EPA-670/2-73-084
October 1973
Environmental Protection Technology Series
An Oil Recovery System
Utilizing Polyurethane Foam «
A Feasibility Study
Office of Research and Development
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
U. 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-670/2-73-084
October 1973
AN OIL RECOVERY SYSTEM UTILIZING POLYURETHANE FOAM
« A FEASIBILITY STUDY
by
R. A. Cochran
J. P. Fraser
D. P. Hemphill
J. P. Oxenham
P. R. Scott
Contract #68-01-0067
Project Officer
J. Stephen Dorrler
Edison Water Quality Research Laboratory, NERC
Edison, New Jersey 08817
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C 20460
For sale by the Superintendent of Documents, U.S. Government Printing Offlw, Washington, D.C. 20402 - Price $2.35
-------
EPA Review Notice
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental Protection
Agency, nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
ii
-------
ABSTRACT
A system has been developed for recovering spilled oil from water surfaces
under a wide variety of environmental conditions and for all types of oils.
The system is designed to recover oil at rates up to 9,000 gal./hr.
This system is based on the use of polyurethane foam, foamed on the job
site, as a sorbent for the spilled oil. The foam is recirculated to
increase efficiency and to lower unit costs. Equipment needed includes
collection booms, an open-mesh chain-link belt for harvesting the oil-soaked
sorbent, and a roller-wringer to remove oil and water from the foam. The
foam is initially comminuted and distributed onto the water by means of a
hay blower (mulcher), and recycled foam is distributed by an open-throat
centrifugal blower. Recovered oil and water are transported to shore in
large fabric bags for further treatment prior to disposal. Used foam is
disposed of by incineration.
This report was submitted in fulfillment of Contract No. 68-01-0067 under
sponsorship of the Water Quality Office, Environmental Protection Agency.
iii
-------
CONTENTS
Section
I Conclusions
II Recommendations
III Introduction
IV Performance of Subsystems
V System Performance
VI Foam Fabrication and Characteristics
VII Sorption
VIII Distribution of Sorbent
IX Collection
X Harvesting of Sorbent
XI Wringing
XII Foam Degradation During Recycling
XIII Foam Disposal
XIV System Design
XV Acknowledgments
XVI References
XVII Nomenclature
XVIII Appendices
Page
1
3
5
11
23
29
45
83
91
103
117
141
155
165
179
181
183
185
-------
FIGURES
PAGE
1 INITIAL CONCEPT OF OIL-SPILL RECOVFRY SYSTEM 6
2 VIEW OF WAVE TANK 9
3 VIEW OF CURRENT TANK 9
4 TEST BOOM ASSEMBLY IN WAVE TANK 16
5 MODEL HARVESTER INSTALLED IN CURRENT TANK 18
6 EXPERIMENTAL WRINGING APPARATUS 19
7 SHELL PIPE LINE MODEL FURNACE USED TO BURN 20
POLYURETHANE FOAM
8 PROPOSED CONFIGURATION 24
9 USE OF COMPONENT MODULES FOR REDUCED CAPACITY 25
SYSTEM
10 TYPICAL ON-SITE GENERATE POLYURETHANE FOAM PRODUCED 31
AT 45°F AND 80$ RELATIVE HUMIDITY JANUARY 11, 1972
(2X2 INCH GRID WITH SUBDIVISIONS OF 0.1 INCH)
11 SINKING RATE OF ON-SITE GENERATED POLYURETHANE FOAM 32
INTO QUIESCENT SYNTHETIC SEA WATER
12 AGING TIME VERSUS COMPRESSIVE LOAD TO OBTAIN 25%, 37
50$, AND 65$ COMPRESSION OF ON-SITE GENERATED
POLYURETHANE FOAM
13 PORTABLE FOAMING EQUIPMENT 38
14 MIXING HEADS USED WITH PORTABLE FOAMING EQUIPMENT 38
15 POLYURETHANE FOAM PRODUCTION UTILIZING A PORTABLE 40
GRACO HYDROCAT UNIT
16 POLYURETHANE FOAM PRODUCED UTILIZING PORTABLE 40
GRACO HYDROCAT UNIT
17 SCHEMATIC OF CONTINUOUS BELT FOR MAKING POLYURETHANE 41
AT SITE OF OIL SPILL
VI
-------
PAGE
18 CONTRACT FOAM EQUIPMENT 42
19 POLYURETHANE FOAM BUN FORMED INTO ROLL 43
20 FOAM BUNS READY FOR MULCHING UTILIZING A REINCO 43
HAY SPREADER
21 ABSORPTION OF OIL FROM A SLICK BY A POROUS 45
OLEOPHILIC BLOCK
22 DIMENSIONLESS VOLUME ABSORBED FROM A SLICK BY A 47
WATER SATURATED OLEOPHILIC BLOCK AS FUNCTION OF
DIMENSIONLESS TIME
23 BENCH-SCALE SORPTION TEST APPARATUS 49
24 VISCOSITY OF TEST OILS 51
25 SPECIFIC GRAVITY OF TEST OILS 52
26 "MULE'S FOOT" SQUEEZER 53
27 OIL VOLUME RECOVERED, CARNEA 15 54
28 PERCENT OIL IN EFFLUENT RECOVERED, CARNEA 15 55
29 OIL VOLUME RECOVERED FROM SORPTION OF CARNEA 15 56
30 PERCENT OIL IN EFFLUENT RECOVERED FROM SORPTION 57
OF CARNEA 15
31 OIL VOLUME RECOVERED FROM 0.057 IN. SLICK BY 58
2-INCH SORBENT CUBES
32 PERCENT OIL IN EFFLUENT RECOVERED FROM 2-INCH 59
SORBENT CUBES
33 RESULTS OF QUALITATIVE EXPERIMENT WITH NO. 6 61
FUEL OIL
34 NATURE OF RECOVERY DEPENDENCE UPON VISCOSITY FOR 62
POROUS OIL SORBENT FOR EXPOSURE PERIODS INSUF-
FICIENTLY LONG TO PERMIT COMPLETE SATURACTION
OF THE FOAM BY MORE VISCOUS OILS
35 APPARATUS AND PROCEDURES 63
36 SPECIFIC SORPTION OF FLUIDS AS A FUNCTION OF THE 66
AREA CONCENTRATION OF FOAM SORBENT
vii
-------
PAGE
37 OIL RECOVERED PER UNIT AREA OF SLICK AND EFFLUENT 67
PURITY AS FUNCTIONS OF AREA CONCENTRATION OF FOAM
SORBENT APPLICATION
38 MULCHED POLYURETHANE FOAM ON SURFACE OF 5-FOOT 68
DIAMETER TANK
39 OIL REMOVED FROM THE SLICK AS FUNCTION OF AREA 70
CONCENTRATION OF FOAM SORBENT
40 SPECIFIC OIL SORPTION AS FUNCTION OF AREA CONCEN- 71
TRATION OF FOAM SORBENT
41 OIL CONTENT OF NET INFLUENT AS FUNCTION OF AREA 72
CONCENTRATION ON FOAM SORBENT
42 EFFECT OF SORBENT APPLICATION CONCENTRATION ON 74
SPECIFIC SORPTION OF OIL FOR NO. 2 DIESEL OIL
43 EFFECT OF SORBENT APPLICATION CONCENTRATION ON 75
SPECIFIC SORPTION OF OIL FOR CARNEA 21
44 EFFECT OF SORBENT CONCENTRATION ON RECOVERY 76
EFFECTIVENESS FOR NO. 2 DIESEL OIL
45 EFFECT OF SORBENT CONCENTRATION ON RECOVERY 77
EFFECTIVENESS FOR CARNEA 21
46 EFFECT OF SORBENT APPLICATION CONCENTRATION 78
ON RECOVERY EFFECTIVENESS FOR NO. 2 DIESEL OIL
47 EFFECT OF SORBENT APPLICATION CONCENTRATION ON 79
RECOVERY EFFECTIVENESS FOR CARNEA 21
48 EFFECT OF SORBENT APPLICATION CONCENTRATION ON 80
OIL CONTENT OF AFFLUENT FOR NO. 2 DIESEL OIL
49 EFFECT OF SORBENT APPLICATION CONCENTRATION ON 81
OIL CONTENT OF AFFLUENT FOR CARNEA 21
50 FOAM PREPARED WITH FITZGERALD BREAKER 84
51 REINCO TM 7-30 POWER MULCHER 85
52 MODIFICATION TO MULCHER 85
53 FOAM TRANSPORT TEST USING POWER MULCHER 88
viii
-------
PAGE
54 TWO POSSIBLE MODES OF FAILURE FOR FOAM SORBENT 92
BOOM IN ABSENCE OF WAVES
55 SHEAR TEST APPARATUS IN USE 93
56 SHEAR TEST APPARATUS 94
57 VARIATION IN LINEAR SHEAR STRENGTH OF A SHEET OF 96
MULCHED FOAM ON WATER AS A FUNCTION OF AREA
CONCENTRATION
58 BRIDGING OF CONVERGING BOOMS BY FOAM MULCH SORBENT 97
59 FAILURE OF BOOMS BY SPLASHOVER AT 1.5 FT/SEC IN 99
PRESENCE OF WAVES
60 BRIDGING TEST ARRANGEMENT 101
61 OCCURRENCE OF BRIDGING OF CONVERGING 12 -FOOT BOOMS 102
WITH GAP OF 3 FEET AND INCLUDED ANGLE OF 90°.
DARKENED SYMBOLS INDICATE OCCURRENCE OF BLOCKING
62 FOAM PARTICLES AGAINST STATIC BELT IN CURRENT 105
63 UPWELLING OF ENTRAINED AIR AND WATER 107
64 1/2" X 1" FLAT BELT HARVESTER WITH EXPANDED METAL 109
FLIGHTS
65 HARVESTER IN WAVE TANK
66 TOWING INTO 14-FOOT LONG WAVES AT 2.0 FT/ SEC 111
67 EFFECT OF WAVE FORM AND FOAM CONCENTRATION 112
68 TOWING INTO 8 -FOOT LONG WAVES AT 2.5 FT/ SEC 113
69 TYPICAL LOADING OF 4- INCH EXPANDED METAL FLIGHT 113
AT HARVESTER ANGLE OF 40°, BELT SPEED = 2 FT/ SEC,
SYSTEM VELOCITY = 3 FT/ SEC
70 PHOTOS OF APPARATUS WRINGING MULCHED FOAM IN 118
ARRANGEMENT TYPICAL OF THAT USED IN THE EXPERIMENTS
71 SCHEMATIC OF WRINGER, CONVEYOR, AND FOAM 119
ARRANGEMENT
IX
-------
PAGE
72 DETAIL OF WIRE MESH CONVEYOR BELT 120
73 VOLUME OF OIL/FOAM MASS VS IMPOSED FOAM PRESSURE 121
74 VOLUME OF OIL/FOAM MASS VS CONVEYOR SPEED 123
75 CORRELATION OF WRINGING DATA FROM EXPERIMENTS USING 124
ONLY OIL AND 2-INCH FOAM CUBES
76 TRANSIENT BEHAVIOR OF SOAKED FOAM DURING RECYCLING 128
77 WRINGING PERFORMANCE FOR DIFFERENT NUMBER OF CYCLES 129
78 WRINGING PERFORMANCE USING 2-FT DIAMETER ROLLER 131
79 LIQUID REMOVED FROM FOAM USING 2-FT DIAMETER ROLLER 132
80 FOAM FAILURE DURING WRINGING RESULTING FROM 135
INCREASING OIL VISCOSITY
81 DRAINING RATE OF OIL AND WATER SOAKED FOAM 138
82 FEEDING MULCHED FOAM FROM TRANSFER BIN INTO REINCO 146
BLOWER
83 NOZZLE USED TO DISTRIBUTE FOAM ONTO WATER SURFACE 146
IN CURRENT TANK
84 FOAM APPROACHING AND BEING PICKED OFF WATER SURFACE 147
BY HARVESTER BELT
85 HARVESTER BELT DISCHARGING ONTO CONVEYOR BELT 147
86 FOAM FALLING DOWN CHUTE ONTO LINK CHAIN BELT OF 148
WRINGING APPARATUS. TWO 24-INCH ROLLS IN WRINGER
87 WRINGER DISCHARGES DRY FOAM INTO TRANSFER BIN 148
88 VISCOSITY OF TEST OILS USED IN RECYCLING TESTS 150
89 POLYURETHANE FOAM REMAINING AFTER EACH CYCLE THROUGH 151
SYSTEM
90 WEIGHT LOSS VS SAMPLE TEMPERATURE - TEMPERATURE OF 159
ON-SITE GENERATED POLYURETHANE FOAM INCREASED AT
9°C PER MINUTE
-------
PAGE
91 WEIGHT LOSS VS SAMPLE TEMPERATURE - TEMPERATURE 160
OF SHELL PIPE LINE POLYURETHANE FOAM INCREASED
AT 6°C PER MINUTE
92 POLYURETHANE FOAM BURNING FURNACE - SCHEMATIC 161
93 SCHEMATIC OF SHELL PIPE LINE DESIGN POLYURETHANE 163
FOAM BURNING FURNACE
94 FLOW CHART PROTOTYPE - EXAMPLE 166
95 MINIMUM BOOM REQUIRED FOR SINGLE BOOM SYSTEM WITH 169
60 SECOND FOAM RESIDENCE
96 FOAM AVAILABLE TO HARVESTER PER UNIT WIDTH 171
97 USE OF COMPONENT MODULES IN HIGH CURRENT (RIVER) 176
xi
-------
TABLES
PAGE
1 DESIGN GOALS 8
2 RECIPE FOR ON-SITE POLYURETHANE FOAM 12
3 COMPARISON OF PHYSICAL PROPERTIES AND SATURATION 13
TIMES FOR FOUR POLYURETHANE FOAMS
4 COMPARISON OF MAXIMUM OIL SORPTION DATA FOR TWO 15
POLYURETHANE FOAMS
5 POLYURETHANE FOAM BURNING RATE FROM FURNACE MODEL 21
6 FLUE GAS ANALYSIS OF EVOLVED GASES WHILE BURNING 22
FOAM CONTAINING NO. 2 DIESEL FUEL
7 ESTIMATED COST OF OIL RECOVERY USING POLYURETHANE 26
FOAM
8 FOAM PRODUCTION RATES AND FOAM PROPERTIES 30
9 SORPTION EVALUATION OF ON-SITE GENERATED FOAM 34
UTILIZING SURFACE-COLLECTING AGENTS
10 COMPRESSIBILITY OF ON-SITE GENERATED POLYURETHANE 36
FOAM
11 PROCEDURES USED TO STUDY OIL SORPTION BY FOAM BLOCKS 50
12 PARTICLE SIZE DISTRIBUTION OF MULCHED FOAM 62
13 PROCEDURES USED FOR SORPTION TESTS IN LARGE TANK 64
14 PARTICLE SIZE DISTRIBUTION OF SLICED FOAM 83
15 FOAM TRANSPORT TESTS 88
16 PARTICLE SIZE DISTRIBUTION RETAINED ON SCREEN - % 89
17 PROCEDURES FOR MEASURING SHEAR STRENGTH OF FOAM 91
MULCH SHEET
18 PROCEDURE FOR CONFINED FOAM BOOM TOWING TEST 98
19 PROCEDURE FOR LOOSE FOAM TEST 98
20 PETENTION OF FOAM ON AN INCLINED STATIC BELT CONVEYOR 104
xii
-------
PAGE
21 SUMMARY OF CURRENT TANK TEST RESULTS - RANKED 107
22 SUMMARY OF CURRENT TANK TEST RESULTS - RANKED 109
23 APPARENT OR ''PACKING DENSITY" OF FOAM WITHOUT 114
COMPACTION
24 WRINGING EXPERIMENTS USING 2-INCH FOAM CUBES 120
25 PERMANENTLY RETAINED OIL IN 2-INCH POLYURETHANE 125
FOAM CUBES
26 EFFECT OF TIME AND WRINGING CYCLES ON WRINGING 125
PERFORMANCE-
27 EFFECT OF FOAM AGING ON WRINGING PERFORMANCE 126
28 WRINGING EXPERIMENTS USING MULCHED FOAM 127
29 PERCENT OIL REMOVED BY SUCCESSIVE WRINGINGS 130
30 BEHAVIOR OF FOAM DURING RECYCLING 133
31 WRINGING PERFORMANCE OF HIGH VISCOSITY OILS 134
32 WRINGING ATTRITION OF MULCHED FOAM 136
33 OIL DRAINED AS OIL-SOAKED FOAM IS LIFTED FROM 137
WATER
34 OIL CONTAMINATION IN WATER REMOVED BY WRINGING 139
35 PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES 142
OF ON-SITE GENERATED POLYURETHANE FOAM THROUGH
REINCO HAY BLOWER—BEATER CHAINS IN PLACE
36 PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES 142
OF ON-SITE GENERATED POLYURETHANE FOAM THROUGH
REINCO HAY BLOWER—BEATER CHAINS REMOVED AFTER
FIRST PASS
37 PROPERTIES OF SCOTT INDUSTRIAL POLYURETHANE FOAM 143
38 COMPARISON OF MAXIMUM OIL RETENTION BY POLYURETHANE 143
FOAMS AFTER FIVE MINUTES DRAIN TIME WHILE, SUSPENDED
IN AIR
xiii
-------
PAGE
39 PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES 144
OF SCOTT INDUSTRIAL POLYURETHANE FOAM THROUGH
REINCO HAY BLOWER- -BEATER CHAINS IN PLACE
40 PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES 144
OF SCOTT INDUSTRIAL POLYURETHANE FOAM THROUGH
REINCO HAY BLOWER— BEATER CHAINS REMOVED AFTER
FIRST PASS OF FOAM THROUGH BLOWER
41 PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES 152
OF ON- SITE GENERATED POLYURETHANE FOAM THROUGH
WHOLE SYSTEM — TEST OIL: NO. 2 DIESEL FUEL
42 PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES 153
OF ON-SITE GENERATED POLYURETHANE FOAM THROUGH
WHOLE SYSTEM — TEST OIL: 35$ NO. 2 DIESEL FUEL
NO, 6 FUEL OIL
43 THERMAL DEGRADATION OF ON-SITE GENERATED 156
POLYURETHANE FOAM
44 THERMAL HISTORY AND OUTGAS PRODUCT ANALYSES OF 157
ON-SITE GENERATED POLYURETHANE FOAM
45 POLYURETHANE FOAM BURNING RATE FROM FURNACE MODEL 158
xiv
-------
SECTION I
CONCLUSIONS
1. A practical system for recovery of spilled oils of all types in a
wide variety of environmental conditions has been developed based on the use
of polyurethane foam, foamed on site, comminuted and distributed by a
hay blower, recovered by readily available equipment, recycled many times,
and disposed by incineration. An entire system for recovery of oil from
the open sea in 30-knot winds, 5-foot seas, and 2-knot currents can be
mounted on a conventional 150-foot barge. Individual components of the
system are modular and may be air transportable. Oil recovery rates up
to 9,000 gal./hr are feasible.
2. A polyurethane foam formulation has been developed for removing spilled
oil from water. This foam has the following characteristics:
a. Specific surface-permeability balance which results in rapid sorption
of oils of widely varying viscosities. Sorption times of one to
two minutes are adequate.
b. May be foamed and ready for distribution in two to ten minutes
at ambient temperatures from 40°F to 120°F and humidities from
20 to 95$.
c. Easily handled by inexperienced workers.
d. Exhibited no toxic effects on F. Similis (a small sea water fish)
in laboratory tests (Appendix 3).
e. Remains buoyant in water if not wrung, and permanently buoyant
after wringing when oil wet.
f. Shelf life of the mixed components of six months.
3. Disposal of used foam can be readily accomplished using a simple
incinerator, easily constructed in the field from available materials.
Water injection is needed to avoid visible smoke. Analysis of flue gases
demonstrated no detectable deleterious nitrogen compounds or chlorides
from incineration.
4. Comminution of foamed polyurethane buns in preparation for use in oil
sorption can be readily accomplished during the initial distribution
process by use of a commercially-available hay blower. Addition of simple
shredder bars improves the mulching process.
5. Recycling of used foam and redistribution onto the oil slick surface
may be accomplished with a hay blower, but with greater attrition of the
foam than would result from the use of an open-throat centrifugal blower.
-------
6. Oil spill booms can be deployed to divert the oil and foam to a
harvesting unit. Necessary boom length depends on the current velocity
or barge speed.
7. An open-mesh continuous belt can be used to harvest oil-soaked sorbent.
Flights are needed to lift the foam particles positively. The proposed
system will work most efficiently if it advances through the water at a
velocity over 1.5 ft/sec (see also "Recovery of Oil-Soaked Absorbents: An
Engineering Study Based on Modification of Existing Device", Ocean Engineering
Corporation, API Committee for Air and Water Conservation, March, 1972).
8. Removal of oil and water from the oil-soaked sorbent can be effected
practically and simply by gravity-loaded rollers operating against a
1/8-inch mesh chain grate continuous belt conveyor system. Efficiency
of oil removal increases with the number of rollers in sequence; two appear
to be necessary and three may be used. The residual oil remaining in the
foam is on the order of three Ib/lb of foam.
9. The rate of oil recovery by this system (as is also true of other
oil spill recovery systems) depends strongly on oil slick thickness.
Therefore, it is desirable to start cleanup operations as rapidly as
possible, before the oil has spread excessively. The use of a surface
collecting agent (surface tension modifier) is desirable to limit the
spreading of the oil.
10. Unit costs for recovery of spilled oil under the design conditions
(9,000 gal./hr, 0.06 in. (1.5 mm) thick oil layer, 30 mph wind, 5-foot waves,
2-knot currents) are estimated to be about $0.15 per gallon of oil
recovered, for a large spill. Processing of oil and water and disposal of used
foam will result in additional costs.
11, The effluent water contains a significant quantity of oil and will
need further treatment before disposal. The most practical system for
handling of both oil and water appears to be temporary storage in flexible
bags or in bolted tanks on the work barge, with subsequent transport to
a land site for processing and treatment.
-------
SECTION II
RECOMMENDATIONS
1. Construction and testing of the proposed oil spill recovery system
based on polyurethane foam is recommended.
2. Means of separating oil from the effluent water onboard the work barge
should be developed which would allow direct disposal of the effluent
water over the side.
3. Further studies of polyurethane formulations are encouraged in order
to develop foams which:
a. Are preferentially oil wettable
b. Are positively buoyant under all conditions after wringing (oil
wet or water wet)
c. Are resistant to mechanical degradation by the recycling and
wringing apparatus
d. Consist of components with extended shelf lives.
4. Further studies are needed to investigate the effects of emulsions on
sorption rates and capacities and overall system performance.
5. Means of reducing residual oil left on the water surface need to be
explored. One means which warrants further study is the use of surface
collecting agents to maintain a relatively constant oil layer thickness
adjacent to the pieces of sorbent.
6. Studies are needed to adapt the system for use on smaller vessels
and for smaller spills including a modular system specifically for use in
harbors.
7. Means of reducing attrition of the sorbent particles during recycling
should be investigated. These should include:
a. Alternate means of distributing the sorbent after initial
comminution. Such systems might take the form of open throat
blowers, air stream eductors, or mechanical conveyors.
b. Changes in wringer design. Reduced wringing pressure and the use
of two opposing belts with gradually decreasing inter-belt gap
are suggested alternatives (see "Development and Preliminary
Design of a Sorbent-Oil Recovery System", Hydronautics, Incorporated;
EPA, September, 1972).
c. Use of higher-strength foam of satisfactory sorption properties.
-------
8. Additional development of subsystem components is needed, not only
engineering design (e.g., of conveyor belts) but also further experimental
studies of continuous foam generation are desirable.
9. Additional studies of the effects of wind and waves on operational
efficiency of the system as affected by vessel size and shape should be
conducted.
-------
SECTION III
INTRODUCTION
Concept of System
Spilled oil can be recovered from water by use of sorbents which immobilize
the oil so that it can be mechanically harvested. An oil spill recovery
system based on the use of sorbents would be usable in a wide variety of
environmental conditions, because it is compliant. Rather than
resist winds, waves, and currents, the system would move readily under
their influence. Further, an oil spill recovery system based on sorbents
would be capable of handling a wide variety of oil types and spill sizes.
A wide variety of sorbent materials have been used for collection of oil,
ranging from native sponges to hay and straw, wood chips, rice hulls,
and expanded vermiculite. All materials used to date have limitations,
but the most severe limitation has usually been low efficiency, that is,
low weight of oil sorbed per unit weight of sorbent.
It has been shown by many investigators (References 1 and 2) that the
efficiency of an oil spill sorbent is inversely related to its bulk density;
and the most efficient sorbent yet studied is low density, flexible, open-
celled polyurethane foam. Polyurethane foam is sufficiently effective as
an oil spill sorbent that its material cost/effectiveness ratio is
comparable to that of straw or hay. Earlier studies had indicated that
transportation of this material to the job site posed serious logistic
problems; however, liquid ingredients may be transported to the job site,
mixed, and the polyurethane foamed on location, mitigating the logistics
problems associated with handling large volumes of low density sorbents.
Thus polyurethane foam, produced on-site, was chosen in the present study as
the basis of an oil spill recovery system.
Other components of a complete system for oil-spill recovery based on use
of a polyurethane foam sorbent include means for distributing particles
of foam on the spill, concentrating the oil-soaked sorbent, harvesting
the sorbent, removing oil (and water) from the sorbent, and redistributing
the foam for another cycle of the recovery process. Final disposal of
the used foam is a necessary consideration in the total system.
Thus, our initial concept of an oil-spill recovery system, illustrated
in Figure 1, included the following components:
1. Polyurethane foam, foamed on site from a two-part mixture
2. Mixing and foaming equipment
3. Hay blower to break up the cured foam and distribute it
4. Collecting-confining system
5. Harvesting device
6. Wringing or separating equipment
7. Foam disposal unit
-------
FIGURE 1
INITIAL CONCEPT OF OIL-SPILL RECOVERY SYSTEM
-------
Other equipment required includes storage vessels for oil and dirty water
and a means of separating oil from the water.
Design Goals
The objective in this project has been to develop each subsystem of the
initial concept of an oil spill recovery system to provide a firm basis
for design of a full scale system. The conditions which this full scale
system must meet are outlined in Table 1.
Development of the subsystems is described in Sections VI through XIII and
is summarized in Section XIV of this report. Performance of the subsystems
is discussed in Section IV. Performance of the total system is discussed
in Section V, together with limitations of the system performance which
can arise from equipment limitations, from physical limitations (e.g., the
rate of sorption of oil from a very thin slick is limited by transport of
oil over the water surface), and from environmental constraints.
The experimental facilities used in this study included a wave tank, shown
in Figure 2, and a current tank, shown in Figure 3. The wave tank is a
fiberglass lined pit, 50 x 125 x 6 feet, equipped to generate waves up to
two feet in height with a steepness ratio of 0.1. This tank is equipped
for towing tests with a variable speed, double-drum winch on either end
of the tank. The current tank can achieve flow velocities of 8 fps
through the test section, which is 6 feet deep x 6 feet wide and has one
transparent wall for subsurface observation.
