WATER POLLUTION CONTROL RESEARCH SERIES
lf>OH()I)JOr> 7O
OIL TAGGING SYSTEM STUDY
.KlMKNi oh IMF. INTERIOR • FEDERAL WATER POLLUTION CONTROL ADMINISTRATION
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe the results
and progress in the control and abatement of pollution in pur
Nation's waters. They provide a central source of information
on the research, development, and demonstration activities in the
Federal Water Pollution Control Administration, in the U.S. Depart-
ment of the Interior, through inhouse research and grants and
contracts with Federal, State, and local agencies, research insti-
tutions, and industrial organizations.
A triplicate abstract card sheet is included in the report to facili-
tate 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, Depart-
ment of the Interior, Federal Water Pollution Control Admin istra-
tration, Room 1108, Washington, D.C. 20242.
-------
OIL TAGGING SYSTEM STUDY
by
Research Division, Environmental
and Applied Sciences Center
MELPAR
An American-Standard Company
7700 Arlington Boulevard
Falls Church, Virginia 22046
for the
Federal Water Pollution Control Administration
Department of the Interior
Contract No. 14-12-500
Melpar Ref. 9059
May 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 20402 - Price $1.60
-------
This report has been reviewed by the Federal
Water Pollution Control Administration and
approved for publication. Approval does not
signify that the contents necessarily reflect the
views and policies of the Federal Water
Pollution Control Administration, nor does
mention of trade names or commercial products
constitute endorsement or recommendation for
use.
ii
-------
ABSTRACT
Several methods of identifying the source of oil pollution are critically
examined. These methods are grouped into two categories: passive tagging and
active tagging. Passive tagging assumes that oils are so chemically diverse that
their contents constitute a stable chemical fingerprint that can be unequivocally
disclosed in the laboratory. Active tagging requires that an inexpensive, coded
material be added to oil; this material must be chemically and physically stable
in both oil and oil slicks; it must also be readily identifiable by available analytical
techniques; and it must have no adverse effect on the oil's subsequent use. Three
methods of passive tagging (trace metals, sulfur-isotope ratios, and paper
chromatography) and three methods of active tagging (halogenated polycyclic
aroma tics, organometallics, and coded microspheroids) have been examined.
Passive tags cannot be recommended because the passive tags are quite likely to
mingle, to evaporate, to be dissolved, or to be oxidized; even if these processes
do not occur, they can create formidable forensic problems for the prosecution
and telling counter-arguments for the defense. Since active tags are designed to
be stable and identifiable, they are satisfactory for the job; and the three types
of active tags reviewed show promise and merit.
This report was submitted in fulfillment of Contract Number 14-12-500
under the sponsorship of the Federal Water Pollution Control Administration.
Key Words: Pollutant identification, tagging, oil wastes, tracers, analytical
techniques, chemical analysis, chromatography, waste identification,
indicators, water pollution control, oil industry.
iii
-------
ACKNOWLEDGEMENTS
The authors are indebted to many people and institutions for helping us in our
research. We are grateful to the Library of Congress for study facilities and the
assistance of many librarians. The American Petroleum Institute allowed us to use
their libraries (in both Washington, D. C., and New York City); thanks are due to
Miss Virginia M. Smyth of API's New York library for reading our bibliography and
helping us correct it. G. J. Schrayer and W. E. Hanson of the Exploration Division
of Gulf Research, Pittsburgh, Pa., called our attention to several significant papers
and sent us copies of them.
Finally, we express our thanks to several publishers for having authorized
us to quote extracts from books that we have found indispensable:
The Elsevier Publishing Company, Amsterdam, The Netherlands, for per-
mission to quote from SCOTT G. (1965), Atmospheric Oxidation and Antioxidants;
and from NAGY B. & COLOMBO U. (editors) (1967), Fundamental Aspects of
Petroleum Geochemistry.
The Pergamon Press, Inc. , Elmford, N. Y., for permission to quote from
MANSKAYA S.M. & DROZDOVA T. V. (1968), Geochemistry of Organic Substances
(SHAPIRO L. & BREGER LA. , translators and editors) Volume 28 of the International
Series of Monographs in Earth Sciences; and BREGER I. A (editor) (1963), Organic
Geochemistry, Monograph No. 16 of the Earth Sciences Series.
The Bureau of National Affairs, Washington, D.C., for permission to quote
from DIEGLER S.E. (editor) (1969), Oil Pollution: Problems and Policies.
The Chemical Publishing Company, Inc., New York City, N.Y., for per-
mission to quote from McCOY J. W. (1962), the Inorganic Analysis of Petroleum.
The St. Martin's Press, Inc., New York City, N.Y., and Macmillan &
Company, Ltd., London, England, for permission to quote from KRISS A. E. et al.
(1967), Microbiological Population of Oceans and Seas (SYERS K., translator,
FOGG G.E., translation editor).
J. H.
G. D. G.
R. G. N.
T. P. M.
IV
-------
CONTENTS
Section Page
I OIL POLLUTION-THE SIZE AND SOURCE OF THE PROBLEM 1
II CHANGES IN OIL SLICKS 7
2. 1 General 7
2.2 The Chemistry of Weathering 8
2.2.1 Introduction 8
2.2.2 Catalytic Oxidation of Petroleum 8
2.2.3 Photochemical Oxidation 11
2.2.4 Metalloporphyrins 16
2.3 Microbial Oxidation 18
IE PASSIVE TAGGING 23
3.1 General 23
3.2 Metals 25
3.3 Sulfur-Isotope Ratios 32
3.4 Special Cases 35
IV ACTIVE TAGGING 41
4.1 General 41
4.2 Soluble Tags 42
4.2.1 Halogenated Aromatics 42
4.2.2 Organometallics 53
4. 2. 3 Summary and Conclusion for Soluble Tagging 55
4.3 Particulate Tags 58
4.3.1 Introduction 58
4.3.2 Palynological Studies 58
4.3.3 Production of Microparticles 64
4.3.4 Retrieval from Spills: Preconcentration 68
4.3.5 Particle Characterization 68
4.3.6 Costs 70
4.3.7 Summary and Conclusions 71
4.4 Radiochemical Tagging 74
-------
CONTENTS (Continued)
Section
V IMPLEMENTATION OF THE TAGGING SYSTEM 77
5.1 Marking and Identifying Oil Shipments 77
5.1. 1 Oil Tagging at Transfer Points 77
5.1.2 Determining Jurisdiction 78
5.1.3 Mechanism of Adding Tags 78
5.1.4 Tag Control 78
5.2 Codes 79
5.2.1 License Plate System 79
5.2.2 Profile System 79
5.2.3 Coding Examples 80
5.2.4 Summary and Conclusions 88
VI CONCLUSIONS AND RECOMMENDATIONS 95
VH REFERENCES 97
vi
-------
FIGURES
No. Page
1 Bond-Dissociation Energies and Double-Bond Activation Energies
for Chemical Groupings Present in Some Autoxidizing Media 12
2 License Plate Coding Showing Oil from Wells W, X, Y Shipped to
Refineries A, B, C and Reshipped to Users I, J, K by Tankers 1
through 10. 84
3 Profile Code with Same Wells, Refineries, Users, and Tankers,
Using a Binary Code 86
4 Profile Codes in Oil from Jurisdiction X to User J 93
5 Profile Codes in Oil from Jurisdiction W to User J 93
6 Profile Codes in Oil Showing a Ten-to-One Dilution at
Each Jurisdictional Change 93
7 Profile of the Added Code 93
8 Code Dilution When Entering Tank Farm 93
vii
-------
TABLES
No.
1 Primary Photochemical Processes 13
2 Resonance Energies of Selected Aromatic Compounds 42
3 Chlorine Isomers of Naphthalene 45
4 Vapor Pressures of Some Selected Aromatic Polynuclear
Hydrocarbons for Use as Tags (25 to 50° C) 46
5 Unit Price of Several Selected Halogenated Aromatic Hydrocarbons 51
6 License Plate Tags 82
7 Numerical Values Corresponding to Figures 4, 5, 6, and 8 90
viii
-------
SECTION I
OIL POLLUTIQN-THE SIZE AND SOURCE OF THE PROBLEM
Before analyzing several suggested methods for identifying spilled oil, a few
words about the size and source of the oil-pollution problem are in order. How much
petroleum (crude and refined) has been spilled on U.S. waters in recent years? How
many spills were there? What sources were responsible? What is the potential
pollution?
Authoritative (but incomplete) answers to these questions have been published
in official Government documents: Congressional Hearings, reports from the Bureau
of Mines, and special reports to the President and to Congress prepared by the Federal
agencies that administer water-pollution laws. There are also private sources of
interest. Some of the highlights are summarized in the following discussion.
The United States carries on an enormous foreign trade in petroleum and
petroleum products. The U.S. Bureau of Mines has estimated that in 1967 this trade
came to more than 1 billion barrels (42 gallons per barrel), 90 percent of which was
imports. Almost all of this trade is conducted by ship. The Bureau also reports that,
of the 3. 6 billion barrels of oil refined in the U.S. in 1967, about 800 million barrels
were sent to the refinery over water. The volume of petroleum transported over
water to U.S. coastal and inland ports each year is astronomic: It has been estimated2
that the tonnage of petroleum moving by water is substantially greater than the com-
bined tonnage of all other commodities. The potential for oil spills is obvious.
Although oil spills are illegal all over the world (by international convention,3
usually supplemented by national statutes), laws are often observed only in the breach.
The U.S. is no exception, even though there are at least six statutory laws that apply:
the Oil Pollution Act of 1924 (33 USC 431 et seq.), the Oil Pollution Act of 1961
(33 USC 1001 et seq.), the Refuse Act of 1899 (33 USC 407), the Dangerous Cargo Act
(46 USC 170), the Tank Vessel Act (46 USC 391a), and the so-called Magnuson Act
(50 USC 191).4
In one winter, November 1, 1968, to February 22, 1969, there were over a
million gallons of crude and refined petroleum spilled on U.S. waters from pipelines,
offshore installations, tankers, and barges. During the same period, on January 28,
1969, Union Oil Company's well 12-A in the Santa Barbara Channel, OCSP-0241, blew
out while a drill pipe was being removed. Gas and oil (300, 000 bbl per day) flowed
from the ocean floor near the well. It was ten days later before the well was killed.
There has been no thorough nationwide survey of the oil spilled into U.S. inland
and offshore waters by vessels and barges, particularly as distinct from spills out of
shore facilities. However, scattered data suggests that spill incidents are frequent,
-------
despite the fact that oil discharges into American waters are prohibited. The U.S.
Army Corps of Engineers estimates that there were over 2000 oil spills within
U.S. waters in 1966, of which 40 percent came from land-based facilities. Recent
Federal Water Pollution Control Administration estimates indicate that illegal dis-
charges range between 7,000 and 10,000 annually. Many of the spills, however, are
unwitnessed and go unreported; then the enforcement of the oil pollution laws becomes
quite a problem.
A particular problem in connection with the Oil Pollution Act of 1924, as
amended, is the extreme difficulty of apprehending and prosecuting offenders at sea.
The Coast Guard, for example, listed 458 violations of the oil pollution statutes
detected in 1967 (of which 63 involved tankers, 56 dry cargo ships, and 56 Navy ships
that reported themselves). Privately, however, Coast Guard officials suggest that the
number of actual violations may have been closer to 5, 000 than 500. The problem is
that an oil spill, such as the one that hit Cape Cod beaches in 1967, often isn't detected
until the offending ship is long out of the area. Unless a witness is on the scene when
a violation occurs or a particular slick can be chemically traced to a particular ship
known to have passed through a polluted area, spills go completely undetected until
the oil washes up on a beach and someone complains. 7
Maritime collisions are known to contribute to oil pollution. The Secretaries'
Report to the President summarizes the most important facts:
The maritime casualty rate, including barge casualties on the
inland waters, fluctuates at a level sufficiently high to justify
serious concern. The following figures summarize the recent
combined casualty record of U.S. -registered vessels world-
wide and foreign vessels in U.S. waters:
FY 1966 FY 1967
Number of casualties, all types 2,408 2,353
Vessels over 1,000 tons 1,310 1,347
Tank ships and tank barges 470 499
Locations:
U.S. waters 1,685 1,569
Elsewhere 723 784
Types of casualties:
Collisions 922 1,090
Explosions 175 168
Groundings with damages 302 282
Founderings, capsizings,
and floodings 315 230
-------
These types of casualties can and in some cases do
cause polluting spills.... H
To put these figures into proper perspective, it must be remembered that in
1965 the world's merchant fleet contained
... over 18,000 vessels of 1,000 gross tons or heavier; almost
3,500 of these vessels were tankers. Of the total, the U.S.
flag was represented by about 2,500 vessels, of which approxi-
mately 400 were tankers. About 1,000 of the American vessels
were regularly engaged in foreign commerce, while the
remainder served in the U.S. coastwise and other domestic
trades. The foregoing fleet was supplemented within the
United States by approximately 36, 000 smaller vessels.
... These included several thousand tank ships and tank barges
on the American inland waterways.
.. .To illustrate the extent to which U.S. waters are used by
vessels in the foreign trade, the following table sets forth the
approximate numbers of visits to U.S. ports in FY 1966 by
vessels over 1,000 deadweight tons:
Average Average
Oil Cargo Bunker Oil
Capacity Capacity
Ship Type
Foreign—Dry Cargo
Foreign—Tanker
U. S. —Dry Cargo
U.S.—Tanker
Number
(tons)
34,000
6,100
7,000
3,000
*****
25,000
*****
14,000
(tons)
1,400
2,500
3,400
1,300
In summary, there were over 50,000 visits to U.S. ports in FY
1966 by medium-size and large ocean-going vessels with a
cumulative capacity of almost 300 million tons of potentially
polluting materials.12
Oil pollution from offshore wells is not a much larger problem. According to
the testimony of Dr. William T. Pecora, Director of the U.S. Geological Survey,
before the Senate Committee on Public Works8
There have been a total of 14,000 wells drilled off shore, both
State and Federal (i.e., on State and Federal offshore lands).
The Federal proportion is 10,243 wells, and the State portion
is 3,847 wells...14
-------
Land-based facilities such as storage tanks and pipelines are the third major
source of oil pollution and are estimated to account for 40% of the oil-pollution
incidents in this country. The Secretaries' Report states that:
There is no readily available compilation of the number,
size, and character of the facilities for moving and
storing these materials [scil., oils and liquid chemicals].. .
Extensive coastal and riverside terminal facilities, now
numbering about 6, 000, have been developed for the
transfer of commodities between water and land [but]...
products arriving at or departing from these water-side
facilities may be handled several times. Both crude and
refined products may travel via pipeline, railroad or tank
truck as well as by marine vessel, producing almost
endless possibilities for the discharge of oil.... This
country is laced by an estimated 200, 000 miles of pipe-
lines operating at pressures up to 1, 000 pounds per square
inch. These lines carried more than one billion tons of
oil and other hazardous substances in 1965. Many sections
of this network are laid in and across navigable waterways
and reservoir systems.... Thus, our pipeline transport
system involves the risk of oil pollution in our watercourses,
port areas, and critical drinking water supply areas.
Pipelines also break under water. Capt. William A. Jenkins, Deputy Chief,
Office of Operations, U.S. Coast Guard, has testified before Congress on the fre-
quency of pipeline breaks in the Gulf of Mexico. The Coast Guard had received
reports from pipeline breaks in the lower Mississippi Delta area and Louisiana
Coast area around the Morgan City area.
Some of these breaks have occurred due to marine casualties or other
unknown sources. They have been of varying sizes, some quite sizable, and have
gone into the swampy bayou areas. Capt. Jenkins supplied the Committee with a
list of six recent incidents of pipeline breakage.
CONCLUSION
The potential for oil pollution exists anywhere oil is handled or used. Though
catastrophic oil discharges usually receive immediate Government and public atten-
tion, the smaller spills and discharges, particularly those that take place during oil
transfer operations, often go undetected until the slick washes up on shore and
someone complains. Because many oil discharges go unnoticed and unreported at the
time they happen, a means to identify the oil discharges from these oil transfer
operations would offer more control. The offender could be identified and official
action taken to ensure that: (1) the cause(s) of the illegal discharge was corrected and
(2) proper cleanup operation was undertaken by the offender to remove the pollutants.
-------
Tagging systems could be of particular value in monitoring oil spills that can
take place when oil is transferred from ship to barge, barge to storage facility, or
storage facility to pipeline operations and indeed where any similar oil transfer oper-
ation is undertaken.
But several criteria must be met if an oil tagging system is to be useful. The
system must be acceptable to producers, processors, shippers, consumers, enforce-
ment agencies, and the courts. Further, if tagging materials are added to the oil they
must be compatible with the oil and not affect its subsequent use. The tagging system
must uniquely and legally identify those responsible for the oil spill and its consequence.
Regulations will have to be developed to govern the handling of oil during shipment and
transfer and the discharge and disposal of oil wastes near port and on the open seas.
Although waterborne sources are estimated to account for more than half the pollution,
land-based facilities account for 40%. Thus, any movement of oil — loading,
unloading, storage, and transfer — constitutes a threat. Accidents, poor maintenance,
carelessness, ignorance and willful neglect, and lack of concern for the environment
contribute to the problem. Though the largest potential sources of catastrophic oil
pollution are few in number and usually quickly identified, the thousands of unidentified
spills that appear "mysteriously" on the water or on the beach must be prevented. The
problem becomes that of identifying these less publicized sources of oil pollution
cheaply and conclusively.
5/6
-------
SECTION II
CHANGES IN OIL SLICKS
2.1 General
When oil is spilled on water, three major variables affect the fate of the slick:
(1) the composition of the oil; (2) the nature of the receiving body of water (whether
fresh, salt, or brackish), its temperature, motion, microbial population, and chemi-
cal contents; and (3) such diverse environmental factors as insolation, latitude, wind,
atmospheric contents, and the like.
Oil, of course, is no simple compound or mixture of compounds: Crude oils
show enormous variation in chemical content, both quantitatively and qualitatively.
The literature on this point is immense, but there is special merit in the reviews of
Bestougeff ' (hydrocarbons on crude oils) and of Costantinides & Arich^ (non-
hydrocarbons). Naturally, variations in chemical makeup are accompanied by varia-
tions in physical properties of the molecular mixture: specific gravity, water solu-
bility, volatility, viscosity, melting point, etc. Perhaps the largest event in the
history of an oil slick is the evaporation of the light fractions. While the rate of
evaporation will be influenced by many variables (wind velocity and turbulence,
temperature of air and water, rate of the slick's spread, wave action, etc.), after
several days at sea the slick will consist almost entirely of compounds whose boiling
point is above about 700°F, ' viz, compounds containing at least 20 carbon atoms,
and probably of complex chemical structure. (For example, while crude oils from
Montgomery County, Texas, contain only a small residue above 600°F, oils from
OO -I Q
Louisiana's Plaquemines Parish are frequently almost half residue. Blokker
has analyzed the spreading and evaporation of several kinds of oil slicks. He reports
that spilled oil rapidly spreads to a film 2 cm thick; although many variables affect
its subsequent thinning, its final thickness will be 10-100 microns on pure water,
but approximately 1 mm on heavily contaminated water.
While evaporation undoubtedly accounts for most of the volumetric loss in the
slick, Dean has suggested that solution in sea water may also play a significant role.
For while the lightest hydrocarbons have finite solubilities in water, some heavier
petroleum constituents, while undoubtedly less soluble, may still go into solution, for
we are considering huge volumes of sea water. Dean notes that if a solubility on the
order of one part per million is assumed, the whole cargo of the Torrey Canyon could
have been dissolved in a patch of sea 20 miles square and 500 feet deep. In this con-
nection, the presence of surfactants in natural crudes (e.g., such surfactants as
vanadyl porphyrins) assumes a new significance. Several compounds of these metals
in petroleum are known to be soluble in water.
As the light fractions and some of the metallic compounds disappear from the
slick, other processes act upon the residue. While many of the common constituents
(such as paraffins) are rather inert, some readily participate in chemical reactions
-------
and others become reactive in the presence of: strong sunlight (especially the ultra-
violet spectrum); reaction intermediates in the slick itself and in the surrounding sea;
atmospheric oxygen and ozone; and miscellaneous sources of chemical and physical
energy. The slick may also be attacked by many microbial species that are common
in both fresh and salt water. These several processes are discussed at length in the
following sections of this report (2. 2-3. 3), Many of these processes are affected by
ambient temperatures, and all of them are undoubtedly affected by the slick's disper-
sion; the degradation of such stable emulsions as "chocolate mousse" is surely impeded
by the small surface area offered to the open sea and sky.
2.2 The Chemistry of Weathering
2.2.1 Introduction
To control violations of water-pollution regulations a system for identifying
violators must be devised and instituted. One approach is to identify the violator
through analysis of the components (intrinsic or added) of the oil slick itself. Analyti-
cal technology and "fingerprinting" (i. e., analytical determination of characteristic
chemical and physical properties of the oil) can be applied to petroleum analyses with
a reasonable degree of sensitivity and precision. However, interpretation of the results
obtained through analytical technology is difficult when the technology is applied to a
continually changing system. These difficulties apply principally to assigning, on the
basis of slick composition, unequivocal legal responsibility for an oil spill: One has
to consider that the composition of the slick may not be identical to that of the unspilled
oil. The changes in chemical composition the oil undergoes is a function of many fac-
tors, the absolute number of which depend on the site of the spill (for example, fresh
or sea water), environmental conditions (continuous exposure to sunlight, microbial
content of the waters), and the time interval between the occurrence of the oil spill and
the analysis of the slick.
At present only a general knowledge exists regarding the changes that crude oils
and other petroleum products undergo when exposed to various natural forces (i. e. ,
weathering). We all acknowledge the fact that the oil changes its character because of
oxidation, solar radiation, and microbial action, but information about the specific
mechanisms and reaction products that result is sorely lacking. This information is
absolutely essential so that a rational and cogent interpretation of the analytical results
can be used to assign liability to the violators of pollution regulations.
2.2.2 Catalytic Oxidation of Petroleum
Oxidation is one of the primary processes in the weathering of petroleum.
Oxidation by air at ordinary temperatures (10 to 40° C) is commonly referred to as
autoxidation; hydrocarbons seem to be oxidized in a chain reaction that involves hydro-
peroxide radicals. The fundamental theories of Bakh2 state that the primary oxida-
tion products are peroxides. Emanuel'2^ and his coworkers25 have shown that the
-------
low-temperature, liquid-phase oxidation of hydrocarbons proceeds through several
peroxides and hydroperoxides . The following mechanisms have been postulated:
(1) RH — »-R* + H
(2) R" + 02— ^RO*
(3) RO" + RH — -ROOH + R* etc.
^
(4) RO* + RO* - «*ROOR + O0
£j Li £t
(5) ROOH - -RO'+'OH
(6) RO* + RH — ^ROH + R*
(7) *OH + RH - «-H O + R*
64
(8) ROOH - ^RH - "-2ROH
(9) ROOH— -R COR + HO
1 U Lt
(10) ROH— *-R COR0
J. ^
(11) R COR — »• acids
J. ^
Reaction (1) starts the oxidation chain and feeds the cycle of reactions (2) and (3). The
chain usually ends with reaction (4), in which RO?> radicals combine. The decompo-
sition of hydroperoxide ROOH to give RO* and *OH radicals (reaction (5)) occurs rela-
tively infrequently since rupture of the O-O peroxide bond requires 30 to 40 Kcal/mole.
(It will be shown later that solar radiation can meet such energy requirements. ) RO*
and *OH yield R*, the basic radical for the oxidative chain, as a result of reactions (6)
and (7). The breakdown of hydroperoxide according to reactions (8) and (9) leads to the
production of alcohols and ketones, which are also obtained by the oxidation of alcohols
according to reaction (10). Finally, acids are produced from ketones according to
reaction (11). Integrating all of these steps, the following scheme emerges.
alcohols
hydroperoxides
^ketones — *-acids
Peroxides and hydroperoxides appear to be the primary products of the liquid-phase
oxidation of unsaturated, alkyl-aromatic, and alicyclic hydrocarbons.
The kinetics of the oxidation depend on the oxygen partial pressure, and tem-
perature, and the hydrocarbon; the total quantity of products formed is a complex
-------
Oe
function of the rate constant for the decomposition of peroxides. The kinetic curve
is S-shaped: After an initial induction period, there is a period of increasing rate
followed by an asymptotic leveling off of the rate of oxidation and finally the reaction
ceases. Tinyakova et al. (1965)2? found that free radical production can occur at low
temperatures (20-50° C) in both aqueous and hydrocarbon media in the presence of an
oxidation-reduction system (reversible system) such as described by the general
equation below:
ROOM + Mn+-^RO% M(n+1)+ + HO"
The oxidation-reduction system can be provided by low concentrations (^0. 002M)
of metal salts. For example, I^C^ decomposes in the presence of ferric iron at
20° C; hydroperoxide can decompose at 50° C under the same conditions. We infer there-
fore that ferric iron in sea water (0. 01%)"" may increase the oxidation of petroleum
hydrocarbons through free-radical mechanisms. (Of course, the situation is not
simple because there are often oxidation- reduction inhibitors in the oil itself; sulfur
compounds in petroleum are sometimes potent inhibitors, though these compounds,
too, may be chemically altered in the slick. ) In this connection, Gureyev and Soblina
(1965)25 found that different hydrocarbon fractions had different sensitivities to oxida-
tion by free radicals. The kerosene fraction was more sensitive to the oxidative
action of metals (Cu, Fe) than the gasoline (benzene) fraction.
The amount of atmospheric oxygen which diffuses from the surface into the
bulk of the liquid affects the reaction. Since the concentration of the oxygen penetrating
into the oil is generally low and the temperature is relatively low, the oxidation is
slow. However, some oxidation occurs rather rapidly. The experience of Arozena^°
is instructive.
A crude petroleum of Syrian origin and a commercial hydrocarbon, gas-
oil, were used. The petroleum was not used directly because, when it is
put in contact with water (whether sweet or salt), its tarry [asphaltenic]
components agglutinate and form small spherical masses; meanwhile, the
oily matter spreads out on the surface of the water, forming a uniform
film without [any phase of] continuous solution .... For this reason we
thought it important to eliminate the semisolid phase of tars and asphal-
tenes .... [The crude oil was distilled at 300° C, the middle distillate
reserved, and the residue rejected. ]
We must report that these fractions, [even] after distillation, were not
refined; they therefore contain sulfur and substances, that can unite with
oxygen to form resins. We analytically verified the presence of sulfur
[in distilled fractions] and observed in the distillate, above all when it
was hot, the formation of new tarry materials upon passage of air through
these fractions. In this last-mentioned case, however, [the tars] do not
agglutinate as in the crude and, therefore, do not create problems of dis-
continuity in the [oil] films; furthermore, the quantity [of tars] is so
10
-------
small that they remain dissolved in the light and middle fractions of
the petroleum when these fractions are put in contact with water.
The composition of the petroleum must be kept in mind when studying
the passage of oxygen through oil slicks of distinct thickness. Accor-
ding to the petroleum's chemical composition, the thickness of the oil
film being held constant, a petroleum will fix more or less oxygen.
In our case, we used only one type of petroleum, whose composition
remained constant. . . .
—JH, trans.
In addition to peroxides, free-radical production can also be initiated by the
ce of acidic gums (non-distillal
of secondary processes (reaction 11).
2.2.3 Photochemical Oxidation
111 auuii/iuu \,\j jjcj.UAIUCO, n cc—i.aui.i/ai jjj.uuuv;i>i.uu v_,au aiou ut iiiinatcia uy LUG
presence of acidic gums (non-distillable resins) and "oxyacids" which are the result
In spite of the absorption of the shorter wavelengths of the sun's rays by the
atmosphere—which reduces the ultraviolet (UV) light of less than 4000 A to no more
than 5% of the total^"—there remains enough energy in the residual radiation to break
a variety of chemical bonds.
Table 1 illustrates several of the many possible primary photochemical pro-
cesses which occur on the excitation of polyatomic molecules. The polyatomic
molecule ABC, in the singlet (S^ or triplet (Tj) states, can undergo modification by
a variety of photochemical processes. Obviously, all these processes are not rele-
vant to the breakdown of water-borne petroleum. Reactions (1), (2), and (6) may be
the principal photoreactive processes responsible for the observed changes in petro-
leum exposed to weathering.
In fact, according to Ellis and Wells, 31 enough energy is provided by the
shorter wavelength solar radiation (2100 to 3100 A) to conclude that several of the
foregoing reactions may serve as the primary initiator of autoxidation of petroleum.
Figure 1 shows the relationship of the energy content of visible light and UV radia-
tion to its wavelength. It can be seen that the energy needed to break a number of
bonds can be provided by solar radiation at the earth's surface.
