EPA-R2-73-221
July 1973 Environmental Protection Technology Series
A Multiparameter
Oil Pollution Source
Identification System
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
t. Environmental Monitoring
5. Socioeconomic Environmental studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-R2-73-221
July 1973
A MULTIPARAMETER OIL POLLUTION
SOURCE IDENTIFICATION SYSTEM
US EPA
Headquarters and Chemical Libraries
EPA West Bldg Room 3340
Mailcode 3404T
1301 Constitution Ave NW
Washington DC 20004
202-566-0556
By
John W. Miller
Contract No. 68-01-0059
Project 15080 HDJ
Project Officer
Bernard Hornstein
Oil Spills Research Branch
Edison Water Quality Research Laboratory, NERC (Cincinnati)
Edison, New Jersey 08817
Prepared for:
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C;« 20460
For sale by the Superintendent of Documents, tr.S. (j-0-v:enW$«ff.Printing Office, Washington, D.C. 20402 - Price $1.85
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency, and approved for publication.
Approval does not signify that the contents
necessarily reflect the views and policies of the
Environmental Protection Agency, nor does mention
of trade names or commercial products constitute
endorsement or recommendation for use.
ii
-------
ABSTRACT
The feasibility of oil pollution source identification is demonstrated
on eighty crude oils from the world's major oil fields. Measurements
of fifteen diagnostic parameters were made on the 600 F fraction of
the crude oil samples. Of the fifteen parameters studied it was
demonstrated that six were sufficient to distinguish among the crude
oils. These parameters are carbon and sulfur isotopic composition,
sulfur, nitrogen, vanadium and nickel contents. A hydrocarbon gas
chromatcgraphic profile was also diagnostic for identification but its
usefulness was reduced for aged samples by the effect of weathering.
The other parameters studied were the saturate, aromatic and asphaltic
contents and the carbon isotopic composition of each of these
fractions, the n-paraffin distribution (odd-even predominance curves)
and the sulfur gas chromatographic profile. The influence of weather-
ing on the parameters was studied.
A statistical procedure based on multivariate normal analysis was
developed to compare an unknown with a data library and to give an
unbiased match of the unknown with a known based on the precision of
the measurement methods.
This report was submitted by Research and Development Department,
Phillips Petroleum Company, Bartlesville, Oklahoma 74004 in fulfillment
of Project Number 15080 HDJ, Contract Number 68-01-0059 under the
sponsorship of the Water Quality Office, Environmental Protection
Agency.
ill
-------
CONTENTS
Section
I Conclusions
II Recommendations
III Introduction
IV Objectives
V Crude Oil Sources
VI Measurement of Parameters
VII Results and Discussion
Hydrocarbon GLC Profile
Carbon Isotopic Composition
Sulfur Isotopic Composition
Sulfur Gas Chromatographic Profile
Silica Gel Fractionations
Carbon Isotopic Composition of Silica Gel Fractions
Normal Paraffins and Odd-Even Predominance Curves
VIII Comparison Techniques
IX Weathering Studies
Weathering Procedure
X Summary Discussion
Practical Evaluation of the Proposed System
XI Acknowledgments
XII References
XIII Appendices A through N
1
3
5
9
13
25
29
36
40
43
51
54
54
58
61
73
73
81
84
89
91
95
-------
FIGURES
Page
1 Geographical Distribution of Crude Oil Sources 17
2 Schematic Diagram of Oil Pollution Source 30
Identification System
3 Typical Gas Chromatographic Profiles of 600 F Residues 38
4 Frequency Distribution of Hydrocarbon GLC Profiles by 39
Classification Type
5 Distribution of Carbon and Sulfur Isotopic Compositions 42
6 Geographical Distribution of Crude Oil Sources and 47
Sulfur Isotopic Composition - Map
7 Typical Gas Chromatographic Sulfur Fingerprints 53
8 Typical Odd-Even Predominance (OEP) Curves 60
9 Effect of Weathering on GLC Profile 600+F Bottoms of 78
Ecuador Crude Oil
10 Effect of Weathering on GLC Profile (600+F Bottoms) 79
vi
-------
TABLES
Page
1 Origin of Crude Oil Imports Into United States, 14
1970, 1971 and 1972
2 Crude Oil Sources Identification by Number 18-20
3 Crude Oil Sources - Alphabetic by Country 21-23
4 Values of Identification Parameters on Crude Oil 32-34
Bottoms
5 Range of Values for Sulfur, Nitrogen, Vanadium and 35
Nickel Contents of 600+F Hesidues
6 Sulfur Isotopic Composition of 600 F Bottoms by the 45-46
Proposed Reductive (Hydrogenation) Procedure
7 Comparison of Sulfur Isotopic Composition With 49
Literature Values
8 Application of Sulfur Isotopic Composition 49
9 Comparison of Sulfur Isotope Ratios by the Oxidative 50
and Reductive Methods
10 Values for Carbon Isotopic Composition on Crude Oil 55-57
Bottoms and Their Silica Gel Fractions
11 Comparison of Unknown 1 With Potential Sources 65
12 Comparison of Unknown 2 With Potential Sources 66
13 Comparison of Unknown 3 With Potential Sources 67
I/, Comparison of Unknown 4 With Potential Sources 68
15 Comparison of Unknown 5 With Potential Sources 69
16 Comparison of Mississippi Crude Oil Sources 71
17 Identification Efficiency of Sulfur Isotopic Composition 72
18 Effect of Weathering - Crude Oil No. 83, Monagas 75
Pipeline Crude Oil, Venezuela
vii
-------
TABLES (CONTD.)
Page
19 Effect of Weathering - Crude Oil No. 84, Ecuador 76
Composite Crude
20 Effect of Weathering of Relative Normal Paraffin 80
Content
21 Summary of Values on EPA Crude Oils and Weathered 85
Samples
•van
-------
SECTION I
CONCLUSIONS
1. The feasibility of oil pollution source identification has been
demonstrated on eighty crude oils from the world's major oil
fields. Measurements of 15 diagnostic parameters were made on
the 600 F fraction of the crude oil samples. The parameters are:
1. Carbon Isotopic Composition 11. Asphaltic Content
2. Sulfur Content 12. Odd-Even Predominance
3. Nitrogen Content Curves
4. Vanadium Content 13. Carbon Isotopic
5. Nickel Content Composition-Saturates
6. Sulfur Isotopic Composition 14. Carbon Isotopic
7. Hydrocarbon GLC Profile Composition-Aromatics
8. Sulfur GLC Profile 15. Carbon Isotopic
9. Saturate Content Composition-Asphaltics
10. Aromatic Content
A multivariate statistical procedure was developed and tested to
match the 15 parameters of an "unknown" with those of each 600 F
fraction from the crude oils. All 15 parameters were not required
for positive identification; the first six listed in (l) above
were sufficient to distinguish uniquely among 48 sources.
2. Our outdoor weathering studies demonstrated that of the 15 diag-
nostic parameters only hydrocarbon GLC profile and saturate,
aromatic and asphaltic contents were affected by weathering. The
change of the value of these four parameters was dependent upon
the length of weathering time (maximum 49 days).
-------
3. Sulfur isotopic composition is an important identification
parameter by virtue of the wide range of its values and its
stability to weathering that were established in this study.
-------
SECTION II
EECCMfENDATIONS
1. It is recommended that the variations in the values of the
proposed identification parameters that will occur in the normal
course of oil production be established on several major coramer-
cial crude oil sources.
2. The proposed multiparameter identification system should be
tested on naturally-weathered crude oils to demonstrate its
validity.
3. It is recommended that the feasibility of this crude oil
identification system be established for other petroleum
products such as fuel oils and refined products that are
"persistent11 in an aqueous environment.
4. An extension of the present studies on sulfur gas chromatographic
profile and sulfur isotopic composition is recommended with the
objective to reduce their measurements to a routine procedure.
-------
SECTION III
INTRODUCTION
The expanded movement of petroleum by water transport to meet existing
and predicted world and U.S. energy demands will increase the accidental
or deliberate discharge of persistent oils into the waterways of the
world. In U.S. waters in 1971 there were 8,736 spills of oil and other
polluting substances as reported by the U.S. Coast Guard. Large
spills that resulted from tank rupture and collisions accounted for
40/6 of the volume spilled (3,537,343 gallons) but only 1.5/6 of the
spill incidents. The source of these catastrophic spills is readily
determined. However, of major concern are the large number of smaller
spills (2,353), the source of which is unknown. The identification of
these spills to establish liability presents a difficult problem to
both the law enforcement and the scientific communities. Ideally,
the basis for source identification should stand up in legal
proceedings without the benefit of circumstantial evidence.
