EPA-R2-73-233
MAY 1973 Environmental Protection Technology Series
Development of A Mobile System
for Cleaning Oil-Contaminated Beaches
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
.. 3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-R2-73-233
May 1973
DEVELOPMENT OF A MOBILE SYSTEM FOR
CLEANING OIL-CONTAMINATED BEACHES
By
Francis X. Dolan
James P. Bowersox
Project 15080 FIG
Project Officer
Richard R. Keppler
EPA - Region 1
John F. Kennedy Bldg.
Boston, Massachusetts 02203
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.0.20402
Price $1.26 domestic postpaid or tl OPO Bookstore
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents
necessarily reflect the views and policies of
the Environmental Protection Agency, nor does
mention of trade names or commercial products con-
stitute endorsement or recommendation for use.
11
-------
ABSTRACT
A system has been developed for the restoration of oil-
contaminated beach sands. The method involves washing of
the sand in a high energy jet-contactor washer and separation
of the cleaned sand from the washing fluid in a conventional
solid-liquid cyclone. Separation and concentration of the
oil-water effluent from the washing process is also accom-
plished in cyclones. The two separate stages of this
process have been demonstrated on a pilot scale equivalent
to about 3 tons of wet, oil contaminated sand per hour.
The sand washing process has been shown capable of removing
over 99% of the contaminating oil from a simulated beach
sand. Oils used were No. 4 and No. 6 fuel oil at 4 to 8%
of the dry weight of the sand. The oil-water separation
tests yielded a highly enriched oil product stream with
less than 20% water, while the water removed from the
system was suitable for recycle to the sand washing system.
A conceptual design for a mobile beach-cleaning system
based on the processes studied is presented and is shown
to be feasible within the state-of-the-art .
This report was submitted in fulfillment of Project FIG,
Contract No. 14-12-830, under the sponsorship of the
Environmental Protection Agency.
111
-------
CONTENTS
Section Page
1 CONCLUSIONS 1
2 RECOMMENDATIONS 3
3 INTRODUCTION 5
4 BACKGROUND 7
5 EXPERIMENTAL PROGRAMS 15
6 RESULTS AND DISCUSSION 31
7 ACKNOWLEDGEMENTS 71
8 REFERENCES 73
9 NOMENCLATURE 75
10 APPENDICES 77
-------
FIGURES
1 OPERATING MODES FOR LIQUID-LIQUID CYCLONES 9
2 PARTICLE SIZE DISTRIBUTION CURVE FOR #50 SAND 13
3 SAND WASHER ASSEMBLY DETAIL 16
4 SAND CLEANING PILOT PLANT FLOWSHEET 19
5 SAND WASHING EQUIPMENT SKETCH 20
6 DORRCLONE TEST APPARATUS SCHEMATIC 27
7 TML DORRCLONE AND 6 INCH DORRCLONE OIL-WATER
SEPARATION FLOWSHEET 28
8 TWO STAGE OIL-WATER SEPARATION FLOWSHEET 29
9 PERFORMANCE OF SAND WASHING PILOT PLANT 39
10 PERFORMANCE OF SAND WASHING PILOT PLANT 40
11 PERFORMANCE OF SAND WASHING PILOT PLANT 41
12 PERFORMANCE OF SAND WASHING PILOT PLANT 42
13 PERFORMANCE OF SAND WASHING PILOT PLANT 43
14 PERFORMANCE OF SAND WASHING PILOT PLANT "AGED"
SAND/OIL TESTS 44
15 PERFORMANCE OF OIL-WATER SEPARATING CYCLONES 51
16 PERFORMANCE OF OIL-WATER SEPARATING CYCLONES 52
17 EFFECT OF FEED CONCENTRATION ON ENRICHMENT 54
18 EFFECT OF FEED CONCENTRATION ON ENRICHMENT 55
19 EFFECT OF PRESSURE DROP AND FEED CONCENTRATION
ON ENRICHMENT 56
vi
-------
Page
20 SAND WASHING PROCESS FLOWSHEET 64
21 BLOCK DIAGRAM OF SYSTEM ARRANGEMENT 67
22 BEACH RESTORATION PROCESS 69
VI1
-------
TABLES
Page
1 RESULTS OF SAND WASHING TESTS 34
2 EFFECT OF WATER TYPE ON PREDICTING ENRICHMENT 58
3 EFFECT OF FEED TEMPERATURE ON PREDICTING
ENRICHMENT 59
4 EFFECT OF OIL TYPE ON PREDICTING ENRICHMENT 60
5 EFFECT OF CYCLONE UNDERFLOW RECYCLE ON
PREDICTING ENRICHMENT 61
6 PREDICTION OF 6 INCH CYCLONE OPERATION 62
7 SYSTEM COMPONENTS 68
Vlll
-------
SECTION 1
CONCLUSIONS
The following conclusions have been reached as a result
of the work conducted in the course of this study.
Sand Washing and Separation
1) The ability to wash oil from contaminated beach sands
and to separate the sand from the washing fluid for
return to the beach has been demonstrated in a pilot
scale system.
2) The apparatus and procedures involved are straight-
forward and are within the state-of-the-art.
3) Requirements for power in the sand washer are small,
of the order of one horsepower for each ton of sand
processed per hour.
Oil-Water Separation
4) Oil water mixtures can be separated in 10 mm cyclones.
The results of the tests can be predicted well enough
for pilot plant design.
5) Underflow (heavy-phase stream) oil concentrations of 0.1%
can easily be obtained but overflow (light-phase stream)
concentrations above 80% cannot be achieved.
6) Test data from the 6 inch cyclone does not match trends
established by the 10 mm cyclones. The 6 inch data do
not fit any pattern and cannot be resolved without
further test work.
7) The small cyclones can be staged and product streams
routed to other cyclones for further processing. No
adverse effects were found from staging the cyclones.
8) The most important variable affecting the cyclone
performance is the split/ i.e. the ratio of the overflow
rate to the underflow rate. Feed oil concentrations,
water type, and temperature have only minor affects.
-------
FuuLl-Scale System
9^ Presently, no problems are seen regarding the extrapo-
lation of the laboratory data to the design of a full-
scale, beach-rated sand cleaning system. This presumes
the use of 10 mm cyclones for the oil-water separation
stage.
10) The operation of such a system should be straight-
forward, with small requirements for power, operating
personnel and supplies.
11) All of the equipment needed for the system is within
the state-of-the-art, and no new breakthroughs are
needed to make the system practical.
-------
SECTION 2
RECOMMENDATIONS
A three-pronged program is recommended for future work
with this beach cleaning concept.
1) Continue test work on oil-water separation in large
cyclones to determine if effluent oil streams with
less than 20% water can be attained. The applicability
of centrifuge, flotation, coalescing and filtration
systems to this separation should be studied, with the
objective of achieving less than 3% water in the final
oil product. Also investigate a wider variety of oil
types at this stage.
2) From the data, design, fabricate and test a complete
beach cleaning system, include the sand washing and
separation and multiple stages of oil-water separation.
The problems of interfacing and controlling the
two processes must be studied. In connection with this
program, a study of instrumentation which can provide
on-line analyses of the effluent streams should be
undertaken. Any attempt to build a full-scale system
must include provisions for monitoring and controlling
the process.
3) Finally, a systems analysis study needs to be undertaken
to permit use of existing*performance data for designing
a beach cleaning plant which will meet specific
operational requirements. This systems analysis should
also be used to optimize the process design from the
standpoint of capital and operating costs and performance.
-------
SECTION 3
INTRODUCTION
The spillage of oil from ships, at offshore drilling sites
and during transfer operations, can result in contamination
of adjacent shore areas. An American Petroleum Institute
study (Ref. 1) analyzed some 38 major spills in the period
1956 to 1969 to determine the size and nature of the oil
spill problem. From that study, the following information
was extracted:
1) 85% of the spills investigated occurred off recreational
shoreline,
2) 70% of the spills were larger than 5000 barrels, the
median size being about 25,000 barrels, and
3) crude or residual oils were involved in 90% of the spills.
Further, a 1968 Report to the President (Ref. 2) presents
several estimates of the frequency of spills, including a
U. S. Coast Guard report of 371 "cases of record" in 1966.
Viewed together, these statistics graphically illustrate
the potential damage associated with oil spills.
Where spills have contaminated beaches, the major efforts
directed toward restoring the shoreline have usually included
physical removal of the oiled sand, adsorption of the oil
with various sorbing material (notably straw) and the use
of chemical dispersants; however, none of these procedures
have been completely satisfactory. Imprudent sand removal
practices can result in permanent changes to the beach
features. The sorption of oil from the sand is a highly
labor-intensive procedure that is slow and only marginally
successful in cleaning the beaches. Chemical dispersants
can be effective when systematically and correctly applied;
however, improper application can result in severe damage
to the littoral ecosystem.
To a large extent, beaches have been left to restore
themselves through the physical and chemical action of the
winds, waves and tide. Biological degradation of oil on
beaches has also been observed. These natural processes can
reclaim a contaminated beach, but the time frame may be
from several months to a year or more.
-------
Thus, it becomes apparent that new means must be devised
for quickly and effectively cleaning oil-contaminated beaches
and to restore them to their natural state.
This study was one of several initiated by the Environmental
Protection Agency (the Federal Water Pollution Control
Administration at the beginning of the project) to investi-
gate new methods for restoring oil-contaminated beaches.
The concept proposed includes washing of the sand in a jet
contacting washer, followed by cyclonic separation of the
sand from the washing water and cyclonic separation of the
oil and water.
The primary objective of this program was to demonstrate
the feasibility of the proposed sand washing and oil-water
separation systems for restoration of oil-contaminated
beach sands. By using pilot scale test apparatus, the
efficiency of and parameters controlling the sand washing
process were to be determined. Also, the effectiveness
,of oil-water separation in cyclones was to be studied, with
the goal of achieving an oil product containing less than
3% water. The results of these investigations were to be
used to form the basis for the conceptual design of a mobile
system for cleaning oil-contaminated beaches and to make
recommendations regarding the operation of the system and
the disposition of the recovered sand and oil streams.
In the next section we will briefly discuss the background
leading to the development of this concept, including the
preliminary bench-top washing tests, and present arguments
for using cyclones in the separation stages. Also, a survey
is presented of the characteristics of sands and oils likely
to be encountered in beach related spills.
Next, the experimental programs aimed at evaluating the
performance of the sand washing and oil-water separation
processes are laid out. Finally, the results of the testing
programs are presented, followed by a discussion and extra-
polation of the data to a full-scale system design.
-------
SECTION 4
BACKGROUND
Sand Washing
A simple method was devised for determining the effective-
ness of different means for washing oil-contaminated sands,
e.g. solvent and detergent washing, hot and cold water
washing etc. The method consisted of vigorously agitating
oiled sand in a container filled with the washing medium.
A successful washing test would show a layer of clean sand
at the base of the container, a layer of clear water above
that, and a free layer of oil at the air interface. For
cases in which detergents or solvents were used, the washing
solution generally contained, at the end of the cycle, a
large fraction of the input oil as an emulsion or in solution.
One of the basic ground rules of these preliminary investi-
gations was that the sand would be water-wetted prior to its
contamination with oil. This appears to be a reasonable
assumption since the oil is transported to the shore area
by water and is able to contact sand only after the water
advances ahead of it through the sand.
Using cold (70°) tap water as the washing fluid in these
"jar" tests, we were able to effectively clean sand which
had been contaminated with both No. 4 and Bunker C fuel
oils. The oil from these tests usually formed a free
floating surface within about 30 seconds after agitation
ceased. Tests using hot water (up to 200°F) and detergent
and solvent washing were similarly successful in stripping
oil from the sand, but the oil was left in a less than
desirable condition—emulsified or in solution. The
determination of cleanliness in these tests was qualitative;
i.e. no direct measurements of oil remaining in the sand
were made. However, sand extracted from the containers
appeared clean and little or no oil would rub off onto
clean filter paper.
Based primarily on these limited tests, we proposed to
investigate a mobile system for cleaning oil-contaminated
beach sand. The process to be studied would incorporate
cold water washing of the sand in a turbulent jet washer/
mixer followed by separation of the sand from wash water in
a cyclone.
-------
Oil-Water Separation
The sand washing system discussed above discharges cleaned
sand and a mixture of oil and water. This oil-water mix must
be separated into its component parts in order to reduce the
volume of liquid to be handled for disposal. In keeping
with the concept of a mobile beach cleaning system, we
suggested the use of cyclones to effect this separation
process. Cyclones appear attractive for this application
because of their high throughput rates and small space
requirements. Also, the performance of a cyclone is in-
sensitive to the orientation of its major axis (horizontal,
vertical, or positions in between) making it ideal for use
on a platform moving across a beach surface.