11
-------
no
C-H (OLEFINS)
C-C (ALIPHATIC)
C-CI
C-H(1, 4- DIOLEFIN)
500 600
WAVELENGTH
Figure 1. Bond-Dissociation Energies and Double-Bond Activation Energies for
Chemical Groupings Present in Some Autoxidizing Media (Reproduced
from Scott)
12
-------
TABLE 1
30
•TAB + C
+ F
-•ABC
*ABC1(S ) or
ABC(S1)|
or
ABC (IV
PRIMARY PHOTOCHEMICAL PROCESSES
Dissociation into radicals
Intramolecular decomposition
into molecules
—-(ABCH) + R
Intramolecular rearrangement
Photoisomerization
Hydrogen-atom abstraction
(1)
(2)
(3)
(4)
(5)
•*ABC
ABC + Products
Photodimerization (photoaddition) (6)
Photoionization (7)
"External" electron transfer (8)
"Internal" electron transfer (9)
For the various general components to be found in petroleum, photochemical
modification (at 2500 to 3300 A) can occur by the mechanisms shown in the following
series of equations:
Aldehydes;
Ketones:
RCHO + hv
+ HCO, R = alkyl or aryl
CH COC H +hv
o Z 5
Organic Acids;
RCOOH + hv
-~CH3 + COC2H5
•R + COOH (or CO2 + H)
•RCO2 (or R + CO2) + H
-RCO (or R +CO) + OH
•RH + CO^
13
-------
Paraffinic Hydrocarbons: RCH R + hv —^RCR + H (1470 A)
^ 2
Aromatic Hydrocarbons: R_(())—CHR + hv—»-R
CHR + R
It can be seen that a number of radicals can be formed from petroleum constituents
(aldehydes, ketones, acids) by photochemical dissociation. These radicals then can
serve as initiators of the autoxidation process. The reaction mechanism shown for the
paraffinic hydrocarbons, however, will be infrequent under ambient conditions since
it occurs primarily at 1470 A. The aromatic hydrocarbons, in general, exhibit a
greater stability to UV and visible radiation than the paraffinic hydrocarbons, even
though they absorb strongly in these regions. Aside from the protection afforded by
the conjugated ring system, this may be due to the fact that photo-decomposition
modes of the aromatic molecules occur only with low quantum efficiencies in the
first absorption bands (0 = 0.0001 or less).30 In those compounds having alkyl
groups attached to an aromatic ring system, exposure to light in the first absorption
band may lead to rupture of the alkyl H-C or C-C bonds to form the benzyl-type
radical. 32 All of the above mechanisms and the degree to which they are operative
are affected by temperature, light intensity (determined by weather conditions, sea-
son, and geographic latitude), wavelength of the incident radiation, and concentra-
tion of reactants. Ellis & Wells31 report that
The color properties of petroleum may be destroyed by solar UV rays.
Petroleum oils intended for use as lubricants and which may have gum-
ming properties due to unsaturated components are improved in quality
by exposure to UV radiation in an inert gas. Apparently polymerization
occurs, resulting in the formation of saturated bodies.
The addition of sulfur (0. 01 to 0.1%) was also deleterious, whether in the presence
or absence of oxygen. In the presence of air or oxygen, it caused a marked increase
in the amounts of color and gum formed. Acid formation and peroxide number were
also greater than in the absence of sulfur, ft also appears "that sunlight lowers the
induction period for gum formation but water has the opposite effect" with stored
samples of petroleum. In addition the rate of gum formation was observed to triple
for approximately every 20° F rise in temperature.
14
-------
33
Tipson has reviewed the various mechanisms of polycyclic hydrocarbon
degradation by oxygen and UV at low temperatures (< 100° C). Some of those reac-
tions which are typical of the autoxidation mechanisms applicable to crude petroleum
are as follows:
On treatment with O2 and UV light for 30 hours at 70° C, fluorene
in benzene gives the stable 9-fluorenyl hydroperoxide:
H,
H
hv
70° C
Treatment of 2a, 3,4, 5-tetrahydroacenaphthene [sic] with O2 and
light for 150 hours at room temperature gives the hydroperoxide:
CH
hv
25° C
H2°
H,
OOH
1,2-dihydronaphthalene at 18° C for 430 hours gives the unstable
para-transannular peroxide with O , which rearranges to a trimer.
A
hv
I
O
O-
J 3
Solutions of anthracene stored in the open air for long periods of
time form the insoluble anthroquinone. Molecular oxygen can also
be added across the central ring to give the 9,10 peroxide. Naph-
thacene and pentacene are photo-oxidized more readily, and the
reactions occur so readily with hexacene and heptacene that solutions
of these hydrocarbons, exposed to air and light, are almost instan-
taneously photo-oxidized. The nonlinear hydrocarbons react with
oxygen less readily than the linear polyacenes, or not at all.
15
-------
Under the action of UV light, transannular addition of oxygen occurs at the most reactive
centers of linear aromatic hydrocarbons if the two reactive centers are para to each
other (as in the anthracenoid hydrocarbons) giving transannular peroxides.
It seems probable that under these conditions the absorbed light
excites the dienoid structure to a diradical which then rapidly reacts
with the oxygen diradical. The overall process involves consumption
of the diradicals with no regeneration; hence these are not chain reac-
tions. The transannular
reaction of oxygen is further facilitated by the presence of electron-
donating substituents (e.g., methyl or phenyl) at the reactive centers.
2.2.4 Metalloporphyrins
Another important component of crude petroleums is the metalloporphyrins.
Although the porphyrins found in petroleum are stable to heat and to concentrated
acids, 4>35'36 porphyrins, as a category, are relatively unstable to light, peroxides,
and to various forms of ionizing radiation37' 38 (X-, B- and \-irradiation). In most
of these cases, the nucleus is unaffected; however, in the case of oxidation initiated
by photochemically generated hydroperoxides, the side chain may be modified so that
the porphyrin can become water soluble. In view of the fact that petroporphyrins
and their complexes, being large, planar aromatic molecules, can act as surface-
active agents, they would tend to concentrate at oil water interfaces, ' ' ' ' '
thereby increasing the probability of cation exchange and oxidative degradation. Such
processes (conferring a change in the partition coefficient of the porphyrin) would
result in an alteration with time on the order of weeks of the final metalloporphyrin
content of the petroleum. The extent of the partition of porphyrin between the oil slick
(organic phase) and the water depends on the hydrophobic nature of the ring and the
hydrophilic nature of the central nitrogens and the polar side chains.42
43
In an interesting publication by Wolsky & Chapman, in which the authors
describe their attempts to isolate and identify vanadium and nickel porphyrin aggre-
gates from Boscan asphaltines by the Groennings' method,44 a few oxidation studies
were performed to check the stability of the porphyrin aggregate. The results showed
that oxygen alone had little effect but, together with UV light, caused rapid decompo-
sition of the porphyrins. Although the authors do not describe the details of their
oxidation study, nor do they give any indication as to the identity of the decomposition
products, it is of more than passing interest that UV radiation demonstrated so pro-
nounced an effect on the putatively stable porphyrins.
16
-------
The preceding discussion points out the physicochemical lability of several
of the natural constituents (aliphatic, aromatic, porphyrin) of crude petroleum. These
facts demonstrate the potential difficulties one is faced with in attempting to identify
unequivocally a specific crude oil solely by virtue of the concentrations of its intrinsic
components, especially with a sample that has undergone prolonged exposure to many
indefinable and uncontrollable environmental and climatic conditions.
17
-------
2.3 Microbial Oxidation
Biochemical degradation of oil slicks is a natural consequence of oil spilled on
anything but distilled water in a sterile inert atmosphere: Both fresh and salt water
normally contain a diverse microfloral population, and many microorganisms of
aquatic habitat are known to metabolize petroleum compounds, especially hydrocarbons,
via oxidative pathways. Only one other generalization should be risked: Even under
optimal conditions for microbial growth, the heavy, complex molecules characteristic
of aged oil slicks are not rapidly metabolized to either stable cellular components or
to oxidative end-products; however, it should be borne in mind that while disappearance
of the slick is one thing, its biochemical alteration is quite another matter. If due
respect is paid to the scope and complexity of oil-pollution environments, to the diverse
species distribution of microflora in open water, to the variations in metabolic com-
petence within even such well-known hydrocarbon-utilizing species as Pseudomonas
aeruginosa, to the chemical diversity of crude oil, and to the numerous reactions that
may in time alter the molecular makeup of the slick, the whole question of microbial
oxidation must be approached with extreme circumspection.
According to a most authoritative treatise on marine microbiology, there are
approximately 700 species of marine microflora. Bergey's Manual of Determinative
Bacteriology,47 the standard American reference, recognizes 183 bacterial species of
marine origin; ZoBell & Upham list 165 species that are not recognized in Sergey;
and Brisou4" lists 57 more species that are not listed in either Bergey or in ZoBell &
Upham. However, as Kriss himself points out; "Because of overlaps and technical
obscurities in the art of microbial systematics, we should not conclude that Bergey's
list should be doubled. " Nevertheless, he suggests that 700 is perhaps a more realistic
number. Kriss may be wrong, but it would be hazardous to argue the point because he
has conducted an immense field survey of marine microflora at all depths and at a
great number of locations in the Black Sea, the Greenland Sea, the Norwegian Sea, the
central Arctic Ocean, the Atlantic Ocean, the Indian Ocean, and the Pacific Ocean.
He has also investigated the biochemical activity of more than 3, 000 strains of marine
microfloral heterotrophs. Although he did not use hydrocarbons as growth substrates,
his findings have great significance because they do reveal the extraordinary metabolic
variations in marine microorganisms as a function of latitude; moreover, these varia-
tions are likely to be paralleled in the ability to utilize hydrocarbons. His findings are
of the first importance, and should be quoted at length:
A world-wide geographical distribution pattern of microbial forms occur-
ring in the sea and using readily assimilable organic matter in their vital
processes clearly emerges [from our world-wide study]. The polar
regions — Arctic, Antarctic, sub-Arctic and sub-Antarctic — are areas
of very low saprophytic bacteria population density. Although the produc-
tivity of the ocean water is conditioned by a rich plant life, the metabolic
products and dead remains of the organisms inhabiting these parts of the
ocean do not create such a concentration of nutrient material for sapro-
phytic microorganisms as is observed in the equatorial-tropical regions
of the ocean.
18
-------
That this is so can be seen from the high concentration of microbial
life (saprophytic bacteria) in the equatorial and tropical zones of the
Pacific, Indian and Atlantic Oceans. These zones are in marked con-
trast to other geographical zones of the world's seas and oceans in
respect to their microbial wealth. It is interesting that a high concen-
tration of bacteria assimilating labile organic matter is observed in
those geographical regions of the ocean which are poorest in plant and
animal life by comparison with the high latitudes. [p. 274]
It is a striking fact that the equatorial-tropical zone is distinguished
by the low concentration of biochemically active microbial forms among
the saprophytes inhabiting it. In the Arctic, sub-Arctic, Antarctic and
sub-Antarctic regions of the ocean the percentage of bacterial strains
causing thorough breakdown of proteins proved to be four to sixty times
greater, and the percentage of those fermenting carbohydrates two to
eleven times greater, than in the equatorial zone.
These data suggest that the process of organic decay and liberation of
nutrient elements is more vigorous in high latitudes. Although the
saprophyte population there is not so great as near the Equator and the
biochemical activity of the saprophytes is affected by the low tempera-
tures, nevertheless, the presence of a comparatively large number of
microbial forms possessing diversified enzymatic activity makes for
more thorough conversion of organic matter in the high latitudes than
in the tropics. [Emphasis supplied. ]
The result is that in the Arctic, sub-Arctic, Antarctic and sub-Antarctic
regions the sea water contains a higher concentration of the nutrient sub-
stances essential for the development of aquatic vegetation.... [pp. 276-
277].
The reasons for the poverty of tropical water in biochemically active
microbial forms and for the higher concentration of these forms in high-
latitude water are still unknown. The relatively high concentration of
allochthonous organic material readily assimilable by microbes in tropi-
cal waters, which accounts for the high density of the saprophytic popu-
lation in the equatorial-tropical zone, may, perhaps, obviate the neces-
sity for enzymatic reactions ensuring total utilization of the organic
material. In high-latitude water the scanty food supply leads to the ela-
boration of adaptations in the saprophytic bacteria, enabling them to make
more economic use of the organic matter by taking advantage of its more
thorough decomposition and conversion, [p. 277]
19
-------
In short, cold waters are much more likely to contain petroleum-metabolizing bac-
teria than warm waters. There is every reason to think that the microbial attack
on the Ocean Eagle's slick (in Puerto Rico) was drastically different from the
Torrey Canyon's (in the English Channel).
Long lists of microorganisms capable of metabolizing petroleum have been
published by Beerstecher and by Davis. We find, however, that the work of
McKenna & Kallio and of Van der Linden & Thijsse is much more rigorous and
reliable; these authors are also the best guides to the literature. It is well to begin
this discussion with McKenna's warning:
Codification of the numerous microbial forms capable of degrading
or assimilating hydrocarbons has not been attempted on several
grounds. Many reports carry descriptions of new isolates that are,
to be kind, meager. Bacteria, yeasts, and fungi possessing the
capability are legion, new members of this group are continually
being found, and such a list would be relegated to senecide [sic] before
it appeared in these pages. Flihs prepared a list of organisms capable
of growth at the expense of hydrocarbons — it comprises over
100 yeasts, bacteria, fungi, and actinomycetes. It is clear that if
attention is paid to detail and if a modicum of experimental imagina-
tion is indulged in, a wide variety of microorganisms, not originally
selected for the character, can be shown to be hydrocarbon-utilizing
organisms. Appropriately selected and cultivated organisms grow at
the expense of diverse hydrocarbons as well as, or better than, on the
common substrates widely used in microbiology. Under a variety of
conditions oxidation products appear in culture fluids and can be sub-
jected to chemical analysis; such chemical identification has permitted
meaningful formulation of microbial hydrocarbon degradation path-
ways. 52
Although there are hundreds of articles on hydrocarbon microbiology
(Van der Linden & Thijsse53 listed 285 of the most important in 1965), most of them
deal with molecules having less than 20 carbon atoms; all the rigorous scientific litera-
ture is suggestive but not relevant. While not every species has been able to metabo-
lize every n-alkane in every experiment, Van der Linden & Thijsse had no difficulty
finding at least one microbe that had been reported to attack each n-alkane at least
some of the time. Isoparaffins have been less studied, but they too (at least the lighter
ones) are metabolized (often in several ways) by several common bacteria of marine
habitat. Van der Linden & Thijsse suggest that olefins may be more readily oxidized
20
-------
than alkanes, and advance several cogent reasons in support of their argument.
Although there have been very few studies of cycloparaffins (naphthenes), they seem
resistant to most microbial attack:
Cycloparaffins seem to be poorly utilizable by microorganisms in
general, and, in consequence, bacteria that would enable the study
of their biodegradation pathways have only been encountered twice. . . .
[In one of these studies, however, ] no experimental details are
given and no analytical methods are mentioned. This study, there-
fore, cannot be critically examined.
The situation is not much better for aromatic and alkyl-aromatic compounds: Little
is known for certain, and very little indeed is known about the rather heavy and com-
plex molecules that are likely to be found in oil slicks that are several days old.
The situation is further complicated by the molecular diversity of crude oils.
While most studies of microbial hydrocarbon metabolism have tested only one hydro-
carbon at a time, there is strong evidence that (paradoxically) two different hydro-
carbons may be more readily oxidized together than separately:
Hydrocarbons seemingly resistant to both oxidation and assimilation
may be oxidized if attacked by organisms concomitantly oxidizing other,
utilizable hydrocarbons. Co-oxidation. . . may be a widely occurring
phenomenon in microbial metabolism. . . . The extraordinarily high
yields of co-oxidation reactions are prime targets for technological
developments in commercial hydrocarbon conversion by microbial
reactions, but no such studies have been published.
Another phenomenon noted during investigations of bacterial hydro-
carbon degradations in oxidation of a substrate (commonly but not
invariably a hydrocarbon), followed by metabolite accumulation in
the face of an apparent inability of the relevant organism to assimi-
late the oxidation product — materials so oxidized are somewhat
inelegantly referred to as "nongrowth" substrates.... Like co-
oxidation, oxidation of nonassimilable substrates may be a wide-
spread phenomenon in the microbial kingdom; together with co-
oxidation, the process complicates studies of hydrocarbon metabo-
lism based on assimilation studies, and raises intriguing questions
of substrate specificity to say nothing of the assimilatory complex
in bacteria. 52
The process of microbial oxidation of oil slicks, known to be rather slow at
sea, also seems to be slow in the laboratory. It has recently been shown that oil
slick collected at sea but cultivated in the laboratory is barely assimilated after
several weeks of microbial growth.
21
-------
Because of the various ways that aquatic bacteria can alter oil slicks, chemical
tracing of slicks to suspected sources may be difficult to prove in court. This proposi-
tion is a fortiori true when added to the evidence for straight chemical changes in oil
slicks (section 2. 2 of this report).
22
-------
SECTION III
PASSIVE TAGGING
3.1 General
Passive tagging is a convenient circumlocution for no tagging at all: Nothing
is added to the oil, but the slick sample is subjected to various methods of physical
and chemical analysis. The sample itself must provide all the evidence for detection
and identification. By an obvious extension of a figure of speech, passive tagging is
sometimes called physico-chemical "fingerprinting. " Petroleum geochemists have
devoted considerable effort to documenting chemical differences among crude oils,
and anyone who cares to take the trouble will have no difficulty finding major dif-
ferences between most Tertiary crudes and most Paleozoic crudes. However, the
problem of legally identifying oil slicks (i. e. , of proving liability for oil pollution
in a court of law) has more to do with a branch of criminology, which, for want of
a better phrase, we might call "forensic chemistry, " than with geochemistry,
analytical chemistry, or petroleum chemistry. For the central problem in this
application of forensic chemistry is: Can this sample of slick be proved irrefutably
to have come from a tanker known to have been in the area, or from a pipeline
nearby, or from a storage tank, or what have you?
The most direct method of proof (and the only one that we have seen suggested)
is to compare the slick with the unspilled cargo (or fuel) of the suspected source of
pollution. However, as we have already seen in section 2, unless the slick is very
fresh, we have every reason to believe that it will differ from the suspect cargo: The
slick will have been irradiated, mixed with complicated aqueous solutions (sea water,
polluted fresh water, or brackish swamp water), exposed to oxygen and air pollutants,
and intimately associated with a variety of aquatic microbes; depending on local con-
ditions, it will have been partially evaporated, emulsified, and photolyzed, and may
have undergone autoxidation, photopolymerization, enzymatic oxidation, and ionic
exchange.
What part of the fingerprint can be stable to all these conditions? Heavy and
highly branched alkanes are excellent candidates; so are the heavy and complex hydro-
carbons of mixed structure that are common in lubricating oil. However, the heavy
branched alkanes are (in general) exceedingly rare in oil. ^4 Bestougeff,55 for
example, has pointed out that:
The distrubtion of branched-chain paraffins in different fractions of
a crude oil is unequal, and, as in the case of n-paraffins (light crude
oils), drops rapidly as a function of the molecular weight.
23
-------
The very complex hydrocarbons that are characteristic of lubricating oils are even
more difficult to identify precisely: Whitehead & Breger56 do not mince their words:
The compounds thus far isolated from crude oil are rather simple
and include basic paraffinic, cycloparaffinic, and aromatic types.
A few more complicated multiple-ring structures, such as adamantane
and some bicyclooctanes, have also been identified, but the com-
plex structures of the lubricating oil fraction can only be described
in very general terms. Here, where the compounds may be com-
posed of mixed aromatic, cycloparaffinic, and paraffinic structures,
the number of possible compounds becomes almost infinitely large
and the likelihood of positive identification of any one becomes
vanishingly small.
Even though several hundred compounds have been identified in crude oil,
the identified compounds "probably account for well under 50% of the constituents
of a crude petroleum. "56 It should be apparent that the hydrocarbons in slicks
of crude and heavy oils will contain an even higher percentage of unidentified
structures. One must therefore conclude that although heavy, complex hydro-
carbons are likely to be stable constituents of oil slicks, they will not serve as
passive tags because they are exceedingly rare, exceedingly difficult to identify,
or both.
What other stable constituents might be used to fingerprint oil? Trace metals
(especially vanadium and nickel) and isotopic ratios have been suggested, and these
will be examined in detail in the next two sections. These two approaches to passive
tagging have been chosen for special study for more than their analytical attractive-
ness: Each in its way illustrates the difference between forensic chemistry and
analytical chemistry.
In general, the requirements on passive tags should not be different from the
requirements on active tags. Both kinds of tags should be unchanged by weathering
and microbial action, soluble or dispersible in oil, insoluble and nondispersible in
water, and involatile. However, there are special circumstances in which some of
the requirements on passive tags may be relaxed. Where oil pollution is infrequent,
where there are few possible sources of pollution, and where weathering may be
assumed insignificant or inconsequential (because of the slick's freshness), several
methods of chemical analysis—supplemented by a great deal of convincing circum-
stantial evidence of an extra-chemical kind—may identify oil slicks by passive tags
in the spilled oil. Some characteristic methods of this sort are discussed in
section 3. 4. The reader is reminded that these methods are valuable only where
the circumstantial evidence is already convincing; they must be considered as
presumptive, not confirmatory, tests.
24
-------
3.2 Metals
Many inorganics have been found in petroleum ash: Ag, Al, As, B, Ba, Ca,
Ce, Co, Cr, Cu, Fe, Ga, Hg, K, La, Li, Mg, Mn, Mo, Na, Nd, Ni, Pb, Pt, Ra,
Rb, Rn, Si, Sn, Sr, Te, Ti, Tl, V, U, Zr, and Zn. The literature on trace elements
in crude oils has been the subject of several very important reviews. " Many of the
elements listed above are not frequently found; others are frequently found in low
concentrations; only a few, however, are frequently found in petroleum ash in concen-
trations above 0.1%, and of these few, special importance attaches to Ni and V.
These two metals are important because of the problems they cause industry:
The volatility of certain metallic compounds may lead to their presence
in distillate fractions, their concentrations increasing with the depth of
the cut. Some of these alter the selectivity of cracking catalysts,
increasing the formation of coke and hydrogen and decreasing the yield
of gasoline. The more common metals of this type are nickel, copper,
vanadium, and iron. Non-volatile vanadium, which concentrates in the
residuum from which heavy fuel oil is produced, is extremely corrosive
to refractories used in furnaces. Its most serious destructive effect is
upon fire-clay bricks. Vanadium oxides form low melting eutectics with
clay, producing hard, glassy slags which are extremely difficult to
remove. 58 [pp. 4-6]
Nickel is probably present to some extent in all crude oils, much of it
in a volatile form. As an active dehydrogenation catalyst it alters the
selectivity of cracking catalysts, and accurate determination of it in
distillate feed stocks is therefore of importance.58 [p. 179, emphasis
supplied. ]
... Vanadium is present in crude oils combined in porphyrins, some
of which have low enough boiling points to permit the element to distill.
Thus, it is found in gas oil cuts used as feed stock for catalytic cracking.
In common with other elements of the first transition series, vanadium
is a dehydrogenation catalyst with an activity from one-tenth to one-
fourth that of nickel. 58 [p. 235]
Their presence also has geochemical significance, though authors of reviews (all of
them magnificently qualified) often disagree on how much, if any. Based on work of
Vinogradov and of Radchenko & Sheshina,59 Manskaya & Drozdova60 have concluded
that
Sulphur-poor petroleums contain significantly less porphyrin than
sulphur-rich petroleums. In the former case, the porphyrins occur as
nickel complexes, in the latter as vanadium complexes, [p. 145]
25
-------
Dunning61 has savagely attacked these conclusions:
These conclusions seem illogical from the philosophical approach and
doubtful even if the data are accepted as being generally applicable
[They] illustrate the danger of drawing generalizations from data involv-
ing incomplete studies or too few oils. [pp. 398-399]
/>o
In a more recent review, however, Hodgson et al. explicitly support the Russian
conclusions:
Radchenko and Sheshina have drawn attention to the relationship which
exists between vanadyl porphyrins and sulfur in crude oils .... Evidence
has been produced by Park and Dunning ... to discount the possibility
. .., but the fact remains that there is some relation between sulfur in
crude oil and the vanadyl porphyrin content .... It is well known that
sulfur and vanadium compounds are very important factors in redox
systems—possibly this could account for the simultaneous entry of the
two substances into crude oil. [p. 247]
Vanadium and nickel in oil are also of interest to students of water pollution
because they have been seriously proposed as valuable means of identifying the origin
of oil spills. Canevari has put the matter most directly:
The method [we propose] simply entails measuring the vanadium and
nickel contents of the unknown oil pollutant. There is a very specific
amount of these metals associated with every crude—in essence con-
stituting an individual metallic trait. There are other chemical and
physical properties for the various crudes, but these are subject to
change during weathering/aging [sic]. The metallic components, how-
ever, are stable and are not affected by weathering. By simply com-
paring the metallic content of the oil spill against [sic] that of the
suspected source, it is possible to limit the probable source to a
specific vessel, pipeline or storage facility. The technical feasibility
of the metallic fingerprinting [sic] was demonstrated by
Dr. George Milliman and Dr. David McMahon in a systematic analyses
[sic] of actual oil spill samples as well as control samples.
Guinn & Bellanca64 have argued a similar but more complex case for vanadium,
sodium, manganese, and cobalt; they have concluded that their method
... offers definite promise for the trace-element characterization
matching [sic] of an actual oil-slick sample with one sample of fuel
oil—out of a number of suspect samples. The broader and deeper
study now being initiated should establish this possibility more
rigorously, and provide a mathematical basis for the interpretation
of results obtained on oil and oil-slick samples obtained by enforcement
26
-------
agencies in the investigation of individual violations of oil pollution laws.
The various characterizing trace elements, occurring in the original
crude oils largely as essentially non-volatile metal porphyrins, remain
in the residue left when the volatile components of the crude evaporate
(as from an oil slick), and when the light ends are distilled off in refin-
ing. Thus NAA [neutron activation analysis] matching of the levels of
these trace elements—between an oil slick and its polluting source—may,
for the first time, provide a method of identification of the source of
oil pollution of waters that will aid in the enforcement of present oil
pollution laws.
The notion that trace metals, especially vanadium and nickel, might be suitable
passive tags is by no means new. In 1955, Johannesson65 reported that he had success-
fully used vanadium and nickel to identify oil slicks in Wellington, New Zealand, and
had used this evidence to prosecute offenders. Because Johannesson's work has not
been cited by current proponents of trace-metal tags (e.g., Canevari63 and Guinn &
Bellanca64)? we quote a section of Johannesson's article here so that the reader may
see how little the argument has changed over the years (although the analytical methods
for determining trace-metal content have, of course, changed with the times):
I have successfully used ... [a] method depending on the nickel and vana-
dium contents of the oils. The amounts of these elements present in oils
varies [sic] considerably, as shown ... by Shirley [in a 1931 paper] ....
Vanadium and nickel are held in the oils as co-ordinated complexes with
porphyrins, and are not affected by contact with sea water; and, further,
the concentrations of these elements in sea water are insignificantly
small compared with those in the oils.
Samples weighing 3 to 5 g are sufficient for analysis.... Ashing was
carried out in a platinum basin, ethanol being added to facilitate
removal of the water and the burner flame being directed, initially,
on the surface of the liquid. [The reader must be warned that this
method (direct wet ashing) has been severely criticized by all current
authoritiesSV, 58,66 On the inorganic analysis of petroleum because
volatile constituents, especially organometallics, are very likely to
evaporate; vanadium and nickel porphyrins are especially emphasized
as compounds likely to be lost by direct ashing. McCoy^® has pointed
out that even in heavy residual oils, nickel porphyrins are evidently
largely lost by direct ashing; see p. 184 of McCoy's excellent text.]
A number of methods for the determination of vanadium and nickel
in oils has been published and any of these may be used....
27
-------
TYPICAL RESULTS
Vanadium Nickel Ratio of
pentoxide, oxide, vanadium
mg per mg per pentoxide
100 g 100 mg to nickel
Source of oil of oil of oil oxide
Recovered from
harbor 14.0 13.9 1.01
From suspect
source 12.0 10.2 1.17
A distillate
fuel oil trace trace
Abadan 6.8 1.2 5.65
Curaqao 40.0 5.1 7.85
San Pedro 11.0 10.7 1.03
These results established that the recovered oil was the same as that
from the suspected source, allowance being made for evaporation and,
further, that they were of San Pedro [Los Angeles County] origin.
Liability was later admitted by the offending party and so the origin of
the oils was confirmed.
