EPA R2-72-045
Environmental Protection Technology Series
September 1972
Restoration of Beaches
Contaminated by Oil
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;
!• Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
<*. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of polluv.4.on sources to meet environmental quality
••standards*, y . ••;-:^:'r\': -v,-- •. '•• •'.••• -AA V-AV/-- ••-•": '..-'•' • : A'. : • ":
-------
EPA-R2-72-045
September 19J2
RESTORATION OF BEACHES CONTAMINATED BY OIL
By
Garth D. Gumtz
Contract No. 14-12-809
Project 15080 EOT
Project Officer
Ralph Rhodes
EPA - Region III
6th & Walnut St., Curtis Bldg.
Philadelphia, Pennsylvania 19106
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20^60
-------
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or
recommendation for use.
ii
-------
ABSTRACT
Based on laboratory studies, a 30 ton per hour pilot plant was built for cleaning
oil contaminated beach sands. The plant utilized the principle of froth flotation.
Extensive field testing considered different oils, feed concentrations, both
brackish and sea water, and a range of processing conditions. Forty one field
tests were conducted at the U. S. Navy's Fleet Anti-Air Warfare Training
Center at Dam Neck, Virginia. These varied from nominal runs with sand feed
rates of 30 tons per hour and oil concentrations of 0.5% to oil/water separations
at high capacity. Using the test results, a mobile,unit was designed, construct-
ed, field tested, and delivered to the Environmental Protection Agency. Data
was obtained on the effects on cleaning efficiency of relevant process parameters:
(1) sand feed rate, (2) feed steadiness, (3) oil type, (4) oil concentration,
(5) sand age, (6) feed homogeneity, (7) water rate, (8) water type, (9)
slurry density, (10) residence time, (11) aeration, (12) temperature, (13)
surfactant effects, (14) organic solids effects, and (15) oil deposition on wet
or dry sand. The mobile unit operated successfully under a wide range of con-
ditions . This device should prove a valuable adjunct to existing oil spill cleanup
procedures.
This report was submitted in fulfillment of Contract No. 14-12-809 between the
Office of Research and Monitoring of the Environmental Protection Agency and
Meloy Laboratories, Inc.
iii
-------
CONTENTS
Conclusions 1
n Recommendations 5
HI Restoration of Beaches Contaminated by Oil 7
3. 1 General Introduction 7
3.2 Laboratory Studies 12
3.3 Plant Construction 27
3.4 Demonstration Studies 41
3.5 Mobile Beach Cleaner 52
IV Acknowledgements 73
V Appendices 75
Appendix A - Laboratory Data 77
Appendix B - Construction Drawings 81
Appendix C - Plant Demonstration Data 105
-------
FIGURES
Page
1 Active Cell Volume vs. Weight Percent Solids 16
2 Weight Percent Passing vs. Particle Size for Three Sands 20
3 Particle Distribution for Dam Neck and Yellow River Sands 21
4 Demonstration Plant From Feed Hopper End 29
5 View Emphasizing the Process Water Tank and Flotation
Machine and Platform 30
6 Oil Recovery Tank and Cyclone with Tripod Support 31
7 Full View of Plant from the Discharge End 32
8 General Layout and Electrical Shed Section 82/83
9 Attrition Scrubber Foundation and Platform 84/85
10 Attrition Scrubber Foundation, Oil Recovery Tank, and
Vertical Pump 86/87
11 Flotation Machine Foundation and Platform, and
Horizontal Pump 88/89
12 Cyclone Tripod Support 90/91
13 General Layout for Electrical Power and Lighting 92/93
14 Electrical Details, Schedules, Legend and Line Diagram 94/95
15 Piping Isometric and Screen Spray Details 96/97
16 Launder, Screen and Oil Recovery Tank Details 98/99
17 Demonstration Plant Flowsheet 100/101
18 Attrition Scrubber and Additional Support Bracing 102/103
19 Mobile Beach Cleaner Schematic 66/67
20 Photographs of Mobile Beach Cleaner 68
vi
-------
TABLES
No. Page
1 Analytical Results for Series One Lab Tests 78
2 Analytical Results for Series Two Lab Tests 79
vii
-------
SECTION I
CONCLUSIONS
The primary conclusion from the entire project (as discussed in detail in the
main body of this report) was that froth flotation can be used to clean oil contam-
inated beach sand and that a mobile beach cleaner as supplied to the Environ-
mental Protection Agency can be used as a cleanup device during appropriate oil
spill emergencies. Other, more detailed, information was obtained in the
course of the project and is summarized briefly in the following section.
1.1 Demonstration Plant
Demonstration studies run with a stationary pilot plant facility at Dam
Neck, Virginia led to several conclusions; the first point made in the following
list is most critical to the overall project objective while the additional four
points represent supplementary information obtained during the course of the
study.
1) Oily sand cleaning using froth flotation is feasible but, process
water recycle and scrubbing or pumping of oil/sand/water slurries (if used) de-
grade the process significantly. Froth flotation should be utilized by feeding oily
sand and process water directly into the feed box of a froth flotation machine
with a minimum of prior agitation.
The system is limited mainly by the total flow of oil through it with a
maximum acceptable level of oil contamination at a 30 ton/hour sand feed rate of
about 1%.
2) Attrition scrubbing of straw with standard process equipment is not
feasible for the purpose of cleaning the straw for disposal or reuse.
3) Oil/water separation at high capacity is possible using froth flota-
tion. Considerable development work beyond standard mineral processing
techniques will be necessary to take advantage of this basic principle to obtain
high efficiencies.
4) In principle, oily sand scrubbing and dewatering is possible as a
cleaning technique; however, necessitites for a very large process water supply,
oil/water separation, and multiple scrubbing operations make the overall
approach infeasible.
1.2 Mobile Beach Cleaner
The second major phase of the project discussed in this report was the
construction, field testing, and delivery of a mobile beach sand cleaner to the
Environmental Protection Agency. The following conclusions pertain:
1) "Typical" operating conditions for the mobile beach cleaner are:
30 tonsAour sand feed rate
-------
No. 4 fuel oil as the contaminant at 0.5%
oil deposition on wet sand
oily sand aged for about one week
sand as found at the Dam Neck test site without the presence of
surfactants, dispersants or organic solids, homogeneously
mixed with the contaminating oil, and fed steadily to the
flotation machine
process sea water rate of 250 gallons/minute
process temperature of about 60°F
aeration rate of 250 cubic feet/minute
residence time (function of sand and water rates) of 4 minutes
160 ppm oil in the effluent water
75 ppm oil in the cleaned sand
130 ppm oil in the total effluent stream
Many factors affect the operation of the mobile beach cleaner, and some of
these can be used to adjust operation from the conditions above which are only
suggested as nominal.
2) The processing cost is about 77 cents per ton of sand. This com-
pares favorably to the estimated costs incurred in removing and disposing of
oily sand during spill incidents such as the one in San Francisco Bay during the
spring of 1971 even though actual spill experiences at 100 to 200 cents per ton
of oily sand do not consider replacement of the sand removed from a beach.
3) The Mobile Beach Sand Cleaner is indeed mobile and can be operated
with minimum logistical support and personnel.
4) The unit does away with the problem (especially in the long run) of
securing and maintaining permanent storage for oily sand.
5) The unit also assures that valuable sand (e.g., Hawaiian sand at
something like $20/ton) on recreational beaches is not depleted.
The following effects of operating conditions should be noted (these
effects are discussed in considerable detail in the bulk of this report);
1) Other factors remaining the same, increased sand feed rate results
in increased residual oil concentrations on almost a proportional basis.
2) Unsteady sand feed degrades performance.
-------
3) The heavier the oil, the easier the separation.
4) Increased feed oil concentration results in increased residual oil on
a proportional basis.
5) Within the limits of the studies, aging of oily sand improved the
separation process.
6) Wide variation in homogeneity of the feed sand degrades perform-
ance.
7) Increasing the water rate decreases residual oil concentrations but
also the overall process efficiency.
8) Sea water enhances separation.
9) Greatly increased slurry density appears to degrade performance.
10) Increasing residence time has an inversely proportional effect on
residual oil concentrations.
»
11) Increasing or decreasing aeration rate about an optimum results
in poorer separations.
12) Increased temperature enhances oil recovery.
13) Dispersants in the contaminating oil may either improve or degrade
the process.
14) Organic solids (e.g., straw) hinder the process of oil separation by
competing for attachment to air bubbles.
15) Oil deposition on wet sand (as opposed to dry) enhances oil recovery.
16) Water can be recycled from the oil recovery tank without signifi-
cantly degrading the sand cleaning process.
17) Settling ponds can be used to decrease the impact of residual oil.
18) Flotation reagents are not necessary in the use of froth flotation for
oily sand cleaning.
-------
SECTION H
RECOMMENDA TIONS
The following major recommendations relating to the Mobile Beach Sand Cleaner
result from the information gathered during the course of the Project:
1) The Mobile Beach Sand Cleaner should be used in cleaning up oil
contaminated beaches. For this, both the results of the demonstration projects
as presented in this report and the operating and maintenance manuals as
supplied with the mobile unit should be studied and followed closely.
2) Oil dispersants and sorbents should not be used on beaches where
the Mobile Beach Sand Cleaner is to be used.
3) Use of the mobile unit must take cognizance of the fact that cleaning
costs (under particular site conditions) greater than $5.00 a ton are in compe-
tition with simply removing the sand, disposing of it, and replacing it with
fresh sand.
4) Mobile Beach Sand Cleaners should eventually be strategically
placed around the country; Hawaii, West Coast, Gulf Coast, Great Lakes, and
East Coast emplacements would seem most reasonable.
5) Design of further units should carefully consider the results of the
studies as presented in this report and experience with the existing Mobile
Beach Sand Cleaner; larger processing capacity using standard equipment and
a conventional flatbed trailer may be possible.
6) Development of a more rugged water supply system should be pur-
sued in order to extend the range of environmental conditions under which the
Mobile Beach Sand Cleaner can be used.
Finally, before the inherent advantages of the froth flotation principle
can be practicably utilized for high capacity oil/water separations, significant
development work must still be done.
-------
SECTION HI
RESTORATION OF BEACHES CONTAMINATED BY OIL
3* I General Introduction
3.1.1 Oil Contaminated Beaches - The Problem
When oil pollutes a sandy beach, no single form of contamination takes place:
it depends on the type of oil, length of time at sea, temperature, time the oil
has been on the beach, and type of sand. Some oils, sufficiently long at sea,
will arrive at the beach as pebbles or streaks, and can be removed easily by a
beach cleaner. Other types of oil (particularly crudes) which have been at sea
for a long time are water-oil emulsions that are somewhat similar to butter,
and look like chocolate mousse. These emulsions, while on the beach, are
altered by environmental and biological impact; they become putty-like and
finally brittle. This type of pollution can also be cleaned up by a beach cleaner
or dry screening. Fresh crude (and many fuel oils) will penetrate the sand,
coating sand particles and filling some of the interstitial voids in the beach.
Beach materials vary greatly in the ease with which oil can wet them. Quartz
sand is difficult to wet with oil in the presence of water, while many shell
materials are more readily wetted. Consequently, when a beach is contaminat-
ed by oil a complex situation may exist where several forms of contamination
occur. Hence, a cleanup which does not consider the broad spectrum of con-
tamination will not be successful.
Experience with liquid oil falls indicates that the depth of penetration and posi-
tion of the contaminant is not easily ascertained from the surface. Uncontamin-
ated sands may bury the contaminated part, and the width and depth of the oil
contamination may vary markedly within short distances; thus, finding the
contaminated sand can be expensive. Modern practice has been to take large
swathes of the beach and, as spots of contamination remain, to either take a
second cut of sand or dig out the contaminated spots by hand. This results in a
large amount of sand in which relatively small sections are contaminated. Thus,
any cleanup procedure must either concentrate contaminated sands or be very
economical in the treatment of the contaminants.
A figure of $5.00 a ton for the replacement of sand on a beach is a reasonable
estimate. This price includes removal of contaminated sand, finding clean
sand, transportation, and addition of the sand to the beach. Any process used
in beach restoration must consider that cleaning costs greater than something
like $5.00 a ton are in competition with simply removing the sand, disposing of
it, and replacing it with fresh sand. In any treatment of the sand to remove the
oil, care must be taken not to trade oil pollution for other types of pollution.
Any burning techniques must have superior combustion and generate few harm-
ful gases or particles; likewise, harmful chemicals must not be allowed to
escape to contaminate the ground water or ocean. Froth flotation appears to be
an inexpensive method of cleaning oil-contaminated beach sand. It has the
advantages of concentrating oil contaminated sand, stripping oil from sand, and
working for a wide variety of oils and conditions.
-------
3.1.2 Froth Flotation - Laboratory Studies
Froth flotation is considered to be one of the most revolutionary developments
in industrial history because it is particularly suited to treating (at very low
cost) large quantities of low-value materials. In 1960 some 200 million tons of
ore were treated by froth flotation in over 200 plants in the U.S. alone.
Because froth flotation is so cheap and so effective in so many separation pro-
cesses, it has been used by many industries all over the world.
Large quantities of sand are cleaned by flotation in the United States. New
Jersey optical sand is cleaned by floating iron-bearing minerals from the bulk
of the siliceous sand; iron stains on the sand surface are removed by the
violent actions occurring in pumps, flotation cells, and cyclones; this sand,
after cleaning, is sold for $3. 00 a ton. In North Carolina, sands are float-
cleaned and scrubbed by pumps and cyclones; they are sold throughout the
country for use in golf traps; this sand is exceedingly white. Dark brown tar
sands in Canada are floated in hot water to remove the oil, using an otherwise
standard flowsheet; these sands come out very white. Flotation has often been
used to clean and to separate oil from sands.
Commercial flotation cells are produced by several manufacturers and are
available in capacities ranging from a few pounds per hour to several hundreds
of tons per hour. Many machines are self-aerating; i.e., air is drawn into the
cell by the action of the agitator (an impeller); with others an external air
supply is required. These units are not expensive, they are essentially self-
contained and they adapt readily to portable operation.
In the usual applications of this process to mineral industries, crushed minerals
are made hydrophobic (usually by chemical treatment), suspended in water, and
then violently agitated. When air is passed through the agitated suspension,
hydrophobic particles become attached to air bubbles and rise to the surface,
where they are collected in the froth. The hydrophilic particles, however,
remain in aqueous suspension. A reagent, such as pine oil, is added to help
form a stable froth.
Oil-soaked sand is an ideal material for cleaning by froth flotation because very
little chemical or physical pretreatment is called for. It does not need crush-
ing because it is naturally finely divided, but it is also relatively free of
"slimes. " The sand is naturally hydrophilic and the oil is naturally hydrophobic.
Many oils froth rather easily. The oil is less dense than either the water or the
sand, thereby facilitating flotation.
Extensive laboratory experiments at Meloy Laboratories indicated that froth
flotation with appropriate scrubbing permits the cleaning of a wide range of
sands contaminated with a wide variety of oils. "Oils", ranging from a very
light crude whose nature was much like gasoline to a baked solid fuel oil, were
successfully removed from mixtures with sands, ranging from Dam Neck beach
with 100% of its grains smaller than 841 microns to a yellow river sand with
almost 10% by weight larger than 1.68 mm. The oils were aged and imaged and
deposited on both wet and dry sand. In every case in which it was attempted, it
was relatively easy to select a combination of operating conditions under which
the cleaning process worked. In short, flotation seemed to have considerable
8
-------
promise for the cleaning of oil contaminated beach sands.
3.1.3 Froth Flotation - Plant Construction
Based on the laboratory experiments and on consultation with individuals and
literature familiar with the glass sand cleaning industry, Meloy Laboratories
proceeded with the design and construction of a beach cleaning demonstration
plant. Preliminary design of the plant involved considerations analogous to the
design of sand cleaning and froth flotation plants in the mining industry. Two
criteria were primary in the initial design: first, the plant was to operate at a
sand feed rate of about 30 tons/hour and be entirely self-contained, and, second,
the cleaning system was to be closed as completely as was practical so that no
extraneous oil contamination could result from the operation of the planttJuring
the demonstration studies. Of course, all the plant components had to .be re-
latively resistant to a marine environment, and provision had to be made for
running analyses of oil concentrations in both sand and water at the site.
A suitable site was found in the vicinity of Virginia Beach, Virginia; specifically,
it was on the U.S. Navy's Fleet Anti-Air Warfare Training Center at Dam Neck,
Virginia. Navy representatives reviewed the proposed project and conferred
with Meloy Laboratories' technical staff before leasing the site in early 1970.
Three problems immediately came to fore concerning the actual site which was
now to be used. First, the actual plant site was some distance from the
Atlantic Ocean; 550 feet to be exact. Second, the area to which the plant was
restricted (about 7,500 square feet) was about half that originally specified.
And, third, the available electric power was further from the site (750 feet)
than had been planned. These problems were eventually overcome by using
wellpoint water for most of the demonstration tests, rearranging the general
layout of the plant and drastically restricting the movement of support vehicles,
and spending more (over and above the original estimate) of the project funds
for a high tension power line.
To obtain fixed price bids on plant construction a detailed final design was
necessary. This final design was made by the Meloy Laboratories technical
staff with major and invaluable assistance from Mr. J. D. Glenn, consulting
engineer, and his engineering assistants in Norfolk, Virginia. The final design
was completed in May of 1970 and bids for plant construction were requested
immediately. The bids were received in June of 1970, and the Welch Contract-
ing Corporation of Norfolk, Virginia was selected on the basis of both the lowest
bid and the earliest guaranteed delivery time. Construction began in early July
and was essentially completed by mid-September of 1970. Meanwhile, the high
tension power line had also been constructed; Woodington Electric Inc., also of
Norfolk, Virginia, performed this task. All the above subcontractors performed
admirably under very short term conditions.
Meloy Laboratories engineering staff supplied the construction subcontractor
with the major items of process equipment. These included an attrition scrubber,
flotation machine and horizontal slurry pump from the Denver Equipment
Company, a vertical slurry pump from the Galigher Company, a belt feeder and
feed hopper from Link-Belt, Inc., a dewatering cyclone from Krebs Engineer-
ing, a process water pump from Worthington Corporation, and a process water
tank. A work-laboratory trailer was also placed at the site under the direction
9
-------
of Meloy Laboratories; this trailer was supplied by the U.S. Government.
Check out of the plant unit operations began in mid-September of 1970 and was
completed in less than the month originally projected. This was so even though
the project engineering staff had to integrate a leased wellpoint system into the
process and obtain appropriate sheaves for the horizontal slurry pump on very
short notice. The first actual plant demonstration took place on October 6,
1970.
3.1.4 Froth Flotation - Demonstration Studies
The demonstration studies took place under conditions comparable to those of a
medium size minerals processing plant. The major differences were due to the
comparative isolation of the site from the, sometimes, necessary support ser-
vices. These differences were, however, themselves a valuable education
since a mobile beach cleaning unit may very well have to be operated under
similar isolation. Five demonstrations were required by the original contract,
but by the time this report was written 41 had actually been performed.
Although directly aimed at demonstrating the efficacy of the froth flotation pro-
cess for cleaning oil contaminated beach sands, the project also provided a
variety of knowledge about other processing schemes and the field operation of
such a system. This knowledge ranged from the problems expected in operating
unmodified heavy equipment on a beach to the deleterious effect noticed for the
attrition scrubbing of a "normal" oil/sand mixture to the possibility of making
high capacity oil/water separations with a standard froth flotation machine.
The results of the tests on oily sand cleaning were briefly as follows. Under
nominal conditions with the closed loop process, sand was cleaned to an accept-
able level (less than 150 ppm oil in the water saturated cleaned sand), and the
change to a heavier oil improved the efficiency of the process only slightly.
Lowering the sand feed rate had significant positive effect on the process water
and little effect on the cleaned sand. The presence of straw demands much
more continuous attention to the belt feeder unit operation; there was a slight
indication that straw also promotes dispersion of oil into water. Finally, the
system is limited mainly by the total flow of oil through it with a maximum
acceptable level of oil contamination at a 30 tonAour sand feed rate of about
1%. The sand cleaning demonstrations are discussed in more detail in section
3.4.2.
An elevating scraper and front end loader were necessary for plant operation.
One of the first problems with the demonstrations was getting this equipment to
operate satisfactorily in loose beach sand; there was no problem on the beach
itself; the problems came when the heavy equipment had to move between the
plant and the beach across the loose sand on the back beach. Reliable trans-
port was finally achieved when pierced steel planking was laid in a single track
from the plant to the beach; some such provision should also be made for a
mobile unit since vehicles with balloon tires or tracks may not be available.
Other observations were made during the project relative to support equipment;
these are discussed in some detail in section 3.4 below
The field work did show that the major items of process equipment are very
dependable under even very severe weather conditions; this portends well for a
10
-------
mobile unit. The components which did give trouble were the pumps and de-
watering cyclone; since none of these items are envisioned for use in actual
emergency operations, problems with them are not particularly relevant. The
hopper and belt feeder, flotation machine, and leased submersible pump oper-
ated very dependably over long time periods; maintenance was also only of
minor concern. Since these three items represent (along with a suitable power
supply) a mobile beach cleaner, the results of the field tests were quite
heartening.
Oil/water separation using the froth flotation machine at the demonstration plant
with high flow rates proved more difficult than was originally anticipated. Such
processing was much more sensitive to the nature of the feed stream than when
used for cleaning oily sand. By implication, the tests indicated that consider-
able active flotation cell bypass occurs when a relatively low density oil is pro-
cessed. In the actual tests, oil recovery efficiencies ranged from about 50 to
80% for total flow rates up to 1000 gallons/minute and oil concentrations as
high as 5%. This performance indicates probable success for the basic separa-
tion principle in a system designed to utilize it fully while the success with
standard processing equipment has to be judged as marginal.
Besides the two basic types of tests mentioned above, straw cleaning by
attrition scrubbing and sand cleaning by scrubbing and dewatering were attempt-
ed. The former, due to both a lack of oil removal and extreme difficulty in
effecting the processing, was judged impractical. Although scrubbing and de-
watering of oily sand was demonstrated to be, in principle, possible by the test
results, the practical engineering problems associated with such an approach
(i.e., multiple scrubbing operations, oil/water separation as necessitated by
the dispersion due to scrubbing, the necessary large process water supply for a
30 ton/hour processing rate, and large oily water storage facilities) make it
infeasible for emergency field use.
3.1.5 Mobile Unit Field Demonstrations
Twenty four full scale demonstrations were run with the mobile beach cleaner in
the field. The results of these demonstrations are discussed in some detail in
section 3.5.4 of this report and tabulated in entirety in Appendix D. Briefly,
the field demonstrations substantiated the performance estimates for the mobile
beach cleaner and provided information for modifications of the unit. Some of
the difficulties inherent in remote, field operation of such a unit were more
clearly delineated by the field tests (e.g., problems with weather conditions on
a beach and with obtaining a reliable supply of water).
The general result of the field tests was that under nominal operating conditions
(see 3.5. 6) the cleaned sand and total effluent stream are expected to contain
approximately 75 and 130 ppm oil respectively; this effluent quality applies to
a sand feed rate of 30 tons/hour with an oil concentration of 0.5%. Other effects
which were studied related to such things as the deleterious effects of straw on
the process, the feasibility of water recycle from the oil recovery tank, the
feasibility of using settling ponds to isolate the process effluent stream, and the
ease with which the mobile unit could be moved to and from a location on the
beach. Indicators were also formulated for the dependence of the effluent stream
quality on the many process parameters; due to the large number of parameters
11
-------
involved, some of these indicators were quantitative and some qualitative in
nature.
3.2 Laboratory Studies
3.2.1 Introduction
The purpose of the laboratory studies in Phase I of this program was twofold.
First, flotation cleaning of oil contaminated beach sand was shown to be feas-
ible in the context of the original proposal (the feasibility studies); experiments
which simulated the "natural" contamination of beaches were also successful
(simulated beach conditions). Second, by attempting to clean severely contam-
inated samples of beach sand, limits were placed upon the probable, successful
operation of the proposed plant. These limits were found to confine a zone of
plant operation which is much broader than originally thought possible.
Specifically, the laboratory studies led to or assisted in leading to a number of
important conclusions:
(1) An attrition scrubber should be considered for the full scale
system.
(2) A belt feeder should be utilized to assure steady-state plant
operation.
(3) The pulp level in the flotation cells should be controlled.
(4) Irreversibly contaminated sand grains will be substantially
removed by the flotation process.
(5) Residual contamination due to pine oil frothing reagent
should be negligible.
(6) At the very least an oil-sand specific, quantitative tech-
nique for measuring sand contamination should be used in
evaluating demonstration tests; such a technique was developed
in Phase I and used in Phase n. A more general analytical
methodology was beyond the scope of the contract.
The following sections discuss the laboratory work which was performed.
Where pertinent, analyses and extensions of this work are also presented.
The laboratory work was intensely practical; such should be the case since the
flotation cleaning of oil contaminated beach sand is an intensely practical
problem.
3.2.2 Feasibility Studies
The feasibility studies in Phase I involved obtaining estimates of the importance
of five systems'.phenomena: (1) aging of crude oils, (2) materials1 recovery,
(3) residence times and pulp density, and (4) the effect of salt water on flotation.
The study of the aging of crude oils must embrace both long and short range
effects as well as whether the aging process occurs on the ocean or on a beach.
12
-------
The short range effects were found to be nil for the three crude oils investi-
gated; contrary to what one might expect, the long range effects appeared to be
most severe for heavy crude oils. Studies of aging were made for oil on sea
water and on beach sand. Inspection of the processing materials in the labora-
tory led to the conclusion that only trace amounts of oil are present in the
cleaned sand or in the sea water accompanying it. Similarly, the recovered oil
carries with it enough sand and water to be important from a handling or dis-
posal standpoint but not from the standpoint of the feasibility of the overall
process. Based on both the laboratory studies and a review of similar process-
ing units, the residence times (calculated) of the various unit operations in the
beach cleaning plant were estimated; in certain cases these turned out to be
different from the values estimated from the preliminary laboratory tests but
not to a degree which could have been considered detrimental to the overall pro-
gram objectives. Due to the results of the laboratory tests, pulp density
estimates were also revised; in all cases these revisions were favorable to the
process. Only actual operation of the plant will show whether such revisions
hold for a full scale system, also. The effect of salt (or sea) water on the
flotation process is of considerable importance in the treatment of ocean beach
sand: it means the difference between having or not having to provide fresh
water in an area where salt water abounds. Fortunately, salt water did not
prove to have an adverse effect on flotation and, in fact, eventually proved
favorable from the standpoints of both frothing and the actual oil separation.