-------
TABLE 1
DESIGN GOALS
Source: Contract No. 68-01-0067, Environmental Protection Agency
Protected Waters
Unprotected Waters
Environment
Wave heights, ft
Wind velocity, mph
Currents, knots
Recovery System
Oil recovery capacity, gal./hr
Oil properties
Viscosity
Thickness
Oil-Sorbent Separation
Characteristics of output:
Oil
Sorbent
Water
2
20
6
1,350
Light diesel to
heavy asphalt
0.06 in. (1.5 mm)
H20
Reusable
< H oil
5
30
2
9,000
Light diesel to
Bunker C
0.06 in. (1.5 mm)
H20
Reusable
< 1$ oil
Vessel
Speed, knots at above environ-
mental conditions
Other
General
Other Design Goals
12
Maneuverable
12
8 knots speed in
10-foot seas with
38 mph wind
Adequate size for equipment and
storage needs
Reject floating solids which would
interfere with the efficiency
of, or damages, the recovery
system.
Complete removal of oil from
the water surface.
8
-------
Figure 2 - VIEW OF WAVE TANK
Figure 3 - VIEW OF CURRENT TANK
-------
SECTION IV
PERFORMANCE OF SUBSYSTEMS
Polvurethane Foam
On a weight basis, flexible open-cell polyurethane foam has been found
to be one of the most efficient materials for sorbing oil from water
surfaces. To minimize the cost of storage and transportation a foam
recipe which would allow the fabrication of polyurethane foam at the site
of usage was formulated and is shown in Table 2. This two-component recipe
produces a foam which is 1) oleophilic, 2) low density, 3) buoyant 4) open
cell, 5) flexible, and 6) sorbs oils rapidly relative to the majority of
preformed foams readily available. Further, the foam has a pore size
distribution which makes It a general-purpose foam which rapidly
sorbs and retains large quantities of oils below about 100 cp (centipoise)
as well as very viscous oils of 1000 cp (Bunker C). The foam can be
reused numerous times with wringing and redistribution systems. The
foam components required for this recipe are both non-irritating and easy
to handle, requiring only precautions similar to those necessary in
handling common, volatile hydrocarbon solvents. Foam produced using
this recipe cures rapidly at ambient temperatures from about 40 F to
120°F and relative humidities from about 20$ to 95%. The foam is ready
for distribution 2 to 10 minutes after mixing. Standard foam equipment
capable of pumping and mixing 500 cp to 1000 cp two-component blends at
a ratio of 2:1 produce foams having densities from about 1.5 to 3.0
pounds/cubic foot. Under ideal conditions, this foam is capable of
sorbing oils equivalent to many times the weight of foam used. The
quantity of oil sorbed depends upon many factors, e.g., thickness of
oil film, viscosity of oil, degree of agitation, length of exposure time,
particle size of foam, areal coverage, etc. Properties and sorption rates
typical of this on-site generated polyurethane foam are compared with those
of two preformed polyurethane foams used in the 'furniture industry in
Table 3. Sorption values are shown in Table 4.
Distribution
The polyurethane foam must be distributed over the surface of the oil
spill at concentrations ranging from 0.04 to 0.1 lb/ft2 and at as high
as 3 Ib/gallon of oil. It is anticipated that the maximum distance the
foam must be delivered from a vessel will be less than 50 feet. In
practice, foam will be placed within containment booms or will be spread
upwind since the sail area of each piece causes it to move downwind faster
than the oil spill. In tests of a small power mulcher (Reinco Model
TM7-30, manufactured by Reinco, Plainfield, New Jersey) it was found that
good, controlled distribution could be obtained at distances up to 60
feet in calm air* Against a 7-knot wind, the range was reduced by half.
The mulcher, discharge may be into a conduit, however, and the foam delivered
to a header directly over the water. The entire foam requirement for the
protected water system could probably be delivered by two of these mulchers.
For the offshore system, four of a larger unit (e.g., Reinco Model M60-F6)
would be needed.
11
-------
TABLE 2
RECIPE FOR ON-SITE POLYURETHANE FOAM
Parts by Weight
Component in Blend
6500 MW Polymeric triol with a functionality
of near three (Polyol)1) 100
Dichloromethane 10
Water 5
Tertiary amine catalyst2'
Polymeric Methylenediphenyldiisocyanate3) (MDI) 50-80
1) Thanol®nSF6500 made by Jefferson Chemical Company, Incorporated
2) Thancat- TAP made by Jefferson Chemical Company, Incorporated
Thancat DD made by Jefferson Chemical Company, Incorporated
3) Papi® made by the Upjohn Company
Rubinate M made by Rubicon Chemical Incorporated
Thanate P-30 made by Jefferson Chemical Company, Incorporated
Note: The 6500 MW polyol, dichloromethane, water and
catalyst are blended to make one component-Component B.
Component A consists of (MDI). The ratio of Component A
to Component B is not critical. However, usable foams are
produced at ratios of 1:1.5 to 1:2.5. The best foam is
produced at a ratio of about 1 Component A to 2 Component B.
12
-------
TABLE 3
COMPARISON OF PHYSICAL PROPERTIES AND SATURATION
TIMES FOR FOUR POLYURETHANE FOAMS
Test Condition:
o
Two-inch cubes of foam were placed on the surface of 77 F test oil
contained in two-liter beakers. A stopwatch was started on contact.
The percent of foam not wetted with oil was recorded at various
time intervals.
Foam Samples Description:
1. On site generated polyurethane foam - Batch made 11/2/71
2. On site generated polyurethane foam - Batch made 12/3/71
3. Commercial polyurethane foam used to make pillows, cushions, etc,
4. Commercial polyurethane foam used to make seat cushions
Sample Sample Sample Sample
Properties: 1 2 3 4
Density, lbs/ft3 1.74 2.23 1.11 1.44
Tensile strength, lbs/in2 3.8 3.1 12.0. 11.5
Cell openings, No./in. 58 56 100 100
Compressibility, lbs/50 in2
25$ compression 7 7 17 33
50$ compression 12 11 20 45
65% compression 19 18 27 64
Note: Samples 3 and 4 represent the major quantity of readily
available foam at foam fabricator usable for absorbing oil.
Test Oil Description:
1. No. 2 Diesel Fuel
2. Blend of No. 2 and Bunker C Fuel
3. Bunker C Fuel
Properties; No. 2 Diesel Blend Bunker C
Gravity °API at 60°F 42.3 24.5 10.8
Viscosity at F, cs
-60 2.6 27 2700
771) 2.2 13 1000
80 2.1 12 890
1) Test temperature
13
-------
TABLE 3 (Continued)
Test Time. Minutes
Foam
Saturated
*v
5
10
20
30
40
50
60
70
80
90
95
99
100
Sample
1
(On-Site)
--
—
0.04
—
0.06
--
--
0.10
--
"
"
0.13
No
Sample
2
(On-Site)
._
—
—
0.03
—
0.07
—
--
0.13
..
--
--
0.19
. 2 Diesel Fuel
Sample
3
(Comra'l)
0.13
0.19
—
~
--
0.50
• ^
1.0
--
--
--
--
49.0
Sample Sample
4 5
0.3
..
0.6
2.7
20
120 0.20
1300
5000
..
9000
--
17,000
0.06
Sample
1
(On-Site)
--
0.07
0.13
0.16
--
0.25
--
0.33
0.43
0.50
0.60
--
0.93
Blend Fuel
Sample
2
(On-Site)
0.03
--
0.09
--
0.16
0.23
0.31
0.25
0.55
--
—
0.70
Sample
3
fCotnm'l)
—
25
40
50
55
65
90
110
--
--
270
--
315
Sample
4
(Comm'l)
60
240
--
—
--
1300
—
--
4300
—
--
--
7200
Sample
5
(Scott)
0.05
0.10
--
0.15
0.25
--
0.33
0.40
--
--
--
0.70
Sample
1
fOn-Site)
1.5
3.0
7.5
13.0
21.5
27.5
30.0
33.5
42.0
49.5
54.0
63.5
68.0
Bunker C Fuel
Sample
2
(On-Sitel
--
3.0
7.5
13.0
20.0
27.0
31.5
41.5
45.0
50.5
55.0
—
67.0
Sample
3
jCpmittVQ
180
1,200
4,300
9,000
18,000
--
--
--
—
—
—
--
--
Sample
4
(Comm'l)
360
9,000
18,000
--
--
--
--
--
--
--
--
--
--
Sample
5
(Scott
2
2.5
5
8
11
—
20
26
--
--
--
39
-------
TABLE 4
COMPARISON OF MAXIMUM OIL SORPTION
DATA FOR TWO POLYURETHANE FOAMS
Experimental procedures are described in Table 3.
Drain Time
5 min
15 min
30 min
1 hr
2 hr
Oil Held by Material lb Oil/lb Material
Commercial Foam
On-Site Generated Foam
No. 6 Fuel Oil Sp Gr = 0.996
5 min
15 min
30 min
1 hr
2 hr
53.6
51.2
50.6
47.0
45.3
Shallow Yates
29.9
29.2
28.8
28.2
28.0
Crude Oil Sp Gr = 0.905
36.3 22.9
35.4 22.1
34.6 22.0
34.1 21.8
33.7 21.5
Foam Description;
1. Commercial Foam TDI - Polyether
2. On-Site Generated Foam MDI-- Polyol
Properties; 1 2
Density, lb/ft3 1.1 2.3
Pores per inch 100 60
15
-------
Tower mulchers are readily available for emergency mobilization (Reinco
has listed several hundred that are in use throughout the Midwestern and
Eastern parts of the United States) and can be modified quickly at the
job site. Recommended modifications before use for mulching and initially
distributing the buns consist of the addition of several studs in the
beater chamber wall to control the foam particle size and the provision
of an extended charging chute as an aid in increasing the throughput
(especially for foam which has been pre-mulched or is being recycled).
An alternate means for distributing the foam during recycling (open
throat blower, air stream eductor, or mechanical conveyor) should be
considered to reduce further attrition of the sorbent particles after
the initial mulching.
Collecting
The foam will be distributed within or directly before the mouth of
a channel formed between two converging booms or one boom and the barge
upon which the processing equipment is located. The required boom
angle to the current will be very shallow (i.e., much less than 30°)
and foam containment should be effective up to a velocity of at least
three ft/sec. From experiments, continuous flow of the oil-soaked sorbent
through this channel should be realized despite high areal concentration
of the sorbent. The sorbent will then flow through the confined channel,
as indicated in the photograph of the test booms shown in Figure 4,
•Figure 4 - TEST BOOM ASSEMBLY IN WAVE TANK
16
-------
to an inclined wire mesh belt harvesting apparatus where it will be
lifted from the water onboard the barge. The experimental harvesting
apparatus is shown in Figure 5.
The tow tension in each boom should be taken up at several points along
its length to allow it to conform to the water surface, minimizing splash-
over and broaching of the booms between wave crests. Sharp corners on
floats or at boom ends and steep inclinations of the boom to the current
should be avoided to minimize wash under of the sorbent. A properly
designed system with an included angle of 30 or less, a freeboard of
two feet and a draft of one foot should be effective to velocities of
three ft/sec if the booms are sufficiently flexible to conform to the
waves.
Wringing
The wringing experiments described in Section XI demonstrate a simple
wringer concept to be capable of removing a satisfactory amount of oil
at conveyor speeds approaching 100 ft/min and at imposed wringer pressures
from 7 to 30 Ib/in. of wringer width using simple pipe rollers. These
experiments also demonstrate that there is a residual amount of oil
remaining in the foam after extensive wringing which must be recycled with
the foam. The ratio of the weight of the residual oil to that of the dry
foam varied from three to six, the largest values being for the most
viscous oils. Some foam attrition was observed for the highest visco-
sity oils tested, but even with 1100 cs oil only 2% to 3% foam per cycle
is reduced to a size smaller than that which will pass through a one-
inch mesh by the wringing process along.
The oil removal system proposed consists of two sets of gravity-loaded pipe
rollers through which the foam is successively wrung. Foam would be
carried to and through each set of rollers on a wire mesh belt conveyor
as shown in Figure 6. Sufficient wringing pressure can be obtained by
making the top rollers of 24-inch diameter pipe with 1/2-inch wall thick-
ness filled with water. The heavy top rollers can be left free to move
vertically within guide bars rather than constrained to operate at a
constant gap. In this way the wringer is flexible and can accommodate
mulched foam layers of varying thickness as well as debris which the
harvester may pick up. Experiments with this design have shown that
satisfactory oil removal can be obtained with layers of mulched foam up
to six inches deep wrung at a speed of 70 ft/min.
Foam Disposal
Generally, it is necessary to dispose of the polyurethane foam after it
has been used to sorb spilled oil. Techniques which might be used
without creating additional pollution problems were considered. Disposal
methods which were investigated included solution, compaction, and
burning. Burning appears to be the most rapid and practical method of
disposing of used foam. To accomplish this a furnace to burn used foam
without producing appreciable quantities of particulate emissions was
17
-------
(a)
(b)
Figure 5 - MODEL HARVESTER INSTALLED IN CURRENT TANK
18
-------
Figure 6 - EXPERIMENTAL WRINGING APPARATUS
19
-------
built and operated. The model furnace is shown in Figure 7. Burning
rates are presented in Table 5.
Figure 7 SHELL PIPE LINE MODEL FURNACE USED
TO BURN POLYURETHANE FOAM
Burning tests in the 6.25-inch diameter model furnace indicate:
1. Polyurethane foam used to sorb Bunker C and No. 2 Diesel fuels
from water can be burned at rates from 10 to 20 pounds per
hour per square foot of grate area (on the basis of dry foam)
without producing smoke.
2. The burning rate for dry, unused foam is about 40 pounds per
hour per square foot of grate area.
3. Water, either added while burning foam or absorbed while sorbing
oil from water, greatly reduces the particulate emissions. When
20
-------
TABLE 5
POLYURETHANE FOAM BURNING RATE FROM FURNACE MODEL
Description of
Foam
Dry
Water Wet
No. 2 Diesel
Wet
No. 2 Diesel
Water Wet
Bunker C
Water Wet
Bunker C, No. 2
Diesel Mix -
Water Wet
Burning Rate, Pounds
of Dry Foam per Hour
per Square Foot of
Grate Area
36
14
9
15
8
14
Flame
Temperature
°F
1400-1500
1300-1400
1400-1500
1300-1400
1200-1400
1400-1500
Stack
Temperature
F
1400-1500
1300-1400
1400-1500
1300-1400
1200-1500
1400-1600
Auxiliary
Fuel Con-
sumpt ion
SCF/Hr
42
42
42
42
68
68
Unburned Fo<
or Oil Lost
as Dripping
$w
Nil
Nil
Nil
Nil
6
1.5
-------
either dry foam or oil-soaked foam is to be burned, water should
be sprayed on the foam prior to burning. The quantity or rate
can be established by trial.
4. No deleterious nitrogen compounds or chlorides were detected
by an analysis of flue gases evolved during foam incineration,
as indicated in Table 6.
5. Additional fuel is required to fire an igniter and an afterburner.
6. A furnace to burn used polyurethane foam can be constructed in
the field utilizing readily-available materials and manpower.
TABLE 6
FLUE GAS ANALYSIS OF EVOLVED GASES WHILE BURNING
FOAM CONTAINING NO. 2 DIESEL FUEL
Component Volume,
Carbon dioxide 5.0
Argon 1.0
Hydrogen sulfide 0.0
Oxygen 14.2
Carbon Monoxide 0.0
Nitrogen 79.3
Hydrogen 0.08
Helium 0.0
Methane 0.08
Ethane and heavier hydrogens 0.04
Acetylene 0.06
Water 0.02
Note: No cyanides, isocyanates, or chlorides were detected in
the flue gas.
22
-------
SECTION V
SYSTEM PERFORMANCE
Prototype Design
The oil recovery system now proposed is fundamentally the same as the
original concept of Figure 1. This study has increased our under-
standing of the processes involved and led to refinements in certain
parts of the system. The most obvious and significant changes have
been the reduction in size of the boom array needed and the use of a
redistribution system other than the mulcher-blower to reduce
attrition during recycling. The reduction in boom length has made it
practical to consolidate the system into a configuration which may be
placed on a single large flat-deck barge, or, with benefit of the
modular design concept, deployed as dual systems aboard smaller barges
or work boats (see Figures 8 and 9). It should be emphasized that
consistent effort to remain conservative in extrapolations from lab-
oratory to prototype has doubtless resulted in a system having a
greater capability than specified.
As shown in Figures 8 and 9, the foam is placed only within the area
confined by a boom, and little or no loss of foam to the sea is to be
expected. Any floating debris which may pass through the boom throat
will be accepted by the harvester. Sufficient time is provided in
transit on the harvester and wringer feed conveyor for inspection and
for manual removal of debris. Small debris will not damage any part
of the system except, possibly, the blower. The quantity of water
contaminate recovered with the oil is strongly a function of slick
depth and oil properties. To assure consistent attainment of the
design goals for effluent purities, recovered liquids are treated at
oil-water separation facilities on shore, where the operation is
efficient and the effluent can be well controlled (see page 137).
With a battery of storage containers manifolded on board, however, it
may be desirable to segregate the oil-rich wringing effluent from the
generally oil-free water drained prior to wringing.
Vessel motions are not expected to restrict operations to any great
extent within the range of sea conditions specified. The recovery
vessel is generally expected to operate approximately parallel to the
direction in which the oil spill is moving and thus be working into
the wind and waves. The pitch and heave motions under these conditions
should not affect the system or the personnel. Consideration must be
given to the proper securing of the liquid storage tanks, however, since
large inertial forces will be generated by vessel motion when the
containers are in use.
As large barges of the type preferred are available for charter in most major
shipping areas, but not in all potential oil spill areas, all components
23
-------
FIGURE 8 - PROPOSED CONFIGURATION
24
-------
FIGURE 9 USE OF COMPONENT MODULES FOR REDUCED CAPACITY
SYSTEM
25
-------
of the system are modular and may be employed in various combinations
to suit the availability of vessels. Two half-size systems can be
assembled on work boats, small barges, etc. without modification to the
equipment itself. The only component of this system that would require
disassembly before truck shipment is the recycling conveyor, and this
can be designed for quick assembly from palletized packages (and could
be stored in that fashion, even on a barge). In practice the entire
system could be maintained for immediate loading on trucks or aircraft
similar to the C-130.
The cost of oil recovery using this polyurethane foam based system
including the transportation of recovered effluents and used foam to
shore is estimated to be on the order of $0.15 per gallon of recovered
oil, as shown in Table 7. The cost of operating shore-based treatment and
disposal facilities will increase this cost somewhat.
TABLE 7
ESTIMATED COST OF OIL RECOVERY USING
POLYURETHANE FOAM
Basic Costs
1 - 200' Barge
1 - 1,000 hp Tug
30 man-days labor/day at $7.50/hr
$7,670
day
Basic System
hr
$ 650/day
900
5,400
$7,670/day
9,000 gal. oil
$0,036/gal. oil
Support Vessels
3 - 600 hp Tugs at $30/hr
3 - Cargo Barges ~ 140*±
2 - Crew Boats
Material (Sorbent) Costs = i-
(Assumes 10$ loss/pass and
3 Ib/lb oil sorption —•
See Section XII and
Appendix 2)
$2,160
1,200±
600±
$3,960/day-$0.018/gal. oil
gal.V(7.5 lb/gal.W0.10).($0.37/lb) _
3 Ib/lb
$0.092/gal. oil
Total
$0.146/gal. oil
26
-------
Limit atjLons of System
Limitations to performance of this oil spill recovery system can arise
from environmental conditions, physical limitations inherent in the
oil/water/sorbent system, and limitations in the equipment items needed.
The rate of transport of oil on water to a sorbent particle and the
rate of sorption of oil by a given sorbent particle depend on the oil
layer thickness. Therefore, the overall rate of oil recovery by the
system will decrease with thickness of the oil layer as is true with
most recovery systems. High oil viscosities will result in a slower
rate of transport of the oil on the water surface and a slower rate of
migration of oil into the foam particles. This may be partially com-
pensated by increased retention of adsorbed oil on the outside of the
foam particles with increased viscosity (see page 62).
A sorbent-based oil spill recovery system may be less sensitive to the
effects of wind, waves, and currents than mechanical systems, because
the sorbent-based system is generally compliant. Nevertheless, adverse
environmental conditions will affect performance. In addition to
problems of ship and boat handling and problems of operating on the
deck of a moving vessel, the following limitations are expected;
a. Wind will affect the ability to distribute foam uniformly
and will affect the distance over which it can be projected
by a blower without significant losses. Further, wind will
tend to move the foam over the water surface until it absorbs
oil and water so as to sink partially, decreasing the free-
board upon which the wind may act and increasing the draft
and drag of the particles. These effects of wind can be
minimized by use of a distribution manifold cantilevered
from the bow of the work barge over the area on which foam is
to be dispensed.
b. Waves may increase the efficiency of the sorbent and the
performance of the harvester. However, the harvester design
should be such that the lower end will not broach out of the
water and the booms used to divert oil-soaked sorbent to the
harvester should be relatively insensitive to waves.
c. Rain will probably interfere with the foaming operation,
although this can be controlled by use of shelters.
d. Temperature will affect the foaming reaction, although our
studies indicate that the selected formulation is usable over
a wide range (at least from 40° to 120°F). At low temperatures,
it may be desirable to heat the foam components prior to
mixing in order to speed up the foaming reaction.
27
-------
e. Currents pose significant operating problems for boom systems.
Failure of booms to contain oil and sorbents and mechanical
breakage may occur at high current velocities. A major concern
for the proposed recovery system at high currents is to provide
enough contact time between the foam and the oil. It should be
noted that there is a relative velocity below which the wet
sorbent will be kept away from the harvester by circulation
patterns produced by motion of the harvester belt (see page 106).
A primar}* limitation on the performance of this system arises from the loss
of foam due to attrition (see Section XII). There are two primary sources
of attrition, the foam transport system and the wringer. If the Reinco
hay blower is used both for preparation of the foam (comminution) and
foam transport, significant generation of fine particles is likely with
each cycle. This attrition can be minimized by use of two separate
systems, one for foam comminution (the hay blower), and an open-throat
centrifugal blower or mechanical conveyor for foam transport during recycling,
Attrition during wringing is especially severe for the more viscous oils,
over 100 centistokes (cs), owing to high internal pressures generated
within the foam as the oil is forced out. The pressure generated is a
function of the oil viscosity and the rate of oil removal; this suggests
that a modified wringer configuration (two opposing open-mesh belt, with
gradually reducing gap between them, as suggested by Hydronautics, Inc.,
see page 3) could be used to minimize this source of attrition,
Sinking of the foam is not expected to be a severe problem, because foam
which has become oil-wet will apparently remain permanently buoyant.
Minor losses of foam can be expected from sinking of foam which has been
exposed to water only, wrung out, and then redistributed onto the water
surface (see page 31).
28
-------
SECTION VI
FOAM FABRICATION AND CHARACTERISTICS
Introduction
To minimize the cost of storage and transportation it is desirable to
fabricate polyurethane foam at the site at which it will be used. The
foam should have the following properties:
1. Oleophilic
2. Low density
3. Positive buoyancy
4. Open cell
5. Flexibility
6. Rapid cure under a wide variety of conditions
7. Require simple and rugged equipment to produce, disperse, and
harvest
8. Be easy and non-hazardous to produce
A polyurethane foam possessing these properties has been developed.
On-Site Foam Formulation
Components in the recipe described in Table 2 are both non-irritating and
easy to handle, requiring only precautions similar to those which should
be taken when handling common volatile hydrocarbon solvents. This formu-
lation cures rapidly at any ambient temperature between 40 F and 120°F
and relative humidities between about 20 and 95%. The foam is ready for
distribution two to ten minutes after mixing and has a density between
1.5 and 3 Ib/cu ft. This foam can be made and distributed at the site
of an oil spill.
The reactions involved are described in Appendix No, 1. About two hundred
blends utilizing these and other components were made and tested. One
formulation was consistently preferred. Details concerning a part of the
formula testing are described in Appendix No. 2.
Foam Properties
The properties of individual batches of foam vary depending upon a) climatic
conditions, b) mixing conditions, and c) component ratio (Table 8); however,
no batch of foam has been produced, using recipe in Table 2, which was
unsatisfactory for absorbing oil from water. The density of the foam is
the property most influenced by the above variables; it varied between
1.7 lb/ft3 and 2.7 lb/ft3. The rise time and cure time to tack-free
increased as the temperature decreased. Foam having a density of about
2.1 lb/ft3 is typical. Properties of this foam are shown in Table 8.
Pore size and distribution are indicated by the foam cross-section shown
in Figure 10.
29
-------
TABLE 8
FOAM PRODUCT!ON RATES AND FOAM PROPERTIES
Conditions: Polyurethane foam produced from recipe shown in Table 2 using
the following equipment:
Graco Hydracat Variable Pumping Unit
Consists of President Model 205-038 Series D Air Motor
which drives two Graco Size 2 displacement pumps
mounted on a portable frame, with associated filters
and hoses.
Binks 18 FM gun with flush equipment consisting of a 5-galIon
Monark Hydra Spray unit, Model 226-153 Series "A".
Foam applied to kraft paper to form bun about IS inches wide.
Components were near ambient temperature except for December 3
production. In this case components were near 70 F.
Nov. 2, Nov. 29, Dec. 3, Jan. 11, June 21,
Date; 1971 1971 1971 1972 1972
Ambient Temperature, °F 90 60 40 45
Relative Humidity — -- — 80
Pour Rate, Ib/hr 307 360 330
Density, lb/ft3 1.7 2.2 2.2 2.1 2.15
Tensile Strength, lb/in2 43 3 3.1 3.2
Pores Per Inch 58 60 56 60 47
Compressibility, lb/50 in2
25$ Compressed 8 7 7 15 5.5
50$ Compressed 13 10 11 27 8.4
65$ Compressed 21 14 18 44 13.6
30
-------
Figure 10 -
TYPICAL ON-SITE GENERATED POLYURETHANE FOAM
PRODUCED AT A5°F AND 80% RELATIVE HUMIDITY
JANUARY 11, 1972 (2 X 2 INCH GRIDE WITH SUBDIVI-
SIONS OF 0.1 INCH)
The foam described in Table 8 and Figure 10 was given to the Edna Wood
Laboratories, Houston, Texas for a bioassay. Foam was added to test tanks
in quantities equivalent to 0.01, 0.11, 0.60, 1.1, and 2.2 inch thick
layers of foam on 3-foot deep water. Killlfish (Fundulus Slmilis, a sea-
water species) in the test tanks were exposed to the foam for 96 hours.
At the end of the test period all fish were normal. Details concerning
these tests are shown in Appendix No. 2.
Tests were undertaken to determine the persistance of the foam buoyancy.
Typical data for the sinking of dry foam into quiescent sea water are
shown in Figure 11. Foam which has been oil wetted by application to an
oil slick on water does not sink after subsequent wringing and reuse cycles.
Dry foam mulched by a modified Reinco hay blower (Model TM 7-30) was
soaked in water and then passed through our model wringer. Upon returning
the foam to the water, 2% sank. After a second wringing cycle, an additional
5% sank. In four subsequent cycles no more foam sank. A thin oil slick
was added to the water surface. The water-wet foam, including that portion
which had sunk earlier, was applied to the oil slick. All foam remained
afloat. The foam was then passed through four cycles of wringing and
31
-------
OJ
to
(-1
3
on
J-i
u
u
cfl
0-
o
OJ
EC
It)
o
100
* 80
60
0
8
12 16
20
24
28
32
Test Time, Hr
FIGURE 11 -
SINKING RATE OF ON-SITE GENERATED POLYURETHANE FOAM
INTO QUIESCENT SYNTHETIC SEA WATER
-------
reapplication to water only (no oil slick). The foam was then left on
the water for 138 hours with no loss due to sinking. These experiments
demonstrate that fully dry foam and foam that has had some exposure to
oil will not sink. Those foam particles which did sink when fully saturated
with water had no surface "skin". Evidently the "skin", which is a result
of the interaction between the exterior foam surface and the air when the
foam is made, provides a sufficient number of closed cells to prevent
sinking.