Data available in the literature (Ref. 3) and supplied
by Dorr-Oliver Incorporated (Ref. 4) pointed the way
to using a system of staged cyclones to promote a high
degree of separation of the oil from the influent wash
water, and to provide a relatively clean water stream
which could be recycled to the sand washing system.
In the staged concept, the oil-water mixture from the sand
washer is used as feed for the first stage of separation.
Here, the cyclone is operated to extract the maximum possible
amount of clean water at the underflow. The oil-enriched
overflow then passes on to become feed for the second stage.
The second stage cyclone can be operated in the same manner
as the first one (clean water underflow with a concentrated
oil overflow) or it can be used to yield a highly enriched
oil overflow with a dirty water underflow. When operation
is in the first of these modes, the overflow is treated further
in subsequent stages and the underflow is recycled to the
washer. If the second mode is used, the dirty underflow
can be recycled to the feed stream for the first stage and
the overflow pumped to storage and/or disposal. Figure 1
schematically shows the general arrangement of this cyclone
staging/recycling system.
Though important information with regard to the operating
characteristics of the cyclone in separating oil-water mix-
tures was still needed, we were sufficiently confident of
the expected performance to propose cyclones for this applica-
tion.
8
-------
(overflow)
Mixed Phases
Feed
I
Heavy Phase
(underflow)
Feed"
Light Phase
F
I
Mixed Phases
Feed
Light
Phase
Recycle Stream
Heavy Phase
Figure 1. OPERATING MODES FOR LIQUID-LIQUID CYCLONES
-------
Characteristics of Beach Sands
As part of the overall study, we were to establish the
characteristics of beach sands that might be encountered in
an oil spill situation. Also, it was necessary to make a
brief preliminary survey to determine the nature of sand
most suitable for the pilot-scale testing.
Probably the most consistent comment found regarding beach
sands of the coastal U. S. is the high degree of sorting
associated with the materials found. The following
quotation (Ref.5) is typical of the discussions found.
"If a mixture of pebbles, sand and fine silt is placed on
a beach foreshore, the fine silt may be carried away from
the shore zone by suspension, whereas the sand and pebbles
are distributed along the foreshore and nearshore bottom
zones. The pebbles tend to concentrate along the seaward
edge of the foreshore or they may become associated with
the shifting portion of the plunge zone. The sand is pre-
dominantly distributed over the foreshore, or it may be
partly shifted to seaward of the low tide line."
Thus, it can be seen that the active hydraulic forces of
the tides and waves work to segregate beach sediments into
restricted size ranges that are very consistent for any
particular beach. And, when sand is present, it is
preferentially deposited where water-borne oil is most
likely to contaminate it — along the intertidal zone and
the wave-washed region.
Extensive sampling programs (Refs. 6 and 7) verify the
predominance of siliceous sand beaches along the Atlantic
coast of the United States, except for the southern tip of
Florida and the Florida Keys. In this region, the beaches
are made up largely of shell deposits and crushed coral.
Sand particles on Atlantic coast beaches range in size from
0.005 inch to 0.020 inch, but the range for any beach was much
narrower than this, with sorting coefficients of the order of
1.3. Sorting coefficient is defined as the square root of the
ratio of the particle sizes at the "one-quarter greater than" and
"three-quarters greater than" distribution points. The mean
particle diameter on these beaches is around 0.008 inch.
10
-------
The Gulf Coast beaches, too, show remarkable sand size
uniformity (Ref. 8) and chemical makeup. Again, with the
exception of Florida's southern tip, the sands are generally
siliceous, although magnetite is present in abundance. Mean
particle size shows some differentiation east and west of
the Mississippi River Delta region. The Texas and Louisiana
beach sands have mean diameters ranging from 0.0035 inch to
0.0085 inch with sorting coefficients around 1.17.
Beach sands along the Pacific Coast have generally larger
mean particle diameters (about 0.014 inch) and sorting
coefficients around 1.3 (Refs. 9 and 10). These sands are
almost entirely free of calcareous deposits (CaCO amounting
to 7%) and the little that is present is made up of broken
shells in a very thin surface layer.
Sand particles, in general, tend to sphericity in shape
and the mean particle size can be shown to influence the
slope of a beach. Temporary alterations to the texture of
a beach can result from severe storms, but equilibrium is
usually rapidly restored by the normal wind, wave and tidal
conditions. There is also a tendency for the beach character-
istics (sand distribution, dunes, slope, etc.) to change
seasonally, especially where there is a wide seasonal
variation in the tides and storm activity. Man-made shore
installations, e.g. piers, jetties, groins, dikes, etc., can
also disturb the natural condition of a stretch of beach and
attempts at restoring beaches (Ref. 11) have altered the
texture significantly from the norm.
The major conclusions regarding coastal beaches in the U. S.
that can be drawn from this review are:
1) beach sand particle mean diameters range from about
0.008 to 0.014 inch, with sorting coefficients around 1.2,
2) sand in any one stretch of beach will be very well
sorted under normal conditions,
3) sand size variations do occur from the low tide line
to the berm (beginning of backshore region) but the
variation is probably very narrow, and
4) the predominant material making up the beach sand is
silica.
11
-------
The discussion above has been restricted to a consideration
of beach sands. Beaches composed of shingles and boulders
could present a problem to some of the elements of the
sand washing apparatus, and we would therefore expect to
screen out these materials (along with ordinary beach
debris and trash) prior to introducing the oily sand to
the washing machine.
Based on the results of this survey, we selected a well
washed and graded silica sand for use in the testing program,
The sand is designated as #50 sand (for the mean mesh size)
and has a mean particle size of 0.0114 inch with a sorting
coefficient of 1.3. Figure 2 shows the size distribution
curve for the sand as supplied.
Characteristics of Spilled Oils
Reference 1 summarizes the types and quantities of oil im-
plicated in a review of major spills. Crude oil represented
the largest volume of oil spilled (80%) as well as being
involved in the greatest number of spills (18 of 35).
Residual oils make up the next largest number of spills
(14 of 35), although they represented only one percent of
the total spilled volume. Light oils make up the remaining
fractions. It appears, based on these limited data, that
crude and residual oil will probably be involved in any
future spills.
The oils selected for use in the sand washing and oil-water
separation test programs were #4 and #6 fuel oil. Crude
oil (South Louisiana and Bachaquero, obtained from EPA,
Edison, New Jersey) was used to a limited extent in the
sand washing program. The fuel oils were obtained from
bulk oil handlers and are presumed to meet all commercial
specifications for these grades. The specific gravity (at
60°F) for the No. 4 and No. 6 oils was measured and found
to be 0.91 and 0.96 respectively. Crude oil in quantities
sufficient to be used in the pilot-scale testing was not
readily available and so all major testing was accomplished
with the No. 4 and No. 6 fuel oils.
12
-------
H
O
•H
-P
40
30
20
N 10
-H
CO
8
6
5
4
3
©
I
0.1 1 10 25 50 75 90
Weight Percent Finer Than (%)
Figure 2. PARTICLE SIZE DISTRIBUTION CURVE FOR #50 SAND
99
99.9
-------
SECTION 5
EXPERIMENTAL PROGRAMS
The experimental portion of this study was carried out in
two separate phases:
1) design and testing of the sand washing and separating
system/ and
2) separation of oil-water mixtures using cyclones.
Because they were conducted as individual efforts, these
programs will be so presented in the discussions following.
A conceptual marriage of the systems will be presented in
Section 6.
Sand Washing and Separation
Design of the apparatus
The primary purpose of this phase of the program was to de-
vise and test a system for cleaning oil from beach sand and
for separating the cleaned sand from the washing fluid. Our
preliminary "jar" tests demonstrated that a turbulent mixer
would be capable of stripping the oil from the sand if
sufficient mixing time and/or energy were provided. Using
this turbulent mixer concept and the basic requirement that
the process under investigation be adaptable to a mobile
platform, we devised the flow-through jet washer shown in
Figure 3.
In this assembly, sand is fed via a vibrating screw feeder
to the inlet of the washing section. The washer consists
.of a cylindrical tube with radial and inclined water inlet
nozzles machined circumferentially around the chamber. A
second tube surrounds the washing section and the annular
space between them forms a plenum for distributing water to
the nozzles. AS oily sand enters the washer, it is violently
stirred by the impinging water jets. This stirring action
serves to strip the oil from the water-wet sand particles
and reduce the size of the oil droplets as determined from
the oil-water interfacial properties and the power input to
the washer. The major source of energy for this mixing process
is supplied as kinetic energy in the water jets. For a
15
-------
-tie rod
Variable Length
Mixer Section
/• /• /• /* _/
Cv\\\\\\
washer nozzle
Washing Chanber
Wash Water
Inlet Connection
tie rod
Figure 3, SATTD I7ASIIER AfSflSI
-------
typical mass ratio of 3 lbm of water per 1 lb^ of sand, the
kinetic energy in the water jets is about 30,000 times greater
than that in the oil-coated sand.
Following the washer is a constant diameter, variable length
section of pipe in which mixing of the sand with washing
fluid continues. In some of the modeling studies performed
in the early phases of this program, it was shown that the
duration of the washing/mixing process might be an important
variable. This variable length section made possible an
approximate 3.5:1 increase in the total residence time for
a fixed volumetric feed rate through the system. For maximum
feed rates, the holdup time in the washer/mixer is about 0.5
seconds per foot of length. This mixing section was fabri-
cated from clear acrylic tubes permitting visual observation of
the washing process as the mixture proceeded downstream.
At the end of the mixing section, an adaptor flange is used
to provide smooth transition from the mixer tube to the
smaller diameter inlet of the sand-separating cyclone. The cy-
clone used for this phase of the program was a 6-inch diameter,
type FR, DorrClone, manufactured by Dorr-Oliver Incorporated.
This unit was purchased with three different vortex finders
(overflow tube), 1-1/4, 1-5/8, and 2-1/2 inch diameter. The
apex valve (underflow port) was supplied as 3/4 and 1-1/2
inch diameter, although it was possible to reduce this by
squeezing the walls of the rubber discharge nozzle with a
hose clamp.
The cyclone was operated to give a rope-like discharge at
the underflow. This condition was set by varying the size
of the apex opening until the discharge alternated between the
solid-appearing rope flow and a splashing, vortex flow. Then
the port size was decreased slightly until the rope discharge
was stable. In this manner, the maximum amount of washing
fluid (along with the oil droplets) was discharged to the
overflow, carrying along little or no clean sand, and the
underflow carried away only enough water to keep the sand
fluid. Under normal1 operation, the underflow was about
equally divided between sand and water (volume basis). This
means that for a case where the mass flow of water was about
three times that of the sand, 85% of the wash water was
recovered at the cyclone overflow.
17
-------
The apparatus for feeding sand to the washer consisted
of a helical wire screw feeder with a vibrating bin and
screw tube. This vibrating action, induced by means of an
eccentric flywheel attached to the drive shaft, aided in
preventing clogging of the feed port and screw. The unit
was manufactured by Vibra Screw Incorporated, and was des-
ignated as a 3" live bin feeder. Through the use of a variable
speed transmission, the sand feed rate was adjustable from
about 7 to 70 cubic feet per hour (one cubic foot of wet
sand weighs about 100 lb).
Coupling of the washer to the sand feeder presented some-
what of a problem, since the screw feeder tube assembly
(partially shown in Figure 3) must be allowed to vibrate
freely about its axis, but a watertight seal was also re-
quired at the interface of the screw tube and washer flange.
The solution shown incorporates an O-ring face seal in
contact with the vibrating tube.
To help maintain alignment of the vibrating screw relative
to the axis of the washer/ a flexible collar was fitted
into the annulus formed by the outside of the tube and a
sleeve.
Figure 4 shows the flow diagram for the sand cleaning
pilot plant. As may be seen, the system was operated closed-
loop with the cyclone overflow returning to the oil water
storage tank. This tank provided a holdup period to allow
oil to separate from the recycled wash water. The water
was gravity fed to the lower tank wash water storage tank,
passed to the inlet of the main feed pump, and then to the
washer inlet plenum. Sand from the underflow of the cyclone
was dumped into a separate collecting tank. A drawing of
the major equipment in this pilot plant is presented in
Figure 5.