The feasibility studies indicated that, at worst, the flotation cleaning of oil
contaminated beach sand would work as initially expected.
3-2.2.1 Aging of Crude Oils
Perhaps the most critical systems' parameter in any scheme for the physical
removal of oil from contaminated beach sand is the physical state of the oil
itself. This state is the result of aging processes on both the ocean's surface
and the beach. The most significant phenomena affecting the oil are evapora-
tion, oxidation, absorption of solids and gases, chemical and microbial
reactions, solubilization, and interaction with sea water. Of interest here is
the effect which aging has upon the flotation cleaning of sand rather than the
various phenomena which comprise the aging process; for this reason, oils
were aged by exposure to the ambient in a relatively uncontrolled manner. The
ease with which this oil may then be floated from a sand-oil-sea water mixture
indicated the effect of aging on the flotation process to the degree which is
necessary for process design and development.
Three crude oils were aged: light, South Louisiana mix; medium, mixed
sweet (high grade crudes); and heavy, Bachequero 17 (all supplied by the
Humble Oil and Refining Company). A 1500 ml sample of each oil was aged
for about four weeks in a perforated cylindrical container nested in a
rectangular plastic pan filled with sea water. In the main, only drastic
changes in the properties of the oil were looked for. Both the light and
the medium crudes (API gravities of about 45 and 35 respectively) showed
little in the way of significant changes; some tendency to form the so-
called "chocolate mousse" emulsions was noted. There were most probably
some changes in density, viscosity, surface character, and state of
emulsion for the two lighter crude oils, but these changes were definitely not
enough to provide a basis for judging the aged oils as similar to Number 6 fuel
13
-------
oil which was being used as a standard for the flotation cleaning up to this
point in time.
The Bachequero 17 or heavy crude oil (API gravity of about 19) showed marked
changes upon aging. In fact, a surface layer, due probably to evaportaion,
reactions, and the deposition of solids, was very apparent. This effect was of
considerable importance since the heavy crude oil was initially very similar to
a Number 6 fuel oil (even in asphaltene content which is estimated to be
about 15% by weight). It should be noted that this aged oil was essentially an
asphalt and that, therefore, the attempt to remove it from an oil-sand-sea
water mixture was an extreme test for the flotation process. Such a separation
cannot be effected with only a laboratory flotation cell because the energy input
per unit volume of pulp is too low. An ordinary kitchen blender allows a more
reasonable simulation of the conditioning process (scrubbing) in the laboratory.
Some such conditioning is necessary for the cleaning of sand contaminated with
asphaltic crude oils or for the flotation removal of the solid hydrocarbon resi-
dues which are sometimes created when oils age upon a beach.
3.2.2.2 Materials' Recovery
Materials' recovery affects the overall processing scheme most importantly in
the residual oil contamination of the process water and the cleaned sand. Since
crude oil is very slightly soluble in sea water (on the order of parts per million)
a certain amount of residual contamination is unavoidable. A similar consider-
ation holds for the adsorption of hydrocarbons on at least some constituents of
the solids matrix which comprises a sand grain. The sensory and perceptual
impact of such tract contamination is, however, negligible. It should be pointed
out that in some tests the processed sand appears to be cleaner than the sand
prior to contamination with oil; this is due to the removal of iron oxides by the
flotation process. The same effect is not observed for sand from the actual
beach site.
Consider that 10 pounds per minute of oil enters the plant (about 1% by weight
oil in sand), the amount of oil which is lost as residual contamination in the
processed sand and water can be calculated quite simply. The oil lost amounts
to 100 to 200 ppm. Since the water (from the dewatering cyclone in the process
plant) exits at about 65 gpm or 540 pounds per minute, it, therefore, carries with
it 0.054 to 0.108 pounds per minute of oil. The fraction of the oil which re-
mains as residual contamination can now be calculated, but it should be remembered
that the rate of oil carry-over is about the same for considerably larger levels of
initial oil contamination. The percentage recoveries which are calaulated in
the following are, therefore, conservative with respect to more severe contamin-
ation. Oil losses are obviously 0.54 to 1.08% by weight; it follows that oil is
recovered at the rate of 98.92 to 99.46% of the rate of oil input. These are
conservative, yet excellent, values when the large processing rates of the
efficiencies of other separation techniques are considered.
The rates at which sand and water are carried over with the recovered oil may
also be calculated. However, these rates are not nearly as critical as that for
residual oil contamination. Furthermore, the sand which is carried over often
should be since it is permanently contaminated with oil, and water carry-
14
-------
over depends greatly on the total oil recovery rate and the height of the oil froth
in the flotation cells. Anyway, the maximum contamination rates for the re-
covered oil were expected to be 300 pounds per hour of sand and 600 pounds per
hour of water based on an oil input of 600 pounds per hour.
3.2.2.3 Residence Times and Pulp Density
Although the laboratory flotation machine was found to operate optimally at a
pulp density (about 11% by weight solids; see Figure 1) like those often used in
the minerals dressing industries, the pulp densities proposed for the beach
plant do not stand up to the same comparison. A cleaning operation as that
envisioned here usually runs at 25 to 35 weight % solids. Fortunately, an
increase in pulp densities leads to increased residence times for the same
process equipment and the same solids processing rates. Based on the results
of the laboratory studies this is a route which was eventually pursued.
Long conditioning times in minerals processing allow for the action of various
reagents; conditioning of the sand (about 7.5 minutes) promotes other effects.
Combined with a high pulp density (about 75 weight % solids), scrubbing allows
the release of the oil from the surface of the sand particles and promotes oil-
in-water dispersion. Although of opposite relevance, these two effects are of
primary importance in the cleaning process-
Since the aeration rate is relatively fixed in a given flotation machine, an in-
creased pulp density in the flotation cells (and the concomitant increase in
residence time) means that the contaminating oil is exposed to possible contact
with air bubbles for a longer period of time. The result is, will be, and was a
considerable increase in the efficiency of the separation. The laboratory
studies have indicated that these increases are, indeed, beneficial. Therefore,
the design pulp density for flotation was increased to 25% by weight sand, and
the residence time in the flotation cells to about 4.4 minutes.
Changes in pulp densitites also had an effect on the overall processing plant.
Pump capacities were lowered, storage facilities made smaller (except for the
oil recovery tanks and the areas for sand piles), and control of the plant made
simpler.
3-2.2.4 Effect of Salt Water on Flotation
There was some reason for concern over the use of salt water in the flotation
cleaning of oil contaminated beach sand. Since surface phenomena are of
primary importance in flotation unit operations and since solutions with highly
ionic character often have peculiar surface properties, this effect was consider-
ed closely. Contact between covalent and ionic media (for particle sizes of
more than a few microns) leads to both more stable dispersions and enhanced
differences in surface properties. Fundamentally, this leads to a rationale for
the information of stable oil-sea water emulsions (chocolate mousse). As far
as flotation processing is concerned, dispersion stability can be bad and enhance-
ment of differences in surface properties can be good.
15
-------
3.25
3.00
2.75
oo
DC
LU
•s.
i
LU
U
<
o
LU-
LU
U
< 2.25
2.50
2.00
TOTAL CELL VOLUME UNDER FLOTATION
WITH NO SOLIDS (3.07 LITER)
DESIRED OPERATING RANGE
10
1
20 30 40
WEIGHT PERCENT SOLIDS
50
60
FIGURE 1
ACTIVE CELL VOLUME VS. WEIGHT PERCENT SOLIDS
16
-------
Qualitative experiments (i.e., using sea rather than fresh water) indicated that
the flotation process is effected only marginally if it is not helped. Separation
of oil droplets from sea water was observed to occur very rapidly (in approxi-
mately 15 seconds); on the other hand, more water was contained in the froth
column when sea water was used (about 10% rather than the previous 1 to 5),
Both jfaese observations are in line with the previous conclusions: the former
with the enhancement of surface properties, the latter with increased dispersion
stability. In principle, the rate of sand processing can be increased because
the separation occurs more rapidly; however, the recovered oil contains more
water. This can be somewhat adjusted by increasing the height of the oil froth
column so that more water drainage occurs.
In conclusion, salt water does not (on the laboratory scale) appear to have an
appreciable effect on the flotation process.
3.2.3 Simulated Beach Conditions
Since this phase of the laboratory studies covered less than half the time which
was spent on the feasibility studies, its scope was considerably more limited.
Four areas of interest were considered: (1) flotation testing different oil types,
(2) finding an oil to replace aged crudes during the demonstration, (3) simulat-
ing beach aging, and (4) testing different sand types. A large number of oils
were cleaned from sands by flotation; these included crude oils (light, medium,
and heavy as previously indicated), fuel oils (both blended and unblended; num-
bers 2, 4, 5, and 6), aged and imaged oils, medium crude oil blended with
selected organic chemicals (e.g., carbon disulfide and benzene), baked heavy
fuel oil, and oil which was heavily dosed with powdered graphite. Significant
flotation sand cleaning was found to occur with each of these substances;
furthermore, in every case where it was attempted a combination of scrubbing
and flotation conditions could be found which produced essentially complete
cleaning of the contaminated sand. The series of fuel oils, as well as blends
of same, were considered as replacements for aged crude during the demon-
stration runs. Since fuel oils figure in a significant portion of the reported oil
spills and because the heavy fuel oils are highly visible, No. 4 fuel oil was
selected as a suitable replacement. The simulated beach aging of various con-
taminating oils showed that prolonged contact with sea water enhances the
flotation cleaning process. Three sands were successfully tested; they varied
considerably in their content of both fine and coarse particles as well as in
appearance. All could be cleaned under appropriately selected conditions.
3.2.3.1 Testing of Different Qil Types
Probably the most obvious way in which oils can vary is in viscosity, In this
portion of the laboratory studies oils with viscosities ranging from that of gaso-
line, about 0.5 centipoise, to that 6f glycerol, better than 100 centipoise, were
successfully removed from sand mixtures. Since flotation is a process based
predominantly on differences in surface properties, the most difficult oils to
clean up were some of those which one would expect to have radically different
surface characteristics; viscosity and density, in fact, often play a relatively
minor role. For instance, a medium crude oil with a high concentration of
aromatic hydrocarbons was blended in the laboratory; it was quite difficult to
separate this oil from sand and salt water in the flotation cell. This should not
17
-------
be too surprising since aromatics have a considerable tendency to maintain con-
tact with both sand and water. Fortunately, the cleaning process can still be
adjusted to be relatively successful, and most aromatics will have vaporized
from actual oil contamination on a beach. On the other hand, a medium crude
oil whose sulfur content was upped to a known value of 15% by adding carbon
disulfide cleaned more easily than the original crude; this occurred in spite of
the fact that the blended crude oil was both more dense and more viscous than
the original crude. Of course, organic sulfides do have less of a tendency to
wet sand and water than do the alkanes and alkenes.
As far as the flotation process is concerned, it makes little difference whether
the hydrocarbon to be separated from a sand-sea water mixture is present as
an oil or as a solid residue. In either case, the primary concern is that the
hydrocarbon be dispersed as small enough particles to allow air bubbles to
float it out of the pulp. Even when the oil or hydrocarbon has a very high
affinity for silica particles (e.g., crude oil which, has been carbonized or liter-
ally baked onto the surface of the particles due to the action of sunlight) beach
sand can still be cleaned; in this case the permanently contaminated sand parti-
cles are floated with"ftreifcydrocarbon. This principle was, in fact, the basis
for some of the first minerals' flotation processes. Since cleaning under more
extreme conditions is possible, provision was made for this in the demonstra-
tion plant design; the attrition scrubber and two screening unit operations do
this. The attrition scrubber both conditions the flotation pulp and reduces most
of the agglomerates to single sand particles coated with hydrocarbons. The
froth flotation process is then able to remove from the cleanable sand both the
crude oil and the silica particles which are "irreversibly" coated with hydro-
carbon. Flotation functions as both an oil-sand separator and a sand sorter.
All the crude oils and fuel oils which were considered could be cleaned from
beach sand; this included whether they were aged or imaged, blended or un-
blended, or deposited on wet or dry sand. A generalization can be made from
the study of different oil types: the drier the sand the oil is found upon and the
more viscous the contaminating oil, the greater the need for scrubbing of some
sort in a successful flotation cleaning process. The extremes of both these
points are illustrated by the cleaning of carbonized oil (almost infinitely viscous
from baked sand (extremely dry).
3.2.3.2 Oil to Replace Aged Crudes
Initially it was thought that there might be some problem in selecting an oil
suitable as the contaminant for the demonstration tests. However, the success
of the process in the laboratory and familiarity with the nature of aged crude
oils eliminated this concern. In fact, other factors were now of primary
importance. For the purpose of demonstrations, high contaminant visibility
was desirable. The oil used was to be fairly standard in its properties so that
lots bought at different times could still be counted upon to behave similarly. It
was also desirable that the oil have some direct relevance to the problem of oil
spills and beach pollution. These points all indicated the use of a fuel oil. The
cost of these oils was not prohibitive, and they are very widely available. A
choice had to be made between numbers 4,5, and 6 fuel oils. The selection of
the fuel oil was finally made on the basis mat the Humble Oil Company in
Norfolk, Virginia was willing to supply No. 4 fuel oil in the quantities necessary
18
-------
for the demonstrations at no charge.
3.2.3.3 Simulated Beach Aging
The simulated beach aging, flotation tests were very successful. Both a heavy
crude oil and Number 5 fuel oil were aged on sea water for several weeks;
they were then deposited on samples of Dam Neck beach sand and aged for
about one more week. During this last week the oil-sand mixture was periodi-
cally wet with sea water to simulate tidal action. The sand was then cleaned by
flotation; essentially no scrubbing (other than that received in the flotation cells)
was necessary to obtain a very clean product sand. The same result occurred
for a medium crude oil which was aged only upon sand, not upon sea water. It
is obvious that when water is readily available for contact with the sand, the oil
may mix with the sand, thereby producing an unsightly mess, but there is little
or no real interaction between the sand grains and the oil.
3-2.3.4 Different Type Sands
Three types of sand were successfully cleaned: a yellow river sand, a commer-
cial white sand, and Dam Neck beach sand. These sands differed markedly in
their distributions of particles; Figures 2 and 3 on the following pages illustrate
this graphically. The yellow river sand was much the coarser of the three;
just a bit less than 10 weight percent was larger than 1.68 mm. The Dam Neck
sand was much finer with 100 percent smaller than 841 microns. The white
sand fell between these two in coarseness. All three were easily cleaned and
handled in the flotation process. The appearance of the sands was effected
differently, however. The Dam Neck sand was darkened by the cleaning pro-
cess ; this was shown by cleaning uncontaminated sand which still resulted in
the darkening effect. The yellow river sand looked cleaner after flotation; this
was because the process removed some of the iron which gave the sand its very
yellow color. The commercial white sand showed very little change in color
upon flotation cleaning; this was true whether or not it was contaminated. Coral
sand was also obtained in the latter stages of the laboratory studies. When con-
taminated with No. 4 fuel oil, this sand also cleaned very easily with only slight
darkening; this cursory analysis indicates that coral beaches will be restorable
using the froth flotation process.
3.2.4 Pine Oil Toxicity and Contamination
The book, Clinical Toxicology of Commercial Products by Gleason. Gosselin,
and Hodge, rates the toxicity of pine oil as the same as turpentine, pinene,
crude petroleum, petroleum distillates, kerosene, and fuel oil to name only afew
stubstances. These materials are all moderately toxic; for instance, the
probable lethal dose is 500 mg to 5 gm of the substance per kilogram of body
weight or about 1 ounce to 1 pound for a 150 pound man. However, the toxicity
of pine oil cannot be related on a one to one basis with these other materials
since it is, due to its aromatic and hydroxylated nature, significantly more
soluble in water (although still infinitely miscible with the various hydrocarbon
liquids). Since pine oil is a complex and variable mixture of organic compounds,
it is even more difficult to attach a value to either its toxicity or solubility in
other liquids. Its great preference for dissolution in No. 5 fuel oil over water
does assure that any residual contamination in effluent water streams due to
19
-------
HI
N
to
111
>
UJ
co
Q
111
<
o
CO
CO
LU
u
CK
X
o
UJ
100
60
30
10
0.6
0.3
0.1
YELLOW
RIVFR SAND
WHITE
COMMERCIAL SAND
DAM NECK
../•' BEACH SAND
A.
I
j
30 60 100 300 600 1000
SIEVE SIZE OPENING (MICRONS)
3000
FIGURE 2
WEIGHT PERCENT PASSING VS. PARTICLE SIZE FOR THREE SANDS
20
-------
YELLOW
RIVER SAND
100 65 48 35
TYLER MESH SIZE
105 149 210 297 420 595
SIEVE SIZE OPENING (MICRONS)
FIGURE 3
PARTICLE DISTRIBUTIONS FOR DAM NECK AND YELLOW RIVER SANDS
-------
pine oil will be small when compared to the high initial (and, relatively, high
final) concentrations of fuel oil .
Laboratory measurements indicate that pine oil is infinitely soluble in No, 5
fuel oil and soluble in water to the extent of about 1 part in 2000. It is reason-
able to assume that solutions of pine oil with both these solvents are essentially
ideal. Water solutions can be considered ideal since they must be quite dilute
while fuel oil -pine oil mixtures are ideal since both materials are very similar
tn physical and chemical properties. The fact that fuel oil and pine oil are not
homogeneous substances raises some doubt about treating them as if they were,
but, for the purposes here, this should be acceptable.
Consider equilibrium of oil (subscript A) and water (subscript B) solutions of
pine oil; the activities, a.« and a,,, must be equal. For dilute enough solutions
(less than about 1% by weight pine oil) these activities may be expressed as the
product of a constant activity coefficient, C. of C_, and the mass fraction of
A U
pine oil, x. or XB; we find, therefore, that
However, when the oil phase is completely pine oil, the water phase is satura-
ted so the ratio of activity coefficients is equal to XT, caturation* ^^ tne to*a
masses of both phases given by m. and m- and the total mass of pine oil in the
system given by m , a mass balance may be made. The mass balance and the
equilibrium expression give us two equations in two unknowns (XA and Xg):
XA = XB'XB, saturation
and
xAmA + xBmB = mp
The general solution to these two equations is a quadratic; however, when both
phases are quite dilute, the total masses may be set equal to the initial masses
of fuel oil and water, m . and m~ respectively. Solving for XB then gives
^ = mpXB, saturation^mA,o * mB,wXB, saturation* '
For the purpose of the beach cleaning plant we may express the masses in the
above equation in terms of mass flow rates. As mentioned previously, the
saturation value for pine oil in water is about 1 part in 2000 or a mass fraction
of 5xlO~ . The expected operating flow rates are about 300 gallons (2500
pounds) per minute of water, 10 pounds per minute of fuel oil, and 0. 0022
pounds per minute (or about 1 cc/minute) of pine oil. Substitution in the
equation for XB shows that the mass fraction of pine oil in the process water is
about 1x10 or 0. 1 parts per million. Considering that this calculation is a
conservative one and that pine oil is only moderately toxic, this residual con-
22
-------
centration in effluent water should be entirely acceptable if pine oil is, indeed,
needed as a frothing reagent.
3-2.5 Quantification of Qil-in-Sand Contamination
The general problem of determining quantitatively the contamination of sand by
oil was beyond the scope of the contract. Both sands and oils are too variable
to expect that there is any easy way to do this. What was needed, however, was
a quantitative technique which worked specifically for a given oil and Dam Neck
sand. A number of possible techniques were considered: photometrically
measuring the reflectance of visible light from beds of sand, spectrometrically
measuring the oil concentration in a solvent which has been used to extract oil
from sand, chemically determining the concentration of carbon in a solution
obtained by digesting a sample of sand with an appropriate chemical, measur-
ing the transmittance of light through an oil-sand-liquid mixture where the
liquid has a refractive index which is the same as the sand's, and using a gas
analyzer (methane or general hydrocarbon) to detect the "odor" of contamin-
ated sand. The second of these was finally settled upon as being most practical
and amenable to field test conditions.
Solvent extraction combined with spectrometric analysis does not have the same
limitations as visual and photometric evaluations. In principle, this technique
may be used to detect oils which are invisible to the human eye. Wavelengths
may also be sought which maximize absorption and, thereby, maximize detec-
tion sensitivity; unfortunately, for complex mixtures like oils no general
analytical scheme is available. Spectrometric analysis is much more suited for
detecting the components of a mixture than the mixture itself. However, in the
case of a known contaminant a suitable correlation can be developed. Much of
the discoloration due to oil pollutants can be related to suspended solids:
asphaltenes, carbenes, carboids, etc.; these solids are not removed by most
solvent extractions and, therefore, are not usually detected by this analysis
technique. For fixed oil, sand, and solvent this method should be successful;
it was, for this reason, used during the demonstration study.
Several solvents were tested for the analysis of sand and water samples.
Benzene was finally selected due to its efficacy with the fuel oils most common-
ly used in the demonstration plant (numbers 4 and 6). The first solvent used in
the extraction of oil from sand was heptane which worked well with light crude
medium crude oil and some No. 4 fuel oils. In preparing No. 5 fuel oil solu-
tions fine, dark particles appeared in the heptane and these did not dissolve
but settled to the bottom. Other solvents were tried (i. e., chloroform, carbon
tetrachloride and benzene). All three were suitable as solvents for the
heavier oils, but benzene was selected for its lower cost and toxicity.
Briefly, the analytical scheme went as follows. A sample of the oil to be used
as a contaminant was put into solution with benzene; known amounts of both
materials were used to make up these standard solutions. The concentrations
of oil in the benzene solutions were then correlated in the usual manner with
transmittance readings from a Spectronic 20 spectrophotometer from Fisher
Scientific Company. As expected, concentration varied linearly with the
logarithm of the transmittance at a fixed wavelength of incident light. Plots or
correlations of this sort had to be made up for each sand-oil combination con-
23
-------
sidered; since such correlations involve standard, classical analytical chemis-
try, no examples of plots are presented here. The wavelength of light was
generally around 450 millimicrons. Before unknowns could be confidently con-
sidered, the standard correlations had to be available; if they were not, the
best that could be done was to express contamination levels in terms of equiva-
lence to some known oil contamination.
When actual process samples were considered, sand samples were usually
dried before solvent extraction at about 65 - 70°C while water samples were
solvent extracted in a wet state overnight. Checks were run on the sand
analysis both in the laboratory and the field by extracting from both dry and
wet samples; the differences were small enough so that the convenience of
working with dry samples could be continued as standard practice (for an
example of this see the analytical results of field demonstration number 4 in
Appendix 12). Knowing the amount of sample (sand or water) analyzed, the
amount of benzene used, and the apparent concentration of oil in the benzene,
the oil contamination in the sample was quite simply back calculated. Since
parts per million oil could be detected in this manner, several interesting con-
clusions were reached in the field analyses. Generally, large variability of
samples was discovered under full scale, field test conditions^ this was not
unexpected. Background contamination in the sand at the field site was not
insignificant. The analytical results of demonstration number 9 indicated a
background contamination of about 20 ppm (expressed as No. 4 fuel oil) in the
Dam Neck beach sand prior to contamination.
Sampling and analysis for the stationary and mobile unit feed sand were identi-
cal. A sample bottle was half filled by physically moving the bottle against the
flow of the feed belt. Care was taken not to visually inspect the sand being
sampled. Portions of these samples from 2 to 10 gms were used for subse-
quent analysis. These samples were not dried. Samples were weighed on a
torsion balance to the nearest 10th of a gram and benzene added (10 to 100 ml
depending upon the oil content). The sample was agitated by stirring and
shaking until the oil was extracted from the sand. A benzene aliquot was placed
in round Bausch & Lomb cuvettes, and the transmittance measured at 450 mja
on the Spectronic 20. Transmittance was then compared to the calibration
curve for the particular oil used, and the oil content of the sand calculated.
Samples were obtained from the sand discharge streams by passing a sample
bottle through the discharge flow, filling the bottle from 1/2 to 3/4. The
sample bottle was passed vertically (from bottom to top) through the discharge
stream to sample all stream components (sand, water and oil) since they were
not completely mixed. Samples were weighed and shaken thoroughly, the
solids allowed to settle for a few seconds, and the liquid poured into another
bottle; the original bottle was then reiyeighed to obtiin the weight of the liquid.
Benzene was added to the liquid, and the oil was extracted by shaking until the
water no longer contained oil. If an emulsion formed, the sample was allowed
to set until enough benzene could be extracted for analysis. Analysis was as
above. The solid portion of the sample was dried and then thoroughly mixed
making sure that all visible oil was removed from the sides of the bottle. A
portion was weighed and extracted.with benzene, and the oil content determined
spectrophotometrically.
24
-------
Samples must be taken so as to be representative and unbiased. Samples taken
at the stationary plant or mobite unit were placed in 16 oz. screw cap bottles
obtained from Fisher Scientific Company of Silver Spring, Maryland. Caps for
the bottles were black plastic and vinyl coated liners. These caps deteriorated
after a few analyses so molded PVC caps should be used instead. Intervals be-
tween samples varied depending upon the length of the particular run. Ten
samples were desired for each process stream during a particular demonstra-
tion; this is in line with statistics since, for large standard deviations, a
standard deviation increase of about 10% results. Due to particular test condi-
tions, at times fewer than 10 samples were acquired. Two to 10 grams from
the well mixed total were analyzed for oil.
Note also that oils with the same classification occasionally differ greatly as to
their color, viscosity and density with light oils being especially so prone.
Calibration curves should be prepared for each shipment. Equal volumes of oil
can be drawn from each barrel and mixed before taking an aliquot for use in ob-
taining the calibration curve. To cite an extreme example, one shipment of No.
5 fuel oil had a viscosity of 175 centipoise and an API gravity of 22.5 compared
to another shipment with a viscosity of 48 centipoise and an API gravity of 25.3.
The latter shipment was nearer in both viscosity and density to a No. 4 fuel oil
than the usually accepted No. 5 fuel ofl.
Since analysis for oils was done colorimetrically, lighter oils were less sensi-
tive to analysis. When using marine diesel (light yellow), a gallon of No. 6 fuel
oil was mixed with each barrel to give it a color detectable with sufficient
sensitivity by the Spectronic 20. These analyses were run at 400 nyi to further
increase sensitivity.