Test data presented in Table 3 show that on-site generated foam absorbed
oils more rapidly than typical commercially-available foam such as used
in the furniture industry. This is due in part to the larger pores and
consequent higher permeability of on-site foam. These data also show
that on-site foam absorbed No. 2 Diesel Fuel about 450 times more rapidly
than Bunker C Fuel which-was about. 450 times more viscous at test condi-
tions. When on-site foam was held stationary in both No. 6 Fuel oil and
Shallow Yates crude oil, capillary forces saturated the foam to a height
of 0.25 ± 0.03 inches above the level of the oil.
Data in Table 4 show that, once saturated, the commercial foam contained
and retained more oil than on-site foam. The difference between the bulk
densities of the two foams is the most likely explanation for the difference
in sorption capabilities.
During tests made to determine the ability of the foam to remove thin oil
slicks from water, it was noted that foam placed on a thin slick removed
the oil in the immediate area. The absorption rate then tended to exceed
the gravitational flow transfer of oil from the surrounding slick to the
foam. In some cases the oil would not approach the foam, leaving it in
an area free of oil. (This was also observed to happen when samples of
furniture-industry foams were placed on a similar thin slick). This
tended to decrease the oil-to-water ratio in the total liquid sorbed and
to decrease the oil-sorbed-to-foam-weight ratio.
Additional tests were completed to:
1. Evaluate on-site generated foam in relatively constant thickness
oil slicks,
2. Evaluate on-site foam when an excess of oil was present.
3. Evaluate the effect of surface collecting agents.
The data in Table 9 show that the on-site generated foam sorbed about
the same quantities of low viscosity oils as did polyurethane foams
tested by others (see Reference 2). The sorption of Bunker C was lower
than values previously reported for similar oils because of the short
exposure time and lack of agitation while obtaining values reported in
Table 9. The primary difference between the data reported here and those
reported by others was that the exposure period, allowed in the present
series of tests is only one minute with no agitation, whereas an exposure
33
-------
TABLE 9
SORPTION EVALUATION OF ON SITE GENERATED FOAM
UTILIZING SURFACE-COLLECTING AGENTS
Foam Description:
Density, lb/ft3 2.1
Pores/Inch 51
Tensile Strength, lb/in2 3.1
Compressibility, lb/50 in2
25$ Compressed 15
50$ Compressed 27
65% Compressed 44
Oil Description;
1. No. 2 Diesel Fuel
2. Blend of No. 2 Diesel and Bunker C Fuels
3. Carnea 21 Oil
4. Bunker C Fuel
Properties:
Gravity, API at 60°F
Viscosity at °F, cs
60
77
80
Test Conditions:
Test Oil No.
2.6
2.2
2.1
27
13
12
25.1
75
40
36
2700
1000
890
Water (75 F) contained in a 3-ft diameter reservoir was treated with
four drops of Oil Herdeir placed 90 apart and at the reservoir wall.
A measured 250 milliliters of oil were poured into the water at the
center of the reservoir. The diameter of the lens was measured after
the oil had collected into a near perfect circle (about 15 minutes).
An accurately weighed quantity of foam (about 10 grams for large pieces)
was applied to the oil lens near the center. A stop watch was started
when the foam was applied. The foam was removed 1.0 + 0.01 minute after
being applied, and placed in a tared container. The total weight of
liquids sorbed was determined immediately. The foam was squeezed vigor-
ously by hand. The recovered water and oil were separated and measured.
The oil remaining on the water was recovered and measured.
34
-------
Ui
Test No.
No.
1 1
2 1
3 1
4 1
5 1
6 2
7 2
8 2
9 2
10 2
lll) 2
12 2
.13 3
14 3
15 3
162) 4
172) 4
182^ 4
Specific
Gravity
0.808
0.808
0.808
0.808
0.808
0.901
0.901
0.901
0.901
0.901
0.901
0.901
0.899
0.899
0.899
0.989
0.989
0.989
Oil
Viscosity,
cs
2.3
2.3
2.3
2.3
2.3
13
13
13
13
13
13
13
43
43
43
1000
1000
1000
Thickness,
mm
0.055
0.082
0.090
0.055
0.082
0.21
0.18
0.20
0.16
0.35
0.35
0.15
0.25
0.25
0.25
0.20
0.20
0.20
Foam Cube
Dimension,
In.
2
2
2
1
1/2
2
1
1/2
2
2
2
1
2
1
1/2
2
1
1/2
Oil /Foam
Gal./Lb
2.34
2.22
2.34
2.34
2.46
3.04
4.45
3.74
3.28
3.04
3.04
4.10
1.76
3.74
3.74
0.35
0.58
1.05
Sorption
Water/Foam
Gal./Lb
1.52
1.05
1.87
1.52
0.82
0.18
Nil
Nil
Trace
Trace
0.70
Trace
0.12
0.09
Nil
Nil
Nil
Nil
Values
Oil to Oil to
Water Foam
Ratio Ratio
(Weight)
1.6 16
2.1 15
1.2 16
1.6 16
2.4 17
17 25
34
28
26
23
4 23
31
14 13
38 28
28
3
5
9
8
PI
ri
o
i-t
|
D.
1) 2" cube completely soaked in 90 seconds. Left in 0.12-inch (thick) slick on water with about 1.8-inch
extending into water for 18 hours.
2) N'ot Collected because oil would be in excess of 1 inch thick. Foam only partially sank; thus, oil-to-foam
ratio is a function of surface area of foam only, because essentially no oil was imbibed into the center
of the cube.
-------
period of fifteen minutes with agitation was used in the tests of
Reference 2. When the foam was applied to the central portion of an oil
lens which was maintained at an equilibrium thickness by a surface-
collecting agent, the surface tension gradient established by the agent
continuously drew the oil into the area occupied by the foam so that the
oil thickness adjacent to the foam remained nearly constant, resulting in
increased oil sorption by the foam. Thus, & dynamic system, utilizing either
diversionary booms or surface collecting agents to continuously concentrate
the oil, will result in more efficient performance of a foam spill recovery
system than data obtained under static conditions.
Data in Figure 12 and Table 10 show compressive values for two samples of
foam at different dates. Though the crosslinking reactions are near
completion in one or two minutes after the foam components are well mixed,
further reactions continue for a long period of time and increase slightly
the rigidity of the foam.
TABLE 10
COMPRESSIBILITY OF ON-SITE GENERATED POLYURETHANE FOAM
Test Date Compressibility, lb/50 in2
July 22 Foam
251
10
13
spl
13
19
651
22
29
November 29
25$ 501
__
7 10
Foam
651
—
14
July 27
December 1
February 9 — — — 12 17 25
Foam Production Equipment
The two-component polyurethane foam recipe described in Table 2 has been
produced both by hand mixing and commercial foaming equipment.
Commercial equipment used was a Graco Hydracat Variable Pumping Unit equipped
with a Binks 18 FM gun (Figure 13). This unit consists of a President Model
205-038 Series D Air Motor driving two Graco Size 2 displacement pumps (one
variable) mounted on a portable frame, with associated filters and hoses.
A necessary part of the equipment is the gun flush equipment consisting
of a 5-gallon Monark Hydra Spray unit Model 226-153 Series "A" and hose
(Figure 13 extreme left).
The displacement pumps are attached to 55-galIon drums of components
mounted on portable barrel racks (Figure 13 rear). The equipment is driven
by compressed air (40 to 140 psi) at a rate of about 15 CFM. With a
Binks 18 FM gun (Figure 14, left) this equipment pours 300 to 360 pounds
of foam per,hour (5 to 6 pounds per minute). Use of a mixing device
(Figure 14) might increase the pouring rate. Untrained personnel have oper-
ated the equipment and have made good quality foam with only 30 minutes of
instruction.
36
-------
50
40
30
•o
td
o
,J
r
o
10
I I I I
t I I
5 6 7 8 9 10 20
Aging Time, Days
30 40 50 60 70 SD 90100
FIGURE 12 - AGING TIME VERSUS COMPRESSIVE LOAD TO OBTAIN 25$,
50$, AND 65$ COMPRESSION OF ON-SITE GENERATED
POLYURETHANE FOAM
-------
^
•
• • • . •
Figure 13 - PORTABLE FOAMING EQUIPMENT
Figure 14 - MIXING HEADS USED WITH PORTABLE
FOAMING EQUIPMENT
38
-------
Larger scale equipment capable of producing 40 to 50 pounds of foam per
minute was also used. Any of the currently-available foam equipment
which can pump 500 to 1000 cp materials and blend through a mixing head
at a ratio of two parts Component "B11 (polyol) to one part Component "A11
(isocyanate) may be used.
Foam Production
A typical foam production operation is shown in Figure 15. On this day
the ambient.temperature was 60 F. Foam was being poured at a rate of
360 pounds per hour. The rise time was about 45 seconds and the foam was
tack free in four minutes. The resulting 2.2 lb/ft3 foam is shown in
Figure 16. When large batches (over one ton) are required the foam might
be poured on a moving belt as described in Figure 17.
Foam was made utilizing a Polymer Services Corporation foam unit (Figure 18)
The mixed components were applied at a rate of about 40 Ib/min to plastic
sheets and paper spread on the ground. The ambient temperature was 55 F
and a light mist was falling. A 2.5 lb/ft3 foam suitable for sorbing oil
was produced.
Once produced, foam can be stored in rolls as shown in Figure 19, until
needed for mulching and distribution as shown in Figure 20.
Natural Degradation
We have noted that polyurethane foams, exposed to sunlight, degrade with
time. Qualitatively, the on-site generated foam appears to degrade more
rapidly than the commercially-available foams, becoming friable and
easily crumbled. Thus, it appears that on-site foam should be more easily
attacked by bacteria and converted to C02, water, etc. We believe this
an advantage, because any foam which is lost from the system will be
decomposed by natural processes in a shorter time than commercially-
available polyurethane foam.
The prepared components for on-site foaming have an observed shelf life
of six months before becoming insufficiently reactive for satisfactory
foaming. This life might be extended by the use of catalysts during the
foaming operation; however, we have not studied possible catalysts.
39
-------
Figure 15 - POLYURETHANE FOAM PRODUCTION UTILIZING A
PORTABLE GRACO HYDROCAT UNIT
Figure 16 - POLYURETHANE FOAM PRODUCED UTILIZING
PORTABLE GRACO HYDROCAT UNIT
40
-------
ADJUSTABLE BELT
FOAM
MIXING
HEAD
FOAM
APPLIED TO
MOVING BELT
FOAM REMOVED
FROM UPPER
BELT BY
DOCTOR BLADE
CURED
POLYURETHANE
FOAM
DOCTOR BLADE
FOR REMOVING
FOAM FROM BELT
BELT PLATFORM
VARIABLE SPEED DRIVE
FIGURE 17 - SCHEMATIC OF CONTINUOUS BELT FOR MAKING POLYURETHANE
AT SITE OF OIL SPILL
-------
Figure 18 - CONTRACT FOAM EQUIPMENT
42
-------
1
Figure 19 - POLYURETHANE FOAM BUN FORMED INTO ROLL
Figure 20 - FOAM BUNS READY FOR MULCHING UTILIZING
A REINCO HAY SPREADER
43
-------
SECTION VII
SORPTION
Fundamental to the operation of a marine oil recovery system utilizing
a sorbent material is the sorption of oil from a slick by a sorbent
floating on or near the water surface. The nature of the process of
absorption of oil from a slick by a water-saturated oleophilic block
may be understood by considering an elementary, two-dimensional analysis
based on the simplified illustration of Figure 21. The foam particle
FIGURE 21 -
ABSORPTION OF OIL FROM A SLICK BY A POROUS
OLEOPHILIC BLOCK
may be assumed to have a circular equivalent of radius, re, into which
e,
oil is absorbed uniformly about the circumference at a total rate Q.
Ignoring the vertical diffusion of oil through the matrix, the oil flux
with the material is
q =
2rrrd,
(1)
45
-------
where r is the local radius, and ds is the depth of the slick, assumed
to be the depth of the oil layer within the block. The pressure
gradient within the block then becomes
dP
dr
2rrrdoK
(2)
where |j, is the oil viscosity and K is the permeability of the sorbent
material. This equation may be integrated, subject to the boundary
conditions
= 0.; P|
r = r,
(3)
where A a is the interfacial driving force, % is the effective specific
surface, and $ is the porosity of the sorbent material (see Reference 3),
When this is done the result may be solved for Q.
2rrdsAo- KSe
Noting that the volume of oil may be found from the relation
Q = TT
-------
and
t =
gdsK2e\
0 r| ) t
This relation is plotted in Figure 22,
(9)
1.0
0.9 -
0.8 -
0.7 -
0.6
0.5 -
0.4
0.3
0.2
0.1
0.05
.10
0.15
0.20
0.25
FIGURE 22 - DIMENSIONLESS VOLUME ABSORBED FROM A SLICK BY A
WATER SATURATED OLEOPHILIC BLOCK AS FUNCTION OF
DIMENSIONLESS TIME
47
-------
A series of experiments was undertaken to characterize the performance of
polyurethane foam as an oil sorbent. During the initial experiments,
material samples consisting of cubes of uniform dimension were soaked in
slicks of No. 2 diesel oil in the glass-walled tank shown in Figure 23.
The experimental procedure used is described in Table 11, Procedure A.
Variations in sorption characteristics of the order of 10$ were observed
between the on-site generated foam materials produced at different times.
During these early experiments with previously unused foam, it was found
that a large fraction of the fluids absorbed (of the order of 20$) is
retained within the blocks of foam even after extensive hand and mechanical
squeezing. This material consists primarily of oil. Some fractionation
of the oil may also occur within the foam, particularly in the case of
crude oils, as in some cases it was found that the residual material
remaining within the foam after normal squeezing had a specific gravity
about 5$ less than the original oil.
In experiments with heavier oils, whose properties are shown in Figures 24
and 25, the soaked foam was squeezed mechanically with a pressure of
170 lb/ft2 in the "mule's foot" squeezer of Figure 26 and the volume of
effluent recovered together with its oil content were measured directly
as outlined in Table 11, Procedure B. No correction was made for the
residual material remaining within the foam matrix after squeezing, as
it was assumed that commercial squeezing operations would be similarly
inefficient. Results of experiments conducted with Carnea 15 (a refined
oil) are shown in Figures 27 through 30. From Figures 27 and 28 it can
be seen that total volumes and oil volumes recovered increased with slick
depth for any soaking time up to the maximum of 20 minutes tested. Oil-
water ratios of the recovered effluent also improved with increasing
slick depth. Improved performance might be expected for thin slicks in
the presence of waves or currents, because it was observed that oil-free
areas may appear about foam particles within thin slicks in the tank.
To determine the effects of recycling on foam performance, foam samples
were presoaked in oil or water and then squeezed near-dry for use in
the sorbent tests. Results from these experiments are shown in
Figures 29 and 30. It may be seen that prior exposure to either oil or
water generally results in relatively poor performance for soaking times
of less than about 15 minutes (900 seconds).
The results of the soaking experiments conducted with the other oils
whose properties are included in Figures 24 and 25 in a slick which was
initially 0,05 in. (1.3 mm) deep are shown in Figure 31. These figures
indicate that recovered volumes of both total effluent and oil may be
generally high for the light oils but may reach a minimum at an oil
viscosity of about 30 to 40 cs. In Figure 32 the volume fraction of oil
varies from approximately 20 to 30 percent. Again, though results are
mixed, performance is generally better in the cases of the lighter oils
tested in this series; however, performance appears to reach a minimum
at an oil viscosity of about 30 - 40 cs and would improve for higher
viscosities.
48
-------
a) Soaking foam
b) Removing and draining foam
.FIGURE 23 - BENCH-SCALE SORPTION TEST APPARATUS
49
-------
TABLE 11
PROCEDURES USED TO STUDY OIL SORPTION BY FOAM BLOCKS
Equipment
1. Cubes cut from on-site generated foam on band saw to desired
size
2. 10-gallon, 10-1/2-inch x 9-1/2-inches x 12-inches aquarium
3. 1/4-inch mesh stainless screen
4. Depth indicator micrometer mounted on a bracket
5. Mettler balance
6. 100-tnl and 500-ml graduated cylinders
7. Mule's-foot squeezer - 4-1/2-inch I.D. x 10-inch
long thin wall tubing. Steel stock 4-inch diameter
x 4-inches long cut at 20° angle used as plunger.
Plunger weight = 14 Ib (170 lb/ft2 pressure applied.)
8. Tretolite C-10 demulsifier.
Procedure A
1. Fill tank with water
2. Put drain screen in tank
3. Measure level with micrometer
4. Using graduated cylinder add measured amount of oil
5. Measure level with micrometer
6. Weigh the foam in a container on Mettler balance
7. Place foam on oil slick and leave for the required
amount of time
8. Pull foam cubes with screen and let drain for 30 seconds
9. Place cubes in weighing container and re-weight
10. With micrometer measure new water-oil level
11. Calculate the amount of oil and water soaked up by the
foam from the change in oil thickness
(
x 100
(i - SPG)
Procedure B
Same as above for steps 1-9.
10. With mule's foot squeezer, remove oil and water from foam
11. Measure volumes of oil and water directly in graduated
cylinders, using demulsifier if necessary to aid oil/water
separation
Environmental Conditions
Indoors, 72°F
No agitation
50
-------
01
1)
Ji
o
JJ
0)
41
U
200
150
100
75
50
40
30
20
15
10
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
10
200
150
100
75
50
40
30
20
15
10
9.0
•8.0
7.0
6.C
5.0
4.0
3.C
20
3C
40 50
60
70 80 90
100 11C
12C 13C 14C
Temperature ( F)
FIGURE 24 - VISCOSITY OF TEST OILS
51
-------
0.94
0.92
0.90
o
u
o
$C.86
0.84
0.82
6C
70
$hailc
80
90
130
110
120
Temperature ( F)
FIGURE 25 - SPECIFIC GRAVITY OF TEST OILS
52
-------
a) Squeezing No. 2 oil from foam
b) After use with No. 6 oil
Figure 26 -• "MULE'S FOOT" SQUEEZER
53
-------
1.6
1.4
1.2
ID
™ 1.0
ra
O
0.8
c
>
0.6
INITIAL
SLICK DEPTH (IN.)
0.2 =°
0.1 =A
0.05 = 0
0.02 = D
0.4
0.2
.o_
I
I
20C
40C 600 800
Soaking Time (sees)
100
12CC
FIGURE 27 - OIL VOLUME RECOVERED, CARNEA 15
Experimental procedure is decribed in
Table 11, Procedure B.
54
-------
100 r-
80
60
£
A
iH
v4
O
40
20
200
400 600
Soaking Time (sees)
800
1000
1200
INITIAL SLICK DEPTH (IN.)
° =0.2
& =0.1
• =0.05
0 »0.02
FIGURE 28 - PERCENT OIL IN EFFLUENT RECOVERED, CARNEA 15
Experimental procedure is described in Table 11,
Procedure B.
55
-------
O = oil soaked
O = dry
A = water soaked
0.5
0.4
0.2
0.1
Dry
O
J.
Oil Soaked
Water Soaked
200
400 600
Time (sees)
800
1000
1200
FIGURE 29- OIL VOLUME RECOVERED FROM SORPTION OF CARNEA 15
Experimental procedure as described in Table 11,
.Procedure B.
Initial slick thickness = 0.05 in.
56
-------
oil
60
50
40
30
20
10
oil soaked
water
= dry soaked
I
A
I
200
400 60C
Time (sees)
80C
100C
1200
FI01JRE 30- PERCENT OIL IN EFFLUENT RECOVERED FROM SORPTION OF CARNF.A 15
Experimental procedure is described in Table 11,Procedure B,
Initial slick thickness = 0.05 in.
57
-------
o Shallow Yates
A Carnea 21
e oarnea 15
ONo. 2 Diesel
Vis. at
77°F. cs
12
36
16
3.8
0.6
eg
60
0.4
a
E
-------
100
80
Shallow Yates
Carnea 21
Carnea 15
#2 Diesel
O
A
a
©
Vis. at
77°. cs
120
36
16
3.8
60
40
20
200
400 600 800
Soaking Time (sees)
1000
FIGURE 32 - PERCENT OIL IN EFFLUENT RECOVERED FROM
2-INCH SORBENT CUBES
Experimental procedure is described in Table 11,
Procedure B.
Initial slick thickness = 0.05 in.
59
-------
Bench-scale sorption experiments were also performed using No. 6 Fuel
Oil of a viscosity of about 5000 cs at 72°F. When 2-inch foam cubes
were placed on a slick consisting of 0.02 gal. of No. 6 Fuel Oil for
approximately two minutes they did not sink through the oil. No signi-
ficant quantity of the oil adhered to the surface of the blocks upon
their removal from the slick (Figure 33(a)). When strips of sorbed material
were forced through the slick, into the water below, and held for two
minutes, the oil coated the foam, preventing water from entering its
interior. (Figure 33(b)). Small quantities of oil were forced into the
sorbent by hydrostatic pressure, but the majority of the interior remained
dry. A sample consisting of five 2-inch cubes was placed in the tank
and stirred into the slick. Where the sorbent was first exposed to water,
the oil did not adhere as readily as to the dry surfaces. Due to the
large quantity of oil which adhered to the exterior surfaces of the
cubes, large absolute quantities of oil were recovered, and a high
percentage oil content of the effluent was observed when the sorbent was
squeezed. These results confirm that, although specific oil recovery and
effluent purity initially decrease with increasing viscosity, this trend
reverses for oils of viscosities above about 200 cs (Shallow Yates) as
indicated in Figures 31 and 32. This trend may be explained by considering
Figure 34. While the quantity of oil absorbed into the sorbent matrix
over short sorption times generally decreased with increasing viscosity
due to a decrease in imbibition rates, the quantity of oil which adheres
to the exterior of the sorbent particles increases with increasing
viscosity. Thus, the total oil recovered, which is the sum of the two,
is represented by the bucket shaped curve of the illustration.
The above experiments demonstrated the effects of slick depth, sorption
time, and prior foam usage upon the amounts of oil and water sorbed by
2-inch polyurethane foam cubes. The oil content of the recovered fluids,
specific oil sorption by the sorbent, and fraction of the oil contained
in the slick that is recovered will vary additionally with the areal
concentration of the sorbent application. Experiments were undertaken
to determine the nature of this dependency using the polyurethane foam
mulch proposed for use as the oil sorbent in the full-scale recovery
system under investigation, to determine the quantity of sorbent which
should be applied to an oil slick to achieve optimum or near optimum
performance of the system. The size distribution of the polyurethane
foam mulch tested is indicated in Table 12. The apparatus is illustrated
in Figure 35 and procedures are described in Table 13. A 0.06 in. slick
of No. 2 diesel fuel was floated on a 5-foot diameter tank of water. A
measured quantity of foam mulch was dumped on the water and spread evenly
over the tank. After a soaking period of two minutes, the sorbent was
lifted from the tank in a net, allowed to drain for 30 seconds, and
placed in the mechanical wringing device, described in Section XI,
page 119ff. The sorbed fluids were wrung from the foam by passing three
times between six-inch diameter steel rollers under a linear wringing
pressure of 7.5.1b/in. The total effluent volume and oil volume recovered
in the wringing process were then measured directly. The entire pro-
cedure was repeated for a second cycle using the same sorbent material.
60
-------
a) Floated on slick
b) Thrust through slick
Figure 33 • RESULTS OF QUALITATIVE EXPERIMENT WITH NO. 6 FUEL OIL
61
-------
Total Oil Sorption
FIGURE 34 -
Increasing Oil Viscosity
NATURE OF RECOVERY DEPENDENCE UPON VISCOSITY FOR
POROUS OIL SORBENT FOR EXPOSURE PERIODS INSUF-
FICIENTLY LONG TO PERMIT COMPLETE SATURACTION
OF THE FOAM BY MORE VISCOUS OILS
TABLE 12
PARTICLE SIZE DISTRIBUTION OF MULCHED FOAM
Conditions;
Foam mulched by passing once through modified Reinco Model TM 7-30
hay spreader.
Foam mulch particle size distribution as determined by sifting
through square mesh screens.
Screen Square Size
4" x 4"
3" x 3"
2" x 2"
1" x 1"
1/2" x 1/2"
Cumulative
Foam Retained (by Weight)
0.7#
16.4$
60.5^
94.7#
98.7%
62
-------
a) Foam Mulch
Application
b) Sorbent
Draining
c) Wringing of
Foam
Figure 35 APPARATUS AND PROCEDURES
63
-------
TABLE 13
PROCEDURES USED FOR SORPTION TESTS IN LARGE TANK
Equipment
1. A 5-ft diameter 3-ft deep circular thin walled tank
2. A 6-ft diameter, 3-ft deep circular thin walled tank
3. 1/4-inch mesh minnow net
4. Mulched foam
5. Roller wringer
6. One 1000 ml graduated cylinder
7. Tretolite C-10 demulsifier
8. A triple beam balance
9. A hanging scale, 20 Ib capacity
Procedure
1. Place 5-ft diameter tank within 6-ft diameter tank
2. Fill inside tank with water
3. Place net in the water with edges out
4. Add a measured amount of oil
5. Weigh the mulched foam
6. Place foam on slick and start the stop watch simultaneously
7. "Pull the foam out with the net at proper time (usually after
two minutes soaking time) and let drain for desired period
(usually 30 seconds)
8. Hang on scale and weigh
9. Use Oil Herder® to surround the slick left in the tank
10. Skim all the oil off the surface
11. Separate the oil and water recovered
12. Use C-10 demulsifier to separate the emulsion (if any)
13. Measure the total amounts of oil and water recovered
14. Calculate amount of oil absorbed by the foam, allowing
for residual in tank
15. Roller wringer is used to recover the oil from the foam
Environmental Conditions
Wringer tests - outdoors, 70°F
Weight tests - indoors, 72°F
No agitation
64
-------
The results of these experiments for the second cycle are illustrated
in Figures 36 and 37. To aid in the interpretation of these and later
results, the foam, applied on the water surface in the 5-foot diameter
tank in various area concentrations, is shown in Figure 38. As shown in
Figure 36 the total effluent volume and oil volume sorbed per unit mass
of foam decrease monotonically with increasing concentration of sorbent
application. The fractional oil content of the effluent by volume
decreased with increasing concentration of sorbent application as shown
in Figure 37. Also included in Figure 37 is a plot of the quantity of
oil recovered per unit surface area of slick as a function of the area
concentration of the sorbent. This curve was obtained by multiplying
points of the smoothed specific oil sorption curve in Figure 36 by the
corresponding concentrations of foam mulch. It should be noted that
the curve represents only the oil recovered from the foam by the squeezing
process and does not account for that which remains within the foam
matrix after wringing. The upper barrier in Figure 37 represents the
maximum oil available per unit area of the 0.06 in. slick. The maximum
portion of the oil is recovered at about 0.063 lb/ft2 sorbent concentra-
tion, and corresponds tc about 65% of that which is available within the
slick. In considering this result, it should also be noted that the
specific fluid volume and oil volume recovered, and the percent by volume
oil content of the effluent, all show increasing trends as functions of
the number of cycles through which the oil sorbent is processed, and
thus the shape of the curves of Figure 37 will be altered for subsequent
cycles.