In the event of system upsets and for startup transients,
it was possible to backflush oil from the sand collecting
tank, allowing the overflow to drain to the wash water
storage tank. After sufficient oil had collected in these
two tanks, it was skimmed from the surface of the water
and transferred to barrels for disposal.
During the shakedown testing of the pilot plant, we found
that wash water would back up the screw feeder tube and
18
-------
vo
Tank Overflow
Overflow
(oil-
water)
Cyclone
Underflov
(sand)
Washer
and
Mixer
Vibrating Bin
Sand Feeder
Sand Collecting
Tank
Figure 4. SAND CLEANING PILOT PLAUT
FLOWSHEET
P
Oil-Water
Storage Tank
Wash Water
Storage Tank
Wash Water
Flowmeter
* tXh
Flowmeter
Main Feed
Pump
Backflush line
for sand tank
-------
Sand
Collecting
Tank
Oil-Water Storage Tank
Water Flowuieter
'Screw Tube
Plan View
Sand Separating
Cyclone
Washer
Main Feed
Pump
Sand
Collecting
Tank
Oil-Water
Storaye
Tank
Oil-Water
Storage Tank
Elevation View
Figure 5. SAND WASHING EQUIPMENT- SKETCH
20
-------
into the hopper, thereby stopping the flow of sand.
Theoretical calculations indicated that the sand in the
screw tube should be capable of supporting a pressure drop
of about 15-20 psi, based on the flow of water through a
uniform bed of packed sand particles. However, the vibratory
nature of the screw feeder prevents the sand from packing
in the tube, resulting in a much reduced pressure drop
and hence the water backup problem. A solution was pro-
vided by fitting a tight cover to the top of the sand
supply hopper and balancing the back pressure on the washer
with air so that a stationary air-water interface was
maintained downstream of the screw tube. This necessitated
operation of the pilot plant in a batch mode, since the
hopper could be loaded only when open to atmosphere. No
problem was encountered with this batch operation because
the hopper contained sufficient sand to permit reaching
steady-state operation at all feed rates.
Range of variables
There were three sets of variables investigated in the
sand washing phase of this program:
1) feed variables
a) sand type
b) oil type
c) oil/sand condition
d) inclusion of sorbers in feed
2) process variables
a) sand and water feed rates
b) ratio of water to sand in feed
c) feed temperature
d) solvents, detergents, etc.
e) cyclone operation
3) washer/mixer variables
a) nozzle geometry
b) mixing length
The first set of parameters is dependent on the nature of
the oil spill under consideration. For our test purposes,
21
-------
we fixed the sand type and used No. 4 and No. 6 fuel oils
to determine if oil type affects the sand washing process.
Regarding the condition of the oil/sand mix fed to the washer,
we generally used freshly mixed (less than 24 hours old)
batches of oily sand. The sand was always wet with water
(and excess water drained away) prior to mixing in the oil.
Some testing was done with oily sand mixtures which had
been allowed to "age" under infrared lamps for two days.
The condition of the mix after this aging was not much
different than the fresh mix, except that no excess water
was present and the sand particles were slightly more
cohesive. "Tar balls" could be formed by rolling a mass of
aged sand in the hands, but with only minor agitation these
clumps broke apart.
Finally, the only sorber utilized in this program was
straw. Mulched straw was mixed in with oily sand (No. 4 oil)
at about 2% of the weight of the sand. The results of
tests with this feed will be discussed separately in Section 6.
In the case of the process variables, the last three re-
mained fixed throughout the program. No attempt was made
to include temperature as a controlled process variable
for two reasons. First, in the preliminary "jar" tests sand
washing was seen to be equally effective at 35°F and 100°F.
Second, the wash water used in the system is recycled, re-
sulting in a net heating effect due to the work input from
pumping. This warmer water, mixed with cold, oiled sand,
will increase the temperature of the sand since the water
has a severalfold larger heating capacity. The feed
temperature varied only slightly from test to test (in the
range from 50° to 65°F). No solvents, detergents or other
agents were added to the feed. Their use was not indicated
by the results of the washing process. As mentioned
previously, the cyclone was operated to give a rop'.. discharge
in all tests, although the vortex finder and ap•..••: valve
were varied in size for several of the tests. Then, for
the set of process parameters listed, only the absolute and
relative feed rate of water and sand were included as variables
in the test program. The water flow rate was varied from
20 to 40 gpm, and the mass ratio of water to sand ranged from
3 to 9, at all water flow rates.
22
-------
The geometry of the washer nozzle and the mixing lengtn
were also included as variables in the test program. Two
sets of nozzles were available (each set was machined into
a different washing section) requiring disassembly of the
washer in order to change them. In the washing nozzle
assembly designated No. 1, six 1/4 inch diameter jets were
drilled into the wall of the chamber and equally spaced
around the periphery. The axis of each jet lies at a 45°
angle to the longitudinal axis of the washer pointing
downstream, and is also rotated 45° toward the side of the
washer, each jet rotated in the same direction (See Figure 3).
Thus, liquid entering the washer chamber through these jets
tends to travel in a helical motion toward the discharge
end. The reasoning behind this nozzle geometry was that the
water, traveling in a downstream direction, would transfer
some of this momentum to the sand and sweep it along. It
was surmised that the helical motion could increase the
residence time of the sand in the washer and thereby increase
the washer efficiency. Also, it was hoped that this geometry
would reduce the back pressure on the sand feeder tube and
prevent water from backing up the tube.
Nozzle No. 2 was similar to No. 1 in that three of the jets
were arranged on the 45° x 45° pattern, with the other three
set to point radially inward at the influent sand. The angled
and straight jets were alternated around the periphery. In
this case, the job of the straight jets was to act as a
uwater knife" to slice up clumps of sand as they entered the
washer.
As mentioned previously, the length of mixing section down-
stream of the washer was variable. Early in the shakedown
testing it was noticed that/when the shortest length of
mixing section (1 inch) was in place, the washer would
tend to plug on startup, due to a piling up of sand in the
mixing tube caused by a decrease in cross-sectional area
at the entrance to the cyclone. Because of this problem,
we abandoned use of the short mixing section and all future
tests were run with either the 12 inch or 24 inch mixing
sections downstream of the washer.
Appendix A lists all of the tests run (except for special
tests using sorbers or aged sand) and the combination of
parameters for each.
23
-------
Experimental Procedure
A typical test was carried out as below:
1) 400 Ib of wet sand were thoroughly coated with oil at
4 to 8% of the sand weight,
2) the sand feeder bin was loaded with the oiled sand and
the feeder run to fill the screw tube/
3) the sealing cover was fitted to the top of bin,
4) main feed pump was started and the desired flow rate
set,
5) simultaneous with (4) the air supply to the bin was
set to match the pressure in the washer,
6) the sand feeder was turned on and the speed control
set to give the desired sand flow rate,
7) the cyclone apex valve port size was manipulated to
give the desired rope discharge, and
8) as conditions stabilized, flow and pressure data were
recorded and samples of the cyclone underflow and
overflow were collected for later analysis (samples of
the feed were collected prior to sealing the bin).
For any given water rate, the sand feed was varied over
its full range. After all data was collected for a single
water feed rate, it was changed and the procedure repeated.
In some cases, it was necessary to shut down for refill of
the hopper in order to collect a full set of data for one
set of washer or cyclone geometries, with one oil.
The critical information desired from any particular test
was the effect of operating variables on the cleaning and
separating of sand. Thus it was necessary to determine
the degree to which the process was able to remove oil from
sand. The analytical technique used to determine the oil
in the feed and cleaned sand and in the overflow water
was taken from Reference (11) and is quoted below:
24
-------
1) Weigh 50 grams of sample (oiled sand) into 250 ml
Erlenmeyer flask.
2) Slurry four times, or until extraction is complete, with
50 ml of 10% acetone in chloroform, which has been
heated to just below its boiling point.
3) Decant solvent after each extraction through fluted
number 4 filter paper into a 250 ml beaker.
4) Evaporate combined extracts on a steam bath to approxi-
mately 25 ml and transfer quantitatively to a tared
50 ml beaker.
5) Evaporate extracts to dryness, then add 5 ml of acetone,
and again evaporate to dryness.
6) Wipe off excess water from outside of beaker, then
dry 10 minutes in an oven at 103°C.
7) Cool in desiccator and weigh (extract).
Using this technique, we were able to determine the per-
centage of oil removed from a given feed sand and thereby
evaluate the performance of the sand washing system.
The results of the sand washing tests and a discussion
of the data is presented in Section 6.
Oil-Water Separation Tests
Design of the apparatus and procedures
In this phase of the study, conventional, commercially-
available cyclones were used to perform the oil-water separa-
tion tests. During the course of the experimental program,
four different test setups were used. Appendix B lists
all of the tests performed and shows the scope of variables
investigated.
Preliminary testing was done with a TML DorrClone (manu-
factured by Dorr-Oliver Incorporated, as were all of the
cyclones in this program) taking feed from a pre-mixed oil-
water storage tank. The TML unit contained four 10 mm Nylon
25
-------
cyclones arranged so that they are fed from the same inlet
and discharge through the same overflow and underflow
connections. Oil and water were added to the baffled tank
and agitated to produce the feed; product streams were re-
turned directly to the tank. A schematic of the test rig
is shown in Figure 6. This apparatus and procedure was
adequate for the preliminary testing, but it did not produce
consistent feed analyses and the continuous heavy agitation
and pumping tended to stabilize an oil in water emulsion.
Tests 1 through 18 in Appendix B were conducted with this
setup.
The next series of tests (19 through 62) continued in the
TML DorrClone, but separate feed pumps were used for the
oil and water. Mixing of the two components was obtained
by blending the two streams upstream of a throttling valve
and taking a controlled pressure drop across the valve. The
pressure drops taken across the mixing valve were of the same
order of magnitude (5 to 15 psi) as the pressure drops
measured across the sand washer and cyclone combination
during the se.id washing program. This system produced a more
consistent feed than the batch feeding approach and also
permitted the use of heated water for the feed. Underflow
and overflow rates were controlled by valving in the product
lines. The apparatus is shown in Figure 7.
When two-stage cyclone testing was initiated (tests 63 through
67, 75 through 78, and 86 through 95), the modified flow loop
shown in Figure 8 was used. In this setup, two TML units were
used; the first one contained nine 10 mm cyclones while the
second had four. This two-stage system was also used for
recycle tests 79, 80 and 81 in which the underflow from the
second stage was returned to the inlet of the water pump and
blended with the raw feed. The recycle system yields only two
product streams, rather than the three in the non-recycled
two-stage setup.
The fourth group of tests used a single-stage system employing
6 inch diameter cyclones. Two series of tests (68 through 74
and 82 through 85) were carried out with the 6 inch units
in the setup shown in Figure 7.
In all the tests, identical procedures were followed.
Pumps were started and control valves set to produce the
26
-------
Overflow
Underflow
Mixer
Apex valve
Shut-off valve
€
Flow control
Feed pump
Baffled Feed Tank
Figure 6. DORRCLONE TEST APPARATUS SCHEMATIC
27
-------
00
IAP,
Water
-OP
oil
Mixing valve
oil
N—-
Overflow
Figure 7. TML DORRCLONE AND 6 INCH DORRCLONE OIL-WATER
SEPARATION FLOWSHEET
-------
Flowmeter
to
vo
Pressure
Regulator
Mixing
valve
Oil
First Stage
Underflow
Second Stage
Overflow
Second Stage
Underflow
Figure 8. TWO STAGE OIL-WATER SEPARATION FLOWSHEET
-------
desired ratio of overflow to underflow volume flow rates
(this ratio is called the "split" and the terminology will
be frequently used throughout the remainder of this report)
or the desired visual clarity of the underflow streams. When
the system was considered to be at equilibrium, flow rate
and pressure data were recorded and samples of the product
streams were collected for laboratory analysis.
Oil-water sample analysis
The fraction of oil in the various feed and product streams
was determined by either of two methods, depending on the
relative amounts of oil and water in any sample. In samples
with low oil concentration, the oil from a sample of known
volume or mass was extracted using chloroform. The
chloroform was evaporated on a steam bath and the remaining
oil volume or mass measured. The fraction of oil in the
sample was then determined by comparing the mass of the
extracted oil to the mass of the initial oil-water mixture.
In samples with low water content, the analysis was
accomplished with a distillation technique. Here, toluene
was added to the oil-water sample (of known volume or mass)
in a distilling flask. The toluene and water were evaporated,
condensed and collected in a trap. Here the two immiscible
liquids separated and the toluene was returned to the
distilling vessel. Water could be withdrawn from the trap
and collected in a graduate, as needed. The distillation
was continued until water flow from the condenser ceased.