When pouring the liquid off the sand in sample bottles, a film of oil sometimes
remained on the sides of the bottle. This would tend to cause the calculated oil-
in-sand concentrations to be higher than the actual concentrations, but, more
importantly, the results for total oil content remain accurate. Furthermore, as
was noted time and again at the demonstration site, oily water runoff from
contaminated sand does leave free oil or oily froth residue on the underlying
sand; the residue on the sides of the bottles can be thought an analog to this
residue. Perhaps more important is the definition of oil concentration in con-
taminated sand itself.. Obviously, the oil is not in the sand. At best it may be
on the surface of sand grains but usually is dispersed in the water within the
interstitial voids of the sand. Given that total oil content measurements for
process streams are accurate, a degree of arbitrariness remains in measuring
sand and water oil concentrations. Selection of a specific analytical technique
settles this arbitrariness as was done in the studies being discussed in this
report.
Towards the end of Phase I of the program, the analytical technique described
above was used in some limited laboratory studies aimed at the effect of vary-
ing froth flotation operating parameters. The results of these studies are
discussed in the following section.
3.2.6 Quantitative Laboratory Studies
The first series of tests in these studies sought to attach numbers to the effect
25
-------
of increasing turbulence in the laboratory flotation cell. The analytical results
for these tests are presented in Table 1 of Appendix A. Aeration rates were
varied in three steps from the minimum rate to the maximum for each of four
impeller speeds; the minimum rate was dictated by the lower limit o£ the rota-
meter used to measure aeration and was, therefore, the same for each impeller
speed while the maximum rate was itself a function of the impeller speed.
Average residual oil concentrations in the cleaned sand increased fairly regular-
ly with increasing impeller speed: averages of 107, 101, 190 and 300 parts per
million by oil by weight for 1000, 1200, 1800 and 2400 rpm impeller speeds
respectively. There is an implication in this averaged data that an optimum
exists with respect to impeller speed; "a priori, " this must be the case since
it zero impeller speed contaminated sand would settle to the bottom of the cell,
never contact air bubbles, and therefore, never be cleaned while at very high
impeller speeds any oil which rose to the top of the flotation cell would immed-
iately be remixed with the sand slurry due to the intense mixing action, and,
therefore, cleaning would once again not occur. The data is not, however,
precise enough to allow a determination of the optimum impeller speed on the
basis of only 16 laboratory tests. Such a result relative to precision is
classically found for froth flotation on any scale; comprehensive test work in
the minerals industry always involves large numbers of repetitive tests on
large numbers of samples.
The series one tests may also be used to gain a semi-quantitative grasp of the
effect of aeration rate on the process. Each impeller speed was tested at four
aeration rates which may be designated as low, medium low, medium high and
high; the averages over the four impeller speeds at each of these levels is 273,
183, 153 and 89 parts per million residual oil respectively. This result shows
clearly that increased aeration decreases residual oil contamination over the
range of conditions considered; this was not unexpected although high enough
aeration rates should eventually promote a degradation in residual oil concen-
trations .
The second series of laboratory tests looked for the effect, on a quantitative
basis, of three variables: magnitude of sand charge, initial oil concentration
and aeration rate. The results of these tests are presented in Table 2 of
Appendix A. The data for increasing sand charge is somewhat inconclusive
although a general trend towards increasing residual oil with increasing charge
can be seen. This makes sense for fixed feed concentration, impeller speed,
flotation time and (relatively) aeration rate; increased sand charge implies
increased total oil in the cell which, in turn, implies increased solubilized
and dispersed oil in the process water, some of which is always recovered with
the cleaned sand.
The twelve tests which considered three fixed aeration rates (see Table 2,
Appendix A) confirmed the semi-quantitative results of the series one tests.
The averages over the four initial oil concentrations were 109, 139 and 160
parts per million of residual oil for aeration rates of 11.8, 6.1 and 1.35 liters/
minute respectively. This is good quantitative evidence that increased aeration
decreases the amount of residual oil in the cleaned sand. Again, the data was
averaged over four tests at the same aeration rate; a more comprehensive test
program would have to involve repetitive testing under-identical conditions to
allow statistical evaluation of the analytical results. Fortunately, the test work
26
-------
discussed here was more cto verify the practicability of the technique used to
measure oil contamination levels rather than to study, in depth, me laboratory
scale cleaning of oil contaminated beach sand by froth flotation. Such an in
depth study could be a project in its own right, although its value relative to full
scale cleaning operations would be dubious.
Averaging over the three aeration rates (see the second part of Table 2) for the
four initial oil concentrations gives a quantitative indication of the effect of feed
oil concentration on the process: 76, 102, 149 and 218 parts per million resi-
dual oil for initial oil concentrations of 1, 3, 5 and 9% respectively. This
result substantiates quantitatively what had been observed qualitatively in pre-
vious tests: for "significant" feed oil concentrations there is apparently a
bottoming out of residual oil concentration; in other words, there is a minimum
(and finite) exit concentration of oil in the cleaned sand which cannot be
eliminated except by decreasing the feed concentration of the oil to a very low
level (this is, from an applications viewpoint, an impractical way to reduce the
residual oil level). This minimum oil concentration is (considering the data
and range of operating conditions presented above) about 60 ppm or 0. 006%}
feed oil concentrations would have to be reduced to about this level to effect any
further large decreases in residual contamination.
The analytical problems observed during the laboratory tests were also to be
of importance during the field demonstrations. Over and above this, the field
tests presented their own special difficulties: lack of, for instance, homo-
geneity of the feed sand, control over operating temperatures, truly representa-
tive sampling, and absolute control over all process materials. Such considera-
tions were important during the laboratory test program; it was mandatory that
tiiey be kept constantly in mind when the analytical results of the full scale test
program were considered.
3-3 Plant Construction
3.3.1 Introduction
This section covers the construction phase of the program which, including the
final design, took approximately four months to complete. Construction can be
expressed as the sum total of work in three categories: 1) preliminary design
and procurement of necessary unit processes and support equipment, 2) final
design and specification of plant structures and materials, and 3) subcontract
construction and unit process emplacement. The following subsections follow
this general outline.
3.3.2 Preliminary Design and Unit Process Procurement
3.3.2.1 Process Flowsheet
The final result of the flowsheet considerations may be found in Figure 17 in
Appendix B. Naturally, many similar flowsheets were generated over the
course of the project. The first such conceptual guideline was prepared by the
Meloy Laboratories' technical staff with assistance from a number of consult-
ants. Figure 17, since it is a final version, includes equipment call outs and
estimated power requirements; the first design flowsheet contained just the
27
-------
basis for calculations (30 tons of sand processed per hour), typical data,
suggested residence times, and general flow requirements for all process
materials through every unit process.
For the flowsheet presented here, sand enters the system through a feed hopper;
a moving belt then transports the sand from the hopper to an attrition scrubber.
The scrubber was added to the originally proposed system as a safety precaution;
it was to prevent any sand from becoming irreversibly contaminated during the
demonstration studies; laboratory studies had shown that, given enough pro-
cessing time, scrubbing could be used to clean essentially any oil-contaminated
sand. Sand leaves the attrition scrubber through a 1/4 inch screen and drops
into a sump. A vertical slurry pump then delivers the sand to the flotation
machine. Sand leaves the flotation machine in two streams; most of the sand
exits through the discharge port while some is carried over with the recovered
oil (indicated as about 10 pounds/minute on the flowsheet). The sand which
exits the discharge enters a sump from which it is pumped by a horizontal
slurry pump to a dewatering cyclone. A small amount of sand (mainly clay and
very fine silica particles) enters the cyclone overflow and is deposited in the
process water tank; the vast majority of the cleaned sand exits from the
cyclone underflow in close to a water saturated state.
The contaminating oil follows essentially the same path as the sand. Most of it
is discharged as froth from the top of the slurry or pulp in the flotation machine.
Of the oil which is not separated by flotation, most is delivered to the process
water tank while some exits with the cleaned sand. Under severe conditions
during certain demonstration tests, the visual impact of the buildup of oil in the
process water tank was quite startling due mainly to frothing of the oil upon
turbulent contact with water in the tank.
Make-up water enters the system either from a submersible pump when sea
water is used or a large, vacuum assisted centrifugal pump when wellpoint
water is used. Water is fed to the plant in general with a pump for that purpose.
The plant design allows for water addition at six points: as a wash at the feed
hopper, as addition water at the scrubber, as a screen wash and addition at the
1/4 inch screen and following sump, at the flotation machine as addition water,
as carrier for frothing reagent and as wash for the oil froth which exits via the
launder oveflow. Some water leaves the system in the flotation machine's over-
flow; this is a natural phenomena. Water which is used to wash the oil froth
into the recovery tank must be included in this stream for the purpose of
material balance. Water also exits in both streams from the dewatering cyclone;
most returns to the process water tank for reuse while some has to exit with
the underflow since this is a water saturated sand stream.
If reagent is used in the process, most of it will leave the system with the re-
covered oil. Some will, however, solubflize or disperse into water and leave
the system with same.
3.3.2.2 Preliminary Design
In essence, this flowsheet in its first form was delivered along with supporting
data to Mountain States Mineral Enterprises, Inc., of Tucson, Arizona.
Mountain States had won the subcontract, at $4,000, to assist Meloy Laboratories
28
-------
FIGURE 4. DEMONSTRATION PLANT FROM
FEED HOPPER END
29
-------
..
•
"
FIGURE 5. VIEW EMPHASIZING THE PROCESS
WATER TANK AND FLOTATION MACHINE AND PLATFORM
30
-------
FIGURE 6. OIL RECOVERY TANK AND CYCLONE
WITH TRIPOD SUPPORT
31
-------
FIGURE 7. FULL VIEW OF PLANT FROM THE
DISCHARGE END
32
-------
in preparing the preliminary design and conceptual specifications for the demon-
stration plant. This task was completed during Phase I of the project. The
general character of all foundations, supports and piping were specified during
tiie preliminary design, the results of which were submitted with the Interim
Report for this project. Although an attempt was made to get bids for plant
construction on the basis of the preliminary design, it failed. Potential con-
struction subcontractors could not quote a firm price without much more
detailed information; upon learning this, the Meloy Laboratories' technical
staff set out to provide such detailed information. This task comprised the
final design which is covered in a later section of this report.
3.3.2.3 Unit Processes
In general, competitive bids were procured for each item of process equipment.
It is beyond the scope of this report to go over the procurement process in de-
tail; therefore, each item of process equipment is considered only briefly with
a rationale for its selection, a short identification and specification summary,
and a description of its operation and relation to the rest of the plant.
The belt feeder and feed bin or hopper were ordered as light duty installations
since it was not expected that long continuous duty would be necessary or that
the plant would be in place for a long period of time. This item, including the
necessary supports and framing, was supplied by Link-Belt, a division of FMC
Corporation on the basis of both low cost and superior design; the cost was
$7,950. The hopper is 100 cubic feet in volume and is equipped with stiffeners,
columns and bracing, a regulating gate, and a 1 inch Grizzley screen. The belt
is 24 inches wide and operates around 31 foot centers; it is powered by a 2
horsepower drive unit at a speed of 23 feet/minute. The combined belHiopper
system is capable of delivering at least 40 tons per hour of damp beach sand
containing oil and salt water. The 2 horsepower, 1800 rpm drive motor is
woundfor 440 volt, 3-phase, 60 cycle electric current. The hopper is filled
with sand by a front end loader; since the belt operates at constant speed, the
sand feed rate is controlled by adjusting the regulatory gate. Sand leaves the
belt going directly into the feed box of the attrition scrubber.
The attrition scrubber was supplied by the Denver Equipment Company, It is a
4-cell machine with a total volume of 100 cubic feet or 25 cubic feet per cell.
Since the scrubber is a standard piece of minerals processing equipment, it is
of heavy welded steel construction and capable of operating under extended con-
tinuous duty. Agitation is supplied to each cell by a shaft with two opposed
pitch propellers; each shaft is driven by a 15 horsepower, 900 rpm, 3-phase,
60 cycle, 440 voltw high starting torque motor. The interior of the machine is
Neoprene lined for protection from both abrasion and oil. The attrition scrubber
cost $9,940. As was mentioned earlier, the scrubber was added to the process
to assure that all the sand contaminated at the site could be cleaned even under
the most unfavorable conditions imaginable; it also permits study of the effect
of scrubbing in a manner independent of the froth flotation process. Sand enters
tiiis machine directly from the feed belt; the sand is mixed with water at the
feed box to form a slurry with at most 80% by weight solids. The slurry exits
the scrubber onto a 1/4 inch screen from which it flows into the vertical pump's
sump.
vertical pump was supplied by The Galigher Company and was selected on
33
-------
the basis of the lowest bid under the required specifications. The pump cost
$1,141 amd was rated at 194 gallons/minute of water-sand slurry (specific
gravity 1.38) against a 34.3 foot dynamic head with an impeller speed of 1430
rpm. The pump includes a variable speed V-belt drive and a 15 horsepower,
1800 rpm, 230-460 volt, 3-phase, 60 cycle drive motor. This pump delivers
the sand-water slurry from the scrubber to the froth flotation machine.
The flotation machine was supplied by the Denver Equipment Company and was
selected on the basis of low bid as well as for more than meeting the process
design specifications. Total cost, including a level controller and compressor
as well as an air supercharger, was $10,065. The flotation machine was a
6-cell, No. 21 machine with a total volume of 240 cubic feet or 40 cubic feet
per cell. The DR shaft assemblies are complete with disc impellers and-
diffusers for the supercharged air. The bottom of the tank is lined with
Neoprene for abrasion and oil resistance. The machine^is equipped with a
built in air header, air valves and flexible connections from the header to the
diffuser mechanisms. The impellers are driven by dual cell V-belt drives
powered by three 10 horsepower, 1800 rpm, 3-phase, 60 cycle, 440 volt motors.
Froth paddles and 1/4 horsepower drive motor are also included. Level con-
trol consisted of a Denver Auto-Flo.t Level Control system containing a controll-
er, resistance probe, air regulator, dart valve; the control system operates
with 60 cycle, single phase, 120 volt electrical power and 50 to 100 psi air.
The air for the control system is supplied by an Oil-Less air compressor (74
psi maximum air pressure) with a ten gallon receiver, pressure gauge, cut-off
controller and 1/4 horsepower, single phase, 60 cycle, 110 volt motor. Air is
supplied to the flotation machine by a Denver Supercharger rated at 280 cubic
feet/minute against a 16 ounce discharge pressure; the supercharger is power-
ed by a 3 horsepower, 3500 rpm, 3-phase, 60 cycle, 440 volt motor; a 6 inch
blast gate and flexaust air duct interface the supercharger to the flotation
machine. Sand slurry is delivered to the machine by a vertical pump and
diluted to the desired pulp density at the feed box. Since this is a right-hand
machine, the concentrated oil froth overflows into a launder on the right when
the machine is viewed from the feed end; each cell is equipped with slots and
weir bars for individual control of the froth overflow. Since most of the re-
covered oil is floated in the first cells of the machine, the weirs for these cells
are usually adjusted to provide most of the machine's overflow; to make this
adjustment, the froth paddles must be turned off. The froth paddles assist in
removing the froth from the'top of the pulp in the machine; an additional
assist from wash water is rarely needed. Processed sand slurry leaves the
flotation machine through either an overflow weir or an underflow discharge
port. Experience with the machine showed that the overflow discharge caused
pulse-like outputs of small amounts of oil froth in the product slurry; this is
due to the slow accumulation of oil froth at the wall of the last cell adjacent to
the discharge box. Periodically some of this accumulated froth breaks loose,
enters the discharge box and passes over the weir; however, if the weir over-
flow is not used, "tramp" froth just collects on the surface of the pulp
-------
being least cost. This 5" x 5" Frame Three SRL pump has Neoprene runner and
casing liners, a water gland, variable speed V-belt drive, and a 15 horsepower,
3-phase, 220/440 volt, 1800 rpm motor. The pump is rated at 405 gallons/
minute for a 25% by weight sand slurry against a 54 foot total dynamic head.
This pump delivers the processed sand slurry to the dewatering cyclone. Al-
though equipped with a variable speed drive, the lowest possible capacity was
often still too high; for this sort of flow situation, additional dilution water is
supplied to the horizontal pump's sump.
The dewatering cyclone was supplied by Krebs Engineers for $1,529. This
D15-B cyclone is of steel construction with molded Neoprene liners, a nickel
hardened vortex finder, and an adjustable apex valve. Included with the cyclone
was a No. 160 Pine Operator for hydraulic control over the apex valve. Al-
though the dewatering cyclone is rated as being able to process 30 tonsAour of
beach sand, the adjustable apex prevented this for the very monosized beach
sand found at the Dam Neck site. When operated with the adjustable apex, the
cyclone sent a large portion of the sand to the system's process water tank.
This problem was resolved by removing the apex and the Pine Operator; after
this was done, the cyclone operated in an acceptable, although somewhat wet,
fashion. Cleaned sand which is discharged in the cyclone underflow is returned
to the beach; the process water which exits via the overflow of the cyclone is
returned to the process water tank.
While tiie oil recovery tank was of wooden construction and supplied by the con-
struction subcontractor, the process water tank was a standard, 24 foot dia-
meter backyard swimming pool; these are widely available and suitable, in a
restrictive sense, for systems which demand portability. This tank was includ-
ed in the plant for study as to how it stood up under relatively rough use. The
only problem comes in trying to remove large amounts of sand from the tank;
great care has to be taken to avoid puncturing the liner. However, during the
whole series of demonstrations in which many tons of sand were removed from
the tank by shoveling, pumping and flushing, the tank was punctured only once.
Such a tank should be useful as an oil recovery tank in a mobile system.
3.3.2.4 Support Equipment
Three pieces of support equipment were required for the demonstration project;
an elevating scraper for picking up and returning sand to the beach, a front end
loader to dump sand into the feed hopper, and a 4-wheel drive truck for general
plant support. In an actual emergency field operation a scraper would be
necessary for about one third of the total plant operating time; considering
shipping charges, day-by-day rental charges, and possible waiting periods for
acquisition, however, it was bom cheaper and more convenient to have the
scraper available on a full time basis for the demonstration project.
The elevating scraper was supplied by the Furnival Machinery Company at
$2,100 per month; it was selected on the basis of a low bid and at the suggestion
of heavy equipment suppliers for the intended task. It is a WABCO D111A
scraper with a capacity of approximately 11 cubic yards or 15 tons of beach sand.
Although one of the smallest elevating scrapers manufactured, this machine is
more than adequate for supplying sand to a 30 tonAour plant; in approximately
20 minutes it can supply and return sand for an hour of plant operation. A
35
-------
larger scraper (say 30 cubic yards) could obviously spend much less time
supporting the plant. Under ideal conditions this scraper could make an effici-
ent 2 to 3 inch cut on the beach. However, usually the Dam Neck beach has
undulations in it which significantly cut down on this efficiency; under "normal"
surf conditions a best cut of 4 to 5 inches is quite acceptable. After severe
storms, a 6 inch cut was often all that was possible. Although the scraper
moves quite easily within the tidal zone on the beach, it has considerable trouble
in the loose sand above the high water mark. Using matting or pierced steel
planking overcomes this problem while restricting the mobility of the heavy
equipment; a mobile system must take this into account.
The front end loader was also supplied by the Furnival Machinery Company at
$1,000 per month; it was selected on a least cost basis. It is a WABCO 150B
front end, loader and was by far the most reliable piece of support equipment at
the demonstration site. This loader has a 1-1/2 cubic yard bucket and is able
to feed the plant at its maximum rate of about 60 tonsAour; at such a high rate
it takes two men to conveniently keep up with the operation. A front end loader
is mandatory for general plant and site housekeeping. It is also very useful
when the last traces of oil contaminated sand are picked up from the beach;
that is, the front end loader in conjunction with a shovel crew can be used to
gather up the oily sand which is necessarily missed in a scraping operation.
Like the scraper, even a 4-wheel drive front end loader with standard tires
needs a bed of pierced steel planking to move efficiently in the loose sand above
the high water mark.
Although a 4-wheel drive truck was very useful during the demonstration pro-
ject, this would probably not be the case during an actual emergency field
operation. Heavy equipment (loaders, scrapers, graders, etc.) should be
sufficient for transport to and from the beach proper and the nearest road;
standard road transportation would, of course, still be necessary. The truck
used at the Dam Neck site is a 3/4 ton, 4-wheel drive, Ford pickup leased from
Ted Britt Ford at $475 per month. It has been used for such varied tasks as
delivering wellpoint sand and fuel oil in 55 gallon drums. In a field project such
a vehicle is invaluable for general plant and project maintenance.
Early in the plant chek out it became obvious that it would not be possible to
reliably obtain water from the ocean1 s surf over a distance of 550 feet; first,
pumping over this distance required more power than the available pumps had,
and second, even with an adequate pump in the surf, the field engineering staff
would not have it under their direct control; the latter could easily result in
losing the pump to the ocean. A wellpoint system was, therefore, installed to
obtain water for general use. This system was supplied by the Moretrench
Corporation at a cost which eventually leveled out to $262 per month. The
water which it supplies is brackish and contains a large amount of unoxidized
iron. In the course of the demonstration tests it was found that this water is
deleterious to the froth flotation process relative to sea water. Whenever
possible, sea water should be used in actual emergency field operations.
Besides the structures, electrical wiring and piping which were .assembled at
the site, an office-laboratory trailer and a tank for diesel fuel were required.
The M-20 Coastal Trailer was supplied by the U.S. Government while the diesel
fuel tank was supplied by the Hampton Roads Oil Supply Company at no cost
36
-------
other than the purchase of diesel fuel.
The rest of the demonstration plant (other than incidental materials and supplies)
was supplied by construction subcontractors; their work is reviewed briefly in
Section 3.3.4 while the following section discusses the final design upon which
the construction work was based.
3-3'3 Final Design and Specifications
Appendix B contains the 11 engineering drawings which resulted from the final
design work. As was mentioned earlier, the final structural design was com-
pleted by Joe D. Glenn, Jr. and Associates, Consulting Engineers. Figures 8
through 12 represent the input of Mr. Glenn to the final design. Figures 13 and
14 present the essence of the electrical specifications for the plant; these draw-
ings were prepared by Mr. Robert L. Payne, an electrical engineer with
Melpar, Inc., and are based upon the recommendations of the Meloy Laborator-
ies' technical staff, the equipment vendors, and Mr. Joe D. Glenn. Figures
15 and 16 were prepared by the project engineers and are based on revisions of
the preliminary design work done by Mountain States Mineral Enterprises, Inc.
Figure 17, the plant flowsheet in final form, was prepared by the Meloy Labora-
tories' engineering staff and is a result of the previous design and consultation
work; the staff also prepared the drawing presented in Figure 18 for additional
support for the attrition scrubber.
The following sections consider each of the drawings in Appendix B in more
detail as a means of explaining the plant design. This then provides a base for
Section 3.3.4 on the plant construction.
3«3.3.1 General Layout and Electrical Shed
First, note that all the drawings have been halved in size for the purpose of this
report; therefore, for the General Layout of the Pilot Plant in Figure 8 the
scale is actually 1/16" = 1' rather than the 1/8" = 1' indicated on the figure.
The trapezoidal area for the plant as represented in Figure 8 is approximately
8000 square feet; this is considerably less than the 15,000 square feet origin-
ally sought. The plant had to be arranged in a "U" shape to fit the available
area which was bounded by a sand road (at the bottom of the figure), a hill and
a target drone launching complex (at the top), and grassy dunes (on both sides);
it was imperative that none of these restricting influences be disturbed. Con-
taminated sand for feeding the plant is placed in the area bounded by the ramp,
scrubber, oil recovery tank, cyclone tripod, and road. Access to the back of
the plant with the truck and heavy equipment was obviously hindered by the
shape of the site. The shed pictured in section in Figure 8 is mainly to protect
the plant's control panel from the elements. It also provides a small storage
area for tools and supplies; due to the sandy nature of the site, wind and rain
protection did not demand a floor for the shed. During construction the shed
was modified to have doors on both ends. Figure 8 also contains general notes
on plant construction and specifications as well as references to other drawings.
3-3.3.2 Attrition Scrubber Foundation and Platform
As indicated in Figure 9, the service and viewing platform for the attrition
37
-------
scrubber ie of wood construction; this platform also allows easy access for
sampling during demonstration tests. The foundation for the attrition scrubber
is also indicated and is just a concrete pad. The foundation details for the belt
feeder are also included in Figure 9; the relative location of these anchor bolts
was supplied by the Link-Belt engineering staff and proved to be correct when
the hopper and belt feeder were finally put into place.
3.3.3.3 Scrubber Foundations and Recovery Tank
Figure 10 shows the lattice arrangement which was used as a structure to
support the attrition scrubber. 10" x 10" beams were spiked together to form
the base upon which the scrubber sets; this base, in turn, sets upon a concrete
pad. Unfortunately, this approach, although simple in concept and seemingly
reliable, produced an unstable support; the scrubber wobbled on the lattice
and would have walked off it if additional support had not been provided. Figure
18 shows the additional support; an extra concrete pad was formed, and steel
"Lrs" were anchored to it and welded to the side of the attrition scrubber. This
steel frame and pad kept the scrubber from moving around on the lattice.
Figure 10 also includes details of the oil recovery tank. It is of wood construc-
tion with steel overflow and drain pipe. Heavy duty posts were used to brace
the sides of the tank not so much to support the weight of the recovered oil as to
prevent the tank from being collapsed by external factors (the front end loader,
for example). The liner of the tank was supplied by the construction subcon-
tractor and is formed of a polymeric vinyl film, 30 mils in thickness; this sort
of thickness is, in general, desirable so that the tank can withstand some
accidental abuse without rupturing. Details for the vertical pump's sump are
also included in Figure 10; this sump is of concrete construction, and its
bottom is approximately 4 feet below finish grade. The pump is supported by
two 8" x 8" timbers which are anchored to the top edges of the sump. The
screen and chute which follow the scrubber are pictured in the same drawing.
