WATER POLLUTION CONTROL RESEARCH SERIES • 1 50SODX E 1 1 770
FEASIBILITY ANALYSIS OF
INCINERATOR SYSTEMS FOR RESTORATION
OF OIL CONTAMINATED BEACHES
ENVIRONMENTAL. PROTECTION AGENCY • WATER QUALITY OFFICE
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the results
and progress 'in the control and abatement of pollution in our Nation's
waters. They provide a central source of information on the research,
development, and demonstration activities in the Water Quality Office
of the Environmental Protection Agency, through 'in-house research and
grants and contracts with Federal, State, and local agencies, research
institutions, and 'industrial organizations.
A triplicate abstract card sheet is included 'in the report to facilitate
information retrieval. Space is provided on the card for the user's
accession number and for additional uniterms.
Inquiries pertaining to Water Pollution Control Research Reports should
be directed to the Head, Project Reports System, Planning and Resources
Office, Office of Research and Development, Water Quality Office of the
Environmental Protection Agency, Room 1108, Washington, D. C. 202/V2.
-------
A FEASIBILITY ANALYSIS OF INCINERATOR
SYSTEMS FOR RESTORATION OF OIL
CONTAMINATED BEACHES
ENVIROGENICS CO., DIVISION OF
AEROJET-GENERAL CORPORATION
9200 EAST FLAIR DRIVE
EL MONTE, CALIFORNIA 91734
for the
WATER QUALITY OFFICE
ENVIRONMENTAL PROTECTION AGENCY
PROGRAM #15080DXE
NOVEMBER 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C., 20402 - Price 75 cents
-------
EPA REVIEW NOTICE
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval, however, 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
of this method of restoring beaches contaminated by oil.
-------
ABSTRACT
The feasibility of employing a combustion process for restoring oil
contaminated beaches was 'investigated. Beach access problems and
the handling characteristics of shore materials limited the potential
application to recreational (sand) sites. Thermodynamic arguments
required that a system design be adopted in which the contaminated
sand would undergo combustive processing 'in a confined arrangement.
The design selected, from those analyzed, proved to be a three-effect
combustor based on the rotary kiln principle. Provided that the sand
to be cleaned is carefully enough collected to furnish a reasonable
(>6$) oil content and is moved away from the surf and drained to an
acceptable moisture level (£&%}, basic processing costs would be
highly attractive. In comparison with uncontaminated sand, the
cleaned product exhibits only a slightly greyish hue.
This report was submitted 'in fulfillment of Contract No. 14-12-595
between the Environmental Protection Agency and the Envirogenics Co.,
Division of Aerojet-General Corporation.
KEY WORDS: Beaches, oil wastes, cleaning, shore protection, shores,
costs, incineration, oily water.
-------
CONTENTS
Section Page
Abstract . iii
Table of Contents iv
List of Figures v
List of Tables vi
I Conclusions 1
II Recommendations 5
III Introduction . 7
IV Properties Compilation 9
V Combustion Considerations 25
VI System Design 45
VII Acknowledgements „. 69
VIII References . 71
iv
-------
FIGURES
Number
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14
15
16.
Ten TPH Combustive Sand Renovator
Moisture Loss of Wet Sand by Capillary Drying . . .
Sand-Oil-Water Mixtures Furnishing Balanced Energy
Rate of Temperature Rise in Sand/Oil Column With
Time/ Temperature Relationship in Cleaning Carbon-
Batch Type Sand Cleaner
Basic Continuous Sand Cleaning Flow Diagram ....
Fresh Air Reheat With Sand Cooler Flow Diagram . .
Rotary Kiln Cleaner With Hot Sand Recycle
Estimated Construction Costs of Basic Sand
Combustor .„
Page
2
33
34
37
38
39
41
48
50
51
53
57
58
59
60
63
-------
TABLES
Number
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Composition of Oil Contaminated Beach Samples
Sand Removal Required for Beach Renovation as
Estimated by URS Research Co0
Summary of Theoretical Combustion Calculations . .
Average Composition of Beach Samples Taken in
Puerto Rico After OCEAN EAGLE Incident
Theoretical Energy Balances for the Adiabatic
Combustion of Various Oil -Sand-Water Mixtures . .
Combustive Sand Cleaning Cost Estimates „.<,„..
Page
10
12
13
18
19
27
29
31
35
66
VI
-------
SECTION I
CONCLUSIONS
Process Feasibility: The renovation of oil-contaminated shorelines
using a combustive process is practical only under certain conditions.
In-place cleaning with flames is not feasible; the contaminated beach
material must be excavated and fed into the combustor. The sand
cleaned by combustion does not entirely recover its original color; a
slightly greyish tone persists. This is due to tiny pockets and platelets
of carbon which adhere to granules that are not of quartz origin. Abra-
sion in the surf would probably remove this source of slight discoloration.
System Configuration: The system recommended is a three-chamber
combustor operating on the rotary kiln principle. In the first chamber,
combustion air, recuperatively heated, would be introduced in con-
current flow with the oily sand. The feed would dehydrate, the oil
become evenly distributed on the granules, and.partial volatilization
and/or pyrolysis of the oil fraction would occur. Before exiting the
first chamber, the sand feed would achieve a carbonized condition and
good flow properties. The granular, carbon-coated sand resulting from
this treatment would then be passed into a higher temperature kiln where
the gas/solid flow would be counter cur rent. Here the material would be
decarbonized and cleaned to output condition. Heat could be extracted
from the cleaned sand prior to discharge in a third, rotating section.
An output of 60 tons/hr is envisioned for a full-scale system. This
should permit, with proper earth-moving practice, the renovation of a
20 ft wide strip of beach to a depth of 2 in. at a rate of one mile per
each 8-hr work day. Based on conceptual drawings (see Figure 1) a
subscale system (10-30 tons/hr) suitable for immediate field testing
would be 40 ft long, 16 ft wide, and 13 ft high (excluding stack); it
would weight about 13 tons. Because of the large size of the system,
self-propulsion would not be practical. The recommended method of
use would be to position the combustor on a level portion of the beach
and have front-end loaders bring material to it. As the clean-up oper-
ation progressed down the beach, the combustor would be periodically
moved by bulldozer or other suitable vehicle. The input and output
material would be handled by suitable conveyor equipment.
In terms of sand cleaning, the performance of the envisioned system
should not be affected when drift trash and other beach debris is fed
into it. The latter material should also emerge oil-free, but probably
not all of it would be fully combusted. Such charred articles should be
screened from the discharge.
Characteristics of the Feed: Only limited data could be obtained on the
composition of sand removed from oil-contaminated beaches. These
resulted from laboratory tests performed by Government technicians
in connection with the clean-up following the OCEAN EAGLE disaster
-------
FIGURE 1. TEN TPH COMBUSTIVE SAND RENOVATOR
-------
at San Juan, Puerto Rico. The many samples taken by these workers
were described as resembling the material that was excavated for dis-
posal. An average of 16 wt-% of petroleum residues was observed.
Estimates by FWQA officials at another spill site ranged even higher,
while others have suggested that the 16 wt-% value is unrealistically
high.
For combustion purposes, an oil content of 6 wt-% is adequate for self-
sustained combustion provided the moisture content is also about 6 wt-%.
Typical beach sand removed from the swash zone, and from which excess
sea water has been poured off, will have a moisture content of about
11 wt-%. If piled on dry sand, however, such wet material will drain by
capillary action to about 6 wt-% in about one hour.
Equipment Deployment and Operation; Because of the envisioned size
the combustor, transportation to the spill site should be handled by
firms specializing in the drayage of heavy equipment. It is doubtful
that transportation by air would be practical. In fact, the width of even
a 10 tph unit is such (13 ft) that rail-haul would not be possible. The
system could be designed so that it could be dismantled for air- or
rail-haul, but this increases capital costs by at least 25%.
Truck drayage appears to be the only suitable means of cross-country
deployment and this would probably require removal of the stack and
the issuance of a special permit in most states because of equipment
width. In moving the combustor between coastal points, serious con-
sideration should be given to the use of barges. An actual landing of
the equipment by amphibious craft at the beach site would also be
possible, but only if the shore terrain were favorable.
It has been concluded that the combustor would best find application on
sand beaches of the recreational type which are reasonably accessible
by good roads. The beach itself should have enough reasonably flat
back-shore area to permit the periodic movement of the machine.
Once on the beach, the combustor would operate as a self-sufficient
plant with the exception of fuel supply. Auxiliary firing would be accom-
plished using diesel oil, which would also fuel a motor/generator set.
The latter would be rated at 45 hp for a 10 tph combustor.
Combustor throughput has been very conservatively set and may well
prove to be low by a factor as high as three. Establishment of this
parameter was based on a thermodynamic analysis, in which several
conservative assumptions had to be introduced.
Air Pollutants: When properly fired, particulate emissions from oil
are quite low. In tests conducted on the present program, essentially
smokeless flames were obtained when oil/sand/water mixtures were
combusted under process conditions. There is no reason why this
should not also be the case in scaled-up systems. Observation of
gaseous emissions was not undertaken. The insignificant levels involved
should be acceptable considering the nature of the combustor's function.
-------
Cost Factors: While in service, the basic operation and maintenance
costs of the combustor are relatively low. Based on an 8-hr day oper-
ation, the cleaning cost would come to $0. 54/ton if the input sand con-
tained 6 wt-% oil and 6 wt-% moisture. If sand of the same moisture
but containing no oil were fired, the cost would increase to $2. 78/ton.
A number of other cost factors must also be considered. Most, how-
ever, are dependent on variables that cannot be readily fixed. Capital
cost recovery, for example, must be set on the basis of some arbitrary
utilization rate to determine chargeback. For example, if it were
assumed that a 10 tph unit would be used only 30 working days per year,
the user would be expected to pay $5. 00 for each ton of sand cleaned.
This would allow the owner to annualize his capital costs at a rate of
return of only 8%.
Transportation costs have also been found to be highly variable. These
will depend on the haul distance, the method of dray age, the need for
exclusive vehicle privileges, and the availability of suitable unloading
equipment and crews at the point of delivery. As a generalization,
overall truck transportation costs would probably run about $25. 00 for
each hour the carrier is away from its terminal.
A final factor is the cost of operational support furnished by other per-
sonnel and equipment working at the site, It is felt, however, that
because of the functions normally performed at a spill site, that no
additional equipment or manpower would actually be required.
-------
SECTION II
RECOMMENDATIONS
Because of the increasing rate of oil spill disasters , an early reduc-
tion to practice of the conceptual system described herein is advised.
It is recommended that the 10-30 ton/hr subscale system be committed
to final engineering design form and that a field test machine then be
fabricated and tested. Design characteristics for a suitably scaled
system could then be generated based on cost-effectiveness analyses
of the various possibilities. It is not recommended, however, that a
single program attempt to eventuate in the development of a complete
and fully detailed design of a correctly scaled system. The level of
effort required for such a task would be too uncertain to plan and budget
for prior to the initiation of pilot work. It is felt, in fact, that the sub-
scale (10 tph) test system recommended in this report may prove quite
satisfactory, perhaps after suitable modification, for routine beach
clean up operations.
Finally, it is recommended that such a test program be conducted so
that an FWQA Oil Spill Coordinator would have an opportunity to par-
ticipate, at least during key, on-site test operations. The experience
of such an individual would be invaluable in insuring that the work was
performed in a manner consistent with realistic beach disaster conditions.
While this report was being written, four major petroleum spills
occurred. These were the tank vessel accidents involving the DELJAN
APPOLLO in Tampa Bay, Fla, the ARROW off Chedabucto Bay, N. S. ,
and the two tank barges in Humboldt Bay, Calif and Jacksonville, Fla.
-------
SECTION III
INTRODUCTION
The clean up of coastal areas contaminated by petroleum spills is as
unpleasant as it is expensive. If not properly conducted, a renovation
effort can result in devasting effects on the biota, serious depletion of
recreational sand inventories, and lingering contamination problems
that are not immediately obvious. The development of safe and more
cost-effective techniques is now being vigorously pursued by the
Federal "Water Quality Administration, and other public and private
agencies. An objective common to these organizations is the develop-
ment of a capability wherein spills can be isolated and collected at sea
before shore areas are befouled. However, some incidents of oil
stranding will be unavoidable, as when the petroleum issues from a
shore facility or beached vessel, or from an off-shore source close to
land.
Recognizing the inevitability of such situations, the FWQA is currently
sponsoring several programs dealing with new shore renovation con-
cepts. A definite trend being established by this new technology is
toward processes that avoid the sacrifice, by disposal, of valuable
natural resources, notable among which is the sand of recreational
beaches. In line with this objective, the present program has been
addressed to the investigation of combustive techniques.
Although any form of oil burn-off will certainly destroy any organisms
present, it is a process that would involve only a limited segment of the
riparian zone. In any case, few totally nondestructive processes can be
conceived of and these are slow. Relying on the washing action of the
surf and/or biological effects to remove stranded oil can usually not be
accepted if recreational areas are to be quickly returned to public use.
Depending on the nature of the problem, a variety of approaches are
possible in considering the use of combustion processes for beach
decontamination. Even when the proper modality is recognized, trade-
offs still must be considered since the conceptual design inputs can be
influenced by many economic, thermodynamic, and operational con-
straints. The approach used in dealing with such factors and in attempt-
ing to realize an optimum system design is summarized below.
The initial task performed consisted of a planning analysis wherein the
project objectives were defined and methods for their accomplishment
were selected. This resulted in the following phased approach being
adopted.
Properties Compilation: The purpose of this task was to acquire data
that would be needed in the feasibility analysis, particularly with
respect to equipment design parameters. Included in this search was
information on the nature of continental U.S. shore areas, the properties
-------
of oil-contaminated beach materials, the nature of oil spills and the
renovation methodology hitherto practiced. This was accomplished by
reviewing the extant literature and by conducting a field survey in which
various specialists, including personnel associated with the clean up of
four major oil spills, were interviewed.
Combustion Considerations: A critical analysis of the combustion pro-
cess involving the sand-petroleum-water system was to be conducted in
order to develop semi-qualitative data needed for system selection and
design. Bench work was then to be performed to verify the conclusions
derived and to determine the properties and behavior of various input
compositions and the quality obtainable in outputs that had been variously
processed.
System Design: Practical engineering design features were to be devel-
oped and incorporated in various system concepts. The information
developed to this point was then to be reviewed and the final conceptual
design hardened.
The work performed on this program was jointly carried out by the
Envirogenics Co. , Division of the Aerojet-General Corporation (prime
contractor) and Hirt Combustion Engineers of El Monte and Montebello,
California, respectively. The program manager was Mr. R. M. Roberts;
the principal contributors were Dr. R. W. Lawrence of Envirogenics Co.
and Messrs. T. S. Hoyt and C. S. Miller of Hirt. The Project Officer
was Mr. G. Burke of the FWQA facility at Alameda, California.