The volume of water collected was then used to determine
the analysis of the original sample.
In the next section, we present the results of the two
experimental programs and discuss these results and their
implications on the feasibility of the proposed beach
cleaning concept.
30
-------
SECTION 6
RESULTS AND DISCUSSION
In this section are presented the results of the separate
sand washing/separating and oil-water separating programs.
A discussion of the results from each phase of the study is
given and a preliminary conceptual design for a mobile
beach cleaning plant is discussed.
Sand Washing and Separation Data Analysis
At the beginning of the sand washing program, a theoretical
model of the washing mechanism was developed in order to
try to understand which variables would be important in
our analysis of the data.
This analysis showed that the oil stripping action is the
result of viscous shearing caused by the relative velocity
between the oil covered sand particles and the surrounding
washing fluid. interparticle and particle/wall abrasion
may also be important to the cleaning process, but their
effects were not included in the model. Neither did the
model consider the problem of re-coating of the sand particles
with the just stripped oil. However, based on the simple
jar tests it did not appear that re-coating was a problem,
since continued agitation of the cleaning mix revealed no
detectable recontamination of the sand. What did result
from this agitation though was emulsification of the oil
in the water and loss of the easily separable nature of oil
and water.
The conclusions drawn from study of the model were that too
many uncertainties remained with regard to the effects of
water film thickness (between the oil and sand particle),
adhesion, and surface tension effects to permit useful
interpretation of the theoretical results. For instances, by
including a water film between the oil and sand, it was
shown (in the model) that the time required to strip oil from
the sand was reduced by several orders of magnitude from the
case where the sand was oil-wetted only.
Though the model did not prove satisfactory in the analysis
of our data, we did find after careful review of the results
31
-------
from preliminary tests, that the degree of oil removal
from the sand was related to energy dissipation in the washer.
Following is a brief discussion of this energy dissipation
relationship and the results of the testing program are
evaluated in terms of this energy expenditure and some of
the feed and geometry parameters presented in Section 5.
The purpose of the fluid jets in the washer is to create
viscous shear between the water and oil films. The level
of shearing is related to the amount of energy dissipated
in the washer. As mentioned earlier, the energy input to
the washer comes primarily from the kinetic energy of the
water jets and so the power (rate of energy utilization)
consumed in the washer is:
4> = Q Apw (1)
where :
$ = power dissipated (ft-lb /sec)
Q = volume flow rate of water (ft /sec)
Ap = static pressure drop from washer inlet plenum
w
to exit of washer (Ib /ft2)
By dividing the power dissipated, for any particular water
flow rate and pressure drop situation, by the mass flow rate
of sand in the washer, we can obtain an expression for the
specific energy consumption.
Q
SE = *
w W W
s s
where :
SE = specific energy consumed in the washer (ft-lb./lb ,)
w r sand
W = mass flow rate of sand (Ib /sec)
s m
32
-------
The cyclone also exacts a pressure drop penalty on the fluid
system, resulting in further dissipation of energy. Here
the specific energy term is:
Q
EEo = -5— <3>
s
where :
SE = specific energy dissipated in cyclone (ft-lb_/lb )
c t sand
2
Ap = pressure drop from inlet to outlet of cyclone (Ib /ft )
c r
The sum of SE plus SE gives the total specific energy
consumed in the sand washing apparatus, for a particular
set of feed parameters.
Oil removal effectiveness is the principal dependent variable
of interest in the washing study and is defined as :
f.
OR = in (~) (4)
o
whe re :
OR = oil removal effectiveness
f = fraction of oil in the sand feed
i
f = fraction of oil in the cleaned sand
o
In terms of the oil left in the sand after washing, an OR
of 4 results in about 182 ppm of residual oil per percent
of oil originally in the feed. OR's of 5 and 6 mean 67 and
25 ppm of oil remaining per percent oil in the feed. Feed
oil concentrations ranged from 4.5% to 7.5% of the dry sand
weight.
Table 1 lists the results of all the tests, including tabula-
tions of the specific energy dissipations in the washer and
cyclone and the oil removal effectiveness.
33
-------
Test No.
1111-01
1111-02
1111-03
1111-04
1111-05
1111-06
1113-01
1113-02
1113-03
1113-04
1113-05
1113-06
1113-07
1123-01
1123-02
1123-03
1123-04
1123-05
1123-06
1123-07
1123-08
1123-09
1123-10
RESULTS
Oil Removal
Effective-
ness
OR
2.9
3.79
2.78
2.11
2.2
2.72
2.23
2.19
2.20
2.28
0.99
1.13
2.49
2.36
2.73
3.28
3.61
3.05
3.06
3.41
3.46
3.33
3.45
TABLE 1
OF SAND WASHING
Washer
Specific
Energy
SE
w
78
111
175
154
98
69
17
23
32
51
150
177
332
61
78
108
169
24
32
46
72
140
163
TESTS
Cyclone
Specific
Energy
SE
(ft-ibf/ibm>
45
62
96
96
62
45
19
25
34
54
61
71
129
17
25
38
58
11
15
12
32
30
38
Mass
Ratio
(Sand/
Water)
0.25
0.176
0.118
0.112
0.176
0.25
0.33
0.249
0.177
0.111
0.246
0.208
0.111
0.328
0.25
0.176
0.112
0.33
0.249
0.177
0.111
0.246
0.208
34
(Continued)
-------
TABLE 1 (Cont)
Test No.
1123-11
1124-01
1124-02
1124-03
1124-04
1124-05
1124-06
1124-07
1124-08
1124-09
1124-10
1124-11
1201-01
1201-02
1201-03
1201-04
1201-05
1201-06
1201-07
1202-01
1202-02
1202-03
1202-04
1202-05
RESULTS
Oil Removal
Effective-
ness
OR
3.46
3.08
2.84
3.27
3.37
3.12
3.19
3.63
2.87
3.02
3.95
7.25
3.27
3.33
3.34
4.13
4.01
3.94
4.37
3.75
4.12
4.48
4.70
4.1
OF SAND WASHING
Washer
Specific
Energy
SE
w
(ft-lbf/lbm) I
301
49
64
91
144
17
23
32
51
131
154
290
70
92
131
205
171
201
379
66
87
124
195
159
TESTS
Cyclone
Specific
Energy
SE
;ft-ibf/ibm)
75
23
30
41
64
19
25
34
54
30
35
64
23
30
41
64
50
56
102
27
35
48
75
71
Mass
Ratio
(Sand/
Water)
0.111
0.328
0.25
0.176
0.112
0.33
0.249
0.177
0.111
0.246
0.208
•0.111
0.328
0.25
0.176
0.112
0.246
0.208
0.111
0.328
0.25
0.176
0.112
0.246
35
(Continued)
-------
TABLE 1 (Cont)
Test No.
1202-06
1202-07
1203-01
1203-02
1203-03
1203-04
1203-05
1203-06
1203-07
1214-01
1214-02
1214-03
1214-04
1214-05
1214-06
1214-07
1214-08
1214-09
1214-10
1214-11
1218-01
1218-02
1218-03
RESULTS OF
Oil Removal
Effective-
ness
OR
5.02
4.23
3.61
2.96
3.15
3.68
3.53
3.61
3.46
3.94
5.17
5.86
5.79
5.6
5.94
5.43
4.52
3.74
3.63
3.85
5.70
5.49
5.49
SAND WASHING
Washer
Specific
Energy
SE
w
(ft-lb,/lb ) i
f m
188
353
66
87
124
195
159
188
353
70
106
131
206
168
199
374
8
37
52
83
70
92
131
TESTS
Cyclone
Specific
Energy
SE
c
83
151
35
45
62
96
92
107
194
47
69
83
130
92
107
194
29
40
55
86
31
40
55
Mass
Ratio
(Sand/
Water)
i
0.208
0.111
0.328
0.25
0.176
0.112
0.246
0.208
0.111
0.328
0.216
0.176
0.112
0.246
0.208
0.111
0.33
0.249
0.177
0.111
0.328
0.25
0.176
(Continued)
36
-------
TABLE 1 (Cont)
RESULTS OF SAND WASHING TESTS
Test No.
Oil Removal
Effective-
ness
OR
Washer
Specific
Energy
SE
w
(ft-lb./lb )
f m
Cyclone
Specific
Energy
SE
c
(ft-lb_/lb )
r m
Mass
Ratio
(Sand/
Water)
1218-04
1218-05
1218-06
1218-07
1218-08
1218-09
1218-10
1218-11
5.10
6.05
5.64
6.23
3.83
4.15
4.95
4.87
205
159
188
353
24
32
45
72
85
92
107
194
27
35
48
75
0.112
0.246
0.208
0.111
0.33
0.249
0.177
0.111
37
-------
These results are plotted in Figures 9 through 14 with the
oil removal effectiveness (OR) as a function of the specific
energy (both the specific energy dissipated in the washer
alone and the total dissipated in the washer plus cyclone
are used). All of the plotted data show the same general
trend of increasing effectiveness with added energy dissipa-
tion. However, a flattening or fall off in the curves is
seen indicating that the sand washing capability of the
system does not increase without limit.
It appears that once a certain level of energy dissipation
has been reached, further mixing does little to enhance the
process and may, in fact, be detrimental. The drop in
efficiency seen for some of the data could be caused by
emulsification of the stripped oil and the subsequent carry-
over of this oil in the water which discharges with the
sand. In the analysis for oil in the underflow samples, we
did not attempt to determine whether the oil fraction was
actually adhering to the sand particles or if it had been
carried through with the water.
In any event, the effectiveness of the sand washing system
is impressive. The maximum OR achieved for sand contaminated
with No. 4 fuel oil was 6.0 at a total specific energy con-
sumption of 500 ft-lbf/lb (Figure 12), while for the No. 6
oil the maximum OR was 5.0 at specific energy of 270 ft-lb_/lb
(Figure 13). Based on these maximums it should be possible
to clean No. 4 oil from a beach and leave behind a trace
of oil at a concentration of about 125 ppm (assuming 5% oil
on the beach initially) with an expenditure of about 0.6
horsepower per ton of sand washed per hour, in the washer/
cyclone combination. If the contaminant had been No. 6 oil,
the residual would be at about 350 ppm (5% oil initially
in the sand), but the power expenditure would drop to 0.3
horsepower per ton per hour. Cleaning of No. 4 oil to
an OR level of 5 would require only about 0.15 horsepower
per ton per hour.
Figure 9 shows the results from tests conducted with No. 6
oil and with both the washer nozzle geometry and the mixing
length as parameters. The best performance was obtained
when nozzle No. 1 and the shorter mixing section (12") were
used in combination. For tests run with the longer mixer
(24"), the data are indistinguishable between nozzle No. 1
and nozzle No. 2.
38
-------
OJ
V£>
O
to
to
CD
>
•H
-P
0
Q)
M-l
0)
-H
O
Mixer Length
© 12 inches
© 12 inches
^ 24 inches
B 24 inches
#b Oil
Washer Nozzle
#1
#2
#1
#2
0
100
200
300
400
Washer Specific Energy, SE (ft-lbVlb )
w f m
500
Figure 9. PERFORMANCE OF SAND WASHING PILOT PLANT
-------
it*.
o
en
0)
>
•H
4J
O
O -,
M-l 3
«M
W
•H
O
#4 Oil
#6 Oil
Mixer Length = 12 inches
Washer Nozzle 11
100
200
300
400
500
Washer Specific Energy, SE (ft-lb /lb )
w
f m'
Figure 10. PERFORMANCE OF SAND WASHING PILOT PLANT
-------
a 4
en
0)
C
s
-I-?
