EPA-R2-73-156
JANUARY 1973 Environmental Protection Technology Series
Development and
Preliminary Design of a
Sorbent-Oil Recovery System
.*«eo sr4^
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
<*. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-R2-73-156
January 1973
DEVELOPMENT AND PRELIMINARY DESIGN
OF A SORBENT-OIL RECOVERY SYSTEM
By
E. Miller
L. Stephens
J. Ricklis
Contract No. 68-01-0066
Project 15080 HEV
Project Officer
Kurt Jakobson
Applied Science and Technology Branch
Environmental Protection Agency
Washington, D.C. 20460
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
For sale by the Superintendent of Doct^ff|,|^M .SoVernment Printing Office, Washington, D.C. 20402
-------
EPA Review Notice
This report has been reviewed by the Office of
Research and Monitoring, EPA, and approved for
publication. Approval;; does1 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 con-
stitute endorsement or recommendation for use.
11
-------
ABSTRACT
A development program was completed and preliminary designs
were prepared for 3000 gallon/hour protected water and
10,000 gallon/hour unprotected water Sorbent Oil Recovery
Systems. The five phases in the development program were:
(l) the characterization of the sorbent material., (2) the
development of the sorbent broadcasting system,, (3) the de-
velopment of the- harvesting conveyor and evaluation of over-
all recovery performance, (4) the. development of the sorbent
regeneration system and (5) model tests of a 1/4-scale model
recovery platform. The development program showed that a
continuous, sorbent-oil recovery system is feasible using 30
or 80 PPI polyurethane sorbent chips. In. one pass about
90 percent of the oil in a 1.5 mm slick can be recovered.
The water content of the recovered fluid is less than 10
percent. The preliminary designs are presented with detailed
descriptions of the system components,, opera'ting procedures,
and costs.
This report was submitted in fulfillment of Project Number
15080 HEV and Contract Number 68-01-0066 under the Sponsor-
ship of the Office of- Research and Monitoring, Environmental
Protection Agency.
ill
-------
CONTENTS
Section Page
I CONCLUSIONS 1
II RECOMMENDATIONS 3
III INTRODUCTION 5
IV DEVELOPMENT PROGRAM 9
CHARACTERIZATION OP THE SORBENT MATERIAL 9
4
DEVELOPMENT OF THE SORBENT BROADCASTING
SYSTEM 13
DEVELOPMENT OF THE HARVESTING CONVEYOR AND
OVERALL OIL RECOVERY PERFORMANCE 17
DEVELOPMENT OF THE SORBENT REGENERATION
SYSTEM 22
MODEL TESTS OF A 1/4-SCALE MODEL RECOVERY
PLATFORM 26
V SORBENT-OIL RECOVERY SYSTEM DESIGN 29
SYSTEM DESIGN PARAMETERS 29
PLATFORM CONCEPT SELECTION 31
CALCULATION OF SYSTEM CHARACTERISTICS 33
VI 3000 GALLON PER HOUR RECOVERY SYSTEM PRELIMINARY
DESIGN 37
GENERAL DESCRIPTION 39
RECOVERY SYSTEM COMPONENTS 39
RECOVERY PLATFORM 48
WEIGHT SUMMARY 57
CREW AND OPERATING PROCEDURES 62
COST ANALYSIS 64
v
-------
Section Page
VII 10,000 GALLON PER HOUR RECOVERY SYSTEM
PRELIMINARY DESIGN ?1
GENERAL DESCRIPTION 73
RECOVERY SYSTEM COMPONENTS 76
RECOVERY PLATFORM 80
CREW AND OPERATING PROCEDURES 91
COST ANALYSIS 91
VIII ACKNOWLEDGMENTS 97
IX REFERENCES 99
X APPENDICES !0l
vi
-------
FIGURES
Page
1 Effect of Viscosity and Residence Time on Oil
Absorption for 30 PPI Polyurethane Foam 11
2 Effect of Viscosity and Residence Time on Oil
Absorption for 80 PPI Polyurethane Foam 12
3 Full-Scale Sorbent Broadcast Experiment 14
4 The Movable Parallel Plate Nozzle and the
Broadcast Pattern Produced by it j6
5 Harvesting Conveyor Test Setup 19
6 Sorbent Chip Distribution Carriage 20
7 Harvesting Conveyor Operating in Waves 21
8 Sorbent Regenerator Test Apparatus 24
9 Overall View of 1/4-Scale Model Sorbent Recovery
Platform 27
10 Protected Water (3000 GPH) Recovery System
Arrangement Drawing 4l
11 Sorbent Regenerator Arrangement Drawing for
3000 GPH System .. . 46
12 Diagram of Fan and Pump Mechanical Transmission
for 3000 GPH System 49
13 Schematic Diagram of Fan and Pump Hydraulic
Transmission for 3000 GPH System 50
14 Schematic Diagram of Sorbent Regenerator and
Conveyor Hydraulic Transmission for 3000 GPH
System 51
15 Internal Arrangement of Platform Hulls 53
16 Midship Section for Aluminum Construction 55
vii
-------
Page
17 Speed Power Curves for Protected Water Oil
Recovery Unit 56
18 Unprotected Water Recovery (10,000 GPH) System
Arrangement Drawing 75
19 ' Diagram of Fan and Pump Mechanical Transmission
for 10, 000 GPH System 8l
20 Schematic Diagram of Fan and Pump Hydraulic
Transmission for 10, 000 GPH System 82
21 Schematic Diagram of Sorbent Regenerator and
Conveyor Hydraulic Transmission for 10,000 GPH
System 83
22 Viscosities of Oils as a Function of
Temperature 105
23 Specific Gravities of Oils as a Function of
Temperature 106
24 Effect of Viscosity and Residence Time on Oil
Absorption j 07
25 Effect of Viscosity and Residence Time on Oil
Absorption 108
26 Effect of Viscosity and Residence Time on Oil
Absorption 109
27 Effect of Viscosity and Residence Time on Oil
Absorption HO
28 Effect of Viscosity and Residence Time on Oil
Absorption HI
29 Effect of Viscosity and Residence Time on Oil
Absorption 112
30 Effect of Viscosity on Oil Absorption 113
31 Chip Broadcast Pattern Achieved in Early Test.... 119
viii
-------
Page
32 Simulated Debris Rake, Flotation . Pods, and
Side Screens Erected at Discharge End of
Broadcast Experiment .............................. 122
33 Chip Distributions Obtained with Passive
External Plate Device ............................. 123
34 Chip Distributions Obtained .with Mechanical
Nozzle ............................................ 124
35 Typical Effect of Strong Winds on Chip Dis-
tribution Patterns ................................ 126
36 The Blower Wheel Design Capable of Handling
Sorbent Chips ..................................... 133
37 Experimental ''Friction Coefficient" for Head
Loss Due to Chips Flowing in Duct ................. 137
38 Experimentally Determined Factors Which Account
for the Effect of Chips on Blower Head and
Power ............ ................................. 139
39 Model of Head and Power Interactions Between
Blower, Duct, and Chips ........................... l4l
40 Harvesting Conveyor Performance ...................
4l Oil Recovery versus Viscosity - (Calm Water)...... 148
42 Percent Oil in Recovered Fluid versus Viscosity
(Calm Water) ................ . ..................... 149
43 Oil, Recovery versus Viscosity (Waves ) ............. 150
44 Percent Oil in Recovered Fluid versus Viscosity
(Waves ) .......... . ................................ 151
45 Oil Recovery versus Residence Time (Calm' Water )... 153
46 Oil Recovery versus Residence Time (Waves) ........ 154
47 Oil Recovery versus Nominal Sorbent Coverage ...... 156
ix
-------
Page
_
48 Sorbent Oil Recovery Performance in Waves ........ 157
49 Water Recovered by Sorbent versus Slick
Viscosity ................. >••<•••, ...... ...........
50 Water Entrainment Rate Due to Conveyor versus
Conveyor Speed ......................... _ .......... J-59
51 Sorbent Regenerator Test Apparatus ............... 162
52 Sorbent Regenerator Test Apparatus Setup... ...... 163
53 Regenerated Sorbent Density versus Squeezing
Force (80 PPI Foam) No. 2 Heating Oil ............ 165
54 Regenerated Sorbent Density versus Squeezing
Force (80 PPI Foam) 011-70 Percent Bunker "C"
30 Percent No. 2 ................................. 166
55 Regenerated Sorbent Density versus Squeezing
Force (30 PPI Foam) Bunker "C" ................... 167
56 Regenerated Sorbent Density versus Viscosity ..... 169
57 Sorbent Chips After Endurance Test ............... 171
58 Oil Sorbent Recovery System; Pontoon Assembly. . . . 174
59 Mechanical Details of Sorbent Recovery
Platform Model ................................... 176
60 Test Setup for Sorbent Recovery Platform Model... 178
6l Sorbent Recovery Platform Deployment Draft -
Towline Drag versus Speed and Sea State .......... 180
62 Sorbent Recovery Platform Deployment Draft -
Towline Drag versus Speed ........................ 181
63 Sorbent Recovery Platform Under Tow at 9-5
Knots ............................................ 182
64 Sorbent Recovery Platform Operating Draft -
Towline Drag versus Speed and Sea State .......... 183
x
-------
Page
65 Sorbent Recovery Platform Drag Coefficient
versus Speed-Length Ratio 18JI
66 Recovery Platform Operating in Calm Water at
3.0 ft/sec 186
67 Recovery Platform Operating in Sea State 1 at
3 ft/sec 18?
68 Recovery Platform Operating in Sea State 3 at
3 ft/sec 188
69 Sorbent Recovery Platform Relative Motion in
Regular Waves 190
xi
-------
TABLES
1 Range of Parameters in Sorbent Material
Characterization Tests. .*.... j 10
2 Typical Limiting Parameters on a Sorbent-Oil
Recovery System. ...... . ....... 30
3 Desirable Recovery System Platform
Characteristics. \ .... L . 31
4 Advantages of Specifically Designed Recovery
Platform. ...... i . k ,....;....... 32
5 Advantages of a Vessel of Opportunity as a
Recovery Platform.. i. i ...... 32
6 Specific Design Goals for Protected Water
System. .' » 37
7 Characteristics of Protected Water Sorbent-Oil
Recovery System* . . . . . . . . ; 40
8 Broadcasting System Characteristics 42
9 Harvesting Conveyor Characteristics 43
10 Transfer Conveyor Characteristics 44
11 Sorbent Regenerator Characteristics 45
12 Platform Characteristics . . . i i i . 52
13 Outfit Items i * » . , i .- i t . i-. . . : 57
14 Weight Estimate for 3000 GPH Protected Water
Sorbent-Oil Recovery System. \ ; . .-
15 Duties of Recovery System Crew*
16 3000 GPH Protected Water Recovery System Cost
Analysis ..........^
xiii
-------
17 3000 GPH Protected Water Recovery System Total
Construction Costs '
18 Detailed Analysis of Protected Water Unit Re-
covery Machinery Cost • • ""
19 3000 GPH Protected Water Recovery System Direct
Operating Costs °9
20 Specific Design Goals for Unprotected Water
System
21 Characteristics of Unprotected Water Sorbent
Oil Recovery System
22 Broadcasting System Characteristics .............
23 Harvesting Conveyor Characteristics ............. 77
21! Transfer Conveyor Characteristics ............... 78
25 Sorbent Regenerator Characteristics ............. 79
26 Unprotected Water System Platform
Characteristics ................................. 84
27 Unprotected Water System Outfit Items ........... 85
28 Weight Estimate for Unprotected Water Sorbent-
Oil Recovery System ............................. 86
29 Unprotected Water Recovery System Cost
Analysis ........................................ 0,2
30 Unprotected Water Recovery System Total Con-
struction Costs ............................ no
31 Unprotected Water Recovery Unit Recovery
Machinery Cost nj.
32 Unprotected Water Recovery System Direct Op-
erating Costs
33 Oil Absorption Fluxes
xiv
-------
Page
34 Sorbent Chip Loss Rates Experienced in Pull
Scale .Model Tests 127
35 Types of System .Loss.es. .131
36 Viscosity of Test Oil Products • -168
xv
-------
SECTION I
CONCLUSIONS
1. Continuous sorbent oil recovery systems are practical
and can be designed to recover a wide range of oil
products.
2. Systems designed for recovery rates of 3000 GPH in pro-
tected waters and 10,000 GPH in unprotected waters will
basically satisfy the general and specific design goals
given by the EPA.
3. Sorbent oil recovery systems can recover about 90 per-
cent of the oil from a slick in a single pass. This
performance can be achieved over a wide range oil
viscosities,, slick thickness and wave heights.
4. Water in the recovered fluid will be about 10 percent
for typical operating conditions.
5. Depending on the options selected and the number pro-
duced,, the initial cost of a 3000 GPH protected water
recovery system will range between $40,000 and $80,000.
For a 10,000 GPH unprotected water recovery system the
cost will range between $55,000. and $105,000. Op-
erating costs for the 3000 GPH system will be between
$100/hour and $200/hour and for the 10,000 GPH system
between $400/hour and $600/hour.
6. Based on the results of the development program,
rational design procedures have been formulated for
the selection of system and sub-system characteristics
and hardware.
7. The sorbent material used in the recovery system should
be open cell reticulated polyurethane foam cut into
chips approximately 3 in. x 3 in. X 1/4 in. 80 PPI
foam should be used for oils with viscosities less
than 1000 cps and 30 PPI foam should be used for vis-
cosities greater than 1000 cps.
8. A pneumatic sorbent broadcasting system using a fan
with a long shavings type wheel and a moving parallel
plate distribution nozzle will provide satisfactory
sorbent distribution patterns and conveying performance,
-------
9. The recovery platform must .provide a sheltered location
for sorbent broadcasting to reduce losses in winds. If
this is done the sorbent lost rate will be small under
normal operating conditions and will not exceed 2 per-
cent of the broadcast rate in winds of 30 mph.
10. The recovery platform should have a catamaran configura-
tion because it provides a sheltered location for broad-
casting the sorbent, a means of herding the sorbent to
the harvesting conveyor and a location for mounting the
harvesting conveyor.
11. In order to achieve high recovery efficiency the sorbent
material should remain in contact with the slick for
periods of 15 to 30 seconds.
12. The harvesting conveyor should be at an angle of 4-5
degrees and run with a linear belt speed equal to the
sweep speed.
13. A converging belt type sorbent regenerator can regenerate
sorbent material at a loading of .30 ft3/ft2 and a belt
speed of 1.0 ft/sec. A squeezing force of about 200
Ib/in. of belt width is required. The density of the
regenerated sorbent after squeezing will be about 6.0
lb/ft3 which is satisfactory for pneumatic broadcasting.
14. The sorbent material can be regenerated with oil vis-
cosities up to 20,000 cps at the time of squeezing.
Heating coils will be required to reduce the oil vis-
cosity to about 1000 cps for satisfactory flow in the
oil collection pans.
15. The sorbent material can be regenerated in excess of
100 times without deterioration.
16. Model tests conducted at 1/4 full, scale for the 10,000
GPH system demonstrated that a continuous, stable,
sorbent broadcasting and recovery cycle is possible
and that it is not sensitive to waves and sweep speed.
2
-------
SECTION II
RECOMMENDATIONS
The development program and design studies carried out in
this project indicate that a continuous sorbent oil recovery
system is feasible and practical. In order to advance such
systems to field applications it is necessary to design..
build and test a prototype system. It is recommended that
such a prototype program be undertaken. Because of its more
general application and lower cost, the 3000 GPH protected
water unit would be a suitable candidate for such a program,
-------
SECTION III
INTRODUCTION
In the event of an oil spill it is desirable from the stand-
point of the total environment to physically remove the oil
from the water. There are several.basic methods which can
be employed for physical removal of the oil. These are:
direct skimming•by means of a special skimming device; the
use of pumps in conjunction with oil-water separation equip-
ment; or oil collection with a sorbent material followed by
mechanical removal of the sorbent material. The sorbent-
material method has the advantages that the sorption process
is not affected by waves or current and that a high percentage
of oil in the recovered fluid can be obtained. As a result,
large volumes of sorbent material have been used in oil spill
clean up operations. However, this method has disadvantages
such as the generation of a large volume o-f waste material
which requires controlled disposal and the lack of mechanical
systems to recover the sorbent material. Thus., the use of
sorbent materials for oil recovery has been an effective but
expensive procedure.
In an effort to overcome the disadvantages associated with
the use of sorbent material, the U. S. Environmental Pro-
tection Agency issued a RFP for the development of Sorbent
Oil Recovery Systems in late 1970. HYDRQNAUTICS, Incorporated
was awarded one of five research contracts in June of 1971
for the development of a sorbent oil recovery system based
on the use of a synthetic foam sorbent material which could
be recycled many times. This report presents the results of
this development effort.-
In order to guide the developmental phase, the Environmental
Protection Agency provided the following general design goals
for an oil recovery system using sorbent materials.
GENERAL DESIGN GOALS
1. Rapid oil recovery rate.
2. Complete removal of oil from the water surface.
3. Minimum amounts of water entering each unit process.
-------
4. Minimum influence of water motion and waves on collec-
tion efficiency.
5. Minimum amount of auxiliary equipment.
6. Reject floating solids of a size which will interfere
with the efficiency of, or damage, the recovery system.
7. High mobility and maneuverability.
8. Compatibility with marine life.
9. Reasonable first cost.
10. Low operating expense.
11. Minimum maintenance requirements.
12. Maximum ease and speed of repairs.
13. Readily available replacement parts.
14. System'independent of physical properties of oil.
15. Readily available sorption material.
16. Sorption material easily transportable to location
of spill.
I?. Sorption material must be compatible with land harvest
methods.
18. Sorption material must not increase flammability of
oil.
In addition, specific design goals were provided for a pro-
tected water system capable of recovering 3000 gallons/hour
from a 1.. 5 mm, thick slick and an unprotected water system
capable of recovering 10,000 gallons/hour from a 1.5 mm
slick. These'specific design goals are outlined in the sub-
sequent sections of this report along with preliminary de-
signs for these two systems.
-------
In the development of a sorbent oil recovery system it is
desirable to consider the basic unit operations involved.
These basic unit operations are:
1. Sorbent Broadcasting
2. Oil-Sorbent Harvesting
3. Oil-Sorbent Separation
]4. Vessel or Platform Configuration
5- Oil Storage or Disposal
6. Sorbent Reuse or Disposal
Based on these unit operations., a development program was
formulated and carried out which included five experimental
tasks and a preliminary design task. The five experimental
tasks were:
1. Characterization of the Sorbent Material
2. Development of the Sorbent Broadcasting
System
3. Development of the Harvesting Conveyor and
Evaluation of Overall Oil Recovery Per-
formance
4. Development of the Oil Sorbent Separation
System, and
5. Model tests of a 1/4-scale Recovery Platform
to determine system stability and operation
in waves.
The preliminary design task included the formulation of de-
sign procedures and criteria based on the five experimental
tasks and the preparation of preliminary designs for the
3000 and 10,000 gallon/hour recovery systems.
The complete results of the sorbent oil recovery system de-
velopment program are presented in this report.. Section IV
presents a brief description of each of the five experimental
tasks in the development program; the details and the data
from each of these tasks are presented in the appendices.
Section V presents a summary of the design methods, criteria
and observations which resulted from the development program.
Sections VI and VII present descriptions and technical de-
tails from the preliminary designs of the 3000 and 10,000
gallons per hour recovery systems, respectively.
-------
SECTION IV
DEVELOPMENT PROGRAM
Five experimental tasks were carried out in support of the
development of an Oil-Sorbent Recovery System. These five
tasks are briefly described in this section and additional
details are presented in the appendices.
CHARACTERIZATION OP THE SORBENT MATERIAL
The objective of this task was to develop data on the oil
absorption capabilities of the sorbent material as a function
of sorbent characteristics, oil characteristics, slick thick-
ness and residence time on the slick. The sorbent material
selected for use in the oil-sorbent recovery system was an
open-cell reticulated polyurethane foam made by the Scott
Paper Company. This material was selected during the pro-
posal stage of this project and was based on previous work
carried out by HYDRONAUTICS, Incorporated using sorbent ma-
terials. This work is reported in Reference 1.
The experiments were carried out in a 4 ft by 4 ft plastic
tank 10 in. deep. This tank was equipped with a temperature
control system, temperature measuring devices, and a microm-
eter for measuring oil slick thickness. Tests were conducted
by establishing a slick of known characteristics and
thickness on the water surface. A chip of sorbent material
of known weight was dropped on the slick and allowed to sit
for a given time. The chip was then picked up, allowed to
drain for 10 seconds and reweighed to determine the amount
of oil picked up. This process was carried out with several
chips and an average value was taken for the amount of oil
absorbed. The range of parameters covered in the tests are
given in Table 1.
In a system in which the sorbent material is reused, the im-
portant thing is not to maximize the quantity of oil absorbed
per unit volume of sorbent, but rather to maximize the total
rate of oil absorption. As a result, a high percentage of
the slick surface must be covered with sorbent material and
individual sorbent chips will not have the opportunity to
absorb oil to their maximum capacity. Thus, it is most
meaningful to express the oil absorption performance of the
sorbent as an absorption ratio rather than as either the
-------
absolute volume of oil absorbed or the maximum volume of oil
that could be absorbed. The absorption ratio is defined as
the ratio of the amount of oil absorbed to the amount of oil
directly beneath the sorbent chip. This ratio indicates the
extent to which oil in the slick can flow in the edges of the
chip and the inverse of this ratio indicates the percentage
of the slick which must be covered with sorbent for complete
oil recovery.
TABLE 1
Range of Parameters in Sorbent Material
Characterization Tests
Parameter
Range
Slick Thickness
Oil Type
Slick Temperature
Residence Time of Sorbent
on Slick
Sorbent Porosity
Sorbent Chip Geometry
Sorbent Reusability
!! rill
0.5 - 3-5 mm
Diesel, No. 4, Crude, Bunker "C
4 - 27°C
0 - 60 sec. Mostly 5, 10, 15,
30 sec.
30,60,80,100, Mostly 30 and
80 PPI
Square and Rectangular
Fresh and Regenerated, Mostly
Regenerated
Most of the tests were carried out using 3 in. x 3 in. x 3/8
in. chips of the sorbent material. Tests with rectangular
chips did not show significant improvements in the absorption
ratio. It was believed that rectangular chips might be more
difficult to broadcast so they were not considered further.
Typical test results are presented in Figures 1 and 2 which
show the effects of oil viscosity and residence time on the
absorption ratio. Both figures apply to slick thicknesses
of 1 to 2 mm. Figure 1 is for 30 PPI (pores per linear inch)
and Figure 2 is for 80 PPI polyurethane foam. The test re-
sults for other conditions as well as a more detailed de-
scription of this task are presented in Appendix A.
10
-------
I—
280
240
200
160
Z
O
$ 120
CD
100
80
40
1 T
POLYURETHANE FOAM
POROSITY - 30 ppi
SLICK THICKNESS 1 -2mm
O 5 sec
D 10 sec
O 15 sec
A 30 sec
DIESEL
CRUDE
BUNKER "C"
10
100
1000
10,000
KINEMATIC VISCOSITY cm /sec
FIGURE 1 - EFFECT OF VISCOSITY AND RESIDENCE TIME ON OIL ABSORPTION
FOR 30 PPI POLYURETHANE FOAM
100,OOC
-------
POLYURETHANE FOAM
POROSITY - SOppi
SLICK THICKNESS 1 - 2mm
O 5 sec
D 10 sec
O 15 sec
A 30 sec
1000
KINEMATIC VISCOSITY cm /sec
10,000
OJ
100,000
FIGURE 2 - EFFECT OF VISCOSITY AND RESIDENCE TIME ON OIL ABSORPTION
FOR 80 PPI POLYURETHANE FOAM
-------
The important results from this development task can be
summarized as follows:
- The absorption ratio decreases with increasing oil
viscosity. For oils such as No. 6 and Bunker "C"
almost 100 percent coverage of the slick will be
required.
The absorption ratio does not increase greatly with
residence time for viscous oils such as No. 6 and
Bunker "C".
30 and 80 PPI foam have equal absorption ratios for
.Oils with viscosities greater than 50 cps. For lower
viscosities., the 80 PPI foam is superior because oil
will not drain out "of it as it .is removed from the
slick. Tests were not conducted with viscosities
.., less than 2 cps.
Square chips have as high absorption ratios as
rectangular chips for the range of slick thickness
and oil viscosities of interest.
DEVELOPMENT OF THE SORBENT BROADCASTING SYSTEM
The objectives of this task were:
To develop and demonstrate a means of evenly dis-
tributing sorbent material chips across the Inner
width of the recovery platform at its forward end.