3.3.3.4 Flotation Machine Foundation and Platform
The platform for the flotation machine is supported by 4 concrete pads and is
comprised of three levels (see Figure 11). The first level is finish grade. The
second level supports the flotation machine; in this fashion, the froth flotation
machine is as separate from the rest of the-system as possible. Although de-
pendent upon the operation of the vertical and horizontal pumps, it is not hydro -
dynamically connected to them. For instance, the horizontal pump's sump can
partially sand-in and overflow without transmitting an upstream effect to the
flotation machine; the overflow from the sump is just carried by a chute into
the oil recovery tank. Similarly, the flow into the machine is positive for both
the feed sand slurry from the vertical sump pump and the dilution water from
the process water pump. Air is supplied to the flotation machine by the
pressure blower which is located on the second level under the third level plat-
form. The third level (a small platform) permits feed of reagents or contamin-
ants to the flotation machine by gravity; during selected tests, for example,
pine oil was fed to the feed box of the flotation machine to assess its ability to
stabilize oily froth. The roof on the flotation platform served to provide some
protection from inclement weather. The entire platform was designed to with-
stand gale force wirids; more severe winds might tear off the roof of the plat-
form but the rest of the structure should remain intact. Regarding this
38
-------
structure in particular, note that the eye of hurricane Ginger hit the East Coast
120 miles south of the demonstration plant site in September 1972; the flotation
machine platform and .the rest of the plant withstood the storm in excellent
condition.
3-3.3.5 De watering Cyclone Tripod Support
The dewatering support frame was constructed substantially of iron pipe. A 21
foot high tripod was selected for this support to allow access to the dewatered
sand from any direction (see Figure 12). Since dewatered sand drains rapidly
to an even drier state, pools were expected to form around the cleaned sand
pile; if access to the sand was limited to a particular direction, these pools
could prevent timely removal and return of cleaned sand to a beach by prevent-
ing the operation of a front end loader or elevating scraper in the area of the
cleaned sand. The tripod support performed admirably during the demonstra-
tion tests and stood up very well under the severe environmental conditions
sometimes encountered at the test site.
3-3.3.6 Electrical Services
Figures 13 and 14 are schematics of the electrical wiring for the demonstration
plant. The first figure shows the general layout for the entire plant from the
main transformers to the various stations. Note that both 110 volt, single
phase and 440 volt, three phase power was required for the plant. 110 volt
electricity served for plant lighting and general support services while the 440
volt power served to power the major plant equipment. In general, the electri-
cal wiring necessary for the demonstration plant was of a very simple nature.
Figure 14 is a schematic of the plant's main control panel; other than the main
disconnect and disconnects and starters for each motor in the plant, there was
a spare disconnect, a welding receptable, and a 15 KVA, 480 volt primary,
240/120 volt secondary transformer with corresponding junction control panel.
3.3.3.7 Piping Run and Detail Callouts
Figure 15 includes an isometric of the piping runs for the demonstration plant;
pipe hangers and supports are specified on other of the construction drawings.
Other than the obvious runs for the process streams (water and oil-sand-water
slurry), extensive wash water lines were provided; additionally, seal, reagent
dilution, and spray water piping was provided.
Figure 15 also includes details for the 1/4 inch stationary screen which was
placed after the attrition scrubber and before the sump for tire vertical pump;
two rows of spray nozzles were included to assist in moving wet sand through
the screen. In actual practice the spray nozzles proved unsuitable and were
simply replaced with 1/4 inch holes in the feed pipe for the original nozzles;
the nozzles continually plugged up due to trash material in the water (sea or
wellpoint) at the site. In its final configuration, the 1/4 inch stationary screen
performed well for all input slurry conditions (including up to 80%solids). The
trash material also caused plugging of the lines for wash and seal water; for-
tunately, the wash water lines were only rarely used and the seal water line to
the horizontal slurry pump could be quite easily cleared.
39
-------
Figure 16 gives additional detail for the launder, oil recovery tank, and station-
ary 1/4 inch screen. As noted, the launder was of simple wood construction;
although recovered oil discharged through the launder, the wooden seams still
swelled up enough after a short period of operation to stop any leaks which could
have produced a continuing mess around the flotation machine platform. The
1/4 inch stationary screen was held in place by wedges to permit easy replace-
ment and servicing; however, even after long term contact with salt water,
this screen only had to be repaired once. The details for the oil recovery tank
include a water drain and overflow trough. Although the water drain functioned
properly in its original configuration, it proved subject to burying and plugging
by me sand at the site and eventually, while buried, was put out of commission
by a front end loader working around the dewatering cyclone. Although the
overflow trough worked, its capacity proved insufficient as compared to either
pumping or dipping oil out of the tank. This was due mainly to the extremely
viscous nature of the oil which, in turn, was due to the presence of a variety of
solid particles, air bubbles, and water droplets in the recovered oil. A skim-
ming slurry pump is, therefore, suggested for removal of oil from such a tank.
3.3.3.8 Final Flowsheet
Figure 19 is the process flowsheet in its final form. Except for the wash water
at the feed hopper and flotation machine launder, the indicated flows are sub-
stantially those utilized under nominal operating conditions. This flowsheet is,
in fact, tepresentative of the one used in finalizing the process unit operations,
the structural design for the plant, and the electrical power requirements and
design. Besides typical flow rates for water, sand, and oil, the flowsheet calls
out major process equipment with vendors' names and relevant equipment
characteristics and specifications. The flowsheet (FS) numbers are for refer-
ence purposes (e.g., see Figure 13 and 14 relating to the plant1 s electrical
services). Typical materials data, estimated power requirements, and esti-
mated residence times are included on the flowsheet. The residence times are
based on input flow rates; this means that the values for the oil recovery tank
and the process water tank are nominal times to fill and empty the tanks
respectively.
3.3.3.9 Additional Support for the Scrubber Platform
In terms of the plant structures, only one major modification was necessary
after construction was essentially completed. The lattice support platform for
the attrition scrubber (constructed of 10 inch timbers) did not furnish stable
enough support; the scrubber tended to wobble and "walk" on this platform due
to an inherent lack of level in the platform. To stop this tendency, side bracing
was welded to the scrubber and anchored to a concrete slab formed at its side.
The result was a very stable structure, still at a relatively low cost as com-
pared to a more elaborate platform (e.g., the platform which housed the froth
flotation machine). Such a simple structure has much to recommend it for use
in demonstration projects in the field,
3.3.4 Conclusions
As is the case with most projects, the result of the construction project repre-
sented the inputs of a large number of people. Since, with the time allotted, the
40
-------
entire plant could not be specified in complete detail, considerable creative
thought had to be applied to tasks at hand during the actual construction work.
This, in turn, necessitated very close cooperation and communication between
all participants in the construction project.
3.4 Demonstration Studies
3-4. 1 Introduction
Forty one tests were run using the demonstration pilot plant. The results of
and operating conditions for these tests are presented in detail in Appendix C.
The tests using the stationary plant can be broken . into five main categories :
(1) Sand Cleaning Using the Entire Plant (16 tests), (2) Mobile Unit Simula-
tion (7 tests), (3) Straw Scrubbing (3 tests), (4) Oil/Water Separation (11
tests), and (5) Sand Scrubbing and Dewatering (4 tests).
The sand cleaning tests using the entire pilot plant were run to assess the basic
feasibility of froth flotation for recovering oil contaminated beach sand. By con
sidering the effects of oil type and condition, operating parameters, scrubbing
and dewatering, and recycle of process water, probable configurations and
specifications for a mobile sand cleaning unit could be made.
Bypassing the attrition scrubber and eliminating the dewatering and water
cycle operations, allowed the probable mobile unit process configuration to be
simulated. The resulting mobile unit simulation tests served both to confirm
the design estimates of the specifications for a mobile unit and to firm up the
process configuration for such a unit. These tests were deemed necessary to
avoid building a complete mobile system on the basis of (perhaps, misleading)
extrapolation of data obtained using the entire stationary plant.
Since the equipment was available at the site and the necessary tests were quite
simple, the feasibility of attrition scrubbing oil from straw was assessed. If
this were feasible, straw could be either reused or disposed of much more
easily.
Based on the success of the sand cleaning tests, 11 oil/water separations were
run at full scale using the process plant in essentially an "as is" configuration.
Although froth flotation is used commercially to separate oil and water, such
separations are usually with low flow rates (very high residence times) and for
low feed oil concentrations. The tests with the stationary plant were designed
to look at conditions just the opposite to these: high flow rates (low residence
times) and high feed oil concentrations. The intention was to determine whether
the stationary plant equipment could be used in essentially an unmodified state
for rapid oil/water separation. Evidence to the contrary might possibly gifoe
leads as to what modifications would give acceptable performance or what
critical factors must be considered in designing new equipment to do the job
using the same basic principles.
Scrubbing and dewatering of oily sand was tested because the sand cleaning
demonstrations using the entire plant had shown that oil was quite easily
scrubbed from beach sand (and thereby dispersed into water") with an attrition
scrubber. Perhaps oily sand could be scrubbed and put in an acceptable state
41
-------
for discharge on a beach by successive dewaterings in conjunction with a large
settling pond(s).
The following sections review the results of the tests in the five categories as
outlined above. The detailed data for and discussion of the tests are presented
in Appendix C. Only the information judged as most crucial in gaining value
from the test results is presented in the following paragraphs.
3.4.2 Sand Cleaning Using the Entire Plant
Demonstrations 1-13, 21, 22, and 41 in Appendix C involved the use of all the
unit processes in the plant. These demonstrations were used to gain first
estimates of the projected operation of a mobile beach cleaner as well as an
idea as to what the typical operating difficulties with such a system would be.
Although the tests followed a logical sequence and there were more tests run
than originally scheduled, much information was obtained by just going through
the operations involved in running a demonstration. For instance, manual
control of the froth level in the flotation machine was found to be most suitable
for field operation, and severe build up of oil in the discharge box of the flota-
tion machine was observed when the machine was operated without overflow.
Demonstration 1 took place with a sand feed rate of 30 tonsAour at a No. 4 fuel
oil concentration of 0.5%. Water exited the system at about 450 gallons /minute.
The residence times in the attrition scrubber (750 gallon volume) and flotation
machine (1800 gallon volume) were 4.8 and 3. 6 minutes respectively. Aeration
was set at 280 cubic feet/minute (47 cubic feet/minute/cell) in the froth flota-
tion machine. Note that with the scrubber in use additional scrubbing was a
always supplied by the vertical slurry pump which delivered the discharge from
the scrubber to the flotation machine's feed box. The oil in the feed sand was
not analytically determined during this test, but, based on observations of the
homogeneity of the feed sand on the feed belt, such analyses were made for the
tests which followed. For this demonstration the cleaned sand contained an
average oil concentration of 133 ppm and the exit water 286 ppm. Note that
brackish water from a wellpoint system was used rather than sea water.
Demonstration 2, because of problems with sanding in of the process water tank,
involved a decreased sand feed rate of 23 tonsAour; the level of No. 6 fuel oil
in the sand was measured at 3,770 ppm. The total exit water rate was 445
gallons/minute, and the scrubber and flotation residence times were 5.2 and
3.8 minutes respectively. Aeration was 280 cubic feet/minute. The cleaned
sand contained 147 ppm oil and the exit water 232 ppm. The test demonstrated
that given essentially the same operating conditions there was no radical differ-
ence in performance between removing No. 6 or No. 4 fuel oil.
Demonstration 3 involved a sand feed rate of 19 tonsAour with a No. 4 fuel oil
concentration of 4100 ppm. The total water rate was 450 gallons/minute with
residence times of 5.2 and 3.7 minutes for the scrubber and flotation machine
respectively. Aeration was again set at 280 cubic feet/minute. During this test
pine oil was fed to the feed box of the flotation machine at a rate of one cc/
minute; this frothing agent was used in an attempt to stabilize the froth and,
thereby, possibly enhance the oil collection efficiency of the flotation machine.
The pine oil appeared to have little effect since the cleaned sand contained 196
42
-------
ppm oil and the exit water 138 ppm. The decrease in oil in the exit water
(which was quite significant) should be attributed entirely to the lower sand feed
rate (compare to the first demonstration).
Demonstration 4 involved a moderate sand feed rate of 25 tons/hour with No. 4
fuel oil as the contaminant. A higher feed rate was feasible for this (and follow-
ing) tests because the adjustable orifice on the dewatering cyclone had been
removed in order to prevent sanding in of the process water tank. The feed
sand contained 5,480 ppm oil by dry analysis and 5,220 ppm by wet analysis;
given the inherent errors in sampling, these analyses compare very favorably
so the simpler dry analysis was adopted as a standard technique. Aeration was
still maintained at 280 cubic feet/minute. The total water rate during the test
was 420 gallons/minute with residence times of 4.9 and 3.9 minutes for the
scrubber and flotation machine respectively. Pine oil was again added at a rate
of one cc/minute to the flotation feed slurry. Note that for this test each half
barrel of contaminating oil was mixed with a bale of wheat straw before contam-
inating the sand. 160 and 806 ppm oil respectively in the cleaned sand and exit
water was the result. Again, the pine oil appeared to have little, if any, effect
while the straw apparently caused an increase in the oil dispersed into the pro-
cess water. This phenomena was also observed in later tests and is best
explained by competition for attachment to air bubbles between the oil and the
straw, the straw being preferred. The very oily appearance of the feed water
during this test led to the inclusion of feed water oil analyses for the following
tests. Oil build up in the process water supply was obviously leading to overly
high oil concentrations in the exit water from the plant.
Demonstration 5 involved a sand feed rate of 30 tons/hour at a concentration of
No. 4 fuel oil of 8,361 ppm. Note that the feed sand was prepared (by estimate)
to contain about 6 times the oil as that in the previous tests; the analyses did not
bear this out, illustrating once again the need for them. The practice of esti-
mating oil concentrations from the procedure used in preparing feed sand was
obviously subject to large errors. Water was fed to the plant at 560 gallons/
minute with residence times in the scrubber and flotation machine of 4.1 and
3.0 minutes respectively. Aeration was 280 cubic feet/minute and pine oil was
again added at one cc/minute. Samples of beach sand were taken prior to con-
tamination and showed an average equivalent oil level of 23 ppm; although not
very significant for this particular demonstration, such a background level could
be for the sort of performance desired and projected for a mobile unit. The
cleaned sand and exit water contained average oil concentrations of, respective -
ly, 711 and 537 ppm while the feed water contained an average of 1,186 ppm oil.
Obviously, the process was working well during this test, but the water recycle
was severely degrading the effluent. The data for demonstration 5 clearly indi-
cated the rise in feed water and residual oil concentrations which took place
with time. The high residual oil levels were not due entirely to the recycle
water, however; the high feed oil concentration, low flotation residence time,
and severe dispersing action by the attrition scrubber and vertical pump also
served to promote unacceptable cleaned sand and exit water.
Demonstration 6 used a sand feed rate of 20 tonsAour with 4,090 ppm No. 4 fuel
oil. Aeration was 280 cubic feet/minute while the total water rate was 320
gallons/minute. Pine oil was again fed at one cc/minute. The scrubber resi-
dence time was very high at 9.4 minutes while the flotation residence time was
43
-------
set at 5.1 minutes. The feed water contained an average of 251 ppm oil while
the cleaned sand and exit water contained 329 and 1, Oil ppm respectively. The
deleterious effect of scrubbing the oily sand is graphically illustrated by the
results of demonstration 6. Of special interest is the fact that the feed water
increased in oil concentration by a factor of about 55 during the test while the
exit water oil concentration increased only by a factor of 2. The results of this
demonstration pointed out the necessity for a scrubber bypass (and possibly
bypass of the vertical pump) to promote better cleaning and also simulate the
operation of a mobile system.
Demonstration 7 involved feeding oily sand (5, 020 ppm, No. 4 fuel oil) which
had been aged for 15 days prior to the test} the feed rate was 25 tons/hour.
Air and pine oil were again fed to the froth flotation machine at 280 cubic feet/
minute and one cc/minute respectively. The total water rate was 550 gallons/
minute while the scrubber and flotation residence times were 3.8 and 3.1
minutes respectively. The feed water, cleaned sand, and exit water contained
an average of 628, 140 and 749 ppm oil respectively. Froth stability was ob-
served to have decreased due to the aging of the oil for this demonstration;
sufficient pine oil to overcome this would have implied adding more pollutant to
the effluent streams. Extra aeration capacity was deemed to be a more reason-
able way to increase the frothing action. Although only in a qualitative sense,
the data does appear to indicate that aging of the oily sand (obviously within
limits) facilitates the cleaning process. Loss of volatile and water soluble oil
fractions accounts for this effect in light of the results for different oil types.
Demonstration 8 involved a sand feed rate of 19 tonsAour with a No. 4 fuel oil
concentration of about 5,000 ppm (see below for exact values). The water rate
to the attrition scrubber was 215 gallons/minute (a residence time of 3.1
minutes) while the total process water rate was 250 gallons/minute. Tests were
run sequentially using the first four, three, two, and one cells of the froth
flotation machine with aeration rates of 280, 210, 140, and 70 cubic feet/minute
respectively; note that aeration was 70 cubic feet/minute/active cell throughout
the demonstration. Pine oil was again fed to the process at approximately one
cc/minute. The feed sand, feed water, cleaned sand, and exit water contained
respectively the following amounts of oil for each of the four tests: (four cells)
5,630, 296, 155, and 477 ppm, (three cells) 4,700, 322, 214, and 529 ppm,
(two cells) 5,480, 257, 159 and 540 ppm, and (one cell) 4,630, 232, 168, and
669 ppm. The active cell flotation residence times for the four, three, two and
one cell tests were, respectively, 4.2, 3.2, 2.2. and 1.1 minutes. The test
results indicate that the major portion of the1 separation took place in the first
cell of the flotation machine while the latter cells served as "cleaners." This
is most clearly illustrated by the fact that the additional oil added to the process
water as a result of the sand cleaning process was 181, 207, 283, and 437 ppm
for the four, three, two and one cell tests respectively; although these values
correlate with the residence times, the effect of the individual cells is obviously
not the same relative to the degree of oil removal attained by each. These tests
gave a most favorable projection for the performance of a trailer mounted
mobile unit involving only froth flotation with sand and water feed.
Demonstration 9 involved a sand feed rate of 30 tonsAour with a No. 4 fuel oil
concentration of 3,590 ppm; the total water rate was 250 gallons/minute and
the scrubber and flotation residence times were 3.1 and 4.2 minutes respecttve-
44
-------
ly. Using the experience gained in the previous demonstration, this entire test
was run with only four active cells, the idea being to obtain long term, "steady
state" data on such an operation. Pine oil was again added to the feed at one
cc/minute. Aeration was set at 210 cubic feet/minute (52 cubic feet/minute/
active cell) to allow better comparison to the tests with all six cells active.
The feed water, cleaned sand, and exit water contained respectively an average
of 260, 167, and 349 ppm on during this test. Such a result was thought to be
quite favorable for a mobile operation since dispersion with both the attrition
scrubber and vertical pump must have degraded the operation significantly.
Demonstration 10 took place with a bypass of the attrition scrubber using a
metal sluice. Sand was fed to the plant at 19 tonsAour (No. 4 fuel oil at 7,020
ppm) with a total process water rate of 440 gallons/minute (a flotation residence
time of 3.8 minutes with all six cells active) and an aeration rate of 280 cubic
feet/minute. The feed water, cleaned sand, and exit water analyzed respec-
tively at 273, 106, and 496 ppm oil for this demonstration. Given the relatively
high feed oil concentration, continuing dispersion by the vertical pump, and a
low flotation residence time, these results are quite good.
Demonstration 11 involved a very high sand feed rate of approximately 60 tons/
hour with a No. 4 fuel oil concentration of 5,110 ppm. The total process water
rate was 395 gallons/minute with scrubber and flotation residence times of 3.4
and 3.7 minutes respectively. Aeration was again set at 280 cubic feet/minute
with all six flotation cells active. Although pine oil had an aesthetic impact in
the previous tests (it definitely masked the smell of residual oil and possibly
decreased the oily feel of the cleaned sand and exit water), its use was discon-
tinued as of this test since it had no apparent effect in aiding in removal of oil
from beach sand. The results of the test were 1,992, 389, and 2,141 ppm oil
in the feed water, cleaned sand, and exit water respectively. Given the high
BUI^ *6ed rate and ^ow ^otation residence time, these values are quite respect-
able. Low water temperatures at me site were also responsible for a degree of
degradation in the cleaning process; the process water was estimated to be at
5°C.
Demonstration 12 involved a decreased sand feed rate of about 8 tonsAour with
a high No. 4 fuel oil concentration at 16,980 ppm. The total process water rate
was 320 gallons/minute with scrubber and flotation residence times of 3.5 and
5.4 minutes respectively. Aeration was 280 cubic feet/minute. The measured
water temperature during this test was 5°C. Feed water, cleaned sand, and
exit water oil concentrations measured 1,200, 392, and 1, 678 ppm respectively.
Given the dispersal action of the scrubber and vertical pump, the low water
temperature, and the high feed oil concentration, the results of this demon-
stration were about as expected.
Demonstration 13 involved the use of sea water rather than the brackish well-
p°*nt water which had been used in the previous tests. This water was obtained
with a submersible pump deposited in the surf 450 feet from the plant and a run
of hose to the plant's process water tank. The water temperature for this test
was 7°C, and the total process water rate was 305 gallons/minute with scrubber
and flotation residence times of 3.6 and 5.1 minutes respectively. Aeration was
again 280 cubic feet/minute. Sand was fed to the plant at 30 tons/hour and con-
tained a somewhat high level of oil, 8,250 ppm. The feed water, cleaned sand,
45
-------
and exit water contained 296, 137 and 353 ppm oil respectively. These are very
good results. Sea water obviously facilitates the removal of oil by froth flota-
tion. Since, based on the results of the previous results, it is known that low
water temperatures degrade the process, higher sea water temperatures should
produce quite high quality effluent streams.
Demonstration 21 involved a medium crude oil as the contaminant; sand with
an oil concentration of 9,550 ppm was fed to the plant at 30 tons/hour. Aeration
was still 280 cubic feet/minute, and the process water temperature had risen
to 10°C which is still 11 Centigrade degrees below the nominal design value.
The total process water rate was 370 gallons/minute with scrubber and flotation
residence times of 3.8 and 4.3 minutes respectively. The feed water, cleaned
sand, and exit water contained 350, 185, and 517 ppm oil respectively. Given
the high feed oil concentration this was quite an acceptable performance, and
the major implication was that a crude oil could also be removed from sand
using the same basic process as heretofore had been used only for fuel oils.
Demonstration 22 involved an increased aeration rate of 350 cubic feet/minute
(58 cubic feet/minute/cell); this increased rate was made possible by the
acquisition of a larger air blower. Sand with 6,920 ppm No. 4 fuel oil was fed
to the plant at 30 tonsAour. The total water rate was 360 gallons/minute with
scrubber and flotation residence times of 3.2 and 4.4 minutes respectively. The
feed water, cleaned sand, and exit water contained 1,970, 243, and 1,780 ppm
oil respectively. Although the water temperature was low and the feed oil
concentration somewhat high during this test, the results still indicate that
(with the pilot plant) this increased aeration was deleterious. The extra
agitation supplied by the higher aeration rate apparently caused more oil to be
"permanently11 dispersed into the process water. An optimum aeration (of a
semiquantitative sort) of 50 cubic feet/minute/active cell had, therefore, been
determined for the stationary plant.
Demonstration 41 was a long term operation run which lasted for 20 hours,
samples of the feed and effluent streams being taken for the first 6-1/2 hours.
Sand with 4,750 ppm No. 4 fuel oil was fed to the plant at 30 tonsAour. The
process water rate was 330 gallons/minute with residence times of 3. 5 and 4.8
minutes for the scrubber and flotation machine respectively. Aeration was at
280 cubic feet/minute. The feed water, cleaned sand, and exit water contained
an average of 2, 050, 703, and 2,460 ppm oil respectively. The major conclu-
sions from this demonstration were: (1) with closed loop operation of the
stationary plant, oil can build up in and hold at very high levels in the process
water, (2) long term operation of such a process plant to clean oily sand
presents no special problems, and (3) night operation of such a process is not
difficult given adequate lighting and stockpiles of feed sand.
3.4.3 Mobile Unit Simulations
Demonstrations 14 - 20 in Appendix C involved the use of only the froth flotation
machine and process water supply of the stationary plant. A leased, portable
belt feeder was used to lift oil contaminated sand from grade level to the first
level of the flotation machine platform where a chute ffirected the oily sand into
the feed box of the flotation machine. The cleaned sand slurry was discharged
with an overflow chute from the horizontal pump's sump to the underflow dis-
46
-------
charge area for the dewatering cyclone. In summary: the feed sand was not
scrubbed by either the attrition scrubber or the vertical pump during these tests.
The feed sand entered the flotation machine directly at the feed box. Process
water was not recycled, and the cleaned sand/water slurry was sampled direct-
ly as discharged from the flotation machine. Sea water was also used through-
out this series of tests.
Demonstration 14 involved a sand feed of 17.5 tonsAour with a No. 4 fuel oil
contamination level of 5,290 ppm. The aeration rate to the flotation machine
was 350 cubic feet/minute (for six active cells), and the water rate was 280
gallons/minute for a flotation residence time of 5.9 minutes (water tempera-
ture: 8°C). The cleaned sand and exit water contained 115 and 276 ppm oil
respectively. Given the use of cold feed water, this result was quite acceptable.
Demonstration 15 involved a sand feed rate of 24.5 tonsAour at 6,130 ppm No.
4 fuel oil. Aeration was again 350 gallons/minute, and the water feed rate was
250 gallons /minute (a flotation residence time of 6.3 minutes). The cleaned
sand and exit water contained respectively 157 and 280 ppm oil. Again, given
the low temperature of the feed water and comparing to me results of the pre-
vious demonstration, this result is quite good.
For demonstration 16 a sand feed rate midway between those of the previous two
tests was selected (22.5 tonsAour); as in the previous test, the feed concen-
tration of No. 4 fuel oil was unintentionally high at 6,680 ppm. Aeration was
again 350 cubic feet/minute while water was fed to the process at 250 gallons/
minute for a flotation residence time of 6.3 minutes. The cleaned sand con-
tained 233 ppm oil, a result which is hard to explain in light of the results of
the previous tests. The exit water contained 260 ppm oil which is lower than
me values for the previous tests and perhaps explains why the apparent residual
oil level in the cleaned sand was high. Although somewhat anomalous, the
results were still fairly good.
Demonstration 17 involved a much reduced process water rate (125 gallons/
minute) which with a sand feed rate of 27 tonsAour (5,320 ppm No. 4 fuel oil)
gave a flotation residence time of 10.9 minutes. The exit oil concentrations in
the cleaned sand and water were 135 and 407 ppm respectively. Since, especial-
ly, the latter value appears high, the conclusion must be that the dense slurry
(about 46% by weight sand) used in this test served to trap oil and hinder separa-
tion in a manner which even the considerably higher residence time could not
overcome. However, note that the field engineers judged (by visual appearance)
the discharge slurry during this demonstration to be much more acceptable than
those of the previous three. A dense slurry more readily traps and holds
residual oil keeping the oil from surfacing as a froth which accounts for this
visual enhancement.
