flow
Incline angle = 1.9°
Upper layer:
Louvers declining in flow
direction
Hole spacing: in flow direction
1.97/cm (5.0/in); normal to
flow direction = 0.98 cm
(2.50,'in)
Lower layer:
Louvers inclining in flow
direction
Hole spacing: in flow direction
1.57/cm (4.0/in); normal to
flow direction = 0.52/cm
(1.33/in)
Layer spacing = 1.90 cm (0.75 in)
71-77
75-81
78
Measured in sump test flow channel with 3.0 m/sec (6 knots) inlet velocity with 30 wt
motor oil (except hydraulic jump -vegetable oil).
Clogging rating is a qualitative judgement of estimated performance with small debris.
40
-------
tends to smooth out velocity profile irregularities and produce a uniform
velocity profile downstream. The scale of the turbulent eddies is roughly
the same size as the rod diameter, much smaller than the eddies produced
by the hydraulic jump, and consequently the eddies decay faster producing a
more quiescent flow downstream. Several geometries were tested to optimize
rod diameter.and spacing. Best results were achieved with 0.63 cm (0.25 inch)
diameter rods speced 0.088/cm2 (0.57/in2) and thruput efficiencies in the
range of 71 to 77 percent were obtained in the sump test flow channel exp-
eriments with a 3.0 m/sec (6 knot) inlet velocity.
One obvious disadvantage of the surface piercing rods is their tendency
to catch debris, particularly long stringy debris such as weeds, rags, sti-
cks, etc. Inclining the rods at a 45° angle reduces the clogging problems.
With inclined rods, the fluid has a velocity component parallel as well as
perpendicular to the rods and debris is carried upward to the free surface
level. Occasionally lowering the array so the rod tips are beneath the
surface would wash away any debris collected. Since the velocity component
perpendicular to the rods is smaller, the drag on each rod is also smaller
and the rod spacing must be decreased to achieve the same energy dissipation.
As expected, best results were achieved with the same rod diameter as the
vertical array and a spacing of 0.12/cm2 (0.76/in2 ). Thruput efficiencies
in the range of 75 to 81 percent were obtained in the sump test flow channel
experiments with a 3.0 m/sec (6 knots) inlet velocity. This is comparable
to the performance of the vertical rods considering experimental errors.
The energy dissipation devices discussed above are similar in that all
of the fluid entering the sump, both oil and water, passes through the
turbulent area simultaneously and as a result when the oil slick is broken
into droplets, the droplets mix with the entire flow. The porous inclined
plate energy dissipation mechanism remedies this problem by allowing the
oil slick to pass through the turbulent zone last so that any oil droplets
are mixed with only the uppermost layer of water. The porous inclined plate
is a double layer of expanded metal louvers slightly inclined to the flow
direction so that the flow moves nearly parallel to the incline.
The upper layer acts as a cascade or blade row directing fluid downward
with little energy loss. The lower layer is coarser, inclined in the opp-
osite direction and acts as a stalled cascade with high drag and energy di-
ssipation. Flat plates hinged to the incline's bottom side redirect flow in
the downstream direction, smooth turbulent eddies and pivot in response to
the local static pressure to maintain a uniform downstream velocity profile.
Experiments in the sump test flow channel gave a thruput efficiency of 78
percent with a 3.0 m/sec (6 knot) inlet velocity, comparable to the arrays
of surface piercing rods. The porous incline is expected to offer less
clogging problems than the arrays of surface piercing rods. Debris larger
than the holes in the expanded metal is carried up the metal surface by the
flow and may be manually moved as required. Small debris may tend to acc-
umulate on the expanded metal surface. However backflushing or brushing the
surface should remove most materials.
41
-------
OIL/WATER SEPARATION
Downstream of the energy dissipation device, the mixture of oil and
water enters the oil/water separation area where the buoyant oil droplets
rise to the surface to form a reconstituted oil slick. The slick is retained
against the aft end of the sump and recovered as required while excess water
is discarded. This portion of the stimp is essentially a gravity oil/water
separator operating with a very high flowrate.
Gravity oil/water separators have been the subject of considerable
research. The principles of operation, scaling parameters and design det-
ails for efficient operation {are all well known and units of varying capacity
are available from many commercial sources. The American Petroleum Institute
(API) recommends* designing oil/water separators to remove oil droplets as
small as 0.015 cm (0.006 inch) and to have a maximum flow velocity of 1.52
cm/sec (0.6 in/sec or 0.03 knots). Applying these specifications to the
streamlined oil boom sump operating point gives a minimum sump depth of 7.6m
(25 feet), impractical for $ streamlined oil boom skimming only 3.8 cm (1.5
inch). Furthermore, the minimum-size droplets to be recovered is the minimum
size droplet produced in the energy dissipation process which may be either
smaller or larger than 0.015 cm (0.006 inch).
This portion of the streamlined oil boom sump was designed to separate
as small a droplet size as was feasible within reasonable sump size constr-
aints. Experiments were conducted in the sump test flow channel testing
various sump geometries, lengths and flow outlet configurations in conjunction
with alternative energy dissipation devices to optimize performance. The
final streamdined oil boom design is shown in Figure 20. The sump is ess-
entially a rectangular chamber with a porous bottom over a triangular plenum
with a pivoting weir skimmer at the aft end. After passing through the
energy dissipation device, the mixture of water and oil droplets enters the
chamber and begins to separate by gravity. As the flow proceeds through the
sump, the excess water (and some oil) is withdrawn through the porous bottom
into the triangular plenum. The fluid progressively withdrawn is the clean-
est (freeist from oil droplets) because it is nearest the bottom. Since the
sump's horizontal velocity component decreases monotonically to zero at the
aft end, the effective sump residence time for oil droplets is increased.
The rising oil droplets migrate to the surface where they form a
reconsituted oil slick. Since the sump's horizontal velocity drops to zero
at the .aft end, the slick is effectively retained without the usual high
current oil boom failure mechanisms such as drainage and droplet entrain-
ment.**
* Manual on Disposal of Refinery Wastes, "Chapter 5 - Oil-Water Separator
Process Design," American Petroleum Institute, 1969.
** However, if the oil is allowed to accumulate until the slick thickness
approaches a large portion of the sump depth, these failure mechanisms
may occur causing some losses.
42
-------
3.8cm (1.5 in.)
LAYER SKIMMED
ENERGY
DISSIPATION
MECHANISM
OIL RECOVERY
SUMP/SKIMMER
OIL SLICK ON
WATER SURFACE
OJ
FLOTATION
ATTACHMENT
POINT
STABILITY
SYSTEM
PIVOT POINT
RETAINED PIVOTING
OIL WEIR
FIBERGLASSED
FOAM FLOAT
CABLE
PULLEY
ATTACHED TO
PIVOTING ARM
.CABLE
-GA'T°ENG PULLEY
ATTACHED TO
•SIDE PLATES
"Srf"
FOAM
FLOTATION
ATTACHED TO
SIDE PLATES
NACA 4424
AIRFOIL
BOW DESIGN DETAILS
48cm (19 in.)
22.9 cm
POROUS BOTTOM PLENUM BOTTOM
2.8 cm (1.125 in.) DIA. PLENUM
HOLES 18% OPEN DIVERGENCE
ANGLE «5.7°
0.32 cm (0.125 in.) NOSE RADIUS
.3'
CHORD LINE
NACA 4424 AIRFOIL
CHORD - 63.5 cm (25 In.)
THICKNESS,- 15.2 cm (6 in.)
SIDE PLATE
OUTLINE
NOTES:
1. Any of the Energy Dinapation
Mechanisms Shown in Flgura
19 Can I)e Used.
2. Oil li Recovered from Sump/
Skimmer with Hon and Pump
System (Not Shown).
3. Dimentions: Overall Length »
6.7m<22 It.)
Side Plate Hieght = 1.0m (40 in.)
4. Coble System Amplifies Gate
.Motion by Factor of 3.
Figure 20. Final streamlined oil boom/skimmer design.
-------
The performance of this sump design in separating oil droplets can be
estimated following the procedures for conventional oil/water separation
neglecting turbulence, short circuiting and the interaction of oil droplets
one with another. Under these conditions it can be shown that all oil drop-
lets with a terminal rising velocity greater than
R ~ L W
where:
QO = sump flowrate
L = sump length
W = sump width
V^ = terminal rising velocity
will rise to the surface to form a reconstituted oil slick regardless of
their initial depth in the sump. For droplets with slower terminal rising
velocities the initial depth is important and all droplets initially a dist-
ance Y_ above the bottom rise to the surface:
nL W
- H J where D = sump depth
These relationsips can be applied to determine the minimum droplet size
collected by the streamlined oil boom sump. This calculation was performed
using the bow design point with a 3.8 cm (1.5 inch) layer entering the sump,
the geometry shown in Figure 20 and an oil with a specific gravity of 0.875.
Figure 21 shows the calculated minimum droplet size completely recovered as a
function of the current velocity. At the 3.0 m/sec (6 knots) design point,
all droplets greater than 0.12 cm (0.047 inch) are recovered. While this is
a relatively small droplet, it is eight times the recommended API design
point.
WATER REMOVAL AND OIL RECOVERY
The porous bottom and plenum beneath it were designed to withdraw fluid
uniformly across the entire sump bottom. This produces a linearly decreasing
horizontal sump velocity as the fluid moves aft and a vertical velocity
proportional to the distance from the free surface. It can be shown that
these conditions approximate potential flow* and thus minimize the formation
of vortices which might entrain oil.
* Potential flow is incompressible flow with zero or constant vorticity;
it is not influenced by viscous effects.
44
-------
.3-
MINIMUM OIL
DROPLET
DIAMETER
(cm)
NOTES
1. OIL SPECIFIC GRAVITY =0.875
2. WATER VISCOSITY = 0.01 cm2/sec2
3. LAYER SKIMMED INTO SUMP = 3.8 cm (1.5 in.)
4. SUMP LENGTH POROUS BOTTOM 2.3m (92 in.)
.2- -
TOW VELOCITY
(m/sec)
Figure 21. Minimum oil droplet size completely removed from
final sump design as a function of tow velocity.
45
-------
The plenum has a constant included angle of 5.7° producing a cross-
section area increasing linearly in the downstream direction. The angle was
chosen as a balance between a substantial passage width and minimum pressure
drag on the flow passing beneath. The porous bottom is an array of 2.86 cm
(1.125 inch) diameter holes arranged to produce a porous plate approximately
18 percent open. The holes are slightly closer together (greater percentage
open) in the front portion to allow for pressure losses in the plenum. The
pressure loss across the porous bottom is approximately 0.58 cm (0.23 inch)
of water at the 3.0 m/sec (6 knot) design point. This small loss is insig-
nificant compared to the free stream dynamic pressure of 47 cm (1.55 feet)
and leaves ample pressure differential to control flow out of the plenum
with a downstream restriction.
Figure 22 shows the operation of the oil removal device. It is a self-
adjusting pivoting weir which automatically adjusts to variations in the
sump water level. The weir edge is pivoted about a horizontal axis and
supported by a float in the secondary sump. As fluid is pumped, from'the
secondary sump, the fluid level drops and the weir pivots allowing a thin
layer to be skimmed. In this manner the weir follows the secondary sump
fluid level maintaining a small differential between the levels in the main
and secondary sumps. The size of the float determines the difference in
levels and the thickness of the layer skimmed measured at the weir edge.
For a small float, the difference in levels is small and the weir edge pivots
substantially allowing a thick layer to be skimmed. For large floats, the
difference in levels is larger and the thickness of the layer skimmed decre-
ases. The minimum thickness is achieved when fluid passes over the weir as a
free overfall. Under these conditions a 1 mm oil slick entering the sump at
3.0 m/sec'(6 knots) would pass over the weir as a 0.98 cm (0.39 inche) oil
slick.* The weir edge was serrated to reduce flow variations over the weir
width.