-------
SECTION IV
PROPERTIES COMPILATION
Overview: Testing the feasibility of and developing designs for com-
bustion-type equipment for renovating oil contaminated beaches requires
the assessment of many factors. An understanding of the properties of
shore terrane materials containing petroleum intrusions is certainly a
necessity. Of almost equal importance, however, is the consideration
of economic constraints which place practical limits on the scope of
clean-up operations- One must consider the self-mitigating nature of
oil damage in terms of esthetic and toxic effects, and the manner in
which damage-control decisions can be influenced by oil type, magnitude
and time-frame of intrusion, availability of equipment and disposal areas,
and the types of beaches on which the oil becomes stranded. Thus, in
formulating design guidelines, abroad overview of the application situ-
ation must be developed. This was the purpose of the first task under-
taken on the program.
Methodology; The somewhat limited literature on oil spill control and
decontamination was collected and reviewed. This was done primarily
to obtain information concerning the physical, chemical, and thermo-
dynamic properties of petroleum-fouled, shore terrane materials.
Additionally, selected source works on geomorphology were consulted
to acquire a general appreciation of the nature of the shoreline of the
continental U.S. This overall literature is listed in Section 8 as refer-
ences 1 through 34. Unfortunately, the Joint (API/FWPCA) Conference
on Prevention and Control of Oil Spills held in New York City in mid-
December 1969 occurred too late in the program to be included in this
review.
In conjunction with this literature review, which was periodically up-
dated, a field survey was conducted. This consisted of interviewing
technical and supervisorial personnel cognizant of or involved in com-
paratively recent, major spill incidents. This included, where possible,
individuals who were on the scene from the peak of contamination until
the ultimate clean up was effected. The information was sought using
the questionnaire check-list shown in Table 1.
-------
TABLE 1
Information Sought in Field Survey
NATURE OF SPILL
Type and estimated quantity of petroleum released by
accident.
Estimated amount of petroleum deposited on shore.
Sea state existing during major portion of slick movement.
Distance between leakage source and affected shore area.
Estimated sea residence time of slick before contacting
beach.
ENVIRONMENTAL CONDITIONS
Water and air temperature.
Cloud cover.
Surf characteristics.
Tidal range during and after incident.
Presence of sea-borne chemical dispersants, oil-
absorbents, and drift-trash (wood, kelp, flotsom, etc. ).
GEOLOGICAL STRUCTURE AND DIMENSIONS
OF AFFECTED AREA
Classification of terranes and minerals.
Dynamic factors (particularly sand movement).
Sand/stone interdistribution.
Excavatability of terrane materials.
10
-------
TABLE 1 (Cont'd)
PROPERTIES OF CONTAMINATED BEACH MATERIALS
• Oil distribution characteristics - film thickness, substrate
penetration, continuity, etc.
• Deposit characteristics - degree of emulsification, mois-
ture content, rheology, variation of properties with depth,
aging effects.
• Characteristics of excavated sand/oil mixtures.
• Mechanical techniques employed.
• Agglomerative and packing.mechanics.
• Typical compositions outloaded - moisture, petroleum,
pebbles/stones, etc.
The personnel contacted are shown in Table 2. The oil accidents selec-
ted for study were: (1) Rehoboth Beach, Delaware - oil barge HESS
HUSTLER; (2) Puerto Rico - S. S. OCEAN EAGLE; (3) Panama Canal
(Atlantic entrance) - S. S., WITWELL; and (4) Santa Barbara, California
Union Oil Company, Platform A.
While considerable valuable information was acquired on this survey,
many items listed in Table 1 could, at best, only be coarsely estimated
using the interview technique. The questionnaire guide was accordingly
pared down by eliminating items of marginal relevance and those which,
to obtain, would require an effort disproportionate to the value gained.
The net body of information acquired in both the literature review and
the field survey was then screened and digested. The broad problem
overview and conclusions developed are summarized below.
NATURE OF THE CONTINENTAL U.S. LITTORAL
Nomenclature: The standardized geological terminology recommended
by King (Ref. 13) will be observed in this report. The term "beach"
includes the zones of the back-, fore-, and off-shore regions. Back-
shore is the area lying just beyond the reach of the swash of the normal
high spring tide and is therefore only wetted under the influence of ex-
ceptional seas or tides. On a rocky coast, the backshore includes the
cliffs and palisades, while on low coasts will include sand dunes and
mature salt marshes. The foreshore (intertidal) zone is that part of
the beach which extends from the swash mark at highest annual tide
to backwash level at lowest annual tide. The offshore zone extends
from the latter point to a distance at which the movement of beach
material is negligible under normal water conditions.
11
-------
TABLE 2
Personnel Contacted on Field Survey
Individual
Visited (1)
J. R. Fraser
T. H. Gaines
D. Craggs
P. G. Mikolaj
H. Bernard
K. E. Biglane
L. P. Haxby
LCDR P. Sova
H. J. Lampl
E. J. Struzeski,
Jr.
L. C. Mendez
G. Trotter
H. Ramirez
L. Hyman
P. Getabert
R. Rodriguez
E. H. Kohn
P. W. Glynn
City
Los Angeles, Calif.
Santa Barbara, Calif.
Santa Barbara, Calif.
Washington, D. C.
Washington, D. C.
New York, N. Y.
New York, N. Y.
Edison, N. J.
San Juan, P. R.
Bayamon, P. R.
San Turce, P. R.
Balboa, C. Z.
Po s ition/Affiliation
Union Oil Company
Union Oil Company
University of California
at Santa Barbara
Chief, Agricultural &c
Marine Pollution Con-
trol Branch, FWQA
Director of Div. of
Tech. Support, FWQA
Chairman, API Com-
mittee for Air and
Water Conservation
Intelligence fe Law
Enforcement Branch,
USCG
Hudson-Delaware River
Basins Office, Edison,
N. J., FWQA
Army Corps of Engineers
Caribbean Refining Corp.
Dept. of Public Works,
Comm. of P. R.
Smithsonian Tropical
Research Institute
(1) Listed in chronological order.
12
-------
o
In terms of the two prevalent shore forms , the composite sand beach
and the composite shingle beach, separate nomenclatures have deve-
loped. This is because of the topographical differences between the
two which arise from the degree of consolidation of the shore materials
involved.
The composite sand beach usually features a terrace, called a "berm, "
which is situated just above the highwater swash. It is of variable width
and may slope somewhat in either the fore- or backshore direction. In
the foreshore zone, crests and depressions are called, respectively,
ridges and runnels. In the offshore zone these corresponding features
are called submarine bara and troughs. Thus runnels, which are high
and dry at low water, tend to trap rubble, drift trash and oil, while
troughs, which are always under water, tend to accumulate pebbles
and tarry deposits that have agglomerated with heavier materials (e. g. ,
sinking agents). Above-water sand bars separated from the shore by
lagoons or channels are referred to as "barrier beaches" or "islands,"
depending on their dimensions.
In the backshore of the shingle beach, prominences are referred to as
"beach ridges," while the separating depressions are described as
"swales. " The smaller features developed in the intertidal zone are
called "foreshore steps, " while the step located at the break point of
low water waves is called the "break-point step. "
The size classification of materials occurring in beaches will follow
the system adopted by Zenkovish (Ref. 14), as shown in Table 3. This
classification is applicable regardless of the origin of the material.
Biotic residues can, however, be further identified by referring to
them as shell gravel or coral shingle, for example.
TABLE 3
Classification of Beach Materials
Description Diameter, mm
Boulders 100 (!)
Shingle, large 100-50
Shingle, medium 50-20
Shingle, fine 20-10
Gravel, large 10-5
Gravel, medium 5-2
Gravel, fine 2-1
Sand, coarse 1.0-0.5
Sand, medium 0.5-0.25
Sand, fine 0. 25-0. 1
Silt, various grades <0.10
As many as eight basic beach types have been classified (Ref. 9), but
for present purposes these two types need only be considered.
13
-------
Geomorphology: According to McGill (Ref. 11), who was personally
contacted and generously furnished advice, and the Morskoi Atlas
(translation in Ref. 12), the U. S. Atlantic and Gulf coasts, all the
way from Long Island, N. Y. , to Corpus Christi, Texas, are geo-
morphically homogenous. With the exception of some delta plains,
associated with the Mississippi and Brazos Rivers, and the riverine
complex emptying into Mobile Bay, this entire shoreline is essentially
a system of wind shaped coastal dunes deposited on sedimentary plains.
Excluding Alaska, this type of riparian perimeter exists over 60% of
the U.S. coastline, presents a high potential for recreational (bathing)
activity, and, unfortunately, is highly vulnerable to oil damage. In ad-
dition to a heavy tank-ship traffic off these coasts, sea-platform drill-
ing is extensive in the Gulf region and some 105 tanker hulks, casaul-
ties of German submarines in World War II, are scattered (Ref. 16)
throughout the entire area.
South of Corpus Christi to the mouth of the Rio Grande, the land form
consists of dune plains, with scattered delta plains. While suitable
for recreation, the beach area is sparsely populated.
North along Long Island Sound to Cape Cod, itself partially alluvial
(Buzzards Bay) and partially of wind-formed coastal dunes, the region
predominates in glaciated complex hills. With the exception of Cape
Cod, this region offers relatively few pleasure beaches. The coast-
line north of Cape Cod is similar to that south of it. This region,
however, underwent extensive post-Wisconsin marine submergence
and presently is undergoing geological emergence due to isostatic re-
bound. This has resulted in a very rocky coast, which, with tidal
ranges in excess of ten feet, has made the area largely unsuitable
for recreational beach activities other than shore fishing.
The west coast of the U.S. is geomorphically very complex. It should
also be noted that unlike the Atlantic side, the Pacific waters abound
with kelp beds all the way north to the Aleutians. Exceptions are short
reaches along the Oregon-Washington coast where barrier beaches exist
and in the areas around Chula Vista and Cape Mendocino, California.
The coast of the latter State from Point Conception south to the Mexican
border is an equally distributed system of stream-eroded complex hills
and mountains, and of alluvial plains. Because of the influence of
water erosion, good beaches exist over most of this coastline, although
many are inaccessible and, therefore, relatively unused. Between
Point Conception and Monterey Bay the coast structure is about 50%
stream-eroded complex hills and alluvial plains and 50% complex
mountains. Relatively few accessible beach areas exist here, with
the notable exceptions of Morro and Monterey Bays. North of the
latter point to about Rockport, California, the coastal system consists
of complex hills which increase to mountain dimensions. The coast
farther up presents a mixture of these two water-eroded forms and
narrow alluvial plains up to the region of The Heads, Oregon. North
of the latter to about Aberdeen, Washington (an alluvial area) there is
a system of complex hills, 50% of the coast exhibiting wind-formed
14
-------
dune hills. From this point north to the Aleutians the coast exhibits
mostly stream eroded mountains having tidal ranges in excess of 10
feet. Some narrow alluvial plains are found, as between Yakutat and
Cordova, Alaska. In general, however, the coast north of San Fran-
cisco is sparsely populated and is used by a significant number of
recreationists at only a few points.
Composition of Recreational Beaches: Following an oil spill disaster,
the priority in damage control and renovation operations will logically
be assigned to bathing beaches. These are typically quartz sand beaches
in the continental U.S. , the Florida Keys (coral sand predominates)
being the only significant exception. In certain parts of the U.S. bathers
do use shingle beaches (e. g. , Long Island Sound) but perhaps without
great enthusiasm.
The term sand beach is misleading in that it implies that the beach
material is totally sand. This might be nearly true with some berms
but only if nature happens to favor the situation and man cleans the
bermal zone frequently with motorized rakes. The foreshore area,
however, is typically littered with shingle or gravel, shell debris,
drift trash, stranded marine vegetation, and (in many parts of the
country) tar deposits of unknown origin. The relative occurrence of
such materials will depend on many factors of course and cannot be
predicted. Because of the classifying effect of the swash and backwash
(Ref. 13 and 14), it is probably safe to expect that the upper reaches of
the foreshore will contain much less than 25% imbedded gravel and
shingle, and usually very little of the latter. This is the zone where
oil deposition is most likely to occur.
THE NATURE OF THE PROBLEM
The Properties of Petroleum Deposits: Useful information on this sub-
ject was found in a number of documents (particularly Refs. 2, 3, 9>
10, 22 and 25). It was also obtained from various individuals involved
in the Santa Barbara (local crude), Rehoboth Beach (bunker C), San
Juan (Leona crude) and Canal Zone (bunker C and diesel oil - 3:1)
spills.
The scope of this review was restricted to spills in which the released
petroleum contained a significant residual oil content. Crude and bunker
oils are the predominate cargoes of tank ships and, in any case, dis-
tillate-fuel spills would probably not require the application of the type
of equipment considered on this program .
Spills will usually arise from a distressed tank ship, barge, drilling
platform, drilling ship, or possibly a shore facility. Because mixed
_ —,
Distillate fuels penetrate even wet sand to form loosely caked masses
which quickly disperse due to wave action and evaporation.
15
-------
cargoes (e. g. , the WITWATER off Galeta Island, C. Z.) or very light
crudes may be released, a low residual oil content in the slick must
be anticipated. Heavy seas are also likely to be running to have caused
the carrier's distress. This might result in some emulsification of
the oil and the casting of it onto backshore zones4. Depending on the
time the slick was carried at sea and the ambient temperature (water
and air), the oil will lose volatiles and tend to congeal. When stranded
on the beach it accumulates at the swash zone in windrow-like patterns
or in liquid patches, depending on its viscosity. The tidal area washed
by the surf is usually free of such deposits. In the absence of applied
(damage control) detergents, the fluid or plastic contaminant usually
penetrates the foreshore very little unless the sand is coral-derived.
Due to wave action, sand will collect on oil deposits and suggest the
appearance of an oil penetration. In cold weather (e. g. , Rehoboth
Beach), bunker oil may become plastic very quickly and form rigid
deposits several inches thick. At warmer San Juan, the OCEAN
EAGLE's Leona crude (60-66% residuals) was estimated to have under-
gone congealment on the beach within about 24 hours. Rapidly con-
gealing oil will usually not adhere to drift trash and the use of ab-
sorbents or adsorbents is not warranted. Deposits that are fluid for
a time represent the opposite case and clean-up must also include the
removal of oil-soaked driftwood and any applied sorbents (straw,
mineral products, etc.). A special case that arises is the presence
of kelp or sea grass. The former will usually not acquire any oil if
it is fresh; this is not true if toe kelp is dead and dried out. Sea grasses
apparently act as good oil-accumulators regardless of their condition.
The easily collected patches of sargassum on the Puerto Rican beaches
proved most helpful in expediting clean-up operations there after the
OCEAN EAGLE disaster.
While the oil deposit is congealing to an asphaltic or tarry state and
acquiring sand from wave action, it typically breaks up into streaky
masses and/or forms small ball-like deposits. Some of this material
may be returned to the sea by backwash and collect in runnels or
troughs. This process apparently also operates if the oil is emulsified
(water in oil) when stranded. Natural emulsions appear to break rapidly
under the influence of sunlight and the presence of a solid phase. The
persistence of detergent-promoted emulsions of the oil in water type
("chocolate mousse") is apparently greater, as are the problem asso-
ciated with its physical removal. Because of restrictions against the
indiscriminate use of detergents, this type of emulsion will not likely
be encountered on this country's shores.