The claims made in these proposals rest on several assumptions that are pre-
sented as fact. The central assumptions are: (1) The metallic content (especially
vanadium and nickel) of crude and fuel oil is not affected by weathering; (2) the metallic
content (more precisely, the ratio of metal to metal—because the metallic content
increases as the slick evaporates) of the slick bears a known or fixed relation to the
corresponding content of the unspilled oil; (3) these trace elements exist in oil as
"essentially non-volatile metal porphyrins" that are chemically and physically stable
in the slick. The facts, however, seem to be rather different from the claims.
Neither vanadium nor nickel is present in petroleum as a single molecular
species; there are a variety of molecules containing these metals in petroleum, and
not all of these molecules are either stable to oxidation or insoluble in water.
Costantinides & Arich66 have reviewed the evidence:
Occasionally, traces of metals are found in the gas-oil fractions. Metal
contents increase regularly with increasing molecular weight of the
28
-------
distillate fractions and reaches [sic] its maximum in the residue. .. . The
occurrence of metals in distillates may be explained by a mechanical
entrainment of residue or by distillation of the metals in the form of
volatile metal-organic compounds. The latter view is generally accepted
as a result of investigations.. .indicating that some metal complexes,
isolated by elution chromatography and by extraction with aqueous
pyridine, are concentrated in the more volatile distillate fractions....
These findings lead to the conclusions that both vanadium and nickel
compounds of widely varying molecula^ weights exist in crude oils.
[pp. 164-165]
Dunning has made enormous contributions to our understanding of metals in
petroleum. He has also made it abundantly clear that many of these metals (especially
nickel) exist in petroleum as water-soluble compounds:^
Work of Erdman and associates (1956) and Dunning (1953) showed that
[petroleum] porphyrins and their complexes generally are interfacially
active. As a result, they might be expected to be concentrated at
interfaces, and Dunning and associates (1954) used the water spray
method.. .in an attempt to concentrate the porphyrin-metal complexes
of North Belridge oil.... The large excesses of metals and nitrogen
over that accounted for by the metal-porphyrin complex contents of
these extracts points [sic] to the postulate that other metal-nitrogenous
substances may be present.... The nickel and vanadium contents
were increased more than the porphyrin or nitrogen contents in the
pentane-insoluble portion of the water-spray extract. Apparently part
of these metals in crude oil was not complexed with large nitrogenous
molecules. This is in agreement with Garner and associates (1953).
The rather high (0. 01 weight %) nickel concentration in the water-
soluble extract provides further evidence that the crude oil contained
uncomplexed nickel. The nitrogen, nickel and vanadium contents of
these water-spray extracts vary in the same manner as do the porphyrin
contents. The ratios of porphyrin contents to the total nickel and
vanadium contents are 3.0, 4. 6, 4.1, and 6. 8 in the propane-insoluble
extract, whole crude oil, extracted crude oil, and propane-soluble
extract, respectively. As mentioned above, the ratio of porphyrin to
vanadium or nickel in a porphyrin complex is about 10:1. Thus,
insufficient porphyrin was isolated in any of these fractions to complex
all of the nickel and vanadium present. Extensive chromatographic
and solvent-extraction procedures may produce petroleum extracts
which contain vanadium primarily in the form of the vanadium-
porphyrin complex. This observation has lead to some misinterpre-
tations in the literature. It should be emphasized that such extracts,
containing vanadium primarily as the porphyrin complex, are rather
commonly obtained from crude oils that do not contain enough por-
phyrin to complex even a major part of the vanadium present. .. . For the
29
-------
present, it must be concluded that much of the vanadium of most crude
oil is not [emphasis in the original] complexed with substances capable
of extraction as porphyrins by current methods, [pp. 395-397]
In every other major review of petroporphyrins or petroleum trace metals,
similar arguments are presented based on a wealth of evidence. Sometimes, the
arguments are even more damning; thus, Hodgson et al.:
Porphyrin complexes are subject to attack by other agents.... For
example, porphyrin metal complexes may be destroyed to some extent
through oxidation if the... oil should come in contact with oxygenated
ground waters. It is difficult to assess the magnitude of this possibility
since few data are available. Costantinides and Batti (1957) reported
that 90% of the porphyrin complexes in an oil were destroyed in a few
hours at about 200°C under mild oxidizing conditions.... [p. 250]
We have offered these extracts from authoritative reviews of the literature so
that the reader may see for himself the great chasm that separates the claims of those
who would use trace metals as passive tags for oil spills from the disinterested
appraisals of the scientific evidence. It is perfectly plain (and thoroughly established)
that much of the vanadium and nickel in petroleum is water-soluble and volatile. Even
the rather stable compounds of vanadium and nickel with petroporphyrins are somewhat
subject to oxidation. Moreover:
Oxidizing conditions usually appear to result in polycarboxylic porphyrin
structures rather than the reduced porphyrin side chains characteristic
of oil environments. 62
While the decarboxylated porphyrins are only very slightly soluble in water, the car-
boxylated porphyrins are known to be rather water-soluble and to have a marked
affinity for hydrocarbon-water interfaces.61' 62 Indeed, Falk67 has demonstrated
an almost linear relationship between the number of carboxyl groups on a porphyrin
and the porphyrin's mobility in a solvent system of lutidine and water.
Thus far, we have concentrated our attack on vanadyl and nickel porphyrins
because these compounds are known to be rather stable; however, many forms
of these compounds are also volatile and water-soluble. It is obvious that volatile,
water-soluble compounds are not going to be stable in oil slicks; furthermore,
vanadium porphyrins seem to be more stable (to oxidation, replacement, water solution,
and evaporation) than nickel porphyrins, so that, as the slick grows older, the vanadium/
nickel ratio in the slick will probably get progressively higher.
30
-------
The notion of using sodium as a trace characteristic in oil slicks is even more
poorly advised. Sodium is common in petroleum but it is also the most common
cationic metal in sea water; it is therefore readily available to participate in exchange
and replacement reactions with the oil slick. The petroporphyrin nucleus is an obvious
target for exchange reactions with sodium ions from the sea; unfortunately, these
sodium porphyrins are not very stable in the presence of other cations:
... Alkali metal ions in porphyrin molecules are readily replaced by small
divalent cations, by large divalent cations or by small alkali metal ions.62
[p. 194]
From studies with lithium, sodium, potassium, divalent lead, mercury,
copper, silver, cobalt, magnesium and tin, two rules for the replacement
reactions [of petroporphyrins] were formulated: (1) the order of increasing
stability is alkali metal, large divalent metal and small divalent metal;
(2) the metal porphyrin with the smallest metal is the most stable.62
[p. 203]
Bonding of protons and metal ions to the [porphyrin] pigment molecule
is largely covalent in character..., suggesting that the cation probably
does not undergo exchange reactions with cations in solution. Isotope
exchange reaction [sic] involving iron, cobalt, copper, zinc, and magnesium
complexes verify this with negligible rates.... As would be expected the
complexes of porphyrins incorporating monovalent cations show rapid
exchange. [Emphasis supplied.] Unfortunately, the exchange reactions
with divalent cations were not exhaustively examined; therefore, although
negative results were obtained, it is possible that metal complexes of
these divalent cations may undergo exchange at low but measurable
rates. 62 [p. 203]
Since sodium is a rather small monovalent cation, it will "show rapid
exchange" and be "readily replaced" by divalent cations, many of which (e. g.,
calcium, magnesium, iron, etc.) are easily available in sea water. Other chemical
reactions between sea-water sodium and the oil slick may be expected to be more
stable; however, one general characteristic of sodium compounds is water solubility,
and water-soluble compounds are not likely to remain in an oil slick for very long.
In summary, the trace metals in petroleum make an unreliable "fingerprint, "
for the fingerprint is very likely to evaporate and be washed away at sea.
31
-------
3. 3 Sulfur-Isotope Ratios
A group at McMaster University in Canada has been studying sulfur-isotope
ratios in a variety of geologically important materials (including petroleum) since
the late 1940's. Their work has particular value for our purpose because it was not
conceived with oil tagging in mind; it was purely a geochemical investigation of iso-
tope fractionation. Thode has summarized the most important arguments and con-
clusions, which we recapitulate here.
Isotope fractionation is an established and important geochemical process:
Ocean carbonate is rich in O-18 with respect to ocean water, plants are enriched in
C-12 with respect to atmospheric carbon dioxide, and sea sulfate is rich in S-34 with
respect to the sulfur found in meteorites and basic igneous rocks. Although there
are nine sulfur isotopes, only four of them occur in natural sulfur (S-32, -33, -34,
and -36); all four are stable, but they are not equally distributed: S-32 predominates
(95. 0%), and S-34 (4. 22%) accounts for most of the remainder. The ratio of these
two most common isotopes is often standardized with respect to meteorites because
this ratio is virtually invariant;71 in the earth's mantle, however, the content of
S-34 varies by approximately 100%. The 34/32 ratio has been studied in many
materials, including petroleum and sea water: In sea water the ratio is +20 ±2%,
while in petroleum it varies from -8% to +33%—indeed, it varies from -8% to +28% in
petroleums from Utah alone, and the apparent significance of this phenomenon has not
been lost on proponents of passive tagging.
The fundamental fact in the geochemistry of sulfur isotopes is that S-34 tends
to sulfate and S-32 tends to sulfide: S-32 is the more readily reduced and S-34 is the
more readily oxidized isotope. The underlying chemical reaction is:
H2(S-34) + (S-32)04 ^==^H2(S-32) + (S-34)C>4
ft
The kinetics of the isotope-exchange reaction are also known: The reaction goes to
the right because k^ is larger than k^. The equilibrium constant (K = 1. 071) tells
us that at equilibrium, the sulfate will favor S-34 by 71%. Similar calculations show
that hydrogen sulfide is 25% richer in S-32 than in S-34. These reactions are not
sensitive to either temperature or concentration.
Enzyme-catalyzed reactions, however, are sensitive to both temperature and
concentration. For example, when the anaerobic bacterium Desulfovibrio desulfuri-
cans reduces sulfate to sulfide, the isotope-fractionation effect is not a constant 25%:
It varies from 0 to 25%, depending on conditions. At low sulfate concentrations or
at rapid metabolic rates, the diffusion of sulfate is rate-controlling and there is little
isotope effect; at slow metabolic rates, however, the kinetic isotope effect approaches
32
-------
25%. Under the usual conditions of anaerobic reduction in shallow muds, the isotope
effects are approximately 15%. The identical argument should also apply when sul-
fide is oxidized to sulfate by photosynthetic and colorless sulfur bacteria (e.g. , the
Thiobacteriaceae). Although Jones & Starkey^ could not demonstrate isotope frac-
tionation when Thiobacillus thiooxidans aerobically oxidized elemental sulfur to sul-
fate, they pointed out that:
This does not eliminate the possibility that there is fractionation
during its oxidation of soluble sulfur substrates, but this was not
tested.
While all the sulfur compounds in petroleum have not yet been identified, all
the identified sulfur compounds are reduced: thiols (mercaptans), disulfides, sul-
fides, and thiophenes. The ultimate source of this sulfur is not definitely estab-
lished, but many authorities have argued that it must usually be oceanic sulfate,
reduced to sulfide by such bacteria as Desulfovibrio desulfuricans; Davis has pre-
pared an excellent summary of the evidence. These bacteria preferentially metabo-
lized the lighter mass isotope, viz, S-32. As we have already seen, this preference
is usually around 15%; the 34/32 ratio should therefore be about 15% lower in petro-
leum than in the sulfates of ancient oceans. The group at McMaster University has
shown that this is usually the case; their arguments are compelling:
Oils from widely distributed pools in the same reservoir rocks, e.g. ,
Devonian, have very nearly the same S content but differ greatly
in total sulphur content.... This constancy of the sulphur isotope
ratio for oils formed about the same time, and found in the same
formations accompanied by wide variations in the actual sulphur con-
tent, is very strong evidence for the suggestion... that as the oils
mature and lose sulphur there is little fractionation of the sulphur
isotopes. In other words, the sulphur isotope ratio for the oil does
not change materially with time but is indicative of the nature of the
source sulphur and environment in which petroleum was formed.
... The fact that oil pools in Devonian rock extending from the 49th
Parallel to the Arctic Circle have nearly the same sulphur isotope
content indicates that there must have been a large reservoir of sul-
phur over a large area with a constant isotope ratio in Devonian time
and that whatever isotope fractionation occurred in the oil formation
process this factor was reasonably constant over the same area. In
other words, the conditions of oil formation must have been similar
for the whole region. It seems that only a large sea would provide the
large reservoir of sulphur uniform in isotopic content and the con-
stant oil-forming conditions. Similar constant conditions and uniform
source of sulphur must have prevailed during the formation of the
heavy Lower Cretaceous oils and the light Upper Cretaceous oils
of Alberta.68
33
-------
There are exceptions and special conditions, and all the evidence is not in,
but, stripped of these scientific reservations, the heart of the argument is very
small and well worth looking at: Sulfur-isotope ratios are an index of geologic age.
Thode et al. as much as say so:
Six of the crude oils examined from West Texas fields are from
Ordovician and Silurian limestone as are four samples from
Ontario. ... It is interesting that with the wide range of sulphur
isotope delta-values found for petroleum samples in general that
the oils from the Ordovician and Silurian rocks of Texas and
Ontario, separated by several thousand miles, should be so
similar in their sulphur isotope content. The results suggest a
uniform S^^/S ratio in Ordovician and Silurian seas over a wide
area.68
And Thode himself, after clearing his conscience with a "there-is-some-indication-
that, " leaves his readers in no doubt
that the sulphur isotope ratios of the sea have changed with time
in a complex, but cyclic, fashion and that the sulphur in petro-
leum has changed in a similar manner but displaced approximately
15% from the contemporaneous sea level. 69
S-isotope ratios are just another general index of geological age; geochemists
have several such indices (e.g., the ratio of complex cyclic structures to simple
chain structures). It is difficult to see, however, how any such index can be of much
use in identifying oil spills. For example, our largest foreign source of oil
(Venezuela) is a major shipper of Tertiary crude; so is Louisiana, our second most
productive oil state. Our second largest foreign source of oil (Canada) sends large
quantities of pre-Mississippian crude to our East Coast refineries; so does Texas,
our largest oil-producing state. Miocene oil is common in Louisiana; it is also
common in Indonesia. Are S-isotope ratios sufficiently precise to distinguish
Miocene from Eocene?—if not, oil from Fresno, California, can't be distinguished
from the oil of Sanga Sanga, Borneo, or Plaquemines Parish, Louisiana.
There is another objection, however, that may be even more important. The
sulfur in oil is reduced, but when oil is spilled at sea, the sulfur is put in an oxidizing
environment. Because of kinetic isotope fractionatio'n, the S-34 will be more readily
oxidized than the S-32, other things being equal. The logic of this process leads us
to conclude that the more the spill is oxidized, the lower its 34/32 ratio will be; a
week-old spill should have a lower isotope ratio than a day-old spill from the same
source. If the fractionation effect were strong enough, the delta-value might go from
highly positive to very negative. It does not seem likely that sulfur-isotope ratios
can be depended on to prove a violation of oil-pollution laws.
34
-------
3.4 Special Cases
While we do not know of any easily identified passive tag that is unaffected by
harsh or extended weathering, several kinds of passive tags can be profitably used
as presumptive tests on the identity of oil spills under special conditions. These
conditions include: infrequent pollution, few possible sources of pollution, no recent
pollution, and fresh (single) slicks. In these special cases, chemical analysis does
not have to bear the whole burden of proof; the circumstantial evidence (of a non-
chemical sort) must be convincing. Ableson75 is both more and less optimistic than
we can be:
With modern instrumentation, it should be possible to catch the
culprits. For instance, a combination of gas-liquid chromatography,
controlled pyrolysis, mass spectrometry, and computer calcula-
tions probably could provide an identification as valuable as a finger-
print. Patterns obtained from oil slicks could be compared with
samples obtained from tankers bound for, and unloading at, United
States ports.
If the spill has not been extensively weathered and if there are not many possible
sources of pollution, the combination of techniques that Ableson suggests is wasteful:
Where the circumstantial evidence is convincing, extensive chemical evidence is
not required. However, where there are many possible sources of pollution, where
pollution is frequent, and where the spill may have been weathered for days or weeks
(or mixed with other spills, some of which may have been weathered), it seems
highly improbable that any combination of analytical techniques could stand up in
court to a carefully reasoned chemical defense. (Some of the chemical objections
to metallic and isotopic passive tags were discussed in sections 2 and 3. 2-3. 3;
these are the characteristic objections that might and perhaps should be used by
defendants in oil-pollution cases.)
The problem of mixed and continuous pollution is apparently common in
Southern California. Ludwig and his associates 76» 7^ have reported that:
Oil and grease from submarine seeps and waste from offshore
drilling operations are not of the same worldwide importance but
do affect the use of nearshore waters and beaches for bathing
and boating in areas located near offshore oil fields. This is
especially true along the coast of Southern California where the
beaches may be polluted both by natural seepage and by tanker
and drilling wastes.
One would expect that similar conditions pertain along much of the Gulf Coast.
Moreover, a very recent newspaper account quotes Howard Sanders, a senior
Scientist at the Woods Hole Oceanographic Institution, as saying:
35
-------
Just in the past few years we're finding we can't sail anywhere
in the Atlantic—even a thousand miles from land—without finding
oil. 78
The methods we shall now discuss cannot be recommended for such areas as
Southern California, nor for areas that are frequently afflicted by oil pollution;
these methods can, however, be recommended for waters that are not near off-
shore wells, oil refineries, or other possible sources of frequent oil pollution.
In the early 1950's several papers on identification of oil spills were
published in the U. S. and England. The first of these (Schuldiner79) reported
a simple method of paper chromatography that produced evidence for prosecuting
and convicting oil polluters in Baltimore, Maryland :
A method was found useful in prosecuting harbor pollution
violators in Baltimore harbor and in the upper Chesapeake
Bay area. It resulted in many convictions and was responsi-
ble for periods during which pollutions ceased. This tech-
nique was also applied to the identification of the source of
petroleum crudes. Various crude, semi-refined, and refined
petroleum products give characteristic spot chromatograms
by this method.
The simplicity and speed of chromatographic fluorescence
analysis, when applied to the field of petroleum products,
are easily demonstrated. A spot chromatogram is adequate
for identifying the source of crude petroleums and their
derivatives. When a crude is fractionated or cracked, a
whole series of products is formed, and, as a rule, each
derivative will produce a different visual or fluorescence
radial pattern. These spot chromatograms identify the
complex products formed. Mixtures of petroleum products
produce composite chromatograms characteristic of the
mixture. [Emphasis supplied. ] This fluorescence method
and its visual pattern have been applied to the forensic
problem of harbor and beach pollution with fuel and lubri-
cating wastes. The source of crude petroleums can be
identified similarly.
The government agencies detailed inspectors who obtained
samples of oil slick found floating in Baltimore harbor, the
upper Chesapeake, and its tributaries. The samples were
obtained by tying a string around the neck of a small wide-
mouthed bottle, and dragging it across the surface of the oil
36
-------
slick. Suspected ships were then boarded. Samples were
obtained in 1-ounce wide-mouthed bottles of bunker fuel oils,
bilges, a loose oil on the decks or sides of the vessel above
the water line, on adjacent docks, fuel intake valves, etc.
Comparison samples should be treated alike. Two or more
samples for comparison are placed side by side on the same
sheet of blotting paper. It is not necessary to use identical
amounts when original materials are compared, but dilutions
should be made as accurately as possible. Samples must be
protected from evaporation. Changes In viscosity will alter
the relative size of the chromatogram structures. [Emphasis
supplied. ]
A large number of tanker shipments of manifested crudes,
imported through the Port of Baltimore over a period of
about 3 years, were compared by means of the spot chromato-
gram method. The chromatograms from known crudes
invariably gave identical radial fluorescence patterns.
[Emphasis supplied. ] These crudes came from many sources,
including Saudi Arabia, Kuwait, Iran, Mexico, Venezuela,
Romania, Dutch East Indies, and Dutch West Indies.
While Schuldiner's spot chromatograms can be effective, they are not
perfect. In the emphasized passages of the preceding quotation, we have under-
scored three troublesome problems. First, confluent multiple spills will give
chromatograms that are characteristic of the mixture, not of the components
of the pollution; it is very doubtful that the chemical mixture can be resolved
into its legal components, or that a given slick can be proved not to be a mixture
unless the circumstantial evidence is overwhelming. Second, any but the freshest
slicks may have become quite viscous; they may have experienced considerable
evaporation, and their asphaltenes have quite probably agglutinated into a discrete
particulate phase; the chromatograms of the slick and the source are likely to
differ unless the slick is fresh, and the size of the difference will depend largely
on the slick's age. Third, it is not an unmixed blessing that identical chromato-
grams are invariably produced by known crudes from a given area; if all (or even
many) Venezuelan crudes give identical chromatograms, and if two or more
tankers carrying Venezuelan crude have recently been in the area of a spill, it is
not clear how one ship rather than the other could be blamed.
80
In 1953, two years after Schuldiner's work appeared, Herd reported
that he had used paper-strip chromatography to identify fuel oil that had been
spilled in the harbor of Wellington, New Zealand. Herd neither cited nor showed
any familiarity with Schuldiner's paper. Nevertheless, Herd's claims are quite
similar to Schuldiner's:
37
-------
On numerous occasions samples of fuel oil collected from the
surface of navigable waters have been submitted to us for com-
parison with samples of fuel oil collected from the tanks of
ships suspected of having made the discharge in contravention
of the Oil in Navigable Waters Act. To obtain incontrovertible
evidence by the conventional methods of oil analysis that such
samples are identical is almost impossible. The sample from
a river has usually lost some of its more volatile constituents
and, in addition, it will often be emulsified to a certain extent
with water. The removal of this water by drying, filtration or
solvent extraction results in a product whose analytical char-
acteristics are significantly different from those of the original
oil.
Attempts to prepare evidence of identity that could be demon-
strated visually by the use of adsorbent columns and by the
normal technique of paper chromatography did not produce
results of the required degree of precision. The paper-strip
method here described, however, gives visual evidence of
complete identity or of non-identity and, if necessary, the
strips can be used as evidence in a court of law.
All the reservations that apply to Schuldiner's methods apply just as strongly
to Herd's. One further reservation merits special attention. Herd's methods
"identify" only the ether-soluble portion of the slick. Although the ether-insoluble
fraction of a slick is admittedly difficult to work with in the analytical laboratory,
it can provide valuable information (see, for example, section 4.3. 2); in any event,
it does not seem prudent to ignore it altogether because it can be a sizable fraction
of the slick. Ludwig & Carter76 have reported that 12 crude oils contained (on the
average) 7. 9% ether-insoluble material, that some crude oils contained as much
as 15. 5% of ether-insolubles, and that one oil film collected at a beach near Santa
Barbara contained 6. 5% ether-insolubles.
Chromatographic methods are relatively simple and cheap. Other methods,
however, may also be used if there is sufficient equipment at hand and if the cir-
cumstantial evidence is already convincing. In section 3. 2, for example, we dis-
cussed Johannesson's methodSl of identification, which was based on analysis of
vanadium and nickel. Obviously, Johannesson's methods could be very much
improved. Canevari, 82 for example, has suggested emission spectroscopy, and
Guinn & Bellanca®3 have suggested neutron-activation analysis for determining
trace metals in oil spills. While we think that none of these methods can con-
clusively and reliably identify oil spills by analyzing only the oil slick,
Ciuffolotti et al. 84 have demonstrated that neutron activation is more accurate
than either spectrography or spectrophotometry:
38
-------
A comparison between neutron activation data and those from
spectrographic and spectrophotometric analysis, limited to
vanadium and nickel, was made. The results ... indicate, in
general, that higher concentrations of vanadium and nickel
were found by the neutron activation method. These higher
values are believed to be closer to the real contents of these
trace metals, since the neutron activation procedure described
does not require drastic thermal treatment which may lead to
partial loss of the elements.
85
McCoy had argued against spectrographic methods as early as 1962, particularly
for nickel analysis:
It will be noted that there is evidence of volatility of nickel
even in these heavy residual materials. The lowest results
were obtained by the spectrographic method which involves
direct ashing...
McCoy's data show that spectrographic determinations of nickel in residual fuel oils
are about 30% lower than colorimetric determinations, and about 20% lower than
gravimetric determinations.
Other methods of passive tagging (such as isotope ratios) might also prove
suitable, but —again—there is every reason to believe that the chemical evidence
alone would not be conclusive. All methods of passive tagging entail this risk, but
the risk seems well worth taking in those special cases where the circumstantial
evidence is powerful.
39/40
-------
SECTION IV
ACTIVE TAGGING
4.1 General
Literature studies have revealed that it is extremely difficult, if not impossible,
to assign, unequivocally, liability for an oil spill solely on the basis of identification of
the intrinsic constituents (organic or inorganic) of the spilled oil. The preceding
presentation has emphasized these difficulties. How then can the violator be identified
from evidence provided by the oil slick ? For if the intrinsic components of the oil
slick do not lend themselves to reliable identification, possibility of introducing suitable
additives (active tags) must be considered. What does one require of an active tag?
Ideally, it should satisfy ten criteria.
1. It must be an unusual material that is never found in the environment and
is found only in the petroleum to which it has been added.
2. It must be compatible with (i.e., chemically unreactive and physically
stable) and soluble or dispersible in the oil.
3. It must be both insoluble and nondispersible in water.
4. It must be relatively involatile.
5. It must be stable to chemical, photochemical, and microbial degradation
in the oil-slick environment.
6. It must be easily detectable in extremely small quantities by readily
available analytical techniques.
7. It must allow for modifications so that, when it is added to a particular
oil shipment, it provides a unique code or "license plate" for that carrier or transporter.
8. It must remain with the oil during the course of its history following a spill.
9. It must in no way interfere with the end-use applications for that oil, nor
complicate further processing and refining.
10. It must be available in quantity at economical prices.
While few tags will be ideal, all candidate tags must satisfy all criteria to some degree.
41
-------
4.2 Soluble Tags
4.2.1 Halogenated Aroma tics
One particular class of compounds that satisfies most of these parameters, from
a theoretical standpoint, are the polynuclear, aromatic hydrocarbons; specifically, the
halogenated aromatics.
Over the past several decades, many studies have been performed which were
aimed at identifying and characterizing the major constituents of petroleum. 8*> Although
many types of aliphatic, aromatic, and heterocyclic compounds have been identified,
there is no evidence of the presence of halogenated, polynuclear aromatic compounds
in crude oils. Because of this uniqueness these substances appear extremely attractive
for use as a label or tag for crude oils. Let us now examine the properties of these
aromatics to see if they could potentially satisfy the remaining criteria listed above.
Stability
It is well known that the aromatic series of hydrocarbons is significantly more
stable to oxidation, compared to the aliphatics, because of the conjugated structure of
the aromatic ring.87 A conjugated compound, a compound which resonates between
two or more structures, is actually more stable than any one of the possible valence-
bond structures. This stability is due, in part, to the resonance and bond-dissociation
energies of the aromatic system. Table 2 compares the resonance energies of some
aromatic compounds.
TABLE 2
RESONANCE ENERGIES OF SELECTED AROMATIC COMPOUNDS88
Compound Kcal/mole
Benzene 37
Naphthalene 75
Anthracene 105
Phenanthr ene 110
The series is characterized by a marked increase in the resonance energy as the
number of rings increases, and this increase is paralleled by a marked and progressive
increase in the stability of the derivative. In addition, the resonance energy is some-
what greater for the nonlinear ring systems than for the linear ones; the former
resonating among more stable valence-bond structures than the latter. The heterocyclic
compounds on the other hand have substantially lower resonance energies and corre-
spondingly lower stabilities.
42
-------
Halogenation of the aromatic compounds has in common with nitration,
sulfonation, and acylation the great advantage of stabilizing the ring against further
attack. 89 The mechanism of the reaction, involving electrophilic substitution,
involves attack of the ring by a positive halogen atom followed by a loss of a proton,
as shown below.
+ X.
An important feature of the use of halogenated polynuclear aromatic hydrocarbons is
that there are many different ways to substitute F, Cl, Br, or I and their combinations,
on the aromatic nuclei, thus providing a large number of specific tagging compounds.
The structures below depict only some of the vast number of stable polynuclear
ring systems that can be considered for halogenation:
CO
Naphthalene
Anthracene
1,2-benzanthracene
Phenanthrene
Pyrene
Perylene
43
-------
3,4-benzopyrene
Coronene
Chrysene
Depending on the particular polynuclear hydrocarbon, a large number of mono-,
and polysubstitution isomers can be produced with F, Cl, I, and Br. For example,
naphthalene alone has two isomeric monosubstitution products (C^oI^X), ten isomeric
disubstitution products of the type CioH6x2» and fourteen of the mixed type (with any
two halogens) C^QHgXY. And a greater number of combinations are obtained with the
tri- and octa- substitution isomers.