To eliminate the effect of foam handling and influence of inefficiencies
of the wringing process, further experiments were conducted in which the
quantity of oil removed from the slick was measured directly by skimming
the tank after the foam was removed and subtracting the quantity of oil
remaining from that originally contained in the slick. In each test a
given sample of foam was used only once, though both dry sorbent and
sorbent which has been presoaked in oil and wrung "dry" were used in
different tests. The fluid which drained from the sorbent when it was
lifted from the tank was examined and found to contain little or no
oil. The results of these tests are shown in Figures 39 through 41.
Figure 39 shows the quantity of oil removed from the slick as a per-
centage of that available.
Performance was better for the dry foam than for that which was oil
presoaked. Specific oil sorption by the foam is shown in Figure 40 and
is seen to decrease with increasing areal concentration of foam due to
the rapid decrease in the specific rate of sorption of oil by the foam
as the increasing number of particles rob one another of available oil
and cause a rapid decrease in the thickness of the slick over the
available soaking time. Again, performance is seen to be poorer for the
presoaked foam. Finally, the concentration by volume of oil in the fluid
contained within the foam matrix is shown in Figure 41, and is seen to
deteriorate steadily for increasing areal concentration of application.
Again, dry foam exhibits superior performance.
65
-------
0.8T
0.7
3, 0.6-
_2 0.5
cj
ji
U 0.4
0.3
0.2
0. 1
Type of Oil: £2 Diesel
Slick Depth: 0.06 in.
Soaking Time: 2 Min
Draining Time: 30 sec
No. of Cycle: 2
0.25 C.1C 0.15 0.2C 0.25
Area Concentration of Applied Sorbent (lb/ft2)
0.30
0.35
FIGURE 36 -
SPECIFIC SORPTION OF FLUIDS AS A FUNCTION OF THE
AREA CONCENTRATION OF FOAM SORBENT
66
-------
0.0036
0.0032
0.0028
a
oo
0.0024
0.0020
« 0.0016
u
o
I 0.0012
a
O
>
0.008
0.004
///////// Oil Available/Unit Surface Area /////,
////////// / y / / / /// / / // / / / / / ' / /- .1 I / J-
Type Oil: #2 Diesel
Slick Depth: 0.06 in.
Soaking Time: 2 min
Draining Time: 30 aec
No. of Cycle: 2
Region of Detectable Foam
Particle Interaction
I 1
J_
70
60
50
20
C
40 -
s
30
10
0.05 0.10 0.15 0.20 0.25
Area Concentration of Applied Sorbent (Ib/ft1)
0.30
0.35
FIGURE 37 - OIL RECOVERED PER UNIT AREA OF SLICK AND EFFLUENT
PURITY AS FUNCTIONS OF AREA CONCENTRATION OF FOAM
SORBENT APPLICATION
67
-------
a) Area concentration » 0.06
b) Area concentration = 0.12
Figure 38 - MULCHED POLYURETHANE FOAM ON SURFACE OF 5-FOOT DIAMETER TANK
68
-------
c) Area concentration =0.2 lb/ft'
d) Area concentration = 0.28 Ib/ft'
• Figure 38 (Continued)
69
-------
100
-J
o
9C
V
v 80
O
41
70
60
1
Type Oil : f2 Diesel
Slick Depth : Q.06 in.
Soaking Time : 2 min
Draining Time: 3C sec
O Dry
9 Oil Presoaked
0.02
0.04
0.06 0.08 0.10 0.12
Area Concentration of Sorbent (lb/ft2)
0.14
0.16
0.18
FIGURE 39 -
OIL REMOVED FROM THE SLICK AS FUNCTION OF AREA
CONCENTRATION OF FOAM SORBENT
-------
0.7
0.6
CO
00
0
•l-l
iJ
a
u
o
o
o
u
V
a.
en
0.5
0.4
0.3
0.2
0.1
\
\
\
\
Type Oil : #2 Diesel
Slick Depth : 0.06 in.
So.iking Time : ? tnin
Braining Time: 30 8ec
O Dry
* Oil Presoak
I
I
J
0.04 0.08 0.12 0.16 0.2
Area Concentration of Sorbent (lb/ftz)
FIGURE 40 - SPECIFIC OIL SORPTION AS FUNCTION OF AREA CONCEN-
TRATION OF FOAM SORBENT
71
-------
60
50
40
o
30
20
#2 Diesel
0.06 in.
2 min
Type Oil
Slick Depth
Soaking Time
Draining Time: 30 sec
O Dry
• Oil Presoaked
10
0.02
0.04
0.06 0.08 0.1 0.12
Area Concentration of Sorbent (lb/ft2)
0.14
0.16
FIGURE 41
- OIL CONTENT OF NET INFLUENT AS FUNCTION OF AREA
CONCENTRATION ON FOAM SORBENT
0.18
-------
Once probable soaking and draining times had been established for the
proposed full-scale sorbent recovery system, further experiments of the
nature of those above were conducted with a soaking time of one minute
and draining time of 10 seconds. For these experiments, No. 2 Diesel
Oil and Carnea 21 were selected as the test oils. Carnea 21 was selected
as the earlier bench-scale work had demonstrated specific sorption and
effluent oil contents for this oil to be near the minimum for the various
available oils tested.
Results of these tests are presented in Figures 42 through 49.
Experiments were again run with sorbent which had previously been soaked
with oil and wrung dry. Results of this work, for a 0.06 in. deep
slick, are shown in the figures as solid data points and broken lines.
From these figures it may be seen that performance again generally
deteriorates as slick depth decreases. In general, little improvement
in efficiency is seen to be achieved for sorbent application concentra-
tions over about 0.1 lb/ft2 in Figures 44 through 46. At these applica-
tion concentrations, Figures 41 and 42 indicate that a specific oil
sorption of two or better may generally be expected for 0.06 in. slicks.
Figures 41 and 42 demonstrate that the effluent oil content under these
conditions ranges from about 20$ to 45$. These figures provided the basis
for performance estimates of the on-barge, full-scale, sorbent recovery
system (see Section XIV).
Compared to the results of previous experiments conducted at this
laboratory and by others (e.g., see References 2 and 4) the specific oil
sorption realized above may appear unduly conservative. However, the
test conditions used here differed considerably from those used by others.
The primary difference is the oil layer thickness adjacent to each piece
of sorbent. For the data reported in References 2 and 4, the oil layer
thickness was either very great (when measuring maximum sorption capacity)
or was maintained constant during the test period, which corresponds
roughly to very low areal coverage of the slick by foam. For the data
reported in the present section of this report, the oil layer thickness
adjacent to the foam pieces decreased during the course of the experiment,
which means that total oil sorption was limited (among other things) by
the rate of migration of the thin layer of oil on the water surface.
This corresponds roughly (or actually) to high areal coverage of the
slick by foam. The results presented in this portion of this report are
felt to be realistic, lending themselves to the design of a practical oil
recovery system of reasonable size.
73
-------
0.7
0.6
0.5
g 0*4
4J
CL
a
CD
0*2
0.1
0
0
Slick Depth (in.)
O0,06
DO.04
A0.02
• 0.06, oil presoaked
Type Oil : 02 Diesel
Soaking Time : 1 min
Draining Time: 10 sec
I
0.04 0.08 0.12 0.16 0.2
Area Concentration of Sorbent (lb/ft2)
FIGURE 42 - EFFECT OF SORBENT APPLICATION CONCENTRATION ON
SPECIFIC SORPTION OF OIL FOR NO. 2 DIESEL OIL
74
-------
0.8
0.7
0.6
§ 0.5
•H
4J
O.
5 0.4
o
V
0.3
0.2
0.1
Slick Depth (in.)
O 0.06
D 0.04
A 0.02
4 0.06, oil presoaked
Type Oil : Carnea 21
Soaking Time : 1 min
Draining Time: 10 sec
0.04
I
0.08 0.12 0.16 0.2
Area Concentration of Sorbent (lb/ft2)
0.24
FIGURE 43 -
EFFECT OF SORBENT APPLICATION CONCENTRATION ON
SPECIFIC SORPTION OF OIL FOR CARNEA 21
75
-------
0.018
0.016
0.014
C 0.012
iH
<0
ao
*~'
•a
m
.a
I 0.010
o 0.08
H
0.06
0.04
0.02
Slick Depth (in.)
0.06
Dry Sorbent
Oil Presoaked Sorbent
0.04
0.02
Type Oil f2 Diesel
Soaking Time 1 min
Draining Time: 10 sec
J_
0.04
C.08 0.12 0.16 0,2
Area Concentration of Sorbent (lb/ft2)
0.24
FIGURE 44 -
EFFECT OF SORBENT CONCENTRATION ON RECOVERY
EFFECTIVENESS FOR NO. 2 DIESEL OIL
76
-------
0.018
0.016
^ 0.012
9)
.0
o 0.010
o
,-J
a
o 0.08
0.06
0.04
0.02
Dry Sorbent
Oil Presoaked Sorbent
Slick Depth
-------
00
95
90
u
w 85
£ 80
o
v
,0
a
CO
75
70
Slick Depth (in.)
O 0.06
D 0.04
A 0.02
• 0.06, oil presoaked
Type Oil : |2 Diesel
Soaking Time : 1 min
Draining Time: 10 sec
I
0.02
0.04
I
I
I
0.06 0.08 0.1 0.12
Area Concentration of Sorbent (lb/ft2)
0.14
0.16
0.18
FIGURE 46 -
EFFECT OF SORBENT APPLICATION CONCENTRATION
ON RECOVERY EFFECTIVENESS FOR NO. 2 DIESEL OIL
-------
IOC
\0
95
Jt
85
80
75
70
Slick Depth (in.)
O 0.06
A 0.02
• 0.06, oil presoaked
Typ« Oil : Carnea 21
Soaking Tin* : 1 min
Draining Time: 10 sec
0.02 0.04 0.06 0.08 O.I 0.12
Area Concentration of Sorbcnt (lb/ft*)
0.1*
0.16
0.18
FIGURE 47 -
EFFECT OF SORBENT APPLICATION CONCENTRATION ON
RECOVERY EFFECTIVENESS FOR CARNEA 21
-------
oo
o
60
50
§ 4C
o 30
•tie.
20
10
0,02
Slick Depth (in.)
O 0.06
a 0,04
• 0.06, oil presoaked
Type Oil : #2 Diesel
Soaking Time : 1 min
Draining Time: 10 sec
1
I
I
0.04
0.06 0.08 0.1 0.12
Area Concentration of Sorbent (lb/ft2)
o.u
0.16
0.18
FIGURE 48 - EFFECT OF SORBENT APPLICATION CONCENTRATION ON
OIL CONTENT OF AFFLUENT FOR NO. 2 DIESEL OIL
-------
70
oo
>°
>.
_D
60
50
40
30
20
Slick Depth (in.)
O 0.06
a o.o4
A Q.02
• 0.06, oil presoaked
lype Oil : Carnea 21
Soaking lime : 1 min
Draining Time; 1C sec
10
0.02
0.04 0.06 0.08 0.1 0.12
Area Concentration of Sorbent (lb/ftz)
0.14
0.16
FIGURE 49 - EFFECT OF SORBENT APPLICATION CONCENTRATION ON
OIL CONTENT OF AFFLUENT FOR CARNEA 21
0.18
-------
SECTION VIII
DISTRIBUTION OF SORBENT
Jntroduction
The distribution and transport of the polyurethane foam begins with
foe cured foam bun and ends as the used foam is either redistributed or
is stored for disposal. Foam buns are prepared by slicing or tearing
prior to being placed in contact with the oil. For the transport of
used foam from point to point, either belt conveyors or pneumatic
pressure conveyors can be used.
Foam Preparation Requirements
The polyurethane foam generated on-site is produced as a bun (see
Section VI). As presently made, the buns have an impermeable skin
of closed cells formed on the upper surface during curing. The lower
surface which forms in contact with a substrate of paper or other
material, is largely open-celled. To increase efficiency as a sorbent,
the bun is torn or sliced in a way that will expose the open cells in
the interior of the bun. Ideally, these pieces should be as small as
can effectively be recovered from the water. A particle size distri-
bution which may be largely retained on a one-inch or larger square
mesh appears practical. Two methods of foam preparation have been
considered in this study.
.Slicing or Breaking
A sample of our foam bun was delivered to one machinery manufacturer
(the Fitzgerald Company) for a factory evaluation using a standard
breaking machine such as used in the food and drug industry. It was
found that the particular machine produced a particle judged to be
well suited to oil 3pill work. The pieces were relatively uniform in
thickness (3/8-in. to 1/2-in.) and of a size easily recovered (see
Table 14 and Figure 50). A sieve analysis of this material is reported
TABLE 14
PARTICLE SIZE DISTRIBUTION
OF SLICED FOAM
Cumulative Percent
Screen Square Size Foam Retained (by weight)
4-in. x 4-in. 22
3-in. x 3-in. 44
2-in. x 2-in. 55
1-in. x 1-in. 82
< 1 100
83
-------
Figure 50 - FOAM PREPARED WITH FITZGERALD
BREAKER
in Table 14. This machine is not considered practical for offshore
use in its present form, requiring major modification or redesign
for adaptation to this service and to increase its capacity. Although
it might be considered for future development, no further work was
attempted in this study.
Mulching
Foam generated on-site is characteristically of relatively low tensile
strength, making it possible to tear or break the buns into suitable
pieces with a common power muleher such as is manufactured by Reinco of
Plainfield, New Jersey (Figure 51). In the standard configuration, control
of the particle size is not completely satisfactory, and it was found
that the addition of studs (or shredder bars) to the beater chamber improved
performance, as illustrated in Figure 52. This simple field modification
could be made on any of the several hundred similar machines now in use. A
typical particle size distribution, as obtained with our modified Reinco
Model TM 7-30 mulcher, is shown in Table 12. For this study, a total of
84
-------
Figure 51 - REINCO TM 7-30 POWER MULCHER
Figure 52 - MODIFICATION TO MULCHER
85
-------
eleven 1/2-in. diameter studs, projecting 2-1/2-in. into the beater
chamber, were installed.
Distribution of New Foam
The initial distribution of fresh foam is easily accomplished with the
power mulcher or, if the foam is already mulched, by a simple centri-
fugal blower.
The small Reinco TM 7-30 mulcher is manually fed and is capable of
discharging foam at velocities approaching 150 mph (5000 CFM air stream).
The unit is rated at four tons of baled straw per hour, but with foam in
buns its rate, is less due to the lower density of the bun. A mechanical
feed or an enlarged chute would be a desirable modification for any
machine purchased for this service.
In tests with the standard mulcher, manned by three men, small buns
weighing five to six pounds were mulched at a rate of 2100 Ib/hr. This
rate can be increased by about 50$ if buns are pre-sized to the width of
the charging chute and of considerable length (say, ten feet). A
further increase in rate (and increased safety) would be obtained by
adding a long charging chute with high sides. However, 3000 Ib/hr is
a reasonable and conservative average rate* When fed to the mulcher as
buns approximately 10-ft x 2-ft x 4-inches, the foam was distributed into
a five to seven-knot wind to a distance of about 35 feet. When distri-
buted downwind the maximum distance about doubled. In a typical experi-
ment, the foam was mulched and blown to cover an area about 30 feet in
diameter, the center of which was 45 feet from the spreader. Time required
to cover the 30-foot diameter area was 15 seconds.
Another larger mulcher from the same manufacturer (Model M60-F6) has a
larger charging chute and is rated at nine tons of straw per hour. Such
a unit should be capable of handling over 6000 Ibs of foam per hour.
Recycling of Foam
In the present concept, essentially all of the foam will be recycled
after wringing. During recycling, the foam will contain enough residual
liquid to raise the actual density to about 10 lb/ft3. We have found
that this foam, loosely piled, without mechanical compaction, occupies
about 0.56 to 0.67 ft3/lb dry weight equivalent. An average ''apparent
density" of 1.64 lb/ft3 (dry weight equivalent) has been used for design
estimates. Depending upon the size of the recovery system the handling
of this foam for recycling may be by container, by belt conveyor, or by
pneumatic systems.
Containers
If the foam is to be handled by batch methods, containers may be used.
Fabric mesh bags, metal fabric bins, etc., might be utilized. Volume
86
-------
as well as package weight will be a significant limitation. A metal-
fabric bin 6-ft x 8-ft x 12-ft might hold about 5500 Ibs (or 1000 Ibs
dry weight equivalent), representing an oil-sorption capacity of from
350 to 700 gallons. Thus, it may be seen that this approach is best
suited to small capacity systems or cases where the material must be
transported over a considerable distance prior to redistribution.
Such containers might be placed on the stern of a support boat, filled
by blowing through a conduit then distributed by opening a tailgate.
Belt Conveyors
As the oil recovery process is continuous, components of the system
will preferably provide for continuous flow. Belt conveyors of
various types are well suited to this type of material handling situ-
ation, where the movement is,confined to a single vessel. The conveyor
can be extended a limited distance outboard for distributing the foam on
the water.
Pneumatic Conveyors
The foam as recycled can readily be conveyed in a stream of
air. Density, particle size, and the continuous nature of the
process are all compatible with this system. In discussions with
suppliers, it was estimated that for moving quantities of this
material distances of at least 150 feet, a minimum air stream velo-
city of 3000 ft/min would be required. Conveying could be by use of
a simple centrifugal blower or by a positive displacement blower with
feed by air lock into the pressure side.
The latter type of system was investigated in discussions with one
manufacturer, Fluidizer, Inc. A preliminary recommendation called
for the following equipment to move 156,000 pounds of material/hr
at least 150 feet, discharging into the atmosphere: 1-75 hp,
electric motor dirven positive displacement blower (rated at 1200 CFM,
10 psi); four foam inlet hoppers; four rotary air lock injection
valves; and an eight-inch transfer line. The system could be .
palletized for storage or shipment. The total system weight might
be approximately nine tons, the maximum single pallet weight 5500
pounds, and the maximum dimension to be the height selected for the
hoppers (which would be nested for storage). A system of this
type would be well suited to use on a large vessel, provided adequate
electrical power is available. A lighter, simpler system, perhaps
better suited to use on smaller vessels would be the simpler blower.
Preliminary estimates from one supplier suggested a 40 hp blower with
lightweight pipe conduit.
In an attempt to evaluate the practicality of the blower concept, a
series of simple tests were carried out using the mulcher as a blower
(Figure 53>. The results of these tests are shown in Table 15.
87
-------
Run*
TABLE 15
FOAM TRANSPORT TESTS
Foam
A-l 64 Ib-wet
A-2 53 Ib-wet
A-3 27-1/2 Ib-wet
B-l 24 Ib-dry
B-2 45-1/2 Ib-wet
B-3 39-1/2 Ib-wet
B-4 19-1/2 Ib-dry
Time
50 sec
35 sec
30 sec
65 sec
75 sec
47 sec
33 sec
Rate
4600
5400
3300
1300
2200
3000
2100
Ib/hr
Ib/hr
Ib/hr
Ib/hr
Ib/hr
Ib/hr
Ib/hr
A series: 50 feet of 16-inch conduit
B series: 98 feet of 16-inch conduit
Expanded Metal
16-inch O.D.
7
Reinco TM7-30
Muleher
Figure 53 - FOAM TRANSPORT TEST USING POWER MULCHER
88
-------
Pre-mulched foam was collected in large bags. For the wet foam tests
the foam was water saturated, wrung with three passes through the
model wringer (see Section XI) and then rebagged. For these tests
the foam was remulched in passing through the mulcher, as the
beater chains and studs were not removed. A discharge conduit of
16-inch pipe was selected on the basis of the rated blower capacity
of 5000 CFM, though a flow slightly below the desired 3000 ft/min
was achieved against the back pressure. By repeated observation
of single particles, an average particle velocity (in Test B-3)
of about 2500 ft/min was recorded.
In no case did it appear that the blower was overloaded, although
during surges in the manual feeding foam remained in the beater
chamber longer than usual. The manual feed operation, which
involved dumping premulched foam from bags into the standard chute,
limited throughput. A suitable hopper or an enlarged chute might
double the throughput.
After the second pass through the mulcher and into and out of the
system, some size degradation was observed, possibly due to minor
losses in the actual handling of the foam. This degradation was
more pronounced in the "B" series, (see Table 15) and is shown in
Table 16. (See also Section XII of this report).
TABLE 16
PARTICLE SIZE DISTRIBUTION
RETAINED ON SCREEN - %
Single Mulch Second Mulch and Transport
Screen Size Dry Dry Wet
4-in. x 4-in. 0.7 0 0
3-in. x 4-in. 15.7 < 0.2 < 0.2
2-in. x 2-in. 44.1 32.5 29.0
1-in. x 1-in. 34.2 65.1 69.0
< 1-in. 5.3 2.2+ 2.0+
Pneumatic transfer of foam during recyclinj appears to be the
preferred method. However, further study is recommended, parti-
cularly with respect to the possible hazard from static charge
accumulation, though this does not appear to be a significant
hazard in a fully grounded system discharging into the open
water (as opposed to discharging into a confined space such as
a tank).
89
-------
SECTION IX
COLLECTION
The proposed oil recovery system utilizes a one or two sided "V" shaped
converging boom system to concentrate oil soaked sorbent at the harvesting
location. Failure of these booms to contain the sorbent in the absence
of waves will generally result from one of the two phenomena shown in
Figure 54. In the case of Figure 54(a), the foam is simply swept beneath
the boom by the rapid current and consequent steep pressure gradient
along the vertical face. In the case of Figure 54(b), the gap between the
converging booms becomes blocked in low currents by bridging of the foam
particles, resulting in a pile up of foam against the boom, along which
additional foam may roll when driven by the current.
To aid in developing an understanding of the behavior and flow of a sheet
of mulched foam in proximity to oblique and converging booms, experiments
were made to determine the linear shear strength of a sheet of the mulched
polyurethane foam floating on water. The apparatus used for these experi-
ments is illustrated in Figures 55 and 56. The cross-hatched test area in
Figure 56 measured 3-ft 10-inches in width and 6-ft in length. The weight
consisted of a container of sand. The test section was covered uniformly
with foam mulch of the size distribution in Table 12, and sand slowly
added to the weight until failure of the sheet was indicated by continuing
movement of the floating baffles suspended from the longitudinal member
connecting the two styrofoam floats (see Figure 56 and Table 17). The
TABLE 17
PROCEDURES FOR MEASURING SHEAR
STRENGTH OF FOAM MULCH SHEET
Equipment
1. 15-ft x 4-ft x 2-ft tank with fabricated partitions
2. Mulched foam
3. Triple beam balance .
4. 18-inch x 12-inch x 6-inch wax-coated styrofoam blocks
5. Weights
Procedure
1. Fill tank with water.
2. Place sliding partition in position and secure.
3. Weigh out desired amount of foam using triple beam balance.
4. Place foam in tank between partitions.
5. Pull on sliding partition by adding sand (weight) to the
pulling mechanism until movement occurs.
6. Weigh the amount of sand that was required to cause movement.
Environmental Conditions
Indoors, 72°F.
91
-------
Boom
Water surface
-&
•&-»
o
0
0
J/
a) Foam sweeping under boom
Current
Boom
Current
b) Fo^m rolling under boom
FIGURE 54 -
TWO POSSIBLE MODES OF FAILURE FOR FOAM SORBENT
BOOM IN ABSENCE OF WAVES
92
-------
Figure 55 - SHEAR TEST APPARATUS IN USE
-------
Styrofoam
Float (2)
Fixed
Baffles
Foam Sample
Shearing
Force
Planes of Shear
Floating
Baffles
Weight
FIGURE 56 - SHEAR TEST APPARATUS
94
-------
results of these tests are shown in Figure 57, in which linear shear
strength in Ib/ft is plotted as a function of the area density of the
foam sheet in Ib/ft2.
These data indicate that no stress can be supported by a foam sheet of an
area concentration less than about 0.1 Ib/ft2. Above this value, stress
may be supported by the sheet, and is seen to increase steadily with in-
creasing foam concentration. This increase is due to both the increasing
depth of the foam, resulting in a dilution of the stress over the vertical
planes of shear, and increasing compaction, resulting from the weight of
the foam above. This threshold concentration is above the shoulder of
the oil recovery curves of Figures 46 and 47, and corresponds to the
assumed application density of the sorbent for the system performance
estimates (page 167).
The relationship of the maximum shear stress supportable by a foam mulch
sheet to the blocking of converging booms during sorbent collection opera-
tions is illustrated in Figure 58. When blocking of the converging section
occurs, the maximum shear stress lies in the vertical planes of the dashed
lines at the gap edges. The probability of the occurrence of blocking is
reduced for larger gaps as the fluid drag of a larger area of mulched
foam must be supported along the lines of shear, resulting in higher shear
stress for a given towing velocity and foam sheet length. The tendency
of foam to bridge across the gap may also be reduced by decreasing the
included angle between the booms, thus decreasing the normal pressure, and
increasing the shear stress at the boom face, causing the foam to slip
toward the gap more easily. The above data indicate that blocking cannot
occur so long as foam concentrations in proximity of the gap are less
than 0.1 Ib/ft2.
Large scale tank tests were run to determine the nature of the flow of
foam mulch between the booms and bridging of the boom gap (see Tables 18
and 19). A pair of twelve-foot converging booms of 13-inch draft and
23-inch freeboard, positioned to form a 3-foot gap, were towed through
foam mulch confined in a channel between two longitudinal floating booms
in the wave tank. Tows were made at speeds from 0 to 1.7 knots, with
included boom angles of 60 and 90 . Visual observations and photographic
records were made of these tests. Tows were made both with and without
waves. Under these conditions, no significant general boom failure was
observed. However, in areas of concentrated vorticity (i.e., the trailing
edge of the boom at the gap) sorbent material was observed to be drawn
below the surface to depths as great as three feet. Ease of failure is
related to saturation of the foam, dry foam floating higher and washing
down with less ease than that which has become water saturated. With dry
foam, speeds to 3 ft/sec were obtained with an included angle of 90° with
no observed failure, in the absence of waves. Waves one foot in height
and 12 feet in^length produced no failure with wetted foam to a velocity
of 1.5 ft/sec. With waves of greater steepness, failure was observed to
occur due to splashover or the lifting of the boom's lower edge above the
wave troughs in the confused seas created within the 90° convergence area,
as shown in Figure 59. Failure due to the lifting of sections of the boom
95
-------
in
O)
l-i
l-i
re
0.7 r
0.6
0.5
0.4 -
0.3 -
0.2
0.1 -
0.1
0.2
0.3 0.4 0.5
Planar Concentration of Foam (lb/ft2)
0.6
FIGURE 5.7 -
VARIATION IN LINEAR SHEAR STRENGTH OF A SHEET OF
MULCHED FOAM ON WATER AS A FUNCTION OF AREA
CONCENTRATION
-------
Relative
Current
Planes of Maximum
Shear Stress
Direct ion
of Tow
Booms
Gap
FIGURE 58 - BRIDGING OF CONVERGING BOOMS BY FOAM MULCH SORBENT
97
-------
TABLE 18
PROCEDURE FOR CONFINED FOAM BOOM TCWING TEST
Equipment
1. Laboratory wave tank, 6 x 50 x 120 feet.
2. Fabricated wing structure and boats.
3. One boat for general use.
4. Winch arrangement with forward and reverse.
5. Polypropylene rope.
6. Boom arrangement. Booms converging at desired included angle.
Booms are 3-ft x 15-ft.