Two general approaches to the identification problem are recognized:
passive tagging and active tagging. Passive tagging is based on the
properties and constituents indigenous to petroleum and assumes that
these properties are sufficiently diverse to distinguish among the
large number of crude oil sources. Active tagging is based on the
deliberate addition of coded materials to petroleum at each point of
liability transfer. The deliberately added materials are chosen so
as to be easily measured. The selection of either passive or active
tagging depends on an evaluation of objectives and their costs.
Whether the objective is to assess liability from an enforcement
viewpoint or simply to eliminate one's coamercial oil from suspicion
must be determined.
-------
The requirements for active tagging and an illustration of its
2
application are described by Melpar Corporation. The application
of magnetic particles to active tagging has been disclosed recently.
Although active tagging appears attractive from the measurement point
of view, the technical problem of homogeneously blending a relatively
few parts per million of coded material into a tanker shipment of
crude oil is difficult and would be costly on a worldwide basis.
Crude oil is not a uniform material but varies in viscosity from a
water-like material to a solid at ambient temperatures depending upon
its source. Often, heated-transport facilities are necessary to permit
crude oil to be moved. The number of codes required for active tagging
is large and the bookkeeping task is not insignificant. It appears
that all petroleum transporters would be burdened with the need to
tag their cargos to establish liability against a few offenders.
Passive tagging is implemented by the analytical determination of
characteristic chemical and physical properties (parameters) of a
unknown crude oil residue. These measured parameters are often
referred to collectively as "fingerprints" and are compared either
with those from a number of Suspected sources or with those from a
library of all potential sources to identify the unknown. Since the
sample itself provides all the evidence for identification, changes
which occur in the composition of a crude oil when it is spilled into
the environment must be considered. These changes are a result of
evaporation, microbial degradation, photolysis, oxidation, dissolution
and adsorption and are referred to collectively as "weathering". The
most important short-term weathering processes of water-borne oils are
evaporative and dissolution losses, which occur at various rates
dependent upon environmental conditions (wind velocity and turbulence
temperature of air and water, rate of spin spread, wave action and
type of water, fresh or saline). Therefore, any passive tagging
-------
system must reduce all petroleum samples, i.e., weathered and un-
weathered, to a common base for measurement of the indigenous
constituents to eliminate or at least minimize the weathering
effects. A convenient common base is a distillation bottoms, the
properties of which form a basis for comparison.
Passive tagging is often more successful in elimination of suspect
sources than in positive identification of an unknown. Therefore,
the strength of any passive system will be determined by its ability
to identify an unknown uniquely with respect to source.
The Western Oil and Gas Association has developed a tentative method
for comparison of unknown petroleum pollutants with known sources
based on five measured parameters: sulfur, nitrogen, vanadium and
nickel contents and gas ehromatography. The success of this com-
parison method is not known. A wide variation in the values for
these parameters exists among crude oils from various sources. In
cases where the differences between the known sources are large, this
tentative method is adequate to identify a pollutant. If two or more
known sources are equivalent within the repeatability of the measure-
ment methods, unambiguous identification is not possible and circum-
stantial evidence must be considered in the final decision. To
strengthen the potential of this tentative method of comparison a
number of additional parameters could be considered. Adlard
reviewed those identification parameters that have been suggested in
the open literature, i.e., infrared spectroscopy, ultraviolet
fluorescence spectroscopy and chromatographic techniques. Up to this
point no one identification parameter or collection of parameters has
been shown to be superior to the others. The particular combination
of parameters depends upon the facilities available and the expenditure
which is justified to identify an unknown sample.
-------
SECTION IV
OBJECTIVES
The objective of the study described in this report was to develop
and test the feasibility of a multiparameter oil pollution source
identification system on a worldwide basis. The system studied
extended the Western Oil and Gas Association method to include ten
additional parameters.
The economic value of a petroleum source depends upon the gross
physical and chemical properties of the crude oil. A large library
of such data for commercial crude oils exists in the literature and
within petroleum industry files and new sources are evaluated con-
tinuously on this basis. In general, however, these compositional
variables are modified by weathering and are not suitable for
identification of an unknown pollutant.
The measured parameters included in this feasibility study were
selected on the basis of our experience with comparison techniques
for the identification of crude oils with respect to geologic source-
rock origin. These comparison techniques were based on an under-
standing of petroleum genesis, the chemical and physical processes
that occur during the migration and accumulation of petroleum in
reservoirs, and the changes that occur during weathering. Stability
to weathering must be evaluated on the basis of changes in real
samples rather than on the basis of changes that are possible with
2
isolated components or compounds which comprise crude oil. The
essential requirement is that an identification parameter be stable
for a few days to several weeks. Stability is defined operationally
as no statistically significant change in the selected parameter.
-------
If the unknown sample is suspected of being non-representative of the
original water-borne oil pollutant, i.e., that the isolated pollutant
is contaminated with components from the water or mixed with a second
oil or other organic material, the analyst must provide guidance on
the interpretation of the data. The ability of the analyst to obtain
a sufficient quantity of a representative oil pollutant is a practical
consideration not to be ignored.
The development of a system for oil pollution source identification
was planned in five phases. Phase I constituted the installation and
evaluation of initial analytical procedures while Phase II established
an identification feasibility of this approach for North and South
America crude oils. Phase III extended this identification feasibility
to eastern hemisphere crude oils. Phase IV provided additional
parameters to improve the reliability of this approach. Phase V
extended the identification scheme to include two additional parameters,
sulfur isotope ratios and sulfur gas chromatographic profile.
Practically, Phases I, II and III were conducted concurrently to
constitute a study that demonstrated the feasibility of an identifica-
tion system based on the measurement of carbon isotopic composition in
addition to the five parameters recommended by the Western Oil and Gas
Association: sulfur, nitrogen, vanadium and nickel contents and the
hydrocarbon gas chromatographic profile of simulated-weathered crude
oil residues. The crude oils that were used are representative of the
major sources of water-transported oils. Then, further characteriza-
tions were made using additional parameters in Phase IV by fractionating
the crude oil residues into three portions: saturates, aromatics and
asphaltics. The carbon isotopic composition was measured on each
fraction. The normal paraffins were removed from the saturates
fraction and their carbon number distribution was measured. The final
10
-------
portion of the work (Phase V) added the measurement of sulfur isotopic
composition and sulfur gas chromatographic profile. The identifica-
tion potential was strengthened with these added parameters. The
combination of these 15 parameters into smaller groups for different
degrees of identification potential was also investigated.
The major tasks necessary to achieve the objectives of this study were:
1. Select identification parameters based on experience in crude oil
correlation studies.
2. Install and evaluate necessary analytical procedures to measure
these parameters.
3. Select crude oils that represent significant worldwide sources
and measure the parameters of these oils.
4. Validate the parameter selection on the basis of the variation in
parameter values observed for these oils.
5. Develop comparison techniques to match measured parameters for a
known and unknown crude oil on an unbiased basis.
6. Evaluate different combinations of parameters with respect to their
ability to distinguish among crude oil sources using the comparison
techniques developed. t
7. Test the influence of weathering on the measured parameters and
their identification potential.
11
-------
SECTION V
CRUDE OIL SOURCES
The crude oils that were included in this study were selected on the
basis of the facts discussed in the following paragraphs.
The increased demand for petroleum products in the United States is
being met by increased imports of both crude oil and products.
Preliminary information for 1972 reported by the American Petroleum
7 &
Institute (API)'* indicates that crude oil imports were up 30$ and
product imports up 9$ over 1971 levels. Crude oil imports in 1972
amounted to an average of 2,189,000 barrels per day (b/d) with 69$
of the 503)000 b/d increase coming east of the Rockies, an area that
accounted for 1,515,000 b/d imports. Every indication points to
significant increases each year in crude oil Imports. Practically
9
all the growth must come from sources in the Eastern Hemisphere.
The origin by country of these imports in 1970, 1971, and 1972 is
g
summarized in Table 1. All the areas except South America showed
an increase in their imports into the United States with the Middle
East and Africa increasing their share of the total at the expense
of Canada and South America. The dependence upon the Eastern
Hemisphere for increased sources of crude oil imports means that
this area and especially the Middle East around the Persian Gulf will
become a predominate source of water-transported crude oil imports
in future years.
