Concept Evaluation: Recovery of Floating Oil Using Polyurethane Foam Sorbent


EPA-R2-72-049
                     Environmental Protection Technology Series
 September 1972
 Concept  Evaluation:

 Recovery of Floating Oil Using

 Polyurethane Foam Sorbent
                               Office of Research and Monitoring


                               U.S. Environmental Protection Agency

                               Washington, D.C. 20460

-------
            RESEARCH REPORTING SERIES
Research reports of the  Office  of  Research  and
Monitoring,  Environmental Protection Agency, have
been grouped into five series.  These  five  broad
categories  were'established to facilitate further
development  and  application   of   environmental
technology.   Elimination  of traditional grouping
was  consciously  planned  to  foster   technology
transfer   and  a  maximum  interface  in  related
fields.  The five series are:

   1.  Environmental Health Effects Research
   2.  Environmental Protection Technology
   3.  Ecological Research
   4.  Environmental Monitoring
   5.  Socioeconomic Environmental Studies

This report has been assigned to the ENVIRONMENTAL
PROTECTION   TECHNOLOGY   series.    This   series
describes   research   performed  to  develop  and
demonstrate   instrumentation/    equipment    and
methodology  to  repair  or  prevent environmental
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..

-------
                                                         EPA-R2-72-OU9
                                                         September 1972
         CONCEPT EVALUATION:   RECOVERY OF FLOATING OIL

                USING POLYURETHAJTE FOAM SORBENT
                               By

                          C.  H. Henager
                          J. D. Smith
                    Contract  No.  68-01-0070
                       Project  15080 HEU
                        Project Officer

                         R. T.  Dewling
            Edison Water Quality Research Division
            National Environmental Research Center
                   Edison, New  Jersey  08817
                         Prepared for

               OFFICE OF RESEARCH AND MONITORING
             U.S. ENVIRONMENTAL PROTECTION AGENCY
                    WASHINGTON,  D.C. 20^60
For Mia by tbe Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.80

-------
U. S. Environmental Protection Agency Review Notice
This report has been reviewed by the U. S. Environ-
mental Protection Agency and approved for publication.
Approval  does not signify that the contents necessarily
reflect the views and policies of the U.  S. Environ-
mental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.

-------
                             ABSTRACT

Individual components of an oil  spill recovery system were evaluated
using Bunker C oil and three crude oils ranging in API gravity from
14° to 42°.  The system was designed to shred and broadcast polyurethane
foam sorbent onto an oil slick,  herd the sorbent to a shipside conveyor
by a water spray boom, squeeze the sorbent to extract the oil  and rebroad-
cast the sorbent.  The initial concept was to build a half-size,  full
scale system; however, initial foam losses indicated the necessity for
a re-evaluation of the program,  and specific studies on the broadcasting
and "squeezing" systems were undertaken.

The shredder-broadcaster, a commercial straw mulcher, produced acceptable
shredding of dry foam.  However, multiple cycling degraded oil-soaked
foam to unrecoverable sizes in relatively few cycles.  With Bunker C
oil, 47 percent of the foam was  reduced to sizes less than 1/2" in 5
cycles.  With light Canadian crude oil, 29 percent was reduced to less
than 1/2" in 100 cycles.

A sorbent-oil separator using perforated rolls, was designed to extract
viscous oils from the foam at 20°C, without heating, at rates  of up to
5000 gph.  This device showed good recovery of oil from foam.   Multiple
cycling of Bunker C-oil-soaked foam through the full scale device resulted
in a small loss of foam by size  reduction (3.5 percent in 77 cycles).   After
77 cycles of extracting Bunker C oil, foam damage by loss of resiliency
reduced oil extraction per cycle to about 25 percent of the initial amount.
No loss of resiliency was observed up to about 50 cycles.

Because of the high sorbent losses in the shredder-broadcaster, the system
as initially proposed is not recommended for use with Bunker C oil.

This report was submitted in fulfillment of Project No. 15080HEU,
Contract No. 68-01-0070 under the sponsorship of the Office of Research
and Monitoring, U. S. Environmental Protection Agency.
                                m

-------
                             CONTENTS

Section                                                           Page
I             Conclusions                                           1
II            Recommendations                                       5
III           Introduction                                          7
IV            Design and Construction of Experimental               17
                 and Prototypical  Subsystems
V             Evaluation of Subsystems                             33
VI            Acknowledgments                                      59
VII           References                                           61
VIII          Glossary                                             63
IX            Appendix                                             65

-------
                             FIGURES

Number                       Title                             Page

  1           Overall   System Concept                             8

  2           Flow Channel  and Apparatus  for  Determining          19
                Conveyor Angle and Dewatering  Rates