7. Half-inch mesh chicken wire.
8. Small crab nets to recover foam.
9. Mulched foam.
Procedure
1. Using the winch, position the wing close to the beach.
2. Put a known amount of foam within the confined area between the
booms.
3. Start the winch and set on required speed.
4. Immediately after the wing starts moving, the screen across the
boom gap is lifted.
5. After the run the wing is pulled back and repositioned close to
the beach.
Environmental Conditions
Outdoors, 72°F, light wind.
Ref. SPLC RSkD Laboratory Notebook No. LR324.
TABLE 19
PROCEDURE FOR LOOSE FOAM TEST
Squipment
1. Same as in confined foam towing test described in Table 18,
excluding chicken wire.
2. Two sections of slickbar boom.
Procedure
1. Same as confined foam towing tests, described in Table 18.
2. Foam is dumped ahead of the fabricated booms between the two
slickbars.
3. Wing is towed toward the foam.
Environmental Conditions
Outdoors, 72°F, light wind.
98
-------
Figure 59 - FAILURE OF BOOMS BY SPLASHOVER AT
1.5 FT/SEC IN PRESENCE OF WAVES
99
-------
above the water may be mitigated in full scale systems by using articulated
strings of boom sections with low longitudinal tension to allow the booms
to conform effectively with the contours of the water surface or by using
booms of greater draft.
Additional tests were run in which closely controlled concentrations of
foam mulch were confined within the included area of the converging booms
prior to the initiation of towing and any tendency of the foam to bridge
the three-foot boom gap upon getting underway was noted. This test arrange-
ment is shown in Figure 60. The results of 11 such runs are shown in
Figure 61. From these data it is expected that bridging will present no
substantial obstacle to the performance of a full-scale recovery system.
During these experiments failure by foam washing beneath the booms in the
absence of bridging was noted only at the highest tow speed, with the boom
draft reduced to 13 inches, and only with foam which had been exposed to
water over long periods of time so as to render it near neutrally
buoyant.
These tests indicate that massive failure should not occur with a properly
designed system of booms at the convergence angles and towing velocities
anticipated for the full-scale system.
100
-------
a) Stationary
b) Underway
Figure 60 - BRIDGING TEST ARRANGEMENT
101
-------
1.2
1.0
6.8
c
0)
o
o
i-l
O
en
0.6
0.4
0.2
o
O
-L
L
0.2
0.4 0.6 0.8 1.0
Towing Velocity (Knots>
1.2
1.4
FIGURE 61 -
OCCURRENCE OF BRIDGING OF CONVERGING 12-FOOT BOOMS
WITH GAP OF 3 FEET AND INCLUDED ANGLE OF 90°.
DARKENED SYMBOLS INDICATE OCCURRENCE OF BLOCKING
102
-------
SECTION X
HARVESTING OF SORBENT
Introduction
In a small system, sorbent can be recovered with nets or mesh drag
line buckets. In some cases, recovery by skimming is possible using
diaphragm or open impeller pumps. For a high volume continuous system,
mechanical belt harvesters appear to be well suited. To avoid the high
water recovery expected in skimming, and to take advantage of the fact
that the first material which drains from the foam (initial gravity
draining) is essentially water, an open mesh belt harvester was inves-
tigated in this study.
Open mesh belts are structurally sound, and open mesh belt conveyors
have been employed in similar situations in the past. Kelp harvesters
used off the California coast and aquatic weed harvesters more recently
developed for cleaning lakes use this concept. A tentative selection
of belt type was made based on strength, relative amount of open space
presented to water flow, and the characteristics of the mulched foam.
The minimum foam particle size that would normally be introduced into
the system would be retained by a one inch by one inch square mesh. Since
the wet foam particles tend to mat and to interlock when in contact with each
other, pieces which are somewhat smaller than a one-inch cube would
probably not be lost when wet.
The harvesting system studied utilized a flat wire belting (manufactured
by Cambridge Wire Cloth Company) with a 1/2-inch x 1-inch mesh. This
belt is 3/8-inch thick and is available in a variety of weights and
materials. A series of experiments was conducted to investigate the
behavior of mulched foam on an inclined conveyor, in air, and in the
presence of a water current.
Retention of Foam on an Inclined Stationary Conveyor
Mulched foam was screened and two size ranges were used for this
experiment: that retained on a 2-inch square mesh and that which passed
through the 2-inch but was retained on a 1-inch square mesh. For each
test, the foam was distributed on a horizontal section of 1/2-inch x
1-inch flat wire belt conveyor, either in a single layer or in multiple
layers (to a total depth of three inch to six inch). The foam was
applied dry, water-saturated, and wrung "dry" by hand.
The conveyor was then inclined while gently shaken. When a significant
amount of tumbling or movement of the sorbent was observed, the angle
of inclination was recorded.
103
-------
The results shown in Table 20 suggest that this limiting angle is
influenced less by particle size or condition (wet or dry) than by the
thickness of the foam layer. This may be explained by the irregular,
angular shape of the mulched foam which permits interlocking of particles,
making the multi-layer condition more stable than the single thickness in
contact with the mesh belt.
TABLE 20
RETENTION OF FOAM ON AN INCLINED
STATIC BELT CONVEYOR
Particle Single Layer Multi-Layer
Size Range Dry Saturated Wrung Dry Saturated
2-inch mesh 32-36° 37° 35° 40° 39°
1-inch mesh 31° 35° 33° 37° 38°
Foam Collection by Static Inclined Belt in Current
The static belt section was next installed in the current tank to study
the behavior of foam approaching in a current and the mode of failure
when foam is moved under the belt. The behavior of the foam was observed
with the belt at 30° and 40° inclination, and with current speeds from
0.7 ft/sec to 2,0 ft/sec, for the two size ranges of particles described
above. Foam reaching the conveyor mesh quickly formed a barrier to the
passage of water, oil, or foam. For foam evenly distributed on the water,
the initial stages resembled the behavior of an oil film at a boom.
Particles at or near the upstream edge of the foam layer were carried under
by the action of current and waves. In this early stage, these particles
tended to return quickly to contact the under surface of the floating foam
layer but would continue to roll until becoming lodged in an opening. A
thickening of the layer, not unlike the oil film headwave, was sometimes
observed until enough foam had rolled to the conveyor (Figure 62(a)).
After a short time, a solid wedge of foam developed at the conveyor
(Figure 62(b)) and particles tended to roll under the relatively smooth
face of the wedge and past the harvester in a manner similar to that
described on page 91. The upstream face of the wedge was approximately
45 degrees from horizontal, regardless of harvester angle. The wedge
sometimes became unstable if an unusually large foam particle was exposed
on the upstream face. If such a particle was carried away by the current,
much or all of the wedge might roll under as a unit before stability could
be re-established by filling this void. Substitution of a very fine con-
veyor mesh had little effect on the behavior of the foam in these tests.
Use of the smaller foam particles resulted in more rapid formation of a
stable wedge and ultimately a more stable, denser, wedge which presented
a smoother upstream face to approaching foam.
104
-------
(a) Initial Stage of Wedge Development
(b) Fully Developed Wedge with Failure by Particle
Rolling
FIGURE 62 - FOAM PARTICLES AGAINST STATIC BELT IN CURRENT
105
-------
The above observations indicate that failure of foam under a harvester
can occur if it stops or slows for even a short period of time. The
time until failure would depend, to some extent, on the depth of the
conveyor tip.
Operating Prototype Harvester
A small prototype harvester was designed and fabricated for testing in
the current and wave tanks. This unit consisted of a 30-inch width of
1/2-inch x 1-inch x 3/8-inch thick flat wire belt in a steel channel
frame. The conveyor was driven by an hydraulic motor, and the angle of
inclination was adjustable from 0° to 45°. The total weight of the
harvester and base was 575 lb, and the effective length (overall) was
approximately 10 feet. The prototype harvester is shown in Figure 5.
The harvester was suspended in the current tank so that the lower end
was opposite the observation window. A steel mesh basket, 30 inches
wide and 7 feet long was suspended over a channel formed by parallel
booms. The required amount of foam was placed inside the basket and it
was lowered into the stream. The foam was then released through a gate
in the downstream end. A screen installed behind the harvester caught
any foam lost under the booms or the harvester. Velocities were deter-
mined by timing the transit of a portion of the foam over a measured
distance in view of the observation window.
Water and air were entrained on the underside of the conveyor and
carried to the lower end as the belt speed was increased. When the belt
speed became high enough, an upwelling, or bubble barrier, was created
at the approach to the belt, often preventing foam from contacting the
harvester. This effect is shown in Figure 63. In general, when the
ratio of belt velocity V_ to current velocity Vg exceeded about 2.5, a
backwash (or counter current) was noticeable on the surface. At a ratio
VB/VS of 4.0, a counter current of two ft/sec was measured and was
observed to persist for ten feet upstream in the narrow channel.
A simple shroud was installed around the lower shaft to divert this
current parallel to the rising side of the belt. This shroud was not
used in any foam recovery tests. Such a device should be investigated
in the design of a harvester to facilitate operation at low current
velocities and high belt speeds.
A series of tests was run with the harvester at 30° and at 40° to
horizontal. In each series, the belt speed and the approach velocity
were varied. The average foam concentration varied from 0.5 to 0.7
lb/ft2. Recovery rate was determined by the time elapsed from first
contact of foam with the belt until the last of the mass was removed.
This rate, was then expressed as the weight of dry foam recovered per
hour. The results are summarized in Table 21. Little or no tendency
for the foam to roll or tumble was noted at the 30 degree inclination.
At 40 degrees, tumbling significantly affected the recovery rates. To
avoid loss of foam by tumbling, four-inch high flights of expanded metal
106
-------
Figure 63 - UPWELLING OF ENTRAINED AIR AND WATER
TABLE 21
SUMMARY OF CURRENT TANK
TEST RESULTS - RANKED
Harvester Date
Angle
30° 11/3
11/2
11/2
11/2
11/2
11/2
11/2
11/2
11/3
11/3
40° 11/4
11/4
11/4
11/4
11/4
11/4
11/4
Run
g
I
5
2
4
6
/
:-;
1
10
18
4
19
11
7
5
12
V
(ft/sec)
2.0
2.0
0.8
2.0
0.8
0.8
2.0
2.0
3.0
1.0
2.5
2.0
3.0
2.0
3.0
2.0
2.5
VS
(ft/sec)
0.8
2.0
0.5
2.0
2.0
0.8>.
0.5
0.8
0.5
0.5
2.0
1.3
2.0
2.0
1.8
1.2
2.0
V /V
B' s
2.7
1.0
1.5
1.0
0.4
1.0
4.0
2.7
6.0
2.0
1.3
1.5
1.5
1.0
1.7
1.7
1.7
Calculated
Recovery
Ib/hr
(dry wt)
1600
1480
1400
1370
1300
1300
1300
1300
1300
1190
1420
1410
1270
1200
1200
1120
1120
Notes
1'
2
2
2
2
2
2
Note 1: Effect of upwelling observed in belt feed.
Note 2: Rate reduced on all runs by tumbling of foam on belt.
107
-------
were added at four-foot intervals along the belt, as shown in Figure 64.
The results obtained with flights are shown in Table 22.
The harvester, with flights installed, was assembled on a catamaran
for towing tests in the wave tank, as shown in Figures 65 and 66. A
towing winch with variable speed drive was used to tow the assembly
into waves of two types. The first was a long wave of about 12 inches
height and 14-foot length (steepness: 14:1). The second was a shorter
wave of the same height, but a length of seven to eight feet (steepness:
8:1). A preweighed quantity (dry weight") of foam was placed in a container
either dry or in a water wet (but drained) condition. This foam was
spread immediately ahead of the harvester, and the times observed in a
manner similar to the' method of the current tank experiments. Concen-
tration of the foam was estimated as it encountered the short booms
at the harvester. Concentrations (area density) of from 0.27 lb/ft2
to as high as 1.2 lb/ft2 were observed. Runs were made without waves
to correlate with current tank results. All runs were made with the
harvester at 40 degrees.
The data suggest that the recovery rate for foam in a calm sea is
delated to the system velocity more than to any other factor. The
curve in Figure 67, shown as "Current Only" represents data
from the current tank tests. In most instances, waves improve
the recovery by ensuring satisfactory feed to the harvester.
With the necessarily small quantities of foam used in these tests,
the steep waves were less effective than the longer waves, largely due
to a cross chop which developed within the short 45 degree boom array
(cross chop would not be so severe in a longer array at lesser angles).
This situation is shown in Figure 68. Longer waves, more nearly
resembling in shape those expected in operation, had a significant
effect on recovery rate.
Recovery rate in waves is affected by foam concentration. This effect
is exaggerated in our experiments, in which the foam batch size was
limited (the feed rate declines at the tail end of a small batch since
no additional foam is present to provide a driving force) .
The manner in which wet foam is retained on the harvester with flights
is shown in Figure 69. The wedge results in part from the tumbling
of foam on the moving belt; a reduction in flight spacing would" increase
the capacity of the harvester. The figure shows a mass of foam of an
estimated equivalent dry weight of 4.8 lb which was picked up at a belt
velocity of 2.0 ft/sec and system velocity of 3.0 ft/sec. Assuming
equal recovery by each flight, the recovery rate might be:
x
3600 Flight = ,
Flight Sec Hr 4.0 ft
108
-------
•
v Si
• _
1/2" x 1" FLAT BELT HARVESTER WITH
EXPANDED METAL FLIGHTS
TABLE 22
Harvester
Angle
30*
SUMMARY OF CURRENT TANK
TEST RESULTS - RANKED
Flat Wire Belt Harvester - 4" Flights
Date Run VD
D
(ft/sec)
11/12
11/11
11/11
11/11
11/11
11/11
11/12
11/11
11/11
11/11
4
6
7
,',
8
-
8
3
13
12
l\
3
»
'I,
2,
1.
!
'I .
5.
V
a
(ft/sec)
1.5
1.0
0.9
0.5
0.8
V /V
V s
2,
3,
.
3,
2
2,
5.0
1.0
3.8
4.0
5.0
5.0
Calculated Notes
Recovery
(Ib Dry Wt./hour)
9400
6200
5200
4600
4700
4300
4100
3600
2000 1
1600 1
11/9
11/9
11/12
11/9
11/9
11/9
11/10
4
(.
3
5
3
2
1
4.0
4.0
3.0
4.0
3.0
2.0
3.0
1.3
1.3
1.5
1.3
1.3
1.3
0.7
3.2
3.2
2.0
3.2
2.4
1.6
0.4
5200
5200
5200
4700
4500
3900
1600
Note 1: Upwelling effect observed,
109
-------
(a)
(b)
Figure 65 - HARVESTER IN WAVE TANK
110
-------
(a)
(b)
Figure 66 - TOWING INTO 14-FOOT LONG WAVES AT 2.0 FT/SEC
111
-------
10
3
cr
w
to
O
O
o>
4->
CO
O
u
0)
PS
/\ 14-Foot Wave Length
(steepness 14:1)
8-Foot Wave Length
(steepness 8:1)
D Foam Concentration - lb/ft2
2 Indicates belt velocity - ft/sec
D = 0.8 ±
Data for Steep^
Wave
Data for long
wave
= 1.5
Data From
Current tank tests
1.0 2.0 3.0
System Velocity--Ft/Sec
FIGURE 67 - EFFECT OF WAVE FORM AND FOAM CONCENTRATION
-------
FIGURE 68 - TOWING INTO 8-FOOT LONG WAVES AT 2.5 FT/SEC
FIGURE 69 -
TYPICAL LOADING OF 4-INCH EXPANDED METAL FLIGHT
AT HARVESTER ANGLE OF 40°, BELT SPEED = 2 FT/SEC,
SYSTEM VELOCITY = 3 FT/SEC
113
-------
which agrees with the experimental results for steep waves or for no
waves as shown in Figure 67. The average volume recovery calculated for
this 2.5 ft belt width is 1380 ft3/hr (actual volume), giving an apparent
or packing density (dry weight equivalent) of:
i • A -4. 8700 lb Hr iv/^s
packing density = —— x 138Q ft.3 « 6.2 lb/ft3
The actual dry density of this foam batch was approximately 2.0 lb/ft3.
Foam Packing Density During Handling
To assist in the design of all conveying components, a series of design
curves was prepared based upon simple bench scale density experiments in
which a wire mesh basket, 24 inches square, fabricated from 1/2-inch square
mesh was filled with foam, without compaction, and the occupied volume
was estimated. The initial measurement was with mulched dry foam. The
same foam was water soaked for three minutes then returned to the basket.
Foam in this "wet" condition would represent saturated foam after some
30 seconds draining. Again, the same foam was soaked, wrung by the
experimental wringer, and returned to the basket. The results are shown
in Table 23.
TABLE .23
APPARENT OR "PACKING DENSITY" OF FOAM
WITHOUT COMPACTION
2.0 lb Sample 4.8 lb Sample Average
Apparent Apparent Apparent
Volume Density Volume Density Density
Foam (ft3) (lb/ft3) (ft3) (lb/ft3) (lb/ft3)
Dry Foam 1.7 1.2 3.3 1.4 1.3
Wet 1.0 2.0 2.3 2.0 2.0
Wrung 1.3 1.5 2.7 1.8 1.6
As a design aid, the unit recovery rate for foam might be defined as the
equivalent dry weight of foam recovered per hour per unit width of
system -
114
-------
Where
Q| = Ib/hr - ft width
t = thickness of foam layer - ft
avg
V« = belt velocity - ft/sec
d =: apparent foam density - lb/ft3
The apparent or packing density of typical batches of mulched foam
has been determined for four conditions:
a. Dry
b. Water-saturated, driven by current
c. Water-saturated, drained
d. Water-saturated, wrung
These values are of interest in estimating mass flow and power
requirements for all system components. An estimate of the average
density during transit from water surface to deck may be of use in
harvester design. The average (condition e) of (b) and (c) was
used in Section XIV.
115
-------
SECTION XI
WRINGING
introduction
The ability to efficiently recycle the polyurethane foam sorbent depends
upon the ability to remove oil quickly from the foam during the relatively
short cycle time. A simple roller wringer appears to best satisfy the
need for a continuous, simple, reliable system. A wringer can be readily
integrated with a conveyor system. Rollers can be easily fabricated from
large diameter pipe and may be filled with water or other fluids to in-
crease wringing pressures.
Wringing Experiments with Oil and Foam Cubes
A wringing apparatus was constructed for study as pictured in Figure 70
and shown schematically in Figure 71. The unique feature of this design
is the wire mesh conveyor belt which supports the foam as it passes through
the wringer. The conveyor mesh (purchased from Cyclone Fence Sales,
U. S. Steel Corporation, Houston, Texas) had openings 1/4-inch x 1/2-inch
and was 1/8-inch thick, as shown in Figure 72. Both rollers in our early
experiments had diameters of 6-3/4 inches and were 36 inches long. The
top roller was free to move vertically so that the foam was wrung under
constant pressure. The wringer was operated in later experiments with a
24-inch diameter top roller (Figure 6). Wringing pressure was varied by
filling the rollers with water or by hanging weights on the ends of the
top roller (Figure 70). When the small diameter wringer roller was used,
the conveyor was driven by a hand crank. For the experiments with the
large diameter roller the conveyor was driven by a 3-horsepower motor
through a variable speed reduction unit.
Initial experiments were performed by wringing oil from 2-inch foam cubes.
Details of the procedures are given in Table 24. Decreasing the conveyor
speed, increasing the roller weight, and increasing the number of passes
through the wringer resulted in more complete oil-sorbent separation.
Significantly, the rate of improvement in the wringing performance dimin-
ished with increasing wringing pressure, decreasing conveyor speed, and
increasing number of passes through the wringer. Thus, the satisfactory
.performance found at manageable roller weights and conveyor speeds coul
not be greatly improved. The experiments showed that there was residual
oil which could not be removed even by extensive squeezing. The amount
of residual oil increased with increasing oil viscosity.
Effect of Wringer Pressure, Conveyor Speed, and Number of Passes Through
the Wringer
Figure 73 shows the oil remaining in the foam as a function of the "pressure"
imposed by the wringer. The foam pressure is equal to the roller weight
divided by the length of the foam presented to the roller. (For the foam
pictured in Figure 71 the length of foam is 12 inches.) For each experiment
117
-------
Figure 70 - PHOTOS OF APPARATUS WRINGING MULCHED FOAM IN
ARRANGEMENT TYPICAL OF THAT USED IN THE
EXPERIMENTS
118
-------
Foam Arrangement
(2-inch cubes)
Floating Roller
(6-3/4-inch dia.)
Fixed Rollers
(6-3/4-inch dia. )
Wire Belt
Conveyor
Side V iew
1
-
"
( '
4
t
1
££
10-ft
^-
1
1 1
'
1
1,1
.
1
J
n
D
*•
c^p
o
Top View
FIGURE 71 SCHEMATIC OF WRINGER, CONVEYOR, AND FOAM
ARRANGEMENT
119
-------
Figure 72 DETAIL OF WIRE MESH CONVEYOR BELT
TABLE 24
WRINGING EXPERIMENTS USING 2-INCH FOAM CUBES
Equipment:
1. Roller wringer using 6-3/4" diameter roller (see Figures 70 and 71)
2. A triple beam balance
3. Supply of 2-inch foam cubes
Experiment Procedure:
1. Determine dry foam weight.
2. Place foam cubes on an excess of oil until fully saturated.
3. Allow oil to drain until free oil is removed.
4. Weigh oil-saturated foam.
5. Arrange foam cubes on conveyor in pattern shown in Figure 71.
6. Accelerate conveyor to desired speed.
7. Weigh foam cubes after wringing.
8. Repeat steps 5-7 desired number of times.
Environmental Conditions :
Experiments performed outside, temperature 72-82 F
120
-------
Initial
Oil Volume/Foam Mass
(gal/lb)
,5 0.12
0)
to
i
,_^
* 0.10
p
\
*. X
>X\ "-I
X x
X ^|
-------
the foam was wrung four times. It may be seen that the advantage of
wringing the foam more than three times is minimal. With increasing foam
pressure there appears to be a limiting quantity of oil retained by the
foam at each wring.
Figure 74 shows that the oil remaining in the foam increases, though not
greatly, with increasing conveyor speed. After four wrings, the amount
of oil remaining in the foam at a conveyor speed of 35 ft/min is only 15$
less than that remaining at 105 ft/min.
The results of these experiments can be correlated by assuming that the
rate of oil removal from the foam is given by
= -GP (Q - q.) (11)
where Q is the volume of oil in the foam, 0^ is the oil permanently retained
or trapped in the foam, P is the weight of the roller divided by the length
of the foam under the roller and 6 is a dimensional constant.
Integrating from the initial oil volume, Q , at t = 0,
Assuming that t, the time the foam is exposed to the wringer pressure, can
be expressed as the product of a dimensional constant and the number of
passes through the wringer, N, divided by the wringer roller RPM, w, we
have
-G'PN
Figure 75 presents a correlation of the data using the parameters PN/w and
(Q ~ °oo)/(Q • QQ) suggested by Equation 13. The data are for a range of
roller weights and conveyor speeds, for four different viscosity oils, and
for each of four passes through the wringer. The data show by virtue of
the correlation that the percent of oil removed is independent of the vis-
cosity of the oil; but, as shown in Table 25, the increased viscosity
affects the value of 0^, the oil volume which is not removed by wringing.
The data in Figure 75 appear to intercept the y-axis at 0.4, but before
wringing the value must be 1.0. This suggests that an appreciable amount
of the oil, possibly the oil on the surface of the cubes, is removed by
some mechanism not described by Equation 13.
Effect of Recycling and Aging of Sorbent
In Table 26 results are presented from a series of tests performed to study
the effect of the number of cycles and the oil-sorbent exposure time on
wringing performance. The table shows that wringing performance is not
122
-------
CO
00
0.71-
0.6 -
Initial
Oil Volume/Foam Mass
(gal/lb)
O 1.61
X 1.76
A 1.85
No. 2 Fuel Oil
Foam Pressure
6.21 Ib/in
I 0.5
0
CK
Q
° 0.4
co
CO
co
33
**-\
CO
£ 0.3
c
'i~t
i-l
•r-i
o
o 0.2
" " "— - •— " "
)" "^
C
i
D
a
K!
^3
C
! 1
^ ^- -*:
k
_ — — - "^
^!
c
-H
1-1
V
ex
u;
•a
i r
i
±
\.
i
20
80
ICi
Conveyor Speed (ft/min)
FIGURE 74 - VOLUME OF OIL/FOAM MASS VS CONVEYOR SPEED
123
-------
1.0
0.1
Q.Oi
Equation (13)\
with G1 -
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
~ - (Ib/in) (number of wrings)/(RPM)
Experimenal
procedure given
in Table 24.
FIGURE 75 -
CORRELATION OF WRINGING DATA FROM EXPERIMENTS USING
ONLY OIL AND 2-INCH FOAM CUBES
124
-------
TABLE 25
PERMANENTLY RETAINED OIL IN 2-INCH
POLYURETHANE FOAM CUBES
Conditions:
Oil retained by foam after 4 wrings using procedure described in
Table 25.
Conveyor Speed = 70.4 ft/min
Foam Pressure = 18 Ib/min
Oil
No. 2 Diesel
Carnea 15
Carnea 21
Shallow Yates
Viscosity at 80 F
(cs)
3.5
14.0
31.0
105.0
(gaL/lb dry foam)
0.33
0.43
0.48
0.63
TABLE 26
EFFECT OF TIME AND WRINGING CYCLES ON WRINGING PERFORMANCE
(All tests performed by recycling foam first used in Run No. 1)
No. 2 Fuel Oil Used for Tests
Run
1
7
8
9*
(Volume of Oil in Foam) /(Mass of Dry Foam)-gal/lb
Initial
1.83
Above |
weighii
1.83
6 hour
1.67
26 day
0.96
1st wring
0.50
procedure
ig and witl
0.58
2nd wring
0.34
[Run No. 1'
i no time :
0.41
time lapse
0.60
0.42
time lapse
0.47
0.33
3rd wring
0.29
4th wrinit
0.27
repeated five times
.apse between tests.
0.32
0.34
0.27
0.29
0.30
0.24
Final Oil After
4th Wring,
$ Increase
over Run No. 1
..
without
8.7
14.8
-10.9
* For Run No. 9 the foam was not soaked in oil for as long a time as in
the previous runs. This resulted in the lower initial oil content in the
cubes.
Foam Pressure « 21.7 Ib/in
Conveyor Speed = 70.4 ft/min
Experimental procedure given in Table 24.
125
-------
greatly affected by recycling nor by storing the foam in contact with oil
for a time period of 26 days. The data of Table 27 show that there is
little or no effect on wringing performance of aging of dry foam for
periods to one month.
TABLE 27
EFFECT OF FOAM AGING ON WRINGING PERFORMANCE
Run
1
2
3*
(Volume of Oil in Foam) /(Mass of Dry Foam) -
(gal../lb)
Initial
1.83
One we
1.73
One moi
1.09
1st wring
0.50
2nd wring
0.46
*k time lapse
0.48 0.34
ith time li
0.46
ipse
0.32
3rd wring
0.29
0.29
0.27
4th wring
0.27
0.26
0.24
Final Oil After
4th Wring
% Decrease Over
Run No. 1
--
4.3$
8.6$
* For Run No. 3 the foam was not soaked in oil for as long a time as
for Runs No. 1 and 2. This resulted in the lower initial oil content.