Within about 24 hours following the stranding of the last material from
an offshore slick, an oil-contaminated beach can usually be worked
mechanically. The deposits and any oil-soaked sorbents or trash
4
At Rehoboth Beach, Delaware, for example, bunker oil from the
beached barge, HESS HUSTLER, reached the board walk and even
a parking lot.
16
-------
should be picked up, with a minimum of sand, using combinations of
road-graders, front end loaders, bulldozers, and even laborers
equipped with rakes, shovels and pitchforks. The collected material
can then be loaded into trucks for disposal elsewhere or, in the present
context, be fed into on-site renovation equipment to give back to the
beach a cleaned sand.
The fuel value of the outloaded material varies considerably, depending
on the nature of the spill and the expertise practiced in removing the
deposits. At Rehoboth Beach, for example, the crews were inexperi-
enced and initially were depleting far too much of the valuable sand
inventory. This situation was corrected by the FWQA oil spill co-
ordinator and it was believed that the outloaded material thereafter
contained, on the average, about 20% petroleum residues. The Public
Works Department of the Commonwealth of Puerto Rico conducted a
beach sampling program. A geologist and a laboratory technician col-
lected samples from various beaches despoiled by the OCEAN EAGLE
cargo. These were taken from within several weeks to several months
after the incident and included samples of apparently unpolluted as well
as heavily polluted sand. These samples were analyzed by the Common-
wealth for moisture and oil content, the balance being expressed as
sand and sargassum. The geologist, L. Hyman, was interviewed to
determine if the samples would resemble material outloaded if the
same areas were subjected to clean-up by e xcavation. She felt that
the compositional correlations would be good, provided that the beach
areas had not been subjected to detergent treatment. Generally,, the
samples were taken to include about a 2-in. depth, although lesser and
greater depths were employed if the oil layer was thin or if oil-sand
layering had occurred, respectively.
An example of the data obtained by the Commonwealth workers is shown
in Table 4, which represents the first sampling effort. This occurred
just two weeks after the breaking up of the OCEAN EAGLE on 3 March
1968. The area sampled involved about 20 miles of prime recreational
beach area of the north (El Dorado) coast of Puerto Rico from Punta
Salinas, which is about 3 or 4 miles east of the mouth of San Juan har-
bor, east to Puerto del Tortuguero. The great variability in the mois-
ture content probably only reflects the position of the sample site with
respect to the surf and tidal conditions that happened to exist at the
time of sampling.
It should be mentioned that as a result of the OCEAN EAGLE spill (3 to
3-1/2 million gals. Leona crude oil), 13,000 cubic meters of material
had to be removed from the Puerto Rican beaches and dumped at dis-
tant disposal sites. Fresh sand fills were required in many places^.
Particularly the Condado resort hotel beaches where the use of
detergents resulted in serious beach erosion.
17
-------
TABLE 4
Composition of Oil-Contaminated Beach Samples (Puerto Rico)
Beach Area
Laguna Tortuguero
Dorado Hilton Hotel
Playa Cerro Gordo
Playa Cerro Gordo
Playa Cerro Gordo
Dorado Beach Hotel
Fuente al Compamente
Tortuguero
Dorado Hilton Hotel
Dorado Hilton Hotel
Dorado Hilton Hotel
Dorado Hilton Hotel
Punta Salinas
Playa Brenas
Playa Brenas
El Umice Restaurant
El Umice Restaurant
Levittown
H2O, wt-%
0.4
2.2
8.8
40.4
29.8
77.4
28.2
10.3
20.4
40.8
16.3
11.0
25.9
17. 7
14.4
13.4
14.4
Oil,, wt-%
43.9
1.4
48.0
13.2
60.9
8.5
54. 7
7.3
13.4
5.0
6.8
12.5
8.1
2.7
6.8
0.3
6.8
Balance
(Sand, etc. )
wt-%
55.7
96.4^
43.2
46.4(1)
9.3(1>
14.lt1>
17.lt1>
82.4
66.2
54. 2(1>
76.9
76. 5(1>
66.0
79.6
78.8
86.3(2>
78.8(1>
'Sargassum present in sample.
No visual evidence of oil contamination.
18
-------
Some of this material had to be brought all the way up from the South
Coast of Puerto Rico.
Beach Material Excavated: An important factor in designing on-site
renovation equipment is the amount of contaminated beach material that
must be removed in order to leave a clean shore line. It can reason-
ably be assumed that this quantity would also represent the amount of
feed material that would have to be processed for the same area by a
combustive sand cleaner. Data of this type were not available in a
readily interpretable form from any of the sources consulted during
the present task. Fortuitously, however, an FWQA program on beach
renovation has been concurrently pursued by the URS Research Com-
pany of Burlingame, California. Their work included a detailed analysis
of the characteristics and capabilities of various types of earth-moving
machines for outhauling oil-spoiled shore material. Based on field
test work, estimates were generated as to the volume of sand that would
have to be picked up under different beach conditions using preferred
types of earth movers. These data (Ref. 34) are shown in Table 5.
The areas employed for these tests were Francis Beach, Tunitas Beach,
and Half Moon Bay Harbour Beach, all on the San Mateo Coast of Cali-
fornia. They were selected as having about the full range of sand pack-
ing characteristics that would be encountered on recreational beaches.
The data represent an averaging of the excavation test results obtained
at all three of the beaches.
TABLE 5
Sand Removal Required for Beach Renovation
as Estimated by URS Research Co.
Sand Removed, Cu Yds/Ac re
Firm Beach
Procedure Backshore Intertidal Zone with Straw(l)
Motor Grader and 130-145 70-100 180-200
Elevating Scraper
Elevating Scraper 300-400 200-250
Motor Grader and - 300-325
Front End Loader
Front End Loader 800-1200^ ^
(tracked)
' 'Approximately 100 bales/acre.
' 'Based on beach where a motor grader or scraper could not operate
because of poor bearing capacity.
19
-------
Assuming that a strip 20-ft wide is removed, it can be seen that any-
where from 170 to 2900 cu yds of material will have to be removed per
mile of beach. This, of course, would depend on the type of equipment
used and the zone from which the material is removed. A majority of
the contaminant can be expected to collect on the hard packed intertidal
zone, however. Thus, if the optimum excavation procedure is used
(motor grader and elevating scraper), the removal and cleaning of 38
to 53 tons per hour would be equivalent to cleaning one mile of beach
per 8 hr work day. This is based on a1 moist sand density of 125 lb/
cu ft (Ref. 35), which is somewhat conservative in that the presence
of oily deposits would reduce the density. Of course, if a swath wider
than 20 ft must be removed, which would probably involve some back-
shore material, a corresponding increase in the outloading rate would
be necessary to realize the same linear rate of clean-up.
The presence of straw can be neglected in this calculation. Although
it would increase considerably the volumetric bulk of the material that
needs to be outloaded, this would not influence the sizing of the com-
bustive renovation equipment. At typical input temperatures, the in-
loaded straw, even if wet and oil-free, would burn down so rapidly that
system congestion could not occur.
Natural Recovery of Oil Contaminated Beaches: Considerable loss of
tourist revenues can occur if oil-spoiled recreational beaches are not
quickly restored. In the case of shores that are little used, because
of their structuring or inaccessibility, the question arises whether
clean-up operations need be employed at all. Experience has shown
that the combined effects of wave action and animal activity will result
in the eventual rehabilitation of an oil-contaminated beach. What, how-
ever, is the overall effect on the biota if petroleum residues are allowed
to disappear in this manner? The consensus (e.g. , Ref. 17 and 18)
appears to be that little permanent damage occurs when oil residues
are left on shore areas and that nature manages to adjust to all aspects
of the situation. This opinion was expressed by many of the officials
and engineers contacted on the field survey and was vividly corro-
borated by an inspection of the spill site at the Atlantic mouth of the
Panama Canal.
The small tank ship WITWATER released about 800,000 gallons of
bunker and diesel oils there on 13 December 1968. Most of this oil
accumulated in the region of Galeta Island and Fort Randolph. This
region features a mixture of mangrove swamps, coral-sand beaches,
riprap breakwaters, sea walls, and shingle beaches. The areas hard-
est hit by the spill were inspected in August 1969 with the knowledge-
able guidance of Dr. P. W. Glynn, marine biologist of the Smithsonian
Tropical Research Institute. With the exception of the Galeta Island
lagoon, where some oil-soaked drift trash had been removed, no
renovation had been attempted other than to burn down and pump off
the encroached heavy slick.
20
-------
The animal and plant life of the area was prolific and obviously quite
healthy. The feared poisoning of the mangrove trees through their
feeder roots had not eventuated. The occurrence of marine animals —
chitons, crabs, sea urchins, and many varieties of tropical fish—was
abundant. With the exception of a (useful) paving of asphalt atop the
sea wall, some patches of hardened tar high on the riprap barriers,
and some occasional oil-contaminated trash on the backshore side of
the sea wall, there was surprisingly little evidence of oil. There was
also no sign of oil on the bottom of the offshore shallows, where the
clarity of the sea water was exceptional.
Dr. Glynn expressed the opinion that most of the oil was removed by
wave action and the activities of chitons and micro-organisms. A
similar appraisal was later made by Dr. P. G. Mikolaj (University of
California, Santa Barbara) concerning the natural recovery of the
rockeries and other zones near that University which were not cleaned
after the Santa Barbara oil disaster.
USE OF COMBUSTIVE TECHNIQUES IN RENOVATION OPERATIONS
Cornish Coast, England: Butane- or propane-fueled torches were used
in an attempt to cleanse sea walls. This re suited in an apparently clean
surface, which, on closer examination, proved to have resulted from
the thermal exfoliation of the concrete surface. Recovered particles
were found to have (partially carbonized) tarry deposits still intact on
the face of the fragments. The method was abandoned, but apparently
because of its cost and slowness rather than its obvious destructive-
ness. Other combustion attempts employed were aimed at firing the
main slick, pockets of oil entrapped on the beach, and the TORREY
CANYON herself. Only the incendiary bombs dropped on the ship,
which caused confined burning of the oil, proved to have any signi-
ficant results.
Santa Barbara, California: Weed-burning torches were employed in
an attempt to clean rocky areas. This was soon abandoned because a
black veneer-like coating persisted on the surface of the rocks which
resisted all subsequent removal efforts.
Another device tested there was a truck-mounted contraption for aug-
menting the combustion of piled, oil-contaminated, drift trash. Con-
siderable amounts of this material accumulated on the beaches after
untimely heavy rainstorms caused disastrous flooding in nearby inland
areas. Heaps of this storm debris were collected and fired; these piles
burned very slowly and emitted considerable quantities of objectionable
smoke. The device pressed into service consisted of a directional
blower and fuel-squirting assembly. Diesel oil was intermittently
played onto the smoldering piles through an 1/8-in. nozzle. The blower
was directed onto the combusting mass. The distance maintained be-
tween the truck and pile is not accurately known, nor are many other
operational details. The net effect was that the burns were greatly
accelerated and the bulky masses reduced to residues which were
disposed of much more conveniently.
21
-------
Rehoboth Beach, Delaware: A military type flame-thrower was em-
ployed here in the hope of igniting and burning off the thick deposits of
Bunker C oil stranded on the beach; this proved ineffectual.
San Juan, Puerto Rico: Torches were also investigated here. Detailed
information concerning the type of equipment employed could not be ob-
tained. The attempt, however, was described as fruitless.
Galeta Island, Canal Zone: A boom-entrapped slick, about 4-in. thick,
was augmented with kerosene and straw (wicking), then fired. About
75% of the oil burned off before sustained combustion became impractical.
The thickened residue was then pumped off, using a vacuum tank truck,
and poured into a prepared pit where it was again laced and ignited. It
burned down to leave a residue that could be covered with earth.
CONFIGURATIONAL CONSIDERATIONS
Benefits: Based on the foregoing discussion, it was believed that the
most valuable functions to be fulfilled by a beach renovation device
(based on the combustion principle) would be as follows:
• Utilize energy available from contaminant to return sand
to its original condition.
• Eliminate costly loss of sand inventories resulting from
disposal operations.
• Eliminate expense of hauling out spoiled sand to distant
dump sites.
• Reduce equipment and support requirements by confining
renovation operations to immediate contamination area.
• Concomitantly dispose of or burn down oil-soaked drift
trash and sea plants fouling the stricken area.
Application Constraints; As previously suggested, the in-place com-
bustive renovation of oil-contaminated beaches appears to be imprac-
tical. This argument is supported by thermodynamic arguments con-
tained in the next section. It would be necessary, therefore, to isolate
(excavate) the superficially deposited contaminant with a minimum of
mechanically occluded substrate from the heat sink (sand-water column)
and incinerate the mixture in a nearby device. Because the art in-
volved in accomplishing the collection operation is somewhat demanding
of even an expert operator of earth-moving equipment, the idea of coup-
ling together the excavator and appropriate combustion equipment to
form a completely mobile system appeared to be unfeasible unless and
until better-designed oil-collectors could be developed.
22
-------
The degree of mobility of the incinerator system would obviously be
predicated on its size, configuration, and functional characteristics.
Delivery of the system to a disaster site should be as rapid as pos-
sible. Because it must be anticipated that only a few or even only one
of these devices might be available and at a great distance from the
site, air transportability would be a desirable capability.
The scene of operation of the envisioned incineration device would be
the accessible recreational sand beach. Material fed to it would be
highly variable in composition. The primarily desired output would
be clean, oil-free sand having the same appearance as the original
uncontaminated sand. The management and fate of larger input in-
gredients, from gravel up to large pieces of driftwood or masses of
vegetable matter, would, of course, constitute an important design
consideration.
^Either by virtue of being self-propelled or of being capable of being
drawn with or without some preparatory dismantling.
23
-------
SECTION V
COMBUSTION CONSIDERATIONS
THEORETICAL ANALYSIS
Introduction: Two basic arrangements, emerge in visualizing the com-
bustion of oil -contaminated beach material. One is the in-place ap-
proach, wherein a flame is applied to stranded deposits until the latter
are completely dispersed in the form of volatile combustion products.
As pointed out in Section IV, attempts at using this method after actual
spills has not proved successful. The other, more logical, procedure
is to remove the contaminated material with some underlying clean
sand and process it through a combustive device, again burning all of
the petro -contaminant to residue -free gases.
Ideally, the in-place method would be applied to deposits that are con-
veniently arranged on the surface of the beach with no covering sand
interposed. In this situation, the system may be treated as a semi-
infinite solid. In the other approach, we can consider the use of a
machine that affords a separation of the dirty sand grains as much as
possible and allows them to flow, cascade, or be driven by a conveyor
in the presence of a hot gas stream containing available oxygen.
Semi-Infinite Solid: Let it be assumed that a flame containing excess
air is played at some fixed velocity on a layer of contaminant lying on
the sand. The surface exposed to the gas may further be assumed to
have already lost its volatile constituents in the heating process or by
prior weathering. The surface now consists of pitch or asphalt of a
density of about 80 Ib/cu ft. As the gas impinges turbulently on this
surface, oxygen is transported through the stagnant gas film to the
carbon monoxide. The surface concentration of oxygen can be assumed
to be zero, because the reaction rate of oxygen with carbon is very fast
at 1000°F and above.