13
-------
TABLE 1
Bolivia
Columbia
Chile
Trinidad
Venezuela
Sc
Algeria
Angola
Egypt
Libya
Tunisia
Nigeria
A;
Kuwait
Saudi Arabia
Iran
Iraq
Qatar
Canada
Far East
ORIGIN OF CRUDE OIL IMPORTS INTO
UNITED
1970
1,000
b/d
2
20
-
_
268
uth American Total 291
6
—
21
47
-
rica Total 122
33
ia 40
33
-
(United Arab Emirates) __6_3_
ddle East Total 169
6?2
70
tal Crude 1,324
Total
0.2
1.5
-
_
20.2
22.0
0.5
-
1.6
3.5
-
^A
9.2
2.5
3.0
2.5
-
4.8
12.8
50.8
5.3
-
STATES -
1970. 1^
1971
1,000
b/d
2
9
1
-
303
315
13
4
19
53
-
184
29
115
106
11
80
341
721
117
1,681
Total
0.1
0.5
0.1
-
18.0
18.7
0.8
0.2
1.1
3.2
-
11.0
1.7
6.8
6.3
0.7
4.8
20.3
42.9
7.0
_
21^1222
1972*
1,000
b/d
-
1
-
18
2£1
272
78
13
7
137
6
231
472
31
166
126
3
4
112
449
835
161
2,189
Total
-
-
-
0.8
11.6
12.4
3.6
0.6
0.3
6.3
0.3
10.6
21.6
1.4
7.6
5.8
0.1
0.2
20.5
38.1
7.4
_
1972 data are preliminary
14
-------
New domestic sources of crude oil that will become of increased
importance to supply energy demands are located in areas where water
pollution is of great concern. Namely, offshore United States in the
Gulf of Mexico, especially Louisiana, and the Pacific Ocean between
the Gulf of Alaska and the Northwest states and California. The
latter ocean area will be crossed with tankers delivering primarily
North Slope Alaskan crude oil to United States refineries. The crude
oil will cross Alaska through a proposed pipeline. Offshore Louisiana
in the deeper waters is in an area of high production potential and
will be explored when acreage and equipment become available. As the
drilling and production activities grow, the oil pollution potential
will increase in the Gulf Coast waters.
Many California crude oils are significantly different in properties
from typical Mid-continent or Pennsylvanian crude oils. California
crude oils have a relatively high nitrogen and naphthenic acid content
and low saturate and normal paraffin content. Their unique character-
istics should be considered in any worldwide identification system.
With these facts in mind, representative crude oils were selected on a
worldwide basis for this feasibility study of multiparameter source
identification systems. The geographic distribution of the selected
crude oils is shown in Figure 1. The crude oil sources are identified
in numerical order in Table 2 and alphabetically by country in Table 3.
In addition to selection of crude oils from the major producing areas
such as Venezuela, offshore Louisiana, Nigeria, Libya, and Saudia
Arabia, crude oils were selected from different producing formations
in the same field and from the same field at different times. The
former is typified by the three Iranian crude oils (Nos. 28,29,30)
from three zones in the offshore Rakhsh field and the latter by samples
-------
from Nigeria (Nos. 35 and 97). Also included in the representative
crude oil samples are two sets of samples from different fields
within the same geologic basin: those from the Interior Salt Dome
Basin of Mississippi (Nos. 15-19) and the South Florida Basin
(Nos. 51-55). This selection of crude oils will allow any identifi-
cation system to be tested under several realistic situations.
16
-------
FIGURE 1 - GEOGRAPHICAL DISTRIBUTION OF CRUDE OIL SOURCES
-------
TABLE 2
H
CO
COUNTRY/
STATE
CRUDE; OIL SOURCES IDENTIFICATION BY NUMBEH
COUNTY OR REGION
FIELD
1
2
3
4
5
6
7
a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
ALASKA
ALASKA
ALASKA
ALASKA
ALASKA
ALASKA
CALIFORNIA
CALIFORNIA
CALIFORNIA
TEXAS
TEXAS
TEXAS
LOUISIANA
LOUISIANA
MISSISSIP
MISSISSIP
MISSISSIP
MISSISSIP
MISSISSIP
ABU DHABI
ALGERIA
ALGfiRIA
ALGERIA
ALGERIA
INDONESIA
IRAN
KUWAIT
IRAN
IRAN
IRAN
KUWAIT
LIBYA
COOK INLET
COOK INLET
COOK INLET
GULF OF ALASKA SHORE
NORTH SLOPE
NORTH SLOPE
OFF SHORE ST BARBARA
LOS ANGELES CO.
ST BARBARA CO.
YOAKUM 6 GAINES CO.
NUECES CO.
BRAZORJA CO.
TIMBALIER OFF SHORE
JACKSON PARISH
WAYNE
WAYNE
CLARK
JONES
JASPER
OFFSHORE
BRUNET
NEUTRAL ZONE
OFF SHORE
OFFSHORE
OFFSHORE
NEUTRAL ZONE
MID. GROUND SHOAL-M&S175
MCARTHUR RIVER-ICARTHUR
TRADING BAY-TYONEK ST175
KATALLA NO. 1
PRUDHOE BAY - WELL NO. 1
W. KUPARUK - W-3
CALIF ST. WELLS 1»2.3,4
WILMINGTON
GATO RIDGE
HASSON
TURKEY CREEK - KENNEDY 8
CHOCOLATE BAYOU GARDINER
BLOCKS 21 AND 28
KELLEY B
E.YEL CREEK-HUMBLE 2 ROE
W.YcL CREEK-AM ££D *1
NANCY-PLACID #1 MENASCO
POOL CREEK-EL ERICKSON
NEST HEIDELBERG-GULF
ABU ARAB A
GASSI TOUIL CUPPER)
NEZLA NORD
RHOURDI EL 8AGUEL
HASSI-MESSAOUND
JERUDCNG FIELD
GACH SARAN
KHAFJI
RAKHSH-WELL ARK 1-ZQNE S
RAKHSH-WELL ARK 1-ZONE A
RAKHSH-fcELL ARK 1-ZONE ft
WAFRA
HOFRA
-------
TABLE 2 (CONTD.)
COUNTRY/
STATE
H
33
34
35
36
37
41
42
43
44
45
46
47
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
LIBYA
LIBYA
NIGERIA
NIGERIA
NORWAY
UN ARAB RP
UN ARAB RP
UN ARAB RP
VENEZUELA
VENEZUELA
VENEZUELA
CANADA
VENEZUELA
FLORIDA
FLORIDA
FLORIDA
FLORIDA
FLORIDA
NORWAY
GABON
GABON
ABU DHABI
ABU DHABI
INDONESIA
ISRAEL
ISRAEL
KUWAIT
QATOR
OATOR
SAUDI ARAB
SAUDI ARAB
SAUDI ARAB
COUNTY OR REGION
CYRENAICA
NORTH SEA
LAKE MARACAIBO
HGNAGAS
QUIKIQUIRE
ALBERTA
ANZOATEQUI
COLLIER
COLLIER
HENDRY
LEE
HENDRY-LEE
NORTH SEA
FIELD
KHUFF
AMOSCAS BEIDE
EBOCHA
AGIOS
EKOFISK-WELL 2-4-5XtDIS4
ALAMEIN-WELL #1X
UMBARKA-WELL 1A
GHROUD FIELD-WELL 1-X
BLOCK 17-WELL SDC-2
MORICHAL
SOUTH SWAN HILLS
OSCUROTt NOTRE
LAKE TRAFFORD FIELD
SUNMLAND FIELD
FELDA FIELD
LEHEIGH FIELD
WEST FELDA FIELD
EKOFISK CRUDE # Z
8ATANGA
TEHtNQUE
BU HASA I
MURBAN-BAB-8U-HASA
PENATANG
HELETZ
KOKHAV
MAGWA-AHMAD1
IDD-EL-SHARGI
MAYDAN-MAHZAN
GHAWAR CHARADH)
QATIF
SAFANIYA
-------
TABLE 2 (CONTD.)
COUNTRY/
STATE
ARGENTINA
COLOMBIA
CUBA
CUBA
CUBA
CALIFORNIA
CALIFORNIA
LOUISIANA
LOUISIANA
LOUISIANA
LOUISIANA
LOUISIANA
NORWAY
VENEZUELA
ECUADOR
COUNTY OR REGION
FIELD
ST. BARBARA
ST. BARBARA
CLAIBORNE PARISH
OFFSHORE
OFFSHORE
OFFSHORE
OFFSHORE
MONAGAS
CERRO DRAGON-CANADQN GR.