44
-------
The table below lists the 40 fully identified, but not all the theoretically
possible, di- and polychloronaphthalenes, for example.
TABLE 3
CHLORINE ISOMERS OF NAPHTHALENE90
1:2 1:3:6 1:2:3:4
1:3 1:3:7 1:2:3:5
1:4 1:3:8 1:2:3:7
1:5 1:4:5 1:2:4:6
1:6 1:4:6 1:2:4:7
1:7 1:6:7 1:2:6:8
1:8 2:3:6 1:3:5:7
2:3 1:3:5:8
2:6 1:3:6:7
2:7 1:4:5:8
1:2:3 1:4:6:7
1:2:4 1:2:3:4:5
1:2:5 1:2:3:4:6
1:2:6 1:2:3:5:7
1:2:7 1:2:3:4:6:8
1:2:8 1:2:3:4:5:6:7:8
1:3:5
All of the above compounds have boiling points in excess of 250° C. The stability
of overcrowded halogen compounds (e.g., octachloronaphthalene) has also been demon-
strated. The total number of combinations obtainable when considering the fluorine
and bromine or iodine substituent on the same molecule as the chlorine, provides (in
the thousands) individual compounds which are easily distinguished from "background"
polynuclear hydrocarbons naturally present in the petroleum. In addition, because of
the halogen substituent, each isomer can be readily identified in parts per billion within
a mixture by electron-capture gas chromatography.
This provides, then, a large series of compounds which can be used in tagging
a large number of separate petroleum samples, each tag being unique to that particular
sample. Since the aromatics are soluble only in the petroleum hydrocarbons, they
would remain with the oil slick and would be recovered with a water grab sample. In
addition, halogenation of heavy condensed, aromatic compounds of the naphthalene type
(and higher: anthracene, phenanthrene, perylene, pyrene, coronene, etc.)92 would
ensure their involatility since the tag must remain with the heavy petroleum fraction
(Table 4). In addition, halogenation lowers the vapor pressure significantly.
45
-------
TABLE 4
VAPOR PRESSURES OF SOME SELECTED AROMATIC
POLYNUCLEAR HYDROCARBONS FOR
USE AS TAGS (25 to 50° C)93
Compound Vapor Pressure (mm/Hg)
Benzene 26
Hexachlorobenzene « 1
lodobenzene ~ 1
Fluorobenzene 100
Anthracene « 1
Naphthalene « 1
Phenanthrene « 1
Active tags must be identifiable within the residuum or asphaltenic fraction of
the spilled oil even after several weeks of exposure and weathering. Petroleum
alphaltenes have a polycyclic structure, comprising mostly aromatic rings94 jn a
pericondensed sheet configuration. 95 In general, this material would present a
considerable background problem. However, the analytical technique suggested
for the detection of the halogenated, polynuclear aromatics (discussed in the
following section) should minimize this problem.
Electron-Capture Gas Chromatography for Halogen Compounds
An attractive feature for using the halogenated, polynuclear hydrocarbons as
tagging compounds is that the halogen substituent can be readily detected and identified
both qualitatively and quantitatively by gas Chromatography using the electron-capture
detector.9^ -p^g electron-capture detector is a substance-specific device that is
sensitive to certain types of molecules containing electrophilic atoms or groups such
as halogens, carbonyls, nitro-groups and certain condensed-ring aromatics, metals,
etc.9' It has a very low sensitivity for hydrocarbons other than the aromatics. ^
Furthermore, and very important, the electron-capture detector is extremely sensitive
to the halogen substituent. It can reliably detect halogenated hydrocarbons present in
a sample in quantities as low as 10~4 to 10~6 mg/l"and, under optimized opera-
ting conditions, a projected sensitivity to 10" mg/1 has been reported.
Electron-capture gas Chromatography has been used successfully to determine
ppb concentrations of chlorinated insecticides in waste water samples, 101 and
chlorinated hydrocarbons in air samples. 102 since gas Chromatography has been
used extensively to separate and characterize the components of petroleum and oil
products, *^ it can be utilized with great reliability and precision to separate the
components of a spilled oil sample; coupled with the extreme sensitivity offered by the
electron-capture detector, gas Chromatography can ascertain the presence of an added
46
-------
tag by virtue of its chemical structure and substituents. This latter aspect can most
readily be performed by prior compilation of a catalog of standardized chromatographic
column retention times of the various tagging compounds. When an oil spill sample is
brought in for analysis, the retention times of the various chromatographic peaks of the
sample are compared with the catalog and the proper component peaks identified. The
correlation between chemical structure and column retention times (or elution values)
is well documented. 104 Halogenated aliphatics, 105 heterocyclic compounds, and
nonhalogenated, polynuclear aromatics as well as inorganic halogens should not
interfere, since these compounds either will not have the same retention times on the
gas chromatographic column as the tagging compound or will not be detected because
of the selectivity of the electron-capture detector. If several peaks should occur near
or overlap the one of interest, these particular fractions can be collected and
re-chromatographed under operating conditions that will give better separation.
Because of the great sensitivity of the electron-capture detector to halogenated
compounds, it should be evident that very little of the tagging material is necessary to
tag a large volume of oil. For example, the largest tankers now on the drawing board
will hold 500,000 tons of oil. How much tag must be added to identify such immense
cargoes ? The practical limit of detection by electron-capture gas chromatography is
now 10-12 gm/gm. Therefore, 0.016 ounce of halogenated tag is (theoretically)
sufficient to tag a half million tons of oil. But why push the limit of detectability ?
Why not allow a safety factor of three orders of magnitude? Why not, indeed, because
then only 1 pound of tag (about 1 pint) will safely do the job. And just how safe is
"safely"? One gram of oil slick (about 0.25 ounce) will contain one nanogram of tag,
and one whole nanogram of tag is extremely easy to detect by electron-capture gas
chromatography.
It is not without interest that electron-capture gas chromatography is several
orders of magnitude more sensitive than emission spectroscopy106 and approximately
two to three orders of magnitude more sensitive than neutron activation techniques.107
By combining gas chromatography and mass spectroscopylOS the halogenated, poly-
nuclear, aromatic tags can be absolutely identified. Just as important, because of the
sensitivity of the electron-capture detector, the tag can be used in extreme dilution;
the tag can therefore be added to the oil in amounts of orders of magnitude below the
level which will not interfere with the end-use applications of oil or oil products.
In the preceding discussion we have pointed out the potential applicability of
utilizing halogenated, polynuclear aromatic hydrocarbons as tracers for identifying
the perpetrator of an oil spill. We have considered their usefulness in terms of the
criteria that an ideal tag must possess to be practicable, and the suggested soluble
tags appear to meet these criteria. As was mentioned earlier, it would be extremely
difficult to identify a particular petroleum solely by virtue of its intrinsic constituents
because of the differences in the stability of these various components to weathering.
This consideration arises when any chemical compound is to be exposed to weathering
and especially applies to the halogenated aromatics considered here as potential tagging
compounds.
47
-------
Stability of the Tag to Oxidation
It is generally acknowledged that the fate of many of the classes of aromatic
compounds, naturally present in petroleum exposed to aging, is condensation or
polymerization of the aromatic rings together with naphthenic compounds into high
molecular weight compounds. This hybrid material is commonly referred to as the
asphaltenic fraction of the petroleum. Although there are other fractions of high
molecular weight in petroleum (such as the resins and the maltenic fraction), the
asphaltenes appear to have the highest aromatic carbon content.109 The mechanisms
responsible for this condensation may be analogous to those in which it was found that
heating for 30-40 hours at 300-350° C resulted in fairly rapid formation of condensed
aromatic systems from the homologues of benzene, methylnaphthalenes, and from
bicyclic aromatic hydrocarbon fractions isolated from various crudes (Romashkino,
Radchenkovo, Khaudag). HO During this thermal processing of the petroleum, the,
peripheral parts of the molecules are destroyed, and the alkyl substituents are split
off the aromatic nucleus; this is accompanied by condensation reactions of the cyclic
parts of the molecule with the formation of progressively condensing polycyclic
compounds. HI The idea that oxidation such as the above may have been responsible
for the conversion of petroleum to asphalts rests on the observation that, in certain
oils, the presence of an asphalt deposit seems to be related to the migration of an
asphaltic petroleum from depth to the earth's surface where it becomes exposed to
evaporation and atmospheric oxidation. H2 since it is known that aromatic hydro-
carbons are extremely difficult to oxidize at low temperatures and pressures, H3
atmospheric oxidation is most probably a result of photochemical action. That
this atmospheric oxidation and polycondensation does occur is evidenced by the
fact that the weathering of petroleum results in the accumulation of a tarry resi-
due rich in condensed aromatic nuclei.114
What then could be expected to happen to a halogenated, polynuclear aroma-
tic tagging compound which has been incorporated into a petroleum sample and
exposed to air, water, and light for extended periods of time? It was discussed
previously that, as a class of compounds, the aromatic structures are extremely
stable to oxidation at ambient temperatures and pressures. ^^ That is, a signifi-
cant degree of oxidation can take place only at temperatures considerably higher
than ambient and preferably in the presence of specific catalysts. Furthermore,
halogenation has the added effect of stabilizing the aromatic nucleus to attack.
We saw earlier that unsubstituted polynuclear aromatic hydrocarbons can add
oxygen, initiated by photochemically generated radicals, across the central ring
to give the transannular peroxide, H6 and that the presence of electron-donating
groups (such as methyl or phenyl) at the reactive centers facilitates this reaction
with oxygen. However, because of the presence of halogen substituents, the very
opposite of electron donors, the general reactivity (oxidation and condensation) of
the polynuclear aromatic tag will be substantially less than that of the condensed
aromatic compounds naturally present in the oil, often in large quantities.
48
-------
Stability of the Tag to Microbial Oxidation
In an earlier section of this report, the microbial utilization of various types
of petroleum hydrocarbons was discussed. These hydrocarbon types encompassed
aliphatic (straight and branched) and aromatic (with straight and branched side-chains)
compounds. Whenever the microorganism demonstrated ability to oxidize the particular
hydrocarbon substrate, that compound was not of the halogen-substituted type. In fact,
the presence of an electron-donating group such as a methyl group sometimes facilitated
microbial oxidation of either the alkyl side-chain or the ring itself. Therefore, for the
same reasons that were invoked for the nonreactivity of the halogenated, polynuclear
aromatics to photochemical oxidation, viz, the presence of highly electrophilic halogen
substituents, it is believed that the microorganisms indigenous to petroleum and sea
water will not attack and oxidize the tagging compound. In general, the aromatic
hydrocarbons are more inhibitory to cell growth than are the paraffinic or cyclo-
paraffinic hydrocarbons.117
The intermediates detected in aromatic hydrocarbon oxidation have been studied
in investigations on the degradation of aromatic compounds by microorganisms that are
not hydrocarbon oxidizers per se. H& It is therefore assumed that no significant degree
of microbial oxidation of the halogenated, polynuclear aromatic tagging compounds will
occur.
Advantage of the Soluble Chemical Tag
The foregoing discussion has attempted to formulate a working hypothesis for
the identification of the perpetrator of an oil spill. This hypothesis requires that a
tagging chemical which most ideally meets the criteria set forth previously be added
to the oil cargo. These tagging chemicals are the halogenated, polynuclear, aromatic
hydrocarbons. The single most advantageous feature of these compounds (indeed, of
all active tags) is that they potentially enable identification of the polluters in a situation
where multiple spills of oil occur at the same location. Under these circumstances
considerable mixing of the various oils will occur. Since a large number of halogenated
compounds are available or can be synthesized, each oil to be transported can be labeled
with its own unique, single compound. In the event of multiple spills, a water grab
sample will contain the tags of all the labeled oils. Since gas chromatography can
resolve the individual tags according to molecular structure,11^ identification of the
violator, even under these conditions, is quite possible.
A combination of materials (soluble or nonsoluble) to label a single oil, although
unique for that particular oil sample, creates complications in multiple spills, where
the various fractions and constituents of the oil including the tags will be well mixed.
49
-------
In actual practice, it may be necessary to perform a preliminary concentration
and fractionation of the recovered petroleum sample to optimize tag analysis and to
separate the aliphatic fractions from the aromatic fractions (also, to remove water).
These preliminary steps would facilitate chromatographic processing for detection
and identification of the tagging compound by considerably reducing the background
peaks which would appear on the chromatographic profile. Since interest lies only
in the aromatic fraction, this preprocessing might, in actuality, be highly desirable
(although not essential, because electron-capture detector is not sensitive to
unsubstituted aliphatic hydrocarbons). In addition, the column parameters and
chromatographic processing would be made simpler, since a narrower range of
compounds would have to be separated. In some cases, where a large amount of the
recovered petroleum sample is in the form of a water-in-oil emulsion, the emulsion
would have to be broken and the oil separated from the water prior to chromatographic
analysis; of the several techniques for resolving oil-water emulsions, addition of an
appropriate surfactant in combination with froth flotation is the least expensive and
simplest method for processing large quantities.
Cost Considerations
It is desirable to devise oil tagging systems that are relatively economical to
prepare and use; however, low cost is not always essential to the system's imple-
mentation and effectiveness, and we believe that, in this case, the resultant benefits
far outweigh and justify the initial costs.
In evaluating the cost of commercially prepared halogenated, polynuclear
hydrocarbons, we found that many derivatives are inexpensive while others are
relatively expensive. The table below illustrates the wide price range of some
selected halogenated derivatives.
50
-------
TABLE 5
UNIT PRICE OF SEVERAL SELECTED HALOGENATED
AROMATIC HYDROCARBONS
Compound Price/Lb
Monochlorobenzene $ 1.10
Bromobenzene 2.60
Hexachlorobenzene 2.50
1,2,4-Trichlorobenzene 1.30
1,4-Dichlorobenzene 1.10
1,4-Dibromobenzene 3.50
Monofluorobenzene 18.00
1,3,6,8-Tetrachloropyrene 30.00
lodobenzene 24.00
Bromofluorobenzene 116.00
2-Fluoronaphthalene 360.00
l-Bromo-2,3, 5, 6-Tetrafluorobenzene 700. 00
9-Bromoanthracene 1000.00
The prices of several of these derivatives, though high, are not disturbing if
it is recalled that extremely small amounts are needed to tag a particular oil cargo.
Approximately 0. 001 Ib of the chemical can be detected by electron-capture gas
chromatography in 100 million Ib of oil (100 million Ib = 50,000 tons). Using
9-bromoanthracene, for example, the material cost of tagging 50,000 tons of oil
would be $1. 00. If other halogen derivatives of anthracene are comparable in price,
then 1000 tankers carrying 50,000 tons each can be labelled for a material cost of
$1000. If one then considers that a single major spill can cause damages amounting
to many millions of dollars, a cost of $1000 to tag a thousand potential pollution sources
becomes insignificant by comparison. If we consider this in terms of the number of
transporters to be tagged, we find that, in 1966, there were approximately
3000-4000 tankers (2000 gross tons and greater) and'approximately 2000-3000
miscellaneous types of petroleum-carrying vessels (less than 2000 gross tons)
traveling the waterways of the continental United States. Let us assume, for ease of
calculation, that at present there are 10,000 vessels, having-an average capacity of
1,000,000 Ib of crude and semi-processed petroleum. Since we know that the
halogenated poly nuclear aromatics can undergo a dilution of 1012 and still be detectable,
51
-------
then, 1 x 10~6 Ib of tag can be added to 1,000, 000 Ib of oil. However, if a dilution
factor of 1000 is introduced because of potential spillage and losses during sample
recovery, then 0.001 Ib of material can be used to tag the average vessel. This
means that even if the cost of every tagging compound is $1000/lb (which is extremely
unlikely), an average materials cost of only $1. 00 per vessel would be incurred.
Conversely, if the cost of every tagging compound is initially $1. 00/lb (equally
unlikely), then it would cost only 0.1/ per vessel. We believe that, in practice,
the materials cost per vessel would be somewhere between these two estimates.
52
-------
4.2.2 Organometallics
Halogenated aromatic molecules are not the only attractive class of oil-soluble
tags. In 1962, a group at the National Bureau of Standards, in collaboration with the
American Petroleum Institute, developed a number of stable, oil-soluble organic
compounds (mostly organometallics) for use as analytical standards in the spectro-
graphic analysis of petroleum products. 12° The trace elements include Al, Ba, B,
Cd, Ca, Cr, Co, Cu, Fe, Pb, Li, Mg, Mn, Hg, Ni, P, K, Si, Ag, Na, Sr, Sn, V,
and Zn; as cyclohexanebutyrates and/or ethylhexanoates, these metals make spectro-
graphic standards that have the following properties: Each is reproducible, stable,
nonvolatile, oil-soluble, and compatible with all the other spectrographic standards
in this series. Many of these metals can be detected spectrographically in extreme
dilution (parts per million or less), so that one gram could be used to tag a ton or more
of petroleum. The 28 compounds that were selected as standards were chosen from an
initial list of over 150 candidates; many of the candidates that were rejected as
spectrographic standards may still be useful as active tags for oil spills: One of the
principal reasons for rejection was extreme chemical purity (e.g., the metal
naphthenates), and this requirement might be somewhat relaxed for oil-spill tags.
These 28 standard samples, prepared in solid form for easy weighing and handling,
may be purchased for $2.00 a gram from the National Bureau of Standards. Obviously,
in mass production (with some relaxation of the requirements for extreme chemical
purity), the price might be significantly reduced.
Even though all this work on spectrographic standards has made the use of
certain organometallics an attractive possibility, there are still several problems
and doubts that must be resolved. First among these is the problem of solubility.
Only a few of the Bureau of Standard's organometallic standards could be dissolved
in oil, without heat or other solvents. If the standard were dissolved in a solubilizing
agent (such as a free carboxylic acid, a diketone, or an ester) and then diluted with
the base oil, however, the resultant solutions had optimum stability and highest
concentrations of trace metal. The solutions were judged satisfactorily stable if a
solution of at least 500 ppm of the trace metal of interest remained "clear and devoid
of visible change during several weeks at room temperature. "120 An excellent
solubilizer and stabilizer was found to be an equimolar mixture of bis(2-ethylhexyl)-
amine and bis(2-ethylhexyl)dithiocarbamic acid in an equal volume of xylene. It is
not clear, however, what the highest stable concentration of metals might be:
A precise determination of the maximum concentration
attainable for each element of interest, either individually
in oil plus solubilizers, or as a mixture of the standard
compounds in oil plus solubilizers, has not been made.
However, solutions having concentrations much higher
than those given have been prepared and found to be stable.
[Emphasis in the original.] All solutions prepared from
the standard compounds (according to the directions given)
have, up to this time, shown no signs of deterioration on
keeping. 12°
53
-------
A related question is more important still: What happens to the analytical
standard once the solubilizers and stabilizers have evaporated, as is likely after
a slick has aged on the sea for several days? Will the standards gel or precipitate?
It is known, for example, that several organometallic salts are used in the pigments
of paints for ships.
A second major problem is background: Most crude oils contain traces of
inorganics (especially metals), and some crudes have more than 0. 5% ash; there
are several excellent reviews of the literature on this topic. Many refined
petroleum products also contain metals, either deliberately added (e.g., components
of additives to impede corrosion and oxidation or to improve lubricating properties),
or adventitiously present (e. g., volatile organometallics in the crude that were not
removed during refining). How can this background be distinguished from the active
tag? Analysis of petroleum shipments for the presence of background interferents
is obviously not the answer: — It is much too costly, for one thing, and for another
there is no guarantee that the metallic suites in the oil slicks will be identical to
those in the unspilled oil. Indeed, there is every reason to believe that the suites
will be drastically different (see section 3.2 of this report for a discussion of this
matter). A better solution is to separate the tag from the slick (by chromatography,
for example) before attempting further identification. Since the tag is a chemically
stable "known," and since the organic moiety of the tag (e.g., nothing like a cyclo-
hexanebutyrate is mentioned in any recent comprehensive review) should be rare
in petroleum, there is no reason to believe that background suites will invalidate
this method of tagging. However, separating and concentrating the tag from each
slick sample adds to the time and cost of identification; since there are several
thousand oil spills in U. S. waters each year (see section 1 for a discussion of the
evidence), the problem of time and cost is not inconsiderable.
54
-------
4.2.3 Summary and Conclusions for Soluble Tagging
The preceding sections have, in some detail, considered the use of soluble
chemicals as tags for oils. A large number of chemical species may be detected in
oils, and a large number of stable chemicals are available which are not naturally
present in petroleum in parts per billion or more; such materials represent the two
basic possibilities for oil tagging: passive and active, respectively. Chemical tags
are, of course, already used in many ways: for tracing the flow of fluids through
the body in medicine, in tagging materials for the study of transport phenomena in
engineering, and in radiochemically analyzing reaction kinetics in chemistry. The
following paragraphs present the general conclusions which have been reached
regarding chemical or soluble tagging of petroleum.
• Passive tagging ("fingerprinting") is not practicable in the cases where
it is most critically necessary.
• Active tagging is, in principle, quite generally applicable.
Passive chemical tagging should be rejected because of the following:
a. Oils, as the term is usually meant, contain a large number of chemical
species. To characterize an oil completely would require that all species be quantified
down to the limits of detection (say parts per billion or less); as general analytical
and control procedures such quantifications are totally unreasonable.
b. Certain individual constituents of oil have been suggested as possible tags
in their own right; without exception, it has been found that the attempted use of such
tags would be fraught with difficulties which at this time appear insurmountable.
Detection of specific trace materials (e.g., sulfur isotopes or metal porphyrins) to
aid in the apprehension of small carriers which have made small spills is not possible
since detection sensitivity is much too low.
c. The effect of weathering on spilled oil is known to be severe; on the other
hand, there is presently no way in which this effect can be accounted for. Since an
identification technique is sought which can be applied to smaller carriers and small
slicks of unknown age, weathering in and of itself tends to rule out the passive tagging
("fingerprinting").
d. Perhaps the biggest problem with the passive tagging techniques is that
oils from the same geological area and era tend to be very similar in chemical compo-
sition; the errors inherent in chemical analyses generally make any marginal differences
in such oils impossible to detect. Anyway, an additional labeling method would still
be necessary if similar oils were the responsibility of different carriers.
55
-------
e. In the case of passive tagging, identification of a particular oil would
necessarily involve errors which would be compounded for multiple spills; obviously,
mixing of similar oils would result in an insoluble dilemma. This inherent difficulty
would, in turn, be compounded by the effects of weathering, cargoes of mixed oils,
and the complexity of the oil transport system.
Active chemical tagging, however, should for certain cases be considered
carefully. The following reasons pertain:
a. Certain stable chemicals are avai'able which are not known to exist
naturally in petroleum (e.g., organometallics and halogenated polynuclear hydro-
carbons); these chemicals can be obtained in very well-characterized forms for from
cents to thousands of dollars per pound. Detection sensitivity sets the amount
necessary in any particular application.
b. Since the chemicals to be used as active tags are readily soluble in oils,
there is little chance for problems in mixing or adding trace materials to oil cargoes;
trace amounts of such chemicals added to fuel or crude oils would not have deleterious
effect(s) as far as either producers or consumers are concerned.
c. Characterization may be as specific as necessary provided that provisions
are made for tagging the cargo of each potential (and relevant) polluter; this necessi-
tates that a large number or family of detectable compounds be available (this is
certainly the case for the halogenated polynuclear hydrocarbons, for instance).
d. Given that point c. is fully complied with, the problem of multiple oil
spills could be resolved.
e. Production of organometallics and halogenated polynuclear hydrocarbons
is well known; in fact, many of these compounds are available as standard reagents
or can be readily synthesized on order.
f. The necessary analytic techniques are available; among these are spectro-
photometry, gas and paper chromatography, and neutron activation; the sensitivity of
these various methods must be kept carefully in mind (little problem is expected with
the halogenated polynuclear hydrocarbons (detectable in the parts per trillion range)
while existing techniques for the organometallics appear to be an order of magnitude
or more from what they should be).
Certain problems can arise with respect to these techniques, however; among
these are:
a. Little is known of the effect of weathering on these materials while they
are contained in an oil slick; it is not unreasonable to postulate that halogenated
hydrocarbons might show a preference for sea water over a relatively nonpolar oil
media.
56
-------
b. Smaller concentrations of trace materials might necessitate samples of
spilled oil which are too large to be obtained practically, while in the laboratory they
might require complex preconcentration procedures.
c. In the absence of preconcentration, detection (especially for the organo-
metallics) could be extremely difficult for reasonable bulk quantities of trace
materials. For instance, a detection sensitivity of one part per million would require
that 200 pounds of tagging material be added to a 100,000-ton tanker; this could be
quite a costly amount of material as compared to grams per tanker for halogenated
hydrocarbon and particulate tags.
d. There is considerable public opposition to the use of halogenated hydro-
carbons in situations where they may be transferred to the environment. However,
the toxicity to marine life of halogenated polycyclic hydrocarbons has not been
established. This contrasts sharply with the halogenated aliphatic hydrocarbons
(pesticides).
Point a. could be resolved by appropriate feasibility studies while b. and c.
obviously depend upon the selection of the right tagging material(s). Point d. can
be resolved only through education of the public as to the harmless nature of trace
quantities of halogenated polynuclear hydrocarbons (less than one part per billion).
57
-------
4.3 Particulate Tags
4.3.1 Introduction
As with soluble chemical tagging of petroleum, insoluble particles may be
used for both active and passive techniques. Passive particulate tagging, however,
is much less attractive (upon cursory examination) than passive chemical tagging.
Although the particles naturally present in crude petroleum can convey a wealth of
information to a careful investigator, their relatively small number and wide variety
preclude a statistically reliable identification of a given crude. Other factors also
invalidate identification: The entrainment of new particles during transport of the
oil, the addition of particulate matter (especially pollen and spores) during exposure
to the atmosphere; the sometimes fragile and ambivalent nature of the solids contained
in crudes; and the fact that much of the particulate material in crude oils is in the form
of asphaltenes, which are identified as suspended solids, colloidal particles, and/or
solute molecules, depending upon the investigator and his techniques. Petroleum
geologists, palynologists, and engineers have shown some interest in the particles that
naturally occur in crude oils; they will probably continue to do so in the future. How-
ever, their interest does not appear promising of a technology for oil tagging.
Adding well-characterized particles to crude oils (active tagging) is, on the
other hand, a promising technique. In the first place, present technology can easily
manufacture such particles; in the second, they can be added to oils in small quantities;
and, in the third, there is good evidence that, if the need arises, they can be isolated
from the oil matrix in quantities sufficient to allow a very specific identification. It
is important to note at this point that active particulate tagging has all the advantages
of oil-soluble active tags (e.g., specificity and small amounts of tagging material) as
well as several that make it more suitable. Foremost among these latter is the ease
with which such particles may be produced using the same basic technique (e.g.,
grinding and spheroidizing of solids produced from an appropriate liquid melt).
4.3.2 Palynological Studies
Although palynological techniques have been used in the oil industry for about
20 years,123 there is still only a small amount of information on the particles in
crude petroleum. The references accompanying this report contain, we believe,
all the pertinent studies which relate to this subject and which are generally available;
this was borne out by people in the oil industry. Unfortunately the oil industry has
less interest in the nature of the solids found in crudes than in the various organisms
and fossils in a given stratum or zone, that might lead prospectors to oil. In fact,
most of "palynology's first 20 years" in the industry has been to aid prospecting.
The rare exceptions have been those few petroleum geologists who have studied the
particles in crude oil in an effort to explain its genesis. For obvious reasons there
is little impetus for financing such studies.
58
-------
4.3.2.1 Petroleum Geology: It is interesting to note that what appears to be
the most definitive study on the nature of the particles in crude oils is also the first.
J.M. Sandersl24 jn nis 1937 paper presented a detailed analysis of his methods and
results for a large number of Mexican and Rumanian crudes. This paper stands as
the_ basic reference to this day. Although it is popular to refer to such studies as
palynological, Sanders and others have also found fragments of tissue, hairs,
microalgae, bits of feathers, microfossils, etc. Other than for slightly more specific
data, recent studies have added little to his general conclusions. Sanders was
interested in the crude-oil microfossils in their own right; he attempted to explain
petroleum genesis by considering the fossil evidence. Most studies, however, have
been aimed at petroleum exploration.