To develop and demonstrate a pneumatic system for
conveying chips from the oil removal section of the
recovery platform to the distribution device.
These objectives were accomplished with developmental model
tests. Due to the specialized nature of the problem,, the
model size and other parameters were maintained at near full
scale to reduce scaling uncertainties in the prototype design
stage. Figure 3 shows an overall view of the experimental
setup. The test program was directed at the development of
a satisfactory broadcasting nozzle and at the measurement of
the pneumatic conveying system performance.
13
-------
INTERIOR OF FEEDER BOX
BASIC NOZZLE WITH TRIAL GUIDE VANES INSTALLED
OVERALL VIEW OF EXPERIMENT
FIGURE 3 - FULL SCALE SORBENT BROADCAST EXPERIMENT
-------
For the purpose of the broadcasting experiments,, it was as-
sumed that a 100 percent nominal coverage of the slick would
be required. Thus, using 3 in. y 3 in. square chips, for a
slick thickness of 1.5 mm. a recovery rate of 3000 GPH would
require broadcasting 21,'iOO chips/ml mite. A rate of 10,000
GPH would require a rate of 7^,000 chips/minute. Initial
design studies indicated that, for the 10,000 GPH recovery
unit, the sorbent broadcast pattern would have to be about
32 ft wide. It was hoped that a single passive nozzle- with
fixed guide vanes could achieve a uniform distribution with
this width. However, the tests showed that, at near the
design chip rate, a simple nozzle with fixed vanes would jam
with chips. To overcome this problem, a movable parallel
plate nozzle concept was developed. This nozzle produced
satisfactory distribution patterns at the design chip rate.
Figure j! shows the nozzle installed on the experimental
setup. Tests In wind showed some distortion of the broad-
cast pattern and Indicated that the chip loss rate would be
less than 2 percent of the broadcasting rate.
The pneumatic conveying system performance parameters were
measured during the tests conducted for nozzle development.
These performance parameters include the air speed necessary
for good conveying, the head loss in the transport duct, the
head, loss across the fan, and the fan power required. These
parameters were determined as a function of the ratio of the
weight of sorbent material to the weight of air. During the
tests, it was determined that the sorbent chips could be fed
directly through a centrifugal blower equipped with a long
shaving type wheel. Preliminary tests revealed that the
chips should be'stiff enough so that, upon contact with the
wall of the transport ducts, they would not collapse and
flatten completely against,the wall. This implies a thick-
ness of 1/4 in. to 3/8 In. for the 3 in. by 3 in. chips.
Based on the results of the test program, a design procedure
was formulated for estimating the size, air flow rate, fan
characteristics and power of a pneumatic conveying system.
A more detailed description of the sorbent broadcasting
system development program, including the test data, the
design procedure and an example problem, is presented in
Appendix B.
15
-------
FIGURE 4 - THE MOVABLE PARALLEL PLATE NOZZLE AND THE
BROADCAST PATTERN PRODUCED BY IT
16
-------
The important results from this development task can be
summarized as follows:
- A sorbent broadcasting system can be designed, based
on a pneumatic conveying principle, which will meet
the performance requirements and yet be of practical
size.
- A movable parallel plate nozzle concept will produce
a satisfactory broadcast pattern width and distri-
bution.
- Wind will distort the broadcast pattern somewhat
but the sorbent loss rate will be less than 2 per-
cent of the broadcast rate.
A rational design procedure was. developed for
selecting components for a pneumatic conveying
system.
•t
DEVELOPMENT OP THE HARVESTING CONVEYOR AND OVERALL OIL
RECOVERY PERFORMANCE
The objectives of this task were:
To determine the conveyor inclination angle and
speed for optimum performance.
To uevelop data on system performance as a function
of the operating parameters, using the optimum con-
veyor operating condition.
The system performance was defined both in terms of the
percentage of the oil in a slick which is recovered and the
percentage of oil in the recovered fluid. The general and
specific design goals supplied by the EPA request 100 per-
cent recovery of the oil from the slick and in excess of
90 percent oil in the recovered fluid. The operating pa-
rameters which were studied in addition to the conveyor
angle and speed included:
- Slick thickness and viscosity (oil type and'
temperature)
17
-------
Sorbent coverage
- Residence time
Calm water versus waves
The objectives of this task were accomplished by means of
an experimental program. In order to avoid scaling problems,
the tests were carried out using typical oil products, full-
size sorbent chips, and the actual conveyor belt material.
The tests were conducted in a tank which is 80 ft long 2 ft
wide and 2 ft deep. In effect this test setup was equivalent
to a 2 ft wide strip along the centerline of the recovery
system. Figure 5 is a photograph of the harvesting conveyor
setup in the 80-ft tank. The conveyor material is 1 in.
mesh-opening flat-wire belt. Figure 6 is a photograph of
the carriage used to distribute the sorbent chips on the slick
in the tank. The tests were conducted in accordance with
the following procedure: First, a slick of known thickness
was established on the water surface in the tank. The sor-
bent material was then distributed on the slick by the sor-
bent distribution carriage. After waiting the required
residence time, the harvesting conveyor was run down the
tank collecting the sorbent chips and dropping them in a
collection box. Figure 7 shows a typical test in progress.
After the run, the chips in collection box were regenerated
and the amount of oil and water measured. The amount of oil
remaining in the tank was also measured to determine the
recovery performance or efficiency.
The first part of the test program was directed at deter-
mining the effects of conveyor-belt angle and speed on per-
formance. Based on these tests, a conveyor angle of ^5
degrees and belt speed equal to forward speed were selected
for the remaining tests. These tests covered the range of
parameters listed above. These data were reduced to pa-
rametric curves showing percent of oil recovery and percent
oil in the recovered fluid as a function of the independent
parameters. These curves may be used directly in the design
and performance prediction of sorbent oil recovery systems.
These curves and a more detailed description of this task
are presented in Appendix C.
The important results from this development task can be
summarized as follows:
18
-------
FIGURE 5 - HARVESTING CONVEYOR TEST SETUP
-------
FIGURE 6 - SORBENT CHIP DISTRIBUTION CARRIAGE
-------
FIGURE 7 - HARVESTING CONVEYOR OPERATING IN WAVES
21
-------
- A 1 in. mesh flat-wire belt will be satisfactory
for the harvesting conveyor.
The harvesting conveyor should be at an angle of
45 degrees and run at a belt speed equal to the
forward speed.
The oil recovery percentage will range from 90 to
95 percent over a wide range of oil types and slick
thicknesses.
- The percent oil in the recovered fluid will meet the
EPA goal of 90 percent for slick thickness of 1.5 mm.
Waves will improve the oil recovery percentage
relative to calm water because they agitate the
sorbent causing it to contact a higher percentage
of the slick surface.
DEVELOPMENT OF THE SORBENT REGENERATION SYSTEM
The polyurethane foam sorbent material can be regenerated
for reuse simply by mechanically squeezing out the recovered
oil. The sorbent regeneration system must perform this
squeezing operation on a large enough volume of sorbent ma-
terial to satisfy the required flow rate over the range of
operational requirements. In effect, the sorbent regenera-
tion system is the key element which makes a continuous cycle
sorbent oil recovery system possible. The sorbent regenera-
tion system is also the most complex mechanical component
in the system.
The concept of a converging-belt squeezing device was selected
for development as the sorbent regenerator. In this concept,
the sorbent material is carried between two conveyor belts
through a series of squeezing rollers. The upper conveyor
belt converges with the lower belt at the entrance to the
first pair of squeezing rollers. The upper belt is solid
and the lower belt is porous so that the oil squeezed out
flows through it into collecting pans. This concept was
selected for the following reasons:
- A large volume of sorbent can be handled between
small-diameter squeeze rollers.
22
-------
- There is no relative movement between the belts
and the sorbent which could damage the sorbent.
- The squeeze rollers, can be mounted to allow large
travel between them. This makes the regenerator
tolerant of debris and drift wood which can run
through the machine without causing damage.
- The physical dimensions of the resulting machine
are compatible with the arrangement of the overall
system.
A development task was carried out based on the converging-
belt regenerator concept. The objectives of this task were:
Develop performance data for a converging belt
sorbent regenerator as a function of the design
parameters.
Determine the effects of repeated cycles on the
sorbent material.
Identify and resolve mechanical problems prior
to the design of a prototype system.
These objectives were carried out by designing, building
and testing a sorbent regenerator test apparatus. This
test apparatus was about the same size and included many of
the features of a sorbent regenerator for a 3000 GPH re-
covery unit. Figure 8 presents photographs showing overall
views of the sorbent regeneration test apparatus and the
system under test. The test is being carried out with 80
PPI foam and an oil mixture of ?0 percent bunker "C" and
30 percent No. 2. The recovered oil can be seen flowing
into the barrel. In a typical operating condition, the
sorbent regenerator will have to remove about 2 gallons of
oil per cubic foot of sorbent.
The broadcasting process which directly follows regeneration
is sensitive to the density of the sorbent. Thus the per-
formance of the sorbent regenerator is defined in terms of
the density of regenerated sorbent from which 2 gallons of
oil per cubic foot have been recovered.
23
-------
Overall View of Test Apparatus
System Under Test
FIGURE 8 - SORBENT REGENERATOR TEST APPARATUS
-------
The test program was carried out to determine the influence
of various parameters on the density of the regenerated sor-
bent. These parameters included:
- Squeezing Force
- Belt Speed
Sorbent Loading
Sorbent Type
- Oil Viscosity
The results were plotted as parametric relationships which
allow the selection of squeezing force, belt speed,, sorbent
loading and sorbent type for a given viscosity and required
density. These parameters determine the mechanical char-
acteristics of the sorbent regenerator. In addition, the
test program covered endurance tests on the sorbent. Ob-
servations during the testing indicated several mechanical
refinements which should be included on the prototype design.
/
Appendix D presents further details with respect to the
sorbent regenerator development, test programs, data, and
observations. The important results of this task can be
summarized as follows:
A converging-belt sorbent regenerator will
satisfactorily regenerate sorbent chips at the
volume rates required.
- The density of the regenerated chips will be-
about 6 lb/ft3 for a squeezing force of 220 lb/
in., a sorbent loading of 0.25 ft3/ft2 and a
belt speed of 1 ft/sec.
In order to maintain a 6 lb/ft3 regenerated
sorbent density, 30 PPI foam should be used for
oil viscosities in excess of 1000 CPS.
- An oil viscosity of about 20,000 CPS at the time of
regeneration is the practical upper limit because
of the excessive density of the regenerated sorbent
- The sorbent material can be cycled over 100 times
without significant degradation.
-------
MODEL TESTS OF A 1/4-SCALE MODEL RECOVERY PLATFORM
The objectives of this task were:
"~y. g. ,., ' •' '",•"*
:•"• ^ To determine,5 whether a-> continuous stable sorbent
broadcast and.recovery cycle is possible.
., ,. ' '»••!•* ,,, •'."
'^/ ^i i a
To'determine if there are any adverse effects of
wave's">nd forward speed on the continuous broadcast
and'1 recovery cycle.; ,'
- To determine the towing resistance and stability
of the recovery platform at the deployment and
operating draft.
*
The objectives of this task were satisfied by means of tests
conducted on a 1/4-scale model of a recovery platform concept
in the HYDRONAUTICS Ship Model Basin (HSMB®). The model was
equipped with an operating broadcasting system and harvesting
conveyor. No squeezing system was fitted on the model and
no attempt was made to actually recovery oil. The recovery
platform model was based on the concept presented in the
original HYDRONAUTICS,, Incorporated Sorbent Oil Recovery
System Proposal. Subsequent preliminary sizing studies in-
dicated that the basic concept and general proportions were
still valid. Thus the model was sized to be a 1/4-scale
model of a system intended to recover 10,000 gallons per
hour. Figure 9 is a photograph showing the overall arrange-
ment of the model.
Tests were conducted for both towing and pushing the recovery
platform over a range of speeds up to 6 ft/sec full scale and
over a range of wave heights up to a low Sea State 3 full
scale. Measurements were made of the towing drag, speed,
wave height and relative motions between the water surface
and the platform. Observations and photographs were obtained
of the sorbent distribution and collection. The data and
detailed observations obtained from these tests, along with
a more detailed description of the model and test procedures,,
are presented in Appendix E.
The important results from this development task can be
summarized as follows:
26
-------
FIGURE 9 - OVERALL VIEW OF 1/4-SCALE MODEL SORBENT
RECOVERY PLATFORM
-------
A continuous sorbent broadcasting and recovery
operation with a uniform stable distribution of
sorbent material can be achieved.
The sorbent broadcasting and recovery operation
will not be degraded by waves up to a low Sea
State 3 and forward speed up to 6 ft/sec. . ;
The movable parallel plate broadcasting nozzle
concept developed in a previous task provided a
uniform transverse distribution of sorbent
material.
The sorbent material did not show any tendency to
plug the inlet to the harvesting conveyor.
The recovery platform can be towed up to a speed
of 9-5 knots in the deployment condition.
The recovery platform is directionally stable
under tow.
28
-------
SECTION V
SORBENT-OIL RECOVERY SYSTEM DESIGN
Based on the results of the development programs described
in Section IV, design procedures have been defined for
sorbent oil recovery systems and the results are presented
in this section. These design procedures apply to con-
tinuous-cycle sorbent broadcast,, recovery,, regeneration sys-
tems. In order to minimize the effects of environmental
parameters such as wind and waves it is intended that the
sorbent material remain in unrestrained contact with the
slick long enough to absorb the oil. The following dis-
cussion does not apply to systems in which the sorbent ma-
terial is intended to form a cake or filter to remove the
oil.
SYSTEM DESIGN PARAMETERS
The independent parameters which determine the character-
istics of a sorbent oil recovery system are:
Oil characteristics (i.e.,, viscosity)
- Slick Thickness
- Total Oil Recovery Rate
Sweep Speed of the Recovery Platform
Sorbent Residence Time
Percent of Surface Covered by Sorbent
These parameters can be used to define the basic geometric
parameters of the platform such as sweep width and active
length,, and operating parameters such as sorbent broadcast
rate,, broadcast distances and water recovery rate. In
general, the first three parameters listed are specified as
part of the specific design goals for the system; the sys-
tem designer can select the last three parameters listed.
The percent of the surface covered by sorbent and the
residence time determines the percent of oil recovered from
the slick for given oil characteristics and slick thickness,
The relationship between these parameters is illustrated in
Figure 48 for typical conditions. With the percent oil
29
-------
recovery, the total slick area that must be swept per unit
time to satisfy the total oil recovery rate can be calculated.
The sweep rate (area per unit time) and the percent sorbent
coverage define the sorbent broadcast rate. The sweep rate
divided by the sweep speed gives the width of the sweep or
sorbent broadcast pattern. The product of sweep speed and
residence time gives the active length of the system. The
active length is defined as the distance between the sorbent
contact point with the slick and the harvesting conveyor.
The above shows that there are an infinite number of combina-
tions of sweep speed, residence time and sorbent coverage
which will satisfy a given set of specific design goals. As
a result, other parameters must be introduced in order to
produce a unique design solution. These parameters may in-
clude:
Minimum cost
Limitations on system or component dimensions
Practical limits on the range of the independent
parameters
General design goals
In general, the procedure is to design for minimum cost with-
in the limitations imposed by the other parameters. Table 2
lists typical limiting parameters except cost.
TABLE 2
Typical Limiting Parameters on a Sorbent-
Oil Recovery System
Dimensions
Practical
General Design
Goals
Limiting component
length, width, height
and weight for highway
or air transport.
Overall length,
width, height or
draft of restricted
areas (under pi^rs,
between ships).
Residence times in
excess of 30 sec-
onds do not im-
prove recovery
percentage.
Sorbent coverage
in excess, of 90
percent does not
improve recovery
percentage.
Maximum practi-
cal oil recovery
percentage.
Others as listed
in the Intro-
duction.
30
-------
In each case, the system designer must identify which of
these limiting parameters apply and assign them a weight.
Ideally the cost that should be minimized is the total life-
cycle cost per gallon recovered discounted to" present''value.
In practice, this is not possible since the operating sce-
nario, the social discount rate, and the social value of a
gallon of recovered, oil are -not known. As >a 'pesult,, the^-
simplified procedure of-minimizing the sum of the-'capital'* •
and operating cost based on an assumed operating cycle has
been adopted. . .
PLATFORM CONCEPT SELECTION
In the design of a sorbent-oil recovery system, a decision'
must be made as to whether to mount the system' on a vessel
of opportunity or on a specifically designed vessel- "or plat-
form. In either case, the platform characteristics, must be
suitable for the recovery system equipment. Some -of- the"
desirable characteristics of a sorbent-oil recover-y s.ystem
platform are listed in Table 3- • • '
TABLE 3
Desirable Recovery System Platform Characteristics;-
Configuration which protects- and confines sorbent' 'ma-
terial from wind effects during broadcasting. --x
Configuration which channels the sorbent material to -*
the harvesting conveyor. _
Location and support structure for harvesting conveyor.
- Unobstructed run for sorbent broadcasting duct.
Location for mounting additional sweeping boom's.
- Minimum relative motion in waves.
Tankage for recovered oil.
- Acceptable towing resistance and directional stability.
Ease of assembly and disassembly for transport.
31
-------
A basic catamaran configuration has most of the desirable
features listed in Table 3. In particular the area between
the hulls provides a sheltered location for broadcasting the
sorbent and channeling the sorbent to the harvesting conveyor,
The harvesting conveyor can be mounted between the two hulls.
The only advantage of a more conventional ship-like platform
would be its lower towing resistance. Based on these con-
siderations, if a platform is to be designed specifically for
a sorbent-oil recovery system, it should be a catamaran con-
figuration.
The selection of a vessel of opportunity or a specifically
designed platform for mounting the recovery system depends
on the design goals. Typical advantages of using either a
specifically designed platform or a vessel of opportunity
are listed in Tables ;4 and 5, respectively.
TABLE ^
Advantages of Specifically Designed Recovery Platform
Smaller Physical Dimensions
Better Control of Broadcast Sorbent and
Lower Sorbent Loss Rate
Shorter Setup and Reaction Time
Smaller Crew Required
Lower Direct Operating Costs
TABLE 5
Advantages of a Vessel of Opportunity as Recovery Platform
Simpler Transportation over long distances
(only the recovery system equipment need be
shipped)
Lower Initial Cost
-------
These tables indicate that, if the system is going to be used
continuously, the operating efficiency and lower operating
cost of a specifically designed platform would dictate its
selection. This is the case for a harbor unit which will be
used routinely to clean up small spills. The selection is
not as clear for an unprotected water system which will be
used only occasionally against larger spills. The specifi-
cally designed platform for the unprotected water recovery
system described in Section VII, alone will cost between
$20,000 and about $40,000 depending on whetner steel or
aluminum construction is used. A system mounted on a vessel
of opportunity would not require the special platform but
would require 60 to 80 feet of open water containment barrier
which could cost $8000. The charter rate of a barge to mount
the system on would be about $100 per day. Thus the charter
fee would equal the additional cost of this platform in be-
tween 120 and 320 operating days. Because of this and the
increased efficiency of a specific platform, which is hard
to quantify, the unprotected water system described in
Section VII was based on a specifically designed recovery
platform.
CALCULATION OP SYSTEM CHARACTERISTICS
To illustrate the procedures for selection of the system
characteristics which are discussed above, the following
example calculation for a protected water recovery unit
is given:
PROTECTED WATER RECOVERY UNIT
•
Given: Recovery Rate, ¥R = 3000 GPH
Slick thickness = 1.5 mm
Oil Viscosity < 6000 cps
Find: System characteristics for minimum cost
Oil volume per ft2 of slick = 0.0368 gals/ft2
Try: Residence time, TR = 15 seconds
Sorbent coverage, C = 0.80
s
From Figure 48
Oil Recovery r}_ = 0.8?
A
33
-------
¥R
-Area to be swept/hour, A , = nQ,gQ~~~r~ = 94,000
s 0.0368 x T)R ft2/hr
Sweep speed - U , ft/sec
1 ' S
Swe'ep width, ¥ = A /U ft
s s s
.Acti'ye' length L. = U X TR = ft
A S xl
•A '> Sorbent chip area in2 t = sorbent chip thickness,
t •*
Sorbent broadcasting rate,
•••• : '' .. A x c x t
S 3600 X 12
Sorbent broadcasting rate,
'' ' A x C .x 144
• , O Q i
N = , chips/sec
C 3600 X A
• V ' = 0.436 ft3/sec
S
N \= 334 chips/sec
VX
In order to 3elect ,a sweep speed it is necessary to have a
relationship for cost as a function of system geometry. It
is assumed that the operating cost is essentially independent
of the sweep speed and broadcasting rate. Thus it is only
necessary to consider those parts of the first cost which
are; functions of sweep speed and broadcasting rate. In order
to 'define the cost relationships it is necessary to carry
out, a. concept .design and cost estimate for the system. This
has been done and the relationship for systems similar to
the Harbor Unit presented in Section VI is:
L.
Relative Cost = 22,000 (0.80 x ^ + 0.20 x (¥s/l6)2)
V W L
+ 15000 X 7-—- (0.55 + 0.25 X t-| + 0.20 X -£)
34
-------
The first term is related to the recovery platform and the
second term is related to the broadcasting, recovery and
regeneration equipment. The relative cost as a function of
sweep speed is:
Sweep Speed, U
ft/sec s
1.0
1.5
2.0
Sweep
Width, W
ft £
26.1
17-^-5
13-0
Active
Length, L
ft
A
Relative
Cost
dollars
15 35,800
22.5 32,050
30.0 33,400
This indicates that for a 15-second residence time and
80-percent sorbent coverage that the optimum sweep speed is
about 1.6 ft/sec. The effects of residence time and per-
cent sorbent coverage can also be studied. For example,
reductions in residence time to 10 seconds will reduce the
relative cost about 2 percent. However, the oil recovery
ratio drops to 0.75 which departs further from the general
design goals of complete oil removal. Thus the optimum
characteristics for a protected water sorbent oil recovery
systems are:
Recovery Platform Type:
Sweep Speed
Sweep Width
Active Length
Sorbent Broadcast Rate
Catamaran
1.6 ft/sec
16 ft
24 ft
0.44 ft3/sec
These characteristics will also satisfy the typical limiting
parameters listed in Table 2. For example, the hull di-
mensions which will result from these characteristics are
such that the two hulls can be carried side 'by side over the
road. The overall dimensions of the assembled platform are
comparable to those of other oil recovery craft, such as the
M.V. Port Service of Baltimore, which work around piers and
ships.
35
-------
Once the basic system characteristics have been established,
the detailed characteristics of the broadcasting and sorbent
regeneration system can be determined using the procedures
detailed in Appendices B and D.
36
-------
SECTION VI
3000 GALLON PER HOUR RECOVERY SYSTEM PRELIMINARY DESIGN
A preliminary design of a sorbent oil recovery system for
.us.e in protected waters was prepared based on the general
and specific design goals provided by the EPA, These
specific design goals are given in Table 6.
TABLE 6
Specific Design Goals for Protected Water System
a. Environment
Perform effectively in harbors and other protected
waters with 2-foot waves in combination with 20-mph
winds and 6-knot currents.
b. Sorbent Recovery System
The sorbent recovery system in combination with neces-
sary oil-water separation facilities shall have the
capacity to recover at least 1,500 gal/hr of oil with
10 percent or less water content. However, a ^,,000
gal/hr with 10 percent or less water content oil re-
covery rate would be a desired optimum rate.
(l) The sorbent system shall be capable of meeting the
recovery rate objective while recovering oils with
viscosities ranging from that of light diesel oil
to near water density heavy asphalt at 20°C.
(2) The thickness of the slick to be recovered at the
specified rate shall be 1.5 mm or less.
"(3) The device may include as an integral part of the
skimming process, a system of booms to aid in
herding oil toward the harvesting device. How-
ever, emphasis shall be given to the skimming
process rather than the design of the boom.
(4) Designed units proposed for temporary attachment
to existing vessels must consider transport need
and ease of installation,wi.th a- minimum of special
equipment. ^^'
s"
s'
37
-------
c. Oil-Sorbent Separation
The unit must be capable of operating under hydraulic
loadings' and ranges of oil concentrations consistent
with the performance of the oil harvesting devices. It
must be capable of producing output streams with the
following characteristics:
1. Oil - 10 percent water or less.
2. Sorbent material - available for reuse or method for
disposal.