I i o
0)
pa
1-1
0)
S i
O
1
14 Oil
#6 Oil
Mixer Length = 24 inches
Washer Nozzle #1
"30(5"
100
200 300 400
Washer Specific Energy, SE^ (ft-lbf/lbm)
500
Figure 11. PERFORMANCE OF SAND WASHING PILOT PLANT
-------
CO
CO
0)
C
> 4
-H
-P
O
CD
.3
W 3
O
e
CD
O Mixer Length = 12 inches
0 Mixer Length = 24 inches
Washer Nozzle #1
#4 Oil
0
100
200
300
400
Total Specific Energy, SE (ft-lb /lb )
MHJ
Figure 12. PERFORMANCE OF SAND WASHING PILOT PLANT
-------
00
OS 5
o
w
CO
0)
G 4
Q) ^
>
•H
-P
U
OJ
«M
M-l _
•w 3
rt
>
O
e-
'
Mixer Length = 12 inches
Mixer Length = 24 inches
Washer Nozzle #1
§6 Oil
•H
O
1
1
100
200 300 400
Total Specific Energy, SET (ft-lbf/lbm)
500
Figure 13. PERFORMANCE OF SAND WASHING PILOT PLANT
-------
r
o 5
en
w
cu
c
CD
4-1
0
CU
•4-)
m
w
rH 3
c
E
a;
-H
O
#6 Oil
Mixer Length = 12 inches
Washer Nozzle #1
0
100
200
300
400
Washer Specific Energy, SE (ft-lb^/lb )
w
m
500
Figure 14. PERFORMANCE OF SAND WASHING PILOT PLANT
"AGED" SAND/OIL TESTS
-------
In Figures 10 and 11, comparisons are made between tests
run with the two types of oil and with variable mixing
lengths. These curves show that No. 4 oil is considerably
easier to clean from sand than the heavier No. 6 oil, as might
be expected. By comparing the two figures, it is seen that
the shorter mixing length gives better results in terms of
oil removal effectiveness, for either oil type, although the
distinction between the 12 inch and 24 inch data is not great
for the No. 4 oil. Figure 12 shows the same data for the
No. 4 oil plotted against the total specific energy consump-
tion, and here the difference in performance between the
tests run with the two mixing lengths is obscured by
the scatter in the data.
The test results are the No. 6 oil are plotted in Figure 13
against total specific energy with the mixer length as a
parameter. Again the shorter mixer length gives better per-
formance.
All of the plotted data show a higher OR for the washer fitted
with the shorter of the two mixing lengths. The original
reason for including this variable was to increase the holdup
time of the washer and hopefully allow more time for the
scrubbing action to take place. For the same specific energy
level, the shorter mixing length (and time of mixing) results
in a more vigorously stirred reactor, dissipating a higher
level of horsepower per pound of sand in the washer. Now, it
appears that the longer holdup time is not needed, rather
a very intense scrubbing action of short duration is sufficient
to effect a very high degree of oil removal.
Special Sand Washing Tests
In Section 5, it was mentioned that several special tests
had been run using "aged" oil contaminated sand and in which
sorbers had been added to the feed.
"Aged" sand tests
The tests listed in Table 1 in the sequence 1201-01 through
1201-07 were run with aged oil-coated sand. The mix was pre-
pared by applying about 5% by weight of a No. 6 oil to the
surface of a 4 inch thick layer of wet sand. Two, 500 watt,
infrared heat lamps were placed above the mix (about 4 feet from
the surface), and over the course of 48 hours the sand was exposed
45
-------
to this light source for a total of about 20 hours. The
surface .of the oily sand did not rise above about 120°F.
At this temperature level, oil seeped through the layer
of sand to a depth of about 2 inches, coating at least
50% of the sand. After the exposure period, the sand-oil
layer was thoroughly mixed into the clean, underlying sand
and allowed to cool to room temperature.
This mix was loaded into the feeder and a series of washing
tests performed. The results of these tests are plotted
in Figure 14 as oil removal effectiveness versus washer
specific energy consumption. The peak OR is not as great
as that shown for similar test conditions with freshly mixed
sand and No. 6 oil (Figure 10), however, the performance is
quite good.
This method of "aging" oil does not exactly duplicate the
characteristics of naturally weathered oil found on beaches
after spills; however, within the time available and the limited
nature of the feasibility study, no further work along these
lines appeared justifiable.
Sorber testing
Only two tests were run with straw mixed into the sand feed
and they proved to be of limited success. Straw, at 2%
of the weight of the sand, was used as the sorber and was
thoroughly mixed into the sand along with No. 4 oil which
was at 5% of the sand weight. With this feed, the unit ran
quite smoothly for about 15-30 seconds after which time the
overflow from the cyclone stopped completely and flow
backed up through the mixing section. During the steady
operation the underflow sand was clean as in previous tests,
but it contained small shreds (1/4" - 1/2" long) of straw
which were quite black with oil. After tearing down the
cyclone, it was found that 2 to 3 inch lengths of straw
had formed a plug at the inlet to the vortex finder
(overflow port), effectively blocking the flow.
The test apparatus was completely cleaned out, a larger
vortex finder (2-1/2") was installed in the cyclone and the
test repeated. Now the unit operated at steady-state for
about 2 minutes processing about 50 pounds of sand per
minute with a water feed rate of 30 gpm. The overflow carried
46
-------
away most of the straw with the oil and wash water, however,
a considerable amount of straw was swept along with the
sand underflow. The effect of straw in the underflow was
to give a dirty appearance to the sand, since the straw still
retained sorbed oil.
Although we did not test other sorbing agents, it is possible
to extrapolate the results of the present program to the
processing of beach sands which contain quantities of other
available sorbers, e.g. talc, polyurethane foam chips
and a large assortment of commercially available products.
One common characteristic of the agents used for sorbing
oil is that when they are mixed with oil their specific
gravity is such that they will float on water. For this
reason, sorbers which enter the sand washing separation
system under study will for the large part pass out of the
system with the oil. As was seen in the tests discussed
above, some uncleaned straw came out in the sand underflow
and this is to be expected since the high volume flow of
sand can "trap" sorber and sweep it along.
It is evident that the sorber should be segregated from the
sand prior to feeding it through the sand washing system,
particularly if the sorber is of sufficient size to cause
plugging of the washer or cyclones. Also, sorber passing
from the sand separation stage with the oily water is in-
cluded as feed for the oil-water separation cyclones and
the effect it might have on this process is not known,
althoAigh it is not expected to be favorable.
Oil Water Separation Results
Evaluation parameters
Before presenting the results of the oil water separation
tests, it might be useful to examine some of the liquid-
liquid separation parameters and discuss their relationships.
The correlating parameter most frequently found in the liter-
ature on liquid-liquid separation in cyclones is the
separation number or efficiency, E . This number considers
s
the cyclone product streams to be pure phases (in the present
case water at the underflow, and oil at the overflow) and
then defines E as the sum of the rates of pure components,
S
expressed as a fraction of the feed flow.
47
-------
yf(i-yf)
where:
Qf = volume rate of feed
Q = volume rate of overflow
o
y = fraction of light component (oil) in feed
y = fraction of light component (oil) in overflow
This efficiency term is quite misleading when the objective
of the separation is to obtain a pure product of either
the light or heavy phase/ since high values of Es can be
reached with neither product stream in the pure state.
Better terminology for the present case might be to look at
the recovery of water at the underflow and the recovery
of oil at the overflow, R and R respectively.
w o J
and,
K0 - ( <>
where:
Q = volume rate of underflow
y = fraction of light component in underflow
Then, an R of unity means that all of the oil in the feed
was discharged to the overflow, while an R of unity means
that all of the water in the feed passed to the underflow.
48
-------
Two other terms of interest in analyzing the separation
data are:
_ . ^ overflow rate o
S = split = — - - ~ - — - = T~- (8)
^ underflow rate Q
and,
volume fraction of oil
in over flow __
C = enrichment factor = volume fraction of oil
in feed
There are various other manipulations which can be made
and which could prove useful in analyzing the data. For
instance, the oil recovery R can be re-written as:
Ro = Clifl1 (10)
and it is seen that increasing either C or S improves the
recovery of oil at the overflow.
The volume split S has been recognized in the cyclone
literature (Ref. 3) as being the primary controlling process
variable, as will be seen in the later discussion of the oil-
water separation data. The enrichment factor C is important
in that it describes the increase in oil concentration
across the stage. This parameter will be of particular
interest in the discussion on the conceptual design of a
full-scale sand washing system.
Evaluation of the data
In evaluating the results of the complete test program,
some data has been used more extensively than other in-
formation. For example, the tests numbered 1 through 18
Appendix B have been eliminated here because the data
did not fit the trends found in later work. There are
several reasons for this inconsistency.
49
-------
1) No attempt was made to control the feed characteristics
in this early series of tests. The oil and water were
placed in a tank and agitated with a small propeller,
also product streams from the cyclone were returned
to this same tank.
2) Feed mixture was pumped from the tank and the cyclone
inlet pressure was controlled via the throttle valve
used to control the flow rate of oil and water.
Varying the drop across this valve would create
significant differences in the feed to the cyclone.
Although this data is not included in the following discussion,
it should be noted that these early tests demonstrated the
feasibility of the concept of cyclone separation of oil and
water and pointed the direction for further work.
Most of the analyses were done with data from the one and
two stage 10 mm cyclone tests, supplemented with information
from the recycle and No. 6 oil tests. The results of the
6 inch cyclone work are shown in several of the figures,
however, they do not correlate as well as would be expected.
It is not certain at this time why this discrepancy exists,
but further test work with the larger cyclones would surely
eliminate some of this difference, or point out the proper
scaling factors.
An overview of all the data shows a pronounced effect on
the separation due solely to the volume split (overflow
rate divided by underflow rate) . In Figures 15 and 16, the
enrichment factor C is plotted as a function of S. For
Figure 15, data points were selected so that curves of
constant feed oil concentration would result. Study of
the figure reveals that there is little dependence on feed
oil concentration, in the range presented, on the enrichment
factor. As a result, all of the No. 4 oil data were replotted
in Figure 16, without regard to feed oil concentration.
This latter figure is indicative of the expected trend of
the data, since oil should preferentially be discharged
at the overflow and water at the underflow. As the per-
centage of feed reporting as overflow decreases (split
decreases), the concentration of oil in the overflow should
increase relative to that of the feed (enrichment increases).
50
-------
o
o
-p
0
-P
«
0)
o
•H
V4
c
H
0 6% oil
EJ 12% oil
18% oil
30% oil
1.5 -
1.0
Split
Figure 15. PERFORMANCE OF OIL-WATER SEPARATING CYCLONES
-------
Ul
to
u
a 3
O
4J
O
-------
And, as the percentage of overflow increases, with smaller
flows removed as underflow, the value of C should decrease
to unity. The enrichment factor can never go below unity,
unless there is something seriously wrong with the operation
of the cyclone. The recovery of oil R has not been
included in these graphs, and a high value of C does not
necessarily mean a high recovery.
From Figure 15 it appeared that feed oil concentration
did not markedly affect the enrichment factor; however,
by replotting C versus y , with constant values of S (Figures 17
and 18), a different conclusion is reached. At high splits
the earlier assumption is seen to hold true, but as the
split decreases to unity and below, feed oil concentration
does begin to influence the enrichment. Generally, the
enrichment factor is insensitive to y_ until the split
approaches 1 and then the enrichment factor decreases with
increasing yf.
The data in Figures 17 and 18 should not be extra-
polated beyond the feed concentrations shown. During the
testing it was found that the maximum overflow oil concentra-
tion attainable with the setup was 75 to 80% and if, for
instance, the S = 0.6 line in Figure 18 was extrapolated
to a y of 0.50 the expected overflow concentration would
be about 85% (C = 1.7). It is doubtful whether even this
small a gain would be possible.
In all tests after the first 18, a throttling valve was
used to blend the oil and water streams. The effect of
mixing valve pressure drop Ap on C is shown in Figure 19.
As expected, C decreases with increasing Ap , although the
drop is not too great. This behavior was also demonstrated
in the two stage testing in which the high shear (and
pressure drop) from the first stage cyclone did not appear
to affect the performance of the second stage.
Some of the early test work seemed to indicate that the
water type (tap, brackish, and sea water) and temperature
(water temperatures up to 155°F were used) affected the
separation. However, a more detailed evaluation of the
53
-------
2.2
u
c
-------
2.4
2.2
2.0
c
0)
e
x;
u
•H
n
c
w
1.8
1.6
1.4
1.2
1.0
I I T
S = 0.6
4 8 12 16 20 24 28 32 36 40 44 48 52 56
Feed Oil Concentration (% Volume)
Figure 18. EFFECT OF FEED CONCENTRATION ON ENRICHMENT
55
-------
2.0,
u 1.5
c
0)
I
o
-H
1.0
-0 5 %
12% Oil in feed
I
I
I
5 10 15 20
Mixing Valve Pressure Drop (psi)
Figure 19.
EFFECT OF PRESSURE DROP A>JD FEED CONCENTRATION
ON ENRICHMENT
56
-------
data refutes this original hypothesis. Although the increase
in water temperature decreases the oil viscosity significantly
and the salt water has emulsion inhibiting and breaking
properties, the gain in performance is small. Even the
higher specific gravity of sea water showed no great benefit.