This weir oil recovery device is only one of several possible devices
to remove oil once it has accumulated in the sump. Other possiblities
include:
a Conventional floating skimmer head mounted in the sump
• Rotating disc or drum skimmer
• Sorbent skimming system
This calculation assumes that the Froude number based on local velocity
and flow depth at the weir is unity.
46
-------
Oil
Removal
Pi pes
Secondary
Sump Fluid
Level
Secondary
Sump Box
Foam
Float
Main Sump
Fluid Level
Hinge and
Elastomer
Seal
Figure 22. Self-adjusting weir and secondary sump.
. 47 -
-------
SECTION 6
FORCES AND STABILITY
The forces exerted by the flow on the streamlined oil boom and its res-
ponse to those forces are important considerations. If it is deployed as a
boom spanning a wide high current area, an integral flotation system must
be provided to counteract any vertical loads without porpoising or other in-
stabilities and a cable system must be provided to oppose the drag. If it
is deployed as a high speed skimming craft, the lift and drag must be con-
sidered in designing hulls and determing powering requirements.
FORCE MEASUREMENT
Vertical forces on the boom are complex combinations of hydrodynamic
and hydrostatic forces. Hydrodynamic forces are produced by changes in
velocity and direction of fluid passing over, around, and through the boom,
and viscous shear stresses. Hydrostatic forces are caused by the difference
in .water level inside and outside the sump. If the level inside the sump
is higher than outside, there is a net force downward due to the weight of
the excess water contained in the sump and vice versa. Although hydrostatic
forces can be estimated from expected water levels, hydrodynamic forces gen-
erally defy analysis where the flow is complex and adjacent to a free surf-
ace. The horizontal force or drag on the boom is entirely hydrodynamic and
increases as the square of the tow velocity (approximately).*
The forces on the streamlined oil boom were measured during the Phase
II series at the Lockheed Towing Basin in San Diego. In these tests a 0.6 m
(2 feet) wide full-scale model was rigidly attached to a carriage and towed
at speeds up to 3.0 m/sec (6 knots) over a range of operating parameters as
forces and water levels were measured. Figure 23 shows the design of the
model used in these tests.** Six load cells were used to measure forces.
Four of these (two forward and two aft) measured lift forces, and the other
two (in series with the forward lift load cells) measured drag. Two water
level transducers were also used, one over the sump inlet and the other just
upstream of the weir. Data from these transducers allowed the relationships
between operating parameters and lift and drag to be determined.
* Force measurements during the Phase I Lockheed tests showed pronounced
dips in the drag versus velocity curve as the bow wave passed through
various flow regimes,
** Since this model was not the final design the numberical values of the
forces measured are not the same as the final design, but the general
behavior is similar.
48
-------
TOP VIEW
3?
/
. — "
\ \
. i
i
T '
! J ,
1 0.6 m (2 ft)
/ t
Rear Vertical / / \
Parallel Flow Front Vertical Plywood /
Straighteners Parallel Flow Sidewall \
10 cm (4 in) Straighteners 1
Spacing 10 cm (4 in) clear P]ast1c
Pipes to
Recovery
Manually and
Automatically
Adjusted
Serrated Edge
Weir
Manually
Adjusted
Slot
Horizontal Flow
Straighteners
5 cm (2 in) Spacing
Sump
Inlet
SIDE VIEW
Front Support
(Lift and Drag)
Figure 23. Model for Lockheed tests.
-------
Perhaps the most important result of these force measurements is that
hydrodynamic forces were found to be small compared to hydrostatic forces.
This is shown in Figure 24, a plot of total vertical force against differ-
ence in free surface level 6 (inside sump minus outside sump). If only
hydrostatic forces were important, the data would show vertical forces dir-
ectly proportional to € . The data exhibit this general trend but with
some data scatter due to hydrodynamic forces which vary with other paramet-
ers not held constant.*
This implies that flotation requirements could be significantly reduced
by maintaining the level of water inside the sump nearly the same as the
level outside the sump. From" Figure 24 it can be seen that in all cases
where the absolute value of e was less than 0.5 cm (0.2 inch), the total
vertical force was less than 89 N (20 pounds). This corresponds to a flot-
ation requirement of 146 N per meter of boom width (10 pounds per foot of
boom width). These small forces imply relatively small buoyancy require-
ments.
The water level in the sump (and hence € ) is a function of the diff-
erence in flowrate into and out of the sump. If the flowrate into the
sump is greater than the flowrate out, excess fluid builds up in the sump
and the water level increases and vice versa. Since the streamlined oil
boom processes a large flowrate, even small percentage flowrate differences
cause large sump weight changes. For example, if the streamlined oil boom
is operating at the 3.0 tn/sec (6 knot) design point with 116 I/sec per meter
width (561 gpm per foot width) entering the sump, and the flowrate leaving
the sump is only 10 percent less, excess fluid will build up in the sump at
the rate of 119 N/sec per meter width (490 pounds/minute per foot width).
This large weight change coupled with the characteristics of a flotation
system can lead to dynamic instabilities.
HYDRODYNAMIC INSTABILITY
Consider the operation of the streamlined oil boom supported by a
buoyant flotation system as shown in Figure 25. The exit slot-width is
adjusted so that at equilibrium the flowrate entering the sump equals the
flowrate leaving the sump and the weight of excess water in the sump is
balanced by the flotation buoyancy. Although this operating point is an
equilibrium condition it 1s not necessarily a stable equilibrium. If the
flotation system is too compliant and for some reason the flowrate entering
the sump slightly exceeds the flowrate out of the sump, excess fluid would
accumulate in the sump forcing the flotation downward and further increasing
the flowrate into the sump and so on until the sump swamped. Stable oper-
ation can only be achieved where the flotation system is large enough to
accommodate these sump weight changes.
Such as velocity, freeboard, etc.
50
-------
TOTAL
\ VERTICAL'
n FORCE
X • IN Ibs. .
\
*
'•'.\
' • \ *
• •
\ . .
-1.2 -1.0 -.8 -.6 -.4 -.2v
.\
-20-
-40-
-60-
-80-
-100-
-120-
-140-
-160'
-180'
-200-
-220'
-240
-120
•100
POSITIVE FORCES
• 80 ARE UPWARD
• 60
• 40
• 20
..2 .4 .6 .8 1.0 1.2 €
DIFFERENCE IN WATER LEVELS
^ INSIDE MINUS OUTSIDE THE SUMP
\
' \
. \ '
\ .
' .' \
\
\
»
•
C x"
LEAST SQUARES •. •
FIT-DATA TREND \
\
\
\
• \
. \
\
Figure 24. Total vertical force vs. difference in free surface levels
(inside minus outside the sump).
51
-------
EXCESS
FLUID
CONTAINED
IN SUMP
BUOYANT
FLOATATION
SYSTEM LIFT
WEIGHT OF
EXCESS FLUID
INLET
FLOW
en
OUTLET
FLOW
FREE SURFACE
LEVEL OUTSIDE
SUMP
BASE OF
HYDRAULIC
JUMP
Figure 25. Schematic diagram of streamlined oil boom
supported by buoyant floatation system.
-------
This instability was observed in the 1975 OHMSETT tests when a 1.5 m
(5 feet) wide full-scale model was rigidly attached to a flotation supported
catamaran work platform and tow tested at 1.5 to 3.0 m/sec (3 and 6 knots).
In these tests the freeboard was adjusted to vary the flowrate into the sump
and a manually adjusted slot was positioned to set the flowrate out of the
sump. At 1.5 m/sec (3 knots), operation was stable. As the boom was brought
up to the test speed all parameters approach equilibrium asymtotically with
no hunting or oscillations.
At 3.0 m/sec (6 knots), however, stable equilibrium could only be
attained with a very low sump water level and a very thin layer skimmed in-
to the sump. The streamlined oil boom could not be adjusted for stable oper-
ation at the design point. All attempts to increase the layer skimmed or
sump depth by decreasing freeboard or closing the slot caused the sump to
swamp by the mechanism described above. In order, to obtain a "quasi;" stable
equilibrium for the 3.0 m/sec (6 knot) tests, the slot width was continously
adjusted during the experiments in response to changes in sump water level.
A stability analysis was conducted to calculate the flotation require-
ments for stable operation. The slope of the force versus vertical dis-
placement curve (as shown in Figure 26) determines stability. For a posit-
ive slope the equilibrium is unstable; and for negative slope it is stable.
The borderline case of zero slope is astable, with many equilibrium states
and the boom "hunts" moving up and down with no net force.
The vertical motion of the streamlined oil boom can be modeled subject
to the following assumptions:
1. Vertical forces exerted on the boom are a result of
the weight of the excess water in the sump and are
proportional to the difference in water levels inside and outside
the sump,6 .
2. Hydrodynamic forces are neglected.
3. Flow into the sump increases linearly with draft. -,•
4. Flow passes out of the sump through a submerged
slot of fixed width Sw.
5- Inertial forces are neglected.
6. Vertical forces are opposed by a flotation system with
stiffness, S.
The first two assumptions follow from the force measurements at the
Lockheed Towing Basin and the third assumption is a direct result of the bow
wave analysis.* Neglecting inertia! forces (the fifth assumption) confines
Neglecting second order effects,
53
-------
NET FORCE
EXERTED ON
OIL BOOM BY
FLOW
UPWARD
FORCE
DOWNWARD
FORCE
•UNSTABLE
ASTABLE
VERTICAL
POSITION
EQUILIBRIUM
POSITION
(ZERO FORCE)
Figure 26. Vertical stability as a function of net vertical force.
54
-------
the analysis to vertical motion much slower than that produced by wave mo-
tion in the sump. Flotation stiffness is the ratio of increased lift to
increased displacement as a flotation device is forced downward. For exam-
ple, a buoyant float with a one square meter of area measured at the water
line would have 98 N/cm stiffness in fresh water (a 1.0 square foot float
would have 5.2 pounds/inch stiffness).
The results of the analysis define a dimensionless stability
number S..:
N
SSw2g
O..
N v1 2 w L YI
where:
SN > 1 Stable
S = 1 Astable
S < 1 Unstable
C~ = Sump exit slot discharge coefficient (assumed - 1.5)
y = Fluid weight density
and the other parameters are evaluated at equilibrium.
Stability numbers for the OHMSETT tests were calculated from the
etnperical data and the results are shown in Figure 27. The analysis provid-
es a good indication of stability but the dividing line between stable and
unstable cases should be about (0.8) instead of 1.0. This difference is
most likely due to the uncertainties in the effective sump length (not
directly measured), and the layer thickness Y, measured only to the nearest
0.63 cm (1/4-inch) - a 25 percent uncertainty on a 2.5 cm (1.0 inch) layer.
The analysis is helpful in understanding stability for various oil boom
designs over a range of test conditions. For example, it shows that incre-
asing velocity tends to make the oil boom unstable (expected and observed at
OHMSETT) and that decreasing the layer skimmed Y-j tends to increase stab-
ility (unexpected and observed at OHMSETT).
This stability problem can be solved in two ways. Either enough flot-
ation stiffness could be provided to achieve stability by "brute force,"
or the flowrate out of the sump could be adjusted to match the flowrate into
the sump. While the stiff flotation method could be used, the stability
analysis shows the minimum stiffness for stability at 3.0 m/sec (6 knots) to
55
-------
J
1
0
•
•
a
©
ID
0
£3 13 GJ
a a
§3
STABLE
N
-6.0
SM = STABILITY NUMBER
c c 2
S S g
w a
Vj2 w L YI cDr
•5-° (SEE TEXT FOR DEFINITION
OF TERMS)
Q] .5 m/sec (3 knots)
"4'° Q3.0 m/sec (6 knots)
•3.0
•2.0
CALCULATED
^ STABILITY
( LIMIT
i.o t.
-...-.
°-° UNSTABLE
Figure 27. Comparison of stability numbers with observed
stability - OHMSETT Tests 1975.