1 OC
In 1941 Waldschmidt reported on examinations of the organic residues from
Permian crude oils. His techniques were apparently very similar to those of Sanders,
but already the motivation for such studies was moving toward correlation with par-
ticles outside of oil deposits rather than of the nature of those occurring naturally in
them. Within the oil industry such a movement continued. By the 1960's studies126' 127« 128
of the organic microparticles in crude oils had evolved into studies of asphaltenes
(asphaltic particles). These materials are of considerable interest in the oil industry
but of little value as well-characterized microsolids; they, in fact, are highly ambig-
uous in their properties. For instance, the residue from petroleum distillation at
350°C (1 to 40 wt %) is the asphaltic fraction; the part of this residue which is insoluble
in normal-pentane is by definition the asphaltenes (1 to 15 wt %). Further, the portion
of the latter which is insoluble in carbon tetrachloride is called the carbenes (0.3 to
1.0 wt %); this will be of more interest later. Materials isolated in this manner from
a substance as complex as crude petroleum can be expected to be complex also. In fact,
the "particles" so refined are small (about 50 A diameter and 1000 to 10,000 molecular
weight), spherical, hard to centrifuge, hard to oxidize, and almost impossible to identify
in any exact sense of that word. These facts defeat any reasonable use of asphaltic
particles as oil tagging materials and, in fact, indicate that they are really macro-
molecules which behave as solutes, colloids, or suspensoids, depending on how and
for what they are being studied. Sergienko dismisses the portions of the asphaltic
fraction which are insoluble in carbon tetrachloride (the carbenes) and in carbon
disulfide (the carboids) as unimportant. This is not surprising. The asphalt industry
would benefit little from a more exact delineation of the molecular nature of asphalt.
Sergienko's lack of interest is manifest in his claim that there are no particles in
petroleum larger than 65 A . Finally, with regard to passive asphaltene tagging,
"... it is practically impossible to identify the individual hydrocarbons. "
A fair amount of palynological data relating to oil exploration and its various
ramifications (little of it representing even an attempt at quantification) has been
published during the last decade. Hedberg reviewed some of this work briefly
in 1964. Contrary to the opinion held by Sergienko, there is a wide variety of micro-
solids in petroleum which are greater than 1 micron in size. In fact, the evidence is
59
-------
indisputable for the ".. .widespread presence of identifiable plant and animal remains
in petroleum. . . "; problems do occur, however, when the origin, character, and
distribution of these remains are considered. In general, emphasis has been upon
the qualitative nature of the spores and pollen in crude oil; few sophisticated techniques
have been advanced since Sanders' paper. The only exception to this is a Russian
paper published in 1963, " this work suggests some of the more common palynological
techniques for analyzing crude oils and ground waters. The literature on palynology
is perhaps a better source for such techniques. It is good to keep in mind, however,
that the selection of appropriate solvents and/or chemical reagents allows the iso-
lation of almost any solid.
More interesting than the work on specific analytical techniques has been that
on the geographical distribution of microfossils in crude petroleum. 131> 132» 133» 134
It has been discovered that certain geological ages left distinctive microfossil remains
in their crude oils. Chepikov & Medvedeva135 demonstrated that the crude oil in the
European U.S.S.R. is distributed in three very distinct zones: the Volga-Urals
region, the pre-Caucasus region, and a transitional region around the Ukraine. Such
a general characterization, however, is of little use for tagging crude oils; why it is
too difficult to extend this technique to smaller geological zones for this purpose will
be discussed in more detail in the following section. Tomor was probably the
first investigator to observe this property of crude oils experimentally; his latest paper
indicates the possibility of its use for age determinations. Bliznichenko observed
"ancient" types of algae in crude petroleums of the West Siberian Lowlands; unfor-
tunately these algae did not correlate with the age of the rock deposits surrounding
the oil. Speculation followed. Again no quantitative measurements were possible
and the palynological character of the oil in a given deposit was not at all consistent
with other facts and/or was not consistently repeatable. In a 1967 paper (Chepikov
& Medvedeva)132 an attempt was made to relate oil and gas migration to the various
spores, pollen, tissue, microalgae and microfossils found in them. If entrainment
of those materials is assured upon contact with a migrating fluid, this technique
makes some sense; however, the indications are once again that the impossibility
of quantitative measurement and the migration of the solid materials themselves
rule out passive particle tagging. deJersey's134 work (1966) leads one to the same
conclusion; he was able to make some general statements about oil and gas migration
and history but no differentiation of similar oils from different deposits. Again there
was evidence that crude oils have and still are picking up spores, pollen, microalgae,
and microfossils; since such a process would most probably be accelerated in an oil
spill, even qualitative determinations would become much more difficult. An example
of this is the different character of oil and natural gas from the same petroleum
deposit. The work of Horowitz & Langozky in Israel utilized some techniques
new to the field, but the results were generally the same as those above. Upper
Triassic oils (small spores, 10-20 microns) could be differentiated from Miocene-
Pliocene oils (larger spores, 20-80 microns), and contamination of the Triassic oil
was indicated. As did Chepikov & Medvedeva, *32»135 they also noted that pollen
migrates quite readily through rock formations; this is especially true when water
60
-------
is present. They were the only research group to notice analytical difficulties due to
contamination by asphaltenes and an abundance of pollen and spores. All in all, the
palynological studies which have been made over the last 30 years indicate that these
techniques are not advanced enough to allow characterization of different lots of crude
oil in the general case.
4.3.2.2 Nature of Microsolids: It was necessary in the previous section to
mention some of the more general types of particles found naturally in crude oils.
This section serves to elaborate briefly on each of these in order to demonstrate
that, in the case of a specific particle, differentiation is usually possible. Again,
J. McConnell Sanders' study provides the most extensive, published cataloging
of the types of particles most often found in crude oils; if the reader is interested
in the detail possible on these, he should definitely obtain a copy of this paper.
As might be expected, most of the observed microparticles have been pollen
or spores; the papers of Sanders,12^ Hedberg, I29 and Chepikov & MedvedevalSS
indicate, however, that many other materials are also present. The fact is that
most investigators do not look for other materials. Besides the spores and pollen
ranging from 5 to 100 microns, diatoms are often observed. These small, unicellular
algae are present in approximately the same size range, they are readily identifiable
and, therefore, present no particular difficulty in microanalysis. Only two types of
particles are usually found which have a well defined spherical geometry: pyrites and
resins. Even these are often not perfectly spherical and can be identified quite
readily (usually by visual techniques). This fact is of importance since carefully
controlled particle tagging would most probably involve spheroids of about the same
size. Also observed have been foraminifera, acarina, colonies of many protists,
antheridia, fungal hyphae, trichomes, etc. The variety is almost endless and,
therefore, the possibility of complete quantitative and qualitative analysis is very,
very slight. Pieces of tissue abound in crude oil; these may be either plant or
animal in origin and are hard to identify in any specific case. Fossilized plant
and animal remains are usually found; these often occur as fragments. Regarding
the environment, it is not surprising that much partly decomposed vegetable matter
is found. Pieces of shells, feathers, hairs, scales, petrified wood, etc., are
generally present; they often interfere with analytical techniques. Many of the
preceding materials are very fragile and are affected by investigative techniques;
besides the variety of materials present, this is the major reason why obtaining
meaningful quantitative data is so difficult. One's attitude to these materials, in
this respect, should be one of elimination rather than consideration.
4.3.2.3 Techniques: Practically, the optical limit for microscopic analysis is
about 0. 5 microns.137 Fine resolution is obtained only for particles quite a bit
larger than this; generally it is preferred to work at or over 5 microns. These
larger particles may be studied with standard laboratory microscopes, and their
fine structure can be resolved down to about 0. 2 microns. Palynology has a
wealth of techniques for the microscopic examination of the solids in crude petroleum,
but little to offer over and above the methods used by Sanders in 1937.
61
-------
Brown's book138 contains a fairly extensive review of palynological methods.
However, the technique he presents for analyzing crude oil (p. 115, Sittler's method)
is quite harsh; that is, much particle destruction is assured. For instance, (1) HF
treatment, (2) boiling in KOH, and (3) acetolysis are only three of the suggested steps.
It is not surprising that analysis such as this often indicates the presence of only very
small amounts of particulate material (e.g., residue from the Rancho La Brea tar
pits and Uvalde asphalt). In contrast, Sanders' methods (and those workers who have
followed his lead) are much less destructive and do not appear to yield inferior results.
The most valuable advice contained in Brown's treatise is that particulate-specific
solvents and reagents should be found and used when much solid matter must be
processed; this is especially relevant for active tagging since small amounts of the
tagged material may be in the presence of (possibly) large amounts of other solids.
Another specific technique is preliminary centrifugation rather than filtration; with
sufficient time for treatment, this is much less liable to promote the loss of important
microsolids. Practically, centrifugation limits the amount of oil which may be con-
veniently processed; but then, so does microfiltration.
Sanders' use of organic wash liquids is the most suitable for the general
examination of the solid residue from crude petroleum. As previously mentioned,
centrifugation would probably be more successful than Sanders' filtration in pre-
concentrating microsolids. More exotic methods than these are indeed available
(see Miziuke 1965),139 but would add little to the process being considered here.
The use of the usual wash liquids (n-C5Hi2, CC14, CS2) would most probably be
sufficient; however, development of an active tagging scheme would involve finding
a specific maceration technique anyway.
A scheme for microscopic examination, based on the works mentioned previously,
is outlined below:
a. Approximately one liter of retrieved, spilled oil is cut with a light organic
liquid (e.g., n-pentane) to reduce its viscosity.
b. The resulting suspension is centrifuged (or filtered) to isolate the solids.
c. After decanting the oil phase, the solids are resuspended in a light
organic liquid and recentrifuged; this is done about 5 times.
d. The particles are then filtered and trapped on a Millipore filter.
e. Depending upon the nature of the tagging particles, the solids are then
washed with appropriate liquids (e.g., CC14, CS2, H2O, KOH, CH3CH2OH, HNOs)
to remove background particles and clean the isolated solids.
f. With a hypodermic syringe, a sample of the particles, suspended in a
highly volatile liquid (e. g., acetone) is transferred to a microscope slide.
62
-------
g. Under the microscope, the relevant tagging particles are isolated further,
and any necessary preliminary tests are made on them.
h. The now completely isolated particles may be transferred from the slide
for more extensive analysis (e.g., to a sedimentometer for density determination,
to a standardized mount for size measurement, or to a microspectrofluorometer for
spectrographic characterization).
63
-------
4.3.3 Production of Microparticles
The production of finely divided solids is a field for study in its own right; for
instance, the minerals and ceramics industries have long been interested in and have
utilized fine grinding to obtain micro-sized particles. Their problems are unique,
however, in that they are concerned with the processing of quite large quantities of
relatively heterogeneous materials. Conversely, the operation envisioned here
involves only relatively small amounts of what is hoped are fairly homogeneous solids;
this homogeneity extends beyond the chemical constituents of the particles to both their
size and shape. The literature which addresses itself to these sorts of criteria is not
very extensive; however, the indications are that with available technology the
problems are soluble (subject, of course, to the results of reasonable engineering
studies). Although particles of widely varying sizes and shapes but of reasonably
similar chemical composition can be obtained, only one shape is convenient for the
production of large numbers of "completely characterized" particles: the spheroid.
Given that homogeneous solid material is obtained by (1) solidification of a liquid melt,
(2) coprecipitation from solution, (3) efficient mixing and sintering of suitable solids,
or (4) deposition from a vapor phase, it has been found that almost any such material
may then be spheroidized.
In addition to characterization and technological feasibility, microspheroids
are advantageous for several other reasons: (1) Their small surface area per unit
volume and lack of areas of large surface curvature make them less susceptible to
chemical reactions; (2) these same qualities make it more unlikely that they will clog
or impact upon passage through small apertures, or adhere to interfaces; and (3) micro-
spheroids are much more easily and reliably sized (e.g., by sieving, sedimentation,
and microscopic methods) than any others. The flowability of microspheroids makes
them ideal for unit operations where mixing is necessarily of importance; flowability
is obviously important where trace solids are to be added to a very large amount of
liquid. The above points also apply to simplifying the preconcentration and isolation
required in any analysis of suspensoids. Finally, small particles (10-50 microns) are
not rare in nature; but very smooth microspheres of this size are much less common.
4.3. 3.1 Techniques: There is a fair amount of literature on the production of
microspheres. Using spheres in experimental studies is often desirable since much
of the theoretical work on particulates has involved assuming such a shape; for
instance, think of the usual approach in fluid mechanics for studying the velocity dis-
tribution around a solid in a moving fluid (i.e., Stokes' analysis), or the basic approach
to the reaction of solids in physical chemistry. Much of the material presented by
Allen1 ° is based on a spherical geometry, although it is applied to similar, non-
spherical particles by introducing proportionality constants. Due to their having the
minimum surface-energy configuration, spheres (especially small ones) may be formed
quite readily; in nature, however, interactions with other particles and the presence
of directional forces often prevent their formation. The most common microspheroidi-
zational process is liquefaction of a particle followed by solidification in a dilute,
slowly moving, gaseous "suspension. " Other techniques have been used: Lucite
64
-------
spheres are produced commercially by tabling; deposition from a liquid solution has
been tried for certain resins; and by mechanical action, the polishing compounds used
in ball mills may be used to produce ceramic or glass spheres.
In 1934 Sklarew141 proposed what was probably the first system for making
glass microspheres, which are now commercially available in a few types of glass.
His apparatus was cumbersome and made a significant number of nonspherical
particles. Since it is difficult to separate micron-sized particles according to shape,
this method was generally rejected. Sollner142 modified Sklarew's approach and
came up with a simple system which produced essentially perfect spheres among all
of the thousands which he observed microscopically. These spheres were 15 to 50
microns in diameter; the smaller size is set, practically, by the limit of the micro-
scope he used and by his preliminary sizing procedure for the rough particles.
Sollner's paper contains a sketch of his apparatus; it is delightfully simple. Oxygen
is passed through a fluidized bed of the sized, rough particles which are to be
spheroidized; particles are entrained and pass into a simple glass torch. The smooth
glass torch is used to prevent unnecessary particle abrasion. If soft glass is used,
the flame may be oxygen-methane; for Pyrex or Kimax, an oxygen-hydrogen flame is
necessary. The particles are trapped in a pool of water toward which the flame is
directed. Besides being almost entirely spherical, the microsolids treated in this
way were found to be strain free, as was shown by analysis with Nicol prisms.
Freedom from strain is highly desirable since it lessens the chances for further
particle degradation and.reaction and reduces the possibility of spectrographic
anomalies due to structural distortions.
Bloomquist & Clark (1940)143 modified Sollner1 s experiments by using a
standard "blast lamp" torch and a stovepipe-funnel-filterbag collection system. They
produced the rough fine particles by wet-grinding Pyrex in a ball mill for several
hours and then sized them with 12 to 15 decantations at 5-minute settling times. The
spheres which resulted from this procedure were remarkably good (see the photo-
micrograph in their paper). For five nominal sizes of 25 microns and less, they made
statistical analyses; about 98% of the particles were found to be spheres, and 90% of
these were ±2 microns of the nominal size. This statistical size distribution can be
significantly improved by decantation after spheroidization, if necessary. These
experiments were run with very inefficient, batch processes: About 300 grams of
closely sized microspheres (in five sizes) were produced in 10 days. Yet, as shown
and discussed in more detail later, this is enough to tag about 500,000 tons of crude
oil'. It is also what amounts to a custom synthesis.
Laan & Nicholls144 constructed a simple apparatus for making 10- to 100-
micron spheres of low-melting metals. It amounts to nothing more than a small,
modified shot tower operated at 10 to 40 psig in an inert nitrogen atmosphere. Liquid
metal is drained from a chamber and passed through a 0.1-inch-diameter nozzle.
The metal solidifies in a 6-foot free-fall region; a rough separation of spherical and
nonspherical particles is achieved by infall on a slightly tilted collection plate.
65
-------
Potter145 describes an apparatus suitable for higher-melting materials; liquefaction
is achieved in the plasma of a direct current-arc column. Better grading is achieved
when previously sized particulates are used; both shot tower and wire sputtering suffer
in this respect. Helium is. used as an inert carrier, and again free fall permits solidi-
fication. The primary advantages of plasma spheroidizing are (1) large throughput
rates (as high as 20 pounds per hour) and (2) adaptability to processing a wide range of
materials: Over a very wide range of melting points, plasma-jet spheroidizing is
successful. A plasma-jet flame reaches temperatures of 50, 000°F; therefore, fine
powders with moderate melting points (e.g., Pyrex) are virtually 100% spheroidized.
An example of the efficiency of this process is found in the treatment of zirconia
(melting point, 2700°C); 80 x 120 mesh (about 150 microns) particles are 95 to 99%
melted and 85% spheroidized. The Thermal Dynamics Corporation uses such a
process. 14*> Microballoons of 25 to 300 microns are produced commercially14? from
a number of glasses and organics; the process involves passing a volatile solvent
solution of a filming material through a spray-drying tower. A suitable vapor-
generating material must also be in solution or spheroids form; these spheroids
represent a possible analytical interference when similar materials are used for
tagging crude oils—note that microballoons are widely used for evaporation control
in storage tanks. Careful microscopic examination should resolve this problem, however.
4.3.3.2 Materials: Of course, soft glass is the most common material from
which microspheres are made. Such glass may be "doped" with a wide range of metallic
salts, and it usually has some present as impurities. Microanalysis for such impurities
allows a tremendous amount of information to be obtained from a single microsphere.
Similar doping capabilities exist for the solids which are mentioned below. As pre-
viously mentioned, the hard glasses, e.g., Pyrex, may also be spheroidized. Plasma
techniques allow the processing of most metals and ceramics. Zirconia has been
mentioned above; alumina, many oxides, nitrides, and carbides have also been made
into microspheres. Laan and Nicholls' apparatus144 is suitable for spheroidizing
materials such as lead, zinc, and cadmium. With regard to the lower density materials
which may be more suitable for oil tagging, spray drying, as well as the other tech-
niques, may be used. This process has been used to make spheres of glass, cellulose,
starches, polystyrene, phenolic resins, and many other organic compounds. Treat-
ment of homogeneous solutions by spray drying would permit simultaneous doping and
formation of microspheres.
One concludes there is no problem with microsphere production. It can be done
on a laboratory scale using standard techniques; doping is relatively straightforward
and so is handling. For the application being considered here, sufficient production
capacity is assured. Once the process has been developed, technicians would be able
to do most of the work involved.
4.3.3.3 Addition to Petroleum: The metering of small amounts of liquids has
been developed by the chemical, pharmaceutical, and processing industries. For
particles of the size which we are considering, the same techniques will suffice. The
problems of contamination and analytical control are of primary importance. Both
66
-------
the metering system and the equipment used in particle production and analysis should
be easily cleaned or inexpensive enough to be disposable. The simpler production
equipment is both; only the modified shot tower and the plasma-jet apparatus would
involve tedious, but necessary, cleaning procedures; the spray-drying column is easily
cleaned since it is, in essence, a liquid processing unit. In no case need the particles
and/or the initial process materials make contact with the analytical equipment; there-
fore, cleaning is not a problem.
Addition of a tracer material to petroleum involves two similar and related
criteria: First, the trace material should be in a homogeneous phase, and, second,
this carrier phase should be injected uniformly into the petroleum so that the resulting
tagged dispersion (or solution) is also uniform. There is ample evidence that relatively
stable dispersions of particles in a petroleum-based liquid are easily produced. It is
important that this stability be maintained when the carrier fluid is injected into the
petroleum. Khomikovskii & Rehbinder148 found that maximum dispersion stability
correlates with maximum surfactant adsorption for pigments in oil. In crude oil,
enough surfactants are usually present to assure stability (at least for the dilute sus-
pensions we are considering); in refined petroleum products, it may be necessary to
select the particle material with great care so that the appropriate surface properties
pertain (i.e., relative hydrophobicity and oleophilicity). Highly polar materials
(e.g., ionic salts) actually stabilize suspensions when the particles are larger than
about 1 micron (Koelmans);^^ the only important exceptions are metal sols, which
are, however, stabilized by metallo-organic molecules (e.g., chelates). According
to Koelmans, "... a system is stable when a crossing of the potential barrier
between the particles by thermal motion alone is very improbable"; this potential
barrier is about 15 kT for a modestly charged double-layer and is much greater than
the usual thermal energies (about 1 kT). This energy difference also applies to
particles approaching an interface: Thermal energies cannot overcome the surface
potential barrier. Very concentrated suspensions (20 to 90 wt %) may be stabilized
by adding small quantities of dissimilar particles;150 this stabilization technique may
be of value in prescribing carrier suspensions.
Once a stable carrier suspension has been produced, it can be metered into
petroleum without further problems. Many metering pumps are available at nominal
cost; some of these (through mechanical action on a disposable diaphragm, bladder,
or flexible tube) do not even contact the metered fluid. Since the volume of the metered
suspension may be increased almost at will, pump capacities are not of special impor-
tance.
In conclusion, sufficiently stable particle suspensions can be produced and
then successfully metered into petroleum which is to be tagged; since this latter step
represents a very large dilution, the resulting suspension is expected to be very stable.
67
-------
4.3.4 Retrieval from Spills: Preconcentration
Although difficult, it is possible to cl llect very large samples of spilled oil;
however, the analytical techniques described in the previous sections involve, ideally,
relatively small amounts (i.e., 200 to 1000 ml). Such volumes of crude oil and
liquids, which have long been exposed to air, contain fairly high concentrations of
particulates (Krueger, pp. 622-633). ^ Reference to Horowitz & Langozkyl^ allows
selection of a typical tagging concentration: A few hundred particles in a few hundred
milliliters of oil are easily detectable. Therefore, we may specify 1000 particles
per liter. As an example, consider medium-density particles (1.0 gm/cc) in a
medium-density suspension (1.0 gm/cc) and calculate the total mass of particles
Mtotal (10 microns in diameter) in 100,000 metric tons of suspension (approximately
the load in a tanker):
Mtotal =
A one-liter sample of oil («50 gin) with only 5% the average number of particles
(i. e., 50 particles) would still be amenable to analysis. Since 10- to 50-micron
particles are uniformly and stably dispersible in oil, there is no problem in obtaining
a well-tagged sample. Probably the most critical point in sample retrieval is that
oil should be sampled from several places; these samples should then be mixed
uniformly and the analytical sample taken from this blend. Dispersed water should
also be processed in the preconcentration stage because some particles may be
found in the water phase of oil emulsions; either filtration or centrifugation will
successfully isolate these particles. It may be desirable to homogenize oil-water
dispersions (say with lanolin) to facilitate separation.
4.3.5 Particle Characterization
After tagged particles have been isolated from an oil spill, many identification
procedures are possible. These range from simple qualitative to more complex
quantitative schemes. Visual identification and counting are the most common
palynological methods; as mentioned previously, these methods are, however, subject
to a number of errors (e.g., operator bias). Visual recognition of particle shape
is definitely qualitative; therefore, the shape should be standardized and easily
recognized (i. e., we consider spherical particles). Since particle shape is set in
this manner, preconcentration and/or isolation are well-defined unit operations.
When the spheroids have been isolated, analysis of the character of individual
particles may begin; physical and/or chemical techniques may be used.
For a spheroid, size measurement presents little or no problem; standard
microscopic gratings which allow accurate and convenient comparative measurements
are available (McCrone et al.151) Particle diameters down to several microns may
68
-------
be determined in this manner. It is more difficult, but possible using standard
techniques, to measure the density of microspheres (Blanchard). 152 jf we consider
particle sizes of from 10 to 50 microns with an error of ±2. 0 microns for a given
size, there are at least 11 distinct and well-defined sizes. For density, with a
range 0. 9 to 1. 50 gm/cc and an error of ±2%, we have, again, at least 11 possible
characterizations. The product of the numbers of these two properties gives 121
distinct particle characterizations. In many cases this would be sufficient for
tagging fluids or solids, but the large number of shipments of petroleum makes it
necessary to expand the characterization. Ultramicroanalytical methods are widely
available (see Kirkl53 or Meinkel54). Microchemical techniques for a single
particle are also available from standard microscopy. Particles which contain
trace amounts of chemically detectable materials may be characterized either
qualitatively or quantitatively. For instance, a standard series of microtests could
easily determine whether or not 20 or so identifiable trace chemicals are present.
With a very simple binary code, this number of trace chemicals would represent
220 (1,048,576) possible particle characterizations. It is important to emphasize
that this tagging scheme is simple to execute and that the number of possible charac-
terizations presented above is a conservative one. Since these chemical techniques
are well known, no further elaboration is necessary.
A fascinating possibility is the use of spectrographic techniques. These may
be qualitative (Sventitskii)155 or semiquantitative (Harvey)156 or quantitative
(Hercules). 157 Furthermore, spectroscopy maybe used to analyze particulate
assemblies or single particles. Regarding the former, standard microanalytical
techniques apply directly; in fact, for particles less than 5 microns in diameter,
suspensions may be treated spectrophotometrically as homogeneous fluids (Lebedev).158
In practice particulate assemblies may also be solubilized, and analysis run
on the resulting solution. In the treatment of either a suspension or a solution, a very
large number of trace materials may be detected; for example, Sventitskii used a
steeloscope and was able to detect nine elements at the concentrations given in the
table below.
Detectable Concentration
Element (weight percent)
Cr 0.01 to 0.10
Mn 0.02 to 0.15
Mo 0.05 to 2.0
W 1.0 to 18.0
Va 0.2 to 2.5
Ni 0.03 to 0.3
Cu 0.05 to 0.6
Mg 0.02 to 0.12
Zr 0.1 to 1.0
69
-------
The detectable concentrations given above are for strictly visual observations. With
photographic assistance, Sventitskii155 was able to detect 17 elements at 0.1 or less
weight percent (p. 240); obviously, although his techniques were qualitative, a degree
of visual quantification was possible. Note that the detection of 10 elements at four
concentration levels yields the same number of characterizations as for the binary
system presented above.
Some of the preceding work was semiquantitative; Harvey's entire book156
considers experimental techniques of this sort. It is best, therefore, to consider
quantitative analyses directly. Since a very specific detection scheme is necessary,
spectroscopy is a logical choice. In fact, a unique spectrographic trace for either
a group of particles or a single particle gives rise to the possibility of a general
"fingerprint" or trace for any tagging material. This has two effects: (1) Although
the technique is quantitative, it is used qualitatively since only the specific particles
need be identified. (2) By adding appropriate chemicals to the particle matrix,
spectrographic traces may be "custom made. " Since these traces are then very
specific, the possible number is almost limitless. For example (although one should
note that what follows is too restrictive), given 20 distinguishable spectrographic
peaks (at different wavelengths) which may be detected at five different intensity
levels, we find that there are 52^ or about lO1^ possible characterizations. This
number is, to say the least, sufficient.
4.3.6 Costs
The materials' cost for particle tagging should be quite small. The matrix of
the particles will probably be a quite common plastic or glass: cost, a few cents to
a few dollars per pound. The trace materials imbedded in this matrix would cost
about the same but would be used in almost negligible quantities. More expense
would be involved for the capital equipment to do the necessary tasks. For instance,
laboratory apparatus to produce the microparticles would run in the neighborhood of
$5000. A microspectrofluorimeter, with a lens system for single-particle measure-
ments, might cost $30,000; note that spectrographic analyses of the desired type may
also be purchased commercially for about $50 per sample. High quality microscopic
equipment would be an absolute necessity; it is available for from $4000 to $20,000
(McCrone).151 A metering system for the tracer fluid would probably cost $20,000
to $100,000; a single metering pump is inexpensive (about $200), but a fairly large
number would be necessary in a practical distribution system. Additional capital
costs would include laboratory and office facilities, field-analytical equipment, etc.
As in many projects of this sort, the cost of labor would be very important.
Considering that a large amount of petroleum might have to be tagged, five laboratory
workers, 50 field workers, and managerial and office help would be necessary;
running costs for this labor would be 3/4 to 1 million dollars per year. This, of
course, is predicated on all the work being done by one specialized group.
70
-------
We should note that the above cost breakdown is for complete fabrication and
operation of a petroleum tagging center. These costs are not applicable to feasibility
studies and the like. Since much of the equipment is available, it could be modified
for a limited study; personnel who are familar with the necessary analytical methods
are also available. A development and demonstration study seems very attractive,
and could be conducted at nominal cost. It should be remembered that the results
of such a study would be applicable to work much divorced from the tagging of petroleum
(e.g., tagging solid wastes and studying particle transport in rivers).
4.3.7 Summary and Conclusions
It is hoped that the preceding sections have given a reasonable picture of the
possibilities of both active and passive particle tagging (insoluble tags). Besides its
obvious application to the tagging of crude oils, active particle-tagging has many
other possibilities: tagging and the concomitant study of solid air-pollutants, labeling
Pharmaceuticals, trace analyses in systems with multi-effect mixing phenomena, and
the growing study of microparticulate systems in general. Since the possibility of a
large number of easily produced and/or procured microsolids has already been illus-
trated, applications using them are now enhanced considerably.
For the particular problem being considered in this report, the conclusions
which follow have been reached. First, generally:
• Passive tagging with microsolids is not at all a realistic approach.
• Active particle tagging is.