Foam Pressure = 21.7 Ib/in.
Conveyor Speed = 70.4 ft/min
No. 2 Fuel Oil used for tests.
Experimental procedure given in Table 24.
Wringing Oil and Water from Mulched Foam
The results of the experiments using both oil and water agree qualitatively
with those described above. The experiments were of a practical nature.
Mulched foam (see Table 12 for the foam size distribution) was placed on a
water-filled 5-foot diameter tank on which oil had been spread. The foam
was allowed to sorb oil and water for approximately three minutes, lifted
from the tank, allowed to drain, and successively wrung. The foam was
recycled in this manner, simulating the recycling process anticipated for
the actual collection system. The detailed experimental procedure is
given in Table 28.
The first experiments were performed using the small diameter roller,
2.2 Ib of dry foam, and 0.65 gal. No. 2 Fuel Oil. The oil was placed on
the tank for each cycle resulting in an initial slick thickness of 0.06 in.
In Figure 76 the transient behavior of the foam during recycling is presented,
The oil, water, and total liquid recovered by the four wrings in each cycle
are plotted as a function of the number of the soaking-wringing cycle. In
each case the fluid volumes plotted have been divided by the dry weight
of the foam. Figure 76 shows that the transient process continues over the
first four cycles, during which oil content of the effluent increases from
126,
-------
TABLE 28
WRINGING EXPERIMENTS USING MULCHED FOAM
Equipment;
1. Roller wringer using 6-3/4" diameter roller or 24" diameter
roller as specified. (See Figures 89, 6, and 92.)
2. A 5-foot diameter 3-foot deep thin walled tank.
3. A hanging scale.
4. Nylon net, 1/4" mesh.
5. Mulched foam (size distribution given in Table 12).
6. 2000 ml graduate.
Experimental Procedure;
1. Determine dry foam weight (typically 2.2 Ib).
2. Place two feet of water in 5-foot diameter tank.
3. Place desired quantity of oil on the water surface (typically
0.65 gallons). Note: This amount of oil is not sufficient to
saturate the foam.
4. Distribute foam evenly over water surface and allow it to
sorb oil for 2-3 minutes until all oil is in foam, by visual
inspection. To promote oil contact, move foam slowly over
water surface.
5. Lift oil-soaked foam from tank using nylon net. Allow free
oil to drain off.
6. Weigh oil soaked foam.
7. Arrange foam on wringer conveyor in a layer of the desired
depth. Figures 89 and 92 show typical foam arrangement.
8. Start conveyor at desired speed.
9. Weigh wrung foam.
10. Repeat steps 7-9 as desired.
11. Measure oil and water volumes removed by wringing.
12. Clean water surface on 5-foot diameter tank.
13. Repeat steps 3-12 using the same foam sample for the desired
number of cycles.
Environmental Conditions;
Experiments performed outside, temperature 55-75 F.
127
-------
00
(D
00
s:
00
•r-l
0)
E
Eb
01
1
3
cr
0.5r
0.4
Oil Plus Water Recovered from Foam
Oil Recovered from Foam
0.1 -
0
Experimental Procedure is given in Table 28 " ReC°Vered from Foam
No. 2 Fuel Oil - 0.65 gal/cycle, dry foam weight = 2.2 Ib
Conveyor Speed = 70.4 ft/min
Foam wringing pressure =7.5 Ib/inch
6.75-inch diameter roller
Foam chunks loosely packed in 2-inch deep layer 2-feet wide
I I I I I
8
Number of Soaking - Wringing Cycles
FIGURE 76 - TRANSIENT BEHAVIOR OF SOAKED FOAM DURING RECYCLING
-------
approximately 10$ to greater than 50$. The amount of water removed
from the foam is approximately constant, suggesting that water is not
accumulating within the foam during this initial period. One encour-
aging result is that the total liquid recovered increases during the
transient period. In Figure 77 the wringing performance during the first,
third, sixth, and eighth cycle is presented. The weight of the liquid re-
tained by the foam divided by the dry foam weight is plotted versus the
number of times the foam is wrung. It is seen that, except for the first
cycle, the weight of the liquid retained after four passes through the
wringer is unchanged. In Table 29 the results are presented of an experi-
ment to determine the percent oil removed during successive wrings. The
experiment was performed using the foam sample which had been cycled eight
times as described above. Table 29 shows that the percent oil removed
increases only slightly during successive wrings.
s:
00
e
CO
o
V
en,
3
T-t
J
O
.C
00
0
Soak-Wring
Cycle No.
X - i
D -
- 6
Experimental Procedure Given in Table 28
No. 2 Fuel Oil - 0.65 gal/cycle, Dry Foam Weight =2.2 Ib
Conveyor Speed =70.4 ft/min.
Foam Wringing Pressure =7.5 Ib/inch.
6.75-Inch Diameter Roller.
Foam chunks loosely packed in 2-inch deep layer
2-feet wide.
1 I I 1
Number of Times Through Wringer
Figure 77 - WRINGING PERFORMANCE FOR DIFFERENT
NUMBER OF CYCLES
129
-------
Oil
2.1 x I0'\
9.4 x 10
5.7 x 10~2
3.5 x 10"
2.1 x 10"2
1.7 x 10"
1.0 x 10"
4.42 x 10"1
Water
2.9 x 10"1
1.1 x 10"1
5.2 x 10"2
3.5 x 10"2
9.9 x 10"3
1.5 x 10 ,
5.2 x 10
5.1 x 10"1
Total
4.9 x 10'J"
2.0 x 10 ~~
1.1 x 10"
7.0 x 10"2
3.1 x 10~2
3.2 x 10 ,
1.6 x 10"
9.5 x 10"1
% Oil
42
47
52
50
68
52
67
TABLE 29
PERCENT OIL REMOVED BY SUCCESSIVE WRINGINGS
Conditions:
Experimental procedure given in Table 28.
No. 2 Fuel Oil, Conveyor Speed, = 70.4 ft/min, Foam Wringing
Pressure =7.5 Ib/in.
6.75-in. diameter roller.
Mulched foam loosely packed in 2-inch deep layer 2-feet wide.
Liquid Removed (gal.)
Wring No.
1
2
3
4
5
6
7
For the experiments discussed above, the mulched foam was arranged in a
2-inch deep layer on the conveyor. As shown in the top photo of Figure 6,
experiments were performed using the 24-inch diameter roller when the mulched
foam was 6 inches deep. The results of these experiments are presented in
Figures 78 and 79. The experiments were performed in exactly the same
manner as described above except that the foam sample had a dry weight of
11 Ib, and 3.25 gaL of oil were used for each cycle. The roller was fabri-
cated from pipe having a 1/2-inch wall thickness, and the pipe plus the
circular steel caps had a combined weight of 400 Ib. For the first four
cycles shown in Figures 78 and 79, the empty pipe was used as a roller,
but for the fifth cycle the roller was partially filled with water and had
a weight of 700 Ib. With the roller completely filled with water, it
weighed close to 1000 Ib. Comparing Figure 77 and Figure 78, it is seen
that the residual oil is approximately the same for the experiments performed
with the 2-inch deep mulched foam layer and the 6 3/4-inch diameter roller
as for the experiments performed with the 6-inch deep mulched foam layer
and the 24-inch diameter roller. Comparing Figures 76 and 79, the transient
process in which oil accumulates within the foam is similar for the two
experiments.
Wringing Experiments_ with_H]Lgh_ Viscosity Oils
Wringing experiments were performed using mulched foam and high viscosity
oils. The technique used is described in Table 28. Two experiments were
performed, the first using 148 cs oil and the second using 1100 cs oil.
Table 30 gives the weight of the foam before and after wringing and the
volumes of oil and water recovered for the two experiments. Table 31 shows
130
-------
x;
oo
ctf
O
•o
01
t-l
0)
JJ
IV
OS
3
cr
x:
oo
Q Cycle No. 1
^ Cycle No. 2
x Cycles No. 3 and No. 4
A Cycle No. 5
Temperature = 67 F
Cycle
No.
1-4
5
Foam
Pressure
Ib/in
16.7
29.2
Conveyor
Speed
f t/min
71
71
No. 2 Fuel Oil - 3.25 gal/cycle, dry foam weight = 11 Ib
Foam chunks loosely packed on wringer conveyor 6-inch
deep layer 2-feet wide.
Experimental Procedure Given in Table 28
I.
J
01234
Number of Times through Wringer
FIGURE 78 - WRINGING PERFORMANCE USING 2-FT DIAMETER ROLLER
131
-------
00
1-1
£
to
o
Eh
D
O4
0.032
0.028
0.024
0.020
0.016
0.012
0.08
0.04 -
Cycle
No.
1 - 4
5
Foam
Pressure
Ib/in
16.7
29.2
Conveyor
Speed
ft/min
71
71
Oil Plus Water
Oil
Water
No. 2 Fuel Oil - 3.25 gal/cycle, dry foam
weight = 11 lb
Foam chunks loosely packed on wringer conveyor
6-inch deep layer 2-feet wide.
Experimental Procedure Given in Table 28
Cycle Number
FIGURE 79 - LIQUID REMOVED FROM FOAM USING 2-FT DIAMETER ROLLER
132
-------
OJ
10
TABLE 30
BEHAVIOR OF FOAM DURING RECYCLING
Conditions:
Procedure given in Table 28.
Temperature = 55°F.
Conveyor Speed =70.4 ft/min
Foam wringing pressure = 7.5 lb/in., three wrings/cycle, 6.75-in. roller.
Foam chunks loosely packed in 2-in. deep layer 2-ft wide.
Final*
Initial Wt (Ib) Water Oil Weight Not*
Cycle Soaked After Weight Recovered Recovered Oil and Water Accounted
No. Wt (Ib) 3 Wrings Loss (Ib) (gal.) (gal.) Recovered (Ib) For (Ib)
Oil:
1
2
3
4
Oil:
1
2
3
60$ No.
20.9
23.9
24.9
23.8
80$ No.
26.9
24.4
24.9
6 Fuel Oil
14.7
15.0
15.0
15.0
6 Fuel Oil
13.4
15.4
15.0
+ 40$ No. 2 Fuel
6.2
9.0
9.9
" 9.0
+ 20$ No. 2 Fuel
7.5
9.0
9.9
Oil--148 cs,
i
5.0 x 10 i
4.2 x 10
7.5 x 10~
4.7 x 10
Oil- -1100 cs
8.3 x 10~x
6.8 x 10
7.8 x 10"
2.2 Ib dry foam,
_ 2
2.6 x 10 .
_i
1.3 x 10
2.1 x 10"X
2.0 x 10
, 2.2 Ib dry foam
2.6 x 10"2
3.9 x 10
6.5 x 10"2
0.65 gaL of
4.4
4.6
8.1
5.7
,0.65 gaL of
7.3
5.9
7.0
oil/cycle.
•1.8
4.4
* 1.8
3.3
oil/cycle.
0.2
3.1
2.9
* Due to oil, water, and foam retained on conveyor and oil collection tray.
-------
TABLE 31
WRINGING PERFORMANCE OF HIGH VISCOSITY OILS
Conditions:
Procedure given in Table 28.
Wringer pressure =7.5 Ib/in., 6.75-in. roller
Conveyor speed = 70.4 ft/min
2.2 Ib dry foam, 0.65 gal oil/cycle
Foam chunks loosely packed in 2-in. deep layer 2 ft wide
Experiment No. 1 Oil Viscosity 148 cs - 60$ No. 6 Fuel Oil and 40$
No. 2 Fuel Oil at 55°F.
Cycle
1
2
3
4
Fluid Weight in Foam/Dry Foam Weight
Initial
8.51
9.87
10.32
9.87
After 3 Wrings
5.7
5.7
5.7
5.7
Experiment No. 2 Oil Viscosity 1100 cs -
No. 2 Fuel Oil at 55°F.
No. 6 Fuel Oil and 20$
Cycle
1
2
3
Fluid Weight in Foam/Dry Foam Weight
Initial
8.5
10.1
10.3
After 3 Wrings
5.1
6.0
5.8
134
-------
the weight of fluid in the foam divided by the dry foam weight before and
after wringing. Comparison of the results presented here with those for
the low viscosity No. 2 Fuel Oil presented in Figures 76 and 79 shows that
the oil recovered is considerably less for the high viscosity oils. The
fluid weight contained in the foam per weight of dry foam after wringing
is 5.1 to 6.0 for the high viscosity oils versus 3.6 to 4.1 for the No. 2
Fuel Oil. This higher residual oil content is consistent, however, with
the results presented in Table 25 for the experiments performed with the
high viscosity oils and the foam cubes.
Wringing high viscosity oil from the foam can result in foam attrition.
Figure 80 shows the effects of oil viscosity on the appearance of wrung
foam. The experiments which produced these results were performed by
wringing individual foam cubes soaked in oil with a wringing pressure
much higher than that which would be encountered in the actual sorbent
wringing system. These experiments illustrate the type of failure that
can occur if the viscosity of the sorbed liquid(s) limits the rate at
which it can flow out the foam pores in response to rapid application of
wringer pressure.
Figure 80 - FOAM FAILURE DURING WRINGING RESULTING FROM
INCREASING OIL VISCOSITY
Wringing No. 2 Fuel Oil and water from mulched foam resulted in little or
no attrition. A 2.2 Ib sample was wrung more than 50 times with no notice-
able attrition. However, for the experiments conducted using the high
viscosity oils and the mulched foam, moderate foam attrition was observed.
135
-------
Table 32 shows the mulched foam particle size distribution before and after
wringing the indicated number of times. It appears that attrition of the
large foam chunks (3" x 3" and 2" x 2") is the most severe, but that the
smaller 1" x 1" foam chunks are better able to'withstand wringing. If we
assume that foam particles smaller than 1/2 " x 1/2" will not be recycled
but removed from the sorbent stream on the work boat deck prior to redis-
tribution, the data in Table 32 indicate that from 2$ to 3$ foam make-up
per cycle would be sufficient to offset the attrition due to wringing for
the highest viscosity oil tested.
TABLE 32
WRINGING ATTRITION OF MULCHED FOAM
Wringing Conditions:
Procedure given in Table 28.
Three wrings per cycle, wringer pressure on foam 7.5 lb/in., conveyor
speed 70.4 ft/min, 6.75-in. diameter roller.
2.2 Ib of dry foam, 0.65 gal oil/cycle.
Mulched foam loosely packed in 2-in. deep layer 2-ft wide.
Foam Retained on Screen. Percent by Weight
Initial After 6 Cycles* After 3 Cycles"*"
Screen Size Distribution Oil Viscosity - 148 cs Oil Viscosity - 1100 cs
4" x 4" 0 0 0
3" x 3" 9.2 0 4.4
2" x 2" 50.0 11.3 23.5
1" x 1" 36.8 66.5 61.2
1/2" x 1/2" 3.5 19.4 9.3
Pass 1/2" x
1/2" 0.5 2.8 1.6
* 60$ No. 6 Fuel Oil + 40$ No. 2 Fuel Oil at 55°F
+ 80$ No. 6 Fuel Oil + 20$ No. 2 Fuel Oil at 55°F
Draining Rate from Mulched Foam
Table 33 presents the results from draining experiments. The experiments
were conducted by confining foam in a two-foot by two-foot wire mesh box which
was lifted from the oil covered water surface after the foam had been
allowed to sorb oil and water for two minutes. Experiments were performed
at two foam-water surface densities for the low viscosity No. 2 Fuel Oil.
For these experiments the oil drained was always less than 3$ of the total
oil removed by both draining and subsequent wringing. For the more viscous
oil, 5.3$ of the liquid removed drained as the foam was lifted from the
water.
136
-------
TABLE 33
OIL DRAINED AS OIL-SOAKED FOAM IS LIFTED FROM WATER
(Initial Oil Slick Depth 0.06 In.)
Experimental Procedure;
1. Distribute foam uniformly over 21 x 2', 1" mesh wire box.
2. Lower box below oil covered water surface until the foam
floats in the confined space formed by the sides of the box.
3. Allow che foam to sorb oil for 3 minutes.
4. Lift wire box from the water in approximately one second.
5. Collect effluent in a tray placed under the wire box.
Foam Surface Oil Viscosity Oil Drained Oil Removed Percent Oil
Density (cs) (gal-) by Wringing Drained
(lb/ft2) (gal.)
0.13 4.0*
0.064 4.0
0.16 148*
0.0026 to 0.0052
0.0026 to 0.0039
0.01
0.24
0.16
0.19
1.1 to 2.1
1.6 to 2.4
5.3
+ No. 2 Fuel Oil at 72°F.
* 60$ No. 6 Fuel Oil + 40$ No. 2 Fuel Oil at 72 F.
Figure 81 shows the rate at which liquid drains from the foam as it is
pulled from the water. The 1.1 Ib sample of mulched foam was confined
by the two-foot by two-foot wire box to an approximate thickness on the
water of two inches and was pulled suddenly from the water (removal time
approximately one second). The rate at which the weight decreased during
draining was observed. It is seen that more than 2/3 of the liquid
which drains is removed in the first 40 seconds.
Oil Contamination J.n Water Removed by Wringing
The experiments summarized in Table 34 were performed to determine the
oil content of the water separated from the wringer effluent. The
experiments consisted of placing 0.65 gaL of oil on the water surface.
2.2 Ib of foam were added to the surface, allowed to soak oil, removed, and
wrung three times. (For the detailed procedure see Table 28.) In
all cases the oil/water mixture removed by wringing was allowed to
separate by gravity for one hour prior to sampling the water phase.
Visually, the separation was observed to cease after 15-30 min. All
water samples were highly turbid, having a brownish hue which indi-
cated they were highly contaminated. As shown in Table 34 the oil
content varied from 630 to 1700 ppm indicating a need for additional
treatment of the effluent water before it is discarded overboard.
137
-------
20 *—
1
The experimental procedure is the same as that reported in Table 33
except that the hanging weight of the wire mesh box containing the
soaked foam is visually monitored using a spring scale.
u
4:
oo
•1-1
CO
o
15
10
OJ
00
t>0
•l-l
g
CD
O
0
Water Alone
-X 0.65 gal No. 2 Fuel Oil
Plus Water at 72°F.
0
40
80
120
160
200
240
Draining Time - seconds
FIGURE 81 - DRAINING RATE OF OIL AND WATER SOAKED FOAM
-------
TABLE 34
OIL CONTAMINATION IN WATER REMOVED BY WRINGING
The wringing procedure used to collect the effluent is described in
Table 28.
Test PPM Oil in
No. Oil Used Foam Used Water Used Effluent
1 No. 2 Fuel Oil Same used in Same used in 980
sinking tests sinking tests
2 No. 2 Fuel Oil Same as No. 1 Clean 1700
3 Shallow Yates New foam Clean 1050
Crude Oil
4 Shallow Yates Same as No. 3 Same as No. 3 730
Crude Oil
139
-------
SECTION XII
FOAM DEGRADATION DURING RECYCLING
Introduction
It is expected that the cost of oil recovery will depend on the
quantity of foam lost during each cycle. To identify sources of
foam degradation during recycling, two test runs were made in which
all parts of the system were operated in sequence through several
cycles, and measurements were made of foam particle size distribu-
tions and foam losses at three points during each cycle. Different
oils were used for the two test runs. The properties of the on-
site generated polyurethane foams used in these tests are shown in
Table 8.
Degradation Caused by Hay Blower
Prior to testing the total system, two runs were made to study
particle size degradation of the polyurethane foam as a result of
passage through the hay blower. One run was made with the beater
chains installed (see Table 35). The beater chains caused progres-
sive reduction in particle size with each successive pass. In the
second run, the chains were removed after the first pass through the
blower. Although some reduction in particle size was observed after
the first pass, the rate of reduction was far less than was observed
with the beater chains installed (compare Table 36 with Table 35).
Limited tests were made to explore the comminution of a higher
tensile strength polyurethane foam for comparison with that used
above. The material chosen was Scott Industrial Foam, polyester
type, reticulated (Scott Paper Company, Chester, Pennsylvania)
having a density of 1,98 lb/ft3 and pore size of 64 ppi. Comparison
of the properties of the sample used, shown in Table 37, with Table 9,
show the Scott foam to have a tensile strength about seven times
higher and compressive strength several times greater than that gen-
erated on site. Pore sizes and densities of the two foams are quite
similar. Oil sorption tests indicated the two foams are comparable
(see Tables 38 and 3).
Two runs were made with the Scott foam to study particle size
degradation as a result of passage through the hay blower. One run
was made with the beater chains installed (Table 39). Although some
reduction in particle size was observed with multiple passes through
the blower, the rate of reduction was far less than was observed with
the foam generated on-site. (Compare Table 39 with Table 35). A
second run was made with the beater chains removed after the first
pass through the blower. As shown in Table 40, virtually no particle
141
-------
1
0
4.9
43.7
43.5
5.3
2.5
2
0
1.3
32.1
55.9
8.7
2.0
5
0
0
13.3
71.3
13.6
1.6
10
0
0
0
64.2
30.6
5.2
TABLE 35
PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES OF ON-SITE
GENERATED POLYURETHANE FOAM THROUGH REINCO HAY BLOWER—BEATER
CHAINS IN PLACE
Percentage of Foam Retained on
Screen after Number of Passes
Screen Size. In. Through Hay Blower
4x4
3x3
2 x 2
1x1
0.5 x 0.5
<0.5 x 0,5
TABLE 36
PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES OF
ON-SITE GENERATED POLYURETHANE FOAM THROUGH
REINCO HAY BLOWER--BEATER CHAINS REMOVED AFTER
FIRST PASS
Percentage of Foam Retained on
Screen after Number of Passes
Screen Size, In. Through Hay Blower
4x4
3x3
2x2
1x1
0.5 x 0.5
<0.5 x 0.5
1
0
14.0
47.8
32.1
3.7
2.4.
2
0
11.7
45.7
34.7
4.1
3.9
_5
0
2.9
40.9
47.1
5.0
4.1
10
0
2.4
26.4
56.8
7.9
6.5
142
-------
TABLE 37
PROPERTIES OF SCOTT INDUSTRIAL POLYURETHANE FOAM
Density, lb/ft3
Pores/Inch
Tensile Strength, lb/in2
Compressibility, lb/50 in2
25% Compressed
50$ Compressed
65% Compressed
1.98
64
23
39
45
55
TABLE 38
COMPARISON OF MAXIMUM OIL RETENTION BY POLYURETHANE FOAMS
AFTER FIVE MINUTES DRAIN TIME WHILE SUSPENDED IN AIR
Test Conditions:
Two-inch cubes of foam saturated with test oil
prior to drain period. After draining, foam
cube was. weighed and then test oil was removed
by successive washing with hexane, VM&P naphtha,
and pentane, and then dried
No. 2 Diesel
Blend
No. 6 Fuel Oil
Oil Held by Foam. Ib/lb Foam*)
On-Site Foam Scott
7.7 6.9
13.2 9.7
30.4 27.6
1) See Tables 3 and 37 for properties.
143
-------
1
39.1
28.0
15.4
12.0
2.7
2.5
2
20.3
41.8
18.3
13.7
2.5
3.1
5
19.7
28.8
28.6
15.1
2.7
4.9
10
8.4
33.4
32.2
18.7
3.0
4.2
TABLE 39
PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES OF SCOTT INDUSTRIAL
POLYURETHANE FOAM THROUGH REINCO HAY BLOWER--BEATER CHAINS IN PLACE
Percentage of Foam Retained on
Screen after Number of Passes
Screen Size. In. Through Hay Blower
4x4
3x3
2x2
1x1
0.5 x 0.5
< 0.5 x 0.5
TABLE 40
PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES OF SCOTT
INDUSTRIAL POLYURETHANE FOAM THROUGH REINCO HAY BLOWER --
BEATER CHAINS REMOVED AFTER FIRST PASS OF FOAM THROUGH BLOWER
Percentage of Foam Retained on
Screen after Number of Passes
Screen Size. In. Through Hay Blower
4x4
3x3
2x2
1x1
0.5 x 0.5
< 0.5 x 0.5
1
55.3
19.2
13.1
8.6
1.4
2.2
2
51.5
22.1
14.8
8.3
0.9
2.5
5
50.2
16.7
16.4
9.2
0.8
6.7
10
49.1
23.8
13.5
9.6
0.9
2.9
144
-------
size reduction occurred after the first pass through the blower. These
data show that higher tensile strength of the foam can significantly
reduce the rate of comminution during distribution using a hay blower.
Foam particle damage during wringing would be expected to be less than
observed with the on-site foam, although no tests were made (see page 135
of Section XI). The Scott Industrial Foam is considerably more
expensive than on-site foam (approximately $4.00 per pound vs 50»{ per
pound for the on-site generated foam) and qualitative tests show the
Scott foam will eventually become water saturated and sink when placed
on quiescent fresh water. Thus the advantages of high tensile strength
are not without compensating penalties.
Tests of Comple.te System
The equipment items used for tests of the system and the arrangement
for the tests are shown in Figures 82 and 87. A sheet metal bin was
used to transfer the mulched polyurethane foam to the hay blower, and
a trough was .provided on the hay blower to avoid loss by spillage.
The sheet metal bin was weighed empty before the test and also when
full of foam in order to ascertain the weight of foam used. Initial
weight of dry foam was approximately fifty pounds for each run. The
foam passed through the Reinco hay blower, equipped with beater chains
for the first pass to mulch the foam. The beater chains were removed
after the first pass so the unit acted only as a centrifugal blower.
The blower discharged into a 16-inch sheet metal duct which conducted
the mulched foam to a diffuser section designed to discharge the
foam uniformly onto the surface of the current tank immediately down-
stream of where oil was being discharged onto the water surface.
Water velocity in the current tank was adjusted so that the transit
time on the water surface was approximately one minute. Oil was placed
on the water in the current tank in a nearly uniform layer, approximately
0.04-in. thick, by use of a feed tank which discharged into a weir set
across the current tank. An inclined board conducted the oil from the
weir nearly to the water surface.
The harvester belt was set at an angle of 40° and adjusted to pick up
the foam at about the rate of its arrival. Downstream of the harvester
in the current tank was a 1/8-inch mesh screen which caught substan-
tially all of the foam particles which by-passed the harvester. This
foam was considered to be that which was lost from the system on each
cycle.
The harvester dumped the foam onto a conveyor belt which in turn
discharged into a chute which conveyed the foam to the wringer equipped
with two 24-inch diameter rolls. The wringer discharged the "dry" foam
back into the sheet metal bin and the liquids were pumped to a waste
oil storage tank. The use of the 24-inch diameter rolls resulted in
very high unit loading during wringing, which is a relatively severe
test condition. No. 2 Diesel Fuel and a mixture of No. 2 Diesel Fuel
(35#) and No. 6 Fuel Oil (65$) were used. These oils,.were chosen to
given an indication of the effects of oil viscosity on foam particle
145
-------
Figure 82 - FEEDING MULCHED FOAM FROM TRANSFER
BIN INTO REINCO BLOWER
Figure 83 - NOZZLE USED TO DISTRIBUTE
FOAM ONTO WATER SURFACE IN
CURRENT TANK
146
-------
JS-
Figure 84 - FOAM APPROACHING AND BEING
PICKED OFF WATER SURFACE
BY HARVESTER BELT
Figure 85 - HARVESTER BELT DISCHARGING
ONTO CONVEYOR BELT
-------
Figure 86 - FOAM FALLING DOWN CHUTE ONTO LINK CHAIN
BELT OF WRINGING APPARATUS. TWO 24-INCH
ROLLS IN WRINGER
Figure 87 -
WRINGER DISCHARGES DRY
FOAM INTO TRANSFER BIN
148
-------
size degradation. The viscosity-temperature relationships for
these two oils are shown in Figure 88.