56
-------
be 2,900 N/cm per meter of boom width (500 pounds/inch per foot width).
This corresponds to an area several times the effective sump area and is
clearly impractical. Consequently, the remaining development focused on
designing an automatic stability control system to continuously adjust flow
out of the sump in response to changing operation conditions.
STABILITY CONTROL SYSTEM
The streamlined oil boom support system was developed as two subsystems:
a stability control system which adjusts the flowrate out of the sump to
match the flowrate into the sump thus minimizing hydrostatic forces, and a
flotation system to oppose the remaining forces.
There are several ways to adjust flow out of the sump but the simplest
and most promising method is to regulate the width of an outlet slot in
response to the weight of the sump. A stability analysis similar to that
conducted for fixed slot widths was conducted on several alternative designs
and the system shown in Figure 20 (page 43) was chosen for experimental
optimization. In this design the streamlined oil boom is pivoted from a
front flotation system with a center of buoyancy located near the sump inlet *
The after portion of the boom is free to move in response to the vertical
hydrodynamic and hydrostatic forces. Flowrate into the sump is determined
primarily by the freeboard at the sump inlet* and flowrate out of the sump
is regulated by a sliding gate moving in response to the float in the sump.
If the flowrate leaving the sump is too small the rising sump fluid level
will raise the gate until the flowrates are equated, and the sump weight is
counteracted by the small block of flotation in the aft end.
This sytem allows adjustment of four key control system parameters:
• operating point
t gain
• damping
a feedback
The operating point (equilibrium sump depth) may be adjusted by Varying the
length of the vertical link, the gain (ratio of change in flowrate to
change in float position) may be adjusted by changing the linkage geometry
and the damping (float response time) may be adjusted by restricting the
passage connecting the main sump and float chamber. Feedback is the rela-
tionship between flowrate into the sump and the operation of the stability
control system and may be adjusted by changing the flotation pivot point.
If the pivot point is at the sump inlet, the stability control system motion
will not affect the flowrate into the sump - zero feedback. If the pivot
point is moved aft, a downward motion of the sump's aft portion will raise
the sump inlet reducing the flowrate into the sump - negative feedback.
* Angle of attack changes produce only small flowrate changes.
57
-------
The operation of this stability control system is a complex nonlinear
problem. Although the simplified stability analysis indicated that stable
operation could be achieved, it did not model the system with enouxjh accu-
racy to permit analytical optimization. Experiments were required.
Since the stability control system responds primarly to gravity and
inertia! forces, Froude number modeling can be used to conduct experiments
on small scale. Proper scaling requires velocity to be scaled as the square
of physical dimensions. One-third scale was chosen for the experiments
since it produced a convenient model size roughly 2.2 m. (7.3 feet) long, and
3.0 m/sec (6 knots) full-scale tow velocities scale to 1.76 m/sec(3.4 knots)*
well within the speed capability of a small boat powered with a small out-
board motor.
A 1/3-scale model of the entire oil boom was constructed from fiber-
glassed foam and resin coated plywood using construction techniques similar
to those employed in the full-scale model. It incorporated an array of
vertical rods in the sump a$ the energy dissipating mechanism and a stability
control system with wide ranges of adjustment. The model was attached to a
force balance and suspended between the hulls of a 4.8 m (16 feet) catamaran
powered by a 7.5 hp outboard motor for tow testing in Newport Harbor.
Since the apparatus could operate at test speed for long periods, both
starting transients and equilibrium operation could be observed. Four test
series were run to observe the control system operation as the control system
parameters were varied. Those combinations of adjustments found to be unsta-
ble were eliminated and the remaining experiments concentrated on measuring
performance under stabe conditions as freeboard was varied.
Overall the stability control system was found to operate simply and
reliably and could be adjusted to give stable operation with the.model
pivoted at the sump inlet and at the middle of the sump incline. Some damp-
ing was beneficial and reduced oscillations under certain conditions. How-
ever, with more damping the time lag between sump water level changes and
float position changes was too great causing an out of phase feedback instab-
ility with large oscillations and porpoising.
When the model was pivoted at the bottom of the incline, stable opera-
tion could not be achieved. Large slot width and sump depth oscillations
were observed with a frequency of approximately 1.0 Hertz. Damping and aft
buoyancy improved performance but stable operation could not be achieved
with any combination tested.
Pivot point location is an important variable. When the model is pi-
voted at the sump inlet, the stability control system balances the weight of
excess water in the sump and consequently minimizes vertical forces on the
front flotation. However, since the freeboard at the sump inlet and hence
the layer skimmed are determined directly by the flotation system, its para-
meters are critical to proper operation. As the pivot point is moved aft of
58
-------
the sump inlet, the amount of sump weight supported by the front flotation
increases but the negative feedback helps to adjust the sump inlet position
minimizing the requirement for accurately setting flotation system parame-
ters.
The force balance measured lift and drag at each operating point, and
the results can be extrapolated to full-scale values through Froude number
modeling. At the full-scale 3.0 m/sec (6 knot) design point with the boom
pivoted at the sump inlet, the rate of change of vertical force with vert-
ical pivot point position is 161 N/cm per meter of boom width (28 poinds/inch
per foot of boom width). The front flotation system must have at least this
amount of stiffness for stable operation. This compares with a flotation
stiffness requirement of 2,900 N/cm per meter width (500 pounds/inch per
foot width) without a stability control system. Stiffness requirements in-
creased as the pivot point was moved aft as expected. The full-scale hydro-
dynamic drag was 14.6 per meter width (86 pounds per foot width) or a drag
coefficient based on frontal area of 0.39. As a comparison, a conventional
oil boom has a drag coefficient of about 1.50.
59
-------
SECTION 7
SYSTEM INTEGRATION TESTS
The previously developed bow, sump and stability control system were
combined in a full-scale 1.5 m (5 feet) wide streamlined oil boom preproto-
type model for a three-week system intearation test series at OHMSETT. The
objectives of these tests were to tune the system for optimum operation and
determine the overall system performance in retaining and recovering oil
slicks at speeds up to 3.0 m/sec (6 knots) and conditions typical of inland
waterways.
MODEL CONSTRUCTION AND TEST RIGGING
Figure 28 is a photograph of the completed model and Figure 20 (page 43)
is a side view drawing. The model was constructed as a two-dimentional
1.5 m (5 feet) wide section held together with vertical side plates. The
materials of construction were chosen for low cost and ease of fabrication
rather than desirability for long term immersion in a marine environment.
The front hydrofoil and sump inlet were constructed of fiberglassed foam,
similar to surfboards, and the remainder of the model was primarily 1..9 cm
(0.75 inch) thick exterior and marine grade plywood coated with resin.
One of the side plates was made of transparent acrylic plastic in the aft
sump area to permit observation of oil movement with underwater video.
The inclining sump inlet was designed for operation with a hydraulic
jump as the basic energy dissipation mechanism and the other mechanism
(arrays of vertical and inclined rods and the porous plate assembly) were
constructed as bolt-on modifications.
The stability control system utilized a cable and pulley system to
control gate position allowing the gain to be adjusted in three increments
(1.2/1, 2.5/1 and 2.8/1) and the gate position to be remotely adjusted
during testing. The float was a fiberglassed polyurethane foam module with
approximately 2,670 N (600 pounds) buoyancy and 91 N/cm (52 poinds/inch)
stiffness. It was made intentionally oversize to reduce hysteresis due to
friction in the cable mechanism at the highest gain setting.
The model was supported from a free floating catamaran work platform
and pivoted above the sump inlet.* The catamaran work platform was con-
structed from two 34-feet Sea and Air Products, Inc. Flattop Catamaran Hulls
Provision was made for pivoting 76 cm (30 inches) aft of the sump inlet
as well, but this arrangement was not tested.
60
-------
Figure 28. Streamlined oil boom model for System Integration Tests
-------
connected together with four wide flange steel beams and decked with 1.9 cm
(0.75 inch) thick plywood to form a U-shaped 6.1 m (20 feet) wide working
platform. The two hulls provided ample buoyancy and 1,300 N/cm (750 pounds/
inch)flotation stiffness. Since the minimum required for a 1.5 m (5 feet)
wide streamlined oil boom module was calculated to be 242 N/cm (140 pounds/
inch), these hulls provide a factor of safety greater than 5.
The streamlined oil boom module and catamaran work platform together
form a high-speed skimming craft. Figures 29 and 30 show how it was rigged
for testing in the OHMSETT channel. The catamaran was towed between the
main bridge and truss near one side of the channel with the video truss in
the center so oil flow in the sump could be observed through the clear
plastic sideplate. Oil was distributed approximately 15 m (50 feet) ahead of
the model through a distribution manifold suspended 0.6 m (2 feet) above the
water surface and cascaded onto the surface over an inclined plywood sheet.
For the majority of tests the oil slick was constrained by trailing ropes
into a 1.2 m (4 feet) wide surface slick thus ensuring essentially 100 per-
cent encounter efficiency.* On two tests the ropes were removed and re-
placed by diversionary boom stretched from the oil distribution manifold to
the streamlined oil boom side plates. The diversionary booms were of the
conventional type with a 30 cm (1 foot) draft and were inclined at 6 degree
angles to the tow direction to divert a 3.7 m (12 feet) wide oil slick into
the streamlined oil boom working section.
Oil was recovered from the secondary sump through two 7.5 cm (3 inch)
hoses feeding two air operated double diaphragm pumps mounted on the cata-
maran. The pump discharges were manifolded together and pumped through a
10 cm (4 inch) discharge hose to a series of 378 1 (100 gallon) recovery
barrels on the rear truss.
TEST PROCEDURES
Three types of tests were run over a range of operating conditions to
adjust hydrodynamic parameters and measure the system's performance in
retaining and recovering oil:
• Stability tests
t Holding tests
• System tests
The following is a brief description of the purposes and procedures
used for these tests. The detailed test procedures are shown in Appendix B.
Encounter efficiency is the ratio of the oil volume encountering the
streamlined oil boom active section to the oil volume distributed
upstream.
62
-------
SOUTH
OIL
DISTRIBUTION
MANIFOLD
UNDERWATER
VIDEO
RECORDER
. .TRAILING ROPZS
!*•—TFOR 1.2
OIL DISTRIBUTION
PUMPS AND
FLOWMETERS
DOUBLE
DIAPHRAM
AIR OPERATED
,. RECOVERY
PUMPS
PUMP CONTROL VALVES
OIL
RECOVERY •
HOSES (2)
ooooooooo
FIRE
HOSE
PUW
ACCESS
LADDER
378 1 (100 gal)
OIL RECOVERY BARRELS
Figure 29. Plan view of OHMSETT testing arrangement - 1976.
63
-------
fcJtL?? ft 'rj -i^
•• *'- 1* ^
Figure 30. Streamlined oil boom in system test at OHMSETT.
-------
Stability Tests
Stability tests were run without oil to monitor the operation of the
stability control system and adjust control system variables for optimum
stability. Although the stability control system had been previously tested
on 1/3-scale, additional full-scale tests were required to tune operation
to the OHMSETT test conditions and provide full-scale stability data. The
variables adjusted included:
t Freeboard at sump inlet
• Gate position
• Gain
• Damping
t Aft buoyancy
The freeboard at the sump inlet and sump depth were remotely adjustable
during the tests. The freeboard was fixed by the position of an electrically
powered linear actuator controlled by a catamaran mounted switch. The sump
depth was controlled by the position of the sliding gate regulating flow out
of the plenum. The Tenth of the cable controlling the gate motion was
adjustable through a winch mechanism mounted on the catamaran deck. These
two remote adjustments greatly simplified the stability testing procedure
since the proper adjustments could be made during the tests rather than
setting the adjustments and conducting a series of tests by trial and error.
The gain, damping and aft buoyancy were not remotely adjustable and were set
at the desired values prior to each test.