It is further assumed that the applied gas has a linear velocity of 10
ft/ sec and is produced by burning a hydrocarbon fuel with 150% excess
air leaving 12% oxygen in the gas (0. 12 atmospheres), which is at
1800°F. If the contaminant surface is at 1000°F, the mean film tem-
perature will be about 1400°F. The gas density, viscosity, and the
diffusion coefficient of oxygen through a stagnant film of nitrogen were
evaluated for these film conditions. A characteristic length, d, of 1.0
ft was chosen and the corresponding Reynolds and Schmidt numbers
computed. The generalized correlation presented by Treybal (Ref. 36)
was used to compute the mass transfer coefficient, j^, for flat plates,
and hence the rate of oxygen diffusion, NO-, was obtained from:
kG = ^DOM/ <"
25
-------
and
NQ = kGpQ , Ib moles/sq ft-hr (2)
£• Lt
where: k_ is the mass-transfer film coefficient,
Lr
GA, is the mass flow rate of gas, Ibm/sec-sq ft,
M
p is the mean inert gas pressure in the film,
BM.
S is the Schmidt number, and
c
p is the difference in the partial pressure of oxygen
2 across the film (0. 12 atm).
The results of the calculation (Case 1, Table 6) show that 1. 32 Ibs of
carbon/hr-sq ft are consumed. Because petroleum asphalt has an ap-
proximate density of 80 Ib/cu ft, the regression rate of the surface is
00 108 inches/hr. This rate is so slow that an extremely high fuel con-
sumption would obviously be required to make the process feasible.
Combustion Around Individual Sand Grains: Let it be assumed that a
20-mesh (coarse) sand of spherical particles 0.05 in. diameter is used
and each sphere is coated with a layer of asphalt 0.01 in. thick, and a
column of coated sand one foot in diameter is considered. Hot flue gas
at 1800°F flows through the column, which is to have a uniform tem-
perature from top to bottom. The gas composition and other properties
are considered to be the same as in Case 1; i. e. , a particle surface
temperature of 1000°F and a mean gas film temperature of 1400°F. In
this case (Case 2) the length applicable for the Reynolds number is the
particle diameter (0.05 inch). Using the correlation curve for mass
transfer to spheres (Ref. 36), oxygen reaches the carbon surface at
0. 90 Ib moles/hr-sq ft of particle area. Since each sand particle has
an area of 54. 9 x 10~" sq ft and each particle contains 2. 04 mg of car-
bon, it will require 8. 7 sec to consume all of the carbon (see Case 2).
Additional computations were made to find the effect of mass velocity
and of temperature. If the velocity were increased four-fold to 40 ft/sec
(Case 4), then only 3. 3 sec would be required to remove the deposit.
If the film temperature were dropped to 700°F, keeping the initial
10 f/sec gas velocity (Case 3), then 6.0 sec would be required; but
700°F is so low a temperature that chemical reaction rates would
probably become the controlling factor rather than gas diffusion.
In general, one may conclude that, if oxygen diffusion is the controlling
process, carbon burn-off will increase with: (1) decreasing particle
diameter; (2) increasing gas mass-velocity, and increasing oxygen
content in the stream. Heat transfer to the particle is also important
because the surface must be brought up to the combustion temperature.
In the absence of detailed calculations, it can be assumed that heat
26
-------
TABLE 6
Summary of Theoretical Combustion Calculations
Gas velocity, ft/ sec
Cone O?, atm
Gas Temperature, F
Gas density, Ib/cu ft
Film temperature, F
Viscosity (film), cp
Viscosity (film),
Ibs/ft-hr
ii i
Reynolds No. (Re )
Gas mass velocity (G),
Ib/hr-sq ft
Molecular diffusivity
(DQ ), sq cm/sec
d
DO (at film), sq ft/hr
£•
Schmidt No. (S )
Mass transfer co-
efficient (JD)
Gas molar mass velocity
(GM), Ib-mole/hr-sq ft
Mass transfer film co-
efficient (k_)
Oxygen diffusion rate
(NO?)» lb-mole/
hr-sq ft
Carbon combustion rate,
Ib/hr-sq ft
Carbon regression rate,
in. /hr
Laminar burning time,
sec(2)
1
10.0
0. 120
1800
0.0193
1400
0.046
0. Ill
6. 300
-
0. 181
4.00
1.43
0.0090
25.0
0.459
0.0550
1.32
0.108
360
2
10.0
0.120
1800
0.0193
1400
0.046
0.111
33.4
890
0.181
4.00
1.43
0.30
31.8
7. SO*1*
0.90
21.6
8.65
Case
3
10.0
0.120
1800
0.310
700
0.0320
0.0773
77.0
1430
0. 181
2.80
0.892
0.20
50.6
10.9
1.31
31.6
_
6.0
4
40.0
0.120
1800
0.0193
1400
0.046
0.111
133.6
3.560
0.181
4.00
1.43
0.20
127.2
20.0
2.40
57.6
_
3.3
5
10.0
0.120
1800
0.0193
1400
0.046
0. Ill
1.036
890
0. 181
4.00
1.43
0.060
31.8
1.49
0.180
4.3
_
43.8
TTT
This mass transfer film coefficient value is for k , wnich does not
express the PBM included in k^.,.
' 'Time required to burn through a layer of carbon 0. 01 in. thick.
27
-------
transfer to each particle will be of the same order of magnitude as
mass transfer, so that for Case 2 an additional 8. 7 sec would be re-
quired to heat the particle to 1000°F<, The overall combustion process
would then require 17 sec, at a minimum.
An important aspect of the process is the loss of vaporized hydro-
carbons from the particles. This flow obstructs the counter-flow of
oxygen to the surface and also results in combustion above the particle.
This process must be nearly complete before the removal of the car-
bonaceous residue can be accomplished; it is not possible to compute
this time interval, except to assume that it too will require 10 to 20
sec. Therefore, a total period of as long as one minute retention may
be required to complete the combustion process for Case 2.
Whether the renovation effect can be achieved under the imposed pro-
cess-rate constraints will obviously depend on the design of the appa-
ratus used to effect combustion. The use of countercurrent flow or of
fluidized bed combustion, for final clean-up, are suggested as possible
means of accomplishing this. Then, too, it should be remembered that
most incineration processes almost unavoidably involve a feed preheat-
ing stage. This would help reduce the required residence time at com-
bustion conditions by effecting a portion of the required heat transfer
and promoting the release of some of the volatiles from the material.
It should be pointed out that the presence of agglomerates of tar and
sand much larger than 20 mesh cannot be tolerated because of the
greatly increased time required to burn off the tar. For example
(Case 5, Table 6), the time to burn an aggregate ball 2.0 inches in
diameter (assumed to be mostly tar) is computed. The temperature
of the film is the same as Case 2. Not only is the rate of oxygen dif-
fusion to the surface slower than for Case 2, but the time to burn the
1-inch thickness, which will consume the agglomerate, is 4, 380 seconds.
In order that no such aggregates of oil or inerts are present in the final
heating of the same, some practical means of separating and/or inti-
mately mixing most of the oil and sand must be devised. Probably the
most practical is to heat the incoming mixture until the oil has a low
viscosity and either coats or drains away from the sand. If this were
done, the sizing of the inert granular phase would not be critical,
since the oil-film thickness developed on the particles would be about
the same in all cases.
THERMODYNAMIC CONSIDERATIONS
Sand Composition: It is not possible to generalize on the range of com-
positions that must be anticipated in renovating oil-contaminated sand
by combustion techniques. The data obtained from the Public Works
office of the Commonwealth of Puerto Rico do, however, appear to
furnish practical guidelines. These data are summarized in Table 7.
Unlike Table 4, which furnishes data only for the earliest (17 March
1970) sampling performed, Table 7 averages the data for nearly all
beach samplings made up to 15 April 1968. Excluded are four "control"
28
-------
samples described as having "little evidence of contamination. " These
four samples were all sargas sum-free and contained an average of 11.2
wt-% moisture. It is noteworthy that, in spite of a clean appearance,
they were still found to have an oil content of 1. 9 wt-%.
TABLE 7
Average Composition of Beach Samples Taken
in Puerto Rico After OCEAN EAGLE Incident
Sargassum-Con- Sargassum-
y.1 Samples taining Samples Free Samples
Water, wt-%
Max. 79. 7 79. 7 25. 9
Min. 0.4 2.2 0.4
Avg. 26.0 33.2 13.4
Oil, wt-%
Max. 60.9 60.9 48.0
Min. 1.4 1.4 7.3
Avg. 16.7 17.1 16.3
Balance, wt-%
Avg. (by diff.) 57. 3 49. 7 70. 3
Number of Samples 33 21 12
It can be seen from Table 7 that the average oil content of all samples
(16. 7 wt-%) is in reasonable agreement with the estimate (20 wt-%)
hazarded by FWQA Oil Spill Coordinators for the sand outloaded after
the HESS-HUSTLER spill at Rehoboth Beach, Delaware.
It is apparent from the moisture data that most of the samples had
been taken from the wet foreshore region, and some,perhaps even
exhibited a supernatant phase. This was verified in a simple test
performed with beach sand obtained from Playa del Rey, California,
the characteristics of which are described later in this report. A
sea water saturated column of this material was briefly drained (by
decanting) and found to retain water in the amount of 11.0 wt-%. This
agrees with the average for the sargas sum-free samples (13.4 wt-%)
shown in Table 7 and the value mentioned earlier for the four (appa-
rently) uncontaminated samples (11.2 wt-%) regarded as controls in
the Puerto Rican samplings. The higher moisture contents found in
the sargas sum-containing samples must almost certainly have been
due to the water in the vegetable fraction.
29
-------
Heat Capacities: The heat (H) that must be supplied the oil-covered
sand to achieve the temperature for complete combustion at film tem-
perature (1400°F) is:
For water (including the heat of vaporization) from 77 to 1800°F
HT-H77 = 1,135 + 9.18 (T-212) Btu/lb
For the sand, over the same range,
HT-H7? = 0. 325 (T-77) Btu/lb
With the exception of Case 3, T will be 1800°F for water, but only
1000°F for sand, due to the difference in the solid and gas tempera-
tures. Because of this and the disparate heat capacities of steam and
sand, the heat contents of the two components at combustion conditions
will be about 50 times greater per Ib of water than for a like amount of
sand.
For the oil, the net heat of combustion, water as vapor, for various
oils does not differ greatly. Thus, the net heat of combustion to be
used from API gravity is:
API Gravity Btu/lb
10 17,400
30 18,200
60 18,900
Based on the Puerto Rico data given above, the energy balances for
possible oil-sand-water mixtures calculate as shown in the following
table.
The assumed bulk gas temperature.
30
-------
TABLE 8
Theoretical Energy Balances for the Adiabatic
Combustion of Various Oil-Sand-Water Mixtures
, . Balance,
Mixture Water, wt-% Oil, wt-%' ' wt-%(2)
18.9
38.7
57.3
49.7
70. 3
75.4
A
B
C
D
E
F
79. 7 (Max)
0.4 (min)
26.0
33.2
13.4
6.0
1.4 (Min)
60.9 (Max)
16.7
17.1
16.3
18.6
Required Heat
Input, Btu/lb
of Mixture
+ 12,342
- 10, 174
+ 1,418
+ 2,459
454
- 1,993
' 'The net heat of combustion of the petroleum fraction is assumed
to be 12,000 Btu/lb.
(2)
v 'Assumed to be sand.
These data do not of course consider systemic heat losses and other
important process-heat factors. They do demonstrate, however, that
the energy demand can change considerably with compositional varia-
tions in the input material.
Of the mixtures considered in Table 8, A and B are intended as ex-
treme cases. Here the minima and maxima for oil and water contents
found in the Puerto Rico samplings have been arbitrarily combined. In
the actual beach material tested, the driest sample taken did not, of
course, exhibit the highest tar content as well, or conversely.
Mixtures C, D, and E are based on Table 7 averages calculated for,
respectively, all of the samples (C), sargas sum-containing samples
(D), and sargassum-free (E) samples. It is interesting to note that
the last mixture, perhaps the most typical that can be expected, has
a net energy balance suggesting that no added heat would be necessary
to support combustion. This value, however, can only be regarded as
an encouraging indicator in that the performance of an actual incinera-
tor will be far from adiabatic. In these calculations no credit was
taken for the heat of combustion of sargassum. The concentration of
this weed was only determined for a limited number of samples by the
Puerto Rican workers and not at all in the case of the data summarized
in Table 7. Thus, in assuming that the balance of material is sand, a
conservative energy value is obtained.
It can be seen from these data that the operation of an incineration
system based on worst-case thermal demand (Mixture A) would be
unattractive in cost. Assuming equipment heat losses of about the
same magnitude as the heat input required by the charge, Mixture A
31
-------
material could easily require an equal weight of fuel to renovate com-
bustively. It will be desirable to avoid altogether any charging of the
renovator with sand that is very wet. A possible pre-processing tech-
nique that might have to be practiced would be to pile wet, contaminated
material near the combustor and allow it to drain for a time before
being processed.
A test was performed wherein 3 Ib of Playa del Rey Beach sand were
saturated with sea water, briefly drained and deposited on 100 Ib of
dry sand^. This material was stored in an air-conditioned room at
70°F so that evaporative moisture loss would be small. Core samples
were removed from the wet cake over a period of 24 hours. Figure 2
shows that the water loss was by far the greatest during the first hour,
when almost half of the water passed down into the supporting dry sand
column. This degree of drying by such a simple operation would doubt-
less be quite advantageous in achieving optimum performance from any
combustive renovator envisioned. Assuming utilization of this effect,
a mixture F was included in Table 8. This composition is based on
the assumption that Mixture E had been "naturally" dried, as just de-
scribed, to a material containing 6. 0 wt-% water. It can be seen that
over a 4-fold increase in energy content results.
Additionally, fuel demand can be considerably reduced by predrying
the feed in a section of the incinerator operated, preferably regene-
ratively, at a temperature well below that of the combustion chamber.
Figure 3 illustrates this, again for the case of an adiabatic effect.
The two curves represent the compositions of sand-oil-water which
are balanced-energy systems. Thus, the area above the curves in-
volves compositions that would theoretically combust without fuel
augmentation, while the areas below involve the opposite situation.
It can be seen that when the heat content for the contained water is
reduced by lowering its process temperature from 1800° to 500°F,
a significantly smaller amount of oil is required to support adiabatic
combustion. At 13% moisture, for example, the amount of oil needed
to provide self-sustained combustion is one-third less.
EXPERIMENTAL STUDY
Materials: A request for samples of bunker fuel oil (BFO) and crude
oil was made to Mr. Harry Morrison, Vice President and General
Manager of the Western Oil and Gas Association. A fifty-five gallon
drum of each type oil was promptly donated to the program through
the courtesy of the Union Oil Company. The crude oil provided was
of Alaskan origin, having a residual oil content of about 40%. Sea
water (320 gals) was obtained through arrangements made with the
U.S. Naval Undersea Research and Development Center, Pasadena,
California. This was also accomplished at no cost to the program.
Moisture content of this material was 0. 05 wt-% as received. It had
not been subjected to any processing other than excavation from the
open beach and packaging.