PAYOA
BACURANOO-CRUZ VEROE
DOS ESTRELLAS (JATIBQN.1
SANTA MARIE DEL MAR
LGMPAC
SANTA MARIA VALLEY
COLQUITT
DELTA WEST (BLOCK 27»
DELTA WEST (BLOCK 411
EUGENE ISLAND (8LK 276}
SHIP SHOAL (BLOCK 176}
WEST EKOFISK
PIPELINE CRUDE
GULF COMPOSITE
-------
TABLE 3
CRUDE OIL SOURCES - ALPHABETIC BY COUNTRY
ro
CRUDE
OIL
NO.
60
59
20
2
I
3
4
5
6
21
22
24
23
70
8
7
9
75
76
47
71
72
74
73
84
52
51
53
55
54
57
58
COUNTRY/
STATE
ABU DHABI
ABU DHABI
A8U DHASI
ALASKA
ALASKA
ALASKA
ALASKA
ALASKA
ALASKA
ALGERIA
ALGERIA
ALGERIA
ALGERIA
ARGENTINA
CALIFORNIA
CALIFORNIA
CALIFORNIA
CALIFORNIA
CALIFORNIA
CAMADA
COLOMBIA
CUBA
CUBA
CUBA
ECUADOR
FLORIDA
FLORIDA
FLORIDA
FLORIDA
FLORIDA
GABON
GABON
COUNTY OR REGION
OFFSHORE
COOK INLET
COOK INLET
COOK INLET
GULF Of ALASKA SHORE
NORTH SLOPE
NORTH SLOPE
LOS ANGELES CO.
OFF SHORE ST BARBARA
ST BARBARA CO.
ST. BARttARA
ST. BARBARA
ALBERTA
COLLIER
COLLIER
HENDRY
HENDRY-LEE
LEE
FIELD
MURBAN-BAB-BU-HASA
BU HASA I
ABU ARAB A
MCARTHUR RIVER-HCARTHUR
MID. GROUND SHOAL-MGS175
TRADING SAY-TYONEK ST175
KATALLA NO. 1
PRUDHOE BAY - WELL NO. 1
M. KUPARUK - W-3
GASSI TOUIL (UPPER*
NEZLA NORD
HASSI-MESSAQUNU
RHOUROI EL BAGUEL
CERRO DRAGON-CANADON GR.
WILMINGTON
CALIF ST. WELLS 1,2,3»4
GATO RIDGE
LOMPAC
SANTA MARIA VALLEY
SOUTH SWAN HILLS
PAYOA
8ACURANOQ-CRUZ VERDE
SANTA MARIE OEL MAR
DOS ESTRELLAS (JATI80N.)
GULF COMPOSITE
SUNNILANO FIELD
LAKE TRAFFORO FIELD
FELDA FIELD
wEST FELOA FIELD
LEHEIGH FIELD
BATANGA
TEHENQUE
-------
CRUDE
OIL
NO.
61
25
26
30
29
28
63
62
64
27
31
34
32
33
77
14
78
80
79
81
13
17
19
18
16
15
36
35
82
56
37
66
COUNTRY/
STATE
INDONESIA
INDONESIA
IRAN
IRAN
IRAN
IRAN
ISRAEL
ISRAEL
KUWAIT
KUWAIT
KUWAIT
LIBYA
LIBYA
LIBYA
LOUISIANA
LOUISIANA
LOUISIANA
LOUISIANA
LOUISIANA
LOUISIANA
LOUISIANA
WISSISSIP
MISSISSIP
HISSISSIP
MISSISSIP
MISSISSIP
NIGERIA
NIGERIA
NORWAY
NORWAY
NORWAY
QATCW
TAB IE 3 (CONTD.)
COUNTY OR REGION
BRUNET
OFFSHORE
OFFSHORE
OFFSHORE
NEUTRAL ZONE
NBJTRAL ZONE
CYRENAICA
CLAIBGRNE PARISH
JACKSON PARISH
OFFSHORE
OFFSHORE
OFFSHORE
OFFSHORE
TIMBALIER OFF SHORE
CLARK
JASPER
JONES
WAYNE
WAYNE
NORTH SEA
NORTH SEA
FIELD
PENATANG
JERUDGNG FIELD
GACH SARAN
RAKHSH-WELL ARK 1-ZQNE M
RAKHSH-WELL ARK 1-ZGNE A
RAKHSH-WELL ARK 1-ZQNE S
KOKHAV
HELETZ
MAGWA-AHMAQI
KHAFJI
WAFRA
AMOSCAS 3EIDE
HOFRA
KHUFF
COLQUITT
KELLEY 8
DELTA MEST C8LOCK 27)
EUGENE ISLANO IBLK 276J
DELTA WEST (BLOCK 41}
SHIP SHOAL (BLOCK 176)
SLOCKS 21 AND 28
NANCY-PLACID fl HENASCO
WEST HEIDELBERG-GULF
POOL CREEK-EL ERICKSON
W.YEL CREEK-AM E£0 *1
E.YEL CREEK-HUMBLE 2 ROE
AGIOS
EBOCHA
WEST EKOFISK
EKOFISK CRUDE * 2
EKOFISK-WELL 2-4-5XtDIS4
MAY0AN-MAHZAN
-------
TABLE 3 (CONTD.)
CRUDE
OIL
NO.
65
69
68
67
12
11
10
41
42
43
50
44
45
46
COUNTRY/
STATE
OATOR
SAUDI ARAB
SAUDI ARAB
SAUDI ARAB
TEXAS
TEXAS
TEXAS
UN ARAB RP
UN ARAB RP
UN ARAB RP
VENEZUELA
VENEZUELA
VENEZUELA
VENEZUELA
VENEZUELA
COUMTY OR REGION
BRAZGRIA CO.
NUECES CO.
YOAKUM £ GAINSS CO.
ANZGATEQUI
LAKE HARACAIBO
MGNAGAS
CUIRIQUIRE
MONAGAS
FIELD
IOO-EL-SHARGI
SAFANIYA
QATIF
GHAWAR (KARAOH1
CHOCOLATE BAYOU GARDINER
TURKEY CHEEK - KENNEDY B
UASSON
ALAHE IN-WELL #1X
UMBARKA-WELL 1A
GHROUO FIELD-WELL 1-X
OSCUROTE NOTRE
BLOCK 17-WELL SDC-2
PIPELINE CRUDE
-------
SECTION VI
MEASUREMENT OF PARAMETERS
Each of the crude oils and weathered samples were reduced to a "topped"
residue using a vacuum distillation procedure before the identifica-
tion parameters were measured. The residues have an initial boiling
point above 600°F and will be referred to as the "600 F bottoms" or
merely "bottoms". The possibility of thermal cracking is minimized
by limiting the distillation temperature to about 100°C (212°F) and
by reducing the duration of heating. The vacuum distillation or
vacuum flash procedure is described in detail in Appendix A.
The crude oil bottoms are analyzed for total sulfur using ASTM D 1552,
a high temperature combustion procedure that converts the organic
sulfur to sulfur dioxide, which is measured by an iodometric titration.
Total nitrogen is measured by a high temperature oxidative combustion
procedure to produce elemental nitrogen which is determined by thermal
conductivity. The details of the procedure are in Appendix B. The
vanadium and nickel contents are measured on a single portion of the
bottoms, which is decomposed by incineration with sulfur to prevent
loss of volatile metalloporphyrins. The resulting metal sulfides
are analyzed for vanadium and nickel content by atomic absorption
spectrometry in nitrogen oxide-acetylene and oxygen-hydrogen flames,
respectively.
The hydrocarbon gas chromatographic (GLC) profile on each bottoms is
obtained by a procedure equivalent to the dual column technique out-
lined by the Institute of Petroleum. The bottoms are flash vaporized
into a gas-liquid chromatographic column, where the components are
separated by increasing boiling point. The procedural details are
described in Appendix C. Interpretation of the chromatograms is
discussed in Section VII.
25
-------
The carbon isotopic composition, i.e., the proportion of carbon present
as the stable isotope carbon-13 versus the predominant isotope carbon-12
13
and commonly referred to as the carbon isotopic "ratio" or 6 C
(pronounced "del C-13"), is measured by means of an isotope ratio mass
spectrometer. This type of mass spectrometer makes precise isotope
ratio measurements by the simultaneous collection of ions composed of
12
either the major or the minor isotope. A schematic diagram and
description of the instrument used for this study are in Appendix D.
Carbon isotopes are measured on carbon dioxide, which is prepared from
13
the crude oil bottoms by their combustion in an oxygen atmosphere.
The over-all standard deviation of the carbon isotopic composition
measurement is 0.1 per mil or better (see Section VII, Equation 5 for
definition).
The sulfur isotopic composition, i.e., the proportion of sulfur present
as the stable isotope sulfur-34 versus the predominant isotope sulfur-32
(sulfur isotope ratio or 6 S), of the 600 F bottoms is measured on
sulfur dioxide with the isotope ratio mass spectrometer (Appendix D).