The tests were run until half of the original foam charge had been
lost from the system. Particle size distribution of the foam was
determined at three points in the system: after the hay blower, after
the harvester, and after the wringer. Thief samples were taken at each
location during each cycle of foam through the system and each sample
was spread on a white board, photographed, and returned to the sheet
metal bin. Particle sizes were measured visually from the photographs
and counted to determine the particle size distributions.
The foam recovered from the current tank ("lost" from the cycle) was
weighed after each cycle of each run. Because this foam was wet with
both water and oil, one sample from each run was cleaned by washing
with hexane, VM&P naptha, and pentane to remove the oil fraction and
then dried overnight in a vacuum oven to remove the water. From this
measurement the proportion of dry foam in the wet samples of "lost"
foam was determined and the amount of foam lost during each cycle
calculated. It should be noted that the foam lost from the system
during each cycle included some which was not buoyant (no more than
half of the total foam lost) and some which was basically buoyant but
was nevertheless swept under the diversionary booms on the harvester
or passed through the holes in the harvester belt.
Tests of foam degradation during re-cycling through the entire system
were made using on-site generated foam produced on June 21, 1972 (see
Table 8 for properties). As shown in Figure 89 over half of the foam
was lost from the system within four cycles for the No. 2 + No. 6 Fuel
Oil mixture and within five cycles for the No. 2 Fuel Oil alone. From
these data, the rate of loss of foam is approximately eight to ten
percent per cycle.
Foam was lost from the system partly due to comminution and partly1due
to loss of buoyancy, presumably because the few closed cells in the
foam were ruptured. To identify the causes of comminution and foam
loss, the particle size distributions measured at three points in
the cycle (after passing through the hay blower, after the harvester,
and after the wringer) are summarized in Tables 41 and 42. By comparison
with the data in Table 36, which shows the effect of the blower alone
on dry foam, it appears that passage through the whole system is more
damaging to the foam than is passage through the blower alone. The
largest single source of damage to the foam is believed to be the
blower, but it appears that wringing partly tears the foam particles,
making them more easily borken during subsequent passage through the
blower. Particle size degradation appeared to be more severe with the
more viscous oil (the No. 2 + No. 6 mixture). This is consistent with
our earlier work on wringing, wherein we found severe damage to the
foam particles at high oil viscosities (see Section XI).
149
-------
W
-------
100
oo
a
to
u
j-i
o>
Dn
80
70
60
50
30
No. 2 Diesel Fuel
65$ No. 6 Fuel Oil
No. 2 Diesel Oil
2 3
Cycle No,
FIGURE 89 -
POLYURETHANE FOAM REMAINING AFTER EACH CYCLE THROUGH
SYSTEM
151
-------
TABLE 41
PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES OF
ON-SITE GENERATED POLYURETHANE FOAM THROUGH WHOLE SYSTEM
— TEST OIL: NO. 2 DIESEL FUEL
Screen Size, In.
Percentage of Foam Retained on Screen
After Number of Cycles Through System
1
2 3
4
5
After Passage Through Blower
> 4 x 4 6.4
> 3 x 3 42.1
> 2 x 2 38.2
> 1 x 1 12.6
> 0.5 x 0.5 0.7
< 0.5 x 0.5 0.0
After Harvester
> 4 x 4 9.7
> 3 x 3 41.0
> 2 x 2 38.1
> 1 x 1 10.9
> 0.5 x 0.5 0.5
< 0.5 x 0.5 0.0
0.0
23.0
38.6
35.5
2.8
0.2
0.0
32,0
39.6
25.9
2.6
0.2
0.0
0.0
20.0
45.9
24.2
10.5
0.0
0.0
23.4
49.4
21.1
6.5
0.0
0.0
0.0
32.7
37.2
31.5
0.0
0.0
0.0
44.8
40.1
16.4
After Passage Through Wringer
> 4 x 4 9.0
> 3 x 3 38.2
> 2 x 2 35.5
> 1 x 1 16.5
> 0.5 x 0.5 0.9
< 0.5 x 0.5 0.0
0.0 0.0
23.3 26.0
39.5 22.1
32.1 46.1
4.9 5.3
0.2 0.6
0.0
0.0
11.1
63.5
19.1
6.8
0.0
0.0
0.0
57.6
32.5
10.6
152
-------
TABLE 42
PARTICLE SIZE DISTRIBUTION AFTER MULTIPLE PASSES OF
ON-SITE GENERATED POLYURETHANE FOAM THROUGH WHOLE SYSTEM
— TEST OIL: 35% NO. 2 DIESEL FUEL
65$ NO. 6 FUEL OIL
Screen Size, In.
Percentage of Foam Retained on Screen
After Number of Cycles Through System
After
> 4 x 4
> 3 x 3
> 2 x 2
> 1 x 1
> 0.5 x 0.5
< 0.5 x 0.5
After
> 4 x 4
> 3 x 3
> 2 x 2
> 1 x 1
> 0.5 x 0.5
< 0.5 x 0.5
After
> 4 x 4
> 3 x 3
> 2 x 2
> 1 x 1
> 0.5 x 0.5
< 0.5 x 0.5
1
Passage Through
21.9
36.2
30.1
11.1
0.7
0.1
Harvester
10.8
56.2
26.0
6.9
0.1
0.0
2
Blower
0.0
24.7
22.5
45.0
6.5
1.4
0.0
46.2
37.2
15.4
1.3
0.1
3
0.0
18.2
13.2
47.5
14.6
6.8
0.0
0.0
24.7
48.6
25.8
1.5
4
0.0
0.0
22.8
37.7
23.8
16.6
0.0
27.3
39.9
31.7
1.4
0.1
Passage Through Wringer
36.9
34.8
22.1
6.2
0.2
0.0
0.0
7.7
47.9
41.9
2.8
0.0
0.0
12.3
40.0
42.9
5.0
0.1
0.0
0.0
57.8
39.8
2.6
0.1
153
-------
SECTION XIII
FOAM DISPOSAL
Introduction
Techniques are needed to dispose of the polyurethane foam which has
been used to sorb oil without creating additional pollution problems.
The majority of our studies have concentrated on burning, though
solution, compaction, and burial were also considered.
Disposal by Burning
Three separate thermal degradation studies were made to determine the
effects of temperature on disposal of the on-site generated foam by
incineration. Results of the first study showed that polyurethane foam
can be fused to form a compact solid at a temperature of about 330°F to
350°F, indicating that the foam begins to melt and run at about 350°F.
A thermogram of the foam produced on November 2, 1971 (see Table 8) was
made by measuring the weight loss as the temperature of the foam was
continuously raised at 9 C per minute. The results are reported in
Table 43 and Figure 90. The major portion of the polyurethane is
vaporized at a temperature of about 800 C. The remaining 25% w appears
to be carbon.
Mass sgectrographic analyses of gases evolved from the November 2 foam
at 200 C and 375 C were obtained. The analyses, reported in Table 44
show that ammonia and water were the only contaminants found that are
not typical of normal, dry laboratory air. Weight loss versus sample
temperature is shown in Figure 91. The weight loss at the outgas
test conditions agrees reasonably well with the weight loss while
constantly increasing the surface temperature at 9°C per minute as
shown in Figure 90. These tests indicate the foam can be burned with-
out the need for high ignition temperatures and further, that the
burning operation not not hazardous.
A model furnace, is constructed after a limited literature search
indicated commercially-available furnaces could burn a maximum of
about 30#w plastic foam mixed with other combustibles (e.g., wood,
paper, etc.). The model furnace used is pictured in Figure 7 and is
shown schematically in Figure 92. Tests were made to evaluate this
furnace for burning polyurethane foam. Dry, water wet, oil wet, and •
oil-water wet foams were burned. One pound lots of our standard foam
were used to. remove No. 2 Diesel fuel, Bunker C fuel and a one to one
mixture of Mo. 2 Diesel and Bunker C fuels from water surfaces. Each
lot of saturated foam was then passed through our model wringer three
times to remove the excess oil and water. Immediately after wringing
the foam was burned in the model furnace. A summary of the burning
rates is shown in Table 45.
155
-------
TABLE 43
THERMAL DEGRADATION OF ON-SITE GENERATED POLYURETHANE FOAM
Sample
Temperature
Weight Loss,
Elapsed Time
uc
25
75
100
150
200
250
300
350
400
450
500
550
600
650
700
800
850
"F
77
167
212
302
392
482
572
662
752
842
932
1,022
1,122
1,202
1,292
1,472
1,562
$W
0
0
1
1
2
4
20
45
53
55
57
60
64
66
69
75
75
Minutes
0
5
8
14
19
25
31
36
42
47
53
58
64
70
75
86
92
1) Density of Foam, lb/ft3 = 1.74
Pores/Inch =60
Fusion Point, PF = 340
2) Temperature increased at 9 C per minute
156
-------
TABLE 44
THERMAL HISTORY AND OUTGAS PRODUCT ANALYSES OF
ON-SITE GENERATED POLYURETHANE FOAM*
Temperature,
F
Weight Loss
Elapsed Time,
minutes
Comment
75 167 0
150 302 1
200 392 2
200 392 5
250 482 7
300 572 25
375 707 40
375 707 47
400 752 48
1) Density of Foam, lb/ft3
Pores/Inch = 60
Fusion Point, °F = 340
2) Temperature increased at
0
12
21
51 Mass Spec, of Gas
59
68
80
110 Mass Spec, of Gas
114
= 1.74
o
6 C per minute
MASS SPECTROME TRY ANALYSES OF CASED SAMPLES
AND NORMAL
Outgas
Composition
N2
02
H20
Ar
CO 2
NH3
LABORATORY AIR
Concentration, Mole H>
200°C 375°C Normal Air (Dry)
68.10 74.61 78.08
25.26 23.17 20.95
5.14 1.13
0.95 1.00 0.93
0.52 0.07 0.03
0.03 0.02
157
-------
Ul
00
TABLE 45
POLYURETHANE FOAM BURNING RATE FROM FURNACE MODEL
Description of
Foam
Dry
Water Wet
No. 2 Diesel
Wet
No. 2 Diesel
Water Wet
Bunker C
Water Wet
Burning Rate, Pounds
of Dry Foam per Hour
per Square Foot of
Grate Area
36
14
9
15
8
Flame
Temperature,
°F
1400-1500
1300-1400
1400-1500
1300-1400
1200-1400
Stack
Temperature,
F
1400-1500
1300-1400
1400-1500
1300-1400
1200-1500
Auxiliary
Fuel
Consumption
SCF/Hr
42
42
42
42
68
Unburned
Oil Lost as
$w
Nil
Nil
Nil
Nil
6
Foam or
Dripping
Bunker C, No. 2
Diesel Mix -
Water Wet
14
1400-1500
1400-1600
68
1.5
-------
Ui
v£>
100
80
60
01
O
j_>
00
40
20
100
200
300 400
o
Temperature, C
500
600
700
800
FIGURE 90 - WEIGHT LOSS VS SAMPLE TEMPERATURE - TEMPERATURE OF
ON-SITE GENERATED POLYURETHANE FOAM INCREASED AT
9°C PER MINUTE
-------
100 f-
80
60
en
09
o
.J
oo
£ 40
20
0
Outgassed
For 30 minutes
Outgassed
For 30 minutes
I
I
100
200 300
Temperature, °c
400
500
FIGURE 91 - WEIGHT LOSS VS SAMPLE TEMPERATURE - TEMPERATURE
OF SHELL PIPE LINE POLYURETHANE FOAM INCREASED
AT 6 C PER MINUTE
160
-------
4" Sch. 10 Pipe
Butterfly Valve
6" X 4" Weldon Reducer
6'
Afterburner
Gaseous Fuel
6" Sch. 10 Pipe
Hopper
3.
Ignition Burner
Fuel
(Methane, Propane,
Or Butane-Propane)
6" Sch. 10 Pipe
Compressed Air
Compressed Air
y- 4" Sch. 10 Pipe
Butterfly
Valve
Expanded Metal Grates
Compressed Air
Compressed Air
3 - 1" Pipe|B
120° Apart
ompressed
Air
FIGURE 92 - POLYURETHANE FOAM BURNING FURNACE - SCHEMATIC
161
-------
Bunker C-water saturated foam burned at about half the rate established
for foam saturated with No. 2 Diesel fuel. More auxiliary fuel was also
required, primarily in the afterburner, to prevent the emission of black
smoke from the stack. Additionally, the Bunker C tended to drip past
the three grates and out the bottom of the furnace. There is a possi-
bility that incomplete combustion of foams soaked with heavy oils will
also occur in a large furnace. A means to remove or recover the
drippings may be necessary.
Water, added during burning or absorbed while sorbing oil from water,
greatly reduced the particulate emissions. When either dry foam or
oil-soaked foam is to be burned, water should be sprayed onto the
foam prior to burning. The proper amount can be established by trial
and error.
Injection of compressed air at various points, as shown in
Figure 93, did not significantly improve either the burning rate or
particulate emission rate. The use of forced draft air appears to be
unnecessary. Flame and stack temperatures were low (Table 45) due in
part to a large excess of air (Table 6). The water in the foam may
also have been a factor. Temperatures in a large furnace may be
higher, due in part to the smaller ratio of furnace surface area to
burning grate area.
These burning tests indicate it is both practical and possible to
construct a furnace at the site of an oil spill and to burn the used
polyurethane foam without producing black smoke. The furnace may be
constructed utilizing welders and normal work crews and readily avail-
able materials such as pipe, sucker rods, propane or butane burners,
and air compressors. No scale-up problems are anticipated. A schematic
of a large furnace is shown in Figure 93.
Disposal bv Compaction and Subsequent Burial
Results from the fusion study show the polyurethane foam can be fused
to form a compact solid at a temperature of about 330 F to 350 F.
This indicates the volume of the foam could be reduced by heating it
to the fusion point and compressing. The volume of dry foam can be
reduced by a factor of about 30. The volume of oil-wetted foam can
be reduced by a factor of about 10 (assuming the oil remains with
the foam). Disposal by burial of compacted foam would require less
volume of earth to be moved. Two major problems are envisioned if
reducing the volume of used foam is necessary: 1) disposal of water
and oil vapors, and 2) heat transfer to the core of foam particles.
Disposal by Solution
The on-site generated polyurethane foam is essentially insoluble or only
slightly soluble in readily available solvents, e.g., aromatics, ketones,
acetetes, glycol-ethers, chlorinated hydrocarbons, etc. For this reason,
disposal by solution does not appear to be practical. Solubility tests
162
-------
FIGURE 93 - SCHEMATIC OF SHELL PIPE LINE DESIGN POLYURETHANE
FOAM BURNING FURNACE
1. Butterfly valve - secondary burning rate control (24-Inch) 20. Pipe furnace - 36 Inch
2. Butterfly valve - control handle 21. Pipe stack - 24 Inch
3. Butterfly valve - control rod
4. Afterburner fuel and air lines
5. Butterfly valve - air intake control valve (24-inch)
6- Butterfly valve - control handle
7. Butterfly valve - control rod latch
8. Pipe leg (12-inch)
9. Split slide valves - air intake control valves (36-inch)
10. Ignition burner mount for variable height burner
11. Ignition burner fuel and air lines
12. Ignition burner
13. Baffle plates to prevent foam dripping from affecting burner and to allow complete evapora-
tion and combustion of foam drippings
14. Adjustable expanded metal grates
15. Shield to prevent foam from dropping out of feed inlet
16. Conveyer to feed used foam to furnace
17. Air Intake annulus. Air keeps conveyer cooled and assists In carrying foam into furnace
18. Short radius 90° ell, 24-inch pipe fitting
19. Afterburner nozzle
163
-------
were made by placing 0.25 gram cubes of the on-site generated foam in
50 ml of each of the following solvents :
1. Benzene
2. Methylethylketone
3. Dichloromethane
4. Butyl cellusolve
5. Normal butyl acetate
None of these solvents dissolved a significant quantity of the foam
in 240 hours.
164
-------
SECTION XIV
SYSTEM DESIGN
Introduction - Offshore Systems
The initial concept of the offshore system, as shown in Figure 1, has
been modified somewhat as a result of this study. While the original
version is acceptable, the preferred offshore configuration is now
represented by Figure 8. A smaller "half size" system, using the same
equipment modules, is shown in Figure 9. The principal differences
result from acceptance of a shorter residence time or slower system
velocities, resulting in a more compact system requiring fewer vessels.
In the preceding sections (Section VI through Section XIII), all of the
basic processes required for the oil spill cleanup system are described
and characterized. Drawing upon these sections, the performance require-
ments for the system may be developed. There are essentially four steps
in this procedure:
1. Establish the parameters of the design oil spill and the
required rate of oil recovery.
2. Estimate material flow in the system to meet these requirements:
(a) Foam required
(b) Oil and water recovered
(c) Foam losses
(d) Foam for recycling
(e) Make up foam required
3. Estimate component performance requirements and approximate
size.
4. Adjust components to satisfy overall system constraints
as to mobility, vessel size, etc.
In this discussion, the above procedure has been used to develop the
performance requirements for a system of the type shown in Figure 8. The
sorption investigation has shown that, of the oils studied, the crude
identified as Carnea 21 displayed the minimum specific sorption and
effluent oil contents (see Figures 31 and 32). This oil, then, has been
selected for preliminary design of a recovery system. In making this
selection, and in applying the results of our laboratory sorption and.
separation studies, it is felt that a practical yet conservative design
will be evolved. A system flow chart is shown in Figure 94. The
following assumptions and estimates are based upon results in the earlier
sections.
165
-------
"Oil Herder"
57,000 Ib/ht
130,000 Ib/hr
r 5'.0°0 lb/hr < w.t«r/0ll<
1 7,000 -
T.nii [ r--«-l :: = —
82.000 lb
.UUU 1O l
,000 gal'
15B.UUU iD/nr i .,i«v. \25,QOO Ib/hr, w
P , 91.000 lb uataJt—^—I V J f
I 67,000 lb oil ' I — J -
11,000 gal water^ \
9,000 gal oil
130,000 Ib ^../..t.^155'000 lb
16,000 g.,<»11'M"f< 19,000 g.l
3000 Ib/hr
-E" Tranafer
Veaaal
!""• ^25,000 Ib/hr,
AD. I n nn« < J W«tBK
/Ttaat
Plant I «atar
c--«
Actual Wt. 100,000 lb
Eiiulv. Dry Ut (20,000 lb)
Equipment
Storaga
9 Tranafar
FIGURE 94 - FLOW CHART PROTOTYPE - EXAMPLE
Oil Recovery System Using Sorbent Materials
Polyurethane Foam - Generated On-Site
-------
System Requirements - Assumptions
— Oil: Carnea 21, specific gravity = 0.889,
viscosity =• 31 cs at 70°F,
— Spill: 0.06 in. (1.5 mm) thick
— Recovery specified; 9000 gal./hr oil (this equates to
240,000 ft2/hr of spill that must be traversed).
— Sorbent: Foamed on-site polyurethane foam, average
density 2.1 lb/ft3.
— Characteristics of foam application:
Residence time: 60 sec (min)
Specific oil recovery: 0.35 - 0.41 gal./lb
Effluent: 45 to 50$ oil
Residual oil in foam after wringing: 0.47 - 0.48 gal./lb
Foam concentration on surface of spill: 0.1 Ib/ft2(max)
Material Flow - Estimates
— Foam Required:
9000 gal.
(oil)
Ib (foam)
hr v—' " 0.35 gal. (oil) '
(This is approximately 12,000 ft3/hr)
— Oil and Water Recovery:
26,000 Ib/hr
a. Foam at contact
with harvester
b. Foam at transfer
to deck conveyor
c. Water drained
to sea
d. Foam at approach
to wringer roll
e. Water/oil drain
to storage
f. Fluid removed by
Unit Quant.
34 lb/ft3
Total
10 sec
30 sec
490,000 Ib/hr
25 lb/ft3 360,000 Ib/hr
(9 lb/ft3) (130,000 Ib/hr)
21 lb/ft3 300,000 Ib/hr
(4 lb/ft3) ( 57,000 Ib/hr)
wringing
Oil (45#) at 0.35 gal./lb « (4.6 lb/ft3)(67,000 Ib/hr)
Water . (6.3 lb/ft3)(91,000 Ib/hr)
Total fluid removed = (10.9 lb/ft3)(160,000 Ib/hr)
167
-------
g. Foam at exit
Total weight - (21 - 10.9) lb/ft3 = 10.1 lb/ft3
Foam weight • = 1.74 lb/ft3
Residual liquid - (8.3 lb/ft3)
(Note that this calculated residual of 8.3 lb/ft3 is in
rough agreement with the 6.3 lb/ft3 determined in some
experiments.)
— Anticipated Foam Losses (distribution, collection, harvesting,
wringing, etc.) ^ 10$
Dry wt. equiv. = 25,000 Ib/hr; actual » 15,000 Ib/hr ±
— Foam available for recycle
From above, 90$
dry equiv. = 25,000 Ib/hr; actual = 140,000 Ib/hr
— Make up foam requirement
From above, 10$ ±, or 2,500 Ib/in.
System Components
Single Barge/Single Boom Configuration
Having values given for slick thickness, area covered per hour, and
residence time; combinations of system velocity, boom length, and boom
deployment angle may be investigated. Such an exercise is represented in
graphical form in Figure 95.
Entering the top portion of Figure 95 with thickness and a system
velocity
d = 0.06 in., V = 2.0 ft/sec
s s
it is seen that the opening in the boom array (the swept width W) must be
at least 27 feet. Continuing to the lower portion of Figure 101, to the
velocity of 2.0, it is seen that a boom 125 feet in length, deployed at
about 15 degrees, will be satisfactory.
However, for this example involving a larger barge, a higher velocity
might improve handling characteristics; for instance, a velocity of
3.0 ft/sec with a sweep of 23 feet and a 180-foot boom would serve.
Concentration of Foam by Booms
In sizing equipment, it is convenient to consider the material flow per
unit width - Q', where:
Q1 - D x V x 3600
S
Q1 = Ib/hr-ft
D = Foam concentration lb/ft2 (or, area density)
V - System velocity .- ft/sec
168
-------
I
g
8
m
X
CO
Swept Width - w (ft)
25 50 75 100 125
150
600
400
200
4.0
3.0
2.0
1.2
FIGURE 95 - MINIMUM BOOM REQUIRED FOR SINGLE BOOM SYSTEM WITH 60 SECOND
FOAM RESIDENCE
169
-------
This expression has been graphed in Figure 96.
For the example, the foam entering the boom array is calculated:
Q' i- 0.1 lb/ft2 x 3.0 ft/sec x 3600 sec/hr
» 1100 Ib/hr-ft
(This quantity can also be found from Figure 96)
As the foam is further concentrated in approaching the harvester, Q1 is
modified by the width ratio Ww, where w = width of harvester belt.
Harvester Width and Operation
Experiments have shown that the foam, when driven onto a conveyor by
water motion, will pack to a density (apparent) of from 4.0 to 6.3 lb/ft3
(dry foam equivalent). For sizing a harvesting conveyor, a conservative
value of 4.1 lb/ft3 has been used as a "design1* value (see explanation of
the relationship between Q' and belt speed V for various packing thick-
nesses in Section X).
Since conservative mechanical design limits flat belt conveyor speeds to four
or five ft/ sec, and since normal practice would be to operate the belt at:
VB - Vg/cos 6
where 6 = angle of inclination,
where 9 » 40°,
V_ - 3.9 ± ft/sec
D
at
V0 = 3.0 ft/sec
9
The figures may be used to estimate the foam thickness for various
harvester widths, resulting in a width selection (tentative design of
harvester has been based on a maximum foam thickness of about 0.2 ft,
although higher values are practical). The equation developed in Section X
QH " (ds )(Vfi) (3600) (4.1)
avg N
where d = avg thickness - ft may be used to find a practical value of
avg
harvester width w by trial and error:
W
23
23
23
w
15
10
5
W/w
1.5
2.3
4.6
Q-,
W/w*
1700
2600
5100
ds
0.05
0.08
0.05
0.08
0.10
0.30
0.40
VB
2.2
1.5
3.4
2.3
1.7
4.8
3.6
* Ql. *= (W/w) (1100)
W/w
170
-------
5.0
System
VS
ft/sec
Use of
Booms
Q = D x Vg x 3600
D = Area Density = lb/ft2
Vc = Relative Vol. = ft/sec
t>
1000 2000 3000
Foam Available Per Foot Width - Q - Ib/hr Note. Where dlverging
booms used to
concentrate foam.
Multiply Q by
FIGURE 96 - FOAM AVAILABLE TO HARVESTER PER UNIT WIDTH ratio W/w'
-------
A five-foot width would exceed the d and V limitations; a 10-foot width
is satisfactory and capable of handling surges in flow.
Wringer Feed Conveyor
The foam being transferred from the harvester to the wringer has drained
some 10-20 seconds while on the harvester. Experiments have shown that
this material, when dropped onto a belt, will pack to an apparent density
of about 2.0 lb/ft3 (dry foam equivalent). The basic equation is
modified:
0' - (d )(V_)(3600)(2.0)
s D
avg
The wringing experiments have shown that foam thicknesses of 0.5 ft and
belt speeds of 100 ft/min (1.7 ft/sec) are acceptable for a simple cylin-
drical roller. At these values, an indicated minimum belt width would be
found as
Q£ = (0.5)(1.7)(3600)(2.0)
= 6000 Ib/hr-ft
60 •
To allow for surges caused by irregular flow, a greater width would be
advisable.
Transfer After Wringing
If a belt conveyor is used for part of the foam recycling operation, its
size will depend again upon an experimentally determined apparent packing
density, found to be about 1.6 lb/ft3 (dry foam equivalent). The flow
equation is again modified:
Q1 = (ds )(VB) (3600) (1.6)
avg
If a width is arbitrarily selected to match that of the other modules,
say 5 feet, a belt speed is assumed and the average thickness found:
n, _ 25.000 _1_ _ 5,000 Ib
y ~ hr 5ft " hr-ft
a reasonable velocity of 3.0 ft/sec gives:
5.000 Ib/hr-ft
* °
s 3.0 ft/sec x 3600 sec/hr x 1.6 lb/ft3
avg
Storage of Liquids
It is proposed that liquids be stored temporarily on deck, either in bolted
steel storage tanks or in fabric storage cells. No attempt is made to
172
-------
provide separation of water on board except that a two-stage liquid
system could by employed where conveyor drain and gravity settling in
tanks might permit segregation of liquid containing little oil. Multiple
tanks in battery, with suitable manifolding/ would give this flexibility.
On a large, stable barge or in protected waters, the conventional oil-field
bolted steel tanks could be utilized in capacities from 250 bbl (10,500 gal.,
8 ft high x 16 ft diameter) to 1000 bbl (42,000 gaL, 8 ft x 30 ft diameter,
or 16 ft x 21 ft diameter). These tanks are available palletized for
storage and are usually fabricated in accordance with API Standard 12B.
Estimated weight on pallets: 0.3 - 0.5 pounds /gal. cap.
Estimated cost on pallets: 12-15 cents/gal, cap.
Assembly time: 12 hours or more.