32
-------
c
0)
00
uo
-0—
0 15
Time, h r s
20
25
FIGURE 2. MOISTURE LOSS OF WET SAND BY CAPILLARY DRYING
-------
100%
100%
100%
0%
Oil
FIGURE 3. SAND-OIL-WATER MIXTURES FURNISHING BALANCED ENERGY
CONTENTS IN ADIABATIC COMBUSTION
34
-------
A ton and a half of beach sand was purchased from the El Segundo Sand
Company. The material supplied was unprocessed material excavated
from the beach at Playa del Rey, California. A sieve analysis (Table 9)
run on this sand showed it to be predominately a medium sand, con-
taining significant amounts of both coarse and fine sand (see Table 3).
TABLE 9
Sieve Analysis, Playa del Rey Beach Sand
Min. Particle dia.
Sieve Size, Tyler % Retained Retained, mm
16 1.0 0.991
28 5.2 0.589
48 62.0 0.295
100 30.8 0.149
200 0.6 0.074
fines 0.4
Sample Preparation; Samples for combustor testing were prepared by
placing 100 Ib quantities of sand in cylindrical polyethylene tubs (23. 5
in. dia x 10 in. depth), with and without added sea water, and spread-
ing weighed amounts of oil over the bed. When sea water was added
to the sand, the mixture was brought to a fully drained condition (mois-
ture content —11%) by tilting the tub and running off the excess water.
The combustor tests were performed by Hirt Combustion Engineers.
These tests were initially of a very basic nature and progressed to
more definitive experiments.
PRELIMINARY TESTS
Test No. 1: Samples (20 wt-% BFO on dry sand - 8. 33 Ib/sq ft) were
placed in a small container; heat was applied at the bottom. After a
period of hours, during which the sample smoked vigorously, the sand
became a hard coke-like mass. Complete cleaning of the sample
proved virtually impossible in this mode of combustion.
Q
Test No. 2: A sample was placed in a container. An open flame bur-
ner was placed above the sand, positioned to fire onto the surface. A
compressed air jet was also directed onto the surface. After ignition,
only localized burning took place, with formation of clinkers.
Test No. 3; A sample 9 was piled onto a surface. An air/gas flame
was directed onto the center of pile. Again; only localized burning
took place; however, no smoke was apparent and the sand fused into
a hard, glassy mass at the point of flame impingement.
_ . —
Of the same composition as used in Test No. 1.
35
-------
Test No. 4: A 6-in. diameter shell was filled with oil-contaminated
sand"? and temperature probes inserted below the surface at 1, 3, 5,
7 and 9-in. depths. The end of the cylinder was then exposed to a di-
rect flame from a burner. The temperature 1 in. below the surface
took 20 minutes to rise from 70°F to 700°F; the temperature 3 in.
below the surface rose to only 390°F in 20 minutes (see Figure 4).
The purpose of this and Tests No. 2 and No. 3 was to verify the
theoretical conclusion made earlier that in-place burning would be
impractical. The time requirement, based on observed thermal
conductivity, verifies this.
Test No. 5 (Series): After completion of the preceding tests, it was
apparent that agitation of the samples would be required in order to
promote rapid heat distribution. Tests were therefore conducted
using the small, rotating (23 rpm) kiln device shown in Figure 5. It
was fitted with a burner, firing into the open end, and a thermocouple
probe for measuring solid phase temperatures. By preheating the in-
terior of the device, an approximation of the effects that would be ob-
tained from a continuous feed system could be simulated.
It was found in using this test apparatus that essentially three effects
take place in the thermal-cleaning process. In the first stage, the as-
phaltic material softens at about 200-300°F and tends to coat all of the
sand particles with a layer of oil. This is advantageous in that it
causes the sand/oil mass to become rather fluid and thus relatively
easy to handle. The time required for this change proved variable
because it depends on the size and shape of the asphaltic lumps, oil/
sand ratio, contact area, etc. Generally, one minute was sufficient
to produce the transition in the samples9 tested, when exposed to
temperatures around 1600°F.
The second stage was found to consist of a volatilization or cracking
(pyrolysis) of the oil fraction; at this point it was either burned off,
if ignited, or escaped as an aero sol-rich vapor if no ignition source
were present. There was no definite threshold between these first
two stages; the combined processes took place in about two minutes
even at relatively moderate temperatures (about 1000°F).
Upon completion of the second stage transition, the contaminated sand
became a dry, black, loose mass; each sand grain was coated with a
hard carbon residue. Removal of this carbon residue (decarbonization)
proved to be the most difficult stage of sand cleaning, as predicted in
the theoretical analysis.
Tests were conducted at temperatures from 900 to 1600°F, and samples
removed at various levels of temperature. To simplify the operation,
the sand was first brought to a stage 3 condition (dry, with a black car-
bon coating) prior to insertion into the machine. Figure 6 shows the
results of these tests. The retention time decreased almost linearly
as the temperature increased, basing process completion on a visual
comparison of the product with clean sand. A residence time of one
36
-------
700 r-
600
500
o 400
-
r;
_ | _ •r" f f f
- 2 HT-2 5" 7" 9
-3 1-11 1
A •
i 1
5
300
0)
Q.
e
4)
200
100
Test
Cylinder
Positions a Nos of
Thermoc ouples
_L
8
12 14
Time, m i n
16
8 20 22
24
FIGURE 4. RATE OF TEMPERATURE RISE IN SAND-OIL COLUMN WITH IMPINGING FLAME
-------
Thermocouple Tube
A ir or Flame Inlet
20 gal. Can
\ v
\
\ \ \
FIGURE 5. ROTARY KILN TEST APPARATUS
38
-------
1400
1300
12 00
00
0>
Q.
E
o>
000
900
_L
j_
I
468
Retention Time , min
10
FIGURE 6. TIME-TEMPERATURE RELATIONSHIP IN CLEANING CARBON-COATED SAND
39
-------
minute at 1400°F for complete combustion proved to be consistent with
the prediction based on theory.
SUSPENDED PHASE TESTS
Test Equipment: Utilizing samples of stage 3 (carbon-coated) sand10,
tests were conducted to determine the time-temperature relationships
involved in producing clean sand by entraining the material in a stream
of hot air. The apparatus (Figure 7) used was a cyclone separator ar-
ranged so that a stream of hot air was drawn through a 10-ft horizontal
run of 10. 5-in. (dia) pipe into the cyclone and out the top through an ex-
haust fan. The sand was introduced into the air stream at midpoint in
the pipe, where it became suspended in the gas, and traveled to the wall
of the cyclone and then settled to the bottom of the cone. In order to
prolong the retention of the sand in the cyclone, the bottom discharge
port was left open so that cold air also entered at this point due to fan
suction and caused the accumulating solid phase to be thrown turbu-
lently back into the combustion zone. The SCFM of air introduced at
the bottom port represented only about 2% of the total gas flowing into
the cyclone. The following tests describe the results.
Test No. 6; As a preliminary investigation, about a pound of tar-
contaminated sand? was injected into the entrance point of the hot air
stream, which was maintained at about 1200°F. It fell to the bottom
of the pipe, quickly shaped itself to the pipe contours and then ignited.
It burned to a hard, black, coke-like mass covered with a thin layer
of loose clean sand. There was no tendency for the loose sand to be-
come entrained in the gas. It was concluded from this that the feed
material would have to be in the Stage 3 form in order to be processed
satisfactorily.
Test No. 7; Samples of Stage 3 sand were injected into the air stream,
which was now maintained at 1300°F, and held suspended for various
lengths of time after entering the cyclone. Clean sand was obtained
in the many runs made after a residence time which varied from 60 to
300 seconds. Removal of a sample of the cyclone's contents was ac-
complished by plugging the bottom discharge port, fully opening the
air valve and, after material had collapsed into the tail-piece, shutting
the valve. When the plug was removed, several ounces of representa-
tive material would drop out of the tail-piece to be collected.
Test No. 8; Test No. 7 was repeated with the inlet gas at 1500°F.
Clean sand resulted after a retention period which ranged from 10 to
90 seconds.
Test No. 9t Test No. 7 was again repeated, this time with the air
stream at 1600°F. Cleaning was completed after a maximum of 40
seconds in the cyclone. The shortest retention time achieved (10 sec)
was no better than that which was obtained at 1500°F.
With the exception of Test No. 6.
40
-------
Air L.
Heoter
Sand in
i r
he rmocoupl e
Viewing For
,J
Scale = 20 = 1
^ to
Blower
Sand out
FIGURE 7. CYCLONE TEST APPARATUS
-------
INTERMEDIATE CLEANING STEP
Test No. 10: After determining the conditions needed to convert Stage
3 sand to usable product, tests were conducted to parameterize the
process required to convert the raw input to Stage 3 material. This
test was an exploratory effort in which various attempts were made to
determine the rate at which hot sand beds would pyrolyze and combust
the oil contaminant present in the feed sand. This was done by adding
small weighed quantities of sand" into the tumbler (Figure 5) which
contained a hot sand bed of known temperature and weight (usually
about 27 Ib). It was soon apparent that the temperature of the bed could
not easily be held constant with the combustor configuration used. How-
ever, the tests did indicate that 500°F was the critical temperature be-
low which combustion was quenched and irreversible cooling resulted,
unless auxiliary heat were applied. In all of the tests conducted, any
smoke generated during steady state conditions could be essentially
eliminated, except at high temperatures, by blowing a gentle stream
of air into the tumbler over the hot sand bed.
Test No. 11; In order to control more evenly the process of producing
Stage 3 sand, the refractory lining of the tumbler was removed to in-
crease heat dissipation and a cover installed to confine the combustion
of the formed gases. The latter was fitted with a hole through which
the air pipe was inserted, leaving a clearance of about 1 in. around the
pipe. The air supply was connected so that both its temperature and
flow rate could be controlled. With this test system it was determined
that:
1. A bed of sand heated to about 650°F converted the feed"
to Stage 3 material almost as fast as the tumbler could be
charged. This rate was found to be about 57 Ib/hr for the 2. 67
cu ft rotating chamber.
2. The heat content of a bed of sand weighing 27 Ibs was
sufficient to support the combustion process at the above-
indicated feed rate without applying external heat.
3. The temperature of the bed could easily be controlled
by varying the quantity of air introduced into the tumbler.
The air temperature was not varied to effect this control,
only the flow rate. Satisfactory results were obtained utili-
zing ambient temperature air.
4. The volatiles produced by the pyrolysis process burned
smoke-free at the mouth of the tumbler with no visible flame
being noted on the inside of the chamber. The burning off-
gases behaved much like a non-aerated natural gas flame.
Test No. 12; Up to this point all of the tests were performed using
dry beach sand? contaminated with BFO. A practical renovation
system would of course have to deal with surf-wetted feed material.
42
-------
Test No. 11 was therefore repeated at 700°F using sand saturated with
sea water (moisture -11 wt-%) to which BFO was added to furnish an
oil content of 20 wt-%, dry basis. Expectedly, the bed could not be
made to sustain the pyrolysis process without the use of supplementary
heat. Thus,^the transition from Stage 1 to Stage 2 sand, dehydration
and tar fluidization and dispersion, will require an energy input; the
transition from Stage 2 to Stage 3 will not, if sufficient oil is present.
The specific energy input for the dehydration effect will obviously be
predicated on the moisture content of the feed and should conform
reasonably well11 with that which was predicted in the theoretical
analysis.
Test No. 13; The residence time required for the dehydration and oil
dispersion step was next estimated. This was done by feeding the com-
bustor variously sized portions of oil-covered, wet sand and observing
the time required to induce combustion. The combustor contents were
held at nearly constant temperature by controlling the flow of natural
gas to a flame impinging on the outside bottom of the rotating drum.
When processed at 700°F, sand samples containing 6 wt-% water and
10 wt-% BFO required about 2 minutes before combustion (Stage 2
condition) was initiated.
Test No. 14; In this series of tests, dry sand containing lower load-
ings of BFO and Alaskan crude were tested to determine the minimum
petroleum content at which the unaugmented conversion to Stage 3 sand
would occur. The test conditions were the same as those employed in
Test No. 11. It was found that, regardless of the form of oil added,
as little as 6 wt-% petroleum would provide sustained combustion.
Miscellaneous Tests; Mr. J. Parsons of Hirt Combustion Engineers
obtained some well-aged, oil-contaminated sand samples from a beach
in the Santa Barbara area. Other than being more asphaltic and ag-
glomerated than the artificial samples prepared, their behavior in the
combustor, per the procedure of Test No. 11, was essential identical.
The residence time before acquiring the Stage 2 condition appeared to
have been only slightly longer.
Another qualitative test involved observation of the behavior of emulsi-
fied Alaskan crude oil on wet sand. The emulsion ( ~ 6 wt-% water)
was prepared by vigorously agitating the crude oil in sea water with
an ordinary mechanical stirrer. After about 4 hrs, the entire mixer
contents, emulsion and water phases, were then poured onto dry sand,
which was then drained briefly to free it of excess water. The dehy-
dration of this (unstable) system required considerable auxiliary heat
input but the emulsion itself collapsed and dispersed almost instantly
after contacting the hot bed.
11
Assuming corrections for systemic heat losses are applied.
43
-------
SECTION VI
SYSTEM DESIGN
Overview: Initial tests proved that oil-contaminated sand can be ther-
mally cleaned to furnish a product that is hardly distinguishable from
sand that has not been contaminated. This testing also provided data
needed for the design of practical incineration systems; this included
information pertinent to handling times, fuel and power demand, heat
penetration, and temperature requirements.
These tests further established that the contaminated sand progresses
through three stages in the thermal cleaning process. The first stage
involves dehydration and oil dispersion. It takes place when the oily
sand acquires a temperature of about 300°F> wherein the oil-fraction
becomes quite fluid and coats the sand and excess oil flows from the
mixture. When the temperature of the system is increased to about
600°F, all of the oil fraction will vaporize and/or pyrolize to leave a
black solid coating on the sand grains.
It was established that if the input material (moisture content <6 wt-%)
is surrounded by hot sand or air at 700°F, the first two effects will be
accomplished in approximately two minutes. This parameter, however,
is a function of heat transfer rate, which in turn is affected by feed rate,
apparatus configuration, air flow, feed material characteristics, and
many other variables.
The vapors, other than water, emitted during the first 2 stages of tran-
sition have considerable fuel value; utilization of this energy source is
of prime importance in apparatus design.
The final cleaning stage consists of exposing the black, dry sand to suf-
ficiently high temperatures and sufficient air to burn off the carbon coat-
ing the sand grains. Tests showed that this could be accomplished in
about 60 seconds at 1400°F. Here again, many variables will influence
the time requirement.
In all of the stages of cleaning it was found that agitation of the oil/sand
mixture was required in order to avoid the formation of undesirable
clinkers and to obtain reasonable cleaning rates. The known methods
of accomplishing the required heating and agitation considered on this
project were: rotary kiln, fluidized bed, and cyclonic cleaning.