The organic sulfur in the bottoms is reduced by hydrogenation to
hydrogen sulfide and recovered quantitatively as solid silver sulfide,
Ag2S. This procedure is an alternate means to measure the total sulfur
content of the bottoms and serves as a check on the more rapid ASTM
procedure. Roasting of the silver sulfide with cupric oxide forms
sulfur dioxide, which is purified before mass spectrometric measure-
ment. These reactions are summarized below:
Hp
Organic Sulfur * > [H SJ
26
-------
This is referred to as the reductive approach to sulfur isotope ratios.
The detailed procedures to accomplish these conversions are described
in Appendices E and F.
An oxidative approach was also used on a few 600 F bottoms. This
embodies combustion of the samples in oxygen to form barium sulfate,
which is converted quantitatively to silver sulfide as follows:
°2 H2°2 Ba*-
Qrganic Sulfur v [SO j * * } [H_SO. j fj _ v BaSO. (3)
A ' ^ HO ' * 4 ; _ *•
2 aq« -
-
BaSO HI v [HJ5] M > CdS -*S_> Ag S (4)
—
Reaction (3) also can serve as an additional procedure to determine
total sulfur but is much less convenient. The procedures to conduct
the oxidative approach are described in Appendices G and H.
The sulfur GLC profiles are obtained in a manner analogous to the
GLC profiles except that a sulfur-selective flame photometric detector
(FPD) is used to determine the approximate boiling point distribution
of the sulfur-containing components of the 600 F bottoms. The exact
operating instructions are in Appendix I.
Further characterization of each crude oil bottoms is accomplished by
measurement of seven additional parameters. The bottoms are fraction-
ated by liquid-solid adsorption chromatography over silica gel into
three portions: saturates, aromaties, and asphaltics. In general,
the material recovery from the silica gel is greater than 85%; samples
giving recovery less than this should be re-run. The detailed
procedure is described in Appendix J. The carbon isotopic composition
is determined on each fraction as previously described (Appendix D).
27
-------
The saturates portion is treated to isolate the n-paraffins (straight-
chain saturated hydrocarbons) by urea adduction. The carbon number
distribution from C.. _ to Cort of the n-paraffins is measured by gas
U Jo
chromatography and is expressed as a normalized weight per cent of
each n-paraffin present. This composition does not represent their
concentration in the crude oil bottoms or the saturate fraction. The
significance in the n-paraffins is in their relative concentrations
to one another and not their absolute concentration in the original
sample. A complete description of these two procedures is in
Appendices K and L.
The data obtained by these measurement procedures and their interpre-
tation with respect to source identification are discussed in
Section VII.
28
-------
SECTION VII
RESULTS AND DISCUSSION
Each of the 80 crude oils selected for study were characterized by the
sequence of steps illustrated schematically in Figure 2 and the
procedures outlined in Section VI.
In any passive tagging approach, the comparison of parameters or
"fingerprints" must be made on a basis that eliminates or, at least,
minimizes the influence of weathering on the measured parameters. In
the earlier stages of weathering most of the changes in the character-
istics of the oil are due to the effects of evaporation and dissolution
of the lighter components of the crude oil. The preparation of distilla-
tion bottoms with an initial boiling point above 518 F is expected to
eliminate the effects of evaporative weathering. 5 Thus, the prepara-
tion of the "600 F bottoms" provides a residue that can serve as a
convenient base for the measurement and comparison of selected
parameters. The 600 F initial boiling point used for this work is
intermediate between that chosen by the Western Oil and Gas Association
(540+F bottoms)^ and the Institute of Petroleum (649+F bottoms).
The first operation in the sequence of steps outlined in Figure 2 is
the preparation of the "600 F bottoms" or merely "bottoms". Since
thermal cracking of crude oils can occur at temperatures as low as
150°C and this cracking could result in a modification of the
indigenous properties of the oil, it is necessary to limit the tempera-
ture and duration of the distillation. The method chosen utilizes a
vacuum distillation, which is more nearly a vacuum flask, in which the
operation is carried out at 101 C and 0.15 nrca Hg pressure.
29
-------
CRUDE OIL
ISOTHERMAL
DISTILLATION 101°C AT 0.15mm
I
RESIDUE, WT. %
DISTILLATE
SILICA GEL 0.25gOIL
MEASURE
GLC PROFILE
SULFUR, NITROGEN
VANADIUM, NICKEL
CARBON ISOTOPE RATIO
SULFUR ISOTOPE RATIO
SULFUR GLC FINGERPRINT
SATURATES, WT. %
AROMATICS, WT. % ASPHALTICS, WT. %
UREA ADDUCTION
MEASURE
CARBON ISOTOPE RATIO
n-PARAFFINS
_MEASURE
CARBON NUMBER DISTRIBUTION
CALCULATE OEP
FIGURE 2 - SCHEMATIC DIAGRAM OF OIL POLLUTION
SOURCE IDENTIFICATION SYSTEM
30
-------
The amount of 600 F bottoms recovered from the atmospheric equilibrated
crude oils (stock tank liquids) varies from 30 to 99 weight per cent
of the original charge. The values obtained for each individual crude
oil are tabulated in Table 4» These values represent approximately
the per cent of a crude oil spill that will persist after it has
weathered a few days. Evaporation of the light ends will represent
the majority of the loss. The magnitude of the clean-up operation in
terms of volume can also be estimated. Except for the comparison of
crude oils for their economic value this parameter serves little value
for identification of a weathered unknown.
The initial portion of this study had as its objective the demonstration
of the identification feasibility based on the carbon isotopic composi-
tion plus the five parameters of sulfur, nitrogen, vanadium, and nickel
contents and the hydrocarbon gas chromatographic (GLC) profile. The
values'obtained for these six parameters plus the values of the sulfur
isotopic composition on the eighty 600 F bottoms are presented in
Table 4. The diagnostic significance of these parameters is disclosed
by an examination of the variation in each parameter. The variations
observed for the four elemental parameters as well as the repeatability
for each measurement are presented in Table 5. Conclusions regarding
parameter values are based on repeatability.
The sulfur values fall into three ranges that allow each of the 600 F
bottoms to be classified as high, medium or low in sulfur. High sulfur
bottoms are characteristic of crude oils from California, Mississippi,
Florida, Venezuela, and the Middle East, and low sulfur bottoms of
crude oils from Alaska, Louisiana, the North Sea and Nigeria.
Exceptions to these generalizations have been observed.
31
-------
TABLE 4
VALUES OF IDENTIFICATION PARAMETERS ON CRUDE OIL BOTTOMS
CRUDE
OIL
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
BOTTOMS
WT. %
56.5
56.7
45.0
45.7
72.9
66.6
29.7
85.2
88.3
59.1
49.2
61.6
70.6
33.0
82.6
76.9
53.4
£4.8
63.6
73.4
58.6
54.8
60.0
48.7
38.0
76.0
72.0
63.0
63.7
69.7
31.2
43.8
59.4
67.9
SULFUR
0.07
0.11
<0.05
<0.05
1.30
0.98
0.07
1.70
6.00
2.90
0.26
0.11
0.29
<0.05
4.91
3.91
1.65
4,08
2.97
2.01
<0.05
0.15
0.45
0.19
0.12
2.10
3.80
2.34
2.39
3.90
3.77
0.48
0.54
1.10
NITROGEN
0.16
0.22
0.11
0.11
0.26
0.21
0.23
0.53
0.73
0.18
0.08
0.06
0.13
0.02
0.11
0.22
0.06
0.13
0.10
0.31
0.06
0.05
0.09
0.07
0.08
0.34
0.20
0.07
0.09
0.23
0.13
0.17
0.25
0.32
VANADIUM
0.5
0,6
<0.2
<0.1
18.8
23,7
0.4
49.1
255,0
11.2
1.1
<0.1
1.0
<0.1
15.6
14.4
<0.1
19.3
12.4
72.3
<0.2
0.5
0.5
<0.2
<0,2
112.0
73.3
4.7
10.0
83,3
47.9
1.0
5.6
10.0
NICKEL
0.5
2.1
1.3
2.0
12.3
15.2
4.6
72.0
123.0
7.5
0.8
1.9
2.3
<0.1
14.1
15,7
2.0
14.1
7.1
26.3
0.1
0.1
0.1
0.1
2.7
42.9
23.8
9.5
2.8
24.8
15.0
9.1
24.5
47.0
DEL C13
-30.2
^30.2
-29.5
-24.9
-29.3
-30.4
-22,4
-22.6
-22.8
-28.7
-26.9
-26.8
-26.6
-27.5
-25.3
-25.5
-24,9
-25.1
-25.0
-26.5
-27.9
-28.1
-28.5
-29.3
-27.8
-26.3
-27.2
-25.9
-26.6
-26.6
-26.7
-27.2
-27.8
-28.3
DEL S34
SULFUR
ISOTOPE
-3.1
-5.4
6.6
17.6
-4.3
-1.8
-7.9
-7.1
-8.2
-9.4
-6.0
-9.6
0.3
2.3
-8.7
-5.8
-5.4
-9.6
2.3
0.0
1.3
GLC
PROFILE
5
5
3
4
6
9
2
2
B
1
B
k-
5
3
2
1
3
2
2
5
0
0
0
1
2
1
0
1
1
5
0
0
4
4
-------
TABLE 4 (CONTD.)