  3           Schematic  of  Sorbent-Oil  Separator                  21

  4           Sorbent-Oil Separator Under  Construction            22
                Showing Internal  Mechanism

  5           Perforated Rollers in the Full  Scale                23
                Sorbent-Oil  Separator

  6           The Completed Full Scale  Sorbent-Oil                24
                Separator

  7           Scale Model Sorbent-Oil Separator                  25

  8           External  Loads  and Reactions on Spray Boom          28

  9           Onload ing  Conveyor Support Frame                    30

 10           Sorption,  Separation  and  Retention of Bunker        34
                C  Oil  in Polyurethane  Foam Recycled in Model
                Sorbent-Oil  Separator  - One-Inch-Thick Foam

 11           Sorption,  Separation  and  Retention of Bunker C      35
                Oil  in  Polyurethane Foam  Recycled in Model
                Sorbent-Oil  Separator  - One-Half-Inch-Thick Foam

 12           Sorption,  Separation  and  Retention of San Joaquin   38
                Crude Oil  in Polyurethane Foam Recycled in
                Model Sorbent-Oil-Separator  -  Spray Mixing

 13           Sorption,  Separation  and  Retention of Light         39
                Canadian Crude  Oil  in  Polyurethane Foam
                Recycled in  Model  Sorbent-Oil-Separator -
                Spray Mixing

 14           Water-Oil  Ratio  Obtained  in Spray Mixing Tests      40
                with Santa Maria Crude Oil - One-Half-Inch-
                Thick Foam

 15           Conveying  of  Foam  on  Simulated Conveyor Belt for    42
                Canadian and  San Joaquin Crude Oils - Velocity
                1.0  Knots

                                vi

-------
Number                      Title

 16         Conveying of Foam on Simulated Conveyor Belt       43
               for Canadian and San Joaquin Crude Oils -
               Velocity 1.5 Knots

 17         Conveying of Foam on Simulated Conveyor Belt       44
               for Canadian and San Joaquin Crude Oils -
               Velocity 2 Knots

 18         Drainage of Fresh and Emulsified Santa Maria       46
               Crude Oil from Polyurethane Foam

 19         Particle Distribution jof One-Inch-Thick Foam       47
               in a Single Pass Through the Mulcher-
               Spreader

 20         Average Particle Area Distribution from the        48
               Mulcher-Spreader - One-Inch-Thick Foam

 21         Total Particle Area of One-Inch-Thick Foam         49
               from a Single Pass Through the Mulcher-
               Spreader - Calm Conditions

 22         Total Particle Area of One-Half-Inch-Thick Foam    51
               from a Single Pass Through the Mulcher-
               Spreader - Quartering Wind 1 to 3 Knots

 23         Total Particle Area of One-Half-Inch-Thick Foam    52
               from a Single Pass Through the Mulcher-
               Spreader - Tail wind 1 to 3 Knots

 24         Reinco Model TM7-30 Mulcher and Test Apparatus     53
               for Recycling Oil-Soaked Sorbent

 25         Particle Size Distribution of Light Canadian       54
               Crude-Oil-Soaked Sorbent from 0 to 100
               Cycles Through the Mulcher-Spreader

A-l         Appearance of Foam after 50 Cycles Through         71
               Squeeze-Roller Processing Bunker C Oil

A-2         Appearance of Foam after 60 Cycles                 72
               Through Squeeze-Roller Processing Bunker
               C Oil

A-3         Appearance of Foam after 70 Cycles Through         73
               Squeeze-Roller Processing Bunker C Oil

A-4         Apperance of Foam after 77 Cycles Through          74
               Squeeze-Roller Processing Bunker C Oil
                              vii

-------
                            TABLES

Number                      Title                             Page

  1          System Cycle Times and Weights                     17

  2          Results of Water Spray Herding Experiments          36
               with Sorbent and San Joaquin Crude Oil

  3          Results of Water Spray Herding Experiments          36
               with Sorbent and Light Canadian Crude Oil

  4          Onloading Conveyor Drainage for One-Half-Inch      45
               Mesh Screen as a Function of Time

  5          Foam Particle Distribution - Mulcher-Spreader      50
               Cycling Test

  6          Foam Particle Size Distribution - Sorbent-Oil       56
               Separator Cycling Test
                             viii.

-------
                          SECTION I

                         CONCLUSIONS

The system, as proposed, included:   (a) an unmodified  straw mulcher
to first shred 1/2-inch to 3/4-inch thick sheets of sorbent to about
5" x 5" pieces, broadcast it onto the water surface, convey the
sorbent in the suction air stream after pressing and rebroadcast
the sorbent, (b), a squeeze-roller to separate the oil (and water)
from the sorbent after the sorbent was  herded  to the side of a
recovery vessel by (c) a water spray boom and transported aboard
ship by (d), a mesh belt conveyor extending into the water.