The stability test procedure allowed the operation of the stability
control system to be observed and the hydrodynamic parameters to be measured
under static and dynamic conditions. The control system variables were first
adjusted to the desired values under static conditions and the water levels
at various stations on the model were recorded. The catamaran was then
accelerated to test speed and the operation of the control system was
observed. The freeboard and sump depths were adjusted as required and when
a stable equilibrium was achieved the water levels, model trim and velocities
in the sump and plenum exit were recorded. The model was then decelerated
to a stop and returned to its initial position.
Comparison of static and dynamic measurement data allowed the effects
of each variable to be determined and the system to be adjusted for optimum
operation. Twenty-nine stability tests were run and best results were
achieved with the following adjustments:
65
-------
• Freeboard at sump inlet - adjusted to achieve a
3.8 cm (1.5 inch) layer entering the sump
• Gate position - adjusted to achieve a 38 cm
(15 inch) sump depth
• Gain - 3.8/1.0
• Damping - none
• Aft Buoyancy - 1,600 N (360 pounds)
The freeboard and gate position for optimum operation were found to be
functions of test speed. The adjustments for best results are shown in
Table 3. The highest available gain setting 3.8/1 and zero damping allowed
rapid changes in gate position in response to changing sump depth. The sump
depth was maintained constant in most tests within +2.5 cm (+1.0 inch) of
the desired value with the gate making exursions of approximately +10 cm
(+.4.0 inches).
Aft buoyancy was used to improve the model's trim. At 3.0 m/sec
(6 knots), the surface disturbances produced by the streamlined oil boom
bow have primary wavelenghts approximately equal to the model's length.
This causes the aft portion of the model to ride in a trough and gives the
model a positive angle of attack. Adding flotation to the aft portion of
the model partially counteracts this trim change.
Oil Tests
The oil tests measured the streamlined oil boom's performance in retain-
ing and recovering oil slicks. Figure 31 shows how oil flows through the
streamlined oil boom and identifies the oil loss mechanisms. A portion of
an oil slick approaching the streamlined oil boom could be lost around the
sideplates. This loss mechanism was reduced essentially to zero in the
OHMSETT calm water tests by directing the oil slick between the side plates
with trailing ropes. A portion of the oil slick passing between the side
plates could also be lost under the sump inlet. Underwater observation at
OHMSETT and previous test series confirmed that this loss mechanism was
essentially zero under calm water conditions with free surface variations
less than 10 cm (4 inches) trough to crest.
The oil and water entering the sump passes through an energy dissipat-
ion mechanism where the flow is slowed to 10 percent of its initial velocity.
In the process, a portion of the oil slick is mixed with the water in the
form of droplets. In the gravity oil/water separation portion of the sump
many of these droplets rise to form a reconstituted oil slick which is
retained in the sump. Those droplets not rising into the reconstituted
slick are lost out the porous bottom. A portion of the surface oil slick
may also be lost prior to its recovery by the weir skimmer.
66
-------
TABLE 3. FREEBOARD AND GATE POSITION ADJUSTMENTS
FOR OPTIMUM PERFORMANCE*
Variable
Freeboard at
sump inlet
Sump depth** for
zero slot width
Test Velocity
2.0 m/sec
(4 knots)
+1.3 cm
(+0.5 in.)
38 cm
(15 in.)
3.0 m/sec
(6 Knots)
+2.5 cm
(+1.0 in.)
43 cm
(17 in.)
* These adjustments give approximately 3.8 cm
(1.5 in.) layer skimmed into the sump measured
at the sump inlet and 38 cm (15 in.) sump depth
measured at the weir skimmer.
** Sump depth measured at center of stability control
system float.
67
-------
BOW AND SUMP INLET
I
ENERGY DISSIPATION
MECHANISM
i
GRAVITY OIL-WATER
SEPARATION
I
SURFACE OIL SLICK
I
WEIR SKIMMER OR OTHER
RECOVERY DEVICE
LOSSES
ENCOUNTER LOSSES OIL
LOST AROUND SIDES
OIL PASSING UNDER SUMP
INLET
OIL DROPLETS NOT
FORMING'SURFACE SLICK
LOST OUT SLOT
OIL DROPLETS LOST FROM
SURFACE SLICK
/RECOVERED A
I OIL SLICK J
Figure 31. Oil flow through streamlined oil boom system.
68
-------
Losses from the oil slick in the sump were measured in the holding
tests. A measured volume of oil was deposited directly into the sump with
the model stationary. The model wasJAen accelerated to test speed, towed
the length of the channel and decelerated to a stop. The recovery pumps
were then started and the oil retained in the sump (and some water) was
pumped into recovery barrels. The volume of the recovered oil was measured,
corrected for water content and compared to the initial preload to determine
the magnitude of the losses.
_Systems tests measured the performance of all portions of the streamlin-
ed oil boom operating as a system recovering oil slicks at high speed. To
reduce starting transients and simulate equilibrium conditions the sump was
preloaded with oil similar to the holding tests. The model was accelerated
to test speed. Then an oil slick was distributed ahead of the model and the
recovery pumps were operated simultaneously. Oil was distributed for a pre-
determined time period corresponding to an oil slick approximately 100 m
(325 feet) long and the recovery pumps were operated until the last of the oil
slick entered the sump. The model was then decelerated to a stop. The total
volume of oil and water pumped into the recovery barrels during the test was
measured and divided by the recovery pump operation time to determine the
recovery flowrate. Next any oil remaining in the weir box, hoses and pumps
was flushed into the recoery barrels. The volume of the recovered oil was
measured and corrected for water content. The recovery efficiency was calcu-
lated as the ratio of the oil volume recovered to the total oil and water vol-
ume pumped into the recovery barrels during the run. The oil volume retained
in the sump was also measured as in the holding tests to complete the oil in-
ventory. Tests where the initial and final sump oil volumes were essentially
equal were considered to be equilibrium.
69
-------
TEST RESULTS
Test Conditions
Following a series of stability tests where the system was adjusted
for optimum hydrodynamic performance 23 oil tests were run with two test
oil, calm and random wave conditions and at two taw speeds.
The test oil characteristics are listed in Table 4. The lube oil was
used for the majority of the tests and was similar to a 30 wt non-detergent
lubricating oil with viscosity similar to number two fuel oil and interfa-
cial surface tension modified with a surfactant to approximate number two
fuel oil.
Tests were conducted in both ca]m and "1 foot harbor chop" surface
conditions. During the calm water tests the OHMSETT wave generator was not
operated and the only surface disturbances were those caused by wind.
Under some conditions, these waves approached 5 cm (2 inches) trough to
crest.* The 1 foot harbor chop is a standard wave pattern used to test
oil spill control equipment at OHMSETT. Waves are produced by two hydrau-
lically powered wave flaps spanning at the south end of the tank. The
flaps are pivoted at the bottom and extend well above the water line. The
1 foot harbor chop was produced by driving these flaps out of phase at
35 rpm with a 3.8 cm (1.5 inch) stroke. The wave generator was operated
for 10 minutes prior to the test to allow a semi-random confused wave state
to build up.
The characteristics of the 1 foot harbor chop were measured by the
OHMSETT facility operator prior to the Streamlined Oil Boom Tests. A wave
probe was mounted in approximately the center of the tank and the free
surface oscillations were recorded on an analog tape recorder. The analog
TABLE 4- TEST OIL PROPERTIES
Parameter
Viscosity
Specific gravity
Surface tension
Interfacial surface
tension
Water and sediment
Lube
Oil
144.8
0.871
31.5
11.8
0.1
#2
Simulant
32.5
0.850
30.2
10.3
0.1
Units
centi stokes
@ 70°F
relative to water
dynes/cm
dynes/cm
percent
Estimated.
70
-------
WefriSJJnlay1e-d -ba(;k °n -a Precis1on Instrument Tape Recorder, the signal was
filtered to eliminate noise and the data was processed digitally. The re-
sults of the analysis on six data runs are shown in Table 5 and show the
average wave height to be approximately 15 cm (6 inches) with the average
1/3 highest and average VO highest wave heights approximately 23 and 30 cm
(9 and 12 inches) respectively.
During the streamlined oil boom wave tests with the 1 foot harbor chop,
the characteristics of the waves were observed to vary with position in the
tank. Near the north end of the tank (opposite end from the wave generator)
nearly standing waves with trough to crest heights of approximately 0.6 m
(2 feet) were observed with substantially smaller waves along the tank cen-
terline a few meters from the north end. Consequently, the wave data in
Table 5 should be considered only a rough indication of the 1 foot harbor
chop wave environment.
Two towing speeds were used - 3.0 and 2.0 m/sec (6 and 4 knots). Three
m/sec (6 knots) was the highest tow speed available at OHMSETT and was used
for a majority of the tests. The 203 m (667 feet) OHMSETT tank length
allowed approximately 30 seconds constant speed running time at 3.0 m/sec
(6 knots) and 45 seconds constant speed running time at 2.0 m/sec (4 knots)
after allowing for acceleration, deceleration and the space between the two
bridges. Two m/sec (4 knots) was chosen as the slower test speed because
the dynamic pressure at 2.0 m/sec (4 knots) is roughly half of the dynamic
pressure at 3.0 m/sec (6 knots).
TABLE 5. AVERAGE WAVE HEIGHTS IN
OHMSETT'S "1.0 FOOT HARBOR CHOP"
Wave Heights in Inches
Run
K-18
K-19
K-22
L-25
L-28
K-40
Average
Standard Deviation
Average
5.691
6.613
5.129
5.977
6.659
5.807
5.904
0.509
1/3
Highest
8.828
9.659
8.195
9.598
10.562
8.904
9.291
0.826
1/10
Highest
10.959
12.059
10.626
12.345
13.091
11.359
11.740
0.926
NOTE: Data from OHMSETT Facility operated - Mason and Hanger -
Silas Mason Co., Inc.
71
-------
Efficiency Calculations
The streamlined oil boom's performance in these oil tests may be des-
cribed in terms of thruput and recovery efficiencies. Thruput efficiency is
an index of the overall performances and is defined as the ratio of oil
volume recovered to oil volume distributed. In the holding tests the thru-
put efficiency 77- is defined as:
n = L (expressed as percent)
where:
VF = final oil volume contained in sump and weir box at the end of the
test
Vp = preloaded oil volume
This thruput efficiciency is a measure of the streamlined oil boom's per-
formance in retaining oil in the sump. If V-r = 100 all of the oil initially
preloaded into the sump was retained and if i}= 0 it was all lost.
In the system's tests, the thruput efficiency must be corrected to
account for changes in sump oil volume as follows:
„ _ R ' F P (expressed as percent)
where:
VR = volume of oil recovered over the weir during the test
Vs = volume of oil slick distributed.
If *?y = 100, no oil was lost; all was either recovered over the weir or
retained in the sump. If Vp = Vp the test operated essentially at equil-
ibrium and the thruput efficiency reduces to the ratio of oil volume re-
covered to oil volume distributed. Thruput efficiencies less than zero are
also possible in cases where the sum of the final sump oil volume and volume
of oil recovered over the weir is less than the initial preload.
Recovery efficiency V% is defined as the concentration of oil in the
recovered oil /water mixture:
. . v
Vj (expressed as percent)
72
-------
where:
VT = the total volume of oil and water recovered.
Recovery efficiency was calculated for the system tests only.
Appendix C discusses the oil handling procedures used in these tests
and the probable experimental error associated with oil volume measurements.
Based on Appendix C, the calculated values of thruput and recovery efficien-
cies are probably within a range of +15 to -10 percent. The range of exper-
imental error is larger on the plus side since some losses of recovered oil
before measurement were inevitable.
Test Data
The numerical results from the 23 oil tests are shown in Table 6. Over-
all the tests demonstrated good performance at both 2.0 and 3.0 m/sec (4 and
6 knots) testing speeds under calm water conditions. In the system tests '
thruput efficiencies as high as 99+ percent and 87 percent were observed at
2.0 m/sec (4 knots) and 3.0 m/sec (6 knots) repectively and losses in the
holding tests were minimal. As expected, thruput efficiencies degraded in
waves but in holding tests a high percentage of the initial preload was
retained.