Passive particle tagging must be rejected for the following reasons:
a. Other than for very general characteristics, it is difficult to attach
specific significance to the particles found in crude petroleum; that is, the element
of control is severely lacking.
b. Many of the solids which occur naturally in crudes are extremely fragile;
the technique of analysis, therefore, bears much too heavily on their detection.
c. Most of the particles found in isolated crude petroleum occur in the
ambient; therefore, in the case of a spill (i. e., exposure) they will be picked up.
This will render almost meaningless any identification based on the distribution of
particulates in the original crude.
d. Identification of a particular oil involves a statistical analysis with an
error factor(s) very difficult both to identify and quantify (especially with spillage,
transport, exposure, varied handling techniques, etc.); in a single case this problem
is a tough one—for multiple spills it would be intractable.
71
-------
On the other hand, active particle-tagging should be carefully considered for
the following reasons:
a. Well-characterized microparticles may be produced from a very wide
variety of materials (e.g., ceramics, glasses, plastics, metals); these materials
are available from a fraction of a cent to $1000 per pound (for a rare metal).
b. Micron sized particles can be injected into a tanker (or the oil as it is
either transported or loaded) in gram quantities or less, depending on the detection
technique; they will not affect the properties of a crude oil as far as either producers
or consumers are concerned (for particles of the size being considered here, injection
into many petroleum products would not be harmful).
c. Depending on how detailed the microanalysis is, the characterization may
be as elaborate as necessary (e.g., in principle, microspectrofluorimetry would
allow a "fingerprint" of a single, 20-micron spheroidl).
d. If the microanalysis is quite detailed, multiple oil spills present no
particular problem.
e. The production of microspheres is already developed technically.
f. The necessary analysis techniques appear to be available; these include
preconcentration, microscopic observation and minipulation, and quantitative measure-
ment.
Certain problems do arise with respect to this technique, however:
a. Spheroidal particles are found naturally, usually as resins and pyrites.
Glass microspheres may also be present when the crude oil has been stored with
microballoon evaporation-control.
b. Generally, the smaller the amount of the desired trace material, the
larger the spillage sample which must be analyzed.
It is expected that the former may be resolved by careful visual observations
during preconcentration. Careful selection of tagging criteria should eliminate the
latter.
The exact methods for analysis depend on many factors; hardly the least of
these is the development of such a particle tagging program itself. A thumbnail
sketch of what, at this time, appears to be a reasonable procedure follows:
a. Microspheres of about 10 microns diameter would be produced by spheroi-
dizing an appropriate, finely ground and sized solid.
72
-------
b. These particles would be characterized in preliminary tests to the
necessary degree.
c. They would then be metered into the specified petroleum (preferably as
it passes through a pipeline or as it is pumped into a tanker). In order to have, on
the average, 10^ particles per liter of oil, only about 5xlO~4 gram of particulates
needs to be added to each ton of oil; this amounts to 50 grams for a 100,000-ton
tanker.
d. Upon spillage, a representative sample would then be taken from the
slick.
e. The particles in about a liter of spilled crude would then be preconcentrated
as follows: (1) centrifuging and washing with a light hydrocarbon, (2) washing with
more digestive liquids (perhaps acids and bases), (3) a terminal filtration and wash
on a Millipore filter, and (4) transfer of the solids to appropriate microscopic mounts.
f. The same analytical steps would be run as in step b. to identify the particles.
These tests could be visual microscopic (e.g., size and/or shape, color, fluorescence,
phosphorescence), or microanalytic (e.g., density, hardness, trace analysis), or
microspectrometric (e.g., light absorption, fluorimetry, reflectance). Microspectro-
fluorimetry could provide a characteristic spectrograph of just one or a number of
particles; such a trace would, indeed, be a "fingerprint. "
g. The data obtained from the particles in the spilled oil would be compared
with that of the preliminary tests (which should be properly cataloged), thereby
identifying the source of the spilled oil.
h. In the case of a multiple spill, the individual offenders could be identified
quite readily. Determining the portion of a spill which each is responsible for would,
however, involve characterizing a larger, representative sample of microparticles;
based on the number density of each type particle, proportional liabilities could then
be estimated.
It is unfortunate that a technique such as that above has not yet been developed;
the tool that such a scheme represents may prove to be of monumental importance.
The necessary technology is available; all that remains is to put the system together
and develop it properly. The fact that such has not been done in the field of petroleum
technology reflects only the lack of knowledge of and interest in particulate systems;
this can also be observed in many other fields of study. This situation is, however,
changing quite rapidly. Perhaps, due to the study proposed here, interest will be
revived in the nature of particulates in crude oil. Anyhow, there is no doubt that
petroleum can be simply, economically, and effectively tagged with microspheres.
Such a study will open up a new area in pollution technology and technology in general.
73
-------
4.4 Radiochemical Tagging
There are two general alternatives in radiochemical tagging: (1) addition of
radioactive isotopes to the potential pollutant, and (2) addition of material that lends
itself to neutron-activation analysis. While both alternatives have been used in
water-pollution studies, the use of radioactive isotopes has been limited (for
excellent reasons) to a very special type of study. Neutron-activation analysis, on
the other hand, has been widely used.
Radioactive isotopes may be used advantageously when the product to be
tagged is not stored. Harremones,168 for example, has routinely used one curie
of bromine-82 (as ammonium bromide) in tracer experiments to select sites for
sewage-disposal plants; these methods are apparently standard in Denmark, and
are used in experiments
before selection of the site for future sewage disposal from
industries and municipalities. ... The movement and dispersion
of the tracer is recorded by a ... boat sailing in and out of
the cloud of radioactivity.... Tracer injections at different
sites permit a selection of the most economical site and of
the required purification of the sewage. *"8
Guizerix169 has reviewed the use of radioactive tracers (principally bromine-82
and iodine-131) in water-pollution research. In all the uses he reports, the radio-
isotopes are released instantaneously or regularly (at short intervals) and
measured regularly. Very "hot" radioisotopes are never added to a product that
will be stored, for the simple reason that the isotopes rapidly decay. The half-life
of bromine-82, for example, is approximately 40 hours, and the half-life of iodine-131
is about 8 days. Indeed, the rapidly-decaying isotopes that have high disintegration
energies are used as tracers precisely because they are detectable in very small
quantities. However, even when such "soft" and long-lived isotopes as tritium are
used as tracers, the experimenter must still comply with strict governmental regu-
lations. Consequently, even tritium (see, for example, Guinn17" and Anguenot171)
is almost always used in carefully controlled laboratory studies, and every effort is
made to ensure that the isotope does not escape into the air or the water: For, if
a rapidly decaying isotope escapes, its danger will be intense but short-lived; but
if a slowly decaying isotope escapes, the danger will be long-lived.
Ideally a radioactive tag should have a half-life of 90 days, be readily
available, be low in cost, and easily detected. To be effective as a tagging system
there should be a minimum of 30 radioactive tags all of whose half-life is close
together.
74
-------
Tags with short half-lives will not remain detectable during long voyages.
Furthermore, freshly activated tags would have to be furnished to the tagging groups
quite often. Tags which do decay, however, are useful because old tags disappear
and the tagging system noise is decreased. For these reasons the ideal half-life
would be over 60 days.
Long half-life isotopes present a serious problem because of their slow
disintegration rates. For the same counting rates far higher concentrations of long
half-life isotopes would be added to the oil. When the oil is subsequently burned
as a fuel the radioactive material would get into the air and into the food chain. This
would be a very touchy political problem not to speak of the medical implications.
With an ideal half-life of 90 days and the requirement that all tagging isotopes
used have a half-life close together, it is instructive to note that there are a total of
11 isotopes with half-lives between 90 and 120 days and 18 with half-lives between 90
and 180 days. In fact, there are only 10 gamma emitting isotopes with half-lives
between 60-180 days, six between 180-365, and 33 with over 365 days (gamma is
the easiest to detect). When the energy spectrum of the gamma emitters is examined
for isotopes over 0.3 MEV there are only 4, 2, and 22.
Radioactive tagging is unfeasible because there are not enough tags of the
right half-life, and if one goes to long half-life tags, the problem of air pollution by
radioactive materials becomes both a political and medical issue.
To circumvent these difficulties and dangers, neutron activation is often used.
Neutron activation is (in a sense) the mirror image of radioisotope tracing: Instead
of converting an element into a radioactive isotope before using it as a tracer, the
element is made radioactive only after the inert tracer has been collected and pre-
pared for analysis. Chatters & Peterson, *72 who used neutron activation to study
polution from a wood-pulp mill, have summarized the advantages of neutron activation
over radioisotope tracing.
This method eliminates those problems usually associated with
the handling of isotopes in [paper] mill. It also obviates the
necessity of having to comply with local, state of federal codes-
of-regulation [sic] relating to the use of radioisotopes within a
mill or their disposal into public waterways.
Neutron-activation analysis is now a widely used analytical technique; it is often us
in the petroleum industry to provide rapid and sensitive elemental analyses of refir1
product consumption, and catalyst poisoning.17^ Neutron activation (using small
accelerators) is routinely used for determining such elements as O, V, Mn, As, Cl,
Br, I, Si, Al, and Na in concentrations of 0.001%; with more powerful equipment,
the level of detection can be made some 10,000 times more sensitive.
75
-------
However, neutron activation is only one analytical technique among many;
it is no analytical panacea. In some cases other techniques (such as chromatogra-
phy and spectroscopy) are more sensitive, or more accurate, or less subject to
interference. Each case must be decided on its own merits. Insisting on neutron
activation would be as pointless as insisting on any other analytical technique; all
such arguments beg the question, which is: What inexpensive, stable, active tag
is suitable for identifying oil spills ? Having found suitable active tags, it is a
relatively small matter to choose the most powerful and sensitive method for
identifying them.
76
-------
SECTION V
IMPLEMENTATION OF THE TAGGING SYSTEM
Up to this point the feasibility of using additives to tag oil has been discussed.
It has been concluded that passive tagging is ineffective, but that active tagging is both
practical and economical. Discussed below are the mechanics and coding of tagging.
To effectively tag, a series of steps must be taken, such as adding the tagging material
to the oil, canceling the tag, checking the quality of the tags in the material, as well as
keeping the codes as flexible as possible. The purpose of tagging is to use a series
of unique codes that are specific and unique for each legal jurisdiction.
5.1 Marking and Identifying Oil Shipments
The mechanics of adding tagging material include a series of steps. Only five
(jurisdictional taggings) are presented here: oil (1) entering a tank farm, (2) entering
a tanker, (3) being received at a refinery, (4) entering another tanker, and (5) arriving
at a user's tank farm. Many of the casual spills are believed to occur at a transfer
point. Thus, tagging oil in these jurisdictions will cover a majority of the spills.
The tagging material will be added by metered pumps ahead of the main pumps
as the oil is on-loaded or off-loaded from a tanker. The tagging fluid for the refinery
will be kept by the refinery and that for the tanker will be carried by the tanker. Each
ship will have its own code and will maintain cognizance over the coding material.
Each time oil is loaded on a tanker, the tanker's coding material will be added to the
cargo. Samples will be taken on the downstream side of the tagging point for the
"record" and for "quality control. " Periodically, grab samples from tank farms and
from tankers will be analyzed to determine the code present, and to ensure that the
material is tagged as required.
5.1.1 Oil Tagging at Transfer Points
Because it is believed that the majority of the casual oil spills occur at transfer
points or from tankers, the initially recommended coding system should start by tag-
ging all oil entering into shipper's tank farm. Thus, all oil in the tank farm's juris-
diction will have its own unique code and any oil spilled on the waters from this
jurisdiction, for whatever reason, can be attributed,to the tank farm. When a tanker
is to load, the harbormaster will obtain from the tanker captain the coding solution
and insert it into a special metering pump ahead of the main loading pump. As the
oil is pumped into the tanker, it will be tagged with the tanker's code and will be in
the tanker's jurisdiction. Any oil spilled downstream of the pumps will be considered
in the jurisdiction of the tanker. When the tanker departs, any oil on the water with
the tanker code will be in the tanker's jurisdiction. As the ships ply the high seas, any
housekeeping adjustments made, causing a spill, will have the tanker's code, or
license plate, in the oil. Should this oil cause damage or be recovered subsequently
from the ocean's surface, the tanker's jurisdiction over the oil and responsibility for
77
-------
damage can readily be established. Upon entering the harbor and off-loading into the
tank farm, which in this case may belong to a user or refinery, the tank farm's coding
material would be put in,by a meter pump upstream of the main off-loading pump on
the tanker. This new code added to the oil will effectively cancel the tanker's juris-
diction and establish the tank farm jurisdiction.
5.1.2 Determining Jurisdiction
If a spill should occur during on- or off-loading, both the tanker code and the
tank farm code will be in the oil. Both parties will have observers, and responsibility
for the incident should be readily affixed. The spill will be clearly tagged with the
time and date confirmed.
5.1.3 Mechanism of Adding Tags
The actual mechanism of adding the tag material is quite simple. Using exist-
ing (or slightly modified existing) metering pumps, small quantities of the tagging
element in a carrier fluid will be pumped into the oil stream ahead of the main pump.
The turbulence of the pump and the subsequent turbulence resulting from the pressure
drop will mix the tagging material throughout the stream. As one or more minute
streams of carefully metered tagging fluid are added to the stream, the shearing
action of the pump and the other turbulence will mix it thoroughly. Subsequent move-
ment of the fluid and normal diffusion will continue to spread the tags throughout the
oil.
5.1.4 Tag Control
Each jurisdiction will carry its own tagging fluid and will be responsible for
safeguarding it. Thus, the shore installations will be responsible for furnishing to
the harbormaster the necessary tagging fluid when material enters their jurisdiction
and, in like manner, the captain of the ship will give the harbormaster his code at the
time of on-loading material. This makes it very unlikely that any false tagging can
occur. Since the codes even in the early part of this program will be less than simple,
it is not anticipated that counterfeit tagging material can be devised. Furthermore,
the problem of canceling (or confusing codes) by adding material to slicks is unlikely
due to the difficulty of mixing the coded material into an extant slick.
Freshly tagged material, undergoing a jurisdictional change, will be sampled
routinely and the sample stored for future reference. In case of a slick where a
question arises, the sample could be analyzed for comparison. Furthermore, the
sample can be routinely analyzed to determine if the tagging material is being pro-
perly added to the oil. If abnormalities occur in the system, they can be detected in
the sample and in the oil.
78
-------
It is also anticipated that samples will be "grabbed" periodically from the
tankers' or the tank farms' jurisdictions to check the efficacy of the tagging. In
the beginning, these will be used more for quality control of the system and for
detecting its effectiveness, and later for checking the system's life. It will be more
or less a legal check to ensure that the oil is being tagged properly.
It is believed that the initiation of the system, as well as the operation of it,
will be done best by one jurisdiction to minimize code confusion and maximize
information. It is further anticipated that there will be unique codes for each juris-
diction, and the codes will be canceled by subsequent addition of tagging material
which cancels the former one. Furthermore, dilution in a tank farm is believed to
significantly change the concentration and profile of the tag material.
5.2 Codes
For tagging oil there are two classes of codes: license plate code and profile
code. In the license plate code, a simple unique tag is added for each jurisdiction,
while, in the profile code, a dominant profile of nonunique tags which mark the
jurisdiction is added. Both have their advantages and disadvantages and both are
practical.
5.2.1 License Plate System
In the license plate system, a unique chemical or unique particle tag would be
added. While the number of unique chemical tags is certainly adequate and readily
analyzable, particles are probably better in this system because they can be concocted
in almost an infinite variety of mixtures and made quite pure. In the license plate
system, each tanker would have its own special material to uniquely identify that tanker.
In any working code, there would also be other redundant information. Such
information would indicate that the oil had been carried in a tanker or that it at one
time had been in the jurisdiction of a specific company's tank farm. In all probability,
there would be other bookkeeping information in the coded material, such as dating
information, coded batch information, and perhaps experimental coded material,
for expansion of the system. However, the unique jurisdictional tag would be present
in the oil and, in case of a spill, the presence of this tag would "state" the responsible
jurisdiction. The examples following this discussion present the methods of establishing
jurisdiction.
5.2.2 Profile System
In the profile codes, a series of analyzable particles or chemical tags is
added. Probably a hundred of these tags would constitute the entire system and the
actual profile of the material would become quite important. This system is more
amenable to chemical tag and is more easily initiated by oil companies with their
extensive knowledge of trace organic analysis. However, for a large system, it is
79
-------
the inferior method because much information will be available as to who are the
likely culprits. Thus, a quick analysis of the slick for the likely candidates will not
be possible and analytical, costs will be too high. Furthermore, in complex repeated
tagging, the profile differences become less distinguishable. Profile codes require
not only the detection of the presence of a code but also the quantity present.
Particulate and chemical tagging can be interspersed without any interaction.
Added quantities of a particular tag will be on the order of one part in 109 to 1011.
Since tags will remain in the oil, being specifically designed to do so, questions of
toxicity and effect on final users' needs should not arise because of extremely low
concentration. In all probability, any material selected as the tag will be no more
toxic that the lighter ends of crude oil itself. The material chosen will have an
optimum combination of ease of detection, cost, and compatibility.
5.2.3 Coding Examples
Examples are included in this section to delineate the actual tagging that will
be done and the characteristics of the codes as they occur in various jurisdictions.
Two examples are given for the license plate tag, and a number of examples are given
for the profile tag. The oil in the examples is considered to originate at an oil field
and be shipped to a tank farm where it is originally coded. Subsequently, the oil
is loaded onto tankers and shipped to a refinery, reloaded onto tankers, and shipped
to a final user, which might be a bunkering station. The final user is considered to
have a tank farm. Therefore, the oil is tagged at the initial well tank farm, when it
goes into the tanker, when it goes into the refinery tank farm, when it is on-loaded
into the second tanker, and when it is finally off-loaded into the user's tank farm-
five laggings. In the example, the actual code is traced through each of these steps
with redundant information being added at each step to further assist in its analysis.
For a "license plate" system, a unique material is added at the jurisdiction.
As the oil moves from one jurisdiction to another (for example, as it is off-loaded
from a tanker into a tank farm), it is usually diluted or mixed with other oil in the
tank. This plays an important role because the last tagging code is in full strength.
For example, an oil from a well tank farm source, loaded onto a tanker, will have
at full strength both the well tank farm's code and the tanker's code. When it is off-
loaded from the tanker into the refinery tank farm, the tanks in which it will be stored
may contain oil with other earlier codes, as well as the tank farm code. This loading
of the tanker's oil into a tank containing oil will dilute the tanker's code but, since
all the oil in the tank farm contains the tank farm's code, there will be no dilution of
the tank farm's jurisdictional code. Thus, in looking at a slick, one finds a pre-
dominant tag and tries to determine if there are other predominant tags. From one
dominant tag an investigator can probably pick up a history of the oil and establish
its absolute history by looking for other trace tags that would substantiate its origin.
In all probability, a tanker spilling oil will have two dominant tags: one of the
refinery and one of the tanker. And, obviously, the tags would indicate that the
history of the oil was from the tank farm to the tanker. Jurisdiction would be assigned
to the tanker unless there were extraordinary anomalies in other parts of the code
pattern.
80
-------
Cancellation of a code cancels jurisdiction, but not the material in the oil.
For example, a tank farm oil will have its jurisdiction, or tag, canceled by the
addition of a tanker code. The tank farm's code remains in the oil but the sequence
of events, which can readily be established from the extensive records kept of oil
shipments, will quickly show the responsible party. That is, tags will show recent
jurisdictions but history will show last jurisdiction in case of ambiguity. In the pro-
file system, cancellation of code profile is done by altering the profile. In any one
jurisdiction the dominant code will be either that of the jurisdiction or a combination
of jurisdictions and previous tank farm. This is because when oil is loaded onto a
tanker there is seldom dilution, but during off-loading, dilution can occur or readily
be arranged.
5.2.3.1 License Plate Tagging: In the first two examples which follow, oil
is shipped from the well tank farms, labeled W, X, and Y, by tankers numbered 1 to
10, to the refineries labeled A, B, and C, and by the same tankers, 1 through 10, to
the users labeled I, J, and K.
The oil passes through five jurisdictions: wells, tankers, refineries, back to
tankers, and to users. Oil is tagged when it enters the well tank farms, when it
enters the tankers, when it enters the refineries' tank farms, when it again enters the
tankers, and finally when it is entered into the users' tank farms. This system can
be extended to tag past the users, but five serialized laggings give the reader enough
of an idea of the codes without being unduly complex.
Table 6 presents a code. The code contains 26 unique tags numbered 1 through
26. In reality the number of tags could be extended to hundreds or even thousands in
number. Each tag position, i.e., 1, 2, 3, 4, ... 26 is a unique chemical or particle.
In the code illustrated in table 6, tag number 1—in position 1—indicates that the oil
has been tagged at a well tank farm; tag number 2—in position 2—indicates that the
oil has at one time entered into the jurisdiction of a refinery tank farm. Tag number 3
indicates that the oil has been delivered to a user's tank farm, and tag number 4 indicates
that the oil, at one point, has been shipped by a tanker. This information in these four
positions gives a quick history of the oil. For example, if tags 1 and 4 are found, it
means that the oil was shipped from the well tank farm via a tanker and then spilled.
The unique jurisdiction tagging, or "license plate" system, occurs after posi-
tion 4. For example, tag 5 indicates that the oil was from well tank farm W; oil from
well tank farm X has tag number 6 in it; oil from well tank farm Y has tag number 7
in it. Oil from tank farms 5, 6, and 7 will also have tag number 1 in the oil to indicate
that it was from a well tank farm. Refinery oil from tank farms A, B, and C are
tagged with tag numbers 9, 10, and 11, respectively, as their unique tags. This
refinery oil is also tagged with tag number 2 to indicate that the oil has been delivered
to a refinery tank farm at one point. Users' tank farms I, J, and K are tagged with
material in position 13, 14, and 15, respectively. In addition, all oil in a user's
jurisdiction will have tag number 3 to indicate that it has been in a user's tank farm.
81
-------
TABLE 6
LICENSE PLATE TAGS
Tag Number Jurisdiction
1 Well TF*
2 Refinery TF
3 User TF
4 Tanker
5 Well W TF
6 Well X TF
7 Well Y TF
8 Spare
9 Refinery A TF
10 Refinery B TF
11 Refinery C TF
12 Spare
13 User I TF
14 User J TF
15 User K TF
16 Spare
17 Tanker 1
18 Tanker 2
19 Tanker 3
20 Tanker 4
21 Tanker 5
22 Tanker 6
23 Tanker 7
24 Tanker 8
25 Tanker 9
26 Tanker 10
"Tank Farm
82
-------
In this example, there are ten tankers, 1 through 10, that are tagged with unique tags,
numbers 17 through 26, respectively. Also, the oil to all tankers will have tag
number 4 to indicate that the oil has been carried in a tanker one or more times.
The upper part of figure 2 indicates the direction and code present in the oil
as it passes from the wells to the refineries and finally to the user. To the left of
center in figure 2 are the tanker codes, and, as explained in the previous paragraph,
4 means that the oil has been in a tanker, and 17 means that the oil is in tanker number 1.
In a like manner, tanker number 8 is a 4:24 code; in this example, 4 means that it has
been in a tanker, and 24 is the unique tag indicating that it has been in, or is in, the
jurisdiction of tanker number 8. Well tank farm W has a 1:5 code, the 1 indicating
that it is a well tank farm code and 5 being the unique code of that tank farm (W).
Refinery tank farm C has a 2:11 code, 2 meaning that it has been in the jurisdiction
of a refinery tank farm and 11 uniquely identifies with the refinery tank farm C. User
tank farm J has a 3:14 code; 3 means it has been delivered to a user's tank farm and
14 is the unique code assigned to the tank farm J.
As the oil progresses from tank farm W to tanker number 1, to refinery B,
back to a tanker—number 9—and finally to user I, the code, or "license plate, "
undergoes a number of changes. At the bottom of the page is a table indicating the
code and the tags in the oil. W means tank farm W, Wl means it moved from tank
farm W to tanker number 1, while W1B means the oil originated at tank farm W, was
loaded into tanker number 1, and then off-loaded into the jurisdiction of tank farm B.
Finally, code W1B9I means originating at tank farm W, shipped by tanker 1 to tank
farm B, reshipped by tanker 9, and off-loaded to user tank farm I. Following the
tags in the oil at tank farm W, it will have a 1:5 code, and after tanker 1 code, it will
have a 1:4:5:17 code. When it is tagged with the refinery tank farm tag it will have a
1:2:4:5:10:17 code. And finally, in the user's tank farm W1B9I, it will have a
1:2:3:4:5:10:13:17:25 code. This code development can be seen at the bottom left of
figure 2. In the example given, no dilution is assumed (the worst case). In the table
at the right is the second example of oil originating at tank farm X, being shipped
by tanker 4, being delivered to refinery A, being reloaded onto tanker 7, and being
delivered to tank farm K at the user's terminal. In this case, it is assumed also
that there is no dilution and that the tags remain unaltered in the residuum.
These codes are unique and can supply information about the history of the oil
and of its jurisdiction. To the left of the double line at the bottom of figure 2, is the
tagging license plate, or code, which contains the generic information: 1 meaning
that the oil has been tagged at a tank farm associated with wells, 2 that it has been
to a refinery, 3 that it has been delivered to a user, and 4 that the oil has been shipped
by tanker. To the right of the lines are the unique jurisdictions.
83
-------
WELLS
X-4-A
wW-l-B-9
•-X-4-A-7
W-1-B-9-I
USERS
X-4-A-7-K
TANKER
NO.
1
2
3
4
5
6
7
8
9
10
CODE
4 17
4 18
4 19
4 20
4 21
4 22
4 23
4 24
4 25
4 26
WELL
TANK
FARM
W
X
Y
CODE
1 5
1 6
1 7
REFINERY
TANK
FARM
A
B
C
CODE
TT
2 10
2 11
USER
TANK
FARM
I
J
K
CODE
3 13
3 14
3 15
u u
a
W
W 1
W 1 B
W 1 B 9
W 1 B 9
a
LU
LU O
a ce.
3 <
§
1
I
1
1
1
°
2
2
2
i
3
O
4
4
4
4
5
5
5
5
5
10
10
10
17
17
17
13 17
25
25
CODE IN OIL (AMPLITUDE)
JURISDICTION SEQUENCE DISREGARDED
X
X
X
X
X
4
4
4
4
A
A
A
7
7 K
1
1
1 2
1 2
1 2
4
4
4
3 4
6
6
6
6
6
9
9
9
20
20
20
15 20
Figure 2. License Plate Coding Showing Oil from Wells W, X, Y Shipped to
Refineries A, B, C and Reshipped to Users I, J, K by Tankers 1
through 10. (Well, Refinery, User, and Tank Codes are Shown. Actual
Code in Oil as it Changes Jurisdiction is Shown in the Two Examples
Above.)
84
-------
In most spills, one or two at most, dominant tags will be present. However,
to take a worst case, when tanker number 9 has jurisdiction of the oil and spills it,
the dominant codes would be 10 and 25. It may very well be that 5 and 17 are present.
One could see immediately, looking in the redundant information, that the oil has
never been delivered to a user's destination and that it was picked up at a refinery on
its way to some delivery point, by looking up the tanker code—25—which corresponds
to tanker number 9. The only suspect is tanker number 9. By checking back in the
records one can see readily that tanker 9 picked up oil from refinery B, thus canceling
B's tag. Tanker 9 is the last jurisdiction and is legally responsible. Should the tanker
captain argue that somehow his code was put accidentally into the oil, the question
will arise as to how. If further arguments should arise, more careful analysis will
reveal tags 5 and 17 present, which, even in this simple case, would uniquely identify
the tanker run that was made.
5.2.3.2 Profile Tagging: In the profile method of tagging, a series of tags is
added, which forms a profile in the oil. In this system, far fewer tags are required-
less than a hundred. With a hundred tags, countless batches could be tagged. However,
part of the code will be used to simplify analytical techniques.
In the following examples, a 12-bit tag is used which admits to 4096 unique
tagging combinations. Part of the tag is a license plate tag which identifies, as in the
previous example, whether the oil has been tagged at a well tank farm, a refinery tank
farm, or a user's tank farm, or if it has been in a tanker. These positions are shown
at the bottom of figure 3. In positions 5 through 12, an 8-bit tag, which allows 255
different combinations, is used to identify ten tankers, three well tank farms, three
refinery tank farms, and three user tank farms. The routing is the same; that is,
the oil goes from wells via tanker to refinery and thence forward via tanker to user.