Conclusions reached in the course of this project, based on tests of
individual components and foam sorption experiments, were:

1.  Polyether-based polyurethane foam  is capable of sorbing high
    amounts of oil per pound of sorbent from floating oil slicks
    with relatively short contact times,  Sorption of up to 53 pounds
    of Bunker C oil per pound of dry sorbent on the third cycle was
    demonstrated with ideal mixing conditions and a contact time of
    3 minutes per cycle.  For other oils, the material sorbed up
    to 32  Ibs/lb of a heavy asphaltic  crude oil and up to 39 Ibs/lb
    of a light crude oil  in the third  cycle of  simulated spray boom
    mixing  tests with a mixing time of 60 seconds  per cycle.   It
    appears that oiled  foam sorbs  oil  more efficiently than dry foam.

2.  A  squeezing operation can be successfully used  to extract  Bunker C
    and crude oils but  the  life of the foam used  in these  tests
    would  be  limited when processing  heavier oils  such as  Bunker C,
    without heating.   See item 11  of  these conclusions for  extrapolated
    foam  losses and replenishment  rates.

3.  The perforated squeeze-roller  type of  sorbent-oil separator showed
    good  results  in separating Bunker  C  oil from  polyurethane  foam
    sorbent without the addition of  heat.  Extraction of  up  to 30
    pounds of oil  per  pound of dry foam  was achieved  with  the  scale
    model  unit at  20°C.   Extraction  of 18  to  20 Ibs/lb and  25  Ibs/lb
    was achieved  for  heavy  asphaltic  crudes and light crudes,  respec-
    tively.  Little or no oil was  extracted on  the first  cycle, but
    extraction increased  with succeeding cycles.   The third  cycle
    showed good extraction  rates.

     For  light oils,  loss  of foam through perforations in  the squeeze-
    roller was restricted to  particles less  than 1/4  inch as determined
     by tests  with the small-scale  squeeze-roller.   For  Bunker  C oil,
     foam  losses in the full-scale  squeeze-roller were restricted  to
     particles of 3/8-inch and  less.   Foam losses attributable to  the
     full-scale squeeze-roller alone,  based on recycling  material
     shredded  by one  pass only of the mulcher, were 3.5% of the initial
     sorbent amount for 77 cycles when processing Bunker C oil.
     If the material  were passed  through the mulcher each cycle,
                               1

-------
    these losses would be much higher.   Foam losses  for  light  oils
    in the squeeze-roller alone should  be less  but this  figure was
    not determined experimentally.

    Repeated squeezing of foam soaked with Bunker  C  oil  caused a
    loss of resiliency and sorption capacity of the  foam.   After  77
    cycles of extracting Bunker C oil,  this loss of  resiliency reduced
    oil extraction per cycle to about 25% of the initial amount.   No
    loss of resiliency was observed up  to about 50 cycles.   Tests
    by SchatzbergO)indicate that this  loss of  resiliency does not
    occur in foam soaked with light oils (e.g.  No. 2 fuel oil) when
    squeezed up to 100 cycles or more.   Thus there is evidence that
    resiliency and sorption capacity of the foam would be less
    affected when processing light oils but this was not investigated
    in this study.

4.  An unmodified straw mulcher with an open, radial blade,  centri-
    fugal blower was  not suitable for multiple  recycling of oil-soaked
    foam in the system concept.  While  the device  performed  certain
    tasks adequately,  e.g.:  70 ft throw, acceptable  particle size
    distribution in one or more passes  of dry foam,  and  adequate
    coverage of broadcast foam, it produced size reduction  of  oil-
    soaked foams judged to be unacceptable.  When  soaked with  a light
    crude oil, 29% of  the foam was shredded to  particle  sizes  less
    than one-half inch in 100 cycles.   For Bunker  C  oil  this figure
    was 47% in 5 cycles.  Although improvements were possible  by
    modifying the blower, it was judged that a  pneumatic broadcaster
    of a different design would be more satisfactory.

5.  The most effective conveyor angle to transport foam  sorbent
    pieces of a general cubical shape soaked in a  light  oil  is 25°
    to 30°.  The angle for foam pieces  soaked in a heavy, viscous
    oil is not critical up to 45°.   Flat pieces (1/2 inch thick with
    rectangular dimensions of 2 inches  or more) soaked in a  light
    or heavy oil will  be transported by a conveyor at angles up to
    45°.  Because sorbent would be reduced in size and shape by
    recycling, an angle of 30° was selected for the  system  concept.

6.  A conveyor belt mesh size of 1/2 inch would provide  a good balance
    between dewatering and sorbent conveying capability. Tests
    showed that smaller sizes tended to plug with  heavy  oil  and debris.

7.  The use of a water spray boom appears to be an effective means
    of mixing sorbent  and oil on a water surface.   However,  still
    to be investigated is the potential problem of driving  oil
    into the water column and possibly  forming  an  emulsion.  Small
    scale experiments  of spray mixing yielded sorption of up to 32
    Ibs of oil/lb of  dry foam witha heavy asphaltic  crude oil  and up
    to 39 Ibs/lb of light crude oil from slicks 0.8  mm thick.   Herding

-------
    of oil-soaked  sorbent by a  water  spray, already  shown to be
    moderately effective in large scale  tests,  was also demonstrated
    in small  scale experiments  in currents  up  to  one knot.  A spray
    angle of  30° relative to the water  surface  was found to be most
    effective in small  scale tests with  a flat  spray nozzle.