The first six tests compared the performance of three energy dissipation
options under similar 3.0 m/sec (6 knots) test conditions with lube oil. A
holding test with a 283 1 (75 gallon) preload and a system test with a 151 1
(40 gallon) preload and a 2.5 mm thick oil slick were run on each on each
option.* The results shown in Table 6 demonstrate that each design performed
well with thruput efficiencies for both holding and system tests in the range
of 80 to 88 percent. Oil losses in the holding tests appeared to be caused
by starting and stopping transients rather than losses during the constant
speed portion of the test. During acceleration, the oil slick sloshed to-
ward the aft portion of the sump with some flowing into the float chamber and
some undoubtedly lost out the porous bottom. During deceleration, the oil
slick sloshed forward toward the sump inlet. Although a pivoting door was
deployed across the sump inlet during deceleration, at the end of each test
some oil was usually observed to escape. These losses could easily total 38
to 57 1 (10 to 15 gallons), and be responsible for the entire losses observed
in the holding tests. If the oil lost in the holding tests is assumed to be
due entirely to transient phenomena and the thruput efficiencies on the thru-
put tests are corrected for these losses, the thruput efficiences for system
tests 4, 5 and 6 would be 96.1, 98.6 and 99.6 percent respectively. Since
each energy dissipation mechanism performed equally well and since the ex-
panded metal incline was expected to perform best in other tests and to be
relatively free fromclogging problems, itwas used for the remainder of the tests.
Tests 6-9 are 3.0 m/sec (6 knot) calm surface systems tests with the
expanded metal incline over a range of slick thicknesses with lube oil.
Tests 20-22 are similar tests with simulated number two oil. The thruput and
* Test input parameters were maintained at approximately these values -
variations were due to limitations of the testing procedure.
73
-------
^^^^^HHMHI^^^
Test Conditions
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Test1
Type
Holding6
Holding?
Holding
System6
System7
System
System
System
System
System8
System8
Holding
System
System
Holding
System
System
Holding
System
System
System
Sys tern
System
on2
Type
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
Lube
#2
12
12
n
Tow
Speed
Knots
6
6
6
6
6
6
6
6
6
6
6
6
6
6
4
4
4
4
4
6
6
6
4
Waves3
Calm
Calm
Calm
Calm
Calm
Calm
Calm
Calm
Calm
Calm
Calm
r H.C.
1' H.C.
1' H.C.
Calm
Calm
Calm
r H.C.
1' H.C.
Calm
Calm
Calm
Calm
Oil Input
Preload
gal.
75.0
75.4
74.9
40.0
40.0
40.5
40.0
40.0
75.2
40.0
40.0
74.9
40.0
40.0
74.9
40.0
40.5
74.9
40.2
40.0
39.5
74.9
41.2
Slick
Thickness
run
2.8
2.8
2.5
I.I
3.7
2.6
0.9
2.6
0.9
2.8
1.4
4.4
4.6
0.9
3.2
5.0
1.6
Volume
gal.
101.1
99.9
89.4
41.5
156.8
101.9
36.5
98.9
31.0
101.3
46.5
147.9
150.9
33.0
115.9
188.8
52.2
Distribution
Rate
gal.
200.8
198.5
176.5
81.8
259.9
202.1
72.4
197.8
60.5
200.6
69.2
220.1
101.1
65.6
231.8
375.7
78.0
— •" — .^ pw^»w«^^B^^^B^^^^^
•^ WBBM •W__^^^^_MM^_^^_^^^^_^»^HV
01 1 Recovery
During Run
Recovery
Rate
gpm
175.4
178.2
128.4
80.5
237.4
249.4
97.9
227.4
88.8
220.3
91.4
239.5
249.3
94.2
266.9
294.2
90.0
Oil
oal.
59.1
54.2
40.3
22.8
91.0
83.4
21.7
52.5
2.5
31.3
41.2
91.4
44.6
23.6
54.8
91.3
28.9
Total
gal.
95.0
98.0
68.0
44.0
158.2
133.0
54.5
133.0
51.5
128.5
71.0
196.0
182.0
51.0
145.0
153.0
65.0
Final
Sump
gal.
65.3
64.1
60.2
68.4
73.0
74.5
49.1
45.7
71.4
24.8
28.3
51.0
10.3
21.7
66.0
46.5
70.3
52.8
26.8
33.9
61.7
90.0
52.6
^•^^^^^•^^•^^^^•^••••^^^•••^^^^^•••••a
^ ^mn^mfn.
Efficiencies
Recovery
%
62.2
55.3
59.3
41.8
58.0
62.7
39.8
39.5
4.9
24.4
58.0
46.6
24.5
46.3
37.8
59.7
44.5
Thruput
%
87.1
85.0
80.3
86.5
87.3
83.1
76.9
62.2
78.1
17.. 8
41.3
68.1
neg.
12.8
88.1
99+
81.9
70.5
20.7
53.0
66.4
56.4
77.2
1. Inclined porous plate energy dissipation method unless otherwise stated.
2. Average oil characteristics - lube: viscosity = 145 centistoke ? 70°F, specific gravity = 0.87, simulated 02
(a lube oil with surfactant added to approximate #2 fuel oil): viscosity = 32 centistoke 9 70°F, specific
gravity - 0.85.
3. Surface condition - calm no waves generated some wind induced surface chop estimated at up to 2 inch trough
to crest, 1' H.C. = 1.0 feet harbor chop semi-random wave condition.
4. Recovery efficiency » oil volume recovered/ total volume recovered.
5. Thruput efficiency for system tests = oil volume recovered during run plus final sump oil volume minus oil
preload volume divided by oil volume distributed. Thruput efficiency for holding tests = final sump oil
volume divided by oil preload volume.
6. Energy dissipation method: 45" array of rods.
7. Energy dissipation method: 90° array of rods.
8. With diversionary boom (see text).
DO
m
CT>
O
m
oo
m
—1
oo
oo
CD
OO
•ya
m
oo
oo
-------
recovery efficiencies for these tests are plotted against slick thickness in
Figure 32. Thruput efficiencies with the higher viscosity lube oil were
generally superior - to the simulated number two oil. The higher viscosity
oil offers more resistance to shearing in the turbulent energy dissipation
process producing larger oil droplets which can be recovered more easily in
the gravity separation portion of the sump. Thruput and recovery efficien-
cies were relatively insensitive to slick thickness. The limited number of
tests and'the probable experimental error range of +15 and -10 percent make
if difficult to accurately assess the effects of slick thickness. The 1.0
to 5.0 mm oil slick thickness corresponds to 2.6 to 13.0 percent of the to-
tal sump flowrate, at most a small percentage. Losses due to oil passing
under the sump inlet would not be expected under calm conditions unless this
percentage were near 100 percent.
Recovery efficiencies for these tests were in the range of 37 to 63
percent. Thus the flow passing over the weir was roughly half oil and half
water. Recovery efficiency should increase with the volume of oil contained
in the sump (the oil layer thickness) and decrease as recovery flowrate in-
creases. However, the range of values covered in the experiments is not
large enough to demonstrate these relationships.
Tests numbers 10 and 11 were 3.0 m/sec (6 knot) calm surface system
tests where a 3.7 m (12 feet) wide lube oil slick was diverted into the
streamlined oil boom active section with conventional diversionary boom
deployed at slight angles to the tow direction. The diversionary angle was
chosen as 5.7 degrees so that the component of velocity normal to the div-
ertionary boom was0.3m/sec (0.6 knots). Since conventional oil booms oper-
ate satisfactorily with normal components of velocity up to this value, the
diversionary boom was expected to divert the oil slick into .the streamlined
oil boom without the usual droplet entrainment and drainage failure mecha-
nisms.
The flow patterns observed in these tests are shown in Figure 33. The
leading edge of each diversionary boom produced a substantial turbulent
breaking bow wave. The bow waves between the booms intersected approximately
midway along the boom lenth and reflected from the opposite sides about 2 m
(7 feet) upstream of the bow. The resulting diamond shaped wave patterns
significantly affected the oil slick. The overall result was a channeling of
the slick between the streamlined oil boom sideplates with no visible losses
due to diversionary boom failure. However, the breaking waves also entrained
a substantial portion of the slick beneath the surface and produced a frothy
turbulent flow at the sump inlet. This accounts for the significantly lower
thruput efficiencies achieved in tests 10 and 11. Comparing these results to
similar 3.0 m/sec (6 knot) calm surface tests where na oil diversion was in-
volved shows that thruput efficiency dropped from 83 to 41 percent for a
2.5 mm slick and from 77 to 18 percent for a 1 mm slick.
Tests 15-17 and 23 are 2.0 m/sec (4 knot) calm surface tests and demon-
strate improved performance over the 3.0 m/sec (6 knot) tests. The kinetic
evergy dissipated in the sump at 2.0 m/sec (4 knots) is only half that dis-
sipated at 3.0 m/sec (6 knot) and as a result the breakup of the oil slick
into small droplets, the principle loss mechanism was significantly reduced
75
-------
100
90
SO
70
60
50
40
30
20
10
0
THRUPUT
EFFICIENCY
PERCENT
D
O
O
D
D
SLICK
THICKNESS
100
90
GO
70
60
50
40
30
20
10 -
0
RECOVERY
;EFFICIENCY
PERCENT
0°
O LUBE OIL
D SIMULATED NO. 2 OIL
n
SLICK
THICKNESS
irm
Figure 32. System Tests Results - Calm Surface at 3.0 m/sec (6 knots)
76
-------
OIL DISTRIBUTION MANIFOLD
DIVERSIONARY
BOOM BOW
OUTSIDE BOW
WAVE CREST
OIL SLICK
DIVERSIONARY
BOOM
INSIDE BOW
WAVE CREST
BOW WAVE
INTERSECTION
BOW WAVE
REFLECTION
POINT
STREAMLINED OIL
BOOM SIDEPLATES
SUMP INLET
Figure 33. System Tests with Diversionary Boom - Flow Pattern
77
-------
resulting in higher thruput efficiencies. Figure 34 shows the thruput and
recovery efficiencies plotted against slick thickness and can be compared to
the similar 3.0 m/sec (6 knot) plots in Figure 32. Again, there is insuff-
icient data to determine a trend but the efficiencies at 2.0 m/sec (4 knot)
thruput are clearly higher than at 3.0 m/sec (6 knots). :
The streamlined oil boom was designed to operate in calm and relatively
low wave conditions typical of small rivers, lakes and other inland water-
ways where the maximum free surface excursions do no exceed about 15 cm
(6 inches) trough to crest. Ideally, the wave tests at OHMSETT should have
included several wave states both below and above this- threshold. Unfortu-
nately >the smallest random wave state available at OHMSETT without a lengthy
calibration procedure was the 1 foot harbor chop with an average 1/10 high-
est wave height of 30 cm (1 foot) though to crest. Since the streamlined
oil boom relies on skimming a thin layer of oil and water into the sump,
performance in this wave state was expected to be poor.
Tests 12 to 14 were the 3.0 m/sec (6 knot) wave tests and tests 18 and
19 were the 2.0 m/sec (4 knot) wave tests. Comparing these data with sim-
ilar calm water data shows that thruput efficiencies were reduced from calm
surface values of 60-100 percent to 0-20 percent. Some of the losses were
due to oil passing around the side plates since the slick was not effectively
contained by the trailing ropes. Other losses occured as the wave trough
(and some oil) passed under the sump inlet and also due to wave action in
the sump. The impact of the wave crests passing into the sump caused sub-
stantial splashing and agitation in the sump. In test 13 one large wave
swamped the sump shortly after the test begain, causing the entire sump pre-
load to be lost.
In summary, performance degraded significantly in the wave tests.