Tables 2 and 3 show that the enrichment factor attained using
brackish (100 ppm NaCl in tap water) and sea waters could
be predicted using the results from the fresh water
testing. The only indicated benefit is a small increase
in oil recovery R over similar tests run with room temperature
fresh water.
Even the type of oil appeared to be of little significance
in determining the enrichment factor C. Table 4 shows
predicted C values for No. 6 oil, using data from the
No. 4 oil tests, against the C values actually attained.
The No. 6 oil tests were run with the oil heated to about
120°F to aid in pumping, and this may have had some
beneficial effect on the performance. Only further testing
with the heavier oil at reduced temperatures will settle
this question.
In Table 5, the predicted performance for the recycle tests
is compared with that actually attained with the two stage
recycle setup. There is some discrepancy in these data, but
the maximum error in predicting C is only 8%.
As may be seen in Figure 16, the data attained with the six
inch cyclones did not agree well with the 10 mm data and
it was not always possible to predict the performance of these
larger units from the available data (Table 6). At high
split values, the enrichment factor could be predicted
but at values near or below 1, that correlation is poor.
Careful examination did reveal that the data from the plastic
cyclone were more predictable than that from the rubber lined
FR unit. This seeming anomaly might be explained by the
closer geometric similarity between the 6 inch plastic and
the 10 mm units than the FR and 10 mm. In any case, the
complete explanation for the behavior of the 6 inch unit
must await further testing and analysis.
In summary, the feasibility of oil-water separation in
cyclones has been demonstrated and sufficient data are
57
-------
TABLE 2
EFFECT OF WATER TYPE ON PREDICTING ENRICHMENT
Oil in Feed
(Vol. %)
Split Water Type Enrichment Factor
Predicted Actual
13.2
8.3
9.0
17.4
8.3
6.5
6.6
9.3
8.4
2.24
1.91
1.54
6.52
1.40
2.69
1.74
1.49
1.44
Salt*
Salt
Salt
Salt
Sea
Sea
Sea
Sea
Sea
1.42
1.51
1.60
1.15
1.62
1.36
1.55
1.62
1.63
1.44
1.53
1.62
1.15
1.65
1.34
1.58
1.67
1.68
* Salt Water - 100 ppm NaCl added to tap water
58
-------
EFFECT
TABLE 3
OF FEED TEMPERATURE ON PREDICTING ENRICHMENT
Oil in Feed Split
(Vol %) (S=Q /Q )
8.1
11.2
6.7
10.0
6.1
6.9
5.7
8.4
5.3
9.0
7.7
8.0
6.2
2.54
1.09
1.98
1.77
6.34
2.34
1.98
1.40
1.36
1.63
1.43
1.25
2.17
Temperature Enrichment Factor
(°F) (C=yb/yf)
Actual Predicted
55
55
55
55
115
115
152
152
122
122
76
76
105
1.38
1.76
1.49
1.55
1.13
1.36
1.47
1.72
1.70
1.61
1.66
1.90
1.45
1.36
1.71
1.45
1.56
1.12
1.38
1.48
1.66
1.70
1.58
1.65
1.94
1.44
59
-------
TABLE 4
EFFECT OF OIL TYPE ON PREDICTING ENRICHMENT
Oil in Feed
(Vol %)
8.4
13.9
7.7
12.1
3.8
6.8
6.2
7.6
Split
(S=Q /Q )
o u
1.47
1.59
1.40
1.08
1.20
0.6
4.30
1.20
Enrichment Factor
(C=yd/yf)
Actual* Predicted**
1.66
1.58
1.57
1.76
1.79
2.20
1.23
1.79
1.64
1.56
1.66
1.74
1.75
2.20
1.24
1.74
*
Data from tests with No. 6 oil
**
Predicted from test results with No. 4 oil
60
-------
TABLE 5
EFFECT OF CYCLONE UNDERFLOW RECYCLE ON
PREDICTING ENRICHMENT
Oil in Feed Split Enrichment Factor
(Vol %) (s = QQ/QU) (C=yb/yf)
Actual Predicted
4.
5.
4.
7.
13.
22.
6
4
7
6
2
4
5.
1.
1.
0.
1.
4.
11
99
28
9
36
20
1.
1.
1.
1.
1.
1.
17
46
62
96
70
19
1.
1.
1.
1.
1.
1.
17
48
72
80
68
24
61
-------
TABLE 6
PREDICTION OF 6 INCH CYCLONE OPERATION
Oil in Feed
(Vol %)
4.9
3.4
5.7
4.6
2.8
3.4
2.7
8.7
4.3
5.0
6.2
Split
(S - QQ/Qu)
3.7
0.24
6.1
0.48
0.44
6.1
0.44
20.0
1.09
1.15
0.65
Cyclone
Type
PI*
PI
PI
PI
PI
PI
PI
FR
FR
FR
FR
Enrichment Factor
(c=y0/yf)
Actual Predicted
1.24
3.79
1.14
1.37
2.75
1.15
1.68
1.05
1.49
1,44
1.63
1.26
3.30
1.15
2.10
2.15
1.10
2.15
1.0
1.78
1.76
2.0
PI - Plastic
FR - Rubber Lined
62
-------
available to predict the performance of a system needed
to process the effluent from the sand washing machine (or
from any other oil-water treatment system, for that matter).
The failure of the present technique to achieve an oil
overflow with only 3% water should not be construed as a
failure for the entire concept. Rather, the testing showed
that the oil could be concentrated up to about 75% of the
volume of the waste effluent stream, opening up many new
possible routes for disposal of the oil, e.g. incineration,
use as low grade fuel, or further processing in centrifuges
or other equipment to achieve the low water content needed
in order to make the oil acceptable for recycle to a refinery,
It is not certain whether this last route is even open to
the oil collected from the sand washing system. As was
pointed out earlier, sorber material is discharged with the
oil-water overflow from the sand separating cyclone and
would likely pass through the final oil overflow stream
from the oil-water separation system. In this event, the
oil would be contaminated with a foreign material which might
make it unacceptable for refinery feedstock.
The possibilities of improving the oil overflow in cyclones
of different geometry, or with changes in the operating
variables still remain and should not be ruled out in the
consideration of future test programs.
Concept for a Complete Beach Cleaning System
The objectives of the extensive test programs of sand
washing and oil-water separation were to demonstrate.that
the processes involved in the proposed mobile, beach cleaning
plant were viable and also to obtain data for a preliminary
design of such a unit. Presented below is the preliminary
design for a mobile system and recommended logistics and
operating features.
System design
Shown in Figure 20 is the process flowsheet for the complete
beach cleaning system. This concept includes screens and
trash racks for removal of sorbers, normal beach debris and
trash from the feed stream. Since problems were encountered
63
-------
IUPUT FKOM AEOCH
TZJ
SOUD/OIL MIXJUCE
3U)(JDD£JJO TD DGP6SAL
Figure 1^0. SAKD WASHING PROCESS FJ.OWSill-'TJT
-------
in processing sand with small quantities of straw mixed in,
it is important that the screening (a flotation system could
be used) system be designed to remove even very fine (down to
1/4") debris to prevent plugging of the process equipment.
The scavenged materials will likely be coated with oil and so
must be properly disposed of. This is discussed further below.
The sand, oil and residual trash passing the screening stage
is loaded into the sand feeder hopper. A screw feeder (or
other type of positive displacement solids feeder) transfers
the oily sand to the washing stage.
As in the pilot scale test rig, the oil is stripped from the
sand in a washer/mixer and then passes to a cyclone for
separation of the sand from the washing fluid and oil. At
the underflow of the cyclone, the sand may be further dewatered,
since it is expected that at least 15% of the influent wash
water will be needed to transport sand from the cyclone. The
water recovered at this point is either returned to the inlet
of the washer feed pump, or is cycled to a disposal tank. The
sand from the washer may be dumped directly onto the beach.
The oil-water mixture reporting at the overflow of the
sand separating cyclone then passes to the second stage of the
process, i.e. the liquid-liquid separation. Shown on the
flowsheet is a two-stage process for recovering oil, although
the ultimate selection of the number of stages is largely
dependent on the concentration of oil desired. If the
recovered oil is to be as high as 75 to 80% of the final
overflow stream (yQ), then 5 to 6 stages of separation
are indicated from the test results. The underflow streams
from these cyclones are shown being returned to the washer
feed pump and blended with makeup. It may be necessary to
recycle the underflow from the first stage (oil-water
separation) cyclone as blowdown from the system to prevent
the buildup of a stable emulsion of oil. Also, instead of
returning the underflow from the second stage to the washer,
it could be recycled directly into the inlet of the first
stage oil-water separator as was done in some of the recycle
tests.
Variations in the process parameters could be optimized
for any particular specified feed characteristics and
required product streams. This optimization might be based
65
-------
on minimizing the capital investment for a given feed rate
or increasing the throughput capacity for a specified set
of equipment.
Figure 21 shows a suggested arrangement for the apparatus
on a transportable skid. Table 7 gives an indication of
the size and capacity of some of the equipment needed to
process 100 ton/hour of oil-contaminated sand.
LogjLstics and ppe^ratj-ons
There are several important areas in regard to the logistics
of cleaning an oil-contaminated beach that need to be
addressed. In the first place, the system as proposed is
mobile, i.e. it can travel to the spill site over land at
trucking speeds (where road access is possible) and proceed
across the beach, under its own power. If direct access
to the spill site is not possible, the mobile rig could be
transported to the shore by a landing craft or similar
vessel. This approach could also be used to enable the
cleaner to "leapfrog" natural and manmade shoreline
obstructions, e.g. mouths of rivers, harbors, cliffs, groin,
jetties, etc.
The collection of the contaminated sand for loading onto
the beach cleaner is accomplished using conventional road
building equipment. This was suggested in our proposals
in early 1969, and has since been proved to be a viable
approach (Ref. 12). Figure 22 is an artist's rendition of
the proposed beach cleaning plant showing some of the features
of the system as discussed here.
Ultimate disposal of effluent streams
Disposal of the effluent oil, wash water and screened
trash streams is a matter of great concern. In the case of
the trash, incineration followed by land-fill disposal of
the residue seems to be the only realistic approach. No
new development work needs to be done, both the technology
and equipment are available, only the logistics of who and
where remain to be settled.
The wash water and blowdown streams, containing small
mass fractions of oil, present a more difficult problem.
66
-------
cycuM/ic.
OIL 77WK
fO.
//
12.
/
2.
3.
-+.
5, i/er GOM7BCKXL LJASHEIL
e.
7.
Figure 21. BLOCK DIAGRAM OF SYSTEM ARRANGEMENT
-------
TABLE 7
SYSTEM COMPONENTS (Refer to Figure 21]
Component
Sand Screens
Conveyors
Hopper
Jet Washer
Cyclone No.l
(Sand Removal)
Cyclone No. 2
No.
2
2
1
1
2
2
Capacity
50 T/hr
50 T/hr
lOOT/hr
lOOT/hr sand
1200 gpm water/
oil
it
1200 gpm water/
Size
4'x4'x6t
S'xS'xlO1
S'xS'xe1
2'x6'x61
24" OD
24" OD
(oil-water separation
stage #1)
Cyclone No.3 3
(oil-water separation
Stage #2)
Sand Dewaterer 1
(Optional)
Main Pump 1
(Diesel or Gasoline)
Transfer Pump 1
(To Tank Truck)
Water Pump (Suction) 1
Oil Storage Tank 1
Control Panel 1
Water Surge Tank 1
oil
120 gpm
lOOT/hr sand
25T/hr water
500 gpm, 10 hp
200 gpm, 5 hp
3000 gallons
600 gallons
6" OD
3'x4'xlOI
1200 gpm, lOOhp 4'x4'xl21
4'x4'x6'
3'x3'x4'
5'x7'xl2
2'x4'x6I
4'x4'x5l
68
-------
-
:
Figure 22. BEACH RESTORATION PROCESS
-------
Dumping of this oiled water at land fills could present
a severe environmental hazard and would likely be in viola-
tion of environmental protection ordinances. Flotation
(straight gravity or froth induced), filtering or coalescing
schemes might prove suitable as a means for removing the
oil from the water stream. If the oil is in the form of
an emulsion, it could be chemically treated to break the
emulsion, followed by separation in a flotation cell. Then,
the clean water can be returned to its source. The re-
covered oil is easily disposed of by incineration since it
would constitute a very small volume.