Demonstration 18 took place under conditions similar to 17 (sand feed of 26 tons/
hour, process water at 116 gallons/minute, and aeration 350 cubic feet/minute)
except for the feed sand oil concentration which was intentionally high at 6,930
PPm. The water temperature had dropped to 6°C for this test, and the flotation
residence time was 11.6 minutes. The cleaned sand and exit oil contained 183
and 524 ppm oil respectively. Again, due probably to the high slurry density,
the analytical results for residual oil do: not appear very good; once again, how-
47
-------
ever, the field engineers reported that the discharge appeared quite acceptable
to the eye.
Demonstration 19 involved No. 6 fuel oil as the contaminant at a level of 5,730
ppm in sand fed to the process at 26 tonsAour. The process water rate was
206 gallons/minute, and the flotation residence time 7.3 minutes. The cleaned
sand and exit water contained 92 and 100 ppm oil respectively. Given the low
water temperature on the day of this demonstration, this was an excellent
result.
Demonstration 20 involved, intentionally, a very high feed concentration of No.
4 fuel oil at 19,810 ppm. To compensate for this high oil input to the system,
the sand feed rate was decreased to 13 tonsAour. The process water rate
was 200 gallons/minute which set the residence time at 7.7 minutes. The
cleaned sand and exit water contained 181 and 599 ppm oil respectively. Given
the low feed water temperature (6°C) and the very high feed oil concentrations,
this is a reasonably good result.
3.4.4 Straw Scrubbing
Demonstrations 23, 24, and 25 were run to determine if oil could be scrubbed
from straw to facilitate disposal of the latter material. Straw saturated with
No. 6 fuel oil was attrition scrubbed for various lengths of time, under various
conditions. Liberation of oil into the carrier water was determined by analyz-
ing for the oil.
Demonstration number 23 involved scrubbing for 6 minutes and resulted in 34
parts per million oil in the exit water.
Demonstration number 24 involved scrubbing for 12 minutes and resulted in 40
parts per million oil in the exit water.
5 gallons of diesel fuel was added to the slurry in the attrition scrubber for
demonstration number 25; after 12 minutes of scrubbing, the exit water con-
tained 64 parts per million of oil. Addition of 5 more gallons of diesel fuel and
scrubbing for 12 more minutes resulted in an exit concentration of 75 parts per
million.
Although the data for the dispersion of oil into the process water moves in the
right direction for these tests* the results are hardly favorable. Perhaps more
important are the extreme operating difficulties which the field engineering
crew had with performing this task. The straw/oil mass tended to wrap around
the impeller shaft in cell number 1 of the scrubber and was then extremely
difficult to remove. This problem is similar to that observed around the im-
pellers of the flotation machine; in this latter case, however, the air diffusing
from the area of the impeller serves to keep straw away, and, therefore, when
such a straw buildup occurs, it occurs very, very slowly. The conclusion from
these tests is that the recovery of oil contaminated straw by attrition scrubbing
is not feasible.
3.4.5 Oil/Water Separation
48
-------
Demonstrations 26-36 considered the separation of oil from water using the
standard minerals processing equipment existent at the plant site. Most of these
tests took place at high flow rates, in some cases elevated by the use of more
than one water supply pump. Generally, the results showed that standard pro-
cessing equipment is not very efficient for oil/water separation; however, for a
rough separation at high capacity such equipment may be of value. The data
illustrates both the potential usefulness of the mobile beach cleaner as an oil/
water separator and the need for redesign and/or modification in operation to
attain very efficient oil/water separations using the same basic principle (froth
flotation).
Demonstration 26 involved a feed water rate of 485 gallons/minute with 5,130
ppm No. 4 fuel oil. The oil was added to the suction side of the vertical pump
which delivered the oil/water mixture to the flotation machine. Aeration was
set at 280 cubic feet/minute or 47 cubic feet/minute/cell. The residence time
jn the flotation machine was 3.7 minutes resulting , in an average of 897 ppm oil
in the discharge. 1>e efficiency of the oil/water separation was, therefore,
about 83%.
Demonstration 27 was run under conditions equivalent to 26 except for an incres-
ed aeration rate of 560 cubic feet/minute or 93 cubic feet/minute/active cell.
The result was an effluent oil concentration of 803 ppm or a recovery efficiency
of about 84.5%. Although the increased aeration improved the separation some-
what, the improvement was not startling.
Demonstration 28 involved a feed water rate of 455 gallons/minute with 5,460
Ppm of a 1:1 mixture of No. 4 fuel oil and No. 260 diesel fuel. The oil was again
fed to the suction side of the vertical pump, air was supplied at 560 cubic feet/
minute, and the residence time was 3.9 minutes. The effluent water contained
1.370 ppm oil for a recovery efficiency of about 75%. The added difficulty in re-
covering lighter oils is again illustrated by this data.
Demonstration 29 was a check on the ease with which oily straw could be remov-
ed from water by froth flotation; an indication was also gained as to the extent
of dispersion of oil into water by the process. The feed stock was prepared by
mixing 3 bales of wheat straw with 14 gallons of No. 6 fuel oil. With a water
rate of 465 gallons/minute to the flotation machine and residence time of 3.9
minutes, this material was placed into the feed box of the flotation machine by.
hand; the feed water stream carried the oily straw into me first cell of the flo-
tation machine. Visual observations both of the cells in and the discharge from
the flotation machine indicated that all the straw separated in the first cell and
none left with the effluent. The test lasted for ten minutes and resulted in an
average residual oil concentration of about 2 ppm. Since, during the test, 4,650
gallons of water and 14 gallons of oil were fed to the machine, the equivalent
feed concentration of the oil was about 3,000 ppm; this implies 99.98% recovery
of the oil which is excellent.
Demonstration 30 was essentially a repeat of 27 with oil added directly to the feed
box of the flotation machine. An input No. 4 oil concentration of 4,830 ppm with
a water rate of 515 gallons/minute and a residence time of 3.5 minutes resulted
in an average effluent oil concentration of 2,870 ppm for a recovery efficiency of
41%. Addition of oil at the feed box had been expected to increase efficiency;
-------
the contrary result points to channeling within the flotation machine which de-
grades operation (i.e., the larger water feed stream apparently keeps the oil
feed stream away from the active cell volume so that it cannot readily be
separated).
Demonstration 31 involved the same conditions as 30 except that four tests were
run sequentially with air to the 6th; 5th and 6th; 4th, 5th and 6th; 3rd, 4th,
5th and 6th cells shut off. Residence times of 2.9, 2.3, 1.7, and 1.2 minutes
and aeration rates of 112, 140, 187, and 280 cubic feet/minute/active cell re-
spectively were the pertinent operating conditions. The four tests resulted in
average effluent oil concentrations of 3,350, 3,430, 3,110, and 2,560 ppm
respectively. This implies oil recovery efficiencies of 31, 29, 36, and 47%
respectively. Obviously the increased aeration per active cell served to count-
eract the effect of decreasing residence time; however, decreased efficiency
with oil feed to the feed box of the flotation machine was again apparent. Due
to this, oil feed to the suction side of the vertical pump was resumed in the
following tests.
Demonstration 32 involved a feed oil prepared by scrubbing to a more dispersed
state in the attrition scrubber; water was fed to the scrubber at 120 gallons/
minute with' a No. 4 fuel oil concentration of 20,000 ppm for a residence time of
6.2 minutes. This dispersion was pumped to the flotation machine with addition-
al water to make a total water rate of 515 gallons/minute and an input oil concen-
tration of about 4,660 ppm. With a residence time of 3.5 minutes in the flotation
machine, the discharge stream contained 1,340 ppm oil implying a recovery
efficiency of about 71%. Performance was substantially degraded by the dis-
persal of the oil into the water with the scrubber (compare to the results of
demonstration 27).
Demonstration 33 involved a decreased feed water rate of 250 gallons/minute
with No. 4 fuel oil from the suction side of the vertical pump at 5,000 ppm.
Aeration was set at 560 cubic feet/minute and the flotation residence time at
7.2 minutes. The discharge water contained 2,7>90 ppm oil for a recovery
efficiency at 44%. This result implies that the increased residence time caused
increased internal circulation within the flotation machine which, in turn,
caused an increase in dispersion of the oil into me water with the resulting de-
crease in efficiency.
Demonstration 34 was similar to 27 except mat the feed stream contained
48,500 ppm No. 4 fuel oil. The result was an average effluent oil concentra-
tion of 22,100 ppm for a recovery efficiency of 54%. Note that the presence of
a quantitatively large amount of oil in the flotation machine appears to decrease
the efficiency of the separation process significantly.
Demonstration 35 was similar to 27 except that No. 6 fuel oil was used as the
contaminant. The flotation residence time was 3.5 minutes with a feed oil con-
centration of 4,850 ppm. The exit water contained an average of 517 ppm oil
for a recovery efficiency of 89.3%; again, the relative ease of separating the
heavier oils was apparent.
Demonstration 36 involved a much increased feed water rate of 1,050 gallons/
minute with a No. 4 fuel oil concentration of 4,760 ppm; the resulting residence
50
-------
time was 1.7 minutes with an applied air rate of 560 cubic feet/minute. The
exit water contained an average of 3, 160 ppm oil for a recovery efficiency of
34%. With all other conditions being approximately the same, the decrease in
residence time obviously decreased the efficiency of the process very signifi-
cantly (see the results of demonstration 27).
3.4.6 Sand Scrubbing and Dewatering
Demonstrations 37 - 40 involved scrubbing oily sand and then dewatering using
tiie Krebs cyclone. The purpose of these tests was to gain a quantitative idea
as to the feasibility of cleaning oily sand by repeated scrubbing and dewatering.
Given a large water supply and sufficient settling tanks or ponds, such an
approach may be feasible.
Demonstration 37 considered a sand feed rate of 30 tons/hour with No. 4 fuel
oil at an average concentration of 4,500 ppm. The oily sand was scrubbed for
3.0 minutes and then dewatered with a total water flow rate of 537 gallons/min-
ute. The cleaned sand contained an average of 278 ppm oil while the feed and
exit water had 1,127 and 1,540 respectively; these latter values indicate an
addition of about 413 ppm oil to the exit water by the scrubbing process. Since
the cyclone underflow contained an average of 64.4% sand, the data indicates
that 88.7% of the feed oil, exclusive of the oil in the feed water, was being re-
moved during the test; therefore, if the water entering the process was entirely
free of oil (i. e., if the process water tank was infinitely large, or the process
was run intermittently to allow the oil to float free from the feed water, or
100% oil/water separation was made on the overflow water from the dewatering
cyclone), the sand would have to pass through the scrubbing/dewatering sys-
tem at least one more time to achieve the same result as was usually achieved
with one pass through the froth flotation machine.
Demonstration 38 also involved a 30 tons/hour sand feed rate but with an in-
creased scrubbing time of 6.7 minutes and a total water rate to the cyclone of
477 gallons/minute. No. 4 fuel oil was present in the feed sand at 3,860 ppm.
Feed water, cleaned sand, and exit water contained 1,280, 271, and 1,710 ppm
oil respectively; these numbers indicate an increase in the exit water oil of
430 ppm due to the scrubbing. With 64.8% sand in the cyclone underflow, 87%
removal of the feed oil is indicated, This is little different than the result of
the previous test, and, again, another pass would be necessary to obtain results
similar to froth flotation separation given that oil free feed water was available.
Demonstration 39 involved No. 6 fuel oil; scrubbing took place for 4.9 minutes,
and the total water flow to the cyclone was 517 gallons/minute. The feed sand
contained 4,120 ppm oil. The feed water, cleaned sand, and exit water con-
tained 625, 222, and 829 ppm oil which indicates an increase of 204 ppm in the
feed water oil due to scrubbing. With 62.8% sand in the cyclone underflow, this
data indicates 95.4% feed oil removal. Again, to obtain results similar to
those with froth flotation removal of No. 6 fuel oil, another pass through the
system would be necessary as well as the use of oil free feed water.
Demonstration 40 involved No. 4 fuel oil in the feed sand at 19,600 ppm. The
sand was scrubbed for 5.6 minutes, and the total water flow to the cyclone was
473 gallons/minute. The feed water, cleaned sand, and exit water contained
51
-------
3,870, 890, and 4,700 ppm oil respectively, indicating an increase in the
water's oil content of 830 ppm due to scrubbing. A 64.0% sand content in the
cyclone underflow, therefore, implies 93.1% oil removal by the process. Again,
another pass through the system would be necessary to obtain results equiva-
lent to froth flotation separation; of course, this projection again assumes an
oil free feed water and implies oil/water separation of the cyclone overflow.
3.4.7 Conclusions
The basic conclusions as evolved from the demonstration studies are as follows:
1) Sand cleaning using froth flotation is feasible although the follow-
ing factors (if included in the overall process) hinder performance.
a) Recycle of water with a reasonably large process
water tank leads to a buildup of oil in the process
loop and, therefore, unacceptable overall per-
formance.
b) Scrubbing or pumping of oil contaminated sand leads
to degradation in process performance since dis-
persion of oil into the carrier water more than
compensates for any additional release of oil from
the sand.
2) Froth flotation should be utilized by feeding sand and water direct-
ly into the feed box of a froth flotation machine with a minimum of prior agitation;
the data from the demonstration studies allowed specification of the operation of
a mobile unit, the result of which is presented in section 3.5.3.
3) Attrition scrubbing of straw with standard process equipment is
not feasible.
4) Oil/water separation at high capacity is possible using froth
flotation although with standard mineral processing equipment the process is
not very efficient; considerable development work would probably be necessary
to take advantage of the basic principle with better separation efficiency.
5) In principle, oily sand scrubbing and dewatering works; however,
necessities for a very large process water supply, oil/water separation, and
most probably multiple scrubbing operations make the overall approach in-
feasible.
3.5 Mobile Beach Cleaner
3.5.1 Introduction
Based on the results of the work with me stationary plant, a mobile beach
cleaner was constructed, tested, and modified. This unit involved basically a
froth flotation machine in combination with a belt feeder for sand, a submersi-
ble pump for water, a supercharger for air, and a diesel generator for electri-
cal power? the entire process plant was mounted on a 40 foot trailer. Intention-
-------
ally. the mobile process involves no scrubbing, pumping, or dewatering of oil/
water/sand slurries. Scrubbing and pumping had been found to degrade the oil
separation process (i.e. , make it much more difficult). Dewatering for water
recycle was deemed undesirable due to buildup of oil in the process water; de-
watering only to segregate effluent water and sand at an actual beach site was
judged to be an unnecessary, additional unit operation.
On January 18, 1972, after extensive field testing and reconditioning, the
mobile beach cleaner was delivered to the Environmental Protection Agency at
the Edison Water Quality Laboratories of Edison, New Jersey.
3-5.2 Design Specifications
Standard over-the-road trailers without special permits are limited to a 40 foot
length and 8 foot width in most States 6f the Union; additionally, the per axle
weight for a tractor -trailer combination is limited typically to a maximum of 8
tons. Therefore, a trailer which conforms to these specifications can generally
travel through the United States without special license or permit, Slight devia-
tions from these limits, however, have little or no effect on transport under
emergency conditions.
basic trailer limitations just mentioned place limits on any processing
operation which is to be mounted on a single trailer. Maximum deck area on a
40 foot trailer is attained by having an elevated deck on the forward section of
tiie trailer of about 12 feet in length. This deck is suitable for placement of low
Profile equipment and storage during transport; it also provides a good work
area during actual plant operation. Control panels, diesel generator, air
supercharger, and fuel tanks were, therefore, placed on the upper deck with
tiie rest of the deck clear except during transit. The 8 x 28 foot lower deck was
available for the larger process equipment plus storage. Since approximately
10 feet was the minimum length necessary to accommodate a 100 cubic foot
feed hopper plus a steeply inclined conveyor belt, an area 8 by 18 feet was left
for a froth flotation machine including open area at the discharge end. The
available area essentially specified the size flotation machine which could be
incorporated into the mobile system. With the data from the demonstrations
using the stationary plant, specific equipment selection could be made accom-
panied by relevant design operating estimates.
Although the design of the mobile unit considered equipment specification and
Performance estimates in a concurrent fashion, these two elements of the design
are separate in the paragraphs which follow. First, the equipment specification
is considered. Including the trailer, there were six major process items to be
specified; these are called out with appropriate identification below.
1) Trailer: Basically a standard 40 foot flatbed trailer with an oak
Plank upper deck; the lower deck was supplied as open to the structural web and
Was eventually filled in by the process equipment and storage areas. The
trailer has dual axles, a reinforced push bar at the rear, both six and seven wire
connectors for the running lights, standard commercial hitch, and mechanical
leveling and support jacks at each of the four corners of the main frame. The
(serial number 20BDH7101S) was supplied by General Engines Co, , Inc.
53
-------
of Thorofare, New Jersey.
2) Hopper and Belt Feeder: To fit into the limited space available,
the hopper and belt feeder had to be specially designed and constructed. As
originally conceived, this unit came with a fixed speed, cleated belt conveyor.
Sand left the approximately 100 cubic foot hopper at the bottom by gravity flow;
a vibrator was supplied within the main hopper volume, attached to the main
structural cross beam, to assist in moving the sand through the bin. A slide
gate was supplied at the bottom exit port of the bin to control the sand feed
rate; sand, in principle, fell through the gate opening onto the conveyor belt
which then lifted it to the feed box of the flotation machine into which it dropped.
A 6 inch bar screen and 1/4 inch wire mesh screen were supplied at the top of
the feed hopper. In practice, the single vibrator proved inadequate; at the
request of Meloy Laboratories, the manufacturer supplied an additional vibrator
for the hopper which was installed on the outside, near the bottom of the hopper,
Similarly, the 6 inch bar screen proved to be an annoyance during actual use of
the unit; therefore, during most operations the screen was removed. The 1/4
inch screen proved inadequate to pass the required amount of oily, wet sand;
Meloy Laboratories1 engineers, therefore, removed this screen and replaced it
with a (nominally) 1-1/2 inch expanded steel screen which eventually proved
adequate. Most importantly, the fixed speed conveyor in combination with the
exit slide gate proved incapable of providing a steady sand fedd rate; further-
more, this combination did not supply a feed rate as low as required by the
system design (about 40 tons/hour). Meloy Laboratories1 field engineers,
therefore, had to replace the original drive unit for the conveyor with a proper-
ly sized variable speed drive unit. A 3 horsepower U.S. Varidrive was pro-
cured and installed in the field; this unit gave steady operation over the requir-
ed sand feed range (10 to 40 tonsAour). Very positive control over the
conveyor belt proved to be the only feasible way to reliably feed oily, wet sand
to the froth flotation machine. The bin and belt feeder, as originally designed
and constructed, was supplied by the American Industrial Corporation of
Virginia Beach, Virginia. The U. S. Varidrive, additional vibrator, and modi-
fied inlet screen were (to correct the aforementioned deficiencies in the equip-
ment as originally supplied) engineered into the system by the Meloy Laborator-
ies staff.
3) Flotation Machine: Given the space available and existing froth
flotation machines similar to the machine used in the stationary plant, a maxi-
mum capacity machine was easily selected. The Denver Equipment Company
of Denver, Colorado supplied a 4-cell D-R 21 froth flotation machine. This
machine has a volume per cell of 40 cubic feet for a total of 160 cubic feet and
is 16 feet 10 inches long, 4 feet 11 inches wide, and 7 feet 4 inches high.
Diffuser impellers in the machine are powered by two 10 horsepower motors,
and a manifold is supplied so that the diffusers can be operated with super-
charged air. The machine was supplied with feed and discharge boxes, and the
latter item had four discharge ports so that the cleaned sand/water slurry could
be piped in three directions away from the machine. A manually operated dart
valve was supplied for control of the rate of sand discharge from the machine.
Meloy Laboratories supplied a sheet metal launder for the oil froth overflow
from the machine.
4) Submersible Water Pump: This item was supplied by Prosser
54
-------
Industries of Anaheim, California. This marine pump (No. 9-16034) is totally
submersible and rated at 15 horsepower, 440 volts, 3-phase. It has a large
Inlet screen and a 4 inch diameter threaded outlet pipe and with its high volume
impeller can deliver the maximum design water rate of 350 gallons/minute
against a 90 foot head and the nominal water rate of 250 gallons/minute against
a 110 foot head. The pump was supplied with waterproof electrical cable and 4
inch diameter flexible hose. Meloy Laboratories mounted the pump on a sled
for deployment into the process water reservoir.
5) Air Blower: This item (size N19P-5) was supplied by the New
York Blower Company of Chicago, Illinois. With a 6 inch discharge pipe this
blower (5 horsepower, 440 volts, 3-phase) can deliver 550 cubic feet/minute of
air against a static head of 30 inches of water. The actual air delivered by the
blower is set with a blast gate.
6) Diesel Generator: North American Engines Co., Inc. of
Alexandria, Virginia supplied the diesel generator. Powered by a Perkins
diesel engine? this unit supplies 50 kilowatts at 440 volts, 3-phase. A 5 kilo-
watt, 110 volt, single phase transformer was supplied as an integral part of the
generator.. Fuel tanks, electrical connections to the generator, and control
panels were supplied by Meloy Laboratories.
In addition to the six main items, a very large amount of support material, as
well as structural additions and modifications, for the basic unit was supplied
by Meloy Laboratories. Ladders, catwalks, stairways, storage compartments,
piping, flow indicators, oil recovery tank, electrical wiring, control panels,
tow hooks and chain, and fuel tanks are major examples of these. Relatively
minor items (such as pipe support tripods, stakes, floodlamp poles and sockets,
shovels, rope, hand tools, hose clamps, and tie-down chains) were supplied in
very large numbers. The aim was (except for fuel supply, front end loader and
elevating scraper support, and major equipment repair) to make the mobile
beach cleaner an entirely self contained processing plant.
3.5.3 Performance Estimates
Since the sizes of the basic unit process were limited by the space available on
a 40 foot flatbed trailer (multiple trailers having been ruled out as unnecessar-
fly elaborate for a first approach to mobilization), the data from the demon-
strations with the stationary plant was used to estimate the performance of the
mobile unit. The data had to be considered without including the effects of
slurry pumping, slurry scrubbing, and water recycle. Sand, water, and aera-
tion rates had to be set based on the observations and measurements made during
the demonstration studies with the stationary plant. Contaminating oil type and
concentration also had to be specified based on prior experience. This may
appear as a somewhat arbitrary approach, but, in the absence of independent
criteria for required performance, no other approach was possible.
Observations with the stationary plant indicated that a 30 ton/hour sand feed
rate to a 4- cell, D-R 21 flotation machine should promote reasonably good sand
cleaning from an aesthetic viewpoint. Note that this was the opinion of the pro-
ject engineers and is subject to much individual interpretation as to what
constitutes acceptable sand cleaning.
55
-------
An oil contamination level of 0. 5% or 5, 000 ppm in the feed sand was also
assumed in making the above judgement. Fresh No. 4 fuel oil was selected as
the contaminant since it had been used extensively in the stationary plant studies
and was known to be intermediate in difficulty of removal as compared to the
lighter (more difficult) and heavier (less difficult) oils. Fresh No. 4 fuel oil
also provides a worst case in terms of the effects of aging oily sand since
observations with the stationary plant had indicated that any aging up to four
weeks improves the recovery process.
Use of sea water at a temperature of 70°F (21 °C) was also assumed in making
the performance estimates; sea water was considered because it is a reason-
ably well defined solution and would be the water present for the majority of
large beach oil spills. The design temperature was selected both for improved
performance (as compared to considering fresh No. 4 rather than No. 6 fuel oil
which allowed for degraded performance) since temperature decreases had been
observed to degrade the separation and for realism in process operation since
decreased temperatures usually imply adverse weather conditions on an ocean
beach.
The water feed rate was assumed to be 250 gallons/minute to give a flotation
residence time of 4.0 minutes. Realizing that its effect (within reasonable
limits) was not drastic, aeration was assumed to be 300 cubic feet/minute or
75 cubic feet/minute/cell.
The sand was, of course, that found at the Dam Neck site; variations from this
process parameter are subject to future experience or testing. The oil was
assumed to be fairly well mixed with the sand (standard deviations! in random one
pound samples of lesr than 10%) since previous observations had indicated that
large variations of the oil concentration in the feed sand degraded performance;
if such variations are the case, the sand feed rate must be decreased. Large
variations in the sand feed rate also degrade process performance and, there-
fore, are to be avoided; attention must be paid to a factor like belt slippage
which can cause feed rates to vary. Large step changes in the other feed rates
must also be avoided.
Given the set of process parameters outlined above, the data from the station-
ary plant studies indicated that the following effluent oil concentrations could be
expected (subject to verification by actual sand cleaning data from full scale
operation of the mobile unit):
SAND 100 parts/million or less
WATER 175 parts/million or less
TOTAL 150 parts/million or less
The following qualitative variations about these effluent values are expected
given that the design criteria are an "optimum:"
(1) With all other parameters held constant, increasing the sand feed
rate will increase the residual effluent levels.
(2) Similarly, increasing the feed oil concentration will increase the
56
-------
effluent oil concentrations.
(3) Increase of the water feed rate should decrease the effluent oil
concentrations but also decrease the percentage of oil in the feed sand which is
recovered,
(4) Either increasing or decreasing the aeration rate should lead to
a decrease in recovery percentage and an increase in residual oil; the former
is due to decreasing oil/air contact and the latter to increased turbulence which
niore than compensates for increased oil/air contact.
(5) Lower process temperatures will lead to increased effluent oil
concentrations while, up to an undetermined limit, higher process temperatures
will result in better oil recovery.
(6) Aging of the oily feed sand (again, up to an undetermined limit)
improve oil recovery.
(7) Separation of heavier oils will be more effective \yhile lighter
°ils will be less easily and efficiently recovered.
(8) Use of fresh or brackish process water or water contaminated
with large amounts of interfering hydrocarbons should lead to decreased pro-
cess efficiency.
The following section discusses actual data obtained with the mobile beach
cleaner in the field under simulated full scale emergency conditions.
3.5.4 MobtLe^Beach Cleaner Demonstrations
Twenty four full scale demonstrations were run using the mobile beach cleaner.