The input material to be processed must be expected to vary consider-
ably in its chemical and physical properties. In terms of mechanical
state, the input may vary from a fairly fluid solid/liquid mixture to a
feed which contains large plates of rigid asphaltic contaminant. The
input should normally contain more than sufficient ( > 6 wt-%) petroleum
to permit the Stage 2 transition to occur without introducing supple-
mentary energy. The moisture content of the input is a critical para-
meter and should be held to a minimum. It is not unreasonable to
45
-------
expect, however, that an average of 6 wt-% moisture can be obtained
if the excavation operation is managed in a specific manner. Another
important characteristic of the feed material will be manifested as
unavoidable inclusions of variously dimensioned foreign and natural
objects. With the obvious exception of gross pieces, the rejection of
such items prior to combustive processing would not be possible. The
incinerator would therefore have to be designed to screen the process
material at some point between input and output, or to permit the com-
plete passage of all noncombustibles that are introduced.
Finally, throughput requirements must be considered. Unfortunately,
the linear rate of beach cleaning cannot be predicted on the basis of the
quantity of sand that can be decontaminated per unit time. A number
of factors preclude such a correlation, as was discussed in a previous
section. It was nonetheless pointed out that under conditions that one
can reasonably expect to find, a feed rate of between 38 and 53 tons per
hour will result in the decontamination of 1 mile of beach per 8 hour
work day. Assuming that this linear renovation rate is an acceptable
one, the input specification for a full-scale system has arbitrarily been
set at 60 tons/hr.
Design Analysis: Stage 1 and 2 transitions of the cleaning process are
the most difficult to accomplish in a short time, primarily because of
configurational constraints. The rotary kiln represents, by far, the
readiest solution to these problems; it provides containment of the mate-
rials, while allowing agitation, conveyance, and gas/vapor containment.
For this reason, process design was developed on the basis of the use
of a rotary kiln type of apparatus for at least the production of Stage 1
and 2 material.
Fluidized bed methods are feasible for the final (Stage 3) cleaning. It
was calculated, for example, that only 15,000 ACFM of air (1400°F)
would be required to fluidize a one-ton sand bed. Assumed here is an
average carbon-coated particle diameter of 0.03 in. , a solid phase
density of 100 Ib/cu ft, and a superficial gas velocity of 4. 5 ft/sec.
This volumetric flow rate (and, thus, the fan power required) is con-
siderably lower than would be necessary for a rotary kiln configuration
of the same output capacity.
A disqualifying drawback of the fluidized process is the fact that the in-
put would have to be screened so that the bottom of the bed would not
accumulate a blockage of larger-diameter noncombustible material.
The utilization of a screening process was studied in some detail but
it was concluded that the incorporation of such components would create
more problems and functional risks than the use of a fluidized bed could
justify. Specifically, no suitable arrangement could be envisioned that
would permit the simultaneous screening and rejection of oversized ob-
jects and the collection of decarbonized sand elutriated from the bed.
Cyclone Cleaning: While considered in the experimental portion of this
project, cyclone cleaning was found to be practical only for combusting
46
-------
Stage 3 sand. Because of the extreme abrasion problems encountered in
handling sand in a cyclone cleaner, and in view of the fact that the initial
cleaning stages would still have to be accomplished in a rotary kiln type
of device, this method was not considered further. This analysis there-
fore considered design parameters and configurations based exclusively
on the employment of a rotary-kiln type of thermal cleaning device.
Batch Kiln Process: This process would consist of the use of a single
rotary kiln, in which an amount of contaminated sand would be loaded,
processed and the clean sand periodically dumped. While production
rate would be low, this type of process does have certain advantages.
For example, control problems are negligible and material handling
would be simplified. This type of processing could be accomplished by
modifying presently available equipment, specifically a cement mixing
truck. By adding a refractor lining, burner, fan, and a fuel atomizing
pump and compressor, it would become a batch-type rotary kiln. If
air pollution problems are to be considered, an after burner would
definitely also be required. Figure 8 illustrates how these modifica-
tions could be accomplished. Considering a 10 cubic yard transit
mixer, at 20% fill12, it is estimated that 2, 800 Ib/hr of sand could
be processed in a single mixer. Thus it would require 7 mixers this
size to produce an output of 10 ton/hr. Fuel requirements for this
processing rate would definitely be higher than for the continuous
process (discussed in the next subsections), since very little of the
heat available in the oil contaminated sand would be utilized. Power
consumption, however, would be somewhat lower. Labor costs would
obviously be higher. Only a single operator will be required for a con-
tinuous system (sub- or full-scale), while at least two men would prob-
ably be needed to tend every four or six trucks. Equipment costs can
be estimated based on two approaches: the purchase and modification
of new equipment1-' or the temporary retrofit of rented vehicles. If
the former practice were followed, the equipment cost would be con-
siderably higher than for a continuous system. The latter (rental)
approach would be far cheaper but would entail problems; e. g. , delay
while retrofitting and acquiring suitable vehicles.
CONTINUOUS KILN OPERATION
Process Description: Rotary kilns, both refractory-lined and unlined,
are common items of industrial equipment. The kiln consists of a tube
rotating on its longitudinal axis, through which solids or slurries are
conveyed with or against a stream of process gas. The interior of the
shell can be equipped with a variety of material lifters, baffles, con-
veyors, or material showering arrangements to provide virtually any
Higher fills can be employed in the batch process than in continuous
kiln operations.
13New 10 cu yd transit mix trucks are priced at about $25, 000.
47
-------
Stack
Refractory Lined
Cement Mixer
Hopper
After
Burner
Di scha rge
Chute
Fuel Pump
FIGURE 8. BATCH TYPE SAND CLEANER
-------
desired material flow characteristics. In general, the function of the
rotary kiln is insensitive to the material being handled; rocks, sand,
bottles, seaweed and driftwood (within reasonable limits) will not af-
fect operation adversely. In the present process context, both counter-
current and concurrent gas/feed flows are applicable. The concurrent
arrangement is indicated for the first portion of the system (Stages 1
and 2) in order to remove volatilized oil constituents and pyrolyzates
from the sand without recooling, while countercurrent flow is suggested
in the high temperature stage, in order to operate at lower gas tempera-
tures. In addition to this two-part rotary kiln, a combustion chamber
is needed to provide heat for the process and to burn oil vapors given
off in the pyrolysis stage.
Figure 9 shows the basic flow in this type of apparatus. The relative
flow rates from the stack and rotating-cleaner blower would be deter-
mined by the speed of the latter, which, in turn, would be controlled by
gas temperature. Normally, the feed for this process from a hopper or
conveyor would be continuous. An apparatus producing the indicated ef-
fects would be sufficient to clean sand. To minimize fuel requirements,
however, several alternatives are available which should be considered
for the final design. These are:
1. Hot sand recycle
2. Hot air recycle
3. Fresh air preheat with sand cooler.
All of these alternatives may be practiced on a continuous feed basis.
Information, data, drawings, etc. , discussed herein are based on a
nominal (minimum) clean sand output of 10 ton/hr, rather than that of
a full-scale system. This was done because the output estimates are
highly conservative and actual results obtainable could well exceed (by
as much as a factor of three) performance expectations. In addition to
avoiding oversizing, it was the intention to provide characteristics for
a system suitable for early field test evaluation which would involve a
more reasonable R&D capitalization.
Hot Sand Recycle: This process would consist of reusing a portion of
the hot, clean sand output as a bed for the contaminated input, thus pro-
viding a heat exchange medium for the Stage 1 and 2 portions of the pro-
cess (see Figure 10). Since the hot recycled sand actually becomes
additional feed material, this process would reduce the input rate of
contaminated material by an amount equivalent to the amount of sand
recycled. The volume of the kiln would therefore have to be enlarged
by a factor representing the square root of the fractional throughput
displaced by recycle sand. The amount of recycle sand needed will
depend on the heat required to produce the first two stages of sand
renovation. If no auxiliary burner is included to produce the first stage
transition, this would almost equal the input of contaminated sand. Be-
cause of the equipment size limitations imposed by the mode of appli-
cation envisioned, this concept has been reserved but not discarded
insofar as the present level of design effort is concerned.
49
-------
Ul
o
Hot Air Discharge
(Flue Gases)
880° F
Stack
t
a mi n a t e d
id Feed
Fuel
Air _
\s U III I
Ai
Hot Air
Sand
1 U
r
N
t
Rotating
Cracker
D 1 1 V II
Fuel _
Hot Gases
-^ i _ , ., „ ^-.
<$£?
Sand 5
T
C ombust ion
Chamber
30° F
V.
R o t a t i
C leane
J
\ 800° F Ga s es
Hot Gases
Sand
ng T
r Cl
1
ean
S<
500°F
FIGURE 9. BASIC CONTINUOUS SAND CLEANING FLOW DIAGRAM
-------
Contaminate d
Sand Feed
Air
Hot Air Discharg e
(Flue Case s)
880° F
Stack
Combustion
d
X.
^
— ^
Ai
Hot Air
Sand^
t i
C
r
v
S
0
i (
Fuel
Hot Gases
^ 1 ^
*[ P ^
*VLS
/^
Sand 5 8
fating
i c k e r
Hot Sar
1
Combust! on
Chamber
o
0 F
R
C
id 1500° F
^
ota
e a
J
!800°F Gases
Hot Gases <^-
Sand
ting A
ner T
i
Clean Hot
Sand
500°F
FIGURE 10. HOT SAND RECYCLE FLOW DIAGRAM
-------
Hot Air Recycle: This effect would consist of utilizing a portion of the
output from an augmented hot gas generator to provide heat, not only
for the final sand cleaning stage but for the initial stages of cleaning.
This arrangement would eliminate the requirement for an additional
burner and make the use of hot sand recycle unnecessary. Figure 11
diagrams the concept. The preheat air temperature can be raised to
any level the physical limitation of the equipment will allow and would
be maintained constant by the addition of tempering (fresh) air ahead of
the fan. This arrangement would necessitate an increase in the size of
the equipment, so that the linear rate of gas flow could be held below
the conveying velocities. For this purpose, 750 ft/min was selected
as the maximum allowable velocity through the kiln.
Fresh Air Preheat with Sand Cooler: If the clean sand were discharged
at final process temperature (~1500°F), certain handling problems and
safety hazards would exist. This difficulty would be considerably re-
duced by the addition of another rotary kiln, through which clean fresh
air were passed countercurrently against the discharging sand flow.
Sand effluent temperatures as low as 250°F could be reasonably ob-
tained in this manner. The exhaust air, which would be heated to about
1000°F in the process, could then either be discarded or, to conserve
fuel, used in the first stages of the cleaning process.
Equipment Description; A complete sand cleaner, as herein envisioned,
consists of: cracker, combustion chamber, cleaner, appurtenant equip-
ment (fans, burners, fuel source, power supply and controls) and an op-
tional sand cooler. A brief description of each follows:
Cracker; This is a rotary shell wherein the first two stages of cleaning
take place. Contaminated sand is fed into one end through a vertical
hopper, along with sufficient concurrent hot gas to complete Stages 1
and 2. The heat is supplied either by a burner, hot recycled sand, or
hot recycled air. Approximately 6. 2 x 10° Btu/hr would be required.
This figure includes no credit for oil combustion energy and assumes an
input of sand of 20, 000 Ib/hr (inert basis), together with the maximum
expectable amount of oil (5, 000 Ib/hr or 20 wt-%, dry basis) to be heated
(but not burned) and water (1600 Ib/hr or 6 wt-%) to be vaporized and
heated in both the liquid and gaseous phases. This would, of course, be
decreased by any combustion heat that is provided by the oil and/or re-
duction in water content. Since, at the probable minimum oil content
(6 wt-%, dry basis), the oil will release about 23 x 10" Btu/hr in the
presence of sufficient oxygen, it will be necessary to limit the amount
of air that is introduced to prevent complete combustion from taking
place in this stage. If the full 6. 2 x 10 Btu/hr required in this stage
could be obtained from the oil combustion, it would require (at 1 cu ft
air/100 Btu) 1,030 SCFM air in this zone. Similarly, if no oil were
present, an auxiliary heater system would be required to furnish a like
amount of air. The off-gases from the vaporizing and pyrolyzing oil
(maximum amount) taken together with the combustion air and water
vapor produced would represent a maximum volumetric flow, at 600°F,
of about 11, 000 ACFM. This would have to be ducted in a kiln diameter
52
-------
Contaminated
Sand Feed
Ul
OJ
Tempering Air
70°F
Hot Air Discharge
(Flue Gases)
880° F
i
Stack
Hot Air 1800 F
Combustion
Air
Fu e
Hot Air
Sand
t
Rotati n g
Cracker
c
ombust ion
C hamber
Sand 580 F
1800 F Gases
Hot Gases
Sand
Rotating
Cleaner
I500F
FIGURE 11. HOT AIR RECYCLE FLOW DIAGRAM
-------
of no less than 4. 5 ft in order to maintain the gas velocity through the
shell at less than 750 ft/min. Length, based on 10% fill (normal for
rotary kilns) and a retention time of 6 minutes (furnishing a 3x safety
factor) would be about 15 ft. Since the temperatures in this portion
would probably not exceed 800°F, the shell could be constructed of mild
steel and without incorporating a liner. Lifters would be provided for
proper agitation and transport.
Combustion Chamber: The combustion chamber for furnishing process
heat would be a non-rotating, refractory lined cylinder, sized to accom-
modate a heat release from combusting petroleum vapors as high as
9x10 Btu/hr, based on a maximum envisioned oil input of 5, 000 Ib/hr.
An auxiliary burner would be included to provide heat for the process
when the oil content of the sand is low. This combustion chamber would
operate at sufficiently high temperatures to avoid air pollution problems.
The discharge gas would be maintained at about 1800°F by the addition
of appropriate quantities of diluent air. A spark ignited continuous
pilot would be provided to insure a source of ignition. Combustion air
for the final cleaning stage would be supplied by a. fan, which would
also remove vapors from the cracking stage.
Sized for maximum required heat release, the combustion chamber
would have a volume of about 530 cu ft. It would be a 20 ft long cylin-
der, 6 ft in diameter, lined with a 4. 5 in. thickness of refractory
material.
Cleaner: The purpose of the "cleaner" is to accept the black, carbon-
coated sand flowing from the cracker and increase its temperature from
about 600°F to 1500°F, thereby oxidizing the carbon from the sand grains
and furnishing a discharge of clean 1500°F sand. Hot (1800°F) air from
the combustion chamber would flow countercurrently with respect to the
sand flow in this kiln. Because of the higher operating temperatures,
the rotating shell for this stage should be constructed of a high tem-
perature alloy. This design feature would probably promote minimum
discoloration of the output. Dimensioning of the shell would be on the
basis of a 10% fill, 3 minute retention time, 750 ft/min gas velocity,
and a temperature drop of,900°F in the hot gases as heat is acquired
by the sand. About 4x10 Btu/hr must be transferred to the sand;
this would require about 4000 SCFM of gas, or 18,000 ACFM at 1800°F.
To maintain a gas velocity of 750 ft/min in the shell would require an
inside diameter of 6 ft. Because, however, the gases are cooled in
passing through the kiln, the volumetric discharge of gas would only
be 10, 000 ACFM (at 900°F). Thus, if the cleaner were sized down
to have, conveniently, the same diameter (4. 5 ft) as the cracker, the
discharged gas velocity would be 630 ft/min. This would preclude the
blow-back of any solids into the cracker. At the (1800°F) sand dis-
charge end of the kiln, however, conveying velocities would be ex-
ceeded. This would result in some material being carried back into
the cleaner, but not its full length. A zone containing a heavier con-
centration of sand fines would be formed but the normal output flow
would be reestablished after steady state conditions were developed.