Finally, disposing of the bulk recovered oil from the
overflow of the final oil-water separation cyclone is a
sensitive area. As it is presently viewed, the oil stream
from this washing process is very likely to be contaminated
with debris and water beyond the point where recycling to
a refinery is advisable. Further, the economic penalty
of trying to get it into an acceptable condition may far
outweigh the value of the oil as a source of feedstock.
If these arguments are valid, then the only obvious solution
to the oil disposal problem is incineration. Incineration
meaning firing of the total waste stream, including its
debris and water, and possible recovery of the heat
energy to whatever use it may be put. The idea of trying
to achieve a useful end for the recovered oil should only
be of secondary importance. The primary objective of the
whole beach cleaning concept is to quickly and effectively
remove the oil from a contaminated beach and restore it
to, as nearly as possible, its natural state.
70
-------
SECTION 7
ACKNOWLEDGEMENTS
All of the test work and report preparation under this
program were carried out by Creare Incorporated, Hanover,
New Hampshire and Dorr Oliver Incorporated, Stamford,
Connecticut, under subcontract to Ecological Research
Corporation, Miami, Florida. Francis X. Dolan of Creare
and James P. Bowersox of Dorr-Oliver were the project
engineers principally responsible for the conduct of the
program.
The authors wish to thank the EPA Project Officers
responsible for monitoring this program, Ralph L. Rhodes
and Richard R. Keppler, for their help, encouragement and
patience throughout this undertaking.
71
-------
SECTION 8
REFERENCES
1) Gilmore, G. A., et al; SYSTEMS STUDY OF OIL SPILL
CLEANUP PROCEDURES, VOL. I: ANALYSIS OF OIL SPILLS
AND CONTROL MATERIALS; Dillingham Corporation, Final
Report to Committee for Air Water Conservation, American
Petroleum Institute, Publication No. 4024, February 1970.
2) Secretaries of Interior and Transportation; OIL POLLUTION;
A Report on Pollution of the Nation's Waters by Oil
and Other Hazardous Substances, February 1968.
3) Bradley, D.; THE HYDROCYCLONE; Pergamon Press, Oxford, 1965.
4) Talcott, R. M. (Dorr-Oliver Incorporated), Private
Communication, September 8, 1969.
5) Krumbein, W. C.; A METHOD FOR SPECIFICATION OF SAND FOR
BEACH FILLS; U. S. .Army Corps of Engineers, Beach
Erosion Board, Technical Memorandum No. 102, 1957.
6) MacCarthy, G. R.; COASTAL SANDS OF THE EASTERN UNITED
STATES; American Journal of Science, Vol. 22, 1931.
7) Martens, J. H. C.; BEACH SANDS BETWEEN CHARLESTON,
SOUTH CAROLINA AND MIAMI, FLORIDA; Bulletin of the
Geological Society of America, Vol. 46, 1935.
8) Marx, W.; WAYWARD BEACHES; Oceans, Vol. 1, No. 3, 1969.
9) Emory, K. 0.; SOME CHARACTERISTICS OF SOUTHERN CALIFORNIA
SEDIMENTS; Journal of Sediment Petrology, Vol. 24, 1955.
10) Twenhofel, W. H.; BEACH AND RIVER SANDS OF THE COASTAL
REGION OF SOUTHWESTERN OREGON WITH PARTICULAR REFERENCE
TO BLACK SANDS; American Journal of Science, Vol. 244,
1946.
11) Anonymous; CLEANING OIL CONTAMINATED BEACHES WITH CHEMICALS;
A Study of the Effects of Cleaning Oil Contaminated
Reaches With Chemical Dispersants, FWPCA, Department of
the Interior, August 1969.
73
-------
12) Anonymous; PRELIMINARY OPERATIONS PLANNING MANUAL FOR
THE RESTORATION OF OIL-CONTAMINATED BEACHES; URS Research
Company for FWPCA, Department of the Interior, February 1970
74
-------
SECTION 9
NOMENCLATURE
C = oil enrichment factor
E = oil-water separation efficiency
s
f = fraction of oil in sand
OR - oil removal effectiveness
Ap = pressure drop
p = pressure
Q = volume flow rate
R — recovery of oil or water
S = volume split
SE = specific energy
W = mass flow rate
y = volume fraction of oil in water
4) = power
Subscripts
c = cyclone
d - mixing valve in oil water separator
f = feed
i = feed stream or initial (in context)
o = oil, overflow stream or final (in context)
s = sand
T = total
u = underflow stream
w - washer
75
-------
SECTION 10
APPENDICES
Page No,
A. RANGE OF VARIABLES USED IN SAND WASHING PROGRAM . . 78
B. OIL-WATER TEST PROGRAM VARIABLES 83
77
-------
APPENDIX A
RANGE OF VARIABLES USED IN SAND WASHING PROGRAM
Oil Type #6
Oil Type #4
Test No.
1111-01
1111-02
1111-03
1111-04
-i 1111-05
00
1111-06
1113-01
1113-02
1113-03
1113-04
1113-05
1113-06
1113-07
1123-01
1123-02
Water
Flow
(GPM)
30
30
30
30
30
30
20
20
20
20
40
40
40
30
30
Sand
Flow
(Ib/min)
62.5
44
28
28
44
62.5
55
41.5
29.5
18.5
82
69.5
37
82
62.5
for tests 1111-01 through 1203-07
for tests 1214-01 through 1218-11
Oil
in Sand
(mass %)
6.2
6.2
6.2
6.2
6.2
6.2
6.65
6.65
6.65
6.65
6.65
6.65
6.65
5.85
5.85
Plenum
Pressure
(psig)
13
13
13
12
12
12
5
5
5
5
22
22
22
11
11
Mixer
Pressure
(psig)
4.5
4.5
4.5
4.5
4.5
4.5
2.5
2.5
2.5
2.5
6
6
6
2.25
2.5
Mixer
Length
(inch)
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
Washer
Nozzle
(No.)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
-------
Test No.
1123-03
1123-04
1123-05
1123-06
1123-07
1123-08
1123-09
1123-10
1123-11
1124-01
1124-02
1124-03
1124-04
1124-05
1124-06
1124-07
1124-08
1124-09
Water
Flow
(GPM)
30
30
20
20
20
20
40
40
40
30
30
30
30
20
20
20
20
40
Sand
Flow
(Ib/min)
44
28
55
41.5
29.5
18.5
82
69.5
37
82
62.5
44
28
55
41.5
29.5
18.5
82
APPENDIX A - (Cent.)
Oil Plenum Mixer
in Sand Pressure Pressure
(mass %) (psig) (psig)
5.
5.
5.
5.
5.
5.
5.
5.
5.
7.
7.
7.
7.
7.
7.
7.
7.
7.
85
85
85
85
85
85
85
85
85
45
45
45
45
45
45
45
45
45
11
11
5
5
5
5
18
18
18
10
10
10
10
5
5
5
5
17
2.
2.
1.
1.
1.
1.
3
3.
3.
3
3
3
3
2.
2.
2.
2.
3
75
75
5
5
5
5
25
5
5
5
5
5
Mizer
Length
(inch)
24
24
24
24
24
24
24
24
24-
12
12
12
12
12
12
12
12
12
Washer
Nozzle
(no.)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
-------
CO
o
Test No.
1124-10
1124-11
1201-01
1201-02
1201-03
1201-04
1201-05
1201-06
1201-07
1202-01
1202-02
1202-03
1202-04
1202-05
1202-06
1202-07
1203-01
1203-02
Water
Flow
(GPM)
40
40
30
30
30
30
40
40
40
30
30
30
30
40
40
40
30
30
Sand
Flow
( Ib/min)
69.5
37
82
62.5
44
28
82
69.5
37
82
62.5
44
28
82
69.5
37
82
62.5
APPENDIX A (Cont)
Oil Plenum
in Sand Pressure
(mass %) (psig)
7.45
7.45
5.53
5.53
5.53
5.53
5.53
5.53
5.53
5.30
5.30
5.30
5.30
5.30
5.30
5.30
5.041
5.041
17
17
13
13
13
13
23
23
23
13
13
13
13
24
24
24
14
14
Mixer
Pressure
(psig)
3
3
3
3
3
3
4.75
4.75
4.75
3.50
3.50
3.50
3.50
7
7
7
4.5
4.5
Mixer
Length
(inch)
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
24
24
Washer
Nozzle
(no.)
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
00
Test No.
1203-03
1203-04
1203-05
1203-06
1203-07
1214-01
1214-02
1214-03
1214-04
1214-05
1214-06
1214-07
1214-08
1214-09
1214-10
1214-11
1218-01
1218-02
Water
Flow
(GPM)
30
30
40
40
40
30
30
30
30
40
40
40
20
20
20
20
30
30
Sand
Flow
( Ib/min)
44
28
82
69.5
37
82
54
44
28
82
69.5
37
55
41.5
29.5
18.5
82
62.5
APPENDIX
Oil
in Sand
(mass %)
5.041
5.041
5.041
5.041
5.041
4.58
4.58
4.58
4.58
4.58
4.58
4.58
4.58
4.58
4.58
4.58
5.091
5.091
A (Cont.)
Plenum
Pressure
(psig)
14
14
26
26
26
16
16
16
16
27
27
27
5
8
8
8
14
14
Mixer
Pressure
(psig)
4.5
4.5
9
9
9
6
6
6
6
9
9
9
3.75
4
4
4
4
4
Mixer
Length
(inch)
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
24
12
12
Washer
Nozzle
(no.)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
00
APPENDIX A (Cont.)
Test No.
1218-03
1218-04
1218-05
1218-06
1218-07
1218-08
1218-09
1218-10
1218-11
Water
Flow
(GPM)
30
30
40
40
40
20
20
20
20
Sand
Flow
(Ib/min)
44
28
82
69.5
37
55
41.5
29.5
18.5
Oil
in Sand
(mass %)
5.091
5.091
5.091
5.091
5.091
5.091
5.091
5.091
5.091
Plenum
Pressure
(psig)
14
14
26
26
26
7
7
7
7
Mixer
Pressure
(psig)
4
4
9
9
9
3.5
3.5
3.5
3.5
Mixer
Length
(inch)
12
12
12
12
12
12
12
12
12
Washer
Nozzle
(no.)
1
1
1
1
1
1
1
1
1
-------
APPENDIX B
OIL WATER SEPARATION IN 10 MM DORRCLONE
oo
Test
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Feed
Rate
(GPM)
2.02
2.54
2.92
3.90
2.89
2.78
2.51
2.98
2.76
2.04
2.05
3.14
3.37
3.33
2.63
3.47
Oil Concentration Under
Feed
(Vol %)
3.50
3.10
2.43
1.59
2.54
2.53
2.81
3.16
3.23
1.96
1.62
1.44
1.45
1.48
3.63
2.51
Overflow
(Vol %)
4.57
4.71
3.29
2.27
3.50
3.23
3.11
3.19
4.62
2.36
1.78
1.82
1.68 .
1.74
6.33
4.24
Flow
(Vol %)
2.62
1.72
1.37
0.49
1.53
1.21
1.16
3.15
2.07
1.18
0.96
0.86
0.66
1.32
0.99
0.38
Split
0.83
0.87
1.24
1.62
1.05
1.90
5.50
0.38
0.84
1.97
4.16
1.51
3.48
0.63
0.98
1.23
Recovery
Oil (%)
59.3
70.4
74.8
88.4
70.7
83.8
93.7
28.0
65.1
79.6
88.6
75.9
90.1
45.6
86.0
93.2
Water (%)
55.2
54.5
45.1
38.6
49.3
34.9
15.7
72.3
55.2
33.9
19.5
40.0
22.5
61.3
52.0
45.8
-------
CO
Test
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Feed
Rate
(GPM)
2.92
3.49
2.74
2.81
2.96
3.08
3.32
2.85
3.10
1.42
1.59
2.74
2.67
1.46
1.43
3.67
3.51
3.63
APPENDIX B
Oil Concentration
Feed
(Vol %)
2.25
3.32
6.9
10.3
11.1
11.8
18.4
6.7
12.3
6.8
14.7
6.0
5.9
8.3
7.9
2.2
3.9
10.7
Over-
Flow
(Vol %)
2.82
4.94
12.8
17.8
19.8
20.5
32.1
11.7
20.3
11.1
25.1
8.9
7.2
11.9
9.6
3.6
7.1
16.0
Under-
Flow
(Vol %)
0.44
0.50
0.9
2.7
2.2
2.8
4.1
1.3
4.1
1.3
3.1
0.6
0.5
0.6
0.3
0.7
0.3
4.8
- (Cont.)