3. Water - 10 mg/1 oil or less.
d. Oil Storage or Disposal
Specify floating or appropriate land-based facilities
which in combination with any on-board storage will
have capacity to store .material to be collected from
spills of at least 200,000 gallons of oil plus agglome-
rates. Agglomerates must be processed before land dis-
posal to preclude leaching. Other appropriate disposal
techniques such as incineration will be considered as
alternatives.
e. Vessel
Recovery equipment may be designed for either permanent
or temporary mounting on an appropriate vessel. Vessels
must be capable of speeds of at least 12 knots under the
environmental conditions listed above (however not per-
forming the oil recovery function). The vessel must
accommodate the oil-sorbent recovery device, any required
oil-water separation equipment, operating personnel, eight
hour fuel supply and all oil specified for storage on
board. It must be sufficiently maneuverable to minimize
time lost in recovering oil from constricted locations
and be capable of operating in waters as shallow as
3-foot depth.
38
-------
The system was designed to satisfy these desired recovery
rates of 3000 GPH of heavy oil at 20°C, For design purpose,
the heavy oil was taken as bunker "C" with a specific gravity
of 0.98 and a viscosity of 4,500 centepoise at 20°C. The
rationale for the selection of the system operating param-
eters such as sweep speed, sorbent coverage .and sorbent
residence time are presented in Section V.
GENERAL DESCRIPTION
The basic sorbent recovery system consists of a pneumatic
broadcasting system with a moving parallel plate nozzle, a
harvesting conveyor, a transfer conveyor, a sorbent re-
generator, the necessary drive systems, and a catamaran type
platform. The basic characteristics of the system are pre-
sented in Table 7. Figure 10 presents an overall arrange-
ment drawing of the complete system. The sorbent material
to be used and the recovery performance over a range of con-
ditions are also given In Table 7.
RECOVERY SYSTEM COMPONENTS; BROADCASTING SYSTEM
The design of the pneumatic broadcasting system and broad-
casting nozzle was based on the design procedures presented
in Appendix B. This system is required to move and dis-
tribute 330 3 in. X 3 In. x 1/4 in. sorbent chips per
second. The technical and performance characteristics of
the broadcasting system are presented in Table 8. The
system uses an off the shelf industrial fan with a long
shavings type wheel. The only non-standard component in
the system is the broadcasting nozzle. The nozzle plates
are oscillated with a mechanical linkage driven by an air
cylinder.
Harvesting and Transfer Conveyors
The harvesting conveyor is mounted between the hulls of the
recovery platform. It has an overall width of 8.0 ft and a
length between sprocket shafts of 9.0 ft. The conveyor is
made up of two 4-ft wide, belts. These belts are supported
on their edges and centerlines by a support frame which
also carries the sprocket shafts. The technical character-
istics of the harvesting conveyor are presented in Table 9.
39
-------
TABLE 7
Characteristics of Protected ¥ater
Sorbent-Oil Recovery System
Design Oil Recovery Rate 3000 gallons/hr
Slick Thickness 1.5 mm
Sweep Speed 1.6 ft/sec
Sweep Width (Normal) 16.0 ft
Sweep Width (Extension Booms) 32.0 ft
Active Length 24.0 ft
Sorbent Material Open cell reticulated
polyurethane foam
Sorbent Form Chips 3 in. X 3 in.
x 1/4 in. nominal
Sorbent Pore Size
Oil Viscosity < 1000 cps 80 PPI
Oil Viscosity > 1000 cps 30 PPI
Sorbent Residence Time 15 sec
Sorbent Coverage 80 percent
Oil Recovery Performance
1.5 mm Slick Thickness
Oil Viscosity Oil Recovery Rate Oil Recovery Water Recovery
cps GPH from Slick Rate
percent GPH
< 6000 3000 87 311
10000 2830 82 270
20000 2580 75 243
Oil Recovery for Viscosity < 6000 cps
Slick Thickness Normal Sweep Width With Extension Booms
MM (16 ft) GPH (32 ft) GPH
0.5 1000 2000
1.0 2000 4000
1.5 ' 3000 6000
40
-------
4=-
h-1
TRANSFER
CONVEYOR
HARVESTING CONVEYOR
38.5 DIESEL ENGINE
STEAM GENERATOR
32'-0"
BROADCASTING
NOZZLE
BROADCASTING FAN
SORBENT REGENERATOR
FIGURE 10 - PROTECTED WATER RECOVERY SYSTEM ARRANGEMENT DRAWING
-------
TABLE 8
Broadcasting System Characteristics
Sorbent Form
Sorbent Broadcasting Rate
Regenerated Sorbent Density
Broadcasting Nozzle Type
Nozzle Rate
Nozzle Angle
Nozzle Drive
Broadcasting Duct Diameter
Broadcasting Duct Length
Air Speed
Air Flow Rate
Fan Type
Diameter (Wheel)
RPM
Fan Power (Total)
Transmission
Total Sorbent in System
Chips 3 in. x 3 in. x I/1!- in.
nominal
0.424 ft3/sec
330 chips/sec
6.0 lbs/ft3
Moving Parallel Plate
1 cycle/sec
±50 degrees
Air Cylinder
1.0 ft
24.0 ft
45 ft/sec
2062 ft3/min
American Blower Series
106
Industrial Fan-Size 17
Long Shavings Wheel
29-5/8 in.
1062
7.8 HP
Mechanical
14.9 ft3
11,450 chips
Sorbent Loss Rate
Wind Speed
mph
0-10
10 - 20
20 - 30
Loss Rate
ft3/hr
3.1
15.5
31
42
-------
The hydraulic drive system can provide an excess of 150
ft-lbs of torque at 47 rpm. This is sufficient to lift a
combined weight of chips and debris of about 400 Ibs. The
belt material and sprockets for the conveyor are off the
shelf items. The support frame will have to be fabricated.
TABLE 9 .
Harvesting Conveyor Characteristics
Conveyor Width
Conveyor Angle
Conveyor Length
Linear Belt Speed
Belt Material
Belt Drive
Power
Transmission
Conveyor Support
Total 8.0 ft - 2 sections
4.0 ft
45 degs.
9.0 ft
1.6 ft/sec
1 in. X 1 in. mesh extra
heavy duty (button head weld)
flat-wire belt galvanized
steel construction
16 Class C sprockets
8-1/4 in. diameter - 23 teeth
1.0 HP
Hydraulic
Integral support frame
The transfer conveyor collects the chips from the har-
vesting conveyor and transports them to the sorbent re-
generator. This conveyor has a belt width of 2.0 ft and a
belt speed of 2 ft/sec. The requirements for this conveyor
can be satisfied with commercial units which are available
off the shelf. The detailed technical characteristics of
the transfer conveyor are presented in Table 10.
-------
TABLE 10
Transfer Conveyor-Characteristics
Conveyor Belt Width
Conveyor Length
Linear Belt Speed
Belt Material
Conveyor Support
Drive
Power
Transmission
2.0 ft
11.5 ft
2 ft/sec
3 ply stitched canvas
Neoprene covered
Troughed bed* with integral
take up and driver pulley
8 in. pulley-friction drive
1/2 HP
Hydraulic
* Hytrol Conveyor Co. Model "TR" Horizontal Belt
Conveyor or Equivalent.
Sorbent Regenerator
The preliminary design of the sorbent regenerator was based
"on the development work and data presented in Appendix D.
The characteristics of the resulting unit are presented in
Table 11. Figure 11 presents an arrangement drawing for
the unit. The various elements are indicated in the figure,
The design closely follows that of the sorbent regeneration
test apparatus. The most important differences are the ad-
dition of a third pair of squeezing rollers and the use of
a reinforced neoprene covered belt in place of the upper
wire-mesh belt with pads. Other mechanical improvements
suggested by the development program and detailed in Ap-
.pendix D.have been included. In .addition^ the collecting
pan between the legs of the lower belt has been deepened
and provisions made for heating coils. The development
tests showed that cold bunker "C" (lO°C 20,000 cps) could
be squeezed from the sorbent and through the lower belt.
However., for satisfactory flow in the collecting pans the
recovered oil must be heated. This will require about
1000 pounds of steam per hour. This steam can be supplied
by a commercial steam cleaning unit.
-------
TABLE 11
Sorbent Regenerator Characteristics
Design Sorbent Rate
Belt Speed
Maximum Squeezing Force
Number of Squeezing Stations
Squeezing Roller Width
Squeezing Roller Diameter
Lower Belt Material
Upper Belt Material
Power
Transmission
Air Supply
Squeezing Force Generation
Squeezing Roller Travel
Heating
0.43 ft3/sec
1.0 ft/sec
220 Ib/in. of belt width
3
24 in.
6 in.
Steel Balanced Belting
Type B-60-32-14
Reinforced Neoprene
2 HP
Hydraulic Drive
Engine Driven Compressor
Air Cylinder
4 in.
Heating coils in collecting
pan 1000 Ib/hr steam
Sorbent Regeneration Performance
Oil Viscosity Sorbent Squeezing Force Regenerated Sorbent
Density
cps
2
300
1000
4500
20, 000
PPI
80
80
30
30
30
Ib/in
220
220
220
220
220
lb/ft=
4.5
5-5
5-0
5.5
7-5
-------
LOWER BELT
STEEL WIRE MESH
UPPER BELT
NEOPRENE
_ . ,-, HYDRAULIC MOTOR
yf^> i s~^\^
AIR ACCUMULATOR.
SQUEEZE ROLLER
FRAME ( 3 )
BELT TENSION ROLLER
OIL DISCHARGE PIPE ( 2 )
ACCUMULATOR
AIR CYLINDER
PPER BELT
LOWER BELT
BRUSH
STEAM LINES FOR
OIL HEATER
FIGURE 11 - SORBENT REGENERATOR ARRANGEMENT DRAWING
-------
Although the sorbent regenerator is the most complex mechani-
cal element in the recovery system and is not a standard item,,
many of its components are commercially available. These in-
clude belts, air cylinder,, shafts and bearings. The con-
struction of the regenerator requires machine shop and light
steel fabrication capabilities. These capabilities are
readily available. !
Miscellaneous Equipment
In addition to the main components of the recovery system,
certain additional equipment are required to complete the
recovery system. These include an oil transfer system to
pump the recovered oil to storage, and a heating system for
the sorbent regenerator and tank heating.
The oil transfer system will use a centrifugal pump driven
mechanically off of the broadcasting system fan .drive. The
pump selected is a centrifugal pump, MP Model 130 or equal,
bronze body, 188 gpm at 15 feet of head.
The heating of the tanks and sorbent regenerator, if-re-
quired, is best provided by steam. Also steam may be re-
quired-: for cleaning. The most economical way of providing
steam for these purposes is with a commercial steam cleaner
unit. A suitable unit would be a Slifer Model 90 or equal
with a capacity of 60 gph at 120 psi.
Power and Transmissions
The various components of the recovery system require a
rugged and reliable power supply. For this purpose an in-
dependent power system"consisting of a diesel engine with
a hydraulic or combination hydraulic and mechanical trans-
mission system was selected. The engine is a 38.5 HP
radiator cooled Model 5034-7201 Detroit Diesel or equal.
The radiator cooling makes the recovery unit independent of
a supply of uncontaminated water for cooling and allows op-
eration in cold weather by the addition of antifreeze.
The recovery system components consist of two types with
respect to power transmission. The broadcasting system fan,
the oil transfer pump, and the bilge and ballast pump (which
is not part of the recovery system) require moderate powers
at moderate rpm's. The harvesting and transfer conveyors
and the sorbent regenerator require low power at low rpm's.
-------
In both cases some independent speed control capability is
desirable. As a result, it is desirable to have two separate
power transmissions. The fan and pumps system could be
either a direct mechanical or a hydraulic system. Diagrams
for the mechanical and the hydraulic transmission systems are
given in Figures 12 and 13, respectively. The mechanical
transmission is less costly but restricts the arrangement of
components. The hydraulic transmission is more costly but
allows a much greater freedom with qomponent arrangement. Both
systems consist completely of commercial components which have
a history of satisfactory operation. The hydraulic lines
would be fitted with quick disconnect fittings which will al-
low rapid assembly and disassembly of the unit. In this case,
a satisfactory "arrangement can be made with the mechanical
system so it is recommended because of the lower cost.
The transmission system for the sorbent regenerator and con-
veyor should be hydraulic. Hydraulic motors are available
which are well suited to the required low rpm, high torque
operation. Also a hydraulic system allows some speed adjust-
ment and is much more easily arranged than a mechanical sys-
tem. Figure 14 presents the flow diagram for the sorbent
regenerator and conveyor hydraulic transmission. Again the
system consists of commercial components with a history of
satisfactory operation.
RECOVERY PLATFORM
The recovery platform is based on a catamaran concept with
asymmetrical hulls configured to confine and herd the sor-
bent material. The harvesting conveyor is mounted between
the hulls. Figure 10 presents an overall arrangement of the
recovery system equipment and platform. The basic charac-
teristics of the recovery platform are presented in Table 12.
Hull
The platform structure is composed of two asymmetrical hulls
joined by two tubes. The tubes are fastened to the hulls
with stainless steel bolts in such a manner that they can be
disassembled for shipment. The hulls are each fitted with
five watertight compartments for machinery, sorbent storage,
recovered oil, and ballast. Figure 15 shows the internal
arrangement of the hulls.
-------
BILGE PUMP-
11
REDUCER
XA
OIL TRANSFER PUMP
TO FAN
•38.5 HP DIESEL
FIGURE 12 - DIAGRAM OF FAN AND PUMP MECHANICAL TRANSMISSION
FOR 3,000 GPH SYSTEM
-------
JYRONE 20200 PUMP
30GPM@2000PSI
DELTA FLOW DIVIDER
WITH SPECIAL 2:1 RATIO
BLOWER FAN
DOUBLE "A" MH39
, MOTOR
510Lb.Jn.@
1240 RPM
DIESEL ENGINE
38*H.P.
OIL TRANSFER
DOUBLE "A" MH-10
MOTOR
83 Lb. ln.@
3000 RPM
BILGE PUMP
DOUBLE "A" MH-10
MOTOR
83 Lb. ln.@
3000RPM
FIGURE 13 - SCHEMATIC DIAGRAM OF FAN AND PUMP HYDRAULIC TRANSMISSION
FOR 3000 GPH SYSTEM
50
-------
TYRONE 20150 PUMP
20GPM@2500PSI
DIESEL ENGINE
•-Q
L-H
SORBENI
REGENERATOR
STAFFA B-30
. MOTOR
400 Lb. Ff.@
38-125 RPM
HARVESTER
STAFFA B-10
MOTOR
142 Lb. Ft.@
47-100 RPM
TRANSFER
STRAFFA B-10
MOTOR
36 Lb. Ft. @
150-250 RPM
FIGURE 14 - SCHEMATIC DIAGRAM OF SORBENT REGENERATOR AND CONVEYOR
HYDRAULIC TRANSMISSION FOR 3000 GPH SYSTEM
51
-------
TABLE 12
Platform Characteristics
Platform Concept
Overall Length
Overall Beam (Normal)
Overall Beam (Extension Booms)
Draft (Maximum)
Hull Beam
Hull Depth
Hull Structure
Light Ship Weight
Service Loads
Recovered Fluid (3000 gallons)
Full Load Condition
Deployment Draft
Full Load Draft
Propulsion 2
Deployment Speed
Full Load Speed
On Board Chip Storage
Catamaran
42.5 ft
16.5 ft
32.5 ft
3.0 ft
4.25 ft
5-5 ft
Aluminum
17,100 Ibs
2,190 Ibs
22,610 Ibs
41,900 Ibs
1.33 ft
2.83 ft
75 HP Diesel I/O Drives
10.6 knots
8.4 knots
240 ft3
Item
Length
ft
2 hulls 42.5
8.0
2 cross
beams
1 deck
house 11.5
Shipping Characteristics
Width Depth Estimated Weight
ft ft Ibs
4.5 8.0 6000 max.
2.0 1.25 500
5.0 7.5 1200
52
-------
QAWT
SORBENT BALLAST
STORAGE
MACHINERY/ SORBENT
SPACE /VSTORAGE
FIGURE 15 - INTERNAL ARRANGEMENT OF PLATFORM HULLS
53
-------
The hull structure can be either aluminum or steel. Aluminum
has the advantages of lower weight for greater onboard oil
storage and shipping, and lower maintenance costs. A steel
hull will have a lower initial cost and higher weight.
Aluminum construction including hulls, cross-beams, and the
deck house will have a weight of 8600 Ibs and a cost of
&25,200 (each of 3). The comparable items in steel will have
a weight of 20,000 Ibs and a cost,, with inorganic zinc coating,
of $17,550 (each of 3). The characteristics presented in
Table 12 are based on aluminum construction. Figure 16 pre-
sents a typical midship section for aluminum construction.
As may be noted, the construction details are very simple so
that the structure can be fabricated by any facility used to
working with aluminum.
Propulsion
It has been assumed that the protected water sorbent-oil
recovery system will be regularly used for cleaning up
routine harbor spills. As a result, it is desirable to
provide some form of self-propulsion rather than to use
another vessel for towing or pushing. The speed-power
curves for the recovery system at its full load and deploy-
ment draft are presented in Figure 17. The specific design
goal of a deployment speed of -12 knots would require a total
of about 240 HP. This power level exceeds the power avail-
able from lightweight propulsion systems. A reduction in
speed to 10 knots reduces1 the required power to 125 HP which
is within the practical rang.e. The available lightweight
and low-to-moderate cost propulsion systems include gasoline
outboard motors or diesel inboard outdrives (l/0's). In-
board gasoline engines could provide higher power levels
but could not satisfy U. S. Coast Guard Regulations for
this type of craft.
The advantages of gasoline outboard motors include light-
weight, low first cost, and good reliability. Their dis-
advantages are high operating cost and short life. For
this application, two mercury Model 650 AC outboard or
equivalent at about 80 HP each would be satisfactory. The
advantages of diesel I/O's include low operating cost, good
reliability and long life. However, they are heavier
(about 1200 Ibs) and more expensive (about $4700) than
gasoline outboards. For this application two Volvo-Penta
54
-------
BULKHEADS 0.160 " R. w /3" X 3/16 " FRAMES ON 12" CENTERS
SIDE 3/16" II
DECK 3/16 " 9. AFT OF BREAK, 0.160 " FWD
3X 3X 3/16" ANGLE
LONG. 6X 3/8" FB
ALL FRAMES WITH 3/16 "I
3"X 1/4" F.B. AND WITH
0.16013"X 3/16 " F.B.,
ALL ON 12" CENTERS
SIDE 3/16" d
BUMPER SPLIT 4" OD
PIPE W/0.318 "WALL
OUTBOARD SIDE ONLY
BOTTOM 3/16" I
STARBOARD HULL LOOKING FORWARD
8'-0" FORWARD OF TRANSOM
ALL WELDED 5086 - HI 11 AND H 32 ALUMINUM
FIGURE 16 -MIDSHIP SECTION FOR ALUMINUM CONSTRUCTION
-------
10 12
SPEED ( Knots )
14
16
FIGURE 17 - SPEED POWER CURVES FOR PROTECTED WATER
OIL RECOVERY UNIT
-------
Inc. Model AQ D32A/270D or equivalent dlesel outdrives at
about 75 HP each would be satisfactory. If the recovery
system is likely to be used on a dally basis it is recom-
mended that the diesel outdrives be fitted. The arrangement
drawing presented in Figure 11 is based on the use of diesel
outdrives.
Outfit
In order to make the recovery platform an operational system,
a certain amount of outfit is required. The typical outfit
items are listed in Table 13.
TABLE 13
Outfit Items
Railings
Hatches
Anchor and Lines
Fenders
Fire Extinguishers
Lights
Lockers
Cleats
Piping
Chemical Toilet
Compass
Radio
The extent of the outfit depends on the service the re-
covery platform will see and the requirements of the par-
ticular owner.
WEIGHT ESTIMATES
A detailed weight estimate was prepared and is presented in
Table l1^. This weight estimate is based on aluminum con-
struction and propulsion with diesel outdrives. If steel
structure is used instead of aluminum, the total structural
weight would increase to about 20,000 Ibs. This would re-
duce the amount of recovered oil which could be carried at
a draft of 2.75 ft to about 1350 gallons.
57
-------
TABLE 14
Weight Estimate for 3COO .GPH Protected Water Sorbent-Oil Recovery System
CALCULATION OF WEIGHT
VJ1
co
1 of 4
*" Protected Water Unit
ITEM Structure
Bottom
Deck
Sides'
Transom
Bulkheads
. . . . . I-or.-itu Jinal Web .
.. Rub Fte i 1 . :
Fv*. Peam . . .
Aft. Beam
Ho.use . .
Gratin?
Miscellaneous . .
Total Structure
Mechanical Systems
Saueezer
Fan
Harvesting Convevor
Broadcast Duct
Hydraulic and Mechanical Drive
Diesel
Total Mechanical Systems
1"
(EIGHT III
Ibs
829
756
2945
126
- 33°.
42-=.
. 1Q4
. .4 1C. ..
218
840 '
465
1000
8600
700
4?5
SOD
320
1150
1200
4335
ABOVE MLO BASF LINE
C.G.
VERT. MOM.
-
REFER.CO TO, Transom
C.G.
rm MCMCNTS
l6o, 2CO
48,205
OM^IU
c.e.
«
AFTER MOHEHTS
wm tioia
-------
TABLE 14 (Continued)
CALCULATION OF WEIGHT
2 of 4
v_n
VD
SHI* Harbor Recovery Unit
ITEM Outfit
Railings
Hatches
Anchor
Fencer?
Pire Extinguishers
Lights
Lockers
Cleats . •
Piolng '.
Chemical Toilet
Comoass
Radio
Miscellaneous
Total Outfit
Propulsion Machinery
2 Diesel I/0's with Dover tilt,
Steering Battery, fuel tank
Mn l WWIM
HEIGHT IN
Ibs
1?C
SCO
125
150
?CO
50
50
50
4^0
SO
10
30
140
1585
1620
JMOVE MLO B*SF LINE
c.e.
VKT. MOM.
ffiFcimcD TO; Transom
c.e.
FHO MOMENTS
31539
3240
CHf^m
c.e.
AFTE* MOMENTS
•01H SIMS
-------
TABLE 14 (Continued)
CALCULATION OF WEIGHT
of
O\
O
9I" Harbor Recoverv Unit
ITtM Service load
Fuel-Diesel
Personnel and Effects
SnrhPnt Chips
Fresh Vfater
Total Service Load
Cars;o - ^COO Gallons at 7.5 IbAal
Recovered Oil
'
^ i^ — — ^ •"— — «^ —
•nn 1
HEIGHT IN
Ibs
600
500
600
160
1Q60
22610
ABOVE MLO BASF LINE
C.G.
VERT. MOM.
REFERRED TO. Transom
C.6.
rm MOMENTS
26000
358000
•^^^••M^BtmvMVM
&.
C.G.
AFTER MOMENTS
*«w*v»«4iwH^viv^"viWb»
wlH VlB^S
-------
TABLE 14 (Concluded)
CALCULATION OF WEIGHT
of
»"• Harbor Recovery Unit
ITE* Summary
Structure
K;-'Ch~p.ical SVP terns
Outfit
Pro--".;! o •' on Kn cM no''"
Margin
,.
Light Shlo
Service L-ia-'?
.Service Condition (Draft - l.?5 ft)
Recovered Oil "COO era] Ions
Full Load Condition (Draft = ? .^ ft)
WEIGHT IN
Ibs
S6ro
•U3S5
1*35
i fipo
71^
16Q55
lo'-o
18915
??^10
^1525 '
ABOVE MU> BASF LINE
C.G.
venr. MOM.
BErtRREo TO: Transom
C.G.
15. f
15..2*
15..'5j
15.6
no MOMCMTS
26^ , c cc
2"6,CCC
2?2, COO
^50,000
650., ooo
C.6.
*f TEd HOMCNTS
•OIK SIBU -
-------
CREW AND OPERATING PROCEDURES
The protected water sorbent-oil recovery system should be
operated by a three-man crew consisting of a captain,
engineer and deck hand. The duties of these crew members '>
are listed in Table 15. Although the sorbent broadcasting,,
recovery, regeneration cycle is an automatic one, some as-
sistance will be required from the crew. The most important
is the removal of large pieces of debris and driftwood picked
up by the harvesting conveyor. These items would be removed
from the transfer conveyor or the inlet of the broadcasting
fan. Also the crew must introduce the sorbent material at
the start of the operation or makeup sorbent during the op-
eration. The sorbent material will be pre-cut and stored
aboard in bags. It is introduced into the system by dumping
it on the transfer conveyor with all components running. If
the slick is thin or discontinuous, the sorbent should be
prewetted with the diesel oil carried onboard. This will
limit the amount of water recovered. At the end of the op-
eration, the sorbent material is recovered by stopping the
broadcasting system and collecting the sorbent in bags as
it comes out of the regenerator.