CRUDE
OIL
NO.
35
36
37
41
42
43
44
45
46
47
50
51
52
*»3
54
55
56
57
58
59
60
61
62
63
64
65
66
t>7
68
o9
70
71
72
73
BOTTOMS
WT. t
SULFUR
MITROGEN
VANADIUM
NICKEL
DEL C13
48.2
49,6
52.8
70.1
45.2
67.7
60.3
98.2
79.7
51.6
49,1
71,7
76,7
76,6
78.7
72.8
60,8
47.5
73.0
54.5
54.1
81.3
70.1
73,3
65.9
62.0
53,9
68.2
70.7
76.1
89.3
63.9
62.0
99.2
0.11
0,10
0,14
1,40
0,05
1.37
1,70
4,40
1,21
0,18
0.17
5.18
4.39
4,65
4.83
4.07
0.16
0.62
0.89
1,33
1.37
0.12
2.30
2.13
3.03
2.86
2.11
2,95
3,62
3.82
0.16
1.25
1.20
1.84
0.12
0.02
0.03
0.12
0.08
0.21
0.17
0.50
0.23
0.16
0,03
0.14
0.24
0.20
0.24
0.16
0.16
0.10
0.16
0.07
0,05
0.14
0.18
0.18
0.19
0.12
0.09
0.14
0.13
0.18
0.36
0.28
0.20
0.12
0.3
0.6
0.4
18.9
-------
TABIE 4 (CONTD.)
CRUDE
OIL
NO.
74
75
76
77
78
79
P.0
81
82
B3
84
BOTTOMS
87.2
78.2
80.6
38.8
79.2
64.9
62.4
57.6
43.7
94.5
78.4
SULFUR
3.52
5.69
5.59
0.17
0.48
0.25
0.89
0.14
0..06
2.40
1.20
NITROGEN
VANADIUM
NICKEL
DEL C13
0.24
0.67
0.73
0.02
0.13
0.11
0.08
0.13
0.05
0..4&
0.28
15.1
246.0
265.0
<0.3
2.4
0.7
3.4
<0.3
0.4
232.0
73.0
17.4
100.0
128.0
0.4
5.9
0.5
3.0
1.1
0.3
70.0
29.3
-25.9
-22.5
-22.4
-25.6
-27.1
-26.7
-26.6
-26.5
-29.1
-27.0
-25.8
DEL S34
SULFUR
ISOTOPE
-4.0
14.4
17.8
-4.1
-9.7
-9.7
-6.0
0.0
7~6
-1.6
GLC
PROFILE
B
0
B
6
7
4
0
2
B
6
-------
TABLE 5
RANGE OF VALUES FOR SULFUR,
NITROGEN, VANADIUM AND NICKEL CONTENTS
OF 600°F RESIDUES
VJl
SULFUR,WT.%
HIGH
MEDIUM
LOW
NITROGEN, WT.%
VANADIUM, WPPM
NICKEL, WPPM
VALUE
2-6
1-2
0.05-1
.24-J3
.15-.24
.03-. 15
112-466
24-92
2-19
60-128
20-47
2-17
NO. OF
CRUDE
OILS
31
15
34
14
25
40
6
21
21
32
9
14
38
19
REPEATABILITY
0.1 6 TO 0.24
0.10
0.05 TO 0.07
• 0.06
5 @ 68 WPPM
2 @ 14 WPPM
0.3 @. 3-2 WPPM
2 @ 2-19 WPPM
-------
Nitrogen content, which covers a smaller range of values than sulfur
content, serves to distinguish the few typically high nitrogen sources
from the remainder. The California, South American and some Middle
East bottoms have the highest nitrogen content. In general, the higher
nitrogen contents are found in crude oil bottoms containing the higher
vanadium and nickel contents.
The trace metals, vanadium and nickel, vary in amounts among the crude
"1 / -I rj
oils of the world. ' Their presence at high levels deactivates
catalysts for hydrodesulfurization, an important refining operation.
These two trace metals and the ratio of their concentrations have been
suggested by others for pollution source identification. ' ' ' Our
data support these suggestions. However, we agree with Kreider that
ratios are useful only if absolute values for individual elements are
also reported. The range of contents for these two elements are
sufficient to be significant identification parameters.
Hydrocarbon GLC Profile
On the basis of the variation in the hydrocarbon GLC profile among the
eighty crude oils conclusions concerning the boiling range of the
bottoms, the presence or absence of n-paraffins, and the presence of
characteristic components can be made. Each profile consists of a
number of peaks, which represent the n-paraffins, standing out above a
broad envelope of incompletely separated components. Typical profiles
are illustrated in Figure 3. The ability to distinguish between
bottoms samples that exhibit profiles as different as 3A and 3F is
obvious. In order to numerically classify the profiles for easy com-
parison a system based on the relative n-paraffin peak heights between
n~C20 anc* n~So WaS deve-L°Ped« A number is obtained by connecting the
peaks of these two n-paraffin components by a straight line and counting
the number of peaks from C through C q inclusive that touch or extend
36
-------
above the line. Each profile is assigned this number for facile
recognition. For example, in Figure 3A through 3D the profile type
is assigned on this basis. The location of the n-paraffin peaks and
their corresponding atmospheric boiling points are giv§n on the
abscissa. If either or both of the two reference peaks do not appear9
the profile is classified as broad and is assigned the letter B.
The n-paraffins are either absent or too small to recognize in the
latter cases (see Figure JE and 3F). The frequency distribution of
classification type among the sources is shown in Figure 4*
37
-------
B. MISSISSIPPI EAST YELLOW CREEK
A. ALGERIA MASS I-MESSAOUND
TYPE 1 PROFILE
SATURATES 64%
TYPE 2 PROFILE
SATURATES 26%
841
755
651
51 9
841
755
C. LOUISIANA COLQUITT
TYPE 5 PROFILE
SATURATES 90%
841
755
651
519 °F
D. UNITED ARAB REPUBLIC UMBARKA
TYPE 9 PROFILE
SATURATES 81!
'20
n-c
i 5
841
755
651
519 °F
E. TEXAS TURKEY CREEK
TYPE B PROFILE
SATURATES 50%
F. VENEZUELA QUIRIQUIRE
TYPE B PROFILE
SATURATES 39%
841
651
651
FIGURE 3 - TYPICAL GAS CHROMATOGRAPHIC PROFILES OF 600+F RESIDUES
-------
11
(/)
-J
0
UJ
Q
D
QC
O
U.
O
nr
NUMBEi
9
10
7777
1 1 !
3
0
10
8
01 23456789B
FIGURE 4 - FREQUENCY DISTRIBUTION OF HYDROCARBON GLC PROFILES BY
CLASSIFICATION TYPE
The classification type is repeatable within - one type for individual
bottoms. The profile serves as an initial screening parameter to
eliminate some potential pollution sources before proceeding to
measure the other identification parameters.
The magnitude of peaks between those for n-paraffins C../ to (Lg are
noted to vary appreciably in height with crude oil source and can be
useful to distinguish between similar profile types. Occasionally
these peaks are larger than the n-paraffin peaks, for example, oil
number 58 from Gabon. The other crude oil from Gabon, No. 57, also
gave two large non-n-paraffin peaks between n~C-,^ and n-C->g and
between n-Cn- and n-Cin. The offshore Louisiana Ship Shoal crude oil,
18 IV i
No. 81, exhibited similar behavior, which has been observed by others.