Split
3.12
1.75
1.01
1.02
1.03
1.04
1.05
1.07
1.01
1.25
1.12
1.88
4.13
2.13
4.50
1.15
1.15
1.10
Recovery
Oil (%)
95.4
94.6
93.6
87.4
90.2
88.7
89.1
90.3
83.0
90.6
90.3
96.6
98.2
97.2
99.8
85.4
97.6
78.4
Water (%)
24.4
37.5
52.6
53.6
54.2
53.9
57.6
51.3
54.2
47.1
53.6
36.6
20.7
34.6
19.7
47.5
48.3
51.0
-------
00
U1
Test
35
36
37
38
20-A
22-A
25-A
29-A
39
40
41
42
43
44
45
46
47
Feed
Rate
(GPM)
3.40
3.20
3.22
3.04
2.71
2.73
2.99
2.96
2.70
3.04
1.43
2.72
2.69
1.53
2.29
1.67
1.73
APPENDIX B (Cont.)
Oil Concentration
Feed
(Vol %)
3.6
3.0
4.7
6,1
5.9
14.3
12.1
12.0
5.1
11.0
7.3
5.3
14.0
7.6
32.1
29.4
52.9
Over-
Flow
(Vol %)
5.2
3.6
8.2
7.4
10.9
26.5
20.7
15.0
8.8
19.6
9.0
10.2
25.9
9.2
63.9
58.8
78.8
Under-
Flow
(Vol %)
0.3
0.1
0.8
0.2
0.9
1.9
3.1
0.9
1.0
2.0
1.1
0.05
1.9
. 1.0
2.2
10.2
31.1
Recovery
Split
2.12
4.93
1.12
4.74
1.01
1.02
1.05
3.70
1.08
1.04
3.77
1.09
1.02
4.28
0.941
0.653
0.840
Oil (%)
97.8
99.7
91.9
99.9
92.4
93.4
87.6
98.4
90.5
90.8
97.2
99.7
93.6
98.0
96.5
79.0
68.1
Water ( % )
33.1
17.4
49.1
18.4
52.3
56.4
53.7
24.1
50.0
53.9
22.5
50.7
56.3
20.3
74.2
76.8
79.3
-------
00
Test
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
Feed
Rate
(GPM)
2.91
3.16
2.90
1.48
1.35
1.49
1.49
2.30
1.55
2.18
1.37
1.33
1.56
2.34
2.12
APPENDIX B
Oil Concentration
Feed
(Vol %)
11.6
4.6
19.1
7.7
8.0
6.2
13.2
8.3
9.0
17.4
8.3
6.5
6.6
9.3
8.4
Over-
Flow
(Vol %)
22.9
21.9
40.6
12.8
15.2
9.0
19.0
12.5
14.6
20.0
13.7
8.7
10.4
15.5
14.1
Under-
Flow
(Vol %)
5.9
1.3
8.3
0.4
2.3
0.05
0.2
0.01
0.3
0.1
0.8
0.5
0.1
0.1
0.3
(Cont.)
Split
0.508
0.188
0.503
1.43
0.80
2.17
2.24
1.91
1.54
6.52
1.40
2.69
1.76
1.49
1.44
Recovery
Oil (%)
66.4
75.2
71.1
97.6
84.4
99.3
99,5
100.0
98.2
99.7
96,4
97.7
100.0
99.9
99.1
Water (%)
70.3
86.7
75.3
44.5
58.9
33.7
35.5
37.4
43.3
16.1
44.9
28.7
39.1
44.2
44.7
-------
APPENDIX B (Cont.)
TWO STAGE OIL-WATER SEPARATION
IN A 10MM DORRCLONE
No. 4 Oil
Tap water (Tdst 67 , 100 ppm NaCl in water)
Flow rates in GPM Split
Test
63
64
65
66
67
First Stage
Feed
4.79
3.74
3.87
4.34
4.38
U'Flow 0'
1.35 3
1.26 2
0.53 3
1.46 2
1.86 2
Second Stage First
Flow U'Flow O'Flow Stage
.44 1.65
.48 0.89
.34 1.00
.88 1.20
.52 0.96
1.79 2.54
1.59 1.98
2.34 6.34
1.68 1.98
1.56 1.36
Second
Stage
1.09
1.77
2.34
1.40
1.63
Test
63
64
65
66
67
Feed
8.1
6.7
6.1
5.7
5.3
Oil
First Stage
U'Flow
0.4
0.3
1.3
0.2
0.3
Concentration
0 ' Flow U '
11.2 1
10.0 0
6.9 0
8.4 0
9.0 0
(Vol %)
Second Stage
Flow 0 ' Flow
.9 19.7
.3 15.5
.1 9.4
.1 14.4
.1 14.5
87
-------
APPENDIX B (Cont.)
TWO STAGE OIL-WATER SEPARATION IN
10MM DORRCLONE
No. 6 Oil
Tap water tests 75, 76
Salt Water (100 ppm Nad) tests 77, 78
Test
75
76
77
78
Flow .Rate in GPM
Feed
3.97
3.77
3.37
3.50
First Stage
U'Flow O'Flow
1.61 2.36
1.57 2.20
1.53 1.84
0.66 2.84
Second
U'Flow
0.91
1.06
1.15
1.29
Oil Concentration
Test
75
76
77
78
Feed
8.4
7.7
3.8
6.2
First Stage
U ' Flow 0 '
0.3 13
1.0 12
0.3 6
0.1 7
Flow
.9
.1
.8
.6
Stage
O' Flow
1.45
1.14
0.69
1.55
(Vol %)
Second
U'Flow
1.0
2.0
1.8
0.3
Split
First
Stage
1.47
1.40
1.20
4.30
Stage
O'Flow
22.0
21.3
15.0
13.6
Second
Stage
1.59
1.08
0.6
1.20
88
-------
APPENDIX B (Cont.)
Recycle Tests with 10 mm DorrClones
No. 4 Oil
Salt water (100 ppm NaCl)
Water temperature 125°F to 140°F
Flow rates in GPM
Test
79
80
81
Test
79
80
81
First Stage
Feed U'
3.36 0
3.24 1
3.70 1
Feed
Fresh
6.1
6.2
14.3
Flow 0
.55
.42
.57
Oil
with
Recycle
4.6
4.7
13.2
Second Stage
Split
First Second
'Flow U'Flow O'Flow Stage Stage
2.81
1.82
2.13
0.
0.
0.
94 1.87
96 0.86
41 1.72
5.11
1.28
1.36
1.
0.
4.
99
90
20
Concentration (Vol %)
First
U'Flow
0.1
0.9
0.7
Stage
0 ' Flow
5.4
7.6
22.4
Second
U'Flow
0.5
1.0
4.7
Stage
O'Flow
7.9
14.9
26.7
89
-------
Test
No.
68
69
70
71
72
73
74
Feed
Rate
( GPM)
42.0
50.8
38.5
50.4
50.7
36.1
50.7
APPENDIX B
OIL WATER SEPARATION IN 6
Plastic Cyclone
No. 4 Oil
Water temperature, 66 °F
72, 73, 74
Tap water tests 68, 69, 7
71, 72, 73, 74
Oil Concentration
Over Under
Feed Flow Flow
(Vol %) (Vol %) (Vol %)
4.9 6.1 0.4
3.4 12.9 1.0
5.7 6.5 0.9
4.6 6.3 0.9
2.8 7.7 0.6
3.4 3.S 0.4
2.7 6.2 1.1
(Cont.)
INCH DORRCLONE
tests 68, 69
0 . tap plus
Split
3.7
0.24
6.1
0.48
0. 44
6.14
0.44
, 70, 71;
100 ppm;
130°F tests
NaCl for
Recovery
Oil
%
98.2
76.6
97.8
94.0
84.2
98.4
71.3
Water
%
22.4
81.7
14.7
33.6
71.0
14.6
70.5
-------
VD
APPENDIX B
OIL-WATER SEPARATION
(Cont.)
IN 6 INCH
DORRCLONE
Rubber lined (FR) cyclone
No. 4 oil
Water temperature 13 0-150 °F
Tap water plus 100 ppm NaCl
Oil Concentration
Test Feed
No. Rate
(GPM)
82 42.4
83 59.4
84 43.4
85 61.4
Feed
(Vol.%)
8.7
4.3
5.0
6.2
Over Under
Flow Flow
(Vol.%) (Vol.%)
0.9 9.1
2.0 6.4
2.4 7.2
3.7 10.1
Split
20.2
1.09
1.15
0.65
Recovery
Oil Water
% %
99.5 5.1
77.7 49.0
77.5 47.7
63.8 62.4
-------
APPENDIX B (Cont.)
HIGH OIL CONTENT TESTS IN
10MM DORRCLONE
(Two
Stage Separation)
No. 4 oil
Salt Water (100
Test
86
87
88
89
90
92
93
95
Test
86
87
88
89
90
92
93
95
ppm NaCl)
in GPM
Feed
2.33
2.43
3.41
2.46
2.24
2.94
3.45
3.04
First Stage
U'Flow O1
0.09 2
0.38 2
1.30 2
0.92 1
0.32 1
0.30 2
0.58 2
0.38 2
Flow
.24
.05
.11
.54
.92
.64
.87
.66
Second Stage First
U'Flow O'Flow Stage
1.39 0.85 24.9
1.28 0.77 5.39
1.32 0.79 1.62
0.73 0.81 1.67
0.77 1.15 6.0
1.20 1.44 8.8
1.37 1.50 4.94
1.41 1.25 7.0
Second
Stage
0.61
0.60
0.60
1.11
1.49
1.20
1.09
1.13
Oil Concentration (Vol %)
Feed
28.6
37.4
22.2
30.3
56.9
91.3
79.6
90.7
First Stage
U'Flow
0.9
1.3
0.3
6.5
42.9
89.8
78.5
89.7
O'Flow
29.7
44.0
35.7
44.5
59.2
91.5
80.0
90.8
Second Stage
U'Flow O'Flow
12.3 58.3
23.6 73.3
18.9 63.8
9.5 76.3
40.5 71.6
88.8 93.8
80.0 80.0
90.5 91.5
*US. GOVERNMENT PRINTING OFFICE: 1973 514-155/316 1-3
-------
SELECTED WA TER i- Report No.
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
2. 3. Accession No.
w
4. Title Development of a Mobile System for Cleaning s- Report Date
Oi 1-Contaminated Beaches 6.
7. Author(s) o
Dolan, Francis, X. and Bowersox, James P.
9. Organization
Ecological Research Corporation
(1) Creare Inc., Hanover, N.H
8. Performing Organization
Report No.
10. Project No.
FIG
11. Contract/Grant No.
14-12-830
(2) Dorr Oliver, Inc. Stamford, Conn. ' 13, Type of Report and
Period Covered
12. Sponsoring Organization Environmental Protection Agency
15. Supplementary Notes
Environmental Protection Agency report
number, EPA-R2-73-233, May 1973.
16. Abstract A system has been developed for the restoration of oi 1-Eontaminated beach
sands. The method involves washing of the sand In a high energy jet contactor washer
and separation of the cleaned sand from the wa shing fluid in a conventional solid-
liquid cyclone. Separation and concentration of the oil-water effluent from the washing
process is also accomplished in cyclones. The two separate stages of this process have
been demonstrated on a pilot scale equivalent to about 3 tons of wef, oil contaminated
sand per hour.
The sand washing process has been shown capable of removing over 99% of the contaminating
oil from a simulated beach sand. Oils used were No.4 and No.6 fuel oil at 4 to 8% of the
dry weight of the sand. The oil-water separation tests yielded a highly enriched oil
product stream with less than 20% water, while the water removed from the system was
suitable for recycle to the sand washing system.
A conceptual design for a mobile beach cleaning system based on the processes studied
is presented and is shown to be feasible within the state-of-the-art.
This report was submitted in fulfillment of Project FIG, Contract No. 14-12-830, under
the sponsorship of the Environmental Protection Agency.
na.Descriptors *0\ 1 Spill Cleanup, *Ljquid-Liquid separation, *Sand Washing
Hydrocyclones, Beach Cleanup
17b. Identifiers
17c. COWRR Field & Group
18. Availability "• Security Class.
(Report)
20. Security Class.
(Page)
Abstractor \Institution
21. No. of Send To:
Pages
22. Price WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 20240
WRSIC 102 (REV. JUNE 1971) G p 0 913.26!
------- |