On arrival at the scene of a spill, the sweeping booms are
extended, if required, and the sorbent material introduced
to the system. The recovery platform is then driven into
the slick and continues sweeping at the selected speed until
the slick is recovered. Since the operation is independent
o-f platform draft, no ballasting is required. If the slick
is In a restricted area, such as between a ship and pier,
a different operating procedure will be required. The re-
covery platform would be positioned in the slick and stopped.
The sorbent material would then be herded back to the
harvesting conveyor by the crew using rakes. Operating in
this manner, the recovery rate would be much less than the
design rate.
When the oil tanks are full, the oil-transfer system would
be used to pump out onto an accompanying barge or at a
shore-side facility. At the conclusion of the recovery op-
eration, the sorbent material is recovered for reuse and the
equipment cleaned and repaired as required. Much of the
sorbent that is lost during the recovery operation may be
expected to remain in the area of the spill. This sorbent
62
-------
TABLE 15
Duties of Recovery System Crew
Captain
Direct the oil recovery operation
Control the speed and location of the recovery
platform
Duty Station: Pilot House
Engineer
Control Recovery Machinery
Repair and Maintain Machinery
Operate Oil Transfer and Ballast System
Assist Deck Hand
Duty Station: Recovery Machinery Controls, Port
Side Aft.
Deck Hand
Introduce Sorbent Material into System
- Remove Larger Pieces of Debris from Transfer
Conveyor
- Repack Sorbent Material
Handle Lines
Duty Station: Grating aft of transfer conveyor.
63
-------
can be recovered by resweeping the area with the broadcasting
system off. This will require an expenditure of personnel
time but may be required and could be economically desirable
if the sorbent losses are large (high winds). The resweeping
operations could be carried out at a higher speed than the
basic recovery operation.
COST ANALYSIS
A detailed analysis was made of the construction and op-
erating costs of the 3000 GPH protected water recovery sys-
tem. The following assumptions were used in this analysis:
- The costs apply only to "following systems" after a
prototype system has been proven and a complete cor-
rected set of specifications and plans are available.
The estimates are based on construction in flights
of 1 and 3 units.
The builder is to be experienced with all types of
fabrications and machinery installation required
and the facility is to be equipped to handle this
size job. Experience is required in steel or
aluminum fabrication, diesel-hydraulic machinery,
and workboat outfitting to USCG regulations.
Only commercial quality construction is required
without extensive reference to government specifi-
cations or special inspection.
Costs are based on Spring 1972 dollars.
The construction cost estimates were made by summing the
total cost of purchased components and the value of the man
hours needed to assemble these components. A man-hour was
valued at $12.00 including overhead and profit. In the
case of the platform structure, budget prices were obtained
from fabricators based on preliminary drawings. The sorbent
material has a basic cost of $6.60 per ftsin bulk form. The
material can be cut into chips with a band saw at the rate
of about 4000 chips/hr. This gives a total cost of $9.20
per ft3 cut into chips.
64
-------
The construction cost will depend on the type of structure,
the type of propulsion, and the extent of the outfit. The
selection of these items depends on the intended duty, and
the importance of initial cost. The construction cost data
are presented in Table 16 for the various system elements.
Various possible cost options are presented in Table 17
ranging from minimum initial cost to the complete system
fitted for Independent operation. The arrangement drawings,
characteristics tables and weight estimates presented above
are for the complete system fitted for independent operation.
The recovery machinery is common to all users and is the
heart of the system. As a result, a detailed analysis of
the recovery machinery construction cost estimate is given
in Table 18.
The operating costs will depend on how and by whom the re-
covery system is operated. The direct operating costs were
estimated on a per hour of effective oil recovery and per
hour of deployment time. The results are presented in
'Table 19 and are based on a crew cost of $-10. per man hour.
Costs are given based on a 10 to 20 MPH wind and no attempt
to recover sorbent and on a resweep of the slick area which
recovers 70 percent of the lost sorbent. If the recovery
operation were carried out by a spill control contractor a
charge would be made to cover the capital cost of the equip-
ment. Typical rates currently charged are 0.0025 times
initial cost per hour. This would amount to between $100
and $175 per hour depending on the unit used.
It is of interest to note that even when 70 percent of the
sorbent is recovered, about 65 percent of the direct op-
erating cost is associated with the recovery and replacement
of lost sorbent. This could pay for a considerable increase
in platform cost to reduce the sorbent loss rate.
-------
TABLE 16
3000 GPH Protected Water Recovery System
Cost Analysis
Summary
Harbor Recovery Unit
Recovery Unit -
Installed and Op-
erating:
Combined Mechanical
and Hydraulic Drive
Full Hydraulic Drive
Hull -
Open Operating Station:
Steel
Steel with inorganic
zinc coating
Aluminum
Enclosed Operating
Station:
Steel
Steel with inorganic
«inc coating
Aluminum
Outfit -
Minimum - Steel
- Aluminum
With enclosed house :
- Steel
- Aluminum
With enclosed house
and self propulsion
- Steel
- Aluminum
Independent Operation
- Steel
- Aluminum
Propulsion
Twin Diesel I/O
Twin Outboards
Supply of Sorbent
Material (?'m ft)
Dollar Total for 1
Material
1^090,
1^050.
13000.
17000.
2';500.
15000. .
19500.
28000.
?590.
3030.
5850.
7580.
676O.
8-50.
73': o.
9900.
7330.
r:870.
2?10.
MH
668.
6M .
-
-
-
-
-
-
2': 8.
2':8.
306.
306.
358.
358.
'400.
•'•00.
SO.
50.
-
Total
22110.
22780.
13000.
'
17000.
2^500.
15000.
19500.
?8ooo.
5570.
6010.
9520.
11P50.
11060.
12950.
126'.0.
1^700.
8790.
3JI70.
2210.
Dollar Total ea of 3
Material
12^00.
132^0.
11700.
15300.
22050.
13500.
17550.
25200.
?280.
2270.
5150.
6670.
5950.
7610.
6890.
8700.
7050.
2730.
2210.
MH
587.
567.
-
-
-
-
-
-
218.
218.
269.
269.
315.
315.
352.
352.
72.
48.
-
Total
19^50
200^0.
11700.
15300.
22050.
13500.
17550.
25200.
iigoo.
5290.
8380.
9900.
9730.
11390.
11110.
12920.
7910.
3310.
2210.
66
-------
TABLE 17
3000 GPH Protected Water Recovery System
Total Construction Costs
Minimum Initial Cost:
Recovery Unit -r Combined Me-
chanical and Hydraulic Drive
Uncoated steel hull
Minimum outfit (steel) w/o
sweep booms
Outboard motor propulsion
Total w/o sorbent material
Sorbent material
Total with sorbent material
Minimum Intermittent Duty:
Recovery unit -combined mech.
and hydraulic drive
Coated steel hull (open op-
erating station)
Outfit (open operating station)
Outboard motor propulsion
Total w/o sorbent material
Sorbent material
Total with sorbent material
Minimum Continuous Duty Unit:
Recovery unit-combined me-
chanically and hydraulic drive
Aluminum (open operating stat. )
Outfit (open operating station).
Twin Diesel Propulsion
Total w/o sorbent material
Sorbent material
Total with sorbent material
Independent Operating Unit:
Recovery unit - Full hydraulic
Drive
Aluminum Hull, enclosed house
Outfit - Independent operation
Twin Diesel Propulsion
Total w/o sorbent material
Sorbent material
Total with sorbent material
Dollar Total
for 1
22110.
13000.
5570.
3^70.
44150. >
2210.
46360.
22110.
i
17000.
5570.
3470.
48050.
2210.
so, 260.
2211C.
24500.
6010.
8790.
61410.
2210.
63620.
22780.
28000.
14700.
8790.
74270.
2210.
76480.
Dollar Total Each
of 3
19450.
11700.
4900.
3310.
39360.
2210.
41570.
19450.
15300.
4900.
3310.
42960.
2210.
45170.
19450.
22050.
5290.
7910.
54700.
2210.
56910.
20040.
25200.
12920.
7910.
66070.
2210.
68280.
6?
-------
TABLE 18
Detailed Analysis of Protected Water Unit
Recovery Machinery Cost
Oil Recovery Machinery -
'Cost of materials and fabrication of components in-
cluding hydraulic system hook-up)
Squeezer
Fan
Harvesting conveyor
Broadcast duct
Broadcast nozzle
Transfer conveyor
Hydraulic drive
Mechanical drive
Diesel
Transfer pump
Total for 1
Total each for 3 @ .
Additional for full
hydraulic drive
Total for 1
Total each of 3 ® .8
&3000.
550.
1400.
200.
50.
800.
5360.
500.
2100.
130.
14090.
12'-! 00.
960.
15050.
13240.
210 MH
50
16
8
48
40
380
334
+24
404
356
fabricated
purchase
complete
fabricated
purchased
fabricated
purchased
purchased
components and
hook-up required
purchased
components and
hook-up required
purchased
complete
purchased
Install and trouble-shoot recovery system (hydraulic and Mech)
288
253 (full hydraulic
system)
240
211
Total for 1
Total for each of 3
Total for 1
Total each of 3
68
-------
TABLE 19
3000 GPH Protected Water Recovery System
Direct Operating Costs
Per Hour of Effective Oil Recovery:
No lost sorbent recovery1
Crew $30.00
Maintenance 10.00
Sorbent 142.50 (10-20 MPH
wind)
Fuel and Miscellaneous 10.00
Total $192.50
70 Percent Lost Sorbent Recovery on Second Sweep:2
Crew 60.00
Maintenance 20.00
Sorbent 42.75
Fuel and Miscellaneous '15-00
Total $137-75
Per Hour of Deployment:
Crew 30.00
Maintenance 4.00
Fuel and Miscellaneous 7-50
Total $41.50
1 No attempt is made to recover the sorbent material lost
due to wind.
2 A second sweep of the slick area is made with the broad-
casting system off to recover sorbent material. It is
assumed that 70 percent of the sorbent which was lost
is recovered by this procedure.
69
-------
SECTION VII
10,000 GALLON PER HOUR RECOVERY SYSTEM PRELIMINARY DESIGN
A preliminary design of a sorbent-oil recovery system for
use in unprotected waters was prepared based on the general
and specific design goals provided by the EPA. These specific
design goals are given in Table 20.
TABLE 20
Specific Design Goals for Unprotected Water System
Unprotected Waters
a. Environment
Perform effectively in waters with up to 5-foot
seas in combination with 30-mph winds and 2-knot
currents,, close to shore in less than 80-foot
depths.
b. Sorbent Recovery System
The sorbent recovery system in combination with
any necessary oil-water separation facilities shall
have the capacity to recover at a minimum rate of
10., 000 gal/hr of oil with 10 percent or less water
content.
(l) The water content objective may be altered if
this is consistent with the method of ultimate
disposal of the oil.
(2) Oil-sorbent separation may be accomplished
either as an inherent characteristic of the
recovery device or an individual unit process.
(3) The device shall be capable of meeting the
recovery rate objective while recovering oils
with viscosities ranging from that of light
diesel oil to Bunker C oil at 20 C.
(4) The thickness of the slick to be recovered at
the specified rate shall be 1.5 mm or less.
71
-------
TABLE 20 (Continued)
(5) The device may include as an integral part of
the skimming process,, a system of booms to
aid in herding oil toward the harvesting de-
vice.
(6) Design of units to be temporarily attached to
existing vessels must consider transport needs
and ease of installation with a minimum of
special equipment.
c. Oil-Sorbent Separation
The unit must be capable of operating under hy-
draulic loadings and ranges of oil concentrations
consistent with the performance of the oil har-
vesting device. It must be capable of producing
output streams with the following characteristics:
(l) Oil - 10 percent water or less.
(2) Sorbent material - available for reuse or
method for disposal.
(3) Water - 10 mg/1 oil or less.
d. Oil Storage or Disposal
Specify floating or other appropriate land-based
facilities which., in combination with on-board
storage, will have capacity to store material to
be collected from spills of at least one million
gallons. Agglomerates must be processed before
land disposal to preclude leaching. Other ap-
propriate disposal techniques such as incineration
will be considered as alternatives.
72
-------
TABLE 20 (Concluded)
e. Vessel
The recovery system must be designed for either per-
manent or temporary mounting on appropriate existing
vessels. The development of vessels designed specifi-
cally for oil recovery use will not be considered.
Vessels to be used must be capable of speeds of at
least 12 knots under the environmental conditions
listed above (however, not performing the oil re-
covery function). They must accommodate any required
oil-sorbent separation equipment, the oil recovery
device and, operating personnel, 24-hour fuel supply
and all oil specified for storage on board. They
must be capable of speeds of at least 8 knots (how-
ever not performing an oil recovery function) in
10-foot seas and winds of up to 38 mph.
The system was designed to satisfy the required recovery rate
of 10,000 GPH of bunker "C" at 20 C. The rationale for the
selection of the system operating parameters such as sweep
speed, sorbent coverage, and sorbent residence time are pre-
sented in Section V. Many elements of the unprotected water
system are similar to those on the protected water system.
These are not presented in detail again so that reference
should be made to the appropriate parts of Section VI.
GENERAL DESCRIPTION
The basic sorbent recovery system consists of a pneumatic
broadcasting system with a moving parallel-plate nozzle, a
harvesting conveyor, a transfer conveyor, a sorbent re-
generator, the necessary drive systems and a catamaran type
platform. The basic characteristics of the unprotected water
(10,000 GPH) Sorbent-Oil recovery system are presented in
Table 21. Figure 18 is an overall arrangement drawing of
this system. The recovery equipment is mounted on a specifi-
cally designed platform rather than a vessel of opportunity.
The reasons for this form of mounting are presented in de-
tail in Section V. Basically they are improved operating ef-
ficiency, reduced reaction time, and potentially lower life-
cycle cost.
73
-------
TABLE 21
Characteristics of Unprotected Water
Sorbent-Oil Recovered System
Design Oil Recovery Rate 10,000 gal/hr
Slick Thickness 1.5 mm
Sweep Speed 2.60 ft/sec
Sweep Width (Normal) 32.0 ft
Sweep Width (Extension Booms) 64.0 ft
Active Length 65.0 ft
Sorbent Material Open cell polyurethane foam
Sorbent Form Chips 3 in.X 3 in. xl/4 in.
normal
Sorbent Pore Size
Oil Viscosity < 1000 cps 80 PPI
Oil Viscosity > 1000 cps 30 PPI
Sorbent Residence Time 25 sec
Sorbent Coverage 75 percent
Oil Recovery Performance
1.5 mm Slick Thickness
Oil Viscosity Oil Recovery Rate Oil Recovery Water Recovery
cps GPH from Slick Rate
percent GPH
< 6000 10,000 90 962
10000 9,^50 85 844
20000 8,600 77.5 752
Oil Recovery for Viscosity < 6000 cps
Slick Thickness Normal Sweep Width With Extension Booms
mm (16 ft) GPH (64 ft) GPH
0.5 33°0 6700
i.o 6700 13400
1.5 10000 20000
-------
Ul
STEAM GENERATOR
PUSHING \ CONTROL STATION
BARGE \ / 75 HORSEPOWER DIESEL
BROADCAST
NOZZLE
\ HARVESTING CONVEYOR
BROADCAST FAN
SORBENT REGENERATOR
TRANSFER
CONVEYOR
EXTENSION
BOOM
AFT PONTOON
FWD PONTOON
FIGURE 18 - UNPROTECTED WATER RECOVERY SYSTEM ARRANGEMENT DRAWING
-------
RECOVERY SYSTEM COMPONENTS; BROADCASTING SYSTEM
The broadcasting system uses larger components but Is other-
wise similar in concept to the system on the protected water
unit. Its technical and performance, characteristics are pre-
sented in Table 22.
TABLE 22
Broadcasting System Characteristics
Sorbent Form
Sorbent Broadcasting Rate
Regenerated Sorbent Density
Broadcasting Nozzle Type
Nozzle Rate
Nozzle Angle
Nozzle Drive
Broadcasting Duct Diameter
Broadcasting Duct Length
Air Speed
Pan Type
Diameter (Wheel)
RPM
Pan Power (Total)
Transmission
Total Sorbent In System
Chips 3 in. X 3 In.x 1/4 in,
nominal
1.3 ft3/sec
1010 chips/sec
6.0 lbs/ft3
Moving Parallel Plate
0.. 5 cycle/sec
±50 degrees
Air Cylinder
21 In.
48.0 ft
45 ft/sec
American Blower Series 106
Industrial Fan Size 21
Long Shavings Wheel
36.5 in.
1099
36 HP
Mechanical
56 ft3
44,000 Chips
Sorbent Loss Rate
Wind Speed
mph
0-10
10 - 20
20 - 30
Loss Rate
ft3/hr
9.4
47.2
93.7
76
-------
Harvesting and Transfer Conveyors
The harvesting conveyor is mounted between the aft pontoons
of the recovery platform. The overall width is 16 ft which
is made up of four 4-ft
-------
TABLE 24
Transfer Conveyor Characteristics
Conveyor Belt Width
Conveyor Length
Linear Belt Speed
Belt Material
Conveyor Support
Drive
Power
Transmission
3.0 ft
20.0 ft
2.5 ft/sec
3-Ply Stitched canvas
Neoprene covered
Troughed Bed* with integral
takeup and drive pulley
8 in. diameter pulley-friction
drive
1 HP
Hydraulic
* Hytrol Conveyor Company Model "TR" Horizontal Belt
Conveyor or Equivalent
Sorbent Regenerator
The operating principles and basic configuration of the
sorbent regenerator for the unprotected water unit are
identical to the protected water unit. These items are
discussed in detail in Section VI. The belt width for the
unprotected water unit regenerator is 60 inches compared
with 24 inches for the protected water unit. To limit the
increase in belt width in the unprotected water unit, the
unit sorbent loading was increased from 0.22 ft3/ft2 to
0.3 ft3/ft2. This slightly increases the density of the
regenerated sorbent but it remains at an acceptable value.
The development program reported in Appendix D covered
values in excess of 0.3 ft3/ft2 unit sorbent loading with-
out difficulties. Table 25 presents the characteristics
of the resulting unit.
78
-------
TABLE 25
Sorbent Regenerator Characteristics
Design Sorbent Rate
Belt Speed
Maximum Squeezing Force
Number of Squeezing Stations
Squeezing Roller Width
Squeezing Roller Diameter
•Lower Belt Material
Upper Belt Material
Power
•Transmission
-Air Supply
Squeezing Force
Squeezing Roller Travel
Beating
1,3 ft3/sec
1.0 ft/sec
220 Ib/in of belt-width
3
60 inches
6 inches
Steel Balanced Belting
Type B-60-32-4
Reinforced Neoprene
4 HP
Hydraulic Drive
Engine DrivenCompressor
Air Cylinder
4 in.
Heating Coils in
Collecting Pan
3000 Ib/hr steam
Sorbent Regeneration Performance
Oil Viscosity
cps
2
300
1000
4500
20000
Sorbent
PPI
80
80
30
30
30
Squeezing Force Regenerated Sorbent
Ib/in
220
220
220
220
220
Density - Ib/ft1
5.0
6.0
5.5
6.0
8.0
Miscellaneous Equipment
Additional components which should be included with the
recovery system are an oil transfer pump with a capacity of
500 GPM at 40 ft of head and a steam generator capable of
evaporating 360 ga.l/hr at 120 psi.
79
-------
Power and Transmissions
The recovery system components are powered by a diesel
engine driving through a hydraulic or combination hydraulic
and mechanical transmission system. The basic concept of
the power train is identical with that on the protected
water unit. For this case, the engine is a 75 HP radiator-
cooled Model 3-53N Detroit Diesel or equal. The broadcasting
system fan and the oil transfer and bilge pump can be driven
either by a direct mechanical or hydraulic system. The
mechanical and hydraulic transmissions for the 10,000 GPH
systems are shown schematically in Figures 19 and 20, re-
spectively. The mechanical system can be made to satisfy
the arrangement requirements and thus is recommended because
of its lower cost. The sorbent regenerator, the harvesting
conveyor and the transfer conveyor require a hydraulic trans-
mission which is shown schematically in Figure 21.
RECOVERY PLATFORM
The recovery platform is based on a catamaran concept with
hulls consisting of four pontoons joined by trusses. The
side trusses are screened to confine the sorbent material
and the aft pontoons are configured to herd the sorbent to
the harvesting conveyor. The harvesting conveyor is mounted
between the aft pontoons. Figure 18 presents an overall
arrangement of the recovery system equipment and platform.
The basic characteristics of the recovery platforms are
presented in Table 26.
Hull
The hull configuration of pontoons joined by trusses was
selected for the following reason:
Simpler to disassemble into units suitable for over
road transport,
- Lower weight than conventional catamarans hulls,
Satisfactory motions in the design sea conditions.
Appendix E presents the results of a 1/4-scale model test
of this platform concept. These tests covered, sorbent
broadcast and recovery in calm water and waves, motions in
waves, and towing stability and resistance.
80
-------
REDUCER
JJ-
BILGE PUMP
75 HP
'DIESEL
t
TO FAN
'OIL TRANSFER PUMP
FIGURE 19 -
DIAGRAM OF FAN AND PUMP MECHANICAL TRANSMISSION
FOR 10,000 GPH SYSTEM
81
-------
TYRONE 20450 PUMP
65GPM @ 750 PSI
BLOWER FAN
TYRONE M25550
MOTOR
TYRONE M20250
MOTOR
I
DELTA FLOW DIVIDER
WITH SPECIAL 2:1 RATIO
BILGE PUMP
TYRONE M2-45
MOTOR
FIGURE 20 - SCHEMATIC DIAGRAM OF FAN AND PUMP HYDRAULIC TRANSMISSION
FOR 10,000 GPH SYSTEM
82
-------
TYRONE 20150 PUMP
20GPM@2500PSI
DIESEL ENGINE
75 H.P.
I .
?
I
Hi-
SORBENT
REGENERATOR
STAFFA B-30
MOTOR
400 Lb. Ft.@
38-125 RPM
HARVESTER
STAFFA B-30
MOTOR
300 Lb. Ft.@
75-150 RPM
TRANSFER
STAFFA B-30
MOTOR
36 Lb. Ft.@
150-250 RPM
FIGURE 21 - SCHEMATIC DIAGRAM OF SORBENT REGENERATOR AND CONVEYOR
HYDRAULIC TRANSMISSION FOR 10,OOOGPH SYSTEM
83
-------
TABLE 26
Unprotected Water System Platform Characteristics
Platform Concept
Overall Length
Overall Beam (Normal)
Overall Beam (Extension Booms)
Draft (Maximum).
Hull Beam (Maximum)
Hull Depth
Hull Structure
Light Ship Weight
Service Loads
Recovered Fluid (10,000 gals)
Full Load Condition
Deployment Draft
Full Load Draft
Propulsion
Deployment Speed
Onboard Chip Storage
Catamaran
82.0 ft
32.67 ft
64.0 ft
4.5 ft
3.5 ft
8.5 ft
Aluminum
29,065 Ibs
2360 Ibs
75.000 Ibs
106,425 Ibs
1.25. ft
4.5 ft
10 knots
500 ft3
[tern
Shipping Characteristics
Length
ft
Width
ft
Height
ft
Weight
Ibs
Aft Pontoon 24.0 8.5
Fwd Pontoon 12.0 3*5
Fwd Beam 27.0 2.08
Side Beams 27.0 8.0
Aft Beams 16.0 7-0
8-5 3950 (each of 2)
8.5 1400 (each of 2)
2.0 1420
8.0 1250 (largest
of 4)
7.5 1200 Ibs
84
-------
The hull and truss structure can be either aluminum or steel
Aluminum has a lower weight for greater onboard oil storage
and easier shipping. A steel hull will have a lower initial
cost and higher weight. The characteristics presented in
Table 26 are based on aluminum construction. The recovery
platform can be fitted with extension booms to increase the
sweep width thus increasing oil recovery rate in thin slicks
Typical extension booms are shown in Figure 18 It is ex-'
pected that these booms could be used only in relatively
calm water.
Propulsion
The unprotected water recovery system is not self propelled.
It may be either towed or pushed by a vessel of opportunity'
or a barge. Figure 18 illustrates how the system can be
pushed by a barge. The recovered oil can then be pumped
directly into the barge.
Outfit
In order to make the-recovery platform an operational sys-,
tern., a certain amount of outfit is required. The typical
outfit items are listed in. Table 27.