The unit was pulled withraTJ7E Caterpillar Tractor to the beach at the Dam Neck,
Virginia demonstration site; sand contamination and cleaning took place on the
beach proper with only a WABCO 150B Front End Loader for support. During
four months of work on the beach, the unit had to be moved back from the beach
three times due to adverse weather conditions (one hurricane and two severe
northeast storms); approximately three weeks downtime, therefore, occurred
due to the weather. The first three demonstrations involved the feed hopper and
conveyor as originally supplied; since there was unfavorable control over the
sand feed rate with this unit, the results of the first three tests should be con-
sidered only in a semiquantftative sense. Except for one demonstration in which
considerable, erratic belt slippage occurred, the last 21 tests may be consider-
ed as quantitatively reliable. The detailed test results for these demonstrations
are presented in Appendix D of this report.
Demonstration 1 involved an average (but instantaneously very erratic) sand feed
rate of 32 tons/hour with a process water rate of 240 gallons/minute, a resi-
dence time of 4.2 minutes, and an aeration rate of 300 cubic feet/minute. No.
4 fuel oil was used as the contaminant in the feed sand at an average concentra-
tion of 5,620 ppm. Hie exit water and cleaned sand contained an average of 307
and 47 ppm respectively. The high feed oil concentration and sand feed rate
plus the erratic behavior of the belt feeder definitely caused a high residual oil
57
-------
concentration in the process; the residual oil in the cleaned sand was low so
that the total oil concentration in the effluent was not far from the design goal
of 150 ppm.
Demonstration 2 was an attempt to repeat the first test but with No. 6 fuel oil
as the contaminant at 3,960 ppm. The water and aeration rates were 245
gallons/minute and 300 cubic feet/minute respectively. The process water flow
plus an average (again, very erratic) sand feed rate of 33.5 tonsAour resulted
in a flotation residence time of 4,1 minutes. The exit water and cleaned sand
from the process contained an average of 40 and 14 ppm oil respectively. Even
adjusted for the relatively low feed oil concentration, these results are sub-
stantially within the performance levels which are to be expected for a heavier
oil.
Demonstration 3 was run under conditions similar to the first test except for an
increased sand feed rate of 39 tonsAour; the sand contained an average of
5,950 ppm No. 4 fuel oil. Water and aeration rates were equivalent to those
used during the first demonstration; the flotation residence time was 4. 0 min-
utes. The exit water and cleaned sand contained respectively 315 and 28 ppm oil.
This result is almost exactly the same as for the first demonstration even
though both the sand feed rate and the input oil concentration were significantly
higher. The less erratic behavior of the feed conveyor at the higher sand feed
rate explains this otherwise anomalous result.
Demonstration 4 was the first to involve very positive control of the sand feed
rate. During the first three demonstrations, problems with the belt feeder
(both in operation and performance) were so severe that modification of the
unit became quite obviously necessary; therefore, a variable speed drive was
installed in place of the constant speed drive motor originally supplied with the
unit. Nominal design conditions were selected for this demonstration but with a
feed oil concentration of 1.0% or 10,000 ppm; the actual measured No. 4 fuel
oil in the sand was an average of 8,900 ppm. Sand, water, and air feed rates
were 28.5 tonsAour, 240 gallons/minute, and 300 cubic feet/minute respec-
tively. The results of the test were 425 ppm oil in the exit water and 74 ppm in
the cleaned sand. The mobile beach cleaner was run for about 4-1/2 hours
during this demonstration; no special difficulties were noted during this ex-
tended operating period. Although the oil concentration in the exit water was
high for this test (due to the high concentration in the feed sand) the results
compare favorably on a proportional basis to the following tests. The import-
ance of the degree of aging of No. 4 fuel oil was becoming quite obvious at the
time of this test. Runs with the stationary plant had involved sand freshly
contaminated with No. 4 fuel oil in the sense that the oil had been mixed with
sand a day or two prior to the test. Ease of operation (given good weather
conditions) made it possible to contaminate sand immediately prior to a demon-
stration with the mobile unit. Effects on performance were noted for aging No.
4 fuel oil on sand for as short a period as half a day; unfortunately, the scope
of the field test program and the restraint imposed by the approach of winter
prevented the acquisition of data sufficient to quantitate the aging effect in an
exact fashion.
Demonstration 5 involved oily sand which had been aged for not quite one week;
other conditions were substantially nominal (except for the water temperature
58
-------
which was 14°C): sand feed rate of 28 tons/hour, water rate of 255 gallons/
minute, aeration rate of 300 cubic feet/minute, and feed concentration of No. 4
fuel oil of 5,740 ppm. The exit water and cleaned sand contained 180 and 76 ppm
oil respectively. Two facts had become apparent by this point in the test series:
(1) Especially for the Dam Neck beach site, and probably for most beaches, the
design water temperature was specified at too high a value; 50 to 55 °F is a
reasonable water temperature range. (2) Completely fresh oily sand is an un-
reasonable criteria since oil washed onto beaches will already be aged and
since the capacity of the mobile beach cleaner is such that, typically, oily sand
will have to be stored in piles prior to cleaning; oily wet sand aged in piles for
at least 4 days (preferably a week or more) is a more reasonable design
criteria. Since the effects of these two factors are counterbalancing (lower
Process temperatures degrade the process while sand aging improves it),, the
results of the field tests are still viewed as quite reasonable.
Pemonstration 6 was essentially a repeat of the previous test; such a repetition
is valuable for gaining at least a semiquantitative grasp of the repeatability of
both process parameters and performance. The sand feed rate was 28 tons/
hour, water rate 248 gallons/minute, aeration rate 300 cubic feet/minute, and
average No. 4 fuel oil concentration in the feed sand of 5,410 ppm. The same
sand was used as during the previous test resulting in exit water and cleaned
sand residual oil of 160 and 67 ppm respectively. The comparison of the results
°f demonstrations 5 and 6 is excellent.
Demonstration 7 involved an increased sand feed rate of 40 tonsAour with a No.
4 fuel oil concentration of 4,480 ppm. Water and aeration rates were 248
gallons/minute and 300 cubic feet/minute respectively. As expected, the
effluent stream contained increased amounts of residual oil; the cleaned sand
has 102 ppm, and the exit water 344 ppm. Although proportional adjustment of
these results does not produce very good agreement with the nominal design
specifications, very fresh oily sand was used for the test, and the process
water temperature was again low.
demonstration 8 water and aeration rates were increased by one half (to
385 gallons /minute) and one third (to 400 cubic feet/minute) respectively. Sand,
with 5,450 ppm No. 4 fuel oil, was fed to the beach cleaner at 29. 6 tons/hour.
The exit water and cleaned sand contained 228 and 97 ppm oil respectively. Pro-
Portional adjustment for the slightly high feed oil concentration and the sand feed
rate does not account for the increased residual oil concentrations. However,
Proportional adjustment for the low residence time during this test (2. 8 minutes
yersus the nominal design value of 4. 0 minutes) does account for the difference
to performance. A major conclusion from this test is mat the 300 cubic feet/
minute aeration rate (as well,, of course, as the 400 cubic feet/minute rate) was
too high and that, therefore, a lower aeration rate may, in fact, promote better
separation.
Demonstration 9 was basically a nominal run but with 3-1/2 bales of wheat
straw added to 2 drums of oil prior to sand contamination. The resulting sand
with 4,940 ppm No. 4 fuel oil was fed to the unit at 29.4 tons /hour. Aeration
and water rates were 300 cubic feet/minute and 255 gallons/minute respectively.
The exit water contained an average of 274 ppm oil which is significantly higher
than the results for previous tests under essentially identical conditions (except
59
-------
for the presence of straw). The cleaned sand contained 48 ppm oil which is
somewhat less than the results of previous tests but not enough to balance the
effect of the increased amount 6f oil in the effluent water. The following are
apparent: (1) the straw holds oil and thereby hinders oil/sand contact and mix-
ing which decreased the residual oil in the cleaned sand, (2) although much
straw is screened from the feed sand, enough enters the flotation machine to
compete for air bubble contact (note that observations indicate that straw air
bubble contact is much preferred since essentially all the straw is floated in
the first cell of the machine) which, in turn, causes an increase in the residual
oil in the exit water, and (3) the presence of straw in the sand makes the feed
operation much more difficult for the front end loader operator. In conclusion,
all other things being equal, the use of straw on a beach which is to be cleaned
up with the mobile beach cleaner is undesirable.
Demonstration 10 involved a moderate increase in the No. 4 fuel oil concentra-
tion in the feed sand to 6,410 ppm. The field engineering staff performing the
demonstrations had major difficulties with conveyor belt slippage during this
test. Unsteady sand feed resulted in additional residual oil contamination over
and above that due to the increased amount of oil in the feed sand. Sand was fed
to the unit at 30 tons/hour, water at 243 gallons/minute, and air at 300 cubic
feet/minute. The exit water and cleaned sand contained an average of 483 and
108 ppm oil respectively. Adjusted for the high feed oil concentration, this
residual oil in the sand is quite reasonable (equivalent to 84 ppm) while that in
the water is not and does not agree with the results of the previous tests. As
already mentioned, the postulate is that the erratic sand feed due to belt slipp-
age caused large pulses of oil to be sent through the froth flotation machine re-
sulting in increased average residual oil.
Demonstration 11 took place on October 15, 1971 with representatives of the
Environmental Protection Agency in attendance. The tests run were under
nominal conditions with No. 6 |uel oil as the contaminant. On request, two sand
feed rates were demonstrated. The feed sand contained an average of 5,120
ppm oil. The first test involved a sand feed rate of 30 tons/hour, a water rate
of 250 gallons /minute, and an aeration rate of 300 cubic feet/minute. The exit
water and cleaned sand contained 79 and 100 ppm oil respectively. The second
test involved a sand feed rate of 16.4 tonsAour, a water rate of 146 gallons/
minute, and an air rate of about 225 cubic feet/minute. The exit water and
cleaned sand contained 14 and 52 ppm oil respectively. The water temperature
at the field test site was still at about 14 °C on the date of this test. The resid-
ual oil levels met or exceeded the design specifications under both sets of
operating conditions although the Agency representatives seemed to feel that the
unit's performance was not completely acceptable. During the first test (with
a residence time of about 4.1 minutes) approximately 95.0% of the oil in the feed
sand was removed by the process. During the second (with a residence time
about 7.5 minutes) approximately 98.4% of the oil in the feed sand was removed.
Note that, if absolutely necessary, an even larger percentage of the oil in the
sand could have been removed by decreasing the sand feed and water rates; that
is, by putting less oil into the flotation machine per unit time residual oil con-
centrations can be reduced proportionally, and by processing the sand with a
larger residence time a larger percentage (up to some limiting value) of the
feed oil can be removed. The major conclusion from this demonstration was
that, although there may be some aesthetic reservations on the part of third
60
-------
party observers, the mobile beach cleaner quite efficiently removes oil from
contaminated sand.
Demonstration 12 involved variation of process residence time with the same
stock of feed sand (contaminated with No. 4 fuel oil at 4,287 ppm and aged for
about two weeks). Aeration was set at 250 cubic feet/minute (a 17% decrease
as compared to the previous tests) throughout the demonstration. Sand and
water feed rates were 10.2 and 83, 15 amd 125, 22.5 and 190, and 30 and 250
tonsAour and gallons/minute respectively for the four consecutive tests. The
residence times equivalent to these four sets of operating conditions are 12, 8,
5.5 and 4 minutes, and the exit water and cleaned sand contained 45 and 56, 61
and 73, 143 and 102, and 162 and 92 ppm oil respectively. With respect to the
residence times of 12, 8, 5.5, and 4 minutes, the total effluent oil concentra-
tions were 49, 65, 129, and 139 ppm. The inversely proportional relation be-
tween residence time and residual oil concentration is very good for this data.
Demonstration 13 was a repeat of demonstration 12 except that very fresh No. 4
fuel oil was used as the contaminant. For the residence times of 12, 8, 5.5,
and 4 minutes, the exit water, cleaned sand, and total effluent oil concentrations
Were 75, 59 and 69, 149, 75 and 125, 284, 122 and 230, and 405, 164 and 325 ppm
respectively. Again, the inversely proportional correlation between residence
time and effluent oil concentration is excellent. Comparison of the results of
demonstrations 12 and 13 illustrates in a very quantitative fashion the effect of
aging oily sand. On the average, the fresh oil produces about 87% more resi-
dual contamination than the oil aged for two weeks. Unfortunately, the data
for both the stationary plant and the mobile beach cleaner indicate that this
aging effect is very dependent on the particular oil being processed and cannot
be linearly interpolated or extrapolated with aging period.
Since, due to the residual oil in the water discharged from the plant, oily froth
had been observed on the beach, demonstration 14 included taking samples
along the beach. Samples were taken at 50 foot intervals from the point where
the discharge entered the surf. Sampling was very selective (perhaps overly
so) in that oily froth was carefully scraped from the surface of the beach where
waves lapped the shore; care was taken to assure that a minimum of sand and
water was picked up with the oily froth. Since "only" oily froth was collected,
the measured oil concentrations are misleading in that they do not reflect the
oil concentration on the beach but only the concentration in the froth; an area
or volume based oil concentration on the beach during a demonstration would
have been very small, indeed, due to the low concentration in the froth and
because there was so little froth on the beach that, in fact, it was quite diffi-
cult to gather an adequate sample. Sand with 3,210 ppm No. 6 fuel oil was fed
to the mobile unit at 30 tons/hour suring the test. Water and aeration rates
were 250 gallons/minute and 250 cubic feet/minute respectively. The exit
Water and cleaned sand contained 55 and 42 ppm oil respectively which is quite
an acceptable result.
Demonstration 15 was a nominal run with settling ponds used to isolate the dis-
charge stream from the process. Relatively fresh oily sand (prepared the day
before the demonstration) which contained 5,450 ppm No. 4 fuel oil was cleaned;
the sand, water, and aeration rates were 30 tons/hour, 250 gallons /minute, and
250 cubic feet/minute respectively. The exit water and cleaned sand contained
61
-------
an average of 225 and 45 ppm oil respectively. Given the low water tempera-
ture on the date of this test (about 13 °C) and the relative freshness of the oily
sand, the results are still reasonably satisfactory; the total residual oil con-
centration was about 165 ppm for a process residence time of 4. 1 minutes.
Settling ponds were found to be quite useful and practical in isolating residual
oil.
Demonstration 16 took place at night; night operation presented no special pror
blems, but the importance of proper floodlight placement was made more
explicit. Samples were again taken along the beach during this test. No. 4 fuel
oil was used as the contaminant during this test at an intermediately high level
(6,830 ppm); sand, water, and aeration rates were 30 tons/hour, 250 gallons/
minute, and 250 cubic feet/minute respectively. The exit water and cleaned
sand contained 362 and 63 ppm oil which, with the process parameters used
plus a residence time of 4.1 minutes, is about the result expected. Analysis of
the beach samples led to three relevant conclusions: (1) the amount of oil along
the beach changed only very slowly with time, (2) the oil physically on the
beach was concentrated near where the process discharge entered the surf, and
(3) the amount of oil on the beach was very small and unstable in that it did not
maintain its presence on the beach as a pollutant given even only several hours
of aging.
Demonstration 17 considered the effect of pine oil on the cleaning process. Two
levels of residence times and pine oil concentrations were used. Both these test
conditions involved No. 6 fuel oil In the contaminated sand at 6, 680 ppm. First,
the sand, water, and aeration rates were 30 tons/hour, 250 gallons/minute, and
250 cubic feet/minute respectively for a residence time of 4.1 minutes. The
first 7 samples involved no pine oil addition while the next 7 involved pine oil
addition at a rate sufficient to give a concentration of about 50 ppm; these two
conditions gave exit water and cleaned sand residual oil concentrations of 102
and 56 and 100 and 49 ppm respectively. Second, sand, water, and aeration
rates were 15 tons/hour, 125 gallons/minute, and 250 cubic feet/minute re-
spectively for a residence time of about 8.2 minutes. Again, the first 7 samples
involved no pine oil addition; the second set of samples involved a pine oil
addition rate sufficient to give a concentration of about 100 ppm. Respectively,
these fwo test conditions gave exit water and cleaned sand residual oil concentra-
tions of 47 and 54 and 45 and 52 ppm. The above results show dramatically
that the addition of pine oil has no significant, measurable effect on the separa-
tion process. That is, pine oil, a frothing agent, would have to be added to the
system at unacceptably high levels to overcome the effect of the contaminating
oil, a froth depressant.
Demonstration 18 was similar to 12 and 13 except that No. 6 fuel oil was used as
the sand contaminant; aeration was set at 250 cubic feet/minute for all 4 tests.
Sand and water feed rates of 10.2 and 83, 15 and 125, 22.5 and 190, and 30 and
250 tons/hour and gallons/minute, led to residual oil concentrations in the exit
water and cleaned sand of 7 and 10, 12, and 25, 24 and 20, and 27 and 39 ppm
respectively. These results are equivalent to total effluent oil concentrations
of 9{n_16, 23, and 31 ppm for residence times of 12, 8, 5-1/2, and 4 minutes
respectively. The results represent, again, an excellent, almost proportional,
correlation.
62
-------
Demonstration 19 was as 18 except that diesel fuel mixed with No. 6 fuel oil in a
ratio of 25:1 was used as the contaminant; the oily sand contained 4, 670 ppm oil
and aeration was again set at 250 cubic feet/minute. The residence times (equiv-
alent to the conditions of the previous demonstration) of 12, 8, 5-1/2, and 4 min-
utes led to total residual oil concentrations of about 64, 113, 122, and 190 ppm
respectively. Again, the inverse correlation implied by these results is reason-
ably good and the relative difficulty of separating lighter oils is illustrated.
Demonstration 20 involved a high concentration of diesel fuel in the feed sand
(15,200 ppm) with samples again taken along the beach. Sand, water, and air were
fed to the plant at 30 tonsAour, 250 gallons/minute, and 250 cubic feet/minute
respectively. For the resulting residence time of 4 minutes, me exit water and
cleaned sand contained 800 and 130 ppm oil respectively; on a proportional basis,
this provides a good comparison with the results of the previous demonstration.
Even with quite a dirty effluent stream, relatively little oil was found in the froth
along the beach. The main effect of the higher oil concentration in the exit water
seemed to be a greater extent of froth deposition along the beach. Again, the oil on
the beach increased only slowly with time, was more concentrated near the point
where the effluent water entered the surf, and did not persist once sand cleaning
had ceased.
Demonstration 21 was run to ascertain the effect of recycling oily water from the
oil recovery tank to the froth flotation machine. Sand, water, and aeration rates
were 30 tons/hour, 250 gallons/minute, and 250 cubic feet/minute respectively;
No. 4 fuel oil was the sand contaminant at 4,960 ppm. The first 6 samples were
taken under conditions without the water recycle and contained 208 and 75 ppm oil
(total 164 ppm) in the exit water and cleaned sand respectively. Samples 7-9 were
for sand cleaning with the water recycle and results in 220 and 49 ppm (total 163
Ppm) respectively. To bracket the results with the oily water recycle, 7 final
samples were taken without recycle resulting in exit water and cleaned sand residu-
al oil concentrations of 190 and 56 ppm (total 146 ppm). Given the standard devia-
tions in the data (about 15%), the conclusion is that recycle of oily water from the
oil recovery tank does not significantly degrade the sand cleaning operation. This
result is crucial for the long term, continuous operation of the mobile beach sand
cleaner.
Demonstration 22 involved addition of dispersants to the contaminating oil; two
different dispersants were used while three separate sets of test data were taken.
All test conditions considered no. 4 fuel oil mixed with dispersant in a ratio of
53:l prior to sand contamination. The first test (22A) used the dispersant "Cor-
exit" and sand, water, and aeration rates of 30 tonsAour, 250 gallons/minute, and
250 cubic feet/minute respectively. A feed oil concentration of 3,860 ppm led to
104 and 76 ppm in the exit water and cleaned sand respectively. This result is
acceptable. The second test (22B) involved the dispersant "Magnus. " A feed sand
oil concentration of 4,980 ppm and a residence time of about 4 minutes led to exit
water and cleaned sand oil levels of 517 and 119 ppm respectively. The conclusion
is that "Magnus" severely effects the separation process. In fact, the effluent
stream was so obviously unacceptable that the field engineer in charge decreased
the sand feed rate and increased the residence time; operation under these con-
ditions constituted test 22C. Sand, water, and aeration rates were 10.2 tons/
hour, 75 gallons/minute, and 200 cubic feet/minute respectively. A feed sand
oil concentration of 4,660 ppm and a residence time of about 13 minutes
led to 59 and 86 ppm oil in the exit water and cleaned sand respectively.
63
-------
The conclusion is that acceptable results can be attained over a wide range of
operating conditions and that the presence of a dispersant in the contaminating
oil does not necessarily mitigate against the use of the mobile beach cleaner.
Demonstration 23 took place on December 16, 1971 for representatives of the
Environmental Protection Agency. Conditions were nominal except that a
reduced sand feed rate and a settling pond were used to minimize the aesthetic
impact of the residual oil. No. 4 fuel oil was the contaminant at 5, 020 ppm in
the feed sand; sand, water, and aeration rates were 22.5 tons/hour, 195
gallons/minute, and 250 cubic feet/minute respectively for a residence time of
5.3 minutes. The temperature was 12°C, and the oil and sand were very fresh-
ly and thoroughly mixed on the morning of the day of this test. The exit water
and cleaned sand contained 130 and 28 ppm oil respectively; corrected propor-
tionally for a sand feed rate of 30 tons/hour, the same residence time, and,
therefore, a 40% by weight sand slurry (which is within the operating capability
of the froth flotation machine), the data implies residual oil concentrations of
about 37 and 127 ppm in the cleaned sand and total effluent stream respectively.
94.2% recovery of the oil was achieved during the test which is quite good con-
sidering the low processing temperature and the fact that very fresh No. 4
fuel oil was used as the contaminant.
Demonstration 24 also took place on December 16, 1971 before representatives
of the Environmental Protection Agency. Conditions were substantially the
same as for the previous demonstration except that No. 6 fuel oil was used as
the contaminant at 3,480 ppm. The exit water and cleaned sand contained 26
and 14 ppm oil respectively for a total oil recovery of 98. 0%. Corrected to a
30 ton/liour sand feed rate and an input oil concentration of 5, 000 ppm (still
with a residence time of about 5.3 minutes) this data implies residual oil in the
cleaned sand at 27 ppm and the total effluent stream at 42 ppm. Compared to
the design criteria, the performance of the mobile beach sand cleaner during
this demonstration was excellent.
3.5.5 Process Operating Costs
Figure 19 on page 66 and page 67 is a schematic of the mobile beach cleaner.
As of the date of this report, the cost for such a unit was estimated to be about
$85, 000. The major elements of this piece of equipment as labeled by the num-
bers on the drawing are as follows:
1) Motor, Froth Paddle
2) Launder
3) Discharge, Recovered Oil
4) Discharge, Cleaned Sand
5) Valve, Flush Water
6) Valve, Wash Water
7) Deck, Trailer
64
-------
8) Tank, Diesel Fuel
9) Generator, Diesel-driven Electric
10) Panelboard, Electrical
1JL) Panelboard, No. 1, Electrical Control
12) Panelboard, No. 2, Electrical Control
13) Pipe, Water Supply
14) Panel, Instrument Control
15) Blower, Pressure
16) Air-duct, Flexible
17) Air Header
18) Pulley, Drive, Flotation Cell #4
20) Motor, Drive, Flotation Cells #3 & 4
20) Pulley, Drive, Flotation Cell #3
21) Pulley, Drive, Flotation Cell #2
22) Motor, Drive, Flotation Cells #1 & 2
23) Pulley, Drive, Flotation Cell #1
24) Elevator, Conveyor Belt
25) Hopper, Sand
26) Push-bar
27) Machine, Flotation
28) Door, Elevator Clean-out
29) Hand Wheel, Process Water Control
30) Valve, Main Process Water
More unit process and equipment detail was supplied as a portion of the package
delivered to the Environmental Protection Agency with the first mobile beach
cleaner in January of 1972.
Given the cost of the basic hardware and operation of the unit for 30 full (24
hour) days per year, the following cost estimate can be made for using the unit;
65
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
RIGHT SIDE VIEW
LEFT SIDE VIEW
FIGURE 20. PHOTOGRAPHS OF
MOBILE BEACH CLEANER
68
-------
Basic Hardware
mobile unit depreciation ($85, 000 over 20 years) $ 4,250
Labor
field engineer, 36 days @ $48 1,730
field engineer, 50% field overhead 860
field engineer per diem, 33 days @ $25 820
travel for field engineer, 3 spill incidents 720
local hire field engineer, 30 days @ $60/12 hour day 1,800
local hire 30 days @ $36/12 hour day 1,080
Support Functions
front end loader rental, 30 days @ $50 1,500
elevating scraper rental, 10 days @ $100 1,000
tracked vehicle support, 6 instances 390
truck tractor transport of mobile unit 1,100
diesel fuel and miscellaneous materials 500
Maintenance
materials, 1% of total unit cost 425
labor, ditto 425
Total Yearly Cost $16,600
If the nominal sand processing rate for the 30 full days of operation is 30 tons/
hour, 21,600 tons of sand will be cleaned per year. The processing cost is,
therefore, 77 cents per ton of sand. This compares very favorably to the esti-
mated costs incurred in removing and disposing of oily sand during spill
incidents such as the one in San Francisco Bay during the spring of 1971. Note
that this comparison is very favorable even though actual spill experiences at
100 to 200 cents per ton of oily sand do not consider replacement of the sand
removed from a beach. In conclusion, the mobile beach cleaner should be use-
ful for treatment of oil contaminated sandy beaches for three primary reasons:
1) Present indications are that this proposed technique will be less
costly than those presently being used.
2) The problem (especially in the long run) of securing and maintain-
ing permanent storage for oily sand is eliminated.
3) The sand on a beach is not depleted by using the mobile beach
cleaner.
3.5.6 Conclusions
The mobile beach cleaner was found to function very well under design conditions.
The following factors, in the quantitative or qualitative fashion indicated, were
found to effect performance; all factors other than the one being considered
specifically are assumed to be held "constant. "
1) Residual oil concentrations vary in direct proportion to sand feed
69
-------
rate.
2) The more stable the sand feed rate, the cleaner are both the
cleaned sand and exit water.
3) More viscous oils result in lower residual oil levels.
4) Residual oil concentrations vary in direct proportion to feed oil
concentrations.
5) Within the limits considered during field testing, the older the
oily sand is, the lower are the residual oil concentrations in me effluent stream.
6) The more homogeneously the contaminating oil is distributed in
the sand, the cleaner the effluent stream.
