&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

-------