8.  An emulsified  oil  would be  expected  to  have a higher draining
    rate from sorbent (e.g. oil lost  by  draining  through a mesh  belt)
    than a fresh oil.   Tests of an emulsified  heavy  asphaltic crude
    showed the drainage rate of the emulsified  oil to be 3 to 4  times
    that of fresh (unemulsified) oil.

9.  Heavy, viscous oils such as Bunker  C increase the rate and amount
    of sorbent degradation and  losses,  as  illustrated in  items 3 and
    4 of this section.  Based on the  data  obtained,  it was concluded
    that the system, as designed would  function most effectively on
    light and medium oils and least effectively on unheated,  heavy,
    viscous oils such as Bunker C.

10. The available contact time for sorption of oil  is 35  seconds for
    operation of the proposed system  at 3  knots.   The required contact
    time for absorption of Bunker C oil  in polyurethane foam  is  several
    hours.  Thus the absorption process, which is immediate on contact
    of the foam with Bunker C,  would  be utilized  in  the system.  This
    contact-adsorption would take place on one side  of foam  particles
    which landed on the oil.   It would be required that the remainder
    of the mixing of oil and sorbent take place  in the spray  boom
    mixing area.  Sorption of 32  Ibs/lb of a heavy,  viscous  oil,
    Santa Maria Crude, 14°API, was demonstrated  in a small  scale
    spray test  with a 60 second mixing time.

    The required contact time for absorption of about 12 Ibs/lb  of a
    light oil,  such as No.  2 fuel oil, by polyurethane foam  is  about
    5 seconds.  Sorption of up  to 39 Ibs/lb of a light Canadian  crude
    was demonstrated  in  the small scale spray mixing tests in 60
    seconds.

    Additional  experiments  on a  large scale are  needed to establish
    the required mixing  times  for both heavy and light oils in  the
    spray mixing area.   Time and  funding limitations did not permit
    such  evaluation  in  this study.

 11. Extrapolating foam  losses  through individual components to  a full-
    size  system indicates  the  following foam losses and replenishment
    rates:

     (a)   For light  oils  with the system as proposed  (using mulcher)--
          In  75  cycles  (one hour of boat operating time) the mulcher
          would  reduce about 6% of the foam to  particles less than
          1/4 inch;  these would be lost  during  the 75 cycles of  the
          squeeze-roller.   Neglecting other system losses, which were
          not determined  in this work, the  foam replenishment rate
          would  be about 6% per 75 cycles  (per  hour)  or about 7  Ibs of
          foam per hour for a full-size  system.
                                 3

-------
(b)   For heavy,  viscous  oils  such as  Bunker C, with the system as
     proposed  (using  mu1cher)--In 5 cycles (4 minutes of boat
     operating  time;  the mulcher would reduce about 47% of the foam
     to particles  less than  1/2  inch; assuming that all of these
     were lost  in  the 5  cycles of the squeeze-roller (this high
     rate would  override any  loss rate attributable to size
     reduction  by  the squeeze-roller) the replenishment rate
     would approximate 47% every 4 minutes or about 800 Ibs
     of sorbent  per hour.

(c)   For light oils using a  different broadcaster producing no
     reduction  in  foam size—Losses through the squeeze-roller
     would amount  to  less than 3.5% for 75 cycles.  Neglecting
     other system  losses, the foam replenishment rate would be
     on  the order of 4  pounds of sorbent per hour.

(d)   For heavy,  viscous  oils  such as  Bunker C, using a different
     broadcaster producing no reduction in foam size—The sorbent
     would degrade by loss of resiliency after about 75 cycles
     through the squeeze-roller and would require replacement.
     The foam replenishment  rate would be one system inventory
     per hour or 114  pounds/hour.

-------
                          SECTION II

                        RECOMMENDATIONS

1.  The sorbent-oil  separator developed under this  project should  be
    used and evaluated in a full-scale operation  on an oil  recovery
    system using polyurethane foam sorbents.   Prior to such use,  some
    minor modifications should be made to improve synchronization  of
    the rollers and  hinging of the internal  oil  scrapers.   Such  a
    system should employ pre-cut, flat-thin  pieces  of foam (from
    3" x 3" x 3/4" to 6" x 6" x 3/4")  or similar sized pieces  shredded
    from large sheets of 3/4" thick foam by  one  pass through the
    mulcher.  The system should use a  different  pneumatic  broadcaster
    to avoid the size degradation which occurred from recycling  foam
    through the mulcher-broadcaster.

2.  Different polyurethane foams should be investigated for resistance
    to damage by pressing and shredding operations  in multiple cycling
    with heavier oils such as Bunker C.  Stronger foams or specially
    processed foams, e.g. reticulated, may be more  resistant to  loss
    of resilience or to tearing.