However, this wave environment was much more severe than the maximum 15 cm
(6 inch) free surface excursions for which the streamlined oil boom was de-
signed. The reduction in performance in these tests demonstrates that the
streamlined oil boom is not suitable for large harbors, bays and other areas
where theses larger waves occur.
78
-------
100
90
GO
70
SO
50
40
30
20 -
10 -
0
THRUPUT
-EFFICIENCY
PERCENT
SLICK
THICKNESS
mm
0
100
90
30
70
60
50 -
40 -
30 -
20 -
10 -
0
RECOVERY
LEFFICIENCY
PERCENT
D
O LUBE OIL
D SIMULATED-NO. 2 OIL
SLICK
THICKNESS
rrm
Figure 34. System Tests Results - Calm Water at 2.0 m/sec (4 knots)
79
-------
REFERENCES
1. Dorrler, J. Stephen, Ray Ayers, and David C. Wooten 1975. High Current
Control of Floating Oil, 1975 Conference on Prevention and Control of
Oil Pollution, March 25-27, 1975.
2. Folsom, B.A. and C. Johnson 1977. Development of a High Current Stream-
lined Oil Boom/Skimmer for Inland Waterways, 1977 oil Spill Conference
Proceedings, March 5-10, 1977.
3. Wooten, David C. 1973. Mechanical Control of Oil Spills Utilizing a
Streamlined Boom, 1973 Conference on Prevention and Cortrol of Oil
Spills, March 13-15, 1973.
4. Wooten, D.C. and B.A. Folsom 1975. "Bow Wave Hydrodynamics for a
Slightly Submerged Hydrofoil at Low Froude Numbers", presented at
American Society of Civil Engineers, Hydraulics Division, 23rd Annual
Hydraulics Division Specialty Conference, Seattle, Washington,
August 6-8, 1975.
80
-------
APPENDIX A
PHENOMENOLOGICAL BOW WAVE THEORY
This appendix discusses a phenomenological theory describing flow over
the upper surface of a slightly submerged hydrofoil at low Froude number with
smooth and turbulent bow waves where both inertia! and gravity forces are
important. The theory is phenomenological in that it does not predict flow
conditions per se but merely assumes that the flow behaves with the same
phenomena as open channel flow over a complex contour.
The flow passing over the hydrofoil was assumed to be totally separated
from flow passing beneath the hydrofoil by an impermeable frictionless boun-
dary consisting of the stagnation streamline and the upper surface of the
hydrofoil. Flow over the contour was assumed to meet the requirements of
frictionless open channel flow and the equation of continuity, energy and
momentum were applied to determine the location of the stagnation streamline
and free surface at several well defined stations.
The following assumptions were made to simplify the analysis:
0 Two-dimensional flow
• Neglect surface tension forces
• Neglect viscous forces
• Assume the velocity to be horizontal everywhere
• Assume all energy loss occurs in a hydraulic jump of infinitesimal
length
The two-dimensional flow assumption requires that the hydrofoil shape be
the same at every section and that the velocity of the flow approaching the
hydrofoil be uniformly constant.
Neglecting surface tension effects is a reasonable assumption for the
streamlined oil boom application since they are responsible for capillary
waves and other small scale effects not important here.
Neglecting viscous forces is justified if the boundary layer thickness
produced by these forces is small compared to the total flow depth passing
over the hydrofoil. Boundary layer thickness starts at zero where the stag-
nation streamline intersects the hydrofoil and increases as flow moves down-
stream. Since the flow is accelerating up to the top of the hydrofoil, the
boundary layer is thin in this region but increases in thickness downstream.
81
-------
Although the velocity is clearly not horizontal everywhere, the impor-
tant factors for this analysis are the reduction in dynamic pressure due to
the neglected vertical velocity component and the pressures produced by
curving streamlines. Since the error in dynamic pressure introduced by neg-
lecting the vertical component is proportional to the square of the sine of
the inclination of the total velocity vector from the horizontal, an 18°
inclination will reduce the dynamic pressure by less than 10 percent. The
pressure produced by curving streamlines is a function of the radius of
curvature and is small when the radius of curvature is large compared to the
thickness of flow passing over the hydrofoil.
Flow over a hydrofoil with a turbulent bow wave modeled by a hydraulic
jump is shown in Figure A-l. A Froude number* based on the local flow con-
ditions can be defined at several stations:
Vi
Fi =
where Fj, V-j and Yj are the local values of Froude number, velocity and flow
depth respectively and g is the acceleration of gravity. Station 1 is the
undisturbed flow far upstream and consequently V] is the freestream velocity
and YI is the thickness of the layer skimmed over the hydrofoil. A hydrau-
lic jump modeling a turbulent bow wave is assumed to form over the horizon-
tal stagnation streamline with Station 1 upstream and Station 2 downstream.
The characteristics of a hydraulic jump require that the local Froude number
upstream be greater than 1.0 (supercritical) and downstream be less than 1.0
(subcritical).
Downstream from Station 2 the flow accelerates continuously and this
requires that the Froude number at Station 3, the highest point on the hydro-
foil, be equal to 1.0 and the Froude number downstream, Station 4, be greater
than 1.0 (supercritical).**
With these assumptions, the flow conditions at stations 1 through 3 can
be calculated by applying the usual equations of open channel flow:
(1) Continuity: V1 Y] = V2 Y2 = V3 Y3 =
(2) Hydraulic Jump: Y2/Y] =
v2
1 + 1
Y '
1/2
- 1
**
Froude number is a dimensionless flow parameter expressing (the square
root of) the ratio of inertia! to gravity forces.
This is equivalent to saying that the flow over the hydrofoil is not affected by
the downstream conditions, observed to be true in applications skimming thin
1ayers.
82
-------
FICTITIOUS BOTTOM FOR OPEN CHANNEL FLOW ANALYSIS
oo
CO
FREE SURFACE
STATION 4
DOWNSTREAM OF
STATION 3
SUPERCRITICAL FLOW
V,
STATION 3
POINT OF MAXIMUM
ELEVATION
CRITICAL FLOW
V
STATION 2
DOWNSTREAM OF JUMP
SUBCRITICAL FLOW
•» <- ^ 1 . A
HYDRAULIC JUMP
(ALL ENERGY
DISSIPATION
OCCURS HERE)
STATION 1
FAgUPSTREAM
SUPERCRITICAL FLOW
V,
Fl -
V/9Y7
ARBITRARY HYDROFOIL SHAPE
STAGNATION STREAMLINE
Figure A-l. Turbulent bow wave flow model.
-------
v 2 v,2
(3) Energy: ^- + Y2 + h2 = ^- + YS + h3
(4) Freeboard Definition: B = h3 - (Y1 + h^
*32
(5) Froude Number at Station 3: -£- = 1.0
9' o
where h is the height of the stagnation streamline or top of the hydrofoil
above an arbitrary horizontal datum line.
Equation 1 is the continuity equation for two-dimensional open channel
flow and Q is the flowrate per unit width. Equation 2 is the relationship
expressing the ratio of upstream and downstream flow depths for a hydraulic
jump over a level contour and may be derived from momentum and continuity
considerations. Equation 3 states that no energy is lost between stations
2 and 3 and Equation 4 defines the freeboard B as the height of the hydro-
foil's highest point (at Station 3) above the undisturbed free surface.
Equation 5 equates the local Froude number at Station 3 to 1.0 as discussed
above.
These five equations can be solved by eliminating the unknown conditions
at Station 2 and expressing the conditions at stations 1 and 3 in terms of
the freestream velocity V] and the freeboard B. The results identify a
dimension!ess independent parameter:
Gn =
which is the reciprocal of the Froude number squared based on freestream
ve"!ocity and freeboard. The reciprocal squared form is used so that positive,
zero and negative values of B may be considered without resorting to com-
plex or infinite terms. Two dimensionless dependent parameters describe
the resulting flow.
v,2
3 ?
V,2
These parameters are also the reciprocal of Froude number squared form. Both
utilize the freestream velocity V-] and the local values of flow depth at
stations 1 and 3.
84
-------
After some manipulation, the results can be expressed as follows:
(6)
2" +
(7).
G
(8) Q =
Flow over a hydrofoil with a smooth bow wave is shown in Figure A-2.
As in the turbulent flow model, this flow is analyzed by applying the equa-
tions of open channel flow to the flow above the stagnation streamline and
top side of the hydrofoil. However, since there is no energy dissipation,
station 2 is nonexistent and the five equations for the turbulent model re-
duce to:
(9) Continuity: V-, Y] = V3 YS = Q
w 2 2
(10) Energy: V1 + Y] + h] = V3 + Y3 + h3
2g 2g
(11) Freeboard Definition: B = h - (Y + h )
(12) Froude Number at Station 3:
= 1.0
Equation 12, equating the Froude number at station 3 to 1.0, is not a
necessary requirement for smooth flow over a hydrofoil. However, it can be
shown that this condition produces maximum flow over the hydrofoil and meas-
urements of flow depth and velocity have confirmed that the Froude number is
very close to 1.0 at Station 3 for shallow submergence. For very deep sub-
mergence*, Froude number is clearly not equal to 1.0 and instead the condi-
tion Y = -B is more appropriate.
Very deep submergence means that the thickness of the flow layer passing
over the hydrofoil is significantly greater than the hydrofoil thickness
85
-------
STATION 3
POINT OF MAXIMUM ELEVATION
FREE SURFACE
00
NOT NECESSARILY
TO 1.0
ARBITRARY HYDROFOIL SHAPE
SMOOTH FLOW
(NO JUMP AND
NO ENERGY
DISSIPATION)
STAGNATION
STREAMLINE
STATION 1
FAR UPSTREAM
Figure A-2. Fictitious Bottom Contour For Open Channel Flow Analysis - Smooth Bow Wave Flow Model
-------
Solving these equations in the same manner as the turbulent case gives
the following results:
(13) G3 = 1(1 - 2GB)
(14) 6, = (G3)3/2 = t(l-2GB)!3/2
V 2
The results of the smooth and turbulent analysis can be displayed in
graphical form. Figures A-3 and A-4 are plots of G3 versus GB and G] versus
63 respectively. These graphs may be interpreted as plots of Y$ and Y-|
against 8 at constant velocity.
The hydraulic jump is only possible for the range - 1.00.07, the flow does not contain enough energy to flow over the body
through a jump, and for GB<-!.O, the local Froude number upstream (at Sta-
tion 1) is subcritical precluding the existence of a jump. As GB—*-1.0,
the jump strength becomes smaller and the turbulent bow wave curves blend
into the smooth flow curves.
For .smooth flow the maximum value of GB is 0.50 and corresponds to the
point at which the energy of the flow is just great enought to lift the
water to the top of the body. These smooth flow curves represent solutions
to the equations with the Froude number at Station 3 equal to 1.0. With
Froude number at the top of the body less than 1.0, the value of 63 will be
greater than that plotted and similarly if the Froude number at the top of
the hydrofoil is greater than 1.0, the value of 63 will be less than that
plotted. For very deep submergence, there will be little disturbance to the
free surface, and 63 will be approximately equal to - GB . This deep sub-
mergence relationship is also shown in Figure A-3.
The phenomenological theory can be extended to include the three-dimen-
sional effects associated with non-uniform velocity profiles. Figure A-5
shows the typical three-dimensional phenomena observed when flow with a
non-uniform velocity profile approaches a two-dimensional hydrofoil. The
velocity profile shown with velocity defects near the end of the hydrofoil's
span is similar to that measured in the small scale experiments and was pro-
duced by hydrodynamic friction along the channel side plates. As flow passes
upward over the hydrofoil some of the kinetic energy, in the form of velocity
head, is converted to potential energy, in the form of the height of the water
surface. If the velocity profile is uniform all of the flow passes over the
model. However, if there is a boundary layer on the sidewall there will be
some region in which the velocity head is not great enough for flow to pass
over the top. Flow with higher velocity will then move into this area and
form a vortex and the flow in the center of the channel will diverge to the
full hydrofoil width.