7) Increasing the process water rate decreases the residual oil
concentrations.
8) Sea water promotes a more effective ofl recovery than does
brackish or fresh water.
9) Denser slurries appear to promote less efficient oil recovery.
10) Residual oil concentrations vary in inverse proportion to froth
flotation residence time.
11) Process effectiveness is a fairly complex function of aeration
rate; for the particular machine which was field tested, an aeration rate of
about 250 cubic feet/minute appeared to be an optimum.
12) Increased temperature results in increased process efficiency.
13) Surfactants or dispersants may either hinder or help oil recovery.
14) Organic solids (e.g., straw or kelp)hinder the flotation process
by competing with the ofl for attachment to air bubbles.
15) Oil deposited on wet sand is easier to recover than that deposited
on dry sand.
Given the results of the field test program, the following can be assumed to be
typical operating conditions for the mobile beach cleaner about which the above
factors operate:
30 tons/hour sand feed rate
No. 4 fuel oil as the contaminant
oil deposition on wet sand
oily sand aged for about one week
70
-------
sand as found at the Dam Neck test site without the presence of
surfactants, dispersants or organic solids, homo-
geneously mixed with the contaminating oil, and fed
steadily to the flotation machine
process sea water rate of 250 gallons/minute
process temperature of about 60°F
aeration rate of 250 cubic feet/minute
residence time (function of sand and water rates) of 4 minutes
160 ppm oil in the effluent water
75 ppm oil in the cleaned sand
130 ppm oil in the total effluent stream
71
-------
SECTION IV
ACKNOWLEDGEMENTS
The author expresses considerable appreciation for the work performed by the
project engineering staff during the course of this project. In a special sense
these individuals are co-authors of this report although the author must accept
sole responsibility for all analyses, conclusions, and recommendations. Mr.
B.C. Langley, Senior Chemical Engineer, was assigned to the project almost
from its inception; he played a major role during the demonstration plant
design and construction and as supervisor of all the field testing, both with the
stationary plant and me mobile beach cleaner. Mr. B. C. Comstock, Senior
Mechanical Design Engineer, stepped into the breach during the design and con-
struction of the mobile beach cleaner; his performance under a very tight time
schedule was outstanding. Mr. K. W. Benson, Field Engineer, saw the project
through all construction and field test work; his performance under sometimes
very adverse conditions was an inspiration to others. Mr. S. J. Rose, Junior
Chemical Engineer, assisted the author and the rest of the engineering staff
throughout the project; he assisted the author during feasibility studies and was
responsible for the oil concentration analyses throughout the project.
Mrs. K. A. Riddle not only typed this report but also coordinated the day-to-day
expenses for the field engineers throughout the project in a typically efficient
manner. Mr. L. K. Eliason, formerly Manager of Meloy Laboratories'
Environmental Sciences Division,helped the project engineers through several
crises during the course of our work. Dr. T. P. Meloy, Vice President, was
a major driving force during the inception and early stages of the project and
provided invaluable assistance in obtaining a project site. Mr. J. E. Riley,
formerly Director of Marketing for Meloy Laboratories, was the project's
most enthusiastic backer and salesman during the first one and a half years of
work.
During the course of this project Meloy Laboratories became indebted to many
parties. Thanks go to those individuals from both government and industry who
encouraged the project staff in their efforts and showed an active interest in the
project by visiting the demonstration site; such attention does much to restore
the flagging spirits of a field crew.
Special thanks go to the U.S. Navy which supplied the site for the project.
Captain Alwyn Smith, Jr., Commanding Officer of the Fleet Anti-Air Warfare
Training Center at Dam Neck, backed Meloy Laboratories continually in re*
quests for lease extensions. Lt. Commanders John Jelkes, Public Affairs
Officer, and Paul Bonham, Public Works Officer, regularly gave their support;
they also gave the project engineers much useful advice and assisted immeasur-
ably in helping to both run the project and display it to interested visitors.
Lieutenant Stevens, Lieutenant MacPherson, and Chief Petty Officer Badners of
the drone launching complex gave a helping hand on several occasions when
problems came up which necessitated outside assistance.
Appreciation is also expressed for the jobs done by the various subcontractors;
Mr. Joe D. Glenn and his staff completed their portion of the final design in a
very short time and still came across with an excellent product. The Welch
73
-------
Contracting Corp. did a superb job in putting up the plant on a two month sched-
ule; Mr. N. Crowder, Vice President, Mr. R. Saunders, Planner, and Mr.
A. Van der Reet, Foreman, were a pleasure to work with and gave valuable
assistance to the field crew not only during construction but also during the
demonstration tests. Woodington Electric Inc. also delivered the high tension
power line under short time conditions; special commendation goes for locat-
ing and obtaining appropriate transformers by a non-standard route. Wooding-
ton Electric also dismantled the demonstration plant and restored the site upon
the completion of the project. Wilbar Truck Equipment, Inc. of Alexandria,
Virginia supplied the space and expertise necessary to assemble the mobile
beach cleaner. Denver Equipment Company, General Engines Co., Inc., and
North American Engines Co., Inc. were suppliers of major unit process items.
Professor A. M. Gaudin provided a valuable input by critiquing the project in
its early stages and giving advice on the future mobilization of the process. Mr.
H. Young of the Humble Oil Company gave valuable assistance by supplying fuel
oil fbr the demonstrations; Mr. Young's cooperation was a credit to both him-
self and Humble Oil.
A multitude of thanks also goes to the many local military personnel and
merchants who gave assistance in the little ways which helped to assure the
successful completion of the project.
Finally, Mr. H. Bernard, Chief of the Office of Research and Monitoring
Environmental Protection Agency, lent continuing support to the project from
its inception. Mr. K. Jakobson of the Office of Research and Monitoring dis-
played an interest in and enthusiasm for the project which heartened the
engineering staff. Mr. R. Rhodes, Project Officer, made numerous critical
evaluations on numerous occasions of the project and was of special assistance
during the feasibility studies and in the process of selecting a site for the field
work. Of course, sincere thanks go to the Environmental Protection Agency
for funding the project which turned out to be an interesting and valuable ex-
perTence for' all concerned.
74
-------
SECTION V
APPENDICES
75
-------
APPENDIX A
LABORATORY DATA
77
-------
Table 1. Analytical Results for Series One Lab Tests*
Oil in Cleaned
Impeller Speed Aeration Rate Sand
(revolutions per minute) (liters per minute) (parts per million
1000
1000
1000
1000
1200
1200
1200
1200
1800
1800
1800
1800
2400
2400
2400
2400
*These tests were run with the Wemco laboratory test flotation machine; cash
involved a 300 gram charge of sand contaminated at 5.00% with number 4 fuel
oil and a froth flotation time of 3.5 minutes. A standard 3 liter cell was used
for each test.
1.35
3.55
6.10
8.85
1.35
4.80
8.85
13.25
1.35
8.85
16.2
19.2
1.35
6.10
11.8
26.8
by weight)
173.1
84.3
82.1
87.5
131.9
118.4
94.5
60.1
483.0
127.6
75.6
72.5
302.2
400.0
360.4
137.1
78
-------
Table 2. Analytical Results for Series Two Lab Tests
Oil in Cleaned
Aeration Rate, Maximum Sand
(liters per minute) (parts per million by weight
Sand Charge
(grams)
100
250
400
550
700
These tests were run with an impeller speed of 1200 rpm, a feed No. 4 fuel oil
concentration of 5.00%, and a 3.5 minute flotation time; again, the standard
test cell was used
12.8
12.5
11.8
11.8
10.4
49.4
107.4
128.9
130.0
82.8
Initial Oil, Concentration
(No. 4 Fuel Oil)
(percent by weight)
***
Oil in Cleaned
Aeration Rate Sand
(liters per minute) (parts per million by weight)
1
3
5
9
1
3
5
9
1
3
5
9
11.8
11.8
11.8
11.8
6.1
6.1
6.1
6.1
1.35
1.35
1.35
1.35
135.1
64.6
77.5
160.1
54.5
110.4
186.1
203.6
39.0
130.2
182.6
289.8
These tests involved an impeller speed of 1200 rpm and a 250 gram charge of
contaminated sand; again, a standard test cell and a 3.5 minute flotation time
were used.
79
-------
APPENDIX B
CONSTRUCTION DRAWINGS
81
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
APPENDIX C
PLANT DEMONSTRATION DATA
105
-------
This appendix contains the basic operating and analytical data relating to each
of the forty one tests which were performed on the demonstration pilot plant.
Relevant flow rates and general operating conditions are included for each test.
The results of the analyses for oil concentrations in the various critical process
streams are presented in total; the statistical averages for these concentra-
tions are presented as well. No attempt has been made to reproduce all data
and calculation sheets; however, this material will be supplied at the direction
of the Project Officer as Government Property. Accompanying the tabulated
operating criteria and analytical results for each demonstration are the special
conditions under which a particular test may have been run and any extraneous
circumstances which are believed to have significance in the consideration of
the data.
Demonstration Number 1
Sand feed rate: 30 TPH Screen spray rate: 20 GPM
Total water feed rate: 450 GPM Launder wash rate: none
Water rate to attrition scrubber: 110 GPM Water rate to horizontal pump:none
Water rate to vertical pump: 100 GPM Rate of pine oil addition: none
Water rate to flotation machine: 220 GPM Oil type: No. 4 fuel oil
Aeration rate: 280 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average n/a n/a 133 286
Standard
Deviation n/a n/a 5% 39%
A "standard run" with a concentration of No. 4 fuel oil of about 0.5%. The term
"standard run" is used to imply a relatively high water rate (about 500 gallons
per minute) and a nominal sand feed rate (25 to 30 tons per hour). Within the
category standard run there is still considerable opportunity for variation in
operating conditions; where applicable, these variations are listed as special
conditions.
Demonstration Number 2
Sand feed rate: 23 TPH Screen spray rate: 20 GPM
Total water feed rate: 445 GPM Launder wash rate: none
Water rate to attrition scrubber: 110GPM Water rate to horizontal pumpmone
Water rate to vertical pump: 100 GPM Rate of pine oil addition: approxi-
Water rate to flotation machine: 215GPM mately 1 cc/minute
Aeration rate: 280 CFM No. 6 fuel oil
Analytical Results
(parts per million by weight oil)
106
-------
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 3,768 measured 147 232
Standard
Deviation 36% 20% 10%
A standard run with No. 6 or Bunker C fuel oil at 0.5%
Demonstration Number 3
Sand feed rate: 19 TPH Screen spray rate: 80 GPM
Total water feed rate: 450 GPM Launder wash rate: none
Water rate to attrition scrubber: 115 GPM Water rate to horizontal pump: none
Water rate to vertical pump: 40 GPM Rate of pine oil addition: approxi-
Water rate to flotation machine: 215 GPM mately 1 cc/minute
Aeration rate: 280 CFM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2
-------
Sand feed rate: 30 TPH
Total water feed rate: 560 GPM
Water rate to attrition scrubber: 137 GPM
Water rate to vertical pump: 66 GPM
Water rate to flotation machine: 277 GPM
Aeration rate: 280 CFM
Screen spray rate: 80 GPM
Launder wash rate: approximately
30 GPM
Water rate to horizontal pump: none
Rate of pine oil addition: approxi-
mately 1 cc/minute
Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Sand Feed Sand
(prior to con-
tamination)
Average
Standard
Deviation
Feed Water Cleaned Sand Exit
(H_O saturated) Water
537
23 8,361 1,186 711
216% 42% 24% 30%
A low sand feed rate with a high level of No. 4 fuel oil contamination (about 3%).
Demonstration Number 6
Sand feed fate: 20 TPH
Total water feed rate: 320 GPM
Water rate to attrition scrubber: 50 GPM
Water rate to vertical pump: 54 GPM
Water rate to flotation machine: 156 GPM
Aeration rate: 280 CFM
Screen spray rate: none
Launder wash rate: none
Water rate to horizontal pump: none
Rate of pine oil addition: approxi-
mately 1 cc/minute
Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Average
Standard
Deviation
Feed Sand
4,088
36%
Feed Water
251
115%
Cleaned Sand
(H2O saturated)
329
29%
Exit Water
1,011
30%
High residence times for both scrubbing and flotation with No. 4 fuel oil at
about 0.5%.
Demonstration Number 7
Sand feed rate: 25 TPH
Total water feed rate: 550 GPM
(546 final)
Water rate to attrition scrubber: 160 GPM
(166 final)
Water rate to vertical pump: 80 GPM
(0 final)
Water rate to flotation machine: 230 GPM
Screen spray rate: 80 GPM (53
final)
Launder wash rate: none
Water rate to horizontal pump: none
Rate of pine oil addition: approxi-
mately 1 cc/minute
Oil type: No. 4 fuel oil
108
-------
Aeration rate: 280 CFM
Analytical Results
(parts per million by weight oil)
Average
Standard
Deviation
Feed Sand
5,024
36%
Feed Water Cleaned Sand Exit Water
(H-O saturated)
628 140 749
37%
19%
23%
A standard run with No. 4 fuel oil at about 0.5% in sand which had been aged in
piles for 15 days.
Demonstration Number 8
Sand feed rate: 19 TPH
Total water feed rate: 250 GPM
Water rate to attrition scrubber: 215 GPM
Water rate to vertical pump: none
Water rate to flotation machine: none
Aeration rate: 280 (Test A), 210 (TestB),
140 (Test C), and 70 (Test D) CFM
Screen spray rate: 35 GPM
Launder wash rate: none
Water rate to horizontal pump:
140 GPM
Rate of pine oil addition: approxi-
mately 1 cc/minute
Oil types: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
A) Average
Standard
Deviation
B) Average
Standard
Deviation
C) Average
Standard
Deviation
D) Average
Standard
Deviation
Feed Sand
5,633
Feed Water
296
Cleaned Sand Exit Water
(H2O saturated)
3
4,696
16%
5,480
57%
4,628
26%
322
19%
257
24%
232
155
29%
214
33%
159
35%
168
14%
477
33%
529
29%
540
11%
669
23%
A standard run using only the first through fourth flotation cells to obtain kinetic
data from a single, continuous demonstration.
Demonstration Number 9
Sand feed rate: 30 TPH Screen spray rate: 35 CFM
Total water feed rate: 250 GPM Launder wash rate: none
Water rate to attrition scrubber: 215 GPM Water rate to horizontal pumpi
Water rate to vertical pump: none HO GPM
109
-------
Water rate to flotation machine: none Rate of pine oil addition: approxi-
Aeration rate: 210 CFM mately 1 cc/minute
Oil type: No. 4 fuel oil
/
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 3,589 260 167 349
Standard
Deviation 34% 36% 36% 22%
A standard run using only the first four flotation cells with a low scrubber
residence time and a high, flotation residence time.
Demonstration Number 10
Sand feed rate: 19 TPH Screen spray rate: none
Total water feed rate: 440 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump:
Water rate to vertical pump: 235 GPM 140 GPM
Water rate to flotation machine: 205 GPM Rate of pine oil addition: approxi-
Aeration rate: 280 CFM mately 1 cc/minute
Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 7,020 273 106 496
Standard
Deviation 34% 26% 14% 40%
A standard run with a low sand feed rate and a scrubber bypass: the initial
attempt at simulating the operation of a mobile beach cleaner.
Demonstration Number 11
Sand feed rate: approximately 60 TPH Screen spray rate: 30 GPM
Total water feed rate: 395 GPM Launder wash rate: none
Water rate to attrition scrubber: 130 GPM Water rate to horizontal pump:
Water rate to vertical pump: 50 GPM 150 GPM
Water rate to flotation machine: 185 GPM Rate of pine oil addition: none
Aeration rate: 280 CFM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
110
-------
Feed Sand Feed Water Cleaned Sand Exit Water
(H0O saturated)
£*
Average 5,112 1,992 389 2,141
Standard
Deviation 49% 43% 35% 24%
A standard run with a very high sand feed rate (about 60 tons per hour).
Demonstration Number 12
Sand feed rate: approximately 8 TPH Screen spray rate: 50 GPM
Total water feed rate: 320 GPM Launder wash rate: none
Water rate to attrition scrubber: 200 GPM Water rate to horizontal pump:
Water rate to vertical pump: 70 GPM 200 GPM
Water rate to flotation machine: none Rate of pine oil addition: none
Aeration rate: 280 CFM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(HgO saturated)
Average 16,980 1,200 392 1,678
Standard
Deviation 20% 37% 23% 37%
A run with low scrubber residence time, high flotation residence time, very
low sand feed rate (about 7 tons per hour), and an increased oil contamination
level (better than 1.5% of No. 4 fuel oil).
Demonstration Number 13
Sand feed rate: 30 TPH Screen spray rate: 50 GPM
Total water feed rate: 305 GPM Launder wash rate: none
Water rate to attrition scrubber: 165 GPM Water rate to horizontal pump:
Water rate to vertical pump: 55 GPM 165 GPM
Water rate to flotation machine: 35 GPM Rate of pine oil addition: none
Aeration rate: 280 CFM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 8,236 296 137 353
Standard
Deviation 40% 44% 23% 17%
A standard run with sea water rather than the brackish water from the well-
point system which had been used previously; the contaminating oil was No^ 4
111
-------
fuel oil at about 0.8%.
Demonstration Number 14
Sand feed rate: 17.5 TPH Screen spray rate: none
Total water feed rate: 280 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 280 GPM Oil type: No. 4 fuel oil
Aeration rate: 350 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 5,290 N/A 115 -276
Standard
Deviation 17% 25%
Demonstrations 14 through 20 were mobile unit simulation studies. Contamin-
ated sand slurry entered the flotation machine directly at its feed box. Process
water was not recycled and cleaned water-sand slurry was sampled directly
from the discharge box of the flotation machine. Essentially nominal conditions
except for the sand feed rate which was set at 17.5 tons/hour. Sand was fed to
the system using a leased, portable belt feeder, and water was obtained from
the ocean using a leased submersible pump.
Demonstration Number 15
Sand feed rate: 24.5 TPH Screen spray rate: none
Total water feed rate: 250 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 250 GPM Oil type: No. 4 fuel oil
Aeration rate: none
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 6,130 N/A 157 280
Standard
Deviation 8% 15% 14%
A mobile unit simulation, as 14, but with a sand feed rate of 24.5 tons/hour.
Demonstration Number 16
Sand feed rate: 22.5 TPH Screen spray rate: none
Total water feed rate: 250 GPM Launder wash rate: none
112
-------
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to notation machine: 250 GPM Oil type: No. 4 fuel oil
Aeration rate: 350 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 6,680 N/A 232 260
Standard
Deviation 19% 19% 7.'
A mobile unit simulation, as 14 and 15, but. with a sand feed of 22.5 tonsAour.
Demonstration Number 17
Sand feed rate: 27. 0 TPH 'Screen spray rate: none
Total water feed rate: 125 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 125 GPM Oil type: No. 4 fuel oil
Aeration rate: none
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 5,230 N/A 138 425
Standard
Deviation 23% 17% 13%
A mobile unit simulation with a sand feed rate of 27. 0 tonsAour and a water
feed rate of half that of the previous tests or 125 gallons/minute (i. e., the
residence time in the flotation machine was considerably higher than for the
previous tests).
Demonstration Number 18
Sand feed rate: 26 TPH Screen spray rate: none
Total water feed rate: 116 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 116 GPM Oil type: No. 4 fuel oil
Aeration rate: 350 CFM
Analytical Results
(parts per million by weight oil)
113
-------
Feed Sand Feed Water Cleaned Sand Exit Water
(H0O saturated)
£»
Average 6,930 N/A 183 524
Standard
Deviation 35% 23% 15%
Similar to 17 with an intentionally increased oil concentration (from 5230 to
6930 parts/million).
Demonstration Number 19
Sand feed rate: 26 TPH Screen spray rate: none
Total water feed rate: 206 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate toflotation machine: 206 GPM Oil type: No. 6 fuel oil
Aeration rate: none
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(HgO saturated)
Average 5,730 N/A 92 100
Standard
Deviation 22% 43% 22%
A nominal mobile unit simulation with No. 6 fuel oil used in place of the No. 4
of the previous tests.
Demonstration Number 20
Sand feed rate: 13 TPH Screen spray rate: none
Total water feed rate: 200 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 200 GPM Oil type: No. 4 fuel oil
Aeration rate: 500 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(HO saturated)
Average 19,810 N/A 181 589
Standard
Deviation 6,2% 16% 24%
A nominal mobile unit simulation with a very high feed oil concentration (about
2%) and a, therefore, reduced sand feed rate (13 tons/hour).
114
-------
Demonstration Number 21
Sand feed rate: 30 TPH Screen spray rate: N/A
Total water feed rate: 370 GPM Launder wash rate: none
Water rate to attrition scrubber: 150 GPM Water rate to horizontal pump:
Water rate to vertical pump: 105 GPM 145 GPM
Water rate to flotation machine: 115 GPM Rate of pine oil addition: none
Aeration rate: 280 CFM Oil type: medium crude oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(HgO saturated)
Average 9,550 350 185 517
Standard
Deviation 12.9% 41% 36.2% 34. i
A nominal run using the entire demonstration plant loop. A medium crude oil
was used as the contaminating medium at 1% concentration.
Demonstration Number 22
Sand feed rate: 30 TPH Screen spray rate: none
Total water feed rate: 360 GPM Launder wash rate: none
Water rate to attrition scrubber: 190GPM Water rate to horizontal pump:
Water rate to vertical pump: 95 GPM 145 GPM
Water rate to flotation machine: 75 GPM Rate of pine oil addition: none
Aeration rate: none Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H2O saturated)
Average 6,920 1,970 243 1,780
Standard
Deviation 8.2% 42% 20% 15.8%
A nominal run using the entire plant with an increased aeration rate to the
flotation machine (350 cubic feet/minute).
Demonstration Numbers 23. 24 and 25
The purpose of these tests was to determine whether or not oil could be scrubbed
from straw to facilitate disposal of the latter material. Straw saturated with
No. 6 fuel oil was attrition scrubbed for various lengths of time, under various
conditions. Liberation of oil into the carrier water was determined by analyz-
ing for the oil.
Demonstration number 23 involved scrubbing for 6 minutes and resulted in 34
Parts per million oil in the exit water.
115
-------
Demonstration 24 involved scrubbing for 12 minutes and resulted in 40 parts
per million oil in the exit water.
5 gallons of diesel fuel was added to the slurry in the attrition scrubber for
demonstration number 25; after 12 minutes of scrubbing, the exit water con-
tained 64 parts per million of oil. Addition of 5 more gallons of diesel fuel and
scrubbing for 12 more minutes resulted in an exit concentration of 75 parts per
million.
Although this data for the dispersion of oil into the process water moves in the
right direction for these tests, the results are hardly favorable. More impor-
tant are me extreme operating difficulties which the field engineering crew had
with performing this task. The straw/oil mass tended to wrap around the
impeller shaft in cell number 1 of the scrubber and was then extremely difficult
to remove. This problem is similar to that observed around the impellers of
the flotation machine; in this latter case, however, the air diffusing from the
area of the impeller served to keep straw away, and, therefore, when such a
straw buildup occurs, it occurs very, very slowly. The conclusion from these
tests is that the recovery of oil contaminated straw by attrition scrubbing is not
feasible.
Demonstration Number 26
Sand feed rate: none Screen spray rate: none
Total water feed rate: 485 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 485 GPM Oil type & rate: No. 4 fuel oil at
Aeration rate: 280 CFM 2.5 GPM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(HO saturated)
£t
Average N/A 897
Standard
Deviation 7.
Demonstrations 26 through 36 were for oil/water separation using the standard
froth flotation equipment of the demonstration plant. The first test involved a
water feed of about 500 gallons/minute with an oil concentration of 1/2% and an
aeration rate of 280 cubic feet/minute. Oil was added at the vertical sump
pump.
Demonstration Number 27
Sand feed rate: none Screen spray rate: none
Total water feed rate: 485 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump:none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 485 GPM Oil type and rate: No. 4 fuel oil
116
-------
Aeration rate: 560 CFM at 2.5 GPM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (H0O saturated)
z
Average 803
Standard
Deviation 3.7%
A repeat of 26 with double the aeration rate.
Demonstration Number 28
Sand feed rate: none Screen spray rate: none
Total water feed rate: 455 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 455GPM Oil type and rate: 1:1 No. 4 fuel
Aeration rate: 560 CFM oil-Diesel 260 at 2.5 GPM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (HO saturated)
Average 1,370
Standard
Deviation 2,9%
As 27 with a 1:1 mixture of No. 4 fuel oil and diesel fuel used as the feed con-
taminant rather than straight No. 4.
Demonstration Number 29
Sand feed rate: none Screen spray rate: none
Total water feed rate: 465 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump:none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 465GPM Oil type: No. 6 fuel oil
Aeration rate: 560 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H0O saturated)
£i
Average 1.96
Standard
Deviation 49.5%
117
-------
Three bales of straw were thoroughly mixed with 1/4 drum of No. 6 fuel oil.
This oily straw was then fed by hand to the flotation machine over a period of
10 minutes with a water rate of 465 gallons /minute and an aeration rate of 560
cubic feet/minute. The straw was floated very readily and little oil was re-
leased to the water.
Demonstration Number 30
Sand feed rate: none Screen spray rate: none
Total water feed rate: 515 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 515 GPM Oil type: 2.5 GPM No. 4 fuel oil
Aeration rate: 560 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated (Ho
Average 2,870
Standard
Deviation 3 .
The same as 27 except that the oil was added directly to the feed box of the
flotation machine rather than at the vertical pump. Channeling within the
flotation cells was very evident.
Demonstration Number 31
Sand feed rate: none Screen spr*ay rate: none
Total water feed rate: 515 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pumpmone
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 515 GPM Oil type: No. 4 fuel oil at 2.5 GPM
Aeration rate: 560 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (H0O saturated)
£i
Average (first sample set, 1-7) 3,350
Standard Deviation 4.1%
Average (second sample set, 8-14) 3,430
Standard Deviation 6.
Average (third sample set, 15-23) 3,110
Standard Deviation 3.
Average (fourth sample set, 24-31) 2,560
118
-------
(total 2,990)
Standard Deviation 11.7%
(total 12.1%)
Similar to 30 except that aeration to the downstream cells was cut off sequen-
tially during the test. Little parallel variation in performance of the total
system was noted which implies that most of the effective separation occurs in
the first one or two cells of the froth flotation machine.