54
-------
The 4. 5 ft (dia) shell would require a length of 7 ft, in order to allow
room for drive mechanisms, trunions, etc. Rotary drier experts ad-
vised that a length of 12 feet would be more suitable for this shell.
Lifters would be provided in this shell to obtain agitation. External
shell temperatures could be reduced with refractory, if required.
Appurtenant Equipment; In addition to the three components previously
discussed, several major items of equipment would be required to com-
plete the system. These would include: air handling equipment, bur-
ners and fuel source, power supply and controls.
Air Handling Equipment: Fans will be required to provide hot air for
the cracker and the combustion chamber, and motive power for air flow
in the cleaner. Two fans will be required to perform these functions.
One fan would supply combustion air (15,000 SCFM) to the combustion
chamber and draw air and vapors through the cracker. The other fan
would draw air through the cleaner (4, 000 SCFM). Both should be of
high temperature alloy construction for maximum service lift and to
permit the handling of the high temperature air. The first fan would
require a 15 HP motor, and the second a 5 HP motor.
Burner and Fuel Source: Because of its availability, diesel oil has been
selected as the auxiliary energy source for the machine. Air-atomized
oil burners would be used to provide heat injection. Two burners would
be required in the basic unit; one in the cracker and the other in the
combustion chamber. The maximum heat required for these stages,
when none is available in the sand influent, would be slightly over 10
Btu/hr, or 72 gph of oil. Diesel fuel would also be used to supply a
motor/generator set to furnish electrical power to the fans, pump,
compressor, kiln rotating motor, and controls. A hydraulic pump and
motors might also be used for this purpose. A total of 45 HP would be
sufficient to provide the required power, which is itemized as follows:
Component Horsepower
Fans 20 HP
Pump 5 HP
Compressor 5 HP
Kiln Drive 10 HP
Control and Miscellaneous 5 HP
Total 45 HP
The fuel consumption of the motor-generator set would be approximately
2 gph.
Controls: Two temperature control systems will be required. One
would control the temperature of the cracker by varying the fuel and
air fed to the burner. The second would control the hot sand discharge
temperature of the cleaner by varying the temperature of the gas led
into the cleaner. This would be done by adjusting the amount of dilution
air mixed with the gas discharged from the combustion chamber. In
55
-------
addition, temperature limited fail-safe devices should be provided in
each chamber to prevent equipment damage in case of malfunction.
Each burner should also be equipped with combustion safeguards to
insure safe operation.
Sand Cooler: The physical size of the optional sand cooler would depend
on the maximum allowable sand discharge temperature. If the cooler
were the same length and diameter as the cracker and cleaner together,
retention time would be 9 minutes. If the air is drawn through the cooler
(as shown in Figure 12), the maximum air temperature rise permissible
through the cooler would depend on the limitations of the fan. Assuming
a maximum fan temperature rating of 1000°F and an input (hot day) air
temperature of 100°F, the AT would be 900°F. If the exiting gas velo-
city were controlled at 750 ft/min, the volumetric flow would be 12, 000
ACFM at 1000°F, or 4, 250 SCFM. Based on the given AT, the heat
transfer to the gas phase (neglecting small heat capacity corrections
for changes in chemical composition) would be about 4. 3 x 10 Btu/hr.
This would be sufficient to lower the sand temperature from 1500°F to
about 415°F. This could be further reduced by increasing the diameter
of the shell and (correspondingly, based on cross-sectional air increase)
the air throughput. Another approach to achieving cooler sand output
would be to relocate the fan from the gas outlet to the air input. This
would allow the outlet gas to come to much higher temperatures. It
might also pose design problems if the system proves to have a high
Reynolds number and considerable air is lost through the sand dis-
charge opening. As previously mentioned, however, sand temperatures
as low as 250°F are not unreasonable to expect. If, in fact, this para-
meter becomes important in terms of operational efficiency or hazard
control, a simple remedy would be to spray sea water onto the piling
output. This could be done with minimum increase in operating cost.
Such a process might also be worth while to minimize thermal damage
to the output conveyor belt.
System Configuration: Figures 13 and 14 illustrate two possible ar-
rangements of the" apparatus and indicate approximate sizes. Figure 13
shows the basic unit, in wnich heat recuperation from the output sand is
not involved. Figure 14 indicates the size increase required for the ad-
dition of hot sand recycle. Addition of hot air recycle is not shown.
However, the size would be approximately that shown in Figure 14.
The system size will increase approximately as the square root of
throughput, so that 100 tph units will be approximately three times as
large as the 10 tph systems illustrated in Figures 13 and 14. The
weights of these combustors have also been estimated and are shown
as a function of throughput in Figure 15. As can be seen, a 10 tph
system would weigh about 13 tons and a 60 tph combustor about 38 tons.
Air Pollution Considerations: The bench tests conducted demonstrated
that, if sufficient oxygen is provided for combustion, the oil in the in-
fluent can be burned with no visible particulate emissions. Dust gene-
ration from air flow through the sand is possible. In this case a small
cyclone dust separator might be required. Fan B (see Figure 13) motor
56
-------
Ai r Bleed
Hot Air Discharge
(Flue Gases)
880°F
Stack
Contaminated
Sand Feed
Ul
C ombusti on
Air
f
Hot Air
Sand.
Fu e
Hot Gases
C
ombusti
C hamb
on
er
Sand 58OF
Rotat i n g
Cracker
Hot Fresh Air
Air
Sand
Cooled
Sand
I 800 F Gases
Hot Gases
Sand
Rotating
Cleaner
Cl ean Sand
1500 F
Rotating
Cooler
FIGURE 12. FRESH AIR REHEAT WITH SAND COOLER
-------
40'-0"
/-Control Panel
jf ^r-Combustion Chamber
Power
Supp ly
Hopper
€
to
Cracker Cleane
Clean Sand Discharge
8'-0"
8J-0"
J.
Plan View
9'- 0'
Cleaner Disch Stack
r
C ombust ion
C hamber
Disch Stack
_ ii _ ii
Fan B
/^i
Side Elevation
FIGURE 13. BASIC ROTARY KILN SAND CLEANER
58
-------
8-0" mg>
8-0" max.
Combus t ion
Chamber
Control
Panel
Plan View
Operator
\ Recycle Sand Conveyer
Elevati on View
Cleaner
FIGURE 14. ROTARY KILN CLEANER WITH HOT SAND RECYCLE
59
-------
80
70
60
50
40
O
<
Q.
30
20
I I
20 30 40
WEIGHT, Tons
I I
50 60
FIGURE 15. ESTIMATED WEIGHT OF BASIC SAND COMBUSTOR
60
-------
horsepower would then have to be increased to 10 HP to provide addi-
tional static pressure for flow through the cyclone.
Other Characteristics: It was felt that the detailed consideration of
certain system characteristics would be premature at this time. En-
gineering aspects dealing with optimum transportability, beach mobility,
and sand handling to and from the system were therefore treated in only
a preliminary manner. Detailed analysis of these design features should
be deferred until after a sub-scale combustor has been built and tested
to provide the needed design specifications for a full-scale system. A
tentative projection of how these other requirements might be managed
is, however, discussed in the following paragraphs.
Transportability: Assuming that the subscale system is capable of ope-
rating at only its minimum rated throughput (10 tons/hr), a full-scale
incinerator (38-53 tons/hr) would, theoretically, have to be 1. 95-2. 30
times larger. The dimensions of the subscale sand cleaner are, how-
ever, already such that transporting it by air would only be possible if
the system were dismantled and moved in several aircraft. If still
larger components are indicated, it is doubtful if air transportation
would be physically possible or economically tolerable.
It may eventuate that the subscale system will satisfactorily perform at
its maximum projected throughput (30 tons/hr) and that this production
rate would be accepted as adequate for routine beach cleaning opera-
tions-'-'*. Even if scale-up proved unnecessary, the question of whether
flying such heavy and bulky equipment would be advisable must still be
considered. In all likelihood, the sand cooler (if used), the combustion
chamber stack, the pump and generator set, and, possibly, the cracker
would have to have detachable features to permit separate or repositioned
shipment. Providing this capability were feasible, it could increase
equipment costs by as much as 25%. A further drawback would be that
a significant portion of the time gained by air freighting the equipment
would be lost in having to reassemble it at or near the use point.
Although trucking the intact equipment to a disaster scene might require
an additional day or two, the device could be put onto the beach directly
from the flat-bed carrier. The money saved per trip by this method of
drayage could amount to as much as $10,000. Thus, a careful analysis
of these factors will have to be made prior to hardening the ultimate
configuration of the renovator and determining its mode of deployment.
Another possibility is the shipment of the combustor from a seaport
storage-point to a spill site by amphibious seacraft. The danger im-
plicit in this kind of delivery is well known even for specially-designed
(combat) vehicles. A tractor-drawn, non-ruggedized machine might
not survive an unfavorable landing situation.
A possible precedent is suggested in Reference 37.
61
-------
Beach Mobility: The conceptual designs developed herein present an
incinerator that is sled mounted. Such an arrangement will doubtless
be adequate for restricted field test purposes. In actual renovation
operations, however, the system will probably have to be moved several
times a day in order to keep the incinerator reasonably close to the ex-
cavation equipment. While it would clearly be impractical to develop a
self-propelled machine, it may be possible, under certain beach condi-
tions, to operate the renovator on the bed of the carrier which delivers
it. However, additional provisions would still be needed to permit its
easy movement along a beach by being drawn by a tractor or similar
vehicle. A sled base may well prove inadequate for this purpose. An
appealing traction form would be the tracked (continuous belt) assembly.
It should be pointed out that, in moving the machine along the contami-
nated area, it should not be allowed to cool. This would cause reheating
delays and additional fuel costs. Thus, while being moved, the com-
bustor would be fired by its auxiliary fuel source.
Sand Handling: The subscale system discussed herein is a hopper-fed
device which discharges its cleaned output directly onto the ground.
For field testing purposes it will be necessary to build up a sand ramp
adjacent to the incinerator so that front end loaders can elevate their
shovels above the 13-ft high hopper. Similarly, the sand discharged
will have to be removed periodically to prevent the incinerator output
opening from becoming blocked. Some type of sand-conveyance device
will have to be provided when a fully operated system is designed. The
configuration this ancillary equipment should assume will be better
recognized after ultimate incinerator requirements are defined. It is
expected, however, that certain of the sand moving techniques now being
evaluated by URS Research Co. for the FWQA will be of value for that
purpose.
Equipment Costs; The costs of fabricating test devices of the type here-
in described will be very close to those of fully functional units, excep-
ting two cost factors. These are the feeding and outloading systems,
and the traction assemblies, which have not been considered. Neither
of these cost items should be excessive, but further design analysis
will be required to verify this conclusion-* ->. The cost function for the
basic system is shown in Figure 16.
Operating Costs: In normal costing, many factors are applied that
would be difficult to consider in terms of the present system. The
equipment described here would be employed for a relatively short
time following an oil spill disaster and costs would be indemnified by
the party found responsible for the damage. The responsible party
could not be expected to pay for the equipment on a fair rate-of-return
basis nor could he be allowed to share pro rata incapital cost annuali-
zation. The situation is somewhat analogous to that wherein the owner
Based on standard pricing practices for chemical feed equipment, an
ROM estimate of $7500 was derived for the constant-feed input system
and output conveyor.
62
-------
80r
70
60
50
Q.
>' 4°
h-
o
Q.
3 30
20
I 0
25
50 75
COST, I03$
100 125 150
FIGURE 16. ESTIMATED CONSTRUCTION COSTS OF BASIC SAND COMBUSTOR
63
-------
of a fire -damaged home would be required to pay the fire department
for services rendered on an annualized basis.
Annual capital costs are determined by annualizing the initial equip-
ment investment, using an appropriate capital recovery factor and then
adding to this rate due compensation for interest, insurance, taxes,
etc. When these costs are returned by users, the equipment use-rate
determines the degree of recovery that must be applied. If the equip-
ment is used one month out of the year, the normal rental rate must
be increased by a factor of 12.
Based on an estimated equipment life expectancy (N) of ten years, the
amortization rate (A) would be 14. 9%, if an 8% rate of return (r) is
expected. That is,
(1 + r) -1
Federal taxes should be set at about 3%, insurance at a minimum of
0.25%, and state and local taxes at about 1. 9%. Annualization would
then be 20%. From Figure 16, it is seen that a 10 tph combustor will
cost about $45, 000 in its basic configuration, and probably about $60, 000
when equipped with a suitable feed conveyance system and an appropriate
traction assembly, neglecting engineering and design costs. Annuali-
zation would therefore be $1,000 per month assuming 100% utilization
(plant factor). If the equipment were only used on spill clean-ups 30
working days per year, which is perhaps conservative, the annualiza-
tion charge would be $5.00 per ton of sand, assuming an 8-hr work
day.
If alternate schemes were developed to insure year -around use of the
combustor, the annualization cost would drop to about $0.42/ton. This
would require of course the utilization of the combustor in other prac-
tical process operations, as discussed later, when not needed for its
primary purpose. Such an arrangement would be entirely feasible
considering the function of the combustor, but probably impractical if
the process in which it were alternatively used had to be- shut down be-
cause of its sudden withdrawal for oil-spill emergencies.
Another aspect of the cost analysis that is difficult to quantify involves
the loading and off-loading operations. Definite capital costs will be
associated with the constant-feed input conveyor and the output system
that will permit the cleaned sand to be disgorged at a reasonable dis-
tance away from the combustor. Design of such machinery was not
attempted on the present feasibility study, except as indicated in
footnote No. 15.
Operating costs will be involved in the loading of the feed conveyor
hopper and removal of clean sand deposited by the output conveyor.
No additional costs can be identified with this operation. At a typical
64
-------
beach restoration site, front-end loaders would normally be deployed
to remove contaminated sand piled up by graders. This material would
then be deposited in trucks for outhaul or, in the present case, at the
working site of the combustor. In the latter situation, the combustor
would be fed directly by the front-end loader, at no cost penalty, if the
load it had acquired had been resting on the berm for an hour or more.
It seems reasonable to expect that the renovation could be managed in
such a manner that this operation could be performed without incurring
additional costs.
Returning cleaned sand to the surf would obviously be done by the same
front-end loader(s) that furnish the input to the combustor. A problem
in this arrangement would be that the shovels of the vehicles would
probably be filthy and tend to recontaminate the renovated material.
This could be prevented by applying a suitable coating to the shovels.
Finally, the actual operating costs of the combustor must be considered.
These are shown in Table 10. There,basic operating costs are presented
under worst case (A and B) conditions, wherein moist, oil-free sand is
processed with and without recuperated heat (from a sand cooler). The
further cost figures, based on realistic sand compositions^", demon-
strate that heat recuperation is not an important economic factor. Thus,
the concept of cooling the output hot sand with sea water is probably
worthy of serious consideration.