TABLE 27
Unprotected Water System Outfit Items
Railings
Hatches
Anchors and Lines
Fenders
Fire Extinguishers
Lights
Lockers
Chocks and cleats
Piping
Radio
Chemical Toilet
Operators Stand
Miscellaneous
Weight Summary
A detailed weight estimate was prepared for the unprotected
water recovery system and is presented in Table 28. This
weight estimate is based on aluminum construction. If steel
structures were used,, the total structural weight would in-
crease to about 45,500 pounds. This would reduce the amount
of recovered oil which could be carried on a draft of 4.5 ft
to about 6400 gallons.
85
-------
TABLE 26
CALCULATION OF WEIGHT
Weight Estimate for 10,000 GPH Unprotected
Water Sor^-ent^OH Recovery System
CD
1 of
*"* UriDrotectel Water Recovery Unit
ITEM Structure - A'utninum
Aft Pontoons
r'.25 in. plate an-1
3 in. •.: 1 in. x 0.25 in. an:;le on 15 in.
centers .
Outroar-; Si-ea
Inr. oar;: Si::es
Transom
Bov: 'Plate
Deck an:! B •ttom
Bulk'neals
Lon.jitu "inal '.;evs
Fopvar ":
Pontoons
Outboard ~i:~es
Inboarii sides
Bo:.'
Aft Enri
Deck ano Bottom
Bulkhea-
Longituiinal v/ebs
•
•ait 1 mmm*
•EIGHT IN
counts
172S
1152
576
5]>
liiiiC
576
?83
792
643
^02
202
216
17^
105
&907
ABOVE M.D BASF LIME
C.6.
VERT. MOM.
mrEBRto TO; Transom
C.G.
12. C
5.c
C
20. C
10.1
LI. "
L6. 0
rs.5
r5.5
Bl
n
^6
76
76
••
m> MOMENTS
OltfRM
C.6.
•FTER MOMENTS
WIN SIMS
-------
CD
CALCULATION OF WEIGHT
TABLE 28 (Continued
2 of 5
*"* Ooen Water
ITEM
Forvard Beam
Side Beams
Upper Longitudinal Tubes
Lower Tubes
Vertical an3 Transvsrse Struts
, ' ? .-< * * 7 :
2>-;4x8
2 x 7 -r. .'4
Diagonals 2. x 5 x in
2 x 5 x 12
2x7x8
— __ . . — f — L ._ _ i __. . .. j. • _ _ — -- . - —
Aft Beam
3 x 16
4 x 5 x ^ x 7
" 6 x 6 x ^ x c
GratinR Aft If y 6
Sid^s. 2 x. 51 :: 5
Sub Total
Margin 20 percent
Total Structure
j™
•€IWT I*
ooun 3 s
118S
•519
n37
17'5
202 .
176
*H7
373
353
2l(.o
151
321
?SO
1377
15220
,3044
18254
ABOVE M.D BASF LINE
C.G.
VERT. MOM.
^W^^^^B^HWOB-IKWWM.-
RErEmo TO; Transom
C.G.
72
45. c
^5.^
47
^••riw^MHVfri
47
47
Ly
-7
7
.0
.0
.0
.0
3.C
4'. 5
4-.=S
rm P40MENT5
525324
630380
OUfRHI
C.6
ArTEKMOMtNTf
p^M*MH«VHM^H^^MM»a^*
W ttDU
-------
TABLE 28 (Continued)
00
oo
CALCULATION OF WEIGHT
3 of 5
SHIP Opfp. Water
ITEM fV. £,!,-„ I <-.- 1 q '
Hull Outfit fn i Krrvi n -' rv n •;
RalUn-v:
Hatclior,
Anchors an 1 Lino
Fv.Tr".c-r~
Firt K:> i < r-.^i'i "ber
Lights
Loclcer-R
Chocka arri Clrats
Misccllancouc
Piping
Radio
Chemical Toilet.
Operators St?jn'l.
SuV-Total
Margin ?0 prrcent
Total Hull Outfit
Mn vmm
WEIGHT IN
8000
3 '10
-AO
160
300
?co
50
50
75
100
?50
30
50
ii-^O
?3?5
Ji75
?800
ADOVC k«.D BASF LINE
C.G.
VERT. MW.
REFEBREO TO-. Transom
c.c.
3.5
;6
il5
T"
i-i
LO
50
'0
!0
?0
10
30
PC
?0
37.-.
37.^
FWO MOMENTS
63,coo
fi^OP
76C3C
Ollf'U
C.G.
ArHR MOMENTS
M1N SIOKS
-------
TABLE '-'8 (Continued)
CALCULATION OF WEIGHT
of 5
00
VD
s"" Open Wstrr
i™" Lori is
Diese] Fuel
Personnel fin-1 Kf c\ elF
Sorhmt
WTcfr
Torai
i
i
i
i
i
Mtt 1 opwrOJI
HEIGHT IN
PQ unrig
!! 00
';CO
1?CO
16C
03 -..-^
ABOVE M.O OAsr LINE
C.G.
VERT. MCM.
Hereunto to. Transom-
C.G.
10
10
10
10
10
FWO MOMENTS
;l 000
C>CCQ
1PCOO
l^CC
•p--^ro
.
O«E^|IM
C.G.
AFTER MOMENTS
•om SIDU
-------
TABLE 23 (Conclude:!}
O
CALCULATION OF WEIGHT
5 of 5
sx" Ooen Wati'r
ITtM Summar;/
Structure - A"1. ur.npuin
yr-ehanlcal Systems
Outfit1 - -.•••-
-
Ll^ht S'-i-ir)
Load's
Service Condition
Recovc-r'3 ."• Oil
Full Loa:i Draft - ;) ft x '^ in.
MR I COMPUTOA
•EIGMT IN
.pounds
KS?'-t
5, CCO
?',600
?PJ c6;i
?, V5o
31','i24
-
75/000
1 ^ri 'i-i'1
L w< ' . A ^.
ABOVE MtO BASF i. (NE
C.G.
VERT. MCM,
>
-
REFESHEO TO: fp ,, ^ ^ r- Q n
C.G.
; •" • 5-
-;i . 5
•;1Y./-
?..; -
"^ O
. ^.' • w
?c; 7-
'- */ • j
^.'75
17. :
FWD MOWCNTS
-.
o.=-C -3-;
65COC
Vf'p.Ap
i - ^ -• ^
77'<-:69 "
2 \-': C 0
7?^C(;S
.,1.06, 55,
L,ccli,3ic
O4Ct;ufl
C.G.
AFTER MOMCNTS
•
•om *iou
-------
CREW AND OPERATING PROCEDURES
The unprotected water sorbent-oil recovery system should be
operated by a four man crew consisting of a captain, engineer
and two deck hands. The duties of these crew members are
similar to those assigned to the crew of the protected water
system which are detailed in Section VI. The operating pro-
cedures for the unprotected water system are similar to those
for the protected water system. The major difference is
that the unprotected water system is not self propelled. It
can be either towed or pushed by a vessel of opportunity. A
particularly favorable procedure would be to have the re-
covery system pushed by a tank barge. The recovered oil
could then be pumped directly to the barge, as mentioned pre-
viously.
COST ANALYSIS
An analysis was made of the construction and operating costs
of the 10,000 GPH unprotected water recovery system. The
assumptions and methods used in this analysis are identical
to those used for the protected water system and are pre-
sented in detail in Section VI.
The construction cost will depend on the structural material
and the extent of the outfit. The selection of these items
depends on the intended duty and the importance of initial
cost. The construction cost data are thus presented in
Table 29 for the various system elements. Various possible
cost options are presented in Table 30 ranging from minimum
initial cost to the complete system. The recovery machinery
is common to all cases and is the heart of the system. As
a result, a detailed analysis of the recovery machinery con-
struction cost estimate is given in Table 31.
The direct operating costs were estimated on a per hour of
effective oil recovery basis and are presented in Table 32.
The charter rates for the tugboat and barge, if used, are
subject to wide variations depending on the exact situation.
-------
TABLE 29
Unprotected Water Recovery System
Cost Analysis
Summary
ftecovery Unit - In
Stalled and operating
Combined mechanical and
Hydraulic
full Hydraulic Drive
Htm - Steel
- Steel with inor-
ganic zinc
coating
- Aluminum
Olitfit - Minimum - Steel
- Aluminum
- Complete - Steel
- Aluminum
Supply of Sorbent
Material
.!,£ J . :l J .
Dollar Total for 1
Materials
20060.
224<">0.
20700.
26500.
49700.
3850.
"150.
12300.
17040.
4600.
MH
666.
670.
-
-
-
382
38?
448
4^8
-
Total
28060.
30500.
20700.
26500.
49700.
8430.
8730.
17r'80.
2174Q.
4600.
Dollar Total ea. of 3
Material
17,660.
18,770.
18,630.
23,850.
44,730.
3390.
3650.
10,810.
15,000.
4600.
MH
585.
589.
-
_
-
336
336
394
394
-
Total
24,680.
26,840.
18,630.
23,850.
44,730.
7420.
7680.
15,540.
19,730.
4600.
92
-------
TABLE 30
Unprotected Water Recovery System
Total Construction Costs
Minimum Initial Cost
Recovery Unit - combined
mechanical and Hydraulic
Steel Hull
Outfit (Minimum)
Total w/o Sorbent
Material
t
Supply of Sorbent Material
Total with Sorbent Material
Continuous Duty
1 ~~
j Recovery Unit - Combined
Mechanical and Hydraulic
Aluminum Hull
Outfit (Complete'!
Total w/o Sorbent Material
Supply of Sorbent Material
Total with Sorbent Material
_____ — — — — — — — — — — — — — — — — — —
Dollar Total
for 1
28060.
20700.
8430.
57190.
4600.
61790.
28060.
49700.
21740.
995CC.
4600.
1CIL10C.
„. — _—— — — — — — —
Dollar Total
each of 3 :
24680.
18630.
7420.
50730.
4600.
55330.
24680.
44730.
19730V
89140..,. .
4600.
93740.
• • '
93
-------
TABLE 31
Unprotected Water Recovery Unit
Recovery Machinery Cost
Oil Recovery Machinery
Sorbent Regenerator
Fan
Harvesting Conveyor
Broadcast Duct
Broadcast Nozzle
Transfer Conveyor
Hydraulic Drive
Mechanical Drive
Diesel
Transfer Pump
Total for 1
Total each 'of 3 at .88
Added for full hydraulic drive
Total for 1
Total each of 3 at .88
Installation and Debug system
(Hydraulic and Mechanical)
Total for 1
Total each of 3 at .88
(Full hydraulic system)
Total for 1
Total for each of 3 at .88
Total Labor to Fabricate and Install
(Hydraulic and Mechanical Drive)
Total for 1
Total each of 3
(Full' Hydraulic System)
Total for 1
Total each of 3
Dollar Total
for 1
5242.
1000.
2800.
750.
100.
1600.
4920.
800.
2720.
130.
20,000
17,655
2400
22,462
19,767
Man Hours
315
150
24
16
88
12
40
330
290
60
390
3^3
336
295
280
246
666
585
670
589
-------
TABLE 32
Unprotected Water Recovery System Direct Operating Costs!
Per Hour of Effective .011 Recovery1
Crew $40.00
Maintenance 10.00
Sorbent 435.00
Fuel and Miscellaneous 10.00
Tug Boat 62.50
Barge 20.00
Total $577-50
70 Percent Lost Sorbent Recovery on Second Sweep2
Crew $80.00
Maintenance 20.00
Sorbent 130.50
Fuel and Miscellaneous 15.00
Tugboat 125.00
Barge 20.00
Total $390.50
1 No attempt is made to recover sorbent material lost due to
winds.
2 A second sweep of the slick area is made with the broad-
casting system off to recover sorbent material. It is
assumed that 70 percent of the sorbent which was lost is
recovered by this procedure.
95
-------
SECTION VIII
ACKNOWLEDGMENTS
This project was carried out by a team of HYDRONAUTICS,
Incorporated personnel headed by Mr. E. R. Miller, Jr. and
Dr. A. Gollan and including Mr. L. Stephens, Mr. J. Ricklis,
Mr. H. Lain and Mr. H. Cartright. Mr. ¥. Lindenmuth also
contributed to the development of the initial concept. Mr.
K. Jakobson acted as the project officer for the Environmental
Protection Agency.
97
-------
SECTION IX
REFERENCES
1. Witmer, Dr. F. E., and Gollan, Dr. A., "Final Report
of Phase 1 Development Program of a Continuous Re-
generating Moving Bed to Remove Oil from Oil-Water
Suspensions/' HYDRONAUTICS, Incorporated Technical
Report 7080-1 (in preparation).
2. Buffalo Forge Company,, "Fan Engineering/1 Buffalo
Forge Company, 1961.
99
-------
SECTION X
APPENDICES
Page
A. Characterization of the Sorbent Material 103
B. Development of the Sorbent Broadcasting System...f 117
C. Development of the Harvesting Conveyor and
Overall Oil Recovery Performance ' 145
D. Sorbent Regeneration System Development , l6l
E. Model Test of a 1/4-Scale Model Recovery
Platform f 173
101
-------
APPENDIX A
CHARACTERIZATION OP THE SORBENT MATERIAL
The objective of this task was to generate data on the oil
absorption capabilities of the sorbent material as a function
of sorbent characteristics, oil characteristics,, slick thick-
ness and residence time on the slick. The sorbent material
selected for use in the oil sorbent recovery system was open-
cell reticulated polyurethane foam made by the Scott Paper
Company.
The data were obtained from the tests of almost 900 in-
dividually tested polyurethane chips. Test conditions in-
cluded the following parameters:
Slick thickness
Slick temperature
Matrix porosity
Matrix reusability
Media geometry
Residence times of media
in slick
Four types of oils:
0-5
- 3-5 mm
4 - 27°C
30, 60, 80, 100, (Mostly
30 and 80 ppi )
Fresh and regenerated
(Mostly regenerated)
Square and rectangular
0 - 60 sec (Mostly 5,
10, 15, 30 sec)
Diesel, No. 4, Crude,
Bunker "C"
The first tests were scouting tests, after which the tests
where standardized in the following way. Only regenerated
sorbent material was used since this represents the real
case. The scouting tests revealed that there were no dif-
ferences in absorption rates or absorbed quantities between
fresh and regenerated polyurethane chips. According to data
reported in Reference 1, absorption of heavy oils proceeded
at a lower rate in the 100 ppi grade, then in the smaller
ppi grades, while the lighter oils did not indicate a basic
change between 100 and 80 ppi. For this reason tests were
carried out on 80 and 30 ppi grades only.
103
-------
Although a range of residence times of the chips on the oil
slick between 0 and 60 seconds were considered for testing,
it was decided after the scouting tests that enough oil
would be absorbed into the chip in periods shorter than 60
seconds to justify a reduction in absorption time. There-
fore, tests were conducted with absorption times of 5* 10,
15 and 30 seconds. Drain time was fixed at 10 seconds, which
was the estimated elapsed time between lifting the chip out
of the slick and regenerating it.
Each combination of variables was checked by three different
chips. Average results obtained from sets of three chips
were used for calculating absorption. A standard measure of
absorption was utilized throughout the tests; i.e., absorption
ratio. This is the ratio, expressed in percent, of the weight
of oil absorbed by a certain chip, to the weight of the oil
directly underneath this chip. These data are best presented
by plotting the absorption ratio as a function of the oil
viscosity, using slick thickness, and absorption times as
parameters. Viscosities and specific gravities of the oils
used, were measured within the range of 1-30°C.
Figure 22 presents viscosity data for the four types of oils,
as determined by England Laboratories of Beltsville, Maryland.
Figure 23 presents specific gravity data as determined in the
laboratory at HYDRONAUTICS, Incorporated.
The data are presented in Figures 24 through 30. It was
found to be very difficult and time-consuming to produce an
oil slick having an exact selected thickness. Thus, the
data, which were distributed throughout the thickness range,
have arbitrarily been divided into three groups: 0-1; 1-2;
2-3-5 mm. Data from each group were plotted separately-using
absorption time as the parameter. Figures 24 through 26
present data for 30 ppi grade polyurethane foam and Figures
27 through 29 present data for 80 ppi grade. Figure 30 is
an example of the data for one specific absorption time -
15 seconds. Although the data show some scatter, there still
is an unmistakable trend which indicates less viscous oils,
(Crude, No. 4 and diesel) are absorbed much better than the
viscous Bunker "C" oil. Also the curves indicate that the
absorption ratios are improved as absorption times are in-
creased. It was noticed that more oil, especially in the
lighter oils group, was lost during the drain period from
the 80 ppi grade. This phenomenon is shown in Figure 26,
104
-------
100,000
10,000 h-
o
1000 t-
A BUNKER C
D CRUDE
O |4 OIL
O DIESEL
TEMPERATURE C
FIGURE 22 - VISCOSITIES OF OILS AS A FUNCTION OF TEMPERATURE
105
-------
£
I
O
u
O
LU
o.
0.96
0.94
0.92
0.90
0.88
0.86
0.84
0.82
0.80
D
A BUNKER "C"
O DIESEL
D CRUDE OIL
O #4 FUEL OIL
10 15
TEMPERATURE °C
20 23 25
30
FIGURE 23 - SPECIFIC GRAVITIES OF OILS AS A FUNCTION OF TEMPERATURE
106
-------
280
H
O
POLYURETHANE FOAM
POROSITY - 30 ppi
SLICK THICKNESS 0 - 1 mm
O 5 sec
O 10 sec
O 15 sec
A 30 sec
1000
10,000
KINEMATIC VISCOSITY cm /sec
FIGURE 24 - EFFECT OF VISCOSITY AND RESIDENCE TIME ON OIL ABSORPTION
100,000
-------
280
240
200
O
2 160
Z
O
120
100
POLYURETHANE FOAM
POROSITY - 30 ppi
SLICK THICKNESS 1 -2mm
O 5 sec
O 10 sec
O 15 sec
A 30 sec
co
O
40
DIESEL
CRUDE
/»ec
FIGURE 25 - EFFECT OF VISCOSITY AND RESIDENCE TIME ON OIL ABSORPTION
100,000
-------
280
240
200
2 160
i
z
o
03
<
120
100
80
40
i r
DIESEL
10
CRUDE
100
T
POLYURETHANE FOAM
POROSITY - 30 ppi
SLICK THICKNESS 2 - mm
O 5 sec
D 10 sec
O 15 sec
A 30 sec
BUNKER "C"
1000
10,000
KINEMATIC VISCOSITY cm /sec
FIGURE 26- EFFECT OF VISCOSITY AND RESIDENCE TIME ON OIL ABSORPTION
100,000
-------
280
r 1
3" x 3" x 3/8" SQUARE
POLYURETHANE FOAM
POROSITY - SOppi
SLICK THICKNESS 0 - 1mm
O 5 sec
O 10 sec
OIL RECOVERED
ABSORPTION RATIO =
VOLUME OF OIL BENEATH FOAM
10
100 1000
KINEMATIC VISCOSITY cm2/sec
10,000
O
H
H
100,000
FIGURE 27- EFFECT OF VISCOSITY AND RESIDENCE TIME ON OIL ABSORPTION
-------
280
POLYURETHANE FOAM
POROSITY - 80 ppi
SLICK THICKNESS 1 - 2mm
O 5 sec
10 sec
O 15 sec
A 30 sec
1000
KINEMATIC VISCOSITY crn/sec
10,000
100,000
FIGURE 28 - EFFECT OF VISCOSITY AND, RESIDENCE TIME ON OIL ABSORPTION
-------
280
POLYURETHANE FOAM
POROSITY - 80 ppi
SLICK THICKNESS 2 - 3.5
O 5 sec
D 10 sec
O 15 sec
A 30 sec
1000
KINEMATIC VISCOSITY cm2/sec
10,000
CVI
100,000
FIGURE 29 - EFFECT OF VISCOSITY AND RESIDENCE TIME ON OIL ABSORPTION
-------
280
240 —
200 —
o
!< 160
z
o
I—
a.
IX.
8 120
CO
100
80
40
D1
O
a
o
DIESEL
10
O
D
CRUDE
100
1 T
i—r
TIME - 15 sec
80 ppi
O 0-lmm
D l-2mm
O2- mm
Oo v
O CO
BUNKER "C"
1000
10,000
KINEMATIC VISCOSITY cm /sec
FIGURE 30 - EFFECT OF VISCOSITY ON OIL ABSORPTION
100,000
-------
where absorption ratios dropped as the viscosity was reduced,
rather than following the increasing trend shown on the other .:
figures. It is also shown on Figure 24 where the points
which represent some diesel oil experiments are very low,
again becaus.e of rapid draining from the, more open structure
of the 30 ppi grade. This phenomenon had a much smaller ef-
fect on the results of the 80 ppi, grade.
No drastic differences were noticed between absorption ratios
from slicks of different thicknesses which actually means
that absorption rates are affected by the oil slick thickness
and are increased as the thickness is increased. The following
table presents absorption influxes for an imaginary oil, having
a kinematic viscosity of 100 cm2/sec, and a specific gravity
of .9 gr/cm3. The fluxes were calculated for 30 seconds of
absorption, where absorption values were picked from the
curves of Figures 24 through 29-
TABLE 33
Oil Absorption Fluxes
Sorbent Type
PPI
30
30
30
80
80
80
Slick Thickness
mm
1
2
3
1
2
3
Absorption Influx*
gr/s ec cm2
5.35*
10.35*
10.80*
4.50*
9.10*
14.78*
lO-3
. ID"3
ID'3
ID'3
lO-3
lO"3
* Area of chip -
103.5 cm2
Except from the flux in the 3 mm 30 ppi case which is low
in comparison because of higher drainage, the fluxes in-
crease almost linearly with the slick thickness.
114
-------
Based on the results of this task it has been concluded that
80 ppi polyurethane foam should be used as the sorbent ma-
terial for low viscosity oils. Either grade may be used on
high viscosity oils such as Bunker "C". The figures re-
lating the absorption ratio to the parameters of oil vis-
cosity, slick thickness, residence time and foam porosity
may be used in engineering estimates of sorbent coverage
required for complete removal or the oil recovered with 100
percent coverage.
115
-------
APPENDIX B
DEVELOPMENT OF THE SORBENT BROADCASTING SYSTEM
The objectives of this task were:
(l) Develop and demonstrate a means of distributing
sorbent material chips, on an oil slick, more or
less evenly across the inner width of the craft at
its forward end.
(2) Develop and demonstrate a pneumatic system for
conveying chips from the oil removal section of
the craft to the distribution device.
These objectives were accomplished with model tests. Due
to the specialized nature of the problem, the model size and
other parameters were near full scale to reduce scaling
problems In the prototype design stage. The resulting design
information Is thus limited to craft employing the same basic
concepts and of the same order of size and capacities as the
model.
Before selecting basic parameters for the model, a small
scale experiment was conducted to examine basic questions.
At that time the intent was established to conduct most of
the experiments with a material having greater density than
the candidate material. This was done to simulate the weight
effects of retained oil in regenerated chips while avoiding
the problems associated with handling regenerated chips.
Only those tests necessary to check for possible problems
associated with oily material were conducted with regenerated
chips. This approach was taken based on experience gained
in previous work with the sorbent material. Indications
were that the porous and fibrous surface of the material did
not permit large areas to come Into contact and it was felt
that stickiness would not be a problem. Weight Is of course
a consideration in conveying and even the small amount of oil
retained in a thoroughly squeezed chip is enough to increase
its weight by a factor of two.
Closed cell polyurethane foam, available in a non-rigid form
and with a density about twice that of the candidate material,
was used for most of the testing.
11?
-------
STAGES IN THE DEVELOPMENT OF THE DISTRIBUTION DEVICE
Passive Nozzle Developed on Small Scale Experiment
The small scale experiment produced a passive distribution
nozzle in the form of a transition .from a circular duct to
a wide, • shallow, rectangular outlet. Thus in the horizontal
plane, diffusion occurred, supplying width to the sorbent
distribution-pattern. The length of the nozzle was minimized
and .internal vanes were installed to reduce the effective dif-
fusion ; angle.. .• .. ;,;.,..,
j . . „
There was concern that the leading edges of these vanes might
trap sorbent chips if the chips were excessively "sheet-like";
that is, thin for a given set of outline dimensions. This
did in fact occur but it appeared that for 3-inch square
chips (candidate size for full scale) a 3/8-lnch thickness
provided enough stiffness so that chips would not fold com-
pletely over the leading edges and be permanently trapped.
Further, it was felt that if the dimensions between leading
edges were kept large relative to chip dimensions, it would
not matter if some chips were trapped.