Demonstration Number 32
Sand feed rate: none Screen spray rate: none
Total water feed rate: 515 GPM Launder wash rate: none
Water rate to attrition scrubber: 120 GPM Water rate to horizontal pump: none
Water rate to vertical pump: 110 GPM Rate of pine oil addition: none
Water rate to flotation machine: 285 GPM Oil type: No. 4 fuel oil
Aeration rate: 560 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (HgO saturated)
Average 1,340
Standard Deviation 9.!
An emulsion of oil in water was prepared as the feed contaminant by scrubbing.
This emulsion plus additional water was delivered to the flotation machine by
the vertical pump. The equivalent oil concentration in the feed to the flotation
machine was about 2%. The water rate to the flotation machine was 515 gallons/
minute and the aeration rate 560 cubic feet/minute.
Demonstration Number 33
Sand feed rate: none Screen spray rate: none
Total water feed rate: 250 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
'Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 250 GPM Oil type and rate: No. 4 fuel oil
Aeration rate: 560 CFM at 1.25 GPM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (HgO saturated)
Average 2,790
standard Deviation 9.3%
separation test with an increased residence time relative to the
Previous tests (250 gallons/minute of water) with a feed oil concentration of
119
-------
1/2% and an aeration rate of 560 cubic feet/minute.
Demonstration Number 34
Sand feed rate: none Screen spray rate: none
Total water feed rate: 515 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 515 GPM Oil type: No. 4 fuel oil at 25 GPM
Aeration rate: 560 CFM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (H~O saturated)
Average 22,100
Standard Deviation 10.5%
A very high feed oil concentration (approximately 5%) with a water rate of 515
gallons/minute and an aeration rate of 560 cubic feet/minute.
Demonstration Number 35
Sand feed rate: none Screen spray rate: none
Total water feed rate: 515 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 515 GPM Oil type and rate: No. 6 fuel oil
Aeration rate: 560 CFM . at 2.5 GPM
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (Ho° saturateti)
Average 517
Standard Deviation 11.1%
Similar to the previous demonstration but with No. 6 fuel oil as the contamin-
ant at a concentration of about 1/2%.
Demonstration Number 36
Sand feed rate: none Screen spray rate: none
Total water feed rate: 1,050 GPM Launder wash rate: none
Water rate to attrition scrubber: none Water rate to horizontal pump: none
Water rate to vertical pump: none Rate of pine oil addition: none
Water rate to flotation machine: 1,050GPM Oil type and rate: No. 4 fuel oil
Aeration rate: 560 CFM at 5 GPM
Analytical Results
120
-------
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(calculated) (HO saturated)
Average A 3,160
Standard Deviation 25.
The last of the oil/water separation tests. A very low residence time was used
due to a water flow of 1050 gallons/minute. No. 4 fuel oil at 1/2% was used as
the contaminant with an aeration rate of 560 cubic feet/minute.
Demonstration Number 37
Sand feed rate: 30 TPH Screen spray rate: none
Total water feed rate: 537 GPM Launder wash rate: none
Water rate to attrition scrubber: 205 GPM Water rate to horizontal pump:
Water rate to vertical pump: 82 GPM 250 GPM
Water rate to flotation machine: none Rate of pine oil addition: none
Aeration rate: none Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H0O saturated)
£.t
Average 4,500 1,127 278 1,540
Standard
Deviation 17.0% 46.1% 11.8% 22.
The first in a series of four scrubbing and dewatering tests. Sand with about
1/2% No. 4 fuel oil was fed to the attrition scrubber at 30 tonsAour; water
was added to the scrubber at 205 gallons/minute. The exit sand was pumped
to the horizontal pump and then to the dewatering cyclone. Total water flow to
the cyclone was 537 gallons/minute.
Demonstration Number 38
Sand feed rate: 30 TPH Screen spray rate: none
Total water feed rate: 477 GPM Launder wash rate: none
Water rate to attrition scrubber: 67 GPM Water rate to horizontal pump:
Water rate to vertical pump: 130 GPM 280 GPM
Water rate to flotation machine: none Rate of pine oil addition: none
Aeration rate: none Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(HO saturated)
Average 3,860 1,280 271 1,710
Standard
Deviation 44.7% 23.4% 9.1% 28.5%
121
-------
Similar to 37 but with an increased scrubbing time of about 7 minutes obtained
by decreasing the scrubber's feed water rate to 67 gallons/minute.
Demonstration Number 39
Sand feed rate: 30 TPH
Total water feed rate: 517 GPM
Water rate to attrition scrubber: 107 GPM
Water rate to vertical pump: 120 GPM
Water rate to flotation machine: none
Aeration rate: none
Screen spray rate: none
Launder wash rate: none
Water rate to horizontal pump:
290 GPM
Rate of pine oil addition: none
Oil type: No. 6 fuel oil
Average
Standard
Deviation
Feed Sand
4,120
3,4%
Analytical Results
(parts per million by weight oil)
Feed Water Cleaned Sand
(HO saturated)
625 272
47.5%
25.
Exit Water
829
24.
Similar to the previous test but with No. 6 fuel oil as the contaminant.
Demonstration Number 40
Sand feed rate: 30 TPH
Total water feed rate: 473 GPM
Water rate to attrition scrubber: 88 GPM
Water rate to vertical pump: 135 GPM
Water rate to flotation machine: none
Aeration rate: none
Screen spray rate: none
Launder wash rate: none
Water rate to horizontal pump:
250 GPM
Rate of pine oil addition: none
Oil type: No. 4 fuel oil
Average
Standard
Deviation
Feed Sand
19,600
17.
Analytical Results
(parts per million by weight oil)
Feed Water Cleaned Sand
(HO saturated)
3,870 890
47.
19.
Exit Water
4,700
20.4%
The last of the scrubbing and dewatering tests.
feed oil concentration was about 2%.
A repeat of 38 except that the
Demonstration Number 41
Sand feed rate: 30 TPH
Total water feed rate: 330 GPM
Water rate to attrition scrubber: 170 GPM
Water rate to vertical pump: 110 GPM
Water rate to flotation machine: 50 GPM
Screen spray rate: none
Launder wash rate: none
Water rate to horizontal pump:
145 GPM
Rate of pine oil addition: none
122
-------
Aeration rate: 280 CFM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Feed Water Cleaned Sand Exit Water
(H_O saturated)
z
Average 4,750 2,050 703 2,460
Standard
Deviation 37.5% 28.3% 26.6% 27.6%
A 20 hour long operational run. Day laborers w,ere used to augment the field
staff. Water was recycled through the plant, and the process parameters were
maintained at essentially nominal values. Adequate lighting and stockpiled feed
sand were crucial for this demonstration.
123
-------
APPENDIX D
MOBILE BEACH CLEANER DEMONSTRATION DATA
125
-------
The following pages contain the results of 24 sand cleaning demonstrations using
the mobile beach cleaner which was constructed during the spring and summer
of 1971. The format for the data given below is essentially the same as that of
Appendix C. Again, no attempt has been made to reproduce all the data and
calculations, but these are available and can be submitted in their raw state as
Government property. Immediately below is a table of contents for the test
data plus a general, relatively qualitative description of each demonstration.
Demonstration Number 1
Sand feed rate: 32 TPH Aeration rate: 300 CFM
Water feed rate: 240 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2
-------
Feed Sand
Average 5,950
Standard Deviation 7.7%
Exit Water
315
15.1%
Cleaned Sand
(HO saturated)
28
47.
Similar to the first demonstration but with a higher sand feed rate, 39 tons/
hour. Sand feed was more reliable and less erratic at the higher feed rate;
this was reflected bybetter removal of the oil from the feed slurry.
Demonstration Number 4
Sand feed rate: 28.5 TPH
Water feed rate: 240 GPM
Aeration rate: 300 CFM
Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water
Average 8,900
Standard Deviation 9,
Cleaned Sand
(H0O saturated)
z
74
425
30.4%
Took place about 3-1/2 weeks after the third demonstration; the drive for the
feed belt had been replaced with a more positive and reliable unit. The feed oil
concentration was high at about 0.9%. Aeration was again 300 cubic feet/
minute which eventually was proved to be too high.
Demonstration Number 5
Sand feed rate: 28. 5 TPH
Water feed rate: 255 GPM
Aeration rate: 300 CFM
Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water
Average 5,740
Standard Deviation 5.2%
180
44.
Cleaned Sand
(H0O saturated)
Ci
76
54.3%
Essentially a nominal run with No. 4 fuel oil. Although the water temperature
was low at 14°C, the performance of the beach cleaner (adjusted) was sub-
stantially within the design criteria. The oily sand had aged for approximately
°ne week prior to this test.
Demonstration Number 6
Sand feed rate: 28 TPH
Water feed rate: 248 GPM
Aeration rate: 300 CFM
Oil type: No. 4 fuel oil
Analytical Results
127
-------
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average 5,410 160 67
Standard Deviation 9.9% 29.2% 49.0%
A repeat of the previous test with sand freshly contaminated with No. 4 fuel oil.
Results were essentially identical to the previous test proving that the modified
system was acting reliably.
Demonstration Number 7
Sand feed rate: 40 TPH Aeration rate: 300 CFM
Water feed rate: 248 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H0O saturated)
£i
Average 4,480 344 102
Standard Deviation 8.9% 13.7% 19.8%
The same as the previous run but with a sand feed rate of 40 tons/hour.
Demonstration Number 8
Sand feed rate: 29. 6 TPH Aeration rate: 400 CFM
Water feed rate: 385 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average 5,450 228 97
Standard Deviation 6.0% 12.9% 23.!
As demonstration 6 but with water and aeration rates increased to 385 gallons/
minute and 400 cubic feet/minute respectively. Performance was degraded
significantly.
Demonstration Number 9
Sand feed rate: 29.4 TPH Aeration rate: 300 CFM
Water feed rate: 255 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(HO saturated)
128
-------
Average 4,940 274 48
Standard Deviation 4.7% 25.4% 38.3%
Contaminating oil was prepared by mixing 3-1/2 bales of wheat straw with 2
drums of No. 4 fuel oil. Other system's parameters were set at nominal
values. Wim straw present, the front end loader operator had trouble main-
taining the sand feed to the unit and overall performance was degraded some-
what.
Demonstration Number 10
Sand feed rate: 30 TPH Aeration rate: 300 CFM
Water feed rate: 243 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average 6,410 483 108
Standard Deviation 8.8% 48.2% 32.1
oil concentration was set at 0.64% with other conditions nominal. Due to
oil which had by this time worked its way into the system, there was trouble
belt slippage. The resulting erratic feed caused a degradation in system
Performance.
Demonstration Number 11
Sand feed rate: 30 TPH (A) 16.4 TPH (B)
Water feed rate: 250 GPM (A) 146 GPM (B)
Aeration rate: 300 CFM (A) 225-250 CFM (B)
Oil type: No. 6 fuel oil (A and B)
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average - (A) 79 100
Standard Deviation 11.9% 19.5%
Average - (B) 5,120 (total) 14 52
Standard Deviation 4.8% (total) 27.4% 20.9%
Representatives of the Environmental Protection Agency witnessed this nominal
run with No. 6 fuel oil. At the request of these representatives sand was fed to
toe unit first at 30 tonsAour and then at 16.4 tons/hour. Although the water
temperature was low, the unit performed substantially within the design criteria.
Demonstration Number 12
129
-------
Sand feed rate: 10.2 (1-7), 15 (8-14), 22.5 (15-21), 30 (22-28) TPH
Water feed rate: 83 (1-7), 125 (8-14), 190 (15-21), 250 (22-28) GPM
Aeration rate: 250 CFM
OH type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand
Average (1-7)
Standard Deviation
.Average (8-14)
Standard Deviation (8-14)
Average (15-21)
Standard Deviation (15-21)
:it Water
45
20.0%
61
11.5%
143
15.5%
Cleaned Sand
(H_O saturated)
Lt
56
37.
Average (22-28)
Standard
Deviation (22-28)
4,287 (total)
22.1% (total)
162
8.
73
21.9%
102
31.
92
15.
Sand feed rates and water rates were varied during this run: 10.2, 15, 22.5,
and 30 tonsAour and 83, 125, 190, and 250 gallons/minute respectively and
concurrently. Aeration was lowered to 250 cubic feet/minute and the feed oil
concentration was somewhat low. The oily sand had been aged for about 2 weeks
and, of course, the water temperature was still low. The relation between
residence times and effluent oil concentrations was shown to be quite consistent
by this demonstration.
Demonstration Number 13
Sand feed rate: 10.2(1-8), 15(9-16), 22.5(17.24), 30 (25-32) TPH
Water feed rate: 83 (1-8), 125 (9-16), 190 (17-24), 250 (25-32) GPM
Aeration rate: 250 CFM
Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand
Water
Average (1-8)
Standard Deviation (1-8)
Average (9-16)
Standard Deviation (9-16)
Average (17-24)
75
37.3%
149
29.5%
284
Cleaned Sand
(H2O saturated)
59
52.5%
75
24.0%
122
130
-------
Standard Deviation (17 -24) 9.9% 26.2%
Average (25-32) 5,740 (total) 405 164
Standard
Deviation (25-32) 11.9% (total) 5.4% 25.
Essentially a repeat of 14 but using sand freshly contaminated with No. 4 fuel
oil. Water temperatures were again low but still a good correlation was ob-
tained between residence time and effluent oil concentration. The use of a
settling pond to isolate the effluent was demonstrated, and once again the effect
of aging No. 4 fuel oil was observed.
Demonstration Number 14
Sand feed rate: 30 TPH Aeration rate: 250 CFM
Water feed rate: 250 GPM Oil type: No. 6 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(HO saturated)
Average 3,210 55 42
Standard Deviation 20.3% 22.7% 26.1%
Analytical Results of Beach Samples
(parts per million by weight oil)
Samples bv Sets Samples by Position
Sample No. Beach Sand Sample No. Beach Sand
Average (3S) 240 Average (SI) 272
Standard Deviation 87.9% Standard Deviation 76.5%
„_ (4S) 18 Average (S2) 21
Standard Deviation 113% Standard Deviation 107%
Average (5S) 197 Average (S3) 87
Standard Deviation 160% Standard Deviation 210%
Average <6S) 150
Standard Deviation 129%
Average (7S) 39
Standard Deviation 110%
e -=- (8S) 164
Standard Deviation 143%
Demonstration 14 was the first during which samples were taken along the beach.
"phis test lasted for about 2-1/2 hours and involved No. 6 fuel oil as the con-
taminant at a somewhat reduced concentration with other process variables at
their nominal values; the results of the cleaning process itself were, as usual,
131
-------
quite good. Due to the prevailing current along the beach (southwards) only
those samples collected south of the plant's discharge contained oil. The code
numbers for the beach samples refer to the following: the "S" indicates south
of the discharge point; the first number refers to the time (multiples of 10
minutes after processing had begun; and the second number refers to the dis-
tance ( in multiples of 50 feet) from the discharge point on the beach. Beach
samples were taken in a very (perhaps overly) selective manner; from an area
of about two feet in diameter centering around the indicated distance from the
discharge point, a sample was scraped from the beach surface at the point where
waves were lapping material onto the sand; great care was taken to assure that
only oily residue was picked up. The major problem with this sampling tech-
nique was the difficulty in obtaining enough material for analysis. The major
advantage was, of course, that the individual doing the sampling was gathering
only material which (visually) appeared to be unacceptable. Average oil con-
centrations over all positions at the same point in time (the left hand column
above) indicate little increase in the oil present on the beach over approximate-
ly a one hour time span and at the positions samples; in fact, the standard
deviations involved make it difficult to say that any increase occurred at all.
Averages at the same position over all the sampling times indicate a decrease
in the oil present on the beach farther from the discharge point; this is as ex-
pected although, except for the sample position right next to the discharge
point, there was really not much oil on the beach (as the analyses indicate). In
making these conclusions, it must be emphasized again that samples were taken
in the very selective manner described above and that in no way do the respec-
tive oil contamination levels represent typical oil concentrations for all the
sand along the beach during a demonstration.
Demonstration Number 15
Sand feed rate: 30 TPH Water feed rate: 250 CFM
Water feed rate: 250 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(HLO saturated)
Lt
Average 5,450 225 45
Standard
Deviation 32.3% 25.1% 54%
A nominal run with No. 4 fuel oil. Dual settling ponds were used and proved
feasible although in a long term continuous operation such an approach to
isolating the effluent would entail the use of additional front end loaders.
Demonstration Number 16
Sand feed rate: 30 TPH Aeration rate: 250 CFM
Water feed rate: 250 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
132
-------
Average
Standard
Deviation
Feed Sand
6,830
6.5%
Exit Water
362
47.
Cleaned Sand
(H2O saturated)
63
37.
Analytical Results of Beach Samples
(parts per million by weight oil)
Samples by Sets Samples by Position
Sample No.
Average (IS)
Standard Deviation
Average (2S;
Standard Deviation
Average (3S)
Standard Deviation
Average (4S)
Standard Deviation
Average (5S)
Standard Deviation
Beach Sand Sample No.
666
94.
211
48.
674
132%
203
46.2%
444
137%
Average (SI)
Standard Deviation
Average (S2)
Standard Deviation
Average (S3)
Standard Deviation
Average (S4)
Standard Deviation
Average (S6)
Standard Deviation
Average (S8)
Standard Deviation
Average
-------
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average (1-7) 102 56
Standard Deviation 24.9% 20.56%
Average (8-14) 100 49
Standard Deviation 11.5% 15.2%
Average (15-21) 47 54
Standard Deviation 13.6% 33.
Average (16-28) 6,680 45 52
Standard
Deviation 17.6% 28.9% 27.5%
A run with No. 6 fuel oil at an, inadvertently, somewhat high feed concentration.
Sand feed rate was changed from 30 to 15 tons Aour and water rate from 250 to
125,gallons/minute halfway through the test. During the second and fourth sets
of data presented above pine oil was metered into the feed box of the flotation
machine at about 60 cc/minute; this rate was sufficient to produce pine oil
concentrations' of about 50 to 100 ppm for the high and the low sand feed rates
respectively.
Demonstration Number 18
Sand feed rate: 10.2 TPH (1-8), 15. 0 TPH (9-16), 22.5 TPH (17-24), 30 TPH
(25-32)
Water feed rate: 83 GPM (1-8), 125 GPM (9-16), 190 GPM (17-24), 250 GPM
(25-32)
Aeration rate: 250 CFM
Oil type: No. 6 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(HO saturated)
£t
Average (1-8) 7 10
Standard Deviation 25.9% 34.
Average (9-16) 12 25
Standard Deviation 19.2% 89.2%
Average (17-24) 24 20
Standard Deviation 13.4% 21.
Average (25-32) 4,600 27 39
Standard
Deviation 17.2% 30.0% 71.
Demonstration 18 took place under conditions identical to those in 12 and 13
except that No. 6 fuel oil was used as the contaminant. Total operating time
134
-------
was 1-1/2 hours. Again, residual oil concentration correlated very well with
residence time in the flotation machine. Total residual oil concentrations of
about 9, 16, 23, and 31 ppm resulted from residence times of 12, 8, 5.5, arid
4 minutes respectively.
Demonstration Number 19
Sand feed rate: 10.2 TPH (1-8), 15 TPH (9-16), 22.5 TPH (17-24), 30 TPH
(25-32)
Water feed rate: 83 GPM (1-8), 125GPM, (9-16), 190 GPM (17-24), 250 GPM
(25-32)
Aeration rate: 250 CFM
Oil type: Marine diesel fuel
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(HO saturated)
Average (1-8) 48 96
Standard Deviation 35.26% 45.5%
Average (9-16) 132 76
Standard Deviation 9.5% 48.9%
Average (17-24) 152 64
Standard Deviation 8.6% 55.3%
Average (25-32) 4,670 251 70
Standard
Deviation 14.1% 10.9% 23.
This test was identical to number 18 except that marine diesel fuel was used
as the contaminant. The diesel fuel was colored (to facilitate its analysis)
with No. 6 fuel oil at a ratio of 1 to 25. Residual oil levels again correlate
quite well with residence time; total residual oil concentrations of 64, 113,
122, and 190 ppm resulted from residence times of 12, 8, 5.5 and 4 minutes
respectively.
Demonstration Number 20
Sand feed rate: 30 TPH Water feed rate: 250 GPM
Aeration rate: 250 CFM Oil type: Marine diesel fuel
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(HgO saturated)
Average 15,200 800 130
Standard
Deviation 18.3% 11.2% 32.1%
135
-------
Analytical Results of Beach Samples
(parts per million by weight oil)
Samples by Sets Samples by Position
Sample No.
Average (IN)
Standard Deviation
Average (2N)
Standard Deviation
Average (3N)
Standard Deviation
Average (4N)
Standard Deviation
Average (5N)
Standard Deviation
Beach Sand Sample No.
81 Average (Nl)
35.2% Standard Deviation
1,010 Average (N2)
78. 0% Standard Deviation
47 Average .(N3)
41.5% Standard Deviation
54 Average (N4)
30.9% Standard Deviation
57 Average (N5)
53.7% Standard Deviation
Beach Sand
81
34.4%
44
20.
43
41%
240
179%
40
21.
Demonstration 20 involved marine diesel fuel (prepared as in the previous test)
at a high concentration in the feed sand (about 1-1/2%). Samples were taken
along the beach but this time in a northwards direction since the current along
the beach had shifted 180 degrees. Due to the high feed concentration of oil,
the effluent stream was quite dirty. Again, there was little, if any, change
with time in the amount of oil found on the beach; in fact, for this very light
contaminating oil there was even little difference in the amount of oil on the
beach at the various positions.
Demonstration Number 21
Sand feed rate: 30 TPH
Water feed rate: 250 GPM
Aeration rate: 250 CFM
Oil type: No. 4 fuel oil
Average (1-6)
Standard Deviation
Average (7-9)
Standard Deviation
Average (10-16)
Standard Deviation
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water
208
9.
220
4,960
16.
7.
190
10.2%
Cleaned Sand
(H_O saturated)
&
75
21.2%
49
16.
56
22.7%
During demonstration 21, dual settling ponds were used while the plant operated
at otherwise nominal conditions. After 6 samples had been collected, recycle
of water from the oil recovery tank to the feed box of the flotation machine was
initiated; this recycle, which is absolutely necessary for long term continuous
136
-------
operation of the system, continued for 30 minutes and then nominal operation
was resumed. The use of dual settling ponds proved feasible.
Demonstration Number 22
TEST 22A
Sand feed rate: 30 TPH Aeration rate: 250 CFM
Water feed rate: 250 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average-22A 3,860 104 76
Standard Deviation 6.3% 14% 25.9%
TEST 22B
Sand feed: 30 TPH Aeration rate: 250 CFM
Water feed rate: 250 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average-22B 4,980 517 119
Standard Deviation 3.4% 10.6% 19.7%
TEST 22C
Sand feed rate: 10.2 TPH Aeration rate: 200 CFM
Water feed rate: 75 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average-22C 4,660 59 86
Standard Deviation 4.2% 52.7% 18.;
22A took place under nominal conditions with approximately 2% of the .dispersant
Corexit added to the contaminating oil. The process was not significantly de-
graded. 22B took place under nominal conditions with approximately 2% of the
dispersant Magnus added to the contaminating oil. Performance was degraded
substantially but brought back within the effluent design criteria by reducing the
sand feed to 10.2 tonsAour and the water to 75 gallons /minute.
Demonstration, Number 23
137
-------
Sand feed ratei 22.5 tons/hour Aeration rate: 250 CFM
Water feed rate: 195 GPM Oil type: No. 4 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(H2O saturated)
Average 5,020 130 28
Standard Deviation 5.5% 13.8% 43.7%
Representatives of the Environmental Protection Agency witnessed this nominal
test on December 16, 1971. The water was quite cold (about 12 °C) and sand
was fed at 22.5 tons/hour. Adjusted to the design criteria, the effluent concen-
trations were within their design values. This is an excellent result for No. 4
fuel oil.
Demonstrtion Number 24
Sand feed rate: 22.5 TPH Aeration rate: 250 CFM
Water feed rate: 195 GPM Oil type: No. 6 fuel oil
Analytical Results
(parts per million by weight oil)
Feed Sand Exit Water Cleaned Sand
(HO saturated)
£t
Average 3,480 26 14
Standard Deviation 27.8% 4.8% 24%
Also took place on December 16, 1971 and witnessed by representatives of the
Environmental Protection Agency. Consisted of a nominal run with No. 6 fuel
oil in sand at 22.5 tonsAour. The feed concentration of the oil in the sand was
unintentionally somewhat low; however, proportional adjustment to 30. 0 tons/
hour and 5000 ppm oil in the feed sand still placed the performance very much
within the design goal, again with a low operating temperature.
138
-------
Organization
Number
2
, 1 n\
,Siib;'ecf Field & Group
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
6715 Electronic Drive
Springfield, Virginia 22151
Title
RESTORATION OF BEACHES CONTAMINATED BY OIL
]Q Authors)
Gumtz. Garth D.
16
21
Project Designation
EPA, ORM Project No.
15080 EOT
Note
22
Citation
Environmental Protection Agency report
number EPA-B2-72-0^5, September 1972.
23
Descriptors (Starred First)
*Oil Pollution, *Beach Restoration, *Water Pollution, Contaminated Beaches,
Sand, Spilled Oil
25 Identifiers (Starred First)
Froth Flotation, Mobile Facility, Field Operation
27
Abstract
Based on laboratory studies, a 30 ton per hour pilot plant was built for cleaning
oil contaminated beach sands. The plant utilized the principle of froth flotation.
Extensive field testing considered different oils, feed concentrations, both brackish
and sea water, and a range of processing conditions. Forty one field tests were
conducted at the U.S. Navy's Fleet Anti-Air Warfare Training Center at Dam Neck,
Virginia. These varied from nominal runs with sand feed rates of 30 tons per hour
and oil concentrations of 0.5% to oil/water separations at high capacity. Using
the test results, a mobile unit was designed, constructed, field tested, and delivered
to the Environmental Protection Agency. Data was obtained on the effects on cleaning
efficiency of relevant process parameters: (1) sand feed rate, (2) feed steadiness,
(3) oil type, (4) oil concentration, (5) sand age, (6) feed homogeneity, (7) water rate,
(8) water type, (9) slurry density, (10) residence time, (11) aeration, (12) temperature,
(13) surfactant effects, (14) organic solids effects, and (15) oil deposition on wet or
dry sand. The mobile unit operated successfully under a wide range of conditions.
This device should prove a valuable adjunct to existing oil spill cleanup procedures.
Abstractor
Garth D. Gumtz
Institution
Meloy Laboratories, Inc.
WR:I02
WRSIC
------- |