-------
Carbon Isotopic Composition
The carbon isotopic composition is expressed by the following equation:
sample - 1
13C/12C
V V ...
standard
lo3 (5)
13
where 6 Cpnn is the parts per thousand difference expressed in units
-JO -tn
of per mil, °/oo> Detween the JC/ C ratio for the sample and that of
an arbitrary standard. All data in this study are referred to the
"PDB" standard, a carbonate fossil Belemni.te.lla. Americana of the Upper
1 3
Cretaceous Peedee Formation of South Carolina. The nominal natural
abundance of these two stable isotopes is earbon-13, 1.1 per cent,
and carbon-12, 98.9 per cent. The variations in the carbon isotopic
composition occur as a consequence of environmental differences during
the genesis of petroleum. The carbon isotopic composition is a
valuable identification parameter because it undergoes little post-
genesis alteration and, therefore, is characteristic for each crude
oil. This fact also limits to a degree the usefulness of this
parameter.
Erdman and Morris demonstrated that crude oils from eighteen reservoirs
in four Mississippi fields have the same carbon isotopic composition
with a mean deviation of 0.1 per mil. This suggests a common source.
This fact and the variation in the elemental composition among the
five Mississippi crude oils, Nos. 15-19, Table 4, demonstrates the
postulate of a common genesis source with subsequent migration into
the present reservoirs where the oils underwent varying degrees of
post-migration change. The elemental compositions and the silica gel
fraction percentages are the only differentiating parameters for these
-------
five crude oils. Identification of an unknown as being one from among
these five crude oil sources provides a rigorous test of any matching
scheme.
The carbon isotopic compositions of the 600+F bottoms that comprise
our crude oil sources range from -22 to -30 per mil (Table 4). The
frequency distribution of this isotope ratio among the bottoms
(Figure 5) indicates that 70% of the values fall within the 3 per mil
range from 25.0 to 27.9. This distribution and range is similar to
19
that found by Eckelmann for 128 crude oils Devonian and younger.
The carbon isotopic composition by itself is not normally sufficient
to provide unique identification, but when applied in combination
with the four elemental composition parameters, it is a diagnostic
parameter that allows two otherwise equivalent crude oil sources to
be distinguished.
The maximum observed differences in carbon isotopic ratios is between
crude oils from Alaska, numbers 1, 2, 3, 5 and 6 and those from
Southern California, numbers 7, 8, 9, 75 and 76. The Alaskan crude
13
oils are isotopically light, i.e., most negative 6 C values, compared
13
to the California crude oils, which have the least negative 6 C
values. The Alaskan values vary from -30.4 to -29.3 per mil and the
California values from -22.4 to -22.8 per mil. For these sources the
carbon isotopic ratio alone is sufficient to distinguish these two
sources; this is fortunate for pollution control in the North Pacific
Ocean. The Alaskan sources are distinguished from one another by the
significant difference in the sulfur and trace metal contents, i.e.,
the Cook Inlet crude oils have values an order of magnitude or more
lower than the North Slope sources. Further applications of this
parameter are discussed in Section IX.
41
-------
A. CARBON ISOTOPE
1
13
26
18
-22 -23 -24 -25 -26 -27 -28 -29 -30
PDB
B. SULFUR ISOTOPE
16 -12 -8 -4 0 +4 +8 +12 +16 +20
FIGURE 5 - DISTRIBUTION OF CARBON AND SULFUR ISOTOPIC COMPOSITIONS
-------
Sulfur Isotopic Composition
Sulfur in petroleum possesses a wide variation in isotopic composi-
20 2.1
tion with different crude oil sources. ' The purpose of this phase
of the project was to examine the feasibility of utilizing the sulfur-
34/sulfur~32 isotope ratio of crude oil bottoms as one of the parameters
in a pollution source identification system. With this purpose in
mind, the following work was performed: (l) techniques necessary to
obtain the data were developed, (2) the data were obtained, (3) the
usefulness of the sulfur isotope ratio as an identification parameter
was demonstrated, and (4) the effect of weathering oil samples under
simulated oil spill conditions on the sulfur isotope data was examined.
In the measurement of sulfur isotopic ratios, a comparison is made of
the ^%/^ S ratio of a sample with the -^S/^S ratio of a standard of
known isotopic composition. The results are expressed in terms of the
6 S value as defined below:
,«"•••
%o = (
sample - 1
standard
103 (6)
6 S is a dimensionless quantity and is usually described in terms of
a "per mil" or "del" value. The standard used as reference is the
troilite phase from the Canon Diablo meterorite (6 Srm = 0) eontain-
O, o, op G1M
ing about k.3% J^S or having a J*Sf*$ ratio of 0.0450045. Samples
containing more -^S than the standard will have positive "del" values.
In practice, most laboratories use a laboratory standard of known
composition and then calculate the 6 ^DM value.
43
-------
The sulfur Isotopic ratios of 52 crude oil bottoms measured using the
reductive hydrogenation approach are presented in Tables 4 and 6.
With present techniques these measurements are limited to bottoms con-
taining greater than 0.25 weight per cent sulfur. The 6 ^ values
range from +19.0 to -15.0 per mil and the distribution of values is
shown in Figure 5B (Page 42). The spread and relatively even
distribution of 6 T3 values make the sulfur isotopic ratios an
useful identification parameter.
The geographic distribution of the crude oils and their tentative
sulfur isotopic composition are shown in Figure 6. The value of this
parameter for the identification of an oil pollution source is illustrat-
ed by the following observations: A pollutant in the North Pacific
Ocean off the Alaskan-Canadian coast could be distinguished as being
from the North Slope fields (Alaska) or California based on 60
values of -3 to -5.4 and +6.6 to +17.8 per mil, respectively; and oils
from offshore Louisiana (numbers JB, 79, 80), Venezuela (numbers 44,
45, 46), and Florida (numbers 51-55) could be distinguished based on
63^S values of -6 to -10, +6.6 to +8.3, and -13.2 to -14.4 per mil,
respectively. A similar situation exists in the Middle East in that
the Israeli crude oils (numbers 62, 63) are isotopically much heavier
than all other area sources measured; this difference ranges from 11.6
to 29 = 5 per mil. The majority of crude oils around the Persian Gulf
are isotopically lighter than the standard, i.e., 6^S values negative.
These observations are in marked contrast to the results predicted by
a previous study. However, another study concluded that sulfur
isotopic ratios could be very valuable as an evidence point in
1 ft
identifying pollution sources.
44
-------
TABLE 6
SULFUR ISOTOPE COMPOSITION OF 600+F BOTTOMS
BY THE PROPOSED REDUCTIVE (HYDROGENATION) PROCEDURE
EPA Crude
Oil Number
5
6
8
9
10
11
15
16
17
18
19
20
26
27
28
30
31
32
33
34
41
43
44
45
46
51
52
53
54
55
Measured
534S. o/gg
- 3.4,-2.8
- 5.9,-4.9,-5.5
7.2,5.9
19.1,15.5,18.1
- 4.3
- 2.2,-1.3
- 7.1,-7.9
- 8.8,-9.8,-5.9
-10.3,-8.5
- 6.1,-5.9
- 9.6
0.3
2.9,1-7
- 8.7
- 5.8
- 4.4,-7.3,-4.6
-10.0,-10.0,-8.8
2.2,2.3
0.0
1.3
-11.2
2.5
6.8,6.8,6.2
9.4,7.1
8.8,8.9,8.3
-13.2
-14.6,-14.0
-13.3,-15.0
-14.4
-12.7,-14.2
Average
63%. O/
- 3.1
- 5.4
6.6
17.6
- 4.3
1.8
7,5
8.2
9.4
0
- 6
- 9.6
0.3
2.3
- 8.7
- 5.8
- 5.4
- 9.6
2.3
0.0
1.3
-11.2
2.5
6.6
8.3
8.7
Source Description
-13.2
-14.3
-14.2
-14.4
-13.5
Prudhoe Bay, Alaska
W. Kaparuk, Alaska
Wilmington, California
Gate Ridge, California
Wasson, Texas
Turkey Creek, Texas
E. Yellow Creek, Mississippi
W. Yellow Creek, Mississippi
Nancy-Placid, Mississippi
Pool Creek, Mississippi
West Heidelberg, Mississippi
Abu Arab A, Abu Dhabi
Gach Saran, Iran
Khafji, Kuwait (Neutral Zone)
Offshore Iran
Offshore Iran
Wafra, Kuwait (Neutral Zone)
Hofra, Libya
Khuff, Libya
Amescas Beide, Libya
Alamein, United Arab Republic
Ghroud, United Arab Republic
Block 17, Maracaibo, Venezuela
Monagas, Venezuela
Quiriquire, Venezuela
Lake Trafford, Florida
Sunniland, Florida
Felda, Florida
Leheigh, Florida
West Felda, Florida
45
-------
TABLE 6 (Cont'd.)