3.  The spray boom concept for mixing  and herding oil  and  sorbents
    should be evaluated on a large scale. A spray  boom is more
    maneuverable and easier to propel  than a mechanical boom and  could
    possibly eliminate problems associated with  hang-up of debris
    and oil-soaked foam on conventional herding  booms.  In addition,
    it should enhance pick up of oil by the  sorbent since  it would
    mix oil and sorbent that would not otherwise come in contact
    with each other.  That is, a sorbent piece broadcast onto  plain
    water in a discontinuous oil slick would not sorb oil, but because
    the spray boom herds both oil and  sorbent, a forced mixing would
    take place at the spray line.

4.  For a water spray boom, additional work  is needed  to define  foam
    losses under or  around the boom, possible choking  at the entrance
    to the onloading conveying system, and the effectiveness of  the
    mixing action at the spray line.  Possible problems of forming
    emulsions and driving oil out of the foam and into the water column
    by the water spray should also be  studied.

5.  A comparative analysis should be made of the economics of  oil
    recovery by a recycling mechanical sorbent harvesting  system
    compared to present methods of oil spill  cleanup,  i.e.,  spreading,
    recovery and disposal of non-recyclable  straw by (largely) manual
    means.

-------
                          SECTION III

                         INTRODUCTION

The physical removal  of oil  from water with the aid  of sorbent materials
has been demonstrated as a desirable method of mitigating  the potentially
disastrous effects of oil pollution arising from tanker and  offshore
drilling accidents or operating errors.   A great number of natural  and
synthetic sorbents have been tried and evaluated(2>3,4).  Many appear
quite promising, but have yet to be effectively utilized for recovering
large quantities of oil, owing to the lack of appropriate  equipment
and/or systems.

This study was concerned with the development and evaluation of a  system
concept for recovering floating oil slicks from unprotected  waters
with the aid of a polyurethane foam sorbent (see Figure 1).

The system was to be capable of recovering floating  oil at the rate
of 10,000 gallons per hour from a slick thickness of 1.5 mm  or less
in five-foot seas with superimposed winds up to 30 mph and 2 knot
currents, close to shore in less than 80-foot water  depths,  with oil
temperatures of 20°C.  Other details of system requirements  are listed
in this section under "Design Goals".

-------
                                                                               SORBENT-OIL
                                                                                SEPARATOR
PILLOW
TANKS
                SORBENT
             MAKEUP STORAGE
D
                                                                             0'1-LOADiriG
                                                                              CONVEYOR

                                                             SPRAY BOOM ASSEMBLY
                                                   Figure  1

                                          Overall  System  Concept

-------
BACKGROUND
             (5)
Prior studies^ ' and experimental  work by Battelle-Northwest and
others(3), later confirmed by SchatzbergT4}} showed that open-cell
polyurethane foam was superior to  other available sorbents  in:  (a)
sorption capacity after multiple cycles; (b) oil  extractability; (c)
speed of sorption; (d) sorption capacity of virgin material; (e)
ability to sorb materials of interest and (f) hydrophobic-oleophilic
characteristics.  Other in-house research also indicated that sorbents
generally would require pre-heating to enable ready release of sorbed
Bunker C oil.

With these findings as a basis, a  system concept was synthesized to
meet specific  design goals described later in this section.  The
project as originally conceived was intended to result in a full-
scale half-system which was to consist of a water spray boom for
herding oil-soaked sorbent on one  side of a test vessel plus onloading,
conveying, extracting and oil storage equipment.   However,  information
from similar ongoing EPA projects  and consultations with EPA's
laboratory personnel indicated that a redirection of research effort
was needed in  order to define the  problems of recycling polyurethane
foams through  mulching and squeezing equipment at least 50 to 100
times.  Degradation to small, unrecoverable pieces and loss of
resiliency were believed possible.  If such stress occurred, system
losses would be high.

Special studies were conducted on  key foam handling equipment pieces
to establish whether or not the foam would perform effectively after
multiple cycling.  If the foam were rapidly degraded in these portions
of the cycle,  or sorbent losses were high, there would be little to
be gained by a complete system test using the particular foam handling
equipment chosen.  That is, failure of one component in the system
to provide reasonable recycling life could negate the entire system
regardless of the merits of the other components.

Re-direction of the program resulted when the sorbent size degradation
was high in the mulcher-spreader,  especially for Bunker C oil, an
oil which was  of major interest for the recovery system.  As a con-
sequence, the full-scale half-system was neither fully assembled nor
tested.  Rather, information was developed on foam behavior in the
separate process steps.