The cost data are based on the conceptual design configuration already
described for a 10 ton/hr system. Although this system may actually
prove to have a capacity of 30 ton/hr, the costs presented for the full-
scale system (60 ton/hr) are based on the minimum expectancy. This
6:1 scale-up involved the assumption that the volumetric increase in the
equipment would be a square root function of the increase in throughput.
It can be seen that, using the full-scale system, with expectable input,
operation and maintenance costs would be 54^/ton to produce clean sand.
If the input contained the anticipated average of 16 wt-% petroleum, this
would represent a clean-up cost of about 1. 3^/gal of contaminant.
About 45% of this cost would be labor; one full-time operator at $7. 16/
hr plus a — generously estimated — one-third time maintenance mechanic
at $7. 36/hr.
Fuel costs are based on the use of diesel oil available at l6£/gal and a
power (motor/generator set) draw of 45 hp for the 10-tph unit and 195
hp for the 60-tph unit. In the normal operating situation (Case C -
self-sustained combustion), an auxiliary fuel allowance was included.
As previously estimated, no auxiliary process heat will be required
when the sand contains as little as 6 wt-% oil (dry basis); the amount
expected from a typical excavation operation is about 16 wt-%. The
difference in energy level is more than sufficient to dehydrate the
sand, assuming it contains 6 wt-% moisture.
65
-------
TABLE 10
Combustive Sand Cleaning Cost Estimates
10 ton/hr System 60 ton/hr System
- - -
$/ton Case: A1 B^ CT Case: A1 B C
Fuel 1.79 1.20 0.29 1.83 1.24 0.30
Labor 0.99 0.99 0.99 0.24 0.24 0.24
Total 2.78 2.19 1.28 2.07 1.48 0.54
System without sand cooler processing oil-free sand (6 wt-% moisture)
System with sand cooler processing oil-free sand (6 wt-% moisture)
3
Either system processing contaminated sand (6 wt-% moisture) con-
taining 6 or more percent of oily material
66
-------
This was provided to permit the incinerator to be brought up to com-
bustion temperature and be moved thrice a day without cooling. The
assumption was that the system would be inoperative (cool) except for
8 hr/day and that it would require 1. 75 hr of auxiliary fuel firing
(under Case A demand) to promote stable operating conditions and 45
minutes of transit time. It is likely that only 10 or 15 minutes will
actually be required for initial equilibration. In this case, the actual
fuel consumption would be only about 10^/ton in Case C operation,
which would bring the total (basic) operating cost down to 34^/ton.
Possible Secondary Uses of the Combustor: The present program has
been aimed at a very specific and acute damage-control problem. The
hardware described herein would hopefully have very limited oppor-
tunities of application. Like an ambulance, however, the annualization
of the capital investment for the owner would have to be passed off as
an unfortunate penalty to the sometime buyer of the equipment's ser-
vice. This would predictably occur between periods of considerable
inactivity. Secondary uses of the combustor thus would seem desirable
to level this burden, but for the same reason that ambulances are not
used to deliver groceries when idle, alternate utilization perhaps should
be carefully considered.
Another constraint is that a combustive beach renovator would neces-
sarily be a low capacity machine in order that it be deployable. It
would therefore be of little interest to large-scale operators dealing
with, say, ore-roasting, calcining, or process-slurry dehydration.
In fact, it would doubtless be uneconomical to use in its primary
function on beaches that had only experienced minor oil-spotting from
the bilges of passing ships or from small, chronic submarine seeps.
Considered as a mobile incinerator, however, a number of specialty
uses can be suggested that might be compatible, particularly on an
interuptable basis. These include the following operations:
• Destruction of classified material
• Incineration of hospital wastes
• Neutralization of low-energy, toxic chemical/biological
agents (military, agricultural, etc.)
• Disposal of contraband (narcotics and other smuggled
goods)
• Rare-ore treatment
• Chemical waste disposal at special sites
• PHS destruction of infected articles in plague areas
• On-site clean-up of littered areas following heavily
attended mass meetings of people.
67
-------
SECTION VII
ACKNOWLEDGEMENTS
The Aerojet-Hirt program staff wishes to express its appreciation for
the cooperation and guidance furnished by the FWQA Project Officer,
Mr. Gerald Burke. We also extend our gratitude to Mr. J. D. Sartor
of URS Research Co. who, on several occasions, furnished us with
vitally needed information. The Program Manager also particularly
wants to extend thanks for the help and suggestions offered by all of the
individuals contacted on the field survey (Table 2).
69
-------
SECTION VIII
REFERENCES
1. Oil Pollution, a Report to the President by the Secretaries of
Interior and Transportation, February 1968.
2. Swift, W. H. , Oil Spillage Study Literature Search and Critical
Evaluation for Selection of Promising Techniques to Control and
Prevent Damage, Battelle Memorial Institute Report on USCG
Contract TCG-15560-A, 20 November 1967.
3. Merz, R. C., Part 1, Determination of the Quantity of Oily Sub-
stances on Beaches and in Near shore Waters, State Water Pollu-
tion Control Board Publication No. 21, Sacramento, California
(1959).
4. Rosen, A. A., L. R. Musgrave and J. J. Lichtenberg, Part 2,
Characterization of Coastal Oil Pollution by Submarine Seeps,
ibid.
5. A Primer on Oil Spill Cleanup, Brochure of the American Petro-
leum Institute Task Force on Oil Spill Cleanup, Committee for
Air and Water Conservation, New York, N. Y. , 1968.
6. El Incidente del OCEAN EAGLE, Office of Petroleum Emergen-
cies, Department of Public Works Report, San Juan, Puerto Rico,
July 1968.
7. The WITWATER Tanker Casualty, Smithsonian Tropical Research
Institute Report, Balboa, C. Z. (undated).
8. Oil Slick Pollution, Minutes of Conference of Oil Slick Pollution
of Harbours and Associated Waters, U.S. Coast Guard Head-
quarters, Washington, D. C. , 18 March 1964.
9. Recommended Methods for Dealing with Oil Pollution, Warren
Spring Laboratory Report No. LR 79 (EIS), June 1968.
10. Dennis, J. V. , Oil Pollution Survey of the U.S. Atlantic Coast,
Division of Transportation, American Petroleum Institute Re-
port, Washington, D. C., 15 May 1959.
11. McGill, J. T. , "Map of Coastal Land Forms of the World,"
Geographical Review, 48, 402-5 (1958).
12. Dubach, H. W. , and M. A. Slessers, "An Evaluation of the
Morskoi Atlas, " Navy Hydrographic Office, March 1958.
71
-------
13. King, C. A. M. , "Beaches and Coasts," Edward Arnold, Ltd.,
London, 1959.
14. Zenkovich, V. P., "Processes of Coastal Development, " Trans-
lation Ed. J. A. Steers, Interscience Publishers, 1967.
15. Putnam, W. C., et al, Natural Coastal Environments of the
World, University of California Los Angeles Report on ONR
Contract Nonr-233 (06), (I960).
16. Sunken Tanker Project Report, U. S. Coast Guard Report to the
Secretary of Transportation (undated).
17. O'Sullivan, A. J. , and A. J. Richardson, "The TORREY CAN-
YON Disaster and Intertidal Marine Life, " Nature, 214, 448-542
(1967).
18. Royce, W. F. , "The Oil Danger is in the Clean Up," National
Fisherman, April 1969.
19. "Slick Tricks," Ind. Bull. (A. D. Little, Inc.), 476, 1J1969).
20. Analytical Characteristics of Oily Substance Found on Southern
California Beaches, Engineering-Science, Inc. , Report to the
Western Oil and Gas Association, July I960.
21. Sampling of Oily Substances on Southern California Beaches,
ibid, November 1969.
22. Combating Pollution Created by Oil Spills, A. D. Little, Inc. ,
Report No. 7138b (R) to the U.S. Coast Guard, 30 June 1969.
23. Fullerton, E. C. , "Marine Enforcement -Oil Spills," paper
presented at Western Assoc. Game & Fish Commissioners,
Honolulu, Hawaii, July 19&7.
24. T. H. Gaines, Notes on Pollution Control Santa Barbara, Union
Oil Co. Internal Report, June 1969.
25. Swift, W. M. , et al, Review of the Santa Barbara Channel Oil
Pollution Incident, FWQV Report No. DAST-ZO, 18 July 1969.
26. Carpenter, C. E. , L. F. Butcher and A. S. Huxley, Laboratory
Examination of Materials Submitted for Treating the TORREY
CANYON Oil Spill, Admiralty Oil Laboratory Report No. 51,
Cobham, England, January 1969.
27. Report of the Ad Hoc Committee, State Regulations and Prac-
tices, Oil and Gas Operations and Oil Pollution, Report of the
Resources Agency of California, June 1969.
72
-------
28. Joint Report of the Resources Agency and State Lands Commis-
sion Requested in House Resolution 77 o± the 1959 Regular Session
California Legislature, (Report source defined in title), May 1969.
29. Montgomery, S. , "Oil Pollution R&D on the Rise, " Undersea
Technology, 1£ (10), 30 (1969).
30. Biglane, K. E., International Aspects of Oil Pollution Control,
FWPCA Trip Report, November 1968.
31. Linsell, R. F. , Disposal of Waste Oils and Solid Hydrocarbons
Contaminated with Sand, Esso Petroleum Co. Internal Technical
Memorandum, Z9 March 1967.
32. Swift, W. H. , et al, "Oil Spillage Prevention, Control, and Res-
toration - State of the Art and Research Needs, " J. Water Poll.
Control Fed. , 4l_, 392-412(1969).
33. Swift, W. H. , C. J. Touhill and P. L. Peterson, "Oil Spillage
Control, " Chem. Eng. Prog. , 65_, 265-273 (1969).
34. Sartor, J. D. (URS Research Co.), Private Communication,
15 December 1969.
35. Perry, J. H. , "Chemical Engineers Handbook, " Fourth Edition,
McGraw-Hill Book Co. , Inc., 1963.
36. Treybal, R. E. , "Mass-Transfer Operations, " McGraw-Hill
Book Co., Inc., 1955.
37. Anon. , "Oil Spills: An Environmental Threat"; Environ. Sci.
& Tech., 4, 97-99 (1970).
73
-------
BIBLIOGRAPHIC:
Znvirogenics Co., a Division of Aerojet-General
Corporation, A Feasibility Analysis of Incinerator Sys-
tems for Restoration of Oil Contaminated Beaches, Final
Report FWQA Contract No. 14-12-595, November 1970.
ABSTRACT
The feasibility of employing a combustion effect
for restoring oil contaminated beaches was investigated.
Beach access problems and the handling characteristics
of shore materials limited the potential application to
recreational (sand) sites. Thermodynamic arguments
required that a system design be adopted in which the
contaminated sand would undergo combustive processing
in a confined airangement. The design selected, from
those analyzed, proved to be a three-effect combustor
based on the rotary kiln principle. Provided that the
sand to be cleaned is carefully enough collected to fur-
nish a reasonable (>6%) oil content and is moved away
from the surf and drained to an acceptable moisture
level (< 6%), processing costs would be highly attractive.
In comparison with uncontaminated sand, the cleaned
product exhibits only a slightly greyish hue.
KEY WORDS:
Beaches
Oil Wastes
Cleaning
Shore Protection
Shores
Coasts
Incineration
Oily Water
BIBLIOGRAPHIC:
Envirogenics Co., a Division of Aerojet-General
Corporation, A Feasibility Analysis of Incinerator Sys-
tems for Restoration of Oil Contaminated Beaches, Final
Report FWQA Contract No. 14-12-595, November 1970.
ABSTRACT
The feasibility of employing a combustion effect
for restoring oil contaminated beaches was investigated.
Beach access problems and the handling characteristics
of shore materials limited the potential application to
recreational (sand) sites. Thermodynamic arguments
required that a system design be adopted in which the
contaminated sand would undergo combustive processing
in a confined arrangement. The design selected, from
those analyzed, proved to be a three-effect combustor
based on the rotary kiln principle. Provided that the
sand to be cleaned is carefully enough collected to fur-
nish a reasonable (>6%) oil content and is moved away
from the surf and drained to an acceptable moisture
level (< 6%), processing costs would be highly attractive.
In comparison with uncontaminated sand, the cleaned
product exhibits only a slightly greyish hue.
KEY WORDS:
Beaches
Oil Wastes
Cleaning
Shore Protection
Shores
Coasts
Incineration
Oily Water
BIBLIOGRAPHIC:
Envirogenics Co., a Division of Aerojet-General
Corporation, A Feasibility Analysis of Incinerator Sys-
tems for Restoration of Oil Contaminated Beaches, Final
Report FWQA Contract No. 14-12-595, November 1970.
ABSTRACT
The feasibility of employing a combustion effect
for restoring oil contaminated beaches was investigated.
Beach access problems and the handling characteristics
of shore materials limited the potential application to
recreational (sand) sites. Thermodynamic arguments
required that a system design be adopted in which the
contaminated sand would undergo combustive processing
in a confined arrangement. The design selected, from
those analyzed, proved to be a three-effect combustor
•based on the rotary kiln principle. Provided that the
sand to be cleaned is carefully enough collected to fur-
nish a reasonable ( >k%) oil content and is moved away
from the surf and drained to an acceptable moisture
level (<6%), processing costs would be highly attractive.
In comparison with uncontaminated sand, the cleaned
product exhibits only a slightly greyish hue.
KEY WORDS:
Beaches
Oil Wastes
Cleaning
Shore Protection
Shores
Coasts
Incineration
Oily Water
-------
1
5
Access/on Number
r* Subject Field & Group
(j)5E
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Envirogenics Co. , Division of Aerojet-General Corporation, El Monte, California
Title
A Feasibility Analysis of Incinerator Systems for Restoration of Oil Contaminated
Beaches,
I Q Authors)
Roberts,
Hoyt, T.
R. M. , and
s.
16
21
Project Designation
Note
22
Citation
23
Descriptors (Starred First)
*Beaches, *Oil Wastes, *Cleaning, Shore Protection, Shores, Coasts,
Incineration, Oily Water
25
Identifiers (Starred First)
*Oil Spills, *Beach Decontamination, Combustion Processes
27
Abstract
The feasibility of employing a combustion effect for restoring oil contaminated
beaches was investigated. Beach access problems and the handling characteristics
of shore materials limited the potential application to recreational (sand) sites.
Thermodynamic arguments required that a system design be adopted in which the
contaminated sand would undergo combustive processing in a confined arrangement.
The design selected, from those analyzed, proved to be a three-effect combustor
based on the rotary kiln principle. Provided that the sand to be cleaned is carefully
enough collected to furnish a reasonable ( > 6%) oil content and is moved away from
the surf and drained to an acceptable moisture level (< 6%) processing costs would
be highly attractive. In comparison with uncontaminated sand, the cleaned product
exhibits only a slightly greyish hue.
This report was submitted in fulfillment of Contract No. 14-12-595 between the
Federal Water Quality Administration and the Envirogenics Co. , Division of
Aerojet-General Corporation.
Abstractor Author
Institution
Envirogenics Company
WR:102 (REV. JUUY 1969)
WRSIC
SEND TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
* GPO: 1969-359-339
------- |