Passive Nozzle Tested on the Full-Scale Model
A similar nozzle for the full-scale model was designed and
fabricated of sheet metal. Vanes were of cardboard,, tem-
porarily taped in place since some trial and error was ex-
pected in determining their location (Figure 31).
This nozzle performed quite well in early tests when ma-
terial feed rates were still relatively low. Distribution
patterns were sufficiently wide and,fairly even (Figure 31).
At rates above about 25 percent, however, the nozzle became
plugged with chips at the vanes leading edges.
Passive External Plate Distribution Device
It was not certain whether the plugging was caused by higher
chip concentration or the reduced speeds which are a con-
sequence of increased chip rate. It was thought that the
conditions which precipitated a plug might be temporary and
that if they did not occur in a closed area, a plug might
not fully develop. Based on this reasoning, a distribution
118
-------
~ "
FIGURE 31 - CHIP BROADCAST PATTERN ACHIEVED IN EARLY TEST
(Grid Spacing = 4 feet)
119
-------
device was constructed which was essentially an open nozzle.
Splitter plates were located several inches away from the
conveying duct outlet. The plates were less efficient than
the nozzle at directing flow to the sides, but with leading
edges located external to the duct, it was hoped a complete
plug would be unable to develop.
Higher chip rates were obtained with this concept (about
50 percent) but eventually, plugs did occur unless the
plates were moved so far away from the outlet that distribu-
tion became unsatisfactory. In fact, distribution patterns
were never completely acceptable with the external device
even at low chip rates.
Flexible Ducting Nozzle
Failure to achieve adequate chip rates with a passive nozzle
forced a decision to accept the penalties associated with an
active distribution device. It is conceivable that a thicker,
stiffer chip would have been compatible with a passive noz-
zle but the extra material would be wasted in terms of oil
recovery and impose acquisition and storage penalties.
The first approach to an active distribution device employed
a section of flexible ducting attached to the conveyor out-
let. The open end of the flexible section was free to swivel
^5 degrees to each side about a vertical axis. This was a
mechanically and structurally neat installation but it failed
by plugging. The ducting was inexpensive, fabric covered,
coiled wire construction, and it is felt the plugging began
in the ridges on the compressed side of the duct in the fully
deflected position.
Movable Parallel Plate Nozzle
In the meantime, while awaiting delivery of the flexible
duct, another concept was developed and tested with more
success. Two rectangular plates were hinged at the vertical
edges of a rectangular outlet section. Their free ends were
connected with a link. A horizontal plate was mounted above
and below this movable plate unit. The result was a 2-foot-
long, rectangular cross section portion of duct which could
swivel JI5 degrees to either side of center. The geometry
is such that cross sectional area is a minimum at the sides
and the increased outlet speed is beneficial In obtaining
lateral throw.
120
-------
This device was capable of passing chips at the highest rates
achievable with the existing chip feeder box. A plug did de-
velop in one test at high chip rates. It occurred in the
sharp corner produced when the nozzle is fully deflected. The
sharpness of the turn will be reduced on the prototype design.
EXPERIMENTAL EVALUATION OP SELECTED DISTRIBUTION DEVICE
(MOVABLE PARALLEL PLATE NOZZLE)
Description of the Model and Experiment
Figure 32 shows the nozzle installed on the full-size model
of the broadcast system. The forward structure of a con-
ceptual craft was also simulated. The flotation pontoon
walls act as windbreaks as well as chip containment devices.
The forward portion of the side screens serve a containment
function and the simulated debris rake serves a certain con-
tainment role since its members must be large enough to pro-
vide significant strength on a prototype. The nozzle move-
ment, which would be powered on a prototype, was performed
by hand in the experiment.
Broadcast distributions were defined by determining the
number of chips deposited in each of eight, 4-foot-wide strips
running parallel to the model centerline. This method was
satisfactory for winds less than about 8 mph. In higher
winds, chips would not stay fixed on the grass surface of the
test area and quantitative measurements could not be made.
Data are available only in the form of photographs taken during
and immediately after these tests.
Broadcast Patterns Calm and Light Winds
Quantitative broadcast pattern data are presented in a
manner showing deviation from the optimum, which is an equal
number of chips in each longitudinal strip. Some data from
the external plate device are presented in Figure 33 to il-
lustrate its ineffectiveness in producing uniform distribu-
tion. Results for the selected nozzle concept are shown in
Figure 34.
Some of these tests were conducted in winds up to about
6 mph at 30 to 40 degrees off the bow and it appears that
irregularities in nozzle motion and chip rate have more ef-
fect on distribution than winds of this strength.
121
-------
I I>
FIGURE 32 - SIMULATED DEBRIS RAKE, FLOTATION PODS, AND
SIDE SCREENS ERECTED AT DISCHARGE END OF
BROADCAST EXPERIMENT
122
-------
>>vu
?50
o. inn
PERCENTAGE OF UNIFORM DISTRIBUTION IN EACH 4 FOOT WIDE STRI
3 o 1 g 8 o 8 ooSoi
1
WIND 330° /5m
• .':'. '£•
Qh
^
11
|
L
i, , i
6 12 8 4 0 4 8 12 1
Wlh
tfV^WWHHIII*
-JD 3301
4 • :'" '':- '.'''.'-'
V2rr
•:';v;>:';;
'.£?%.
*:'\/ '''-'
ph
•;:,• '^-'^- -
;?'••'•'<•'!.
'•' -.-:.'•'•.'-•
'•*'•':-•'"• • '-':
1 •/,''. •''';'••
!:' V'Vr': :{'
:.-.;:::;.>:?:':::;:::-
:••":'- • >•:':+•'••:
; ;,.'• ;.•-.'••.;'•-•
ft§
• 'is'-:-: .- ' '.'•':
•K-
t ^,::,
'\|v-;:p"' '. :-;:
;•;•;•'.•:•:' v.
.'." /:'!''S-
300
250
150
1 rif\ i^
50
0 LOWEI
6 PROVI
IMPRC
300
250
200 . .
150
50
0 <^
t?
}
*ING ROOF
DED SOME
>VEMENT
10
16 12 8 4 0 4 8 12 16
LATERAL DISTANCE FROM DUCT CENTERLINE FEET
PLAN VIEW Of
SPLITTER PLATE
ARRANGEMENT
FIGURE 33- CHIP DISTRIBUTIONS OBTAINED WITH PASSIVE EXTERNAL
PLATE DEVICE
123
-------
o. 200
QL
1—
to
% 150
o 100
O
u.
x 50
;;:;£:;::;:;:;>:-:
£££S?y^
9lfy;g:&,
ifl|4|:
'-'•.'-V '.%''••"' .*-.'''"'-'
iW
;;®^
•&%m*
•^mm
::-.:-':-:'''.'-'-:-:'.-'-'."::":-;:
t>-.v;:|5|
llSH
•i
IfHi
6 12 8 4 04 8 12 1
LATERAL DISTANCE FROM DUCT (? FFFT
200
150
100
50
0
6
CO
(B) NOZZLE MOVEMENT: RAPID IN CENTER,
SLIGHT PAUSE AT EACH SIDE.
PERCENTAGE OF UNIFORA
— • — • ISJ
ui o tn o
o o o o o
til|70^ OF FULL
RATE
REPEAT
i
i
•
r~~~m
200
150
100
16 12 8 4 04 8 12 16
LATERAL DISTANCE FROM DUCT £ , FEET
(A) NOZZLE MOVEMENT: STEADY;
NO ATTEMPT TO WEIGHT ANY AREA.
FIGURE 14 - CHIP DISTRIBUTIONS OBTAINED WITH
MECHANICAL NOZZLE. (Wind from calm
to 6 mph at 30 degrees off the bow)
-------
The data show that if nozzle motion is slowed near the sides,
distribution satisfactorily close to uniform can be achieved'
in cross winds up to about 6 mph., If the wind is nearly on
the bow., distribution should remain satisfactory in con-
siderably stronger conditions.
Strong ¥ind Patterns
Strong crosswinds had a pronounced effect on distribution
patterns. An examination of photographs taken of tests con-
ducted in 14 and 20 mph winds at 45 to 60 degrees off the
bow indicate distributions similar to that shown in Fig-
ure 35. Winds from this direction probably .have more effect
than head or beam winds. Head'winds,, of course,, produce no
asymmetry,, and crosswinds are more effectively blocked by
the pontoon sidewalls.
The non-uniformity in broadcast patterns caused by winds is
not an easy problem to attack since any structure added as a
shield carries its own penalties. The best approach may be
to provide for manual control of the nozzle motionVand use
this means to achieve best possible slick coverage in strong
winds.
Sorbent Loss Rates
Some sorbent material will be lost in operation and the
magnitude of the losses must not be unreasonable in terms
of sorbent material make-up capacity. In all broadcast tests
conducted in conjunction with the simulated bow, the number
of sorbent chips deposited outside of the enclosure was de-
termined (chips thrown through the debris rake were not con-
sidered lost unless deposited outside of the projected line
of the pontoon wall). These losses are presented in Table
34 as a percentage of total chips broadcast in each test.
Except for tests 47 and 52, a conservative average loss rate
is about 2 .percent. ";.• • • • ' '
Tests 47 and 52 are considered:unrealistic compared to the
prototype counterpart for a combination of reasons. Chip
rates were low and because of the high speeds associated
with low rates, chips were thrown a maximum distance. High
speed alone does not produce loss but for these tests, the
nozzle was mounted horizontally for convenience of assembly.
125
-------
C£
O
u_
z z
O
-------
TABLE 34
Sorbent Chip Loss Rates Experienced
in Full-Scale Model Tests
Test No.
45
46
48
49
50
51
52
53
¥ind
Conditions
Calm.
Calm
Calm
330°/6 mph
045°/l4 mph
045°/8 mph
24o°/5 mph
060/1 4 mph
060/20 mph
Chips Lost
percent
2
0.15
,,,•
1.7
1.9
1.6
0.15
5
1.6
Comments
Losses occurred over top
of pontoon. Canted noz-
zle should reduce this
somewhat.
73 percent of losses
were fwd through rake.
Test -was at relatively
low rate/high speed.
Canting nozzle should
reduce this type loss.
50 percent of losses
were fwd thrqugh rake.
About same lost on
both sides.
Fairly low chip rate
and high speed probably
contributed to high
losses
Nozzle canted downward.
All chips lost downwind
side.
1. Rates are percentages of total chips broadcast in that
tests.
2. Wind directions are relative to model centerline.
127
-------
Later, the nozzle was canted downward about 10 degrees with
the Intention of reducing this type of loss. That this was
fairly successful is shown by the 1.6 percent loss rate in
test 53 conducted at nominally the same rate as test 52 with
its 5 percent loss rate. It is therefore concluded that a
loss rate of 2 percent of the number of chips being broad-
cast should be consider"ed typical in strong cross winds. It
is unlikely that it can be lower than 1 percent. In winds
of 10 mph or in headwinds, the loss rate should drop to about
1 percent and probably 0-5 percent in winds of less than 5 mpl-
Based on 1170 - 3 x 3-inch chips per second required for
nominal 100 percent area coverage to recover 10,000 gallons/
hour from a 1.5 mm slick, these loss rates translate into
makeup feeds of 84,200 chips/hour (165 cubic feet/hour) in
strong crosswinds to 21,040 chips/hour (4l cubic feet/hour)
in winds of less than 5 mph.
DEVELOPMENT OF THE SORBENT CONVEYING SYSTEM
Pneumatic conveying is the selected method of transporting
sorbent material from the regenerating apparatus at the
after end of the craft to the forward end where it is dis-
tributed on the slick and begins the sorption process. The
conceptual design envisioned sorbent chips discharging from
the regeneration device into the intake of an axial flow fan
which exhausted into a circular duct, the duct then con-
veying the chips to the distribution device forward.
I
Development of Conveying System Concepts
The literature pertaining to pneumatic conveying which could
be located offered little guidance with regard to low density
material in a chip form and so as mentioned previously, a
small experiment was devised. The primary purpose was to
investigate air speeds required to properly convey the chips.
A 12-foot-long/9-inch diameter duct was constructed from
clear plastic so that observations could be made. A one
horsepower centrifugal blower with a materials handling
wheel was obtained (a centrifugal blower for this size model
w,as more readily available and less expensive than an axial
flow fan).
128
-------
It had been hoped to feed chips through the blower but it
was found that the wheel progressively accumulated chips
until either it became effectively plugged or unbalanced.
The wheel was observed with a stroboscope and it was ob-
served that some chips folded over the blade leading edges and
that aerodynamic forces were great .enough to keep them there.
The blades incorporated swept leading edges so that a compo-
nent of centrifugal force acted along the leading edge.
Stringy material for which the wheel was designed, should
slide out and around the open tips. The sorbent material,
however, exhibited too large .a friction coefficient for this
to take place.
Two venturi injectors were constructed. The first had a
circular cross section which was too inefficient in terms of
feed area compared to cross sectional area. Thus a rec-
tangular cross section injector was constructed.
A series of tests was conducted to determine the effects of
air speed on conveying characteristics. Speed was varied by
choking the blower inlet. Material properties were also
varied in several directions. Two material densities and two
chip sizes(3 by 3 inches and 1-1/2 x 1-1/2 inches) were used.
The material was 1/4 to 3/8-inch thick. High-speed films
were taken of a section of the clear duct.
Examination of the films indicated that a speed of about
30 feet per second was required if the small chips were to
be evenly distributed vertically .in the duct. The visual
effect of speed was not distinct however, and conveying was
not noticeably poor until a speed of about 25 feet per
second.
Results for the larger (3 by 3 inches) ^chips were more diffi-
cult to interpret as were differences between the two material
densities. The large chips were too large for the duct and
collisions with the walls were too frequent to judge con-
veying quality.
It was judged that 30 feet per second for 1-1/2 by 1-1/2 in.
chips with a density of 3-6 Ib/cubic foot was the minimum
speed for good conveying. A dimensionless parameter similar
to the Froude number was used for scaling. This parameter
is defined as
129
-------
V
V
V
where
V = duct speed-ft/sec,
Y = density of material - lb/ft3,
Y = air density - lb/ft3, and
a
^ = chip thickness, ft.
It was felt this relationship could be applied as long as
chip size became no greater than about 1/6 of the duct
diameter. This was the situation with the smaller chips in
the small experiment when wall collisions did not seem to
dominate the mechanism. On this basis, a minimum speed of
about 42 feet per second was required for the full-size model
using 3 by 3 inch chips.
An important effect revealed in the :films was that large,
thin chips are too limp for the process and upon contact with
walls, tend to bend and collapse and sometimes flatten com-
pletely against the wall. The result is a very inefficient
transport mechanism and it was decided,to make full-scale
chips stiff enough to reduce this type of loss. A thickness
of 3/8 in. for 3 by. 3 inches square chips was selected.
Design of Full-Size Model
Table 35 shows the type of losses which were accounted for
in designing the model. Reference 1 indicated an axial
flow fan is sufficient to handle material to air ratios (by
weight) of about 2. Initial calculations called for a fan
output of about 400 cfm at a static pressure of about 3 inches
of water. This pressure is near the limit of a basic axial
flow fan. Straightening vanes behind the fan can increase
the capability, but it was felt the leading edges of the vanes
would greatly reduce the chance that chips could be fed through
the fan. The latter was important since a centrifugal blower
wheel design was located which seemed to show promise as a
130
-------
1. AIR FRICTION IN DUCT
V2/2g r
. - ,OA f x L/ D
air 69.4 |_
2. CHIP FRICTION IN DUCT
CxW xL
h - m
Chips W x69.4
a
3. ACCELERATION OF CHIPS
W x V2 / 2g
i _ m m
accel W x69.4
a
4. POWER INCREASE IN BLOWER
W xV2t.
HP = m f'P
2g x 550
+ 1.5 f = 0.013 FROM PIPE FLOW DATA
J 1 .5 = EXPERIMENTALLY DETERMINED
NOZZLE LOSS (When deflected )
C IS A "FRICTION COEFFICIENT"
AND « 2. 5 FROM EXPERIMENTS
IF ACCELERATION TAKES PLACE AT
BLOWER OUTLET. OTHERWISE THIS
SHOWS UP AS MOTOR POWER INCREASE.
EXPERIMENTS SHOW NET EFFECTS OF
CHIP + AIR FLOW IN BLOWER INCREASE
THIS POWER
TYPICAL CALCULATED
VALUES L=70, D=1.5
V=45,W 9 Ib/sec
m
0.75
INCHES OF
WATER
1.70
INCHES OF
WATER
0.67
INCHES OF
WATER
4
HP
TABLE 35 - TYPES OF SYSTEM LOSSES
-------
chip handler and renewed the hope that a venturi would not
be needed. The wheel design is shown in Figure 36. Es-
sentially all the centrifugal force acting on a chip which
has folded over a leading edge is directed along the edge.,
which hopefully would be sufficient to clear the blades.
A centrifugal flow blower with the wheel mentioned was
specified for the model. It was powered with a 7~1/2 horse-
power motor.
The material ratio of 2 and 4-5 feet per second speed yield a
duct diameter of about 16 inches which was increased slightly
to 18 inches. Sheet metal ducting was obtained and a 7-foot-
long clear plastic section was constructed for observation
purposes. Total duct length 'was about 58 feet and the noz-
zle was expected to throw 10 feet to complete the 67-1/2 foot
length established in the conceptual design. The duct was
supported 5 feet above ground level to model prototype wave
clearance.
Because of pessimism regarding feeding through the blower,
an injector was also constructed. Initial tests with the
blower indicated that chips were captured by the wheel and
so the model was erected with the injector installed.
A sheet metal nozzle similar to that developed on the small*
model was obtained and trial vanes taped' in place.
Sorbent chips were fed into the system with a batch con-
veyor apparatus. A 10-foot-long by 2-foot-square open box
was constructed and the bottom and both sides lined with
canvas strips attached at one end to the sides and bottom
edges of a 2-foot plywood square. At the open end of the
box., the bottom canvas was routed under the box and the side
canvases around the sides of the box. When the free ends
were simultaneously pulled toward, the. closed end of the box,
a three-sided conveyor action took place, carrying up to
20 cubic feet of chips gradually over the lip of the box
into the system inlet (Figure 3).
Chips were broadcast onto a 20 by 40-foot plastic tarpaulin
marked with a grid.
The full-size experiment as originally designed and erected
is shown in Figure 3.
132
-------
FIGURE 36 - THE BLOWER WHEEL DESIGN CAPABLE OF HANDLING
SORBENT CHIPS
133
-------
Tests and Problem Solving on the Model
It soon developed that the system could achieve only 10 per-
cent of the required chip rate, limited by the injector.
Various improvements increased this limit to about 33 per-
cent. At this point it seemed that an injector was too in-
efficient in extracting power from the blower and further
tests were made of the ability of the blower to pass chips.
There was no progressive buildup of chips in or on the wheel
and the model was rearranged to feed into the blower through
the 90-degree inlet elbow and bell mouth which had been de-
signed and built with the rest of the model components.
Wheel RPM was also increased 12 percent to make more power
available. With these two changes, feed rate increased to
66 percent before a new barrier was reached, choking in the
blower inlet.
The 90-degree turn within a short distance of the blower
has been identified as a potential trouble spot in the de-
sign phase. Rather than investigate methods of reducing
the problem, it was decided to eliminate the turn by in-
stalling the blower on its side.
In this configuration, feed rates up to 95 percent were
achieved and although speed dropped below good conveying,
there was no indication of reaching a feed-rate limit. It
is felt that 100 percent rate could have been achieved if
the feed box had been larger.
Having demonstrated that sorbent chips could be conveyed
the required 67-1/2 feet but faced with higher losses than
had been anticipated, it was decided to shorten the duct
to obtain higher air speeds at similar chip rates. The
modified duct was 27 feet long from blower to nozzle outlet
with the center of a broadcast pattern being about 10 feet
further forward.
Tests Conducted with Regenerated Chips
As discussed earlier, the foregoing tests were conducted
with heavier, dry chips. To verify the assumption of no
differences between dry and regenerated chips, approximately
Jl gallons of number 6 fuel oil were dumped into the blower
inlet. The inside of the blower and the nozzle were
thoroughly coated and the ducting was coated to a somewhat
-------
lesser extent. The nozzle was locked to one side to produce
the worst condition.
Regenerated chips from Task 3 were conveyed and broadcast
several times at high rate. No chips were' retained in the
blower or ducting .and only two or three remained in the
corner of the nozzle (this occasionally occurred with dry
chips for that matter).
Engineering Requirements for.the Conveyor
Enough chips were obtained to provide a 6-second test at
100 percent rate. This was a compromise between time enough
to make measurements and the problems of working with a very
large number of chips. Unfortunately, the chips exhibited
a clumping property and entered the intake at a slightly
non-uniform rate so that the system operated at a slightly
unsteady condition. -Data were not very accurate .for these
reasons.
Static pressure measurements were made in the blower inlet
and just downstream of the blower outlet. Inlet pressure
measurements were not of much value since it was not practi-
cal to provide a long enough straight section of duct to
produce straight streamlines. All static pressure measure-
ments had to be made with flush taps since a probe is not
usable in the presence of chips.
Motor current was measured to supply data relating to power
requirements.
Measurement of speed in the duct is a problem due to the
chips. The inside of the duct could have been necked down
to create a venturi but diameters and lengths required to
produce adequate pressure differences looked to be impracti-
cal. The method used relied on a two-bladed paddle-wheel
mounted in a slot in the duct wall. Paddlewheel RPM was
measured with a servotach DC generator and calibrated with
a pitot-static tube in the absence of chips. The flow is
not homogeneous and chips move slightly slower than the air
but it was assumed adequate to take the speed determined
from paddlewheel RPM as both air and chip speeds.
135
-------
Required Air Speed in Duct
Observations at the duct outlet and in the clear duct section
installed during part of the model testing indicated that if
speeds dropped below 40 feet per second, the chips no longer
appeared as a uniform mass and gave the impression of moving
in the lower part of the duct. This is in general agreement
with the results of the small-scale experiment.
It is concluded that minimum air speed should be 45 feet per
second to obtain good conveying.
Head Losses in the Duct
The head loss in the duct and nozzle is caused by air fric-
tion and chip friction. For engineering purposes, it was
decided to account for air friction loss on the basis of
standard pipe flow data as if there were no chips. Measured
head losses were reduced by a pipe flow air friction loss and
the remainder attributed to chip friction.
In accordance with Reference 2, chip losses were thought of
as being the energy involved in a frictional force acting
over the length of the duct in a time interval determined
by chip speed. The frictional force is the product of a
friction coefficient and a normal force equal to the weight
of chips in the duct at any instant.
The friction coefficient was expected to be about 0.4 but
the test data produced a value of about 2.3 (Figure 37). In
hindsight, the tumbling and the unsteady speed of individual
chips observed in the high speed films of the small model
may be indicative of greater losses from air drag on in-
dividual chips than if the sorbent were in particle form.
Also, some chips no doubt come to a stop against the duct
wall and must be re-accelerated. Both types of losses appear
in the experimentally determined "friction coefficient" in
addition to material sliding along the duct.
Characterization of the system is simplified by the fact
that nozzle effects seem to average out with nozzle movement
as shown by the data points on Figure 37.
136
-------
. C x W x L
hchips = m
W
h =HEAD LOSS IN INCHES OF WATER
C = FRICTION COEFFICIENT
L = DUCT LENGTH IN FEET
Wm =LBS OF MATERIAL/SECOND
WQ = LBS OF AIR/SECOND
-------
Engineering Requirements for the Blower
The blower should be centrifugal rather than axial flow since
in the head range required, an axial flow fan would require
swirl recovery vanes which would prohibit feeding sorbent
chips through the fan. It is important that feeding be done
through the fan or blower since the alternative,, an injector,
is inefficient.
The blower wheel must incorporate certain features if it is
to successfully handle the sorbent chips. The blades must
extend radially into the hub to eliminate the axially aligned
leading edge present on air handling wheels and some materials
handling wheels. There must be no rim or front plate on the
wheel so that chips are free to slide along the radial leading
edges and clear the wheel.
Chip Effect on Blower Head,
The presence of sorbent chips in the blower has the effect
of reducing, static head produced at constant RPM and dis-
charge. The magnitude of this reduction is shown in Figure
38, where F is the ratio of effective static head delivered
li
with chips to static head capability of the blower alone at
a given RPM and discharge.
The effect of ?„ on the prototype design will probably be
ti
a reduction in the material ratio to about 1 rather than 2 as
originally intended.
Chip Effect on Blower Power
Enough power must be transferred to the chaps to give them
wheel tip speed. Some chips will no doubt slide along blade
surfaces which should show up as a power increase.