CONCEPT DESCRIPTION

The system concept utilized water  spray booms, extending outward and
forward from a support vessel, to  concentrate oil-soaked polyurethane
foam on the water surface.  The foam would be broadcast in front of
the vessel by mulcher-spreaders; contact time for the foam-oil
sorption process was 35 seconds.  Additional sorption was provided
by the water spray boom which would concentrate and mix the oil and
foam before pickup.  The concentrated material would be withdrawn

-------
from the water surface by means of partly submerged conveyors and
deposited on crossfeed conveyors and then into a sorbent-oil  separator.
This unit would extract the oil by squeezing between two perforated,
counterrotating drums.  Recovered oil would then flow by gravity to
a receiving tank from where it would be pumped to moderately-sized
pillow tanks which, when filled, would be dumped overboard for pickup
or towing by other vessels.  The sorbent from which the oil  had been
extracted would be conveyed forward by means of pneumatic conveyors
to the mulcher-spreaders which would re-distribute the sorbent on the
water surface.  Figure 1 shows the overall system concept.

The initial concept included a heating system with the sorbent-oil
separator to allow extraction of viscous oils such as Bunker  C.
Because of the large amount of heat required, and the size requirements
for such a heater, the design was modified to allow a pressing
operation without heat.  Development of a sorbent-oil separator to
function and extract Bunker C at 20°C without heating is described
in a subsequent section of this report.

The use of water spray jets for concentrating the oil-soaked  sorbent
was selected because it would make it possible to achieve the desired
rate of oil recovery without excessive support vessel speed and with
minimal equipment penetrating the water surface.  Control and con-
centration of oil slicks by hydraulic streams is described in a
previous EPA report(6).

DESIGN GOALS

The system concept was designed to meet the following goals:

     System Capacity - The system was designed to be capable  of coping
     with a spill of at least one million gallons with an oil  recovery
     rate of 10,000 gallons per hour under the maximum environmental
     operating conditions.  Recovery of oil  was to be from a  slick
     thickness of 1.5 mm or less.

     Environmental Conditions - The system was designed to effectively
     recover oil  and oil products from the surface of unprotected
     waters under operational conditions of:   five foot seas  (average
     wave height 5 ft) with superimposed winds up to 30 mph and 2
     knot currents; close to shore in less than 80-foot water depths;
     and oil temperatures of 20°C.

     Transportability and Deployment - The system was designed to be
     transportable to the scene of a spill by air, rail, highway, or
     water and to be readily adaptable to normally available  vessels
     at the scene.

     Spectrum of Petroleum Products to be Recovered - The system was
     designed to recover floating oil products ranging from light

                              10

-------
     diesel  oil  to Bunker C oil  over the range of  operational  environ
     mental  conditions previously described.
       ^ Product Quality - A design goal  for  the recovered  oil  was
     that it contain not more than 10 percent by volume  of  seawater.
     The effluent water discharged to the environment was to  contain
     not more than 10 mg/£ of oil .

MODEL DEVELOPMENT FOR DESIGN INPUT

Prior to the start of design of the full-scale,  half-system prototype,
a number of unknowns required developmental or experimental input.
These unknowns included:  an effective onloading conveyor angle,  foam
sorption rates, contact times and  draining characteristics, required
conveyor belt size, effect of spray jet flow  and angle on mixing  and
herding of sorbent and oil, and evaluation of a  squeeze-roller  concept
to extract Bunker C oils from sorbent without the use of heat.   Experi-
ments were conducted with scale models and an indoor  testing  basin
to provide this needed data.

To determine oil and water draining rates (from  a mesh belt onloading
conveyor retrieving oil -soaked sorbent) and the  suitable angle  for  an
onloading conveyor, experiments were made on  an  apparatus simulating
a conveyor belt.  Sections of varying size wire  mesh  were drawn out
of a circulating water stream at measured velocities  to  simulate the
action of a mesh conveyor belt traveling through the  water  and  retrieving
oil -soaked sorbent.  This apparatus, installed in an  indoor testing
basin, was used to determine the required angle  for an onloading
conveyor, optimum mesh size and draining rates.

Work to determine other data in the indoor basin included experiments
in sorption rates and contact times of different oils with  polyurethane
foam, optimum sorbent dimensions for conveyor belt transport  and
sorption efficiency, effectiveness of spray jets in mixing  oil  and
sorbent, and the effect of varying the spray  jet flow and angle on
retention (herding) of oil -soaked  sorbent. The  apparatus and test
results from these experiments are discussed  in  Sections IV and V
of this report.

For the initial evaluation of the  concept for oil extraction, a model
of a squeeze-roller type of sorbent-oil separator was constructed
using a 12" diameter perforated drum rolling  against  a non-perforated
neoprene-covered drum.  Information from the  performance of this  device
provided design information for the larger (3-foot-diameter rolls)
full-scale squeeze-roller.