As in the previous example, well tank farms are identified by W, X, and Y, refineries
by A, B, and C, and user tank farms by I, J, and K. Tankers are identified as 1
through 10. In figure 3, associated with each identifiable unit, is a binary code along-
side the identifiable unit. The first four bits (see bottom of figure 3) of the binary code
are really a use-unique code indicating the type of jurisdiction (e.g., well tank farm,
refinery tank farm, user tank farm, or tanker) used in the tagging. This is the same
as the first four bits in the license plate tagging system. The last eight places in the
example are unique to the vehicle. Figure 3 shows the routing of the oil. The designa-
tion W6A5J is the same as in the previous example—that is, from tank farm W at the
wells via tanker 6 to refinery A tank farm via tanker 5 to user J. Figures 4 through 8
(page 93) show the code profiles as the oil changes jurisdiction, and table 7 shows the
binary codes in numerical tabular form corresponding to the profiles in the figures.
Figure 4A shows the profile code as added at tank farm X. Figure 5A is the
profile at tank farm W as is figure 6A. Figure 7A shows also the code as added at
tank farm W, but subsequent steps in 7 merely show the profile of the tagging code
added for comparison to the historic codes containing tags from previous jurisdictions.
Returning now to figure 4, the oil from the X tank farm is loaded onboard without
dilution to tanker 9. Figure 4B shows the code profile with certain of the bits being
85
-------
TANKER
NO.
1
2
3
4
5
6
7
8
9
10
TANKER
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
a:
s ^
Od ^
iJ <
\ S
t- UJ
_l LL
* Ot
1 2
0
0
0
0
0
0
0
0
0
0
2
LL
z
t-
OL
LU
3
3
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
1
1
1
PROFILE CODE
0
0
0
1
1
1
1
0
0
0
0
1
1
0
0
1
1
0
0
1
PROFILE
a
LU
z
i-
4
~~
5
1
0
1
0
1
0
1
0
1
0
1
0
1
1
0
1
0
1
0
1
CODE
678
PROFIL
1
0
1
0
1
0
1
0
1
1
0
1
0
1
0
0
1
1
1
1
9 10 11
E REGION
1
1
0
0
1
1
0
1
1
0
12
WELL
CODE PROFILE
w
X
Y
1
1
I
0
0
0
0
0
0
0
0
0
REFINERY
A
B
C
USER
1
J
K
0
0
0
0
0
0
1
1
1
0
0
0
0
0
0
1
1
1
0
0
0
0
0
0
1
0
0
1
1
1
1
1
1
0 0 1
1 0 1
0 1 I
0 0
0 1
1 1
0 1
1 0
1 1
1
1
1
0
1
1
1
1
0
1
0
0
1
0
1
1
0
0
0
1
0
0
1
0
0
1
0
0
0
1
0
1
0
1
1
0
Figure 3. Profile Code with Same Wells, Refineries, Users, and Tankers, Using
a Binary Code (Note that Positions 1 through 4 Denote a Use While
Positions 5 through 12 in Profile Region Denote Tanker.)
86
-------
twice the normal height common to both the tank farm and the tanker code. When the
oil in tanker 9 is off-loaded to tank farm A, it is assumed to undergo a dilution of ten
to one and the resulting code profile is shown in figure 4C. Some background tag
noise will be added to oil being diluted—not shown in this example—but it is not impor-
tant because the dominant profile is that of the added code. Also, it is the profile,
not the dominant profile, that counts in this system. In figure 4D the oil is loaded
back onto a tanker, number 4, and the resulting profile can be seen. Trace amounts
of other codes still remain which cause minor variations in the profile. In figure 4E
the ten-to-one dilution occurs when the oil is unloaded into tank farm J of the user.
The very low values of the certain bits, such as those in position 2, 4, 8, and 10, are
the results of this dilution.
Figures 5 and 6, showing the origin of different oil shipped by overlapping
routes, are used to illustrate the difference in profile that results when there are
differences in commonality. For example, figures 4C and 5C are both in the same
jurisdiction and their profiles are quite similar except for minor perturbations.
However, figure C is in a different refinery and the profile is very different.
Figures 4D and 5D are in different tankers and have different profiles but they
have been both unloaded to J and, therefore, figures 4 and 5E are quite similar.
Figures 5C and 6C are in different refineries and, therefore, look quite different,
but are then loaded onto the same tanker and, due to the back tagging of the two
different refineries, the tag profiles look different. Both can be identified, however,
and, with a little information processing, can be shown to have been in tanker 5.
They unload at different terminals, so the two codes are different and can be shown
to be different.
Figure 7 is simply the profile of the code that is introduced at the point of
loading for figure 6. It can be shown that these profiles look fairly similar to the
profiles in the oil by comparing figure 7 to figure 6. As previously pointed out,
however, this can be misleading and it takes a little more sophisticated code reading
to uncover the differences.
The code in figure 6 differs from the codes in figures 4 and 5 because at every
step of tagging a ten-to-one dilution occurred. Thus, the code profile in figure 7 is
created by a situation which would more resemble the bunkering of merchant marine
ships. It is difficult to conceive how the tank farm code could be diluted ten to one
when loading on a tanker. On the other hand, it is possible to have a tanker code ten
times as concentrated as a tank farm code so, essentially, loading on a tanker is
equivalent to dilution. Figure 8 is assuming the ten-to-one dilution on off-loading
a tanker into a tank farm. The constraints in figure 8 are the same as those in
figures 4 and 5, but the code in figure 8 is identical to the code in figure 6, with
different dilution constraints. The relative profile differences may be seen. Table 7
shows the computed codes for each one of the steps. In the rows are the code profiles
and in the extreme left-hand corner is the actual coding history. To the right of the
code index, such as in the upper left-hand corner, the code for figure 4 is X9A, followed
by a symbol 10 indicating division by ten. This means that the previous codes are
87
-------
assumed to have undergone a ten-to-one dilution. In the code for figure 6 there are
four ten-to-one dilutions in a row, and, as explained earlier, this is why figure 8
compares so closely to figure 7 in code profiles.
With the profile system, when a spill occurs there will be a history of tankers
in the immediate area. By comparing the anticipated coding profiles with the various
tankers, a large percentage of the tankers can be eliminated by vacant bit position
analysis. For example, in figures 4D and 5D, the 11-place bit is vacant in 5D but
occupied in 4D. This means that if tankers 4 and 5 traversed the same path, loaded
with oil from refinery A, and a spill occurred, tanker 5 would be eliminated from a
possibility of having the spill. However, tanker 4, under the analysis of this one-bit
place, would be a potential candidate as author of the spill.
However, elimination of a tanker is not enough to convict another tanker of
guilt. Here the quality-control samples taken at the time of tagging would be brought
to bear. Upon elimination of all tankers whose code profiles are simply not compatible
with the spilled profiles, the choice between several would have to be made. By
obtaining the quality-control sample taken at the time of tagging and comparing this
profile to the tanker profile, a one-to-one correspondence should be obtainable. In
this case, it is a stable fingerprint of tagged material, which will then do for the
conviction. The profile code system requires that, not only should a tag be identified
as present, but also, its magnitude be known within some percentage which depends
on the dilution ratios.
5.2.4 Summary and Conclusions
To obtain a conviction via the tagging process, it is necessary to use tagged
oil. Two methods have been proposed and both will be satisfactory. It is believed
that the individual license plate technique for the larger system is more satisfactory,
but, for immediate tagging purposes, profile-chemical tagging is probably simpler.
There is ample material to tag all the necessary jurisdictions.
Mixtures of oil spills comprise a more difficult problem. In a spill containing
only two sources of tags in about equal proportions, both systems can clearly distinguish
between the two spillers and assess a roughly proportional guilt. The "license plate"
method is more explicit in this case. However, when there is a ten-to-one mixture
(ten parts of one party and one part of another), finding the minor spiller becomes more
difficult. There will be times when this can be done and there may very well be times
when there is too much ambiguity. At the present time, it is not known how many
multiple spills have occurred. In fact, use of a tagging system should provide the
first evidence for or against this phenomenon. The authors believe, however, that
multiple spills occur rarely and are of little consequence; should a tagging system
be put into effect, they would be so rare as to seldom be a problem in the assessment
of guilt.
88
-------
Because of the problem of tag interaction and masking from one jurisdiction
to the next as well as the ambiguities in spill mixtures, one central body should con-
trol the tagging code and the amounts added, at least in the initial stage of the develop-
ment of oil tagging. Envisioned is a bonded service group serving the oil industry by
devising codes for each jurisdiction, and doing quality control work and analysis of
spill mixtures. Outside laboratories could check analyses for the establishment of
court information in legal cases. However, for general quality control work and
system improvement one control group would be essential as the tagging system is
expanded.
89
-------
CO
o
TABLE 7
NUMERICAL VALUES CORRESPONDING TO FIGURES 4, 5, 6, AND 8
Code for figure 4
§
33
.2
a
X
X9
X9A ± 10
X9A4
X9A4J -=- 10
d
o
s
w
W6
W6A * 10
W6A5
1V6A5J + 10
1
1
0.1
0.1
0.01
1
1
0.1
0.1
0.01
0
0
1
1
0.1
0
0
1
1
0.1
0
0
0
0
1
0
0
0
0
1
0
1
0.1
1.1
0.11
0
1
1.1
1.1
1.11
Code for
0
1
1.1
0.11
1
1
1.1
1.11
101
102
1.1 1.0 1.2
2.1 1.0 1.2
1.21 1.1 0.12
1010
1121
1.1 1.1 0.2 0.1
2.1 1.1 1.2 0.1
1.21 0.11 1.12 1.01
figure 5
001
111
2.1 1.1 2.1
1.21 1.11 0.21
1100
2101
211 (\ n i
i. 1 U U.I
1.2 2.1 0 1.1
1.12 0.21 1.0 1.11
-------
TABLE 7
to
to
NUMERICAL VALUES CORRESPONDING TO FIGURES 4, 5, 6, AND 8 (Continued)
Code for figure 6
a
£
3
W
W9 4 10
W9C f- 10
W9C5 -r 10
W9C5K 4 10
d
o
33
a
W
W9
W9C 4 10
W9C5
W9C5K 4 10
1 0
0.1 0
0.01 1
0 0.1
0 0.01
1 0
1 0
0.1 1
0.1 1
0.01 0.1
0
0
0
0
1
0
0
0
0
1.0
0
1
0.1
1.01
0.1
1
1.1
1.11
0
0
1
1.111 1.1
1.01
Code for
0
1
0.1
1.1
0.11
1
2
1.2
1.2
1.12
1.11
figure 8
0
0
1
1
1.1
0
0
0
0
1
0
0
0
0
1.0
1
1.1
0.11
1.01
1.1
1
2
0.2
1.2
1.12
0.1 1.1
0.01 1.11 0.1 0.1
1.111 0.01 1.01
0.11 0
0.1
0.1 1.2 0.1 0.1
0.1 2.2 0.1 1.1
0.01 0.22 0.01 0.11
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
SECTION VI
CONCLUSIONS AND RECOMMENDATIONS
There are several thousand oil spills on U.S. waters each year; only a small
fraction of these are reported to the authorities. In view of this situation, it is
desirable to be able to identify (by analyzing oil slicks) those who are responsible
for oil spills.
There are two approaches to identification: passive tagging and active tagging.
Passive tagging is a convenient circumlocution for no tagging at all: Nothing is added
to the oil; the slick itself must provide all the evidence for its identification. Active
tagging involves adding some known and readily identifiable material to the oil. The
added material must satisfy several criteria: It must be soluble or dispersible in
oil and insoluble and nondispersible in water; it must be chemically and physically
stable in both spilled and unspilled oil; and it must not interfere with the commercial
uses of petroleum. One additional criterion: It must not be too expensive. Two
figures of speech are helpful in distinguishing active tagging from passive tagging:
Passive tagging amounts to a "fingerprint, " active tagging amounts to a chemical
or physical "license plate. "
Passive tagging is not at all attractive, for the fingerprint is very likely to
be changed or to be washed away as the slick is exposed to the abundant sources of
chemical and physical energy in the sea and sky. The easily identifiable passive tags
of petroleum (e.g., isotope ratios, trace metals) cannot be assumed to be stable in
the slick — indeed, there is little doubt that they are unstable — and the more stable
passive tags in the slick (e.g., heavy aromatic and alkylaromatic hydrocarbons) are
prohibitively difficult to separate and identify. In short, fingerprinting seems neither
technically feasible nor scientifically valid.
Active tagging, however, offers several attractive possibilities. It appears
that making a good license plate is much easier than finding a good fingerprint. We
have examined two types of active tags: oil-soluble tags (organometallics and halogenated
polycyclics) and particulate tags (coded microspheroids). Several very attractive
organometallics have been developed by the National Bureau of Standards in conjunction
with the American Petroleum Institute; they seem to have all the properties of good
active tags (compatibility with oil, insolubility in water, chemical and physical stability,
ease of identification), and they are sold by the Bureau for $2. 00 a gram. We strongly
recommend that pilot studies of oil tagging be conducted with these organometallics.
The halogenated polycyclics also have desirable properties, and it would be well to
subject a few candidates from this class to preliminary feasibility tests. Although the
feasibility of particulate tags has not yet been demonstrated, coded microspheres
appear to have exceptional promise: Many are already commercially available at very
low prices; an astronomic variety of coded microspheroids can be made easily and
inexpensively by custom synthesis; and all the particulate tags can be quickly separated
from samples of slick and readily identified by available analytic••:.! techniques.
95/96
-------
SECTION VII
REFERENCES
1. KIRBY JG & MOORE BM (1968). Crude petroleum and petroleum products;
pp. 837-947 in: U. S. DEPARTMENT OF THE INTERIOR, BUREAU OF
MINES, Minerals Yearbook 1967, Volume I-II — Metals, Minerals, and
Fuels. U.S. Government Printing Office, Washington, B.C.
2. LUDWIGSON JO (1969). Oil pollution at sea; pp. 1-20 in: DEGLER SE [ed],
Oil Pollution: Problems and Policies. Environmental Management Series,
Bureau of National Affairs, Washington, D. C.
3. U.S. CONGRESS, HOUSE OF REPRESENTATIVES (1967). Report on inter-
national control of oil pollution: Report of a Congressional delegation to the
Third Extraordinary Session of the Intergovernmental Maritime Consultative
Organization. 90-1, Union Calendar No. 250, House Report No. 628. U.S.
Government Printing Office, Washington, D. C.
4. U.S. CONGRESS, HOUSE OF REPRESENTATIVES, COMMITTEE ON MER-
CHANT MARINE AND FISHERIES (1969). Hearings before the Committee,
90-1, onH.R. 6495, H. R. 6609, H. R. 6794, andH.R. 7325, bills to amend
the Oil Pollution Act, 1924 Serial No. 91-4. Testimony submitted by
Comdr. Clifford F. DeWolf, Chief, Legislative and Regulations Division,
U.S. Coast Guard, pp. 96-97.
5. Ibid., p. 100.
6. U. S. SECRETARY OF THE INTERIOR AND U. S. SECRETARY OF TRANS-
PORTATION (1968). Oil Pollution: A Report to the President. U.S. Govern-
ment Printing Office, Washington, D. C.
7. LUDWIGSON JO (1969). Op. cit.
8. U.S. CONGRESS, SENATE, COMMITTEE ON PUBLIC WORKS (1969). Water
Pollution — 1969; Hearings before the Subcommittee on Air and Water Pollution,
91-1, on S. 7 and S. 544, bills to amend the Federal Water Pollution Control
Act..., Part 4. Pp. 1010-1011.
9. Ibid., pp. 1027-1028.
10. Ibid., p. 1442.
11. U. S. SECRETARY OF THE INTERIOR AND U. S. SECRETARY OF TRANS-
PORTATION (1968). Op. cit.
97
-------
12. Ibid.
13. U.S. CONGRESS, SENATE, COMMITTEE ON PUBLIC WORKS (1969),
Op. cit., p. 931.
14. Ibid., p. 961.
15. U. S. SECRETARY OF THE INTERIOR AND U. S. SECRETARY OF TRANS-
PORTATION (1968). Op. cit.
16. U.S. CONGRESS, HOUSE OF REPRESENTATIVES, COMMITTEE ON MER-
CHANT MARINE AND FISHERIES (1969). Op. cit., p. 122.
17. BESTOUGEFF MA (1967). Petroleum hydrocarbons, pp. 77-108 in NAGY B &
COLOMBO U [eds], Fundamental Aspects of Petroleum Geochemistry. Elsevier
Publishing Co., Amsterdam, N. Y.
18. BEYNON LR (1967). The Torrey Canyon incident. British Petroleum Company
Limited, September 1967. Cited on pp. 4-9 and 4-11 of PACIFIC NORTHWEST
LABORATORIES DIVISION OF BATTELLE MEMORIAL INSTITUTE (1967),
Oil Spillage Study Literature Search... .Distributed under accession number
AD 666 289 by the Clearinghouse for Federal Scientific and Technical Information,
U.S. Department of Commerce, Springfield, Virginia 22151.
19. BLOKKER PC (1964). Spreading and evaporation of petroleum products on
water, pp. 911-919 in Verlagboek van het Vierde International Havenkongres,
Antwerpen, 22-27 June 1964. Apparently published by the Koninklijke Vlaamse
Ingenieursvereiniging, Antwerp, Belgium.
20. COSTANTINIDES G & ARICH G (1967). Non-hydrocarbon compounds in petroleum,
pp. 109-175 in NAGY B & COLOMBO U [eds], Fundamental Aspects of Petroleum
Geochemistry. Elsevier Publishing Co., Amsterdam, N.Y.
21. DEAN RA (1969?). The chemistry of crude oils in relation to their spillage on
the sea, pp. 285-289 in U.S. CONGRESS, HOUSE OF REPRESENTATIVES,
COMMITTEE ON MERCHANT MARINE AND FISHERIES (1969), Hearings
before the Committee, 91-1, on H. R. 6495, H~ R. 6609, H.R. 6794, and
H. R. 7325, bills to amend the Oil Pollution Act, 1924 Serial No. 91-4.
U.S. Government Printing Office, Washington, D. C. This paper was submitted
to the Committee by Vice Admiral Paul E. Trimble, Assistant Commandant of
the U. S. Coast Guard, as "the best comprehensive report known to the Coast
Guard on the physical and chemical changes that occur to an oil spill... ".
22. SMITH HN (1968). Qualitative and quantitative aspects of crude oil composition.
U.S. Department of the Interior, Bureau of Mines, Bulletin 642. See especially
pp. 130-132 (Oil I, Oil K, and Oil N).
98
-------
23. BAKH NA [ed] (1955). Symposium on Radiation Chemistry. Academy of
Sciences of the USSR, Division of Chemical Science, Moscow, 1955; in English
translation (no translators names given). Consultants Bureau, Inc., N.Y.
24. EMANUEL' NM (1965). The problem of the control of the chain reactions
taking place in the liquid-phase oxidation of hydrocarbons. Pp 1-31 in Ref. 25.
25. EMANUEL' NM [ed] (1965). The Oxidation of Hydrocarbons in the Liquid
Phase, DOBSON KR & HAZZARD BJ [trans, from Russian], HOPTON JD
[ed of trans.]. Pergamon Press, Oxford, N. Y., etc.
26. REICH L & STIVALA SS (1969). Autoxidation of Hydrocarbons and Polyolefins:
Kinetics and Mechanisms. Marcel Dekker, Inc., N.Y.
27. TINYAKOVA Yel et al. (1965). Oxidation-reduction systems as oxidation
initiators in hydrocarbon media and their mechanisms of action, pp. 130-139
in Ref. 3.
28. CHERTKOV YaB & ZRELOV VN (1965). The oxidation of hydrocarbon fuels
under storage conditions, pp. 351-361 in Ref. 25.
and
AROZENA AA (1967). Contaminacion del agua del mar por el petroleo. El
paso del oxigeno atmosferico a traves de capas de petroleo de distintos
espresores. Ministerio de Marina, Subsecretaria de la Marina Marante,
Institute Espanol de Oceanografia, Boletin No. 131, August 1967.
El Ministerio, Madrid [N. B. In all earlier citations of this author's surname
in the Boletin, the name appears as AROCENA. ]
29. SCOTT G (1965). Atmospheric Oxidation and Antioxidants. Elsevier Publishing
Co., Amsterdam, N. Y., etc. See especially p. 93ff.
30. CALVERT JG & PITTS JN (1966). Photochemistry. John Wiley & Sons, N. Y.
31. ELLIS C & WELLS AA, revised and enlarged by HEYROTH FF (1941). The
Chemical Action of Ultraviolet Rays. Book Department, Reinhold Publishing
Corp., N.Y.
32. ANDERTON EJ et al. (1960). Photolysis of benzene in viscous solvents.
Proceedings of the Chemical Society (London) 1960 (October): 352.
33. TIPSON RS (1965). Oxidation of polycyclic, aromatic hydrocarbons: a review
of the literature. U.S. Department of Commerce, National Bureau of Standards,
Monograph 87.
99
-------
34. ERDMAN JG et al. (1956). Synthesis and properties of porphyrin-vanadium
complexes. Journal of the American Chemical Society 78(4):5844-5847.
35. ERDMAN JG et al. (1957). The stability of porphyrin-metallo complexes.
American Chemical Society, Division of Petroleum Chemistry, Preprints
2(l):259-267.
36. ROSSCUP RJ & BOWMAN DH (1967). Thermal stabilities of vanadium and
nickel petroporphyrins. American Chemical Society, Division of Petroleum
Chemistry, Preprints 12(2):A-77.
37. FALK JE (1964). Porphyrins and Metalloporphyrins: Their General, Physical
and Coordination Chemistry, and Laboratory Methods. Elsevier Publishing
Co., Amsterdam, N. Y. , etc.
38. FABUSS BM & KOROSI A (1967). Thermodynamic properties of sea water
and its concentrates. Desalination 2(3):271-278.
39. DODD CG et al. (1952). Metalliferous substances adsorbed at crude
petroleum-water interfaces. Industrial and Engineering Chemistry 44:2585.
40. DUNNING HN (1953). The interfacial activity of mesoporphyrin DC and some
derivatives. Journal of Colloid Science 8(3):279-287.
41. DUNNING HN et al. (1953). Interfacial activities and porphyrin contents of
petroleum extracts. Industrial and Engineering Chemistry 45:1759-1765.
42. DUNNING HN (1963). Geochemistry of organic pigments, pp. 367-436 in
BREGER IA [ed], Organic Geochemistry. Earth Science Series, Monograph
No. 16. Symposium Publications Division, Pergamon Press, Oxford, N. Y. etc.
43. WOLSKY AA & CHAPMAN FW Jr. (1960). A study of an intrinsic contaminant
in petroleum — the nature of vanadium in asphaltenes. Proceedings of the
American Petroleum Institute 40(Section III):423-427.
44. GROENNINGS S (1953). Quantitative determination of the porphyrin aggregate
in petroleum. Analytical Chemistry 25:938-941.
45. SVERDRUP HV et al. (1942). The Oceans. Prentice-Hall, N.Y.
100
-------
46. KRISS AE (1963). Marine Microbiology (Deep Sea). SHEWAN JM &
KABATA Z [trans, from Russian]. Interscience Publishers, N.Y.
and
KRISS AE et al. (1967). Microbiological Population of Oceans and Seas.
SYERS K [trans, from Russian]; FOGG GE [trans, ed]. St. Martin's Press,
N. Y. The original Russian edition, Mikrobnoye Naseleniye Mirovogo Okeana,
was published in 1964 by the Akademiya Nauk SSSR, Izdatel'stvo "Nauka,"
Moscow, 1964; the quoted passages of the English translation correspond to
pp. 281 (paras 3 and 4), 285 (paras 1-3), and 285 (para 5) of this Russian
edition.
47. BREED RS et aL. [eds] (1957). Sergey's Manual of Determinative Bacteriology,
7th Edition. Williams & Wilkins, Baltimore.
48. ZoBELL CE & UPHAM HC (1944). A list of marine bacteria including
descriptions of sixty new species. Bulletin of the Scripps Institution of
Oceanography, Technical Series 5(2):239-292.
49. BRJSOU J (1955). La microbiologie du milieu marin. (Collection de
1'Institut Pasteur.) Flammarion, Paris.
50. BEERSTECHER E (1954). Petroleum Microbiology: An Introduction to
Microbiological Petroleum Engineering. Elsevier Publishing Co., Houston.
51. DAVIS JB (1967). Petroleum Microbiology. Elsevier Publishing Co.,
Amsterdam, N. Y., etc.
52. McKENNA EF & KALLIO RE (1965). The biology of hydrocarbons. Annual
Review of Microbiology 19:183-208.
53. Van der LINDEN AC & THIJSSE GJE (1965). The mechanisms of microbial
oxidations of petroleum hydrocarbons. Advances in Enzymology and Related
Subjects of Biochemistry 27:469-546.
54. ROSSINI FD et al. (1953). Hydrocarbons from Petroleum: The Fractionation,
Analysis, Identification...; An Account of the Work of the API Research
Project 6. American Chemical Society Monograph Series No. 12. Reinhold
Publishing Corp., N. Y.
55. BESTOUGEFF MA (1967). loc. cit.
101
-------
56. WHITEHEAD WL & BKEGER IA (1963). Geochemistry of petroleum, pp. 248-
332 in BREGER IA [ed], Organic Geochemistry. Earth Science Series,
Monograph No. 16. Symposium Publications Division, Pergamon Press,
Oxford, N. Y., etc.
57. In addition to Refs. 58, 60, 61, 62, and 65, the following reviews are highly
recommended:
MILNER OI (1963). Analysis of Petroleum for Trace Elements. International
Series of Monographs on Analytical Chemistry, vol. 14. A Pergamon Press
Book, The Macmillan Co., N. Y.
and
WHITEHEAD WL & BREGER IA (1963). Geochemistry of petroleum, pp. 248-
332 in BREGER IA [ed], Organic Geochemistry. Earth Science Series,
Monograph No. 16. Symposium Publications Division, Pergamon Press,
Oxford, N. Y., etc.
and
ERDMAN JG (1962). Geochemistry of the high molecular weight and non-
hydrocarbon fraction of petroleum, pp. 215-237 in COLOMBO U & HOBSON GD
[eds], Advances in Organic Geochemistry. International Series of Monographs
on Earth Sciences, Monograph No. 15. Symposium Publications Division,
Pergamon Press, Oxford, N. Y., etc.
58. McCOY JW (1962). The Inorganic Analysis of Petroleum. Chemical Publishing
Co., Inc., N. Y.
59. RADCHENKO OA & SHESHINA LS (1955). O genezise porfirinov v neftyakh.
Akademiya Nauk SSSR, Doklady 105(6): 1285-1288.
60. MANSKAYA SM & DROZDOVA TV (1968). Geochemistry of Organic Substances,
SHAPIRO L & BREGER IA [trans, and ed. from the Russian]. International
Series of Monographs in Earth Sciences, Vol. 28. Pergamon Press, Oxford,
N. Y., etc.
61. DUNNING HN (1963). Geochemistry of organic pigments, pp. 367-430 in
BREGER IA [ed], Organic Geochemistry. Earth Science Series, Monograph
No. 16. Symposium Publications Division, Pergamon Press, Oxford, N. Y.,
etc.
102
-------
62. HODGSON GW et al. (1967). Geochemistry of porphyrins, pp. 177-259 in
NAGY B & COLOMBO U [eds], Fundamental Aspects of Petroleum Geochemistry.
Elsevier Publishing Co., Amsterdam, N. Y., etc.
63. CANEVARI GP (1968). A method to identify oil slicks of unknown origin. Esso
Research and Engineering Company, Esso Engineering, Technology Department.
Unpublished memorandum dated 8 November 1968.
64. GUINN VP & BELLANCA SC (undated; 1968 or later). Neutron activation
analysis identification of the source of oil pollution of waterways. Gulf
General Atomic Inc., San Diego, California. Unpublished manuscript.
65. JOHANNESSON JK (1955). The identification of fuel oils polluting coastal
waters. The Analyst (London) 80(956):840-841.
66. COSTANTINIDES G & ARICH G (1967). loc. cit.
67. FALK JE (1964). loc. cit.
68. THODE HG et al. (1958). Sulphur isotope abundances in petroleum and
associated materials. Bulletin of the American Association of Petroleum
Geologists 42(11):2619-2641.
69. THODE HG (1963). Sulphur isotope geochemistry, pp. 25-41 in SHAW DM
[ed], Studies in Analytical Geochemistry. The Royal Society of Canada,
Special Publications, No. 6. University of Toronto Press, Toronto.
70. HOBSON GD & LOUIS MC [eds] (1964). Advances in organic geochemistry
1964; proceedings of the International Meeting [of the 2nd International
Congress on Organic Geochemistry] in Reuil-Malmaison [France], 1964.
International Series of Monographs on Earth Sciences, vol. 24. Symposium
Publications Division, Pergamon Press, Oxford, N.Y., etc.