Also, even though chips reduce head at constant capacity, it
seems that efficiency is reduced even more so that power in-
creases considerably more than might be allowed for.
Measured motor currents were converted to horsepowers. A
horsepower due to air only was obtained from manufacturer's
data at the test RPM and capacity. The difference between
measured and air horsepower was attributed to the total effect
138
-------
Fp = ( POWER REQURIED - AIR ONLY POWER) ^-(CALCULATED CHIP POWER)
CALCULATED CHIP POWER = W x V2 /2 g x 550
m tip
o
u
at.
LU
I
LU
CO
O
I
LU
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Wm /WQ - LBS OF MATERIAL PER SEC * LBS OF AIR PER SEC
EFFECT OF CHIPS ON BLOWER POWER REQUIREMENTS
FH =
u
2
a
&
x
u
CO
i
(EFFECTIVE STATIC HEAD PRODUCED) * ( STATIC HEAD EXPECTED WITHOUT CHIPS)
1.00
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
W /W - LBS OF MATERIAL PER SECOND * LBS OF AIR PER SEC
m a
EFFECT OF CHIPS ON BLOWER STATIC HEAD
FIGURE 38 - EXPERIMENTALLY DETERMINED FACTORS WHICH ACCOUNT
FOR THE EFFECT OF CHIPS ON BLOWER HEAD AND POWER
-------
of chips. These powers were compared to the power required
to accelerate the chips to tip speed to produce a measure
of power efficiency to be applied in conjunction with the
pressure efficiency discussed earlier. Values of this factor,
F , are shown in Figure 38. This approach to power require-
ments Implicitly includes the chip acceleration loss defined
in Table 35.
Model of System Operation
Figure 39 presents a graphical model of a concept of the
system which accounts for losses and the effects of chips
on blower head and power.
Validity of Head and Power Factors
The sorbent- broadcast model is a very specialized applica-
tion of pneumatic conveying and the correlations produced
by the tests cannot be expected to hold if applied to a
system much different from the model. It was for this
reason,, of course., that the model was made as near the con-
ceptual design as possible.
0
-------
STATIC HEAD AT RPM 2
SYSTEM OPERATES AT POINT 1 WITH NO
CHIPS BEING CONVEYED
IF CHIPS DID NOT AFFECT FLOW IN BLOWER,
THEN WHEN CHIPS ARE CONVEYED, THE
CAPACITY DROPS DUE TO DUCT LOSSES,
POWER INCREASES OVER AIR HORSEPOWER
TO IMPART TIP SPEED TO CHIPS, AND "PM
DROPS VERY SLIGHTLY. SYSTEM WOULD
OPERATE AT THE HYPOTHETICAL POINT 2
SINCE CHIPS DO AFFECT THE BLOWER, HEAD
DELIVERED DROPS AND POWER INCREASES
AND THE SYSTEM ACTUALLY OPERATES
AT POINT 3 . POINT 3 MUST FALL ON A
LINE WHICH DESCRIBES DUCT LOSSES.
4. IN DESIGN, POINT 3 IS THE KNOWN
STARTING POINT.
CAPACITY, Q
FIGURE 39 - MODEL OF HEAD AND POWER INTERACTIONS BETWEEN BLOWER, DUCT, AND CHIPS
-------
Sample Calculation
I. Design Conditions
1. Nominal slick coverage 4400 ft2/min (conceptual
design)
II.
2. Density of regenerated
chips
3. Chip dimensions
4. From 1,2, and 3, chip
rate
5. From 2 and 3 chip
weight
6. Duct speed with chips
7. Duct length
8. Weight sorbent/weight
air
9. Density of air
Duct size calculation
3.6 lb/ft3
3 x 3 x 3/8 in.
1172 chips/sec
0.00705 Ib/chip
45 ft/sec
67.5 ft (conceptual design)
1.0 nominal
0.075 lb/ft3
1. Weight of material, W = 1172 x 0.00705 = 8.26 Ib/sec
2. Weight of air,
W =1.0x8.26 = 8.26 Ib/sec
a '
3. Total volume flow rate Q = volume of sorbent
+ volume of air
=
60 - 67*6
4. Duct sectional area
Q = Area x Speed
Area = 6746/45 x 60 = 2.50 ft2
„. , -\ ' 4 x Area
Diameter = \.
ir
\ - 4 x 2.50
'v 7T
= 1.78 ft
(21.36 in.)
142
-------
III. Duct losses (Table 35)
.1... Air friction, h = * /^ .f x - + 1.5!
, _ (45)g 67 5
f ~ 69.4 x 64.4 °'013 x TlQ + 1>5i = °'9° ln' of
/ water
W x L
2. Chip friction h- . = C x m
chips W x 69.4
3
hohlps • 2-3 * o x : - 2-23 in. of water
3. Total duct loss = air + chips
h = 0.90 + 2.23 = 3.13 in. of water
IV. System operating point (point 3 on Figure 39)
Q = 6746 ft3/min
h = 3.14 in. of water
V. Inclusion of chip effect on head
From Figure 38, when ^m/^a =1.0
Fh « 0.30
to deliver h = 3.14, blower must operate at
an apparent condition,
Q = 6746
H = 3.14/0.30 = 10.47
(this is point 2 on Figure 39 ).
143
-------
VI. Blower RPM and air horsepower at point 2. (Based on
American Standard Size 21 long shaving type blower
with a 36-1/2 inch wheel diameter)
1204 RPM
17.41 h
Pair
VII. Ideal chip horsepower and allowance for chip effect
on chip horsepower.
1. Wheel tip speed = ^'\^J^ 12°4 = 192 ft/sec.
V2 /2g x W
2. Ideal chip power = — P „ (Figure 38 )
550
hn - (192)2 X 8.26 _ o ^
hpideal - 64.4 x 550 = 8'6°
3. Prom Figure 38 when W /W =1.0
Fp • 3-3
Actual chip power = Ideal chip power x F
""actual ' 8'6° x 3-3 = 28'38
4. Actual operating power = actual chip power + air
power
H = 17.41 + 28.38 = 45.99
144
-------
APPENDIX C
DEVELOPMENT OP THE HARVESTING CONVEYOR AND
OVERALL OIL RECOVERY PERFORMANCE
The objectives of this task were:
1. Determine the conveyor inclination angle and speed
for optimum performance.
2. Using the optimum conveyor operating condition,
develop data on system performance as a function
of the other operating parameters.
The system performance was defined in terms of the per-
centage of oil recovered from the slick and the percentage
of oil in the recovered fluid. The performance goals given
by the EPA were 100 percent recovery of the oil from the
slick and in excess of 90 percent oil in the recovered fluid.
The operating parameters which were studied in addition to
the conveyor angle and speed included:
*
Slick thickness and viscosity' (oil type and
temperature)
Sorbent coverage
Residence time
Calm water versus waves
The harvesting conveyor test setup was installed, in a
80-ft towing tank along with the sorbent chip distribution
carriage. Figure 5 shows a picture of the harvesting con-
veyor which can be run at speeds up to 3-5 ft/sec with the
variable speed drive installed. The conveyor consists of
1 in. mesh opening flat wire belt. Figure 6 shows a picture
of the sorbent ctyLp distribution carriage. This carriage
is equipped with a moving belt feed device which feeds the
chips to a rotating paddle wheel spreader. This device is
used to distribute the chips in the tank in a random fashion.
The test procedure is to cover a 55-ft length of the tank
with a known Quantity of oil. The sorbent chip distribution
carriage is then run down the tank at a fixed speed to dis-
tribute a known amount of sorbent on the slick. After waiting
-------
the required residence time, the harvesting conveyor follows
down the tank and picks up the chips and drops them in a col-
lection box. After the run, the chips in the box are re-
generated and the.amount of oil and water recovered is
measured. The oil remaining in the tank is collected and
measured to determine the amount of oil not recovered. Fig-
ure 7 shows a typical test In progress. . ;
v. "'' '' V" ' i
The initial tests were directed- at.determining the optimum
conveyor angle and conveyor speed relative to forward speed.
Typical results are shown in Figure 4'0. In this figure,
the total percent of oil recovered and the percent of oil in
the recovered fluid are plotted as a function of the conveyor
speed ratio. ' The conveyor speed ratio Is defined as
(Conveyor Speed) x (Cosine Conveyor Angle)
Forward Speed
The test conditions are noted in the figure. The total per-
cent of oil recovered does not seem to be a .function of the
conveyor angle. However, it is a function, of the conveyor
speed ratio. At low speed ratios, the chip distribution
density increases In front 'of.the conveyor, during collection.
This slightly increases total percentage-recovery of oil.
This condition also forces more water into the ..chips so that
the percentage of oil in the recovered fluid drops. At high
conveyor speed ratios the percentage of oil in the recovered
fluid also drops because of the entrained water droplets
carried by the conveyor. The 45-degree conveyor angle is
superior to the 30 degree angle since less water is recovered
with the oil. The 45-degree angle results. In less water be-
cause less water is entrained with the conveyor and there is
a greater tendency for water to run off the chips. Tests
In waves indicate they cause an increase in the percent of
oil recovered. This Is because the waves agitate the sorbent
chips causing contact with a greater area of the slick. Ob-
servations indicate that the conveyor angle cannot be in-
creased much above 45 degrees without the use of flights.
Thus, based on these tests a conveyor angle of 45 degrees
and a conveyor speed ratio of 0.7 seem to be optimum.
The system performance as a function of slick thickne'ss and
oil viscosity is presented in Figures 4l through 44. Fig-
ures 4l and 43 give the percent oil recovered and Figures.42
146
-------
UJ
U
ae.
UJ
BL.
100
90
80
70
60
50
40
30
45 CONVEYOR ANGLE
30° CONVEYOR ANGLE
O = PERCENT OF OIL RECOVERED
D = PERCENT OIL IN TOTAL RECOVERED FLUID
NOTES:
1.5 mm SLICK OF #4 FUEL OIL
30 SECOND RESIDENCE TIME
3"x3"x3/i3" 80ppi SORBENT CHIPS
NOMINAL CHIP AREA/SLICK AREA = 0.875
FORWARD SPEED = 2.25 ft/sec
0.40 0.60 0.80 1.00 1.20 1.40
HORIZONTAL CONVEYOR VELOCITY/FORWARD SPEED
FIGURE 40 - HARVESTING CONVEYOR PERFORMANCE
14?
-------
100
Q
UJ
O£
UJ
>
o
U
UJ
cz
90
80
70
60
50
40
30
20
10
0
1.5mm SLICK THICKNESS
0.5 mm SLICK THICKNESS
\
OPEN SYMBOLS = 1 .5 mm SLICK THICKNESS
SOLID SYMBOLS = 0.5 mm SLICK THICKNESS
30 sec.
°
RESIDENCE TIME:
CONVEYOR ANGLE: 45
RECOVERY SPEED: 2.25 ft/sec.
NO. 4 OIL
-h-— -4
NOMINAL SORBENT
COVERAGE
O 87^
D 65*
l
BUNKER "C
100
1000
10,000
VISCOSITY, cm /sec
FIGURE 41 - OIL RECOVERY VS. VISCOSITY (CALM WATER).
00
-------
Q
5
u.
Q
O
u
O
100
90
80
70
60
50
40
30
20
10
_ 0.5 mm SLICK THICKNESS
e-
a
1 .5 mm SLICK THICKNESS
NOMINAL SORBENT
COVERAGE
A 107*
O 87*
RESIDENCE TIME: 30 sec.
CONVEYOR ANGLE: 45°
RECOVERY SPEED: 2.25 ft/sec.
OPEN SYMBOLS = 1.5 mm SLICK THICKNESS
SOLID SYMBOLS = 0.5 mm SLICK THICKNESS
20
100
1000
10,000
100,000
VISCOSITY CM* / SEC
FIGURE 42 - PERCENT OIL IN RECOVERED FLUID VS. VISCOSITY ( CALM WATER)
-------
f—
U1
o
Q
LU
O
u
LU
100
90
80
70
60
50
40
30
20
10
RESIDENCE TIME: 30 sec.
CONVEYOR ANGLE: 45°
RECOVERY SPEED: 2.25 ft/sec.
OPEN SYMBOLS = 1.5 mm SLICK THICKNESS
SOLID SYMBOLS = 0.5 mm SLICK THICKNESS
NOMINAL SORBENT
COVERAGE
A 107$
O 87^
O 65%
J I
100
1000
VISCOSITY, cm2/sec
10,000
FIGURE 43 - OIL RECOVERY VS. VISCOSITY (WAVES)
-------
I—
u_
Q
UJ
Q£
UJ
O
u
UJ
z
_J
o
100
90
80
70
60
50
40
30
20
10
0
P_ 1.5 mm SLICK O A
- o 0
A
~ • .__—• — —' —
_ A
NOMINAL SORBENT
- COVERAGE
A
o
RESIDENCE TIME:
107*
87*
65*
30 sec.
— CONVEYOR ANGLE: 45°
RECOVERY SPEED:
2.25 ft/sec.
— OPEN SYMBOLS = 1 .5 mm SLICK THICKNESS
SOLID SYMBOLS = 0.5 mm SLICK THICKNESS
1
20
100
1000
VISCOSITY CM /SEC
10,000
100,000
FIGURE 44 - PERCENT OIL IN RECOVERED FLUID VS. VISCOSITY (WAVES)
-------
and 44 give the percent oil in the recovered fluid. Fig-
ures 4l and 42 are for calm water and Figures 43 and 44 are
for waves of 3 inch height and 4 foot length. These figures
are based on a sorbent residence time of 30 seconds. It
should be noted that the tests were carried out with a two-
dimensional section of the actual conveyor at prototype speeds,
oil type and slick thicknesses with the prototype sorbent.
Consequently, the results are applicable to the actual unit
and scaling of the data is not required. The performance in
terms of the percent of oil recovered and the percent oil
in the recovered fluid is better in a slick of 1.5 mm thick-
ness than in a slick of 0.5 mm thickness. This is as ex-
pected since the thicker slick provides more flow area into
the sides of the chips and the larger volume of oil recovered
tends to improve the percent oil in the recovered fluid. The
percent recovery is better in waves because the waves agitate
the chips causing them to contact a higher percentage of the
surface. Figure 43 presents data which indicates that with
small waves and a uniform sorbent distribution, the actual
unit will recover over 90 percent of the oil in 0.5 mm and
1.5 mm slicks for oil viscosities up to 10,000 cm2/sec. The
8 to 10 percent of the oil which is not recovered is due to
the portions of the slick not contacted with sorbent, losses
from chips on the harvesting conveyor and oil washed off the
harvesting conveyor. Tests indicate that about 15 percent
of the oil not recovered is washed off the harvesting con-
veyor and less than 5 percent is lost from the chips on the
harvesting conveyor. Operation in waves also increases the
amount of water recovered due to the agitation of the sorbent
chips. Figure 44 indicates that in small waves the require-
ment for 90 percent oil in the recovered fluid can be satis-
fied in a 1.5 mm slick but not in a 0.5 mm slick. A further
analysis of the water recovery rate Is presented in a fol-
lowing paragraph.
The system performance in terms of percent oil recovery as
a function of sorbent residence time is presented in Fig-
ure 45 for calm water and Figure 46 for waves. The important
conclusion from these two figures is that there is little
or no increase in the percent recovery for residence time in
excess of 30 seconds for the slick thicknesses and oils
tested. The percent of oil recovered does not tend to de-
crease rapidly until the residence time is reduced to 10
seconds or less.
152
-------
LU
O
u
UJ
O£
O
100
90
80
70
60
50
40
30
20
10
1.5 mm SLICK
0.5mm SLICK THICKNESS
10
45° CONVEYOR ANGLE
2.25 FT /SEC RECOVERY SPEED
O # 4 OIL AT 20° C
a f 6 OIL AT 20° C
OPEN SYMBOLS = 1.5 mm SLICK THICKNESS
SOLID SYMBOLS = 0.5 mm SLICK THICKNESS
87$ NOMINAL SORBENT COVERAGE
SYMBOLS WITH TAILS = 107$ NOMINAL COVERAGE
I I I I I I I I I
20
30
40
50
60
RESIDENCE TIME SEC
FIGURE 45 - OIL RECOVERY VS. RESIDENCE TIME (CALM WATER)
153
-------
100
90
80
70
UJ
O
u
=! 50
O
v-
40
30
20
10
0
45° CONVEYOR ANGLE
2.25 FT /SEC RECOVERY SPEED
O * 4 OIL AT 20° C
D # 6 OIL AT 20° C
OPEN SYMBOLS = 1.5 mm SLICK THICKNESS
SOLID SYMBOLS = 0.5 mm SLICK THICKNESS
67% NOMINAL SORBENT COVERAGE
SYMBOLS WITH TAILS = 107$ NOMINAL COVERAGE
I I I I I I I I II I
10
20 30 40
RESIDENCE TIME SEC
50
60
FIGURE 46 - OIL RECOVERY VS. RESIDENCE TIME (WAVES)
154
-------
The percent oil recovery as a function of nominal sorbent
coverage is presented In Figure 47. Data are presented for
calm water and waves for residence times of 15 and 30 seconds
The nominal sorbent coverage is defined as the total projected
area of the sorbent chips divided by the area of the slick
Even with a 100 percent nominal sorbent coverage, small local
areas of the slick are not covered because some of the sor-
bent chips will overlap each other. Small waves counter this
effect by moving the chips around on the surface. This wave
action also reduces the required nominal sorbent coverage for
maximum oil recovery. For example, Figure 47 indicates that
at 30 seconds residence time in calm water, a 90 percent
coverage is required for maximum oil recovery. In waves,
maximum oil recovery can be obtained with between 70 and'80
percent coverage.
The basic format used in Figure 47 is suitable for synthesizing
data presented in previous figures in a form convenient for
system design studies. The result is a cross plot which pre-
sents percent oil recovery as a function of residence time and
percent sorbent coverage. Several plots of this type could
be constructed to cover the range of oil viscosity and slick
thicknesses of interest. An example is presented in Figure
48. This figure presents percent oil recovery in waves and
is appropriate for slick thickness from 0.5 mm to 3 mm and
oil viscosities less than or equal to 6000 cm2/sec.
Previous figures (42 and 44) presented data on the percent
of oil in the recovered fluid. It was observed in the tests
that the water recovered was from two sources. Some of the
water is absorbed in the sorbent along with the oil or re-
mains on the surface of the sorbent in the form of droplets.
The remainder of the water is entrained by the harvesting
conveyor and sprayed onto the collected sorbent. It-'is im-
portant for prototype predictions to separate these two
sources of recovered water since the first is a function of
the amount of sorbent and the second is a function of the
width and speed of the harvesting conveyor. Figure 49 pre-
sents data on the water recovered which is absorbed in the
sorbent material itself. These data are scattered but do
indicate that high viscosity oil in the sorbent tends to re-
duce the amount of water recovered. The agitation of small
waves tends to increase the amount of water recovered. There
is no consistent indication that more water is recovered by
the sorbent in thin slicks (0.5 mm) than in thicker slicks
(1.5 mm). Figure 50 presents data on the rate at which the
155
-------
100
90
80
70
60
50
40
30
20
10
0
1.5 mm OF #4 OIL
D 15 SEC RESIDENCE TIME
030 SEC RESIDENCE TIME
SOLID SYMBOLS = OPERATION IN WAVES
45° CONVEYOR ANGLE
2.25 FT / SEC RECOVERY SPEED
I I i I J I
I !
10
20 30
40
50 60 70 80 90 100 110
NOMINAL SORBENT COVERAGE - PERCENT
FIGURE 47 - OIL RECOVERY VS. NOMINAL SORBENT COVERAGE
156
-------
120
110
100
90
>. 80
a:
HI
8 70
UJ
O 60
»—
Z
ua en
U 50
80 PP! POLYURETHANE CHIPS 3"x3"x 3/8"
SLICK THICKNESS 0.5-3mm
OIL VISCOSITY < 6000 CM2/SEC
LU
a.
40
30
20
10
0
10 20 30 40 50 60 70 80 90 100 110 120
PERCENT SORBENT COVERAGE
FIGURE 48 - SORBENT OIL RECOVERY PERFORMANCE IN WAVES
157
-------
o
to
0.03
NOMINAL SORBENT COVERAGE
A 107*
O 87*
OPEN SYMBOLS = CALM WATER
SOLI D SYMBOLS = WAVES
WITH TAILS = 0,5 mm SLICK
WITHOUT TAILS = 1.5 mm SLICK
0.02 —
WAVES
on
O
0.01
20
O
A
CALM WATER
I
100
1000
VISCOSITY CM* /SEC
10,000
100,000
FIGURE 49 - WATER RECOVERED BY SORBENT VS. SLICK VISCOSITY
-------
u
iu
to
u
UJ
to
to
z
o
o
LU
Z
8.0
7.0
6.0
5.Ox 10
4.0
3.0
2.0
KOx 10
0.0
.-4
\
X30° CONVEYOR ANGLE
-45° CONVEYOR ANGLE
234
CONVEYOR SPEED FT / SEC
FIGURE 50 - WATER ENTRAPMENT RATE DUE TO CONVEYOR VS. CONVEYOR SPEED
155
-------
harvesting conveyor entrains water and sprays it as droplets
onto the collected sorbent. It should be noted that this rate
is greatly reduced by increasing the conveyor angle to 45 de-
grees. This is a major reason for selecting as steep a con-
veyor angle as possible.
Summary
As a result of the work carried out under this task an optimum
conveyor angle and speed ratio has been selected and the data
necessary-, to predict the performance of the system under a
wide range of conditions have been collected and analyzed.
These data indicate that the oil recovery can range from 90 to
95 percent over a wide range of oil types and slick thicknesses
The percent oil in the recovered fluid will meet the EPA goal
of 90 percent for slick thicknesses of 1.5 mm.
160
-------
APPENDIX D
SORBENT REGENERATION SYSTEM DEVELOPMENT
The objectives of this task were:
Develop performance data for a converging belt*type
sorbent regenerator as a function of the design
parameters.
Determine the effects of repeated cycles on the
sorbent material. ' '
Identify and resolve mechanical problems prior to
the construction of a prototype system.
These objectives were satisfied by designing, building and
conducting tests with a sorbent regenerator test apparatus.
The function df the sorbent regenerator is to squeeze the
recovered oil out of the sorbent chips so that the chips are
ready for broadcasting. In a typical case using 3 in. by
3 in. x 1/4 in. sorbent chips, a cubic foot of sorbent ma-
terial will recover about 2 gallons of oil. The sorbent re-
generator must remove the oil from the sorbent and maintain
the density of the sorbent at a low enough value for broad-
casting. Thus the performance of the sorbent regenerator
can be defined in terras of the density of the regenerated
sorbent after the removal of 2 gallons of oil per cubic foot.
The parameters which influence the density of the regenerated
sorbent are:
Squeezing Force
Oil Viscosity
Sorbent Loading
Sorbent Type
Belt Speed
Test Apparatus and Procedures
A converging belt sorbent regenerator test apparatus was de-
signed and built. Figure 51 shows an outline drawing of this
apparatus and Figure 52 shows a photograph of it set up for
testing ' The test apparatus consists of two 18-inch wide
converging conveyor belts which run between two pairs of
squeezing rollers. The lower belt is an open steel mesh
161
-------
AIR CYLINDERS
STEEL BELTS
BELT
TENSIONER
\
i — r
o
i^vfts^v^a
,
J11.A .-./ /
v IS>I
*' . r
OIL COLLECTING^-"
PAN SPOUTS
UPPER BELT
SPROCKET
CHAIN
LOWER BELT SPROCKET
AND BELT DRIVER
"V IDLER AND CHAIN
TENSIONER
CM
SCALE 1:11.4
FIGURE 51 - SORBENT REGENERATOR TEST APPARATUS
-------
FIGURE 52 - SORBENT REGENERATOR TEST APPARATUS SET-UP
163
-------
(Type B-60-32-14 Balanced Belting) and the upper belt is a
similar material covered with segmented neoprene pads. The
two belts are driven by a chain drive system in which the
speed can be adjusted by changing sprocket size. The squeez-
ing pressure is applied to the upper roller in each pair by
an air cylinder acting through a yoke. The squeezing pres-
sure can thus be adjusted by simply adjusting the air pressure.
Oil collecting pans are located under both legs of the lower
belt. The recovered oil flows by gravity out of the collect-
ing pan spouts to a storage barrel.
In order to avoid scaling problems with the data and to
identify mechanical problems, the test apparatus was designed
to be near full scale for the 3000 GPH recovery unit. It is
intended to scale the system up for the 10,000 GPH unit by
simply making it wider.