SYSTEM DESIGN

A major effort of the program involved the design of  a full-scale
half-system.  Using input from the scale model and experiments  per-
formed on the indoor testing basin, a design  was generated  for  the

                               11

-------
full-scale half-system.   This design provided  a  full-size  spray  boom
mounted on one side of the 35 ft test vessel  in  the outdoor  vessel
testing basin plus other equipment for an operable system  to recover
5000 gallons per hour of oil.  Other equipment consisted of  onloading
and crossfeed conveyors, a squeeze-roller, a  pneumatic  conveyor,  a
mulcher-spreader, oil and sorbent storage, water pumping equipment,
oil pumping equipment and a gasoline-engine-driven hydraulic power
supply.

Individual components were designed (or, in the  case of the  mulcher-
spreader, procured as commercially available  equipment) to be installed
on the test vessel.  This design effort, discussed in Section IV  of
this report, included calculation of the system  capacity requirements,
interfacing of components with respect to a support vessel and water
surface motions, structural analysis of the spray boom, and  preparation
of design drawings.

EVALUATION PROGRAM DESCRIPTION

Following design of the system, certain key equipment pieces were
procured for evaluation.  These were the spray boom and framework to
support the onloading conveyor, the mulcher-spreader, the  squeeze-
roller and a hydraulic power supply to drive  the squeeze-roller.

Since the system concept required re-use (recycling) of the  polyurethane
sorbent, it was important to determine the degradation in  foam particle
size and potential loss rate in the system.  The amount of sorbent lost
in recycling through the mulcher-spreader and  the squeeze-roller  would
significantly affect the practicability and economics of the system as
well as requirements for onboard storage of sorbent for replenishment.
Also, additional work is undoubtedly needed to define losses via  the
spray boom and pneumatic conveyor; however, limited funds  prevented
such an evaluation.

Accordingly, the evaluation program was directed toward determining
potential sorbent degradation and losses in the  mulcher-spreader and
squeeze-roller.  Experiments were performed with the mulcher-spreader
to determine foam particle size, throw distance  and coverage for one
pass.  In this work 1/2" to 1" thick sheets of dry (un-oiled) foam
were run through the mulcher-spreader in one pass and data taken on
the resulting size distribution, coverage, and distance.

Further experiments to determine sorbent degradation such  as reduction
of size and reduction of sorption capacity were made using oil-soaked
foam.  For the mulcher-spreader a quantity of foam was soaked in a
light Canadian crude oil and recycled through the mulcher-spreader
100 times, simulating approximately two hours of operating time.   A
similar test was performed using foam soaked in Bunker C oil.  This
test was terminated at 5 cycles because of excessive size  reduction.
                               12

-------
 In the evaluation of the squeeze-roller, foam pieces of the size
generated by one pass through the mulcher-spreader (typically 3/4"
x 6" to 7" maximum rectangular dimension) were soaked in Bunker  C
and cycled through the squeeze-roller.  The foam was soaked each
cycle and the testing was terminated at 77 cycles at which time  foam
damage from loss of resiliency reduced oil recovery to about 25% of
the original value.

CONCEPT EVALUATION

Evaluation of the mulcher-spreader, a Reinco-Model TM7-30 straw
mulcher, showed that without modification, it produced acceptable
sorbent sizes in one pass of dry foam from sheets of polyurethane
foam 1/2-inch to 3/4-inch thick, 15 inches wide and up to 6 feet
long.  The majority of pieces produced were flat, thin pieces ranging
from 2 inches to 7 inches in maximum rectangular dimension, a shape
ideally suited for rapid oil sorption and stable travel up an inclined
conveyor.  Size reduction of the dry foam was negligible in 3 passes
through the mulcher, producing a 10 percent increase in number of
particles in the second cycle and almost no change for the third cycle.
Distribution (surface coverage) of the foam was good and the maximum
throw was 70 feet in calm air.  However, the tests using oil-soaked
foam revealed that size degradation occurred in multiple cycles.  When
soaked with light Canadian crude oil (42° API) 29 percent of the foam
was shredded to particle sizes less than 1/2 inch in 100 cycles.  For
foam containing Bunker C fuel oil, (Venezuela based, 22.9° API)  47
percent of the foam was reduced to particle sizes less than 1/2  inch
after 5 cycles through the mulcher-spreader.  This was judged to be an
unacceptable rate of size reduction because tests with the scale
model squeeze-roller indicated that particles of foam less than  1/2
inch would pass through the perforated roller along with the recovered
oil when processing Bunker C oil.

The mulcher-spreader blower wheel is an open, radial blade, centrifugal
design.  It is believed that the foam particles are sheared by the
rotating wheel blades where they interface with the stationary front
and rear scroll housing plates.  Addition of front and rear blade
shrouds on the blower wheel or selection of a blower better suited
to handle the foam could improve this subsystem; however, such modi-
fication or redesign was beyond the scope of this project.