71. VINOGRADOV AP (1958). Isotopic composition of sulphur in meteorites and
in the earth, pp. 581-591 in EXTERMANN RC [ed], Radioisotopes in Scientific
Research, II. Pergamon Press, Oxford, New York, etc.
and
THODE HG etal. (1961). Sulphur isotope geochemistry. Geochimica et
Cosmochimica Acta 25(3):159-174.
103
-------
72. HARRISON AG & THODE HG (1958). Mechanism of the bacterial reduction of
sulphate from isotope fractionation studies. Transactions of the Faraday
Society 54(l):84-92.
and
JONES GE & STARKEY RL (1957). Fractionation of stable isotopes of sulphur
by microorganisms and their role in native deposition of sulfur. Applied
Microbiology 5(2): 111-118.
73. COSTANTINIDES G & ARICH G (1967). loc. cit.
74. DAVIS JB (1967). op. cit. See especially pp. 75-83.
75. ABLESON PH (1967). Oil pollution. Science 156(3778): 1037.
76. LUDWIG HF & CARTER R (1961). Analytical characteristics of oil-tar
materials on Southern California beaches. Journal of the Water Pollution
Control Federation 33(11): 1123-1139.
77. LUDWIG HF et al. (1964). Characteristics of oil and grease found in the
marine environment. Water & Sewage Works 111(11):486-491.
78. MAPES G (1969). Troubled waters: pollution of the seas, beaches by oil
poses major global problem. The Wall Street Journal 174(104):!, 20;
26 November 1969.
79. SCHULDINER JA (1951). Identification of petroleum products by chromato-
graphic fluorescence methods. Analytical Chemistry 32(11):1676-1680.
80. HERD M (1953). A paper-strip method of examining fuel oils suspected of
being identical. The Analyst (London) 78(927):383-384.
81. JOHANNESSON JK (1955). loc. cit.
82. CANEVARI GP (1968). loc. cit.
83. GUINN VP & BELLANCA SC loc. cit.
84. CIUFFOLOTTI L et al. (1964). Trace metal analysis of oils and asphalts by
neutron activation techniques, pp. 433-446 in COLOMBO U & HOBSON GD
[eds], Advances in Organic Geochemistry. International Series of Monographs
in the Earth Sciences, Monograph No. 15. Symposium Publications Division,
Pergamon Press, Oxford, N. Y., etc.
104
-------
85. McCOY JW (1962). The Inorganic Analysis of Petroleum. Chemical Publishing
Co., Inc., N. Y. See Especially pp. 184ff.
86. ROSSINI FD et al.' (1953). loc. cit.
87. PAULING L (1960). The Nature of the Chemical Bond, 3rd Edition.
Cornell University Press, Ithaca.
88. BADGER GM (1954). The Structures and Reactions of the Aromatic Compounds.
The Cambridge University Press, Cambridge, England.
89. FUSON RC (1962). Reactions of Organic Compounds: A Textbook for the
Advanced Student. John Wiley & Sons, N. Y.
90. DONALDSON N (1958). The Chemistry and Technology of Naphthalene
Compounds. Edward Arnold Publishers, Ltd., London.
91. GAFNER G & HERBSTEIN FH (1960). The crystal and molecular structures of
overcrowded halogenated compounds. I. Introductory survey. Acta Crystallo-
graphica 13 (9): 702-705.
92. CLAR EJ (1952). Aromatische Kohlenwasserstoffe, polycyclische Systeme.
Springer-Verlag, Berlin.
93. WE AST RC et al. [eds] (1968). The Handbook of Chemistry and Physics,
46th Edition. The Chemical Rubber Company, Cleveland. P. D120.
94. Le MAIRE GW (1958). A study of structures of molecular complexes found
in heavy petroleum fractions. Quarterly of the Colorado School of Mines
52(4):l-79.
and
WINNIFORD RS (1950). The evidence for association of asphaltenes in dilute
solution. American Chemical Society, Division of Petroleum Chemistry,
Preprints 5(4):A11-A21.
95. SERGIENKO SR (1965). High-molecular Compounds in Petroleum,
SCHMORAK J [trans, from Russian], Israel Program for Scientific Trans-
lations, Jerusalem. Daniel Davey, N. Y., U.S. agent.
105
-------
96. LOVELOCK JE & LIPSKY SR (1960). Electron affinity spectroscopy — a
new method for the identification of functional groups in chemical compounds
separated by gas chromatography. Journal of the American Chemical Society
82(2):431-433.
and
MITCHELL J Jr. (1961). Gas-chromatography-detector systems, in
CHERONIS ND [ed], Microchemistry Journal Symposium Series, Submicro-
gram Experimentation 1:201-218. Interscience Publishers, N. Y.
and
PIOZZI F & VITA-FINZI P (1960). Gascromatografia di idrocarburi fenantrenici.
Atti della Accademia Nazionale dei Lincei, Serie Ottava, Rendiconti, Classe di
Scienze fisiche, matematiche e naturali 29 (6):549-554.
and
ROUAYHEB GM et al. (1962). Quantitative gas chromatographic analysis of
hydrocarbon systems using the Lovelock diode detector and capillary columns.
Analytica Chimica Acta 26 (4):378-390.
97. ALBERT DK (1967). Determination of nitrogen compound types and distribution
in petroleum by gas chromatography with a coulometric detector. Analytical
Chemistry 39(10):1113-1117.
and
CREMER E et al. (1961). Anwendung eines hochempfindlichen selektiven
Halogen-Detektors in der gas-chromatographie. Chemie-Ingenieur-Technik
33(9):632-633.
and
SCHULEK E & REMPORT-HORVATH Zs (1960). Die Bestimmung von Halogenen
in organischen Stoffen. Annales Universitatis Scientiaruns Budapestinensis de
Rolando Eotvds Nominatae, Sectio Chimica, 2:123-128.
and
SOLO AJ & PELLETIER SW (1961). The gas-liquid chromatography of
phenanthrenes; its application to the separation and identification of
dehydrogenation mixtures from natural products. Chemistry & Industry
1961 (43): 1755-1756.
106
-------
and
TUEMMLERFD (1967). Petroleum. Analytical Chemistry 39(5):157R-190R
and
ZLATKIS A [ed] (1969). Advances in Chromatography: Proceedings of the
5th International Symposium, 20-23 January 1969, Las Vegas, Nevada.
Preston Technical Abstracts Co., Evanston, Illinois.
98. ETTRE LS (1968). Theory and applications of gas chromatography, pp. 28ff.
in KROMAN HS & BENDER SR [eds], Industry and Medicine. Grune & Stratton
co., N.Y.
99. CREMER E et al. (1961). loc. cit.
and
LANSDOWNE RA & LIPSKY SR (1962). Electron capture spectroscopy, an
adjunct to gas chromatography. Analytical Chemistry 34(7):726-730.
100. LOVELOCK JE & LIPSKY SR (1960). loc. cit.
101. KAWAHARA FK et al. (1968). Microanalyses of 14 chlorohydrocarbons in
wastewater by thin layer and gas chromatography. Journal of Gas Chromatog-
raphy 6(l):24-27.
and
ROSEN AA & MIDDELTON FM (1959). Chlorinated insecticides in surface
waters. Analytical Chemistry 31 (10): 1729-1732.
102. SMITH JE (1961). Gas chromatographic investigations of chlorinated hydro-
carbons and attendant air-sampling techniques. Ph. D Thesis, University of
Colorado, L.C. Microfilm 61-809. Dissertation Abstracts 21(10):2871-2872.
103. BUTLER PE & MUELLER WH (1967). The use of combined instrumental
techniques in the structure elucidation of petrochemical products. Abstracts,
American Chemical Society, Division of Petroleum Chemistry, 153rd Meeting.
and
v
GASPARIC J (1968). The systematic use of chromatography in structure
elucidation of organic compounds by chemical methods. Advances in
Chromatography 6:3-59.
107
-------
and
KLAAS PJ (1961). Gas chromatographic determination of sulfur compounds
in naphthas employing a selective detector. Analytical Chemistry
33(13):1851-1854.
and
LUCAS DM (1960). The identification of petroleum products in forensic
science by gas chromatography. Journal of Forensic Sciences 5:236-247.
and
SCHULTZE GR & KIRCHOFF K (1968). Analysis of petroleum and its
derivatives. Erdoel, Kohle, Erdgas, Petrochemie 21(3):148-154.
and
WHITHAM BT (1959). Recent applications of gas chromatographic techniques
in the petroleum industry. 31— Congres International de Chimie, Liege, 1958.
Compte Rendu [in two volumes], Industrie Chimique Beige, Bruxelles.
Volume I, pp. 563-570. Library of Congress call number TP5.C6.
104. SCHOLLY PR & BRENNER N (1961). Comparative retention values of
representative sample types on standard gas chromatography columns.
2nd International Symposium on Gas Chromatography, East Lansing, Mich.,
1959:263-309.
and
SCHOMBURG G (1968). Identification by retention and response values,
KELLER RA [trans, from German]. Advances in Chromatography 6:211-245.
105. KAINZ G & WACHBERGER E (1968). Determination of chlorine traces in
crude oil by combustions and potentiometric titration in nonaqueous media.
Mikrochimica Acta 1968(3):596-601.
and
LARSON CM & LARSON R (1960). Additives for petroleum products, in
GUTHRIE VB [ed], Petroleum Products Handbook, 1st Edition. McGraw-
Hill, N.Y.
108
-------
106. ISBELL HS et al. (1962). Analytical standards for trace elements in petroleum
products. U.S. Department of Commerce, National Bureau of Standards,
Monograph 54.
and
PINTA M (1966). Detection and Determination of Trace Elements: Absorption
Spectrophotometry, Emission Spectroscopy, Polarography. BIVAS M [trans.
from French, orig, published in 1962]. Israel Program for Scientific Trans-
lations, Jerusalem.
107. AMERICAN PETROLEUM INSTITUTE, DIVISION OF REFINING (1964).
Activation analysis. Proceedings of the American Petroleum Institute 44
(Sect. III):235-277. The Institute, N. Y.
and
BAKER CA (1967). Gamma-activation analysis, a review. Analyst (Cambridge)
92(1099):601-610.
and
BOWEN HJM & GIBBONS D (1963). Radioactivation Analysis. Oxford University
Press (at the Clarendon Press), London.
and
TAYLOR D (1964). Neutron Irradiation and Activation Analysis. Van Nostrand,
Princeton, N. J.
108. BAKER EW (1966). Mass spectrometric characterization of petroporphyrins.
American Chemical Society, Division of Petroleum Chemistry, Preprints
11(2) :B-145.
and
CLERC RJ & O'NEAL MJ Jr. (1961). Mass spectrometric analysis of asphalt:
A preliminary investigation. Analytical Chemistry 33(3):380-382.
and
HENNEBERG D (1961). Eine Kombination von Gaschromatograph und
Massenspektrometer zur Analyse organischer Stoffgemische. Frensenius'
Zeitschrift fur Analytische Chemie 183(1): 12-23.
109
-------
and
PETTERSSON B & RYHAGE R (1967). Mass spectral data processing;
identification of aliphatic hydrocarbons. Analytical Chemistry 39(7):790-793.
and
SIMMONS MC & KELLEY TR (1961). Quantitative analysis of complex samples
by gas-liquid chromatography and mass spectrometry. 2nd International
Symposium on Gas Chromatography, East Lansing, Mich., 1959:225-236.
109. ERDMAN JG (1962). Geochemistry of the high molecular weight and non-
hydrocarbon fraction of petroleum, pp. 215-237 in COLOMBO U & HOBSON
GD [eds], Advances in Organic Geochemistry. International Series of Mono-
graphs on Earth Sciences, Monograph No. 15. Symposium Publications
Division, Pergamon Press, Oxford, N. Y., etc. also see Ref. 10.
110. SERGIENKO SR (1965). loc. cit.
111. GRUSE WA & STEVENS DR (1960). Chemical Technology of Petroleum,
3rd Edition. McGraw-Hill, N.Y., etc.
112. SACHANEN AN (1945). The Chemical Constituents of Petroleum. Reinhold
Publishing Co., N. Y.
113. JONES JH et al. (1960). Vapor phase oxidation of aromatic hydrocarbons.
American Chemical Society, Division of Petroleum Chemistry, Preprints
5(2):C-5.
114. CAMPBELL P et al. (1962). The effect of temperature and humidity on the
oxidation of air blown asphalts. American Chemical Society, Division of
Petroleum Chemistry, Preprints 7(3): 123.
and
FUJIWARA S & WAINAI T (1961). Analysis of tar products by nuclear magnetic
resonance spectrometry: 1- and 2-methylnaphthalenes. Analytical Chemistry
33(8):1085-1087. Also see Ref. 10.
115. JONES JH et al. (1960). loc. cit.
116. TIPSON RS (1965). Oxidation of polycyclic, aromatic hydrocarbons: a review
of the literature. U.S. Department of Commerce, National Bureau of Standards,
Monograph 87.
117. DAVIS JB (1967). Petroleum Microbiology. Elsevier Publishing Co.,
Amsterdam, N. Y., etc.
110
-------
118. Van der LINDEN AC & THIJSSE GJE (1965). The mechanisms of microbial
oxidations of petroleum hydrocarbons. Advances in Enzymology and Related
Subjects of Biochemistry 27:469-546.
119. ZIELINSKI WL et al. (1967). Relationship of structure to sensitivity in
electron capture analysis. II. Chlorobenzenes, anilines and related
derivatives. Journal of Chromatography 30:77ff.
120. ISBELL HS et al. (1962). Analytical standards for trace elements in
petroleum products. U.S. Department of Commerce, National Bureau of
Standards, Monograph 54. U.S. Government Printing Office, Washington, D. C.
121. SOUTHWICK SH (1951). Inorganic constituents of crude oil. Ph. D. Thesis,
Massachusetts Institute of Technology, Department of Geology, Hayden
Library, Cambridge, Massachusetts.
and
MILNER OI (1963). Analysis of Petroleum for Trace Elements. International
Series of Monographs on Analytical Chemistry, Vol. 14. A Pergamon Press
Book. The Macmillan Company, N. Y.
and
McCOY JW (1962). The Inorganic Analysis of Petroleum. Chemical Publishing
Company, Inc., N. Y.
122. WHITEHEAD WL & BREGER IA (1963). Geochemistry of petroleum,
pp. 248-332 in BREGER IA [ed], Organic Geochemistry. Earth Science
Series, Monograph No. 16. Symposium Publications Division, Pergamon
Press, Oxford, N. Y., etc.
I
and
COSTANTINIDES G & ARICH G (1967). loc. cit.
123. HOFFMEISTER WS (1959). Palynology's first 10 years as an aid to finding oil.
Oil & Gas Journal 57(34):246-250.
124. SANDERS JM (1937). The microscopical examination of crude petroleum.
Journal of the Institution of Petroleum Technologists 23(167): 525-573.
125. WALDSCHMIDT WA (1941). Progress report on the microscopic examination
of Permian crude oils. Abstracted from the 26th Annual Meeting, Bulletin of
the American Association of Petroleum Geologists 25(23):934.
Ill
-------
126. SERGIENKO SR (1965). High-molecular compounds in petroleum, 2nd edition.
SCHMORAK J [trans, from Russian]. Israel Program for Scientific Translations,
Jerusalem. Daniel Davey & Co., N. Y., U. S. agent.
127. WITHERSPOON PA & MUNIR ZA (1960). Size and shape of asphaltic particles
in petroleum. Producers Monthly 24(10):20-31.
128. WITHERSPOON PA & WINNIFORD RS (1967). The asphaltic components of
petroleum, pp. 261-267 in NAGY B & COLOMBO U [eds], Fundamental
Aspects of Petroleum Geochemistry. Elsevier Publishing Co., Amsterdam, N. Y.
129. HEDBERG HD (1964). Geologic aspects of the origin of petroleum. Bulletin
of the American Association of Petroleum Geologists 48(11): 1755-1803.
130. ARTAMONOVA SV & MEDVEDEVA AM (1963). Methods of extracting spores
and pollen from oils and the waters of oil deposits. International Geology
Review 5(11): 1510-1511.
131. BLIZNICHENKO SI & KRASNOVA LY (1964). Spores and pollen of the West
Siberian Lowland. Geologiia i Geofizika no. 12:116-119. [courtesy of
WE HANSON, Exploration Division of Gulf Research, Pittsburgh, Pa. ]
132. CHEPIKOV KR & MEDVEDEVA AM (1963). New data on the microfossils in
the crude oils of European USSR. Akad. Nauk S. S.S. R., Doklady, Earth
Sciences Sections 153 (l-6):66-67. [courtesy of WE HANSON, Exploration
Division of Gulf Research, Pittsburgh, Pa. ]
133. HOROWITZ A & LANGOZKY Y (1965). Preliminary palynological study of
hydrocarbons in Israel. Geologic en MijnBouw 44(2): 59-63.
134. deJERSEY NJ & ALLEN RJ (1966). Jurassic source for oil of Surat-Bowen
Basin, Queensland, Australia. Bulletin of the Australian Association of
Petroleum Geologists 50(11):2479-2481. [courtesy of GJ SCHRAYER,
Exploration Division of Gulf Research, Pittsburgh, Pa. ]
135. CHEPIKOV KR & MEDVEDEVA AM (1967). Plant microfossils as possible
indicators of oil and natural gas migration. USSR All-Union Oil and Gas
Genesis Symposium (Moscow, Feb. 1967) Proceedings:394-397 (in Russian).
[courtesy of GJ SCHRAYER, Exploration Division of Gulf Research,
Pittsburgh, Pa.]
136. TOMOR J (1964). The genesis and age of Hungarian oils. Zeitschrift fur
angewandte Geologie 10(11):579-585 [in German].
112
-------
137. KRUEGER EO [ed] (1969). Liquid borne particle metrology. Annals of the
New York Academy of Sciences, 158(Art. 3):589-579. [courtesy of RG
NEMCHIN, Melpar, Inc., Falls Church, Va.]
138. BROWN CA (1960). Palynological techniques. Louisiana State University -
Library of Congress 1960.
139. MIZIUKE A (1965). Separations and preconcentrations. pp. 103-159 in
Trace analysis; physical methods, MORRISON GH [ed]. Interscience
Publishers, N. Y.
140. ALLEN T (1968). Particle size measurement. Chapman & Hall, N. Y.
141. SKLAREW S (1934). Preparation of microscopic glass spheres. Industrial
& Engineering Chemistry - Analytical Edition 6 (2): 152-153.
142. SOLLNER K (1939). A simplified method of preparing microscopic glass
spheres. Industrial & Engineering Chemistry - Analytical Edition ll(l):48-49.
143. BLOOMQUIST CR & CLARK A (1940). Preparation of microscopic glass
spheres. Industrial & Engineering Chemistry - Analytical Edition 12(1):61-62.
144. LAAN H van der & NICHOLLS RW (1960). Simple apparatus for making 10-
to 100-|_i spheres of low melting point metals. Review of Scientific Instru-
ments 31(2): 212.
145. POTTER CC (1966). Direct current plasma... a new spheroidizing process.
Metal Progress 90(5): 127-130.
146. ANONYMOUS (1961). Spheroidization of particles, ceramics, metals,
refractories, and radioactive materials... [editorial information]. Industrial &
Engineering Chemistry 53(2):95A.
147. ANONYMOUS (1963). Prefabricated foams: outlet for tiny spheres [editorial
comment]. Chemical Engineering 70(17):92-94.
148. KHOMIKOVSKY P & REHBINDER P (1938). Dependence of stabilization and
wetting of particles suspended in oil medium on the quantity of surface-active
substance adsorbed. Comptes Rendus (Doklady) de 1'Academie des Sciences
de 1'URSS 18(8):575-578 [In English].
149. KOELMANS H (1955). Suspensions in non aqueous media [sic]. Special
printing, doctoral thesis, Univ. of Utrecht [Eindhoven, Netherlands] 36 pp.
113
-------
150. KNOWLES EC et al. (1967). Stable suspensions of powders. U.S. Patent:
US 3,328, 238(C1.167-16), June 27, 1967, Appl. Dec. 22, 1965.
151. McCRONE WC et al. (1967). The particle atlas. Ann Arbor Science
Publications Inc., Ann Arbor, Michigan.
152. BLANCHARD MB (1967). Method for determining the density of microsize
spherical particles. Geological Society of America, Bulletin 78(3):385-404.
153. KIRK PL (1950). Quantitative ultramicroanalysis. John Wiley & Sons, Inc.,
N.Y.
154. MEINKE WW & SCRIBNER BF (1967). Trace characterization; chemical and
physical. U.S. National Bureau of Standards, Monograph 100.
155. SVENTITSKII NS (1965). Visual methods of emission spectroscopy. VARDIN P
[trans, from Russian]. Israel Program for Scientific Translations, Jerusalem,
Daniel Davey & Co., N. Y., U.S. agent.
156. HARVEY CE (1964). Semiquantitative spectrochemistry. Applied Research
Laboratories, Inc., Glendale, Calif.
157. HERCULES DM [ed] (1966). Fluorescence and phosphorescence analysis.
Inter science Publishers, N. Y.
158. LEBEDEV VJ (1969). Suspensions of rocks and minerals in flame spectro-
photometry. Zh. Anal. Khim 24(3):337-341 [in Russian).
159. CURIE D (1963). Luminescence in crystals. John Wiley & Sons, Inc.,
N.Y. 1963.
160. DEMENT JA (1942). Fluorescent chemicals and their applications. Chemical
Publishing Company, Inc., Brooklyn, N. Y.
161. DRUSHEL HV & SOMMERS AL (1964). Direct determination of inhibitors in
polymers by luminescence techniques. Analytical Chemistry 36(4):836-840.
162. FISCHER EK (1950). Colloidal dispersions. John Wiley & Sons, Inc., N.Y.
163. FORSTER CF & RICHARD EF (1963). Phosphorescence in plastics. Nature
197 (4873): 1199-1200.
164. LLOYD FE (1923). A method of ultramicroscopy whereby fluorescence in
the Cyanophycae and Diatomacae may be demonstrated. McGill University
Publications, Series II(Botany), no. 22, 3 pp.
114
-------
165. MELHUISH WH & HARDWICK R (1962). Lifetime of the triplet state of
anthrancene in Lucite. Transactions of the Faraday Society 58(10): 1908-1911.
166. OLSON RA (I960).. Rapid scanning microspectrofluorimeter. Review of
Scientific Instruments 31(8):844-849.
167. REISBERG J & DOSCHER TM (1956). Interfacial phenomena in crude-oil-
water systems. Producers Monthly 21(1):43-50.
168. HARREMONES P (1967). Tracer investigations of effluent dispersion,
pp. 207-219 in Ref. 174.
169. GUIZERK J et al. (1967). Traceurs radioactifs et etude de la dispersion
des elements polluants en mer, dans les fleuves et dans les lacs,
pp. 221-238 in Ref. 174.
170. GUINN VP (1963). Special features of tritium as a tracer in industrial
research, pp. 227-233 in ROTHCHILD S [ed], Advances in Tracer Methodology,
Volume 1. The Plenum Press, N. Y.
171. ANGUENOT F (1967). Mesure continue au moyen d'une chambre d'ionisation
de la consommation d'huile d'un moteur a explosion a partir d'huile triti&e,
pp. 289-299 in Ref. 174.
172. CHATTERS RM & PETERSON RL (1967). Use of tracers in in-plant evaluation
of processes and in pollution control of effluent from pulp and paper mills into
rivers, pp. 251-258 in Ref. 174.
173. THE AMERICAN PETROLEUM INSTITUTE (1964). Twenty-ninth midyear
meeting session on activation analysis [a collection of 5 papers]. Proceedings
of the American Petroleum Institute 44(Sect. III):233-276.
174. INTERNATIONAL ATOMIC ENERGY AGENCY (1967). Radioisotope Tracers
in Industry and Geophysics; Proceedings of the Symposium held by the IAEA
in Prague, 21-25 November 1966. The Agency, Vienna.
115/116
* U. S. GOVERNMENT PRINTING OFFICE : 1910 O - S87-BI
-------
BIBLIOGRAPHIC:
Melpar, a division of American-Standard, Oil
Tagging System Study, FWPCA Contract No. 14-12-500,
May 1970.
ABSTRACT
Several methods of identifying the source of oil pol-
lution are critically examined. These methods are grouped
into two categories: passive tagging and active tagging.
Passive tagging assumes that oils are so chemically
diverse that their contents constitute a stable chemical
fingerprint that can be unequivocally disclosed in the labo-
ratory. Active tagging requires that an inexpensive, coded
material be added to oil; this material must be chemically
and physically stable in both oil and oil slicks: it must also
be readily identifiable by available analytical techniques;
and it must have no adverse effect on the oil's subsequent
use. Three methods of passive tagging (trace metals,
sulfur-isotope ratios, and paper chromatography) and three
ACCESSION NO.
KEY WORDS:
Pollutant Identification
Indicators
Tagging
Oil Wastes
Tracers
Wqt.r Pollution Control
Oil Industry
Wast* Identification
Analytical Techniques
Chemical Analysis
Chromatography
BIBLIOGRAPHIC:
Melpar, a division of American-Standard, Oil
Tagging System Study, FWPCA Contract No. 14-12-500,
May 1970.
ABSTRACT
Several methods of identifying the source of oil pol-
lution are critically examined. These methods are grouped
into two categories: passive tagging and active tagging.
Passive tagging assumes that oils are so chemically
diverse that their contents constitute a stable chemical
fingerprint that can be unequivocally disclosed in the labo-
ratory. Active tagging requires that an inexpensive, coded
material be added to oil; this material must be chemically
and physically stable in both oil and oil slicks; it must also
be readily identifiable by available analytical techniques;
and it must have no adverse effect on the oil's subsequent
use. Three methods of passive tagging (trace metals.
sulfur-isotope ratios, and paper chromatography) and three
ACCESSION NO,
KEY WORDS:
Pollutant Identification
Indicators
Tagging
Oil Woit.t
Tracers
Water Pollution Control
Oil Industry
Wast* Identification
Analytical Techniques
Chemical Analysis
Chroniatograpky
BIBLIOGRAPHIC:
Melpar, a division of American-Standard, Oil
Tagging System Study, FWPCA Contract No. 14-12-500.
May 1970.
ABSTRACT
Several methods of identifying the source of oil poj-
lution are critically examined. These methods are grouped
into two categories: passive tagging and active tagging.
Passive tagging assumes that oils are so chemically
diverse that their contents constitute a stable chemical
fingerprint that can be unequivocally disclosed in the labo-
ratory. Active tagging requires that an inexpensive, coded
material be added to oil; this material must be chemically
and physically stable in both oil and oil slicks; it must also
be readily Identifiable by available analytical techniques;
and it must have no adverse effect on the oil's subsequent
use. Three methods of passive tagging (trace metals,
sulfur-isotope ratios, and paper chromatography) and three
ACCESSION NO.
KEY WORDS:
Pollutant Identification
Indicators
Tagging
Oil Woste.
Trocars
Woter Pollution Control
Oil Industry
Wast* Identification
Analytical Techniques
Chemical Analysis
Chromatography
-------
methods of active tagging (halogenated polycyclic aromatics,
organometallics, and coded microspheroids) have been
examined. Passive tags cannot be recommended because
the passive tags are quite likely to mingle, to evaporate, to
be dissolved, or to be oxidized; even if these processes do
not occur, they can create formidable forensic problems
for the prosecution and telling counter-arguments for the
defense. Since active tags are designed to be stable and
identifiable, they are satisfactory for the job; and the three
types of active tags reviewed show promise and merit.
This report was submitted in fulfillment of Contract
Number 14-12-500 under the sponsorship of the Federal
Water Pollution Control Administration.
methods of active tagging (halogenated polycyclic aromatics,
organometallics, and coded microspheroids) have been
examined. Passive tags cannot be recommended because
the passive tags are quite likely to mingle, to evaporate, to
be dissolved, or to be oxidized; even if these processes do
not occur, they can create formidable forensic problems
for the prosecution and telling counter-arguments for the
defense. Since active tags are designed to be stable and
identifiable, they are satisfactory for the job; and the three
types of active tags reviewed show promise and merit.
This report was submitted in fulfillment of Contract
Number 14-12-500 under the sponsorship of the Federal
Water Pollution Control Administration.
methods of active tagging (halogenated polycyclic aromatics,
organometallics, and coded microspheroids) have been
examined. Passive tags cannot be recommended because
the passive tags are quite likely to mingle, to evaporate, to
be dissolved, or to be oxidized; even If these processes do
not occur, they can create formidable forensic problems
for the prosecution and telling counter-arguments for the
defense. Since active tags are designed to be stable and
identifiable, they are satisfactory for the job; and the three
types of active tags reviewed show promise and merit.
This report was submitted in fulfillment of Contract
Number 14-12-500 under the sponsorship of the Federal
Water Pollution Control Administration.
------- |