Experiments with dye injection confirmed this behavior. Referring to
Figure A-5, when dye was injected near the surface between points A and B
the dye passed over the top of the hydrofoil. When dye was injected at any
depth outside this range it passed under the hydrofoil. The width W] between
points A and B was found to be a function of submergence. For deep submer-
87
-------
00
CO
THEORETICAL RANGE
WHERE HYDRAULIC
JUMP IS POSSIBLE
DEEP SUBMERGANCE
APPROXIMATION
(-B = Y
FLOW WITH
HYDRAULIC JUMP
-1.4
-1.2
0.4 0.5
H-
Figure A-3. Theoretical Correlation 63 vs. GB
-------
CO
1.4
1.2
RANGE WHERE
HYDRAULIC JUMP IS
THEORETICALLY
POSSIBLE
1.0
FLOW WITH
HYDRAULIC JUMP
SMOOTH FLOW °'8"'
WITH F3 = 1.0
(MAXIMUM Y1 CONDITION)
0.6-1
-1.4
Figure A-4. Theoretical Correlation 6-j vs.
-------
Stream]Ines
Velocity
Velocity
Profile
ID
O
Velocity Defect
Oue to Boundary
Layer Friction
Plan
View
Elevation
View
Figure
A-5. Three-Dimensional Flow Over Model
-------
gence, with little conversion from kinetic to potential energy, W-, was es-
sentially equal to the total hydrofoil width W3. As the freeboard was in-
creased the width Wi decreased and the corner vortices became more pronounc-
ed.
The phenomenological theory was extended to describe these three dimen-
sional effects by assuming that the flow passing over the hydrofoil was con-
tained in a diverging channel of width W-| upstream and W3 over the top of
the hydrofoil and downstream. The continuity equation for the turbulent and
smooth bow wave analyses were then modified as follows:
(16) Turbulent Bow Wave Continuity:
Vl Yl Wl = V2 Y2 Wl = V3 Y 3W3 =^3
(17) Smooth Bow Wave Continuity:
Vl Yl Wl = V3 Y3 W3 ' QW3
where Q is the flowrate per unit hydrofoil width.
The equations can be reduced in the same manner as the two-dimensional
analyses and give the same results if G] is generalized to Gc[* as follows:
G* =
' w1 '
and the flowrate Q becomes:
(19) Q = G,* ]_
The relationship between G3 and Gg remains unchanged independent of W-|/W3.
Thus if there are no three-dimensional effects, the results revert to the
two-dimensional case:
W1/W3 =1.0
Figure A-6 shows a comparison of data from the dolphin-nose hydrofoil
model with the phenomenological theory correlations neglecting three-dimen-
sional effects. As expected, the data points for gY3/V] 2 are close to the
theoretical model and the data points for gY-]/V-|2are substantially above
theoretical predictions due to three-dimensional effects.
The ratio Wi/W3 was measured for several data points by dye injection
and a* was calculated directly. Thin plastic sideplates were also attached
to the model to isolate the flow in the model's center (with uniform velo-
city) from the low velocity flow near the channel walls. Data points mea-
sured on this configuration and those corrected for vortices are shown in
Figure A- 7 where G*f is plotted against GB. The agreement is quite close.
91
-------
Turbulent
Bow Wave
Correlation
Theoretical
Maximum
Smooth Flow
5
Turbulent
Bow Wave
Correlation
-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Figure A-6. Dolphin Nose Model - Correlations
92
-------
_aL
-1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2
0.2 0.4
o Corrected for Vortices
°Sidewa11s W-|/W3 = 1.0 Assumed
Figure A-7. Dolphin Nose Model Data Corrected For
Vortices And Data Taken With Sidewalls
93
-------
APPENDIX B
DETAILED TEST PROCEDURES - OHMSETT SYSTEM INTEGRATION TESTS
Stability Test Procedure
(1) Test director obtains test number and parameters from Ultrasystetns
personnel and gives test conditions to control room operator.
(2) Recovered oil from previous test is decanted, sampled measured and
then discarded.
(3) ALL residual oil is cleaned from the channel.
(4) Catamaran personnel adjust boom parameters as required.
(5) The skimmer bar is raised.
(6) Test director verifies test conditions with Ultrasystems' personnel.
(7) Wave generator is started, if required.
(8) All test personnel assume their positions.
(a) Test director
(b) Photographer
(c) Control room operator
(d) Catamaran personnel
(e) Instrumentation specialist
(9) Test director gives three blasts on the air horn to signal ready for
test.
(10) Instrumentation specialist starts wave recorder.
(11) Test director gives one blast on the air horn to signal the start of
test.
(12) Test director checks with control room operator for obtaining test
speed.
(13) Test director signals when full test speed has been reached.
(14) Catamaran personnel adjust streamlined oil boom parameters if
required.
(15) Catamaran personnel record hydrodynamic data.
(16) When bridge nears end of channel, control room operator is notified
to bring bridge to a stop.
(17) Wave generator is stopped.
(18) Wave recorder is stopped.
94
-------
Stability Test Procedure (Continued)
(19) Catamaran personnel raise model stern for return trip.
(20) Bridge is returned to starting position.
(21) All test data is recorded
(22) System is prepared for next test.
Holding Test Procedure
(1) Test director obtains test number and parameters from UHrasystems
personnel and gives test conditions to control room operator.
(2) The recovered oil from previous test is decanted, sampled, measured
and then discarded.
(3) ALL residual oil is cleaned from the channel.
(4) Catamaran personnel adjust boom parameters as required.
(5) The oil preload to be used for the test is measured and deposited
in the primary sump.
(6) The skimmer bar is raised.
(7) Test director verifies test conditions with Ultrasystems personnel.
(8) Wave generator is started.
(9) All test personnel assume their positions.
(a) Test director
(b) Photographer
(c) Control room operator
(d) Catamaran personnel
(e) Instrumentation specialist
(10) Test director gives three blasts on the air horn to signal ready for
test.
(11) Instrumentation specialist starts videotape and wave recorder.
(12) Test director gives one blast on the air horn to signal the start
of test.
(13) Test director checks with control room operator for obtaining test
speed.
(14) Test director signals when full test speed has been reached
(15) Catamaran personnel adjust streamlined oil boom parameters if requir-
ed.
(16) Catamaran personnel record hydrodynamic data.
(17) When bridge nears end of channel, control room operator is notified
to bring bridge to a stop.
95
-------
Holding Test Procedure (Continued)
(18) As bridge slows to a stop, catamaran personnel lower front dam, hold
automatic weir upright and close rear slot as required to hold oil
sump.
(19) Wave generator is stopped.
(20) Videotape and wave recorder are, stopped.
(21) Skimmer bar is lowered.
(22) Firehose pumps and air compressor are started.
(23) Firehose and recovery pumps are operated as required to pump all oil
in main sump and weir into recovery barrels.
(24) Firehose and recovery pumps are stopped on Ultrasystems1 command.
(25) Catamaran crew raises model stern for the return trip.
(26) Bridge is returned to starting position skimming any oil not pre-
viously recovered.
(27) All test data is recorded on bridge data sheet.
(28) System is prepared for next test.
System Test Procedure
(1) Test director obtains test number and parameters from Ultrasystems
personnel and gives test conditions to:
(a) Control room operator
(b) Distribution pump operator
(2) The recovered oil from previous test is decanted, sampled, measured
and then discarded.
(3) ALL residual oil is cleaned from the channel.
(4) Catamaran personnel adjust boom parameters as required.
(5) Recovery pumps are operated pumping until hoses, etc. are fully fill-
ed with water.
(6) The oil preload to be used for the test is measured and deposited in
the primary sump.
(7) The oil distribution pumping rate, and the oil recovery pumping rate,
are calibrated as required for the test and the flowmeter reading is'
recorded.
(8) The skimmer bar is raised.
(9) The air compressor used to drive the recovery pumps is started.
(10) Test director verifies test conditions with Ultrasystems personnel.
(11) Wave generator is started.
(12) All test personnel assume their positions.
96
-------
(a) Test director
(b) Distribution pump operator
(c) Recovery pump operator
(d) Recovery tiose operator
(e) Photographer
(f) Control room operator
(g) Catamaran personnel
(h) Instrumentation specialist
(13) Test director gives three blasts on the air horn to signal ready for
test
(14) Instrumentation specialist starts videotape and wave recorder.
(15) Test director gives one blast on the air horn to signal the start
of test.
(16) Test director checks with control room operator for obtaining test
speed
(17) When test speed is obtained, test director signals to start oil dis-
tribution.
(18) Oil distribution operator distributes oil for time period specified
in testing matrix as measured with stopwatch.
(19) Oil distribution operator records final flowmeter reading and dis-
tribution time from stopwatch.
(20) Oil recovery pump operator starts recovery pumps and stopwatch when
first portion of deposited slick reaches sump inlet.
(21) When last of oil slick reach sump inlet:
(a) Oil recovery pump operator stops pumps and stopwatch and
records time.
(b) Test director starts timing post distribution running time.
(22) At end of post distribution running time, test director signals
control room operator to slowly stop bridge motion.
(23) As bridge slows to a stop, catamaran personnel lower front dam,
hold automatic weir upright and close rear slot as required to hold
oil in sump.
(24) Wave generator is stopped.
(25) Videotape and wave recorder are stoped.
(26) Recovery hose operator measures total volume recovered in collection
barrels.
(27) Skimmer is lowered.
(28) Firehose pumps are started.
(29) Firehose water is directed into weir box and recovery pumps are op-
erated as required to pump any oil remaining in weir box, hoses, etc.
into collection barrels while holding weir upright to prevent mixing
of oil in sump with oil in weir box.
97
-------
System Test Procedure (Continued)
(30) Upon UItrasysterns' command, recovery pumps are stopped and recovery
hose is transferred'to second set of recovery drums and suction hose
is used to clean inventory oil from sump.
(31) Firehose and recovery pumps are operated as required to pump all oil
in main sump into second set of recovery barrels.
(32) Firehose and recovery pumps are stopped on Ultrasystems' command and
suction hose to weir box is reconnected.
(33) Catamaran crew raises model stern for return trip.
(34) Bridge is returned to starting position skimming any oil not pre-
viously recovered.
(35) All test data is recorded on bridge data sheet.
(36) System is prepared for the next test.
98
-------
APPENDIX C
OIL HMDLIMG AND EXPERIMENTAL ERRORS - 0HUSETT
SYSTEMS IRIEGMTIOJI TESTS
This appendix describes the oil handling procedures used in perform-
ing the oil tests and measuring the input and recovered oil volumes. It
also assesses the probable source of inaccuracy and nonrepeatability due
to systematic and experimental errors as opposed to variations in the stream-
lined oil boom's performance.
Oil Handling Procedures
The oils used for testing the streamlined oil boom were part of
OffffSETT's standard oil stocks and had been used previously for testing other
oil spill control equipment at OHftSETT, Following its previous use,
it was skinned from the tank and allowed to settle for several days as a
preliminary oil/water separation. The emulsion on the surface of the tank
was then processed through a vacuum distillation unit where nearly all of
the emulsified water was renoved* and then transferred to storage.
At the beginning of each testing day or as required, oil was pimped
into the bridge storage tanks and a sample of the oil was withdrawn and
analyzed for specific gravity, viscosity, surface tension, interfacial sur-
face tension and water content. The values changed little from day to day
and Table 4 (page70 ) gives the average properties.