Evaluation of the squeeze-roller was conducted initially on the  scale
model and later on the full scale prototype.  Initial results with
the scale model (12-inch diameter rolls; one perforated, one neoprene
covered), showed good extraction of oil, up to 30 pounds of oil  per
pound of dry foam with Bunker C oil, up to 18 Ibs/lb with San Joaquin
Crude (14° API), up to 20 Ibs/lb with Santa Maria Crude (14.7° API)
and up to 25 Ibs/lb with light Canadian Crude (42° API).  Losses of
foam containing light Canadian crude oil were negligible for the entire
particle size range (less than 1/2 inch to the 7-inch size) in the

                              13

-------
small scale model.  However, for foam containing the Bunker C oil,  the
material less than 1/2 inch passed through the 1/4-inch-diameter per-
forations of the small scale model along with the oil.   Material
greater than 1/2 inch was recovered for reprocessing.   (Note:  the
small scale model showed loss of 1/2 inch foam with Bunker C oil
whereas the full scale prototype showed loss of 3/8 inch foam with
Bunker C).

Tests of the full scale squeeze-roller (36" diameter rolls, both per-
forated with 1/4 inch holes) were made by multiple cycles of foam
pieces soaked in Bunker C oil for each cycle.  This evaluation was
directed primarily at determining foam damage from size degradation
and loss of resiliency or sorption capacity.  The oil  extraction rate
achieved was 4.6 Ibs of oil per pound of sorbent which  is lower than
the design value of 10 Ibs/lb.  However, the sorbent was not completely
saturated for each cycle because of time limitations.   It is believed
that the squeeze-roller can extract at least 10 Ibs/lb  from sorbent
saturated with Bunker C in the same manner that the scale model
accomplished rates in excess of this figure.

Results of the evaluation for size reduction and loss  of resiliency
showed that after 60 cycles with Bunker C oil, moderate size reduction
had occurred and the amount of oil extracted had reduced to about 50
percent of that originally extracted.  At 77 cycles, when the test
was terminated, the amount of oil able to be extracted  had decreased
to 25 percent of that originally extracted and the total foam volume
had reduced to about 25 percent of its original volume  from loss of
resiliency.  The fact that the sorbent recovered its original volume
after washing in solvent indicated that the viscous Bunker C oil  was
causing the foam to retain the volume reduction noted.   This indicates
that internal foam damage (destruction of cell walls)  occurred.  It
is probable that the Bunker C caused the damage while  being pressed
out of the tortuous internal paths.  Lighter oils, such as diesel
or light crudes would likely not cause this type of damage for several
hundred cycles, but this effect was not evaluated.

The total amount of sorbent lost to the system in the  squeeze-roller
evaluation was very low, amounting to 3.5 percent for  the 77 cycles
when using foam that was pre-shredded by only one pass  through the
mulcher.  These losses were 3/8-inch and smaller particles which
passed through the 1/4-inch perforations and were trapped in the
recovered oil.  No particles larger than 3/8 inch were  found to have
passed through the perforations.  It was concluded that 1/2 inch and
larger particles would be recovered as re-usable sorbent under
operating conditions similar to those used in the evaluation.  It
should be pointed out that if the foam were run through the mulcher
each time it was run through the squeeze-roller (i.e.  as in the proposed
system) the losses would be much higher.
                              14

-------
Size reduction in the squeeze-roller resulted primarily from a shearing
action of the two rollers because of failure of the hydraulic motors to
synchronize the rollers.  The size range of the original foam was
from 1 inch to 7 inches, maximum rectangular dimension (3/4 inches
thick) with 77.3 percent being in the 1  to 5 inch range and 18.5
percent in the 1 to 2 inch range.  After 77 cycles, 54.5 percent of
the pieces were in the 1 to 2 inch range and the maximum size was
3 inches.
                               15

-------
SECTION IV

DESIGN AND CONSTRUCTION OF EXPERIMENTAL AND PROTOTYPICAL SUBSYSTEMS

GENERAL-SYSTEM CAPACITY REQUIREMENTS

The design goal for the components of the full-size half system was
a capacity of 4000 pounds per hour of sorbent material or a sufficient
volume of sorbent to absorb oil at the rate of 5000 gallons per hour,
whichever amount was smaller. The following criteria and data were
used for the review of the system capacity requirements (mass, volume,
energy requirements).

Sorbent recovery capacity - 10 Ib oil/lb dry sorbent
Oil density - 8 Ib/gal
Oil collection rate - 5000 GPH/side
Dry sorbent density - 1.0 Ib/ft3 (conservative)
Speed of advance - 3 knots
Sweeping width - 20 ft/side

From the above data, the rate of sorbent processing is 4000 Ib/hr-side
of dry sorbent.

The following table presents the cycle duration for each system
component and the sorbent inventory of the component.

TABLE 1

SYSTEM CYCLE TIMES AND WEIGHTS

Foam
Time-Min. Wt-Lbs
Mulcher-spreader flight - 60 ft 0.017 1.13
@ 60 FPS

Floating travel* - 180 ft 9 304 FPM 0.59 39.30

Onloading conveyor - 3 ft x 24 ft 0.08 5.32
@ 304 FPM