The tests were conducted by taking a known volume of sorbent
material and saturating it with between 1.5 and 2 gallons of
oil per cubic foot. The sorbent was then run through the
regenerator and its density determined by weighing. This
process was repeated until the density of the regenerated
chips reached an equilibrium value. The tests were conducted
for a range of belt speeds, squeezing force, oil type, sor-
bent type and .sorbent loading. The sorbent loading is defined
as the volume of sorbent per unit belt area and has the di-
mensions of feet. The sorbent loading times the belt width
times the belt speed gives the sorbent regeneration rate in
cubic 'feet/second. The effect of more than two pairs of
squeezing rollers was simulated by quickly recycling the re-
generated sorbent.
Preliminary bench tests, during the design phase, indicated
that a squeezing force of 50 pounds per inch of belt width
would be satisfactory. However, preliminary testing showed
that this was not sufficient to reduce the sorbent density
to the desired range (i.e. 5 to 7 lbs/ft3). As a result it
was necessary to increase the shaft diameters on the squeezing
rollers to allow testing at higher forces.
System Performance
The performance of a converging belt sorbent regenerator, in
terms of the density of the regenerated sorbent, is presented
in Figure 53 thru 55 as a function of the test parameters.
164
-------
15.0
10.0
UJ
CD
a:
o
to
a
UJ
UJ
o
Ul
0£
U_
O
t
UJ
O
5.0
SORBENT LOADING P*3/Ff2 SORBENT VOLUME
~ 5 j~ BELT AREA
O 0.092 Ft/Ft*
V 0.233 Ft3/Ft2
D 0.428 Ft3/Ft2
A 0.316Ft3/Ft2
OPEN SYMBOLS 0.43 Ft/Sec BELT SPEED
SOLID SYMBOLS 1.00 Ft/Sec BELT SPEED
50 100 150 200
SQUEEZING FORCE Lb/ln OF BELT WIDTH
FIGURE 53 - REGENERATED SORBENT DENSITY VS. SQUEEZING
FORCE (80 PPI FOAM) NO. 2 HEATING OIL
16 5S
-------
15.0
10.0
Z
LU
09
O
to
Q
UJ
LJJ
O
O
£
Z
IU
O
5.0
s
c
s
ORBENT LOADIh
O 0.233 Ft3/
D 0.362 Ft3/
A 0.31 6 Ft3/
)PEN SYMBOLS (
OLID SYMBOLS
DPY 5
OR Ft3 /Ft2 SORBENT VOLUME
Jii l-t ,rt BE
Ft2
Ft2
Ft2
5.43 Ft/Sec BELT
1 .00 Ft/Sec BELT
n
-------
15.0
10.0
o;
O
to
O
z
LU
O
LL.
O
z
LU
Q
5.0
— — -.-,
'"V
"
D,
SORBENT LOADI
_
80 PPI FOAM
^»
*--*.
O
Mr,F,3/F,2 so
O 0.180Ft3/Ft2
OPEN SYMBOLS 0.43 Ft/Sec BELT
SOLID SYMBOLS 1 .00 Ft/Sec BELT
ALL DATA AT 22° C UNLESS NOT
_
— — »^>
RBENT VOLUME
BELT AREA
SPEED
SPEED
ED
,
10°C
O
/>.18
111
•
4 ROLLERS
^DRY SORBENT
0 50 100 150 200 25
SQUEEZING FORCE Lb/ln OF BELT WIDTH
FIGURE 55 - REGENERATED SORBENT DENSITY VS. SQUEEZING
FORCE (30 PPI FOAM) BUNKER "C"
167
-------
Figure 53 is for 80 PPI roam and No. 2 heating oil. Figure
54 is for 80 PPI foam and an oil product made by mixing 70
percent Bunker "C" and 30 percent No. 2 heating oil. Fig-
ure 55 is for Bunker "C" and presents points for both 80
and 30 PPI foam. The viscosities of the oil products used,,
as a function of temperature, are given in Table 36. These
figures indicate that the density of the regenerated sorbent
is most sensitive to squeezing force, oil viscosity and sor-
bent type (80 or 30 PPI). The sorbent loading also has a
significant effect on the density. The results were basically
insensitive to belt speed in the range from 0.4 to 1.0 ft/sec.
TABLE 36
Viscosity of Test Oil Products
Temperature
degrees, C
30
20
10
5
Viscosity: Centipoise
No. 2
2.2
2.6
3.2
4.0
30 percent No. 2
70 percent Bunker "C"
120
800
2300
12000
Bunker C
2120
4500
20000
153000
Figure 56 presents a summary of regenerator performance
as a function of oil viscosity for a typical design point.
The design point parameters are a squeezing force of 220
Ib/in. a sorbent loading of 0.25 ftfft2 and a belt speed
of 1.0 ft/sec or less. This figure shows the need for
using 30 PPI foam when dealing with oil viscosities in
excess of 1000 ops. At this squeezing force it is possible
to regenerate sorbent which contains oil with viscosities
up to 20,000 cps. However, the density of the regenerated
sorbent increases above the desired level and heating will
be required to get satisfactory flow in the oil collecting
pans. Figure 56 also illustrates the benefits of In-
creasing the number of squeezing rollers.
Mechanical Details
The tests conducted with the heavy oils indicated some
tendency for the sorbent chips to stick to the lower belt
and to be carried into the lower collecting pan. A simple
168
-------
10.0
o
UJ
<
O£
UJ
UJ
o
a.
o
UJ
a
5.0
SQUEEZING FORCE =220 Lb/ln
SORBENT LOADING =0.25 Ft3/Ft*
BELT SPEED ^ 1.0 ft/sec
O 80 PPI FOAM
D 30 PPI FOAM
2 SQUEEZE ROLLERS UNLESS NOTED
X
80 PPI FOAM
"*-30 PPI FOAM
4 SQUEEZE
ROLLERS
SQUEEZE
ROLLERS
DRY SORBENT
1000
10,000
100,000
VISCOSITY C / Sec
m
FIGURE 56 - REGENERATED SORBENT DENSITY VS. VISCOSITY
-------
mechanical scraper is not sufficient to remove these chips.
As a result, a rotating brush driven by the belt drive should
be installed on the prototype. This should remove the chips
without difficulty. Also, since some sorbent chips will find
their way into the collecting pans, provision should be made
for periodic cleaning of these pans.
It was also noted during early tests that sorbent chip's
were caught in the drive, chains at the edges of the upper
and lower belts and ultimately ground up. Simple wooden
chain guards were added to the test apparatus and solved
the problem. Similar guards made of a more durable material
should be fitted on the prototype device.
In the test apparatus the upper and lower belts were the
same material and the upper belt was made solid by attaching
neoprene pads to it. This will not be a satisfactory ar-
rangement for the prototype regenerator. The upper belt
should be simple reinforced neoprene belt or a wire mesh
belt of the type used impregnated with neoprene.
The air cylinders used to provide the squeezing force gave
excellent, service and provide a simple and reliable means
of applying and controlling the squeezing force. The shafts
in the squeezing rollers should be designed to withstand
the loads applied by a piece of drift wood or debris going
through the regenerator. For the 3000 GPH unit 1-1/2 inch
diameter high strength steel shafts should be satisfactory.
An endurance test was conducted to determine the effects of
repeated regenerations on the sorbent material. These tests
were carried out by saturating a group of chips (about 20)
and running them through the sorbent regenerator test ap-
paratus at a high enough squeezing force to reduce their
density to 6 Ib/ft3. This process was repeated 100 times
for 80 PPI chips using a mixture of 90 percent Bunker "C"
and 30 percent No. 2 oil and for 30 PPI chips using Bunker
"C". Figure 57 presents a photograph of chips that were
regenerated 1, 50 and 100 times. There was no significant
degradation of the sorbent chips after 100 cycles.
Results
The important results of this task can be summarized as
follows:
170
-------
Heavy numbers indicate the number of times
sample sorbent chips were regenerated
7 8 i I!
FIGURE 57 - SORBENT CHIPS AFTER ENDURANCE TEST
171
-------
A converging belt sorbent regenerator will satis-
factorily regenerate sorbent chips at the volume
rates required.
The density of the regenerated chips will be about
6 lb/ft3 for a squeezing force of 220 Ib/in. a
sorbent loading of 0.25 fts/ft2 and a belt speed
of 1 ft/sec.
In order to maintain a 6 lb/ft3 regenerated sorbent
density, 30 PPI foam should be used for oil vis-
cosities in excess of 1000 cps.
An oil viscosity of about 20,000 cps at the time
of regeneration is the practical upper limit be-
cause of the excessive density of the regenerated
sorbent.
The sorbent material can be cycled 100 times without
significant degradation.
172
-------
APPENDIX E
MODEL TEST OP A 1/4 -SCALE MODEL RECOVERY PLATFORM
The objectives of this task were:
1. Demonstrate that a continuous stable sorbent
broadcast and recovery cycle is possible.
2. Determine any adverse effects of waves and forward
speed on the continuous broadcast and recovery cycle,
3- Determine the towing resistance and stability of the
recovery platform at the deployment and operating
draft.
The objectives of this task were carried out by means of
tests on a 1/4 scale model of the recovery platform in the
HYDRONAUTICS Ship Model Basin, (HSMB®). The model was
equipped with an operating broadcasting system and harvesting
conveyor. No squeezing system was fitted so no attempt was
made to actually recover oil.
The most important results of the test program were the
qualitative observations with respect to the feasibility of
a continuous., stable, broadcasting and recovery cycle under
different operating conditions. Quantitative data were ob-
tained on the towing resistance of the recovery platform
under different conditions. A description of the model, the
test procedure and the results are presented in the following
paragraphs.
Description of Model and Test Facilities
The basic concept for the recovery platform was developed
in the original HYDRONAUTICS, Incorporated Sorbent Oil Re-
covery System Proposal. Subsequently, preliminary efforts
showed that the basic concept and general proportions were
still valid. As a result, the model design was based on
the original concept and was sized to be a 1/4 scale model
of a system intended to recover 10,000 gallons per hour.
A 1/4 scale aluminum model of the recovery platform equipped
with an operating sorbent broadcasting and recovery system
was constructed. Figure 58 presents an arrangement drawing
for the model with its dimensions.
173
-------
n
/^
S ^ ! ^ "
-------
The broadcasting system on the model was based on the work
reported in Appendix B. The movable parallel plate nozzle
was a scaled down version of the one tested in Task 2, and
equipped with a simple mechanical drive. No attempt was .
made to scale the actual blower. Rather, the blower avail-
able from the small scale experiments conducted under Task 2
was used. The air flow necessary for good sorbent distribu-
tion was obtained by choking the blower inlet.
The harvesting conveyor angle of 45 degrees was based on
the 'results of Task 3. It was not practical to obtain a
conveyor belt material that was a 1/4 scale model of that
used in Task 3. However,, it was felt that this was not im-
portant. Instead, a light weight wire belt with 1/4 inch
high wooden flights was used. The harvesting conveyor was
driven with a-variable speed drive so that the linear con-
veyor speed could be adjusted to equal the towing speed of
the platform. A transfer conveyor was fitted between the
aft pontoons of the platform to collect the sorbent chips
from the harvesting conveyor and feed them to the broad-
casting system blower. The transfer conveyor was also
fitted with a variable speed drive. Figure 59 presents
photographs of the mechanical systems described above.
The sorbent material was simulated with a white color,
closed cell foam. This foam was cut to 1/8 in. thick, 3/4
in. by 3/4 in. squares.
The tests were carried out in the HYDRONAUTICS, Ship Model
Basin at HYDRONAUTICS, Incorporated. This towing tank is
equipped with a towing carriage with a top speed of 20 ft/sec,
and with Instrumentation to measure forces, motions and wave
height. A plunger type wavemaker is fitted at one end of
the towing tank which Is capable of generating regular and
long crested irregular waves. The dimensions of the towing
tank are:
Length 308 feet
Width 25 feet
Depth 13 feet
Test Procedures
The towing tests were carried out with the model towed be-
hind the towing carriage on about 30 feet of towline. The
towline ran from a 60 degree bridle between the forward
175
-------
Broadcasting Nozzle and Drive
Transfer Conveyor and Drive
Harvesting Conveyor and Drivs
FIGURE 59 - MECHANICAL DETAILS OF SORBENT RECOVERY
PLATFORM MODEL
176
-------
pontoons, under the carriage, to a force measuring block
gage mounted on a post at the forward end of the carriage.
A sonic wave height probe was mounted on the forward end
of the towing carriage to measure the height of the incident
waves. A capacitance type wave height probe was mounted
outboard on the aft port pontoon at the same longitudinal
location as the lower end of the harvesting conveyor. The
purpose of this probe was to measure the relative motion be-
tween the waves and the recovery platform. The outputs from
the force gage and the two wave probes were recorded on paper
strip charts. Also the output of the force gage was inte-
grated to'give the average drag.
At the start of a typical test, the sorbent material chips
were dumped into the center of the recovery platform, as
shown In Figure 60, and the harvesting conveyor, transfer
conveyor and broadcasting systems were started'. The towing
carriage then towed the model down the tank at the desired
speed while the towing drag, wave height and relative motion
were recorded. Also, pictures and observations were made
of the sorbent distribution and recovery. After the first
test run of the day, the sorbent material was simply allowed
to remain in the recovery platform.
In addition to the towing tests, some tests were carried
out with the model pushed by the towing carriage. This was
to simulate the type of operation in which the recovery
platform Is pushed by a barge. The pushing was done with
3 foot long steel arms pivoted at both ends. In these tests,
no measurements were made of pushing force. This test set-
up is shown in Figure 60.
Towline Drag and Towing Stability
Towing tests of the sorbent recovery platform were carried
out over a range of speeds and sea states for the deployment
and nominal operating conditions. The drafts and salt water
displacements for these conditions are:
Condition Draft DlB£lacejnent
Deployment 2' - 0" ' 57,500 Ibs
Nominal Operating 3' - 0" 80,000 Ibs
177
-------
Adding Sorbent Material at Start of Test
Arrangement of Pushing Arms
FIGURE 60 - TEST SETUP FOR SORBENT RECOVERY PLATFORM
MODEL
178
-------
Tests were conducted in long crested irregular waves cor-
responding to Sea State 1 (significant wave height =1.1
ft) and a low Sea State 3 (significant wave height = 2.6 ft).
The significant wave height is defined as the average height
of the 1/3 highest waves.
The towline drag data were expanded to the full scale value
using Proude scaling. Because of the low scale ratio
(\ = 4.0) and the bluff shape of the body, no correction was
taade for changes in the skin friction in expanding the drag
data to full scale. The resulting towline drag for calm
water and waves over the range of speeds expected for sorbent
recovery are presented in Figure 6l for the deployment con-
dition. Figure 62 presents the towline drag in calm water
for the deployment condition up to the highest practical
towing speed. This speed is about 9-5 knots and is limited
by excessive sinkage the trim of the platform. Figure 63
presents a photograph of the model under tow at a speed
equivalent to 9.5 knots. Figure 64 presents the towline
drag for the nominal operating condition over the expected
range of speeds for sorbent recovery in calm water and waves.
The towline drag data for calm water were also reduced to
nondimensional form so that they can be used to estimate
the drag of similar platforms of different size. The results
are presented in Figure 65 in terms of a towline drag coef-
ficient as a function speed length ratio. The drag coeffi-
cient is defined as:
where
D = towline drag, ilbs
p = mass density of water, slug/ft3
V = displaced volume of platform, ft3, and
u = towing speed, ft/sec.
The speed length ratio is defined as:
Speed length ratio =
179
-------
7000
O
Q
O
z
.O
6000-
5000 -
4000 -
3000 -
2000
1000 -
DRAFT = 2' - 0"
DISPLACEMENT = 57,500 Ibs
CONDITION SEA STATE SIG. WAVE
HEIGHT
O CALM
D S.S. 1 1.1 ft
A S.S. 3" 2.6ft
0
4.0 6.0
TOWING SPEED ft/sec
10.0
0
1
3
KNOTS
FIGURE 61 - SORBENT RECOVERY PLATFORM DEPLOYMENT DRAFT
TOWLINE DRAG Vs SPEED AND SEA STATE
180
-------
50,000
T
T
40,000
30,000
u>
_D
O
<
C£
O
O
Z 20,000
O
10,000
DRAFT = 2' - 0"
DISPLACEMENT = 57,500 Ibs
CALM WATER
MAXIMUM PRACTICAL
TOWING SPEED-
^T 4.0 6~TOO 10.0 12.0 14.0 16.0 18.0 20.0
TOWING SPEED ft/sec
0 1 2 3456 7 8 9 10 11 12
FIGURE 62 - iSORBENT RECOVERY PLATFORM DEPLOYMENT DRAFT
TOWLINE DRAG Vs SPEED
1.81
-------
FIGURE 63 -
SORBENT RECOVERY PLATFORM UNDER TOW AT
9.5 KNOTS
182
-------
7000
6000 -
DRAFT = 31 - 0"
DISPLACEMENT = 80,000 Ibs
0
I
2.0
i
4.0 6.0
KNOTS
i I i
8.0
I
10.0
1
0
1
2345
TOWING SPEED ft/sec
FIGURE 64 - SORBENT RECOVERY PLATFORM OPERATING DRAFT
TOWLINE DRAG Vs SPEED AND SEA STATE
183
-------
u
8
o
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
\
\
\
0.5
3 FT DRAFT
0.2
0.5
1.0
SPEED LENGTH RATIO
(SPEED IN KNOTS)
V,
-
2 FT DRAFT
I t
1.5
2.0
[LENGTH OF AFT PONTOON ( ft )J t
FIGURE 65 - SORBENT RECOVERY PLATFORM DRAG COEFFICIENT VS. SPEED
LENGTH RATIO
-------
where
VK = towing speed, knots, and
L = length of aft pontoon, ft.
The length of the aft pontoon was selected for the length
scale in the speed length ratio since it is the longest
continuous component in the recovery platform. The rise
in the drag coefficient above a speed length ratio of 1.2
indicates that energy is being dissipated in surface waves
as well as in form drag.
The towing stability was checked by giving the model a
lateral displacement as it was being towed down the tank.
The model rapidly returned to its original position directly
behind the towing point on the carriage. This indicates the
platform has directional stability when under tow and thus
will tow satisfactorily without yawing.
Sorbent Broadcasting and Recovery
The sorbent broadcasting and recovery tests were conducted
for conditions equivalent to full-scale speeds of 3.0 and
6.0 ft/sec in calm water,, Sea State 1 and a low Sea State 3.
The data obtained in these tests were qualitative in nature
and consisted of observations of performance supported by
still and motion pictures. Typical photographs from the
test program are presented in Figures 66, 67 and 68. All of
the photographs are for a speed equivalent to 3 ft/sec full-
scale which is the nominal design speed. The range of sea
conditions from calm to a low sea state 3 are covered by the
photographs. The important observations from the broad-
casting and recovery tests are presented below:
1. The sorbent distribution within the system is
stable. The tests showed that concentrations
of sorbent material in the system tend to become
spread out,, and that ultimately the sorbent dis-
tribution throughout the system tends to become
uniform. This is because of the broadcasting
system which spreads the sorbent over a sub-
stantial (about 20 feet) longitudinal distance
as well as lateral distance. As a consequence
of the stability of the sorbent distribution,
185
-------
FIGURE 66 - RECOVERY PLATFORM OPERATING IN CALM WATER AT 3.0 ft/sec
-------
Broadcasting
Harvesting and Transfer Conveyor
FIGURE 67 - RECOVERY PLATFORM OPERATING IN SEA STATE
1 AT 3 ft/sec
187
-------
Recovery Platform Being Towed
I
Recovery Platform Being Pushed
FIGURE 68 - RECOVCK,' ?' *TFQRM OPERATING IN SEA STATE 3
AT 3 ft/sec
-------
it is possible to maintain a continuous sorbent
broadcasting and recovery process without con-
stant adjustments by the operating crew.
2. The sorbent broadcasting and recovery"operation
is not degraded by waves up to a low sea state
3. The tests showed that the broadcasting and
recovery processes are not greatly effected by
waves up to the height tested. The freeboard
and draft of the platform are sufficient to
prevent sorbent loss due to waves breaking over
or washing under the aft pontoons and harvesting
conveyor. The relative motion probe showed that
relative motions in short waves do not greatly
exceed the wave height and in moderate and long
waves, the platform contours the waves with only
small relative motions. This is illustrated by
relative motion data obtained in regular waves
which are presented in Figure 69. These data
show that relative motions will be small in
wavelengths equal to or greater than the plat-
form overall length.
3. The broadcasting nozzle developed in Task 2
will provide a uniform distribution of sorbent
over the width of the recovery platform. These
tests further confirmed the validity of the
broadcasting nozzle concept developed in Task 2.
The quality of the distribution pattern across
the width of the recovery platform is illustrated
in Figures 66 through 68.
4. The sorbent material can be recovered by a
harvesting conveyor half the overall sorbent
pattern width. During the tests, there were
no indications that the sorbent material would
form a bridge or plug across the inlet to the
harvesting conveyor. In waves, the inboard
sides of the aft pontoons.reflect waves back to
the center of the recovery bay. These reflected
waves tend to locally herd the sorbent material
to the center of the harvesting conveyor. Thus,
in waves, there is no chance that the sorbent
material could form a plug across the inlet to
the harvesting conveyor.
189
-------
2.0
X
O
LU
X
<
I
§'•"
I—
o
25
T
T
FORWARD SPEED = 3.0 ft/sec
RELATIVE MOTION AT FOOT OF HARVESTING CONVEYOR
50 75
WAVE LENGTH , feet
100
125
150
FIGURE 69 - SORBENT RECOVERY PLATFORM RELATIVE MOTION IN
REGULAR WAVES
190
-------
Results
The important results of this developmental program can be
summarized as follows:
1. The test program showed that a continuous sorbent
broadcasting and recovery operation with a uniform
stable distribution of sorbent material can be
achieved.
2. The sorbent broadcasting and recovery operation
was not degraded by waves up to a low sea state 3
and forward speeds up to 6 ft/sec.
3. The broadcasting nozzle concept developed in
Task 2 provided a uniform transverse distribution
• of sorbent material.
4. The sorbent material did not show any tendency to
plug the inlet to the harvesting conveyor.
5. The recovery platform can be towed up to a speed
of 9-5 knots in the deployment condition.
6. The recovery platform is directionally stable
under tow.
191
*US. GOVERNMENT PRINTING OFFICE: 1973 514-154/259 1-3
-------
1
Accession Number
w
5
j Subject Field & Group
05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
organization
TTTrpl'O^MlTA Timi-/Nn -i- •
— — — -*- •«••—- j -J-i* Vw-j. £SW4. <-* U V\A
Pindell School Road, Laurel, Maryland 20810
Development and Preliminary Design of a Sorbent-Oil
Recovery System
1 Q Authors)
IWM 1 ~\ p-p TH
rLL X -L OJL , Ju
Stephens,
Ricklis,
1
L.
J.
16
21
Project Designation
15080-HEV
Note
22
Citation
Environmental Protection Agency report
number, EPA-R2-73-156, January 1973.
23
Descriptors (Starred First)
*
Oil Spills, *Water Pollution Treatment, *Design Criteria,
^Laboratory Tests
25
Identifiers (Starred First)
*Sorbent-Oil Recovery
27
Abstract
Abstract
A development program was completed and preliminary designs were pre-
pared for 3000 gallon/hour protected water and 10,000 gallon/hour un-
protected water Sorbent Oil Recovery Systems. The five phases in the
development program were: (l) characterization of the sorbent material,
(2) the development of the sorbent broadcasting system, (3) the de-
velopment of the harvesting conveyor and evaluation of overall re-
covery performance, (4) the development of the sorbent regeneration
system and (5) model tests of a 1/4-scale model recovery platform.
The development program showed that a continuous sorbent-oil recovery
system is feasible using 30 or 80 PPI polyurethane sorbent chips.
In one pass about 90 percent of the oil in a 1.5 mm slick can be re-
covered. The water content of the Recovered Fluid is less than 10
percent. The preliminary designs are presented with detailed de-
scriptions of the system components, operating procedures, and costs.
This report was submitted in fulfillment of Project Number 15080-HEV
and Contract Number 68-01-0066 under the sponsorship of the Office of
Research and Monitoring, Environmental Protection Agency.
Abstract'
I. Miller
Inc. Laurel, Md. (Miller-HYDRONAUTHS)
WR:(02 (REV. JULY 1969)
WRSIC
SEND, WITH COPY OF DOCUMENT,
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 20240
* SPO! 1870-389-930
------- |