On any vessel, and particularly where assembly time is critical, the
fabric containers would be recommended. In most cases, careful attention
to tie-down provisions will be necessary but this should not be a problem
on a steel vessel. Less efficient use of deck area is achieved, but ease
of installation and convenience in storing tend to offset this. Considera-
tion should be given to providing a mix of 20,000 gallon and, perhaps,
10,000 gallon sizes to facilitate installation. Typical sizes and estimated
costs for the usual fabric oil storage container are:
20,000 gallon: 28 ft x 28 ft x 4 ft high
Shipping crate: 410 Ib
5 ft x 4 ft x 2 ft
Unit weight: 0.02 Ib/gal. cap.
Est. unit cost: 25-30 cents/gal. cap. (including special tie down
straps)
10,000 gallon: 20 ft x 20 ft 4 ft high
Shipping crate: 260 Ibs
5 ft x 4 ft x 1 ft
Unit weight: 0.026 Ib/gal. cap.
Est. unit cost: not estimated
Adjust Component Designs to Suit System Requirements
To arrive at a preliminary system design that will satisfy the performance
requirements, be modular in configuration, and be adaptable to convenient
transport and assembly, the components described in the foregoing may be
modified. Such adjustments at this stage are somewhat arbitrary and a
matter of judgment.
To provide for the modular concept, provide a certain amount of redundancy
as well as surge capacity, components may be somewhat oversized. When
the basic performance estimates are also somewhat conservative, it is
probable that the system described in this example is, indeed, oversized.
173
-------
To review the example calculations of the preceding pages, both the material
flow chart of Figure 94 and an estimate of the material inventory will
be useful. Once the system has reached its full operating condition, the
total throughput of foam (dry weight equivalent) is 25,000 Ib/hr, of which
about 2,500 Ib/hr will be lost from the stream of useful sorbent, (see
Section XII), requiring a like capacity for manufacture and introduction
of new foam. This make-up rate is well within the capacity of the commer-
cial scale foam generators and the single large size mulcher (Section VIII).
During startup of the system, it would not be necessary to manufacture
and mulch at the overall throughput rate, since it is only necessary to
introduce an amount equal to the inventory required for a single cycle.
The foam inventory may be estimated from the assumed or calculated
throughputs in the example. Allowing for variations and surges, we esti-
mate the length of time required for one complete cycle to be:
Residence on the surface 60 - 120 seconds
Time on harvester 10 - 15 seconds
Wringer feed conveyor 10 - 40 seconds
Wring and separate 5-10 seconds
Recycle by belt conveyor 65 - 80 seconds
Total cycle time 150 to 265 seconds
Then, without consideration of storage in the cycle, the foam inventory
is found:
Foam inventory = 25,000 Ib/hr x hr/3600 sec x 150 sec = 1000 Ib (min)
and the larger estimate = 1800 Ib (max)
To allow for short interruptions of, say, no more than two minutes in the
recycling process, the foam inventory might be selected as 1800 Ib. A
summary review of the system in this example results in the following:
-- Foam throughput: 25,000 Ib/hr, 420 Ib/min, 14,000 ft3/hr
Foam inventory in system: 1800 Ib
— Foam manufacture: One commercial unit: 2500 Ib/hr
— Foam preparation: One Reinco Model M60-F6: 5-6,000 Ib/hr
— Foam initial distribution: above mulcher: 6000 Ib/hr
— Boom: single boom, 180 feet ± with 23-foot opening
-- Harvester: twin 5-foot wide flat belt modules, each 33 feet ± long
— Wringer feed and wringer: twin metal belt 5-foot modules with
dual cylindrical rolls
— Recycle belt conveyor: single 5-foot fabric belt conveyor, in
sections for total length of approximately 200 feet (NOTE: The
pneumatic systems described in Section VIII would be preferred
and are indicated in the figures).
174
-------
-- Foam storage for shutdownr (dry wt. equiv.)
a. Recycle belt
420 Ib/min x min/180 ft x 200 ft «= 460 Ib
b. Other*
1800 Ib - 460 Ib « 1380 Ib
(volume » 1380 Ib x ft3/!.6 Ib = 840 ft3)
* All or part of this may be included in the reject foam
storage area.
-- Foam reject storage (assume 4-hour capacity)
500 Ib/hr x ft3/!.6 Ib x 4 hr - 1220 ft3
(dry wt. equiv.)
-- Liquid storage
Stage I water tank battery » 80,000 gal.
9900 gal./hr
Stage II oil/water
16,000 gal./hr
Protected Waters System
The performance requirements for this system include operation in 6-knot
currents, two-foot waves and 20 mph winds; an oil recovery rate (minimum)
of 1350 gal/hr net oil, with a desired rate of 2700 gal./hr.
All of the sorbent handling equipment modules of the offshore system may
be applied directly to this situation. All components, as sized in the
preceding section, have capacities in excess of these requirements.
Longer booms are needed, since a foam-on-oil residence time of 60 seconds
would require slightly more than 600 feet. It is likely that somewhat
shorter residence times would be acceptable in this application, since
excess foam and liquid handling capacities are available. Very few
experimental data were obtained for times less than 60 seconds during this
investigation, however.
The significant problem in operating any system in high currents is the
deployment of the boom. It has been demonstrated that almost any boom
will "fail" when the water velocity component normal to the boom exceeds
1.2 to 1.5 ft/sec. Higher current velocities will require deployment as
a diversionary boom, maintaining a shape that does not permit the normal
velocity component to approach failure. Deployment as a catenary (or
parabola) might be assumed. Control of the boom angle at the downstream
end would be by the tension in the boom. Boom design was not investigated,
but recent in-house studies have shown that a diversionary boom having a
3-foot draft, when deployed in a 3.5rknot current so as to sweep only a
50-foot width, would require about 12,000 Ib of tension to maintain the
proper downstream angle. This would indicate that an oil spill-from a
concentrated source could be diverted toward a recovery system as shown
in Figure 97. Note that the recovery, using sorbent in a contained area,
175
-------
,DD
FIGURE 97 - USE OF COMPONENT MODULES IN HIGH CURRENT (RIVER)
176
-------
requires only a very narrow channel when the current velocity is high
(the specified 2700 gal/hr spill, when diverted so as to enter the
recovery channel at a thickness of 0.06 in. would be only two feet in
width). '
The operation of the system in the manner shown in the Figure 97 is little
different from the offshore situation. The component booms and vessels
might be moored (note that the boom tension alone, in even the 3.5-knot
current mentioned above, equates to approximately 500 horsepower), the
harvester could be operated at a much smaller angle of inclination, and
the recycling conveyor would be used to deliver foam to small barges (or
trucks) for return to the distribution point.
Equipment Notes
General
Only limited consideration has been given in this investigation to specific
or detailed equipment design. The concepts and the experimental units
presented appear to be practical and capable of execution by experienced
designers and fabricators of similar machinery; no new technology is
involved, although some further experimental work is indicated in connection
with the detailed design, such as eliminating or reducing the risk of
explosion in the pneumatic conveying system by grounding and by safety
panels, etc.
Power Supply
Wherever possible, the use of hydraulic drives, powered from packaged
engine-driven power units, is recommended. This approach provides flexi-
bility, redundancy, preserves the modular concept, and reduces or elimi-
nates dependence upon shipboard supplies.
Component Modules
All system components have been described in a modular concept. Each
module is suited to movement over the highway, although the pressurized
penumatic conveyor must be palletized (its normal condition for storage
and transport). Module and pallet size have been discussed briefly in
this report. It appears that only certain modules will require disassembly
for shipment by air (other than the palletizing already mentioned). These
might be the longer belt conveyors and the pressurized pneumatic system
storage hoppers. Disassembly would not be necessary for a C-130 type
aircraft but would be required for side door loading in the more commonly-
available Boeing 727 QC type aircraft.
Discussions with aircraft companies, cargo carriers, and reference to
the AIR CARGO GUIDE (published by Reuben H. Donnelley) have established
the following suggested limitations:
177
-------
C-130 Hercules
Cargo compartment: 10 ft x 10 ft x 40*ft
Approximate payload cargo: 20,000 Ibs
»
^Loading: Rear ramp, full compartment length is practical.
Boeing 727 - QC
Cargo compartment: 10 ft x 6 ft (curved overhead) x 19*ft
Approximate payload-cargo: 35,000 Ibs
•^Loading: side cargo door entry limits package length, depending
on package height and width; the maximum length is
230 inches, if height is less than six feet with a width
less than two feet.
Douglas DC-8
(Stretched version in cargo service)
Cargo compartment: 10 ft x 6-1/2 ft x 126*ft
Approximate payload: 80,000 Ibs
*Loading: side door with restrictions on length similar to 727,
but depends on particular airframe configuration.
All carriers contacted will accept palletized packages up to 10 ft x
7 ft x 4 ft ± high. Heights to six feet are possible, depending upon
shape of package.
178
-------
SECTION XV
ACKNOWLEDGMENTS
Key personnel who were responsible for this study:
R. A. Cochran - Wringer
D. P. Hemphill - Distribution; Harvesting; System Design
J. P. Oxenham - Sorption; Sorbent Collection
P. R. Scott - Foam Development; Sorbent Disposal
J. P. Fraser - Project Coordination
The Pipeline Research and Development Laboratory is directed by
E. A. Milz
This project was supported by the Office of Research Monitoring of
the Environmental Protection Agency. Mr. J. S. Dorrler was
Project Officer.
179
-------
SECTION XVI
REFERENCES
1. Milz, E. A. and J. F. Eraser, "A Surface-Active Chemical System
for Controlling and Recovering Spilled Oil from the Ocean",
Journal of Petroleum Technology, 24, March 1972.
2. Schatzberg, P. and K. V. Nagy, "Sorbents for Oil Spill Removal",
Proceedings, Joint Conference on Prevention and Control of Oil
Spills, June 15-17, 1971, American Petroleum Institute, pp. 221-234.
3. Cochran, R. A., VT. T. Jones, and J. P. Oxenham, "A Feasibility
Study of the Use of the Oleophilic Belt Oil Scrubber", Final
Report to the U. S. Coast Guard, AD 723598, October 1970.
181
-------
SECTION XVII
NOMENCLATURE
d = apparent foam thickness
d =• depth of slick
s
D = foam concentration, lb/ft2
G ~ dimensional constant
G'= dimensional constant
K = permeability of the sorbent
AM = total mass recovered
N - number of passes through the wringer
P = weight of roller
q = oil flux
Q = volume of oil
O = volume of oil permanently retained in the foam
Q = initial oil volume
Q = total oil absorption rate (volume flow rate)
A _ Q
V -nr J
e
Q'= equivalent dry weight of foam recovered per hour per unit width of system
Subscripts:
C = in current
H = on harvester
F = on wringer feed
R = recycle
r « local radius
r = circular equivalent radius
e
r = local radius to edge of sorbent block
SPG = specific gravity of oil
t = time
Av = total volume recovered
V., = belt speed
B
V_ = system speed
&
183
-------
w = width of belt
W = swept width
M> = oil viscosity
ACT = interfacial driving force
I! = effective specific surface
0 = porosity of the sorbent material
184
-------
SECTION XVIII
APPENDIX 1
POLYURETHANE FOAM REACTIONS
There are numerous methods for the preparation of polyurethanes. The
most widely used is the reaction of di or polyfunctional hydroxyl com-
pounds with di or polyfunctional isocyanates. Linear polyurethanes are
produced when difunctional polyethers or polyesters react with diisocy-
anates as shown below.
(polyether) (diisocyanate)
HO-R-OH + 0=C=N-R'-NM]=O-»
{0 0 -|
O-R-O-c'-NH-R'-NH-C 4-
(DOIvurethane^ -I n
(polyurethane)
The linkage
-NH-C-0-(urethane)
characterizes polyurethanes although other groups, such as ether, ester,
biuret, allophanate, amide, and other groups may be present in the poly-
mer molecule.
Crosslinked polyurethanes are formed if the functionality of the hydroxyl
or isocyanate component is increased to three or more.
The properties of the various types of urethane polymers depend largely
upon molecular weight and degree of crosslinking. Urethanes are ver-
satile polymers. They include fibers, elastomers, adhesives, thermo-
plastics, thermosetting plastics, rigid foams, and flexible foams. The,
latter are of interest for sorbing oil spills. Open cell, low density
polyurethane foams have been evaluated as oil sorbents by other
researchers and have been used in the field for removing spilled oil.
The production of polyurethane foam at the site of an oil spill is
desirable for several reasons. The development of the proper formula-
tion to produce good quality foam in the varied ambient conditions
existing at oil spills has been time consuming.
185
-------
The open cell flexible polyurethane currently being used is made using
the following components:
(1) Trifunctional polymeric polyol
(2) Polymeric methylene diphenyldiisocyanate
(3) Methylene chloride and
(4) Trimethylaminoethylpiperazine
The two most important reactions are those between the diisocyanate and
trifunctional polyol (the chain propagating reaction) and between the
diisocyanate and water (the foaming reaction). The characteristic
urethane linkage is formed by the first reaction
OH
HO~^~OH + 0=C=N-~3-»
(6500MW triol) (Polymeric diisocyanate)
H
HN-C-0 0-C-N~
0
6=0
*H
I
The reaction between the diisocyanate and water results in the liberation
of carbon dioxide with simultaneous formation of an amine.
Diisocyanate Water Amine Carbon Dioxide
H20-» ~R* -NH2 + C02 t
The amine immediately reacts with additional isocyanate to form a
substituted urea.
0
~R' -NH2 + ~R-N=C=0-» ~R-NH-fc-NH-R~
Substituted Urea
Other reactions also occur which may cause crosslinking, chain propagation,
etc. but for this purpose are considered minor.
The methylene chloride is added as an auxiliary blowing agent. The
density of a flexible foam can be progressively lowered by increasing
the diisocyanate and water levels, but this changes the polymer structure
of the foam. An auxiliary blowing agent is added to decrease density
without increasing crosslinking. The above reactions are exothermic.
Sufficient heat is liberated to vaporize the methylene chloride. In-
creasing the auxiliary blowing agent decreases both the density and load
bearing ability of the foam.
186
-------
The above primary reactions are too slow for the production of urethane
foams for practical purposes. The catalyst trimethylaminoethylpiperazine
is employed to bring about faster rates of reactions. This catalyst
not only brings about faster rates of reaction but also establishes a
proper balance between the chain-propagating reaction (primarily the
hydroxyl-diisocyanate reaction) and foaming reaction (diisocyanate-water
reaction). A balance has to be established between the polymer growth
and gas and vapor formation (1) prevent the development of sufficient
strength in the cell walls to entrap large quantities of gas (2) develop
sufficient polymer strength to prevent the collapse of the structure
when the gas escapes the ruptured cells. When the chain extension and
crosslinking reactions are predominant, the number of closed cell faces
becomes greater and the porosity decreases. When the foaming reactions
predominate, the foam collapses upon release of the gas and vapors.
Tertiary amines catalyze both the isocyanate-hydroxyl and isocyanate-
water reactions. However, all tertiary amines are not good catalyst.
The efficiency of the tertiary amine generally increases as the basicity
increases and as the steric shielding of the amino nitrogen decreases.
When the above components are intimately blended, a number of reactions
take place very rapidly. A polymer is formed and expands to a density
of one to three lb/ft3 in about 45 seconds at 77°F. The timing of the
polymerization and expansion are critical and are controlled by the
catalyst and the relative concentrations of the diisocyanate and water.
The relative quantitites of components are usually critical; however,
usable foam is produced when the quantity of any component in this recipe
is varied about ±20$.
187
-------
APPENDIX 2
POLYURETHANE FOAM FOR ABSORBING SPILLED OIL
The laboratory has tested numerous formulations for making flexible,
open cell polyurethane foam.
The following polyurethane foam formulation produces 1.5 to 3 pounds
per cubic foot density foam when made in beaker quantities in the lab-
oratory, poured utilizing the laboratory portable foam equipment and
when poured utilizing a commercial foam machine.
Component A Parts
Jefferson Chemical Company 100
Thanol SF 6500
Dichloromethane (Dow) 10
Water 5
Jefferson Chemical Company 2
Thaneat TAP
Component B*
Rubinate-M 50-80
(Jefferson Thanate P-30)
* Quantity of Component B depends upon ambient
temperature, humidity, etc.
The material cost is about $0.37 per pound when drum lots of chemicals
are purchased.
The ratio of Component A to Component B is not very critical; however,
the ratio needs to be varied to form the best foam at the existing
ambient condition. Good quality foams have been made at ambient temper-
atures from 40 to 120°F and relative humidities of from about 20 to 95$.
The following tests were designed to simulate conditions that may exist
at an oil spill site.
1. Foam was made by pouring the mixed foam components directly
on both wet and dry sand. The foam quality was good.
2. The mixed components were poured on newsprint paper. The foam
quality was good.
3. The mixed components were poured on Teflon repeatedly to
simulate a moving belt operation (see Figure 17). The foam
189
-------
5.
quality was excellent and the foam removed one minute after
pouring had 80 to 90$ open cells at the foam-Teflon interface.
Before the rise was complete another Teflon sheet was placed
on top to simulate foam formed between two moving Teflon belts.
Open cell structure was found at both interfaces. This operation
simulates a more elaborate but practical technique. The moving
belt concept would allow one to produce long thin strips of
polyurethane with open cells on both sides. The long strips,
4-feet to 1000-feet or more could be applied to large or thick
spills. The harvesting would be simplified.
The foam components were mixed and then applied directly to
water surfaces. Reasonable quality foams were produced at
times. The quantity of isocyanate and the stage of reaction
before contact with the water are both critical. With practice,
reasonable quality foam with open cells on the water interface
side can be made.
The foam components after mixing were applied directly to oil
floating on water surfaces. The above statements concerning
success and application to water apply.
Additional batches of polyurethane foam components were made and evaluated
during this study. Batches were made using Freon-11 as an auxiliary
blowing agent, and batches were made to evaluate catalysts listed below:
1. Thancat® DD, Jefferson Chemical Company, Incorporated
2. Dabco®, Houdry Process and Chemical Company
3. T-9, M & T Chemicals, Incorporated
4. T-12, M & T Chemicals, Incorporated
® ®
Thancat DD and Dabco are tertiary amines, T-9 is stannous octoate, and
T-12 is dibutyltin dilaurate.
Results of our tests with different catalysts are summarized below:
®
1. Thancat DD and Thancat TAPwere found to be interchangeable.
About 25$w less DD is required in otherwise comparable blends.
2. Tin catalysts were not satisfactory because (a) the tin catalysts
lost activity when blended with water, (b) the quality of foam
produced was very sensitive to the quantity of tin catalysts
used. Tin catalysts are not recommended since it appears they
would have to be injected as a carefully-metered separate
component.
®
3. Dabco is not recommended for use with Thanol SF-6500 MW Polyol
and Polymeric MDl, The reactions catalyzed by this catalyst
were not balanced. The reactions were too slow when a small
190
-------
quantity of catalyst was used. When sufficient catalyst was
used to produce a tack-free foam in five minutes, the foaming
reaction (C02 liberation) exceeded the crosslinking reaction
to such an extent the resulting foam slumped.
No foam formulation tested was better than the original available at
the beginning of testing. However, we gained considerable experience
in the production of polyurethane foam and knowledge concerning the
flexibility of operation which is available to us in the use of poly-
urethane foam for oil sorption.
The foam formulation described below produced a good foam for absorbing
oil. However, there are insufficient closed cells to give it good
buoyancy after water wetting.
Component Parts
Thanol SF-6500 100
Freon-11 10
Water 6
Thancat DD® 1.5
Rubinate M 60
A 100-pound quantity of foam was made from this formulation using our
foam machine. The Thanol, Freon-11, water, and Thancat were blended
to make one component (B component). This B component was then mixed
with the isocyanate (A component) with a Binks 18 FM gun. The ratio
of A to B was 1:2 by weight. The resulting 1.75 lb/ft3 foam had a rise
time of one minute and was tack-free in 1-1/2 minutes. This foam has
tear and compression characteristics which make it an excellent foam
for spreading with the hay spreader but at the cost of greater attrition
during recycling.
191
-------
APPENDIX 3
TOXICITY TESTS OF POLYURETHANE FOAM GENERATED ON SITE
One concern with the use of on-site generated polyurethane foam as an
oil sorbent is the ecological damage which might be caused by the foam,
due to the leaching of unreacted components from the foam into the water.
To test this possibility, the Edna Wood Laboratories, Houston, Texas
have made 96-hour acute toxicity tests using F. Similis, a small sea-
water fish. Results are shown in the attached report. Fish ate the
foam with no apparent ill effects. These bioassays are considered to
be a more sensitive measure of possible leaching effects than any
chemical analyses we could make. The foam used in these tests is
described in Table 9.
193
-------
F, Si, 5.
',A - 3.7«^r«ii«*- *• '
EDNA WOOD LABORATORIES
4»20 Old Spanish Trail Houston, Texas 77021
Bioassay No.
_ . ^_ Bioassay Work Sheet
«w»
Received 13 ^fa.m j^^l-
Sample Concent i
t7-Un l*i7Z.
17-iO
tfi-Jan t*r7*
I**
i1-i
-------
EDNA WOOD LABORATORIES
4820 Old Spanish Trail Houston, Texas 77021
Bioassay No.
Bioassay Work Sheet
Physio- Chemical Observations
Date & Time
1 -j J*LA 1971-
Test
Tank
Polarograph
amps
mfl/L
PH
7?
Sp. Res.
ohms
Temp.
°C
/*?
r.
to.
7-7
-7.1
-££.
1*0
195
-------
EDNA WOOD LABORATORIES
4820 Old Spanish Trail Houston, Texas 77021
Bioassay No.
Bioassay Work Sheet ' '
Physio-Chemical Observations
Date & Time
if J-io
*r
is
if
IS
•^
v
^
*•
tf
«•
K
graph
os
mq/l
?-i3^"
•^
•*"
IS
^
•^
•^
«X
•V
•^
•^
fc-"
pH
7.09*
7.I
7>^
7.*?-
7-3L
7.^
7-J-
7./JT
7-3
7»r
7.0?
Y.<*~
Sp. Res.
ohms
•ST*
^
v
**
^
*f
*r
^
*f
W
£L
Si
Temp.
°C
A»
ft*1
-
aV
P*
*•*-
-
-
^
-
~
•»•*
_
196
-------
EDNA WOOD LABORATORIES
4820 Old Spanish Trail Houston, Texas 77021
Bloassay No.
Bioassay Work Sheet
Physio-Chemical Observations
Date & Time
If JA*» 17*11-
l«*00
Test
Tank
£.
*»*_
7,
77,
*•
*..
1,
•TV
lo,
i*v
c
C-
Polarc
;i amps
>»o
•*•
^
V
•*•
-
V
^
^
*r
r
^
>graph
o,
ma/1
•^S-S"
•x"
•^
**•"
•»
V
^
i^1
fc-
•^
«*-
««*•
pH
t.tr''
fc-*?
7.^^
7.^>
<-1
7.«9
t.*?
*•?
7-«0
.4.^
4.1
£>*
Sp. Res.
ohms
5"f
•**
*•
•^*
eV
<**
**•
*~
+r
•V
i-r
5-S~
^?^
Temp.
°C
JLO
W
W
•^
•**
fc^
fc-*
f
•u*
•>-
*X»
**
197
-------
EDNA WOOD LABORATORIES
4820 Old Spanish Trail Houston, Texas 77021
Bi
Bioassay Work Sheet
Physio- Chemical Observations
Date & Time
&.C .lcn tST*-
1 ~2-X fc
Test
Tank
fc,
^
->t
7i_
f,
fri.
«,
«?t
(0,
•a*.
c
c
Polarc
fj amps
J>.r
-
-
-
-•
^
-
^
-
^
*..
^
•
>graph
Oa
mg/1
^,*.s-
-
-
-
~
-
-
-
_
-
-
*'
pH
7.3
7V
7-4
1^5-
T.fc
7. i"
7-^'
7.J-
7-1«T
**!-
7 .r
7.ru
Sp. Res.
ohms
5-W
s~H
s-s-
»-«
«-H
6-4-
«-4
ri
«-•*
«••/
*•«•
^
Temp.
.AO
-
^
-
-
~
-
• ~
--
~-
-
t.
oassav No.
P'lf
198
-------
EDNA WOOD LABORATORIES
4820 Old Spanish Trail Houston, Texas 77021
Bioassay No.
Bioassay Work Sheet
Physio-Chemical Observations
Date & Time
3*1 Jan i«t7S-
i^lfc
Test
Tank
6,
fcj-
•7.
~1^
*,
*^
«,
?x-
<0,
tot
<:
C-
Polarc
/j amps
>to
•*•
tff
**
^
9**
-
•X
-
-
•^
>*•
>graph
02
mq/1
>^i3-5"
-
•^
*^
>*
A-
**•
4^
•X
k^
•^-
-
pH
7 <
7,^r
7.3,\
7.3-
TH
7-f
T.'J^
7.A*-
7.3^
-7-1
7-3T
?.-/
Sp. Res.
ohms
*-1
i^
-
**
^
s
IS
^
•^
^*
5-6
sit,
Temp.
°C
lo
S
s
**
t*
•*
^
*r
-
fc^
•
IS
199
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
1, RepoxtNo.
3, Accession Jfo.
w
4. Title
AN OIL RECOVERY SYSTEM UTILIZING POLYURETHANE FOAM
— A FEASIBILITY STUDY.
7. Authot(s) R. A. Cochran, J. P. Fraser, D. P. Hemphill,
J. P. Oxenham, P. R. Scott
$, RepoitVate
6. . -••• • :•:.-.';'.
& Performing Organization
RepotiSo,
9. Organization
Shell Development Company
Pipeline Research and Development Laboratory
W. Project No.
15080 HES
11. Contract/Grant Jfo.
68-01-0067
Type si Report sad
Period Covered
2. Spvaaoiinr Organ? atfoa
IS. Supplementary Notes
U.S. Environmental Protection Agency Report No. EPA 670/2-73-084,
October 1973
16. Abstract
A system has been developed for recovering spilled oil from water surfaces under
a wide variety of environmental conditions and for all types of oils. The system
is designed to recover oil at rates up to 9,000 gal./hr.
This system is based on the use of polyurethane foam, foamed on the job site, as
a sorbent for the spilled oil. The foam is recirculated to increase efficiency
and to lower unit costs. Equipment needed includes collection booms, an open-mesh
chain-link belt for harvesting the oil-soaked sorbent, and a roller-wringer to
remove oil and water from the foam. The foam is initially comminuted and dis-
tributed onto the water by means of a hay blower (nulcher), and recycled foam is
distributed by an open-throat centrifugal blower. Recovered oil and water are
transported to shore in large fabric bags for further treatment prior to disposal.
Used foam is disposed of by incineration.
This report was submitted in fulfillment of Contract NO. 68-01-0067 under sponsorship
of the Water Quality Office, Environmental Protection Agency.
17tt, Descriptors
*011 Pollution, *011 Spills, *Water Pollution, Incineration, Water Pollution
Control
17 b. Identifiers
*Sorbent, *0il Skimmer, *Polyurethane Foam, Oil Spill Recovery, Hay Blower,
Chain-Link Belts, Recycled Sorbent
I7c. COWRR Field & Group
IS. A v sit ability
19, Security Class. '.
fRspor,)
'a. Se^ 'jrityC! is.
(Page)
st. it Ot of
Pages
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, O. C 2O24O
Abstractor
R- A. Cochran
[ institution Shell Development Company
WRS1C 1O2 (REV. JUNE 1971)
------- |