Handling of the oil inputs to the tests is shown schematically in
Figure C-l. All hoses, pumps, etc were completely filled with oil prior to
the start of the tests. The sump preload was punped from the bridge oil
tank through a positive displacement flowneter and a 5.0 cm (2 inch) dis-
charge hose directly into the sump. A ball valve on the dischage end of the
hose was used to naintain the hose full of oil at all times and eliminate
inaccuracies due to drainage and syphoning. The floMimeter was set to zero
at the beginning of each pumping cycle and the ball valve and pump were
started simultaneously. linen the required volune had been pumped, the ball
valve and pump were shut off and the flometer voluae was recorded to the
nearest 0.5 gallon.**
* Residual water and sediment control was approximately 0.1X
** All oil volumes were measured in gallons. One gallon is equivalent to
3.7853 liters.
99
-------
BRIDGE
OIL
STORAGE
TANK
POSITIVE
DISPLACE
MENT
PUMPS
POSITIVE
DISPLACE-
MENT
FLOWMETER
DISTRIBUTION
MANIFOLD
5cm (2 in.)
FLEXIBLE HOSES
OIL
RECIRCULATION
CIRCUIT
VALVES
SUMP
PRELOAD
WEIR EDGE
DISTRIBUTION
TANK
Figure C-l. Oil Input - Schematic Diagram
TOO
-------
When oil was to be distributed as a slick, the pumps were adjusted
to the required flowrate by circulating the oil through the flowmeter and
back to the storage tank. The pumps were then stopped, the flowmeter set
to zero and the valves adjusted to distribute oil into the distribution
manifold and weir edge distribution tank. During the test, the pumps were
operated for a predetermined time period and the volume distributed was re-
corded to the nearest 0.5 gallon.
Handling of the recovered oil is shown schematically in Figure C-2.
Two air operated double diapharagm pumps were used in parallel to recover
the oil/water mixture from the weir skimmer.* The pump outputs were man-
ifolded together and discharged into a series of 100-gallon recovery bar-
rels mounted on the rear bridge. When recovery efficiency was to be mea-
sured, the pumps, hoses, etc. were preloaded with water and the total volume
of oil and water pumped during the run was measured as the volume collected
in the recovery barrels.
To insure that all oil was recovered from the pumps, hoses, etc.,
approximately 80 gallons of water were pumped through the system after the
last visible oil entered the suction hoses.**
The recovered oil/water mixture was allowed to settle for at least
1/2 hour or until a clearly visible oil/water interface was visible. The
excess water was then drained off and the depth of the remaining mixture
was measured to the nearest 1/8-inch and converted to a volume measurement.
One eighth inch corresponds to approximately 3/8 gallon. A motor driven
propeller was then inserted into the mixture and operated to thoroughly mix
the oil and water. One or more samples of this mixture were then withdrawn
for laboratory analysis.
The percentage of water and sediment in the sample was measured
according to ASTM standards*** and used to correct the recovered volumes.
Systematic Errors
The systematic errors associated with oil volume measurements can be
calculated from the precision of the oil handling methods. The two main
error sources are the measurement of the sample water content and the mea-
surement of the total mixture volumes.
The ASTM method used to measure water content utilized a centrifug-
ing technique where the oil sample was mixed with an equal volume of toluene
and centrifuged in pre-calibrated pear-shaped tubes as shown in Figure C-3.
The percent water and sediment in the sample was calculated by doubling the
separated water volume. The accuracy of this procedure is limited by the
calibration on the tube. The error in corrected oil volume associated with
the minimum calibration divisions as a function of percent water in the
sample as shown in Figure C-4. These errors may be interpreted as the
* A portion of the final sump oil volume was recovered by pumping direct-
ly out of the sump to retrieve the oil trapped underneath the stability
control system float.
** The volume of the pumps and hoses was approximately 50 gallons.
*** ASTM designation D 96-73 Standard Methods of Test for Water and Sediment
in Crude Oils.
101
-------
7.5 ai (3 in.)
FLEXIBLE
SUCTION HOSES
I
[« in.
FLEXIBLE
DISCHARGE
HOSE
1
DOUBLE
"All OPERATED
POMPS
1
IIWEDIATELf
UTTER RECO¥ER¥
OIL + MATER
MIXTURE
AFTER
STANDING
FOR 1/2
AFTER DRA1MING MIXING
EXCESS MATER
FLUID LEVEL
IS MEASURED
CLEARLY
fISIBLE
INTERFACE
REHAIMIIIG
OIL AMD
HATER
BEFORE
SAMPLING
Figure C-2. Oil Recovery - Schematic Diagram)
102
-------
-1MM I.D
44 TO 47 MM
SCALE
36 TO-43 MM
157 TO I6O
MAX.
58± I MM O.D.
SANDBLASTED SPOT
(FOR MARKING)
Figpne C-3. Fear-Shaped Cerotrifiuige Tu|be from AS1M Oeslgmatiom P §6-73
;/yPI Standamdl 2S42
'"Staindlandl Net tod Of Test For Mater
Amd SedimEint In Crade
-------
r- f-\
50-
40-
30-
20-
1 rs
1U-
n
ERROR IN
CALCULATED
OIL VOLUME
DUE TO
CALIBRATION
MARKS
RECOVERED ,
32;- .
I
I
1
1
1
1
i 1
— -I-. i i i 1 . j
' '
10
20
3Q
40
50
60
% WATER
IN SAMPLE
Figure C-4. Error Associated With Minimum Calibration
Divisions On Centriguge Tube ASTM D96-73
104
-------
maximum error due to misreading the calibrations by one of the smallest grad-
ations. The percentage of water in the oil inputs to the tests were 0.1 per-
cent on the average and the error associated with this water content is 0.2
percent.
Figure C-5 is a histogram showing the distribution of the percentage
water in the recovered oil samples. The average water content of the 32 per-
cent corresponds to an error in oil volume of 10 percent.
The errors associated with the measurement of oil volumes may be es-
timated from the precision of the volume measurements and the magnitude of
the'Volumes measured.
The oil inputs were measured to the nearest 0.5 gallon with a posi-
tive displacement flowmeter. Two oil preload volumes were used, 40 gallons
and 75 gallons with an average of 49 gallons. A range of oil slick volumes
were used with a minimum of 31 gallons, a maximum of 189 gallons and an av-
erage of 94 gallons. The percentage errors in oil volumes associated with
measurement errors of 0.5 gallon are shown in Figure C-6. The average error
for the preload volume was T.O percent and for the oil slick volume it was
0.5 percent.
Recovered oil volumes were measured by recording the depth of the
oil layer in 100 gallon recovery barrels* to the nearest 1/8 inch. Up to
five barrels were used for each test. Figure C-7 is a histogram showing the
distribution of recovered oil volumes measured. The 1/8 inch uncertainty in
oil layer thickness corresponds to approximately 0.375 gallon and the per-
centage error as a function of sample volume is shown in Figure C-8. The
average 48 gallon oil volume corresponds to an error of 0.8 percent.
Table C-l summarizes these systematic errors for average test con-
ditions. The majority of the total error is due to recovered oil and is pri-
marily due to the limitations of the water content measurement procedure.
Other Errors
There are several other non-quantifiable sources of error. These
include losses and uncertainty due to:
• Oil volume in weir edge distribution tank
• Oil distributed beneath the surface instead of as a slick
• Oil not encountering the streamlined oil boom
A 15-gallon barrel was used for two samples and a 500-gallon barrel was
used for one sample.
105
-------
NlUWBEiP
25-
15- -
82 SAMPLES
32?
HIS!
Fifyre C-5. Histopraii of % Mater lira Recovered Oil Samples
106
-------
2.0H -
1.5
1.0- -
MMJUIME MEASURED
IBWE5T 0.5 gal.
OIL VOLUME
•gal.
Error Due To Oil Input
foUune
-------
15- -
10- -
5-
NUMBER
OF SAMPLES
AVERAGE VOLUME
48 gal.
NOTE:
MEASURED IN
100 gal. BARRELS
(NOMINAL CAPACITY;
10 20 30 40 50 60 70
80 90 100 110 120 130
SAMPLE VOLUME
IN gal.
Figure C-7 Histogram Of Recovered Oil Volumes
108
-------
RECOVERED VOLUME MEASURED
TO NEAREST 3/8 gal.
50 60 70 80 90 100 110 120
SAMPLE VOLUME
IN GAL.
Figure C-8. Error Due To Recovered Volume Measurement
109
-------
Table C-1. OIL fOLWE ttJISMROEMT SUMMARY -
ERRORS DUE TO MINIMUM CALIBRATION DIVISIONS
Percent Error
fclwK
Volume Measuraient
Preload 1.0
Slick 0.5
Recovered Oil 0.8
Mater
Content
Measurement
0.2
0.2
10.0
Total
1.2
0.7
10.8
• Kesidbal oil IM the channel
« Oil coating puips, noses, streamlined oil boon, etc.
* Oil lost im switdhiirag recovery hose toe&ueen [barrels
* Start-imp and shut dowi transients
• Oil lost iira draiifiiiirag excess Mater from recovery barrels
Most of these sources of error are oil losses associated with the
test procedure and as a resmlt, the calculated values of thruput and reco-
very efficiencies probably tend to be lower than the actual volumes.
Considering all factors, the total probable experimental error for
and recovery efficiencies is estimated to be +15 to -10 percent.
110
-------
^^^^TBajWO^RB^MlTDATA 1
1. REPORT NO. |i ~
EPA-60O/2-80-140 |
4L TITLE AMD SUBTITLE
Development of a High Current Oil boon/Skinner
Blair A. Folson
». PERFORMING ORGANIZATION NAME AMID ADDRESS
Ultrasystens, Inc.
Irvine, CA 92715
12. SPONSORING AGENCY NAME AMD ADDRESS
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
I. RECIFI'ENTS A.CC£SSIi0S#B*O_ 1
I, REPORT DATE 1
JKME IfSffl ISSDUie; DAXE. I
L PERFORMMIte OAGANIZATtON CODE 1
1. PERFCTHtlMBWiG OR&AMHZATIONI REPORT MO. 1
lOt PBIOGRAM ELEMENT MO. 1
1BB610 I
68-4B-0403 [
13L TYPE OF REPORT AND PERIOD COVERED I
Final , 1
114, SPOWSORWilWe AGENCY CODE 1
EPA/600/12 1
15. Stm-PLEWEMTARY NOTES
16. ABSTRACT
A low drag oil spill recovery device has Ibeen developed to retain and recover
floating oil slicks in currents up to 3.0 o/sec (§ knots) ami low nave coinditions
typically encountered in small lakes and harbors, smooth flowing streams and other
inland Matemays. The streamlined oil boop/skimer utilizes a slightly submerged
slotted hydrofoil to skin a thin layer of oil and water into an attached sunp Mfaere
the kinetic energy of the high speed flow is dissipated and the oil and water are
separated by gravity. The excess water is discarded and tlte oil is retained in the
sunp under conditions similar to a conventional low speed oil boon. The retained
oil is recovered as required by a built-in secondary skiner. Because of the
systen's siqriicity and low drag, it can be deployed as a high speed skimner
supported by a catanaran or as a high current bmn with several adjacent sections
spanning a high current area.
The development effort focused m bow, swp and stability control systen design
and the testing of the conplete streamlined oil boon systen at OWBETT. Thruput
efficiencies as high as 99* percent and 87 percent Mere measured under calm surface
conditions at 2.0 and 3.0 «/sec (4 and 6 knots) respectively. Performance degraded
in tests with waves.
|7 KEY WORDS AMD mMmmCEMVAMML'ffSIIS I
L EMESCRHrpORS
Hater Pollution
Oils
SkiiKrs
Harbors
Performance Tests
ISL DBSTBIWUmOWl STATtMtMT
RELEASE TO PUBLIC
iiL£DEB*TEFKERS/aFEW ERSEJED TERMS
Equipment Evaluation
Oil Spill Clearap
Protected Maters
19u SBDUWIlT^ dASS flMusSera*tf
WCLASSIFIED
ant SECURITY CLASS fttmftgcj
WOLASSIFIED
c, COSATII meaSI&amf |
&BD 1
2t.mO.OtF WMBES I
121 1
Z2-PRBCE 1
111
: i*ni—sKy-U
------- |