SULFUR ISOTOPE COMPOSITION OF 600^ BOTTOMS
BY THE PROPOSED REDUCTIVE (HYDROGENATION) PROCEDURE
EPA Crude
Oil Number
57
58
59
60
62
63
64
65
66
67
68
69
71
74
75
76
77
78
79
80
83
84
Measured
634S. %o
5.0
2.6,4.3
- 4.6
- 3.7
13.4,14.4
14.6
-10.0
-15. 7, -14.0
- 8.3
- 2,5,-2.6,-3.0
- 4.1,-5.3
- 9.9,-8.7,-10.1
5.9
- 3.9,-4.0
14.4
17.8
- 4.4,-3.8
- 9-7,-9.1,-10.2
- 9.7
- 6.0
7.6
- 1.6
Average
634S. o/oo
5.0
3.5
- 4.6
- 3.7
13.9
14.6
-10.0
-14.9
- 8.3
- 2.7
- 4.7
- 9.6
5.9
- 4.0
14.4
17.8
- 4.1
- 9.7
- 9.7
- 6.0
7.6
- 1.6
Source Description
Batanga, Gabon
Tehenque, Gabon
Bu Hasa I, Abu Dhabi
Murban-Bab-Bu-Hasa, Abu Dhabi
Helet z, Israel
Kokhav, Israel
Magwa-Ahmadi, Kuwait
Offshore, Qatar
Offshore, Qatar
Ghawar, Saudi Arabia
Qatif, Saudi Arabia
Safaniya, Saudi Arabia
Payoa, Colombia
Santa Marie del Mar, Cuba
Lompac, California
St. Maria Valley, California
Colquitt, Louisiana
Delta West, Louisiana (Offshore)
Delta West, Louisiana (Offshore)
Eugene Island, Louisiana (Offshore)
Monagas, Venezuela
Ecuador
46
-------
Noixisodwoo oidoiosi Hnjins ONV
saonnos no aonno 10 NoiinaiHisia
- 9
a" i
—1.2.3
5, 6
— 3 01 , -5.4
7 , 8, 9, 75, 76
6.6, 17.6, 14.4, 17.8
13, 78, 79, 80, 81
-9.7, -9.7,-6.0
15, 16, 17, 18, 19
-7.1 , -8.2, -9.4 ,-6.0,-9.6
ff '
27 , 28, 29 , 30
'-8.7 ,-5.8 ,-5.4
21,22,
23, 24
© -4.0
72,73,74
51 , 52, 53, 54, 55
-13.2, -I 4 .3, -14.2, -14.4, -13.5
20, 59, 60
0 3 . -4 .6 , -3 .7
32, 33, 34
2 .3 , 0 .0 , 1 ,3
67,68,69
-2.7,-4.7,-9.6
NUMBERS REFER TO CRUDE OILS
IDENTIFIED IN TABLE 2.
-------
To be used as an identification parameter, the sulfur isotopic ratio
should be independent of the other measured parameters. Plots of the
6^^S values against the 6^C values, the sulfur and vanadium concentra-
tion were made and the 6 ^S values appeared to be independent of these
other parameters.
An analysis of the data obtained in cases where duplicate measurements
were made indicates that in most cases the data has a precision of -1
per mil which is sufficient for these exploratory studies. Using
refined techniques it should be possible to improve the precision to
- 0.2 per mil since this precision has been attained in several
instances. We have determined that in several cases where the precision
hoped for was not attained, the SO samples contained traces of im-
purities that probably interfered with the isotope ratio measurement.
Experiments have indicated that a modification of the combustion
technique results in improved purity of the SO- samples.
A direct comparison between the sulfur isotopic ratios obtained in
this study and those measured in published studies is not possible
because of the different oil sources and measurement techniques used
in this study. Previous workers did not make their measurements on
the 600 F bottoms but it is expected that the sulfur isotopic ratio of
22
the total crude would be very similar. The general agreement of the
data in Table 7 indicates that our results are equivalent to those
found by others and supports both our choice of measurement procedures
and the application of sulfur isotopes as an identification parameter.
The sulfur isotopic ratios measured by other workers are summarized in
Appendix M in order to provide a convenient, readily-accessible source
of data for crude oils. These data can be used to evaluate the over-
all identification potential of this parameter for sources and circum-
stances outside the scope of this study.
-------
TABIE 7
COMPARISON OF SULFUR ISOTOPIC COMPOSITION
WITH LITERATURE VALUES
Oil Source
California, Wilmington
Venezuela, lake Maracaibo
Venezuela, Lake Maracaibo
Kuwait, Wafra
Saudi Arabia, Ghawar (Haradh)
Saudi Arabia, Safaniya
This Work
EPA No.
8
44
46
31
67
69
11
1.7
1.7
1.2
3.8
3.0
3.8
£*S
6.6
6.6
8.7
-9.6
-2.7
-9.6
Previous Work
1.5
6.8
6.1
23
18
3.9 -9.8 23
2.1 -1.9 23
3.0 -8.5 23
The application of 6 & values to distinguish between pairs of crude
oils in which the other identification parameters are similar is
described for oil numbers 30 and 69 and for numbers 65 and 67 by the
data in Table 8. There is a clear distinction between the oils only
when the sulfur isotopic data are included.
TABLE 8
APPLICATION OF SULFUR ISOTOPIC COMPOSITION
EPA Crude Oil Number
Oil Source
613C
6 CPDB
Sulfur (Wt. %)
Nitrogen (Wt. %)
Vanadium (W ppm)
Nickel (W ppm)
Vanadium/Nickel Ratio
Sulfur/Nitrogen Ratio
6 ^S (Average)
vs
69
vs
Iran Saudi Arabia Qatar Saudi Arabia
-26.6
3.90
0.23
83.30
24.80
3.36
16.96
-26.6
3.82
0.18
63.40
23.20
2.73
21.22
-26.6
2.86
0.12
23.90
8.44
2.83
23.83
-26.6
2.95
0.14
25.60
6.97
3.67
21.07
- 5.4
- 9.6
-14.9
- 2.7
49
-------
In experiments to determine what the effect of weathering is on the
sulfur isotopic ratios, the 6^S values were obtained from several of
the weathered samples from oils 83 and 84. (Data presented in
Section IX.) There was no trend observed in the 6 S values versus
length of time weathered and it is evident the 6 S values did not
change under the conditions of our weathering experiments.
The 63/*S values obtained for five oils using the oxidative route and
the reductive route are compared in Table 9. The differences observed
are outside the expected range of repeatability and in each case the
value from the oxidative route is more positive. A systematic error
is suspected in one of the techniques used. There is no a priori
reason the results should not agree but the source of disagreement
was not isolated.
TABLE 9
COMPARISON OF SULFUR ISOTOPE RATIOS BY THE
OXIDATIVE AND REDUCTIVE METHODS
Crude Oil Number 634S (Oxidative) 534S (Reductive)
9 22.4 o/oo 17.6 %o
15 - 5.7 - 7.1
17 - 5.1 - 9.4
68 - 2.0 - 4.7
* 74 - 2.2 - 4.0
BKE Elemental Sulfur -11.2
The expected 6 S value, regardless of method, is -11.8 °/oo.
50
-------
The measured values of these preceeding selected parameters on the
600 F bottoms of crude oils from the major producing areas of the
world cover a sufficient range and distribution to offer strong
evidence for oil pollution source identification. The techniques
to apply these variations to the identification problem are discussed
in Section VIII.
Sulfur Gas Chromatographic Profile
Sulfur is the most abundant element other than carbon and hydrogen
present in crude oils. The total sulfur content varies from less
than 0.05 to greater than 6.0 weight per cent and is a useful identifi-
cation parameter. This sulfur is present as a wide boiling range of
organic sulfur compounds; up to 200 individual sulfur compounds have
f\ t
been identified by U.S. Bureau of Mines chemists. These include
thiols (mercaptans), sulfides, disulfides and thiophenes. The
development of a sulfur-selective gas chromatographic detector led
Adlard to propose the measurement of a sulfur gas chromatographic
(GLC) profile analogous to the hydrocarbon GLC profile as an aid to
pollution source identification. The potential value of this profile
was demonstrated in the initial publication. The variation in the
sulfur GLC profile was investigated on our collection of crude oils
with the objective to evaluate its general usefulness.
25
The sulfur-selective detector is based on flame emission photometry.
The radiant energy emitted at 394 nm when the gas chromatograph
effluent is burned in a hydrogen flame arises predominantly from the
sulfur atoms present. The background signal from hydrocarbons is
relatively small. The response of this detector gave a reproducible
and unique profile for each crude oil sample. Forty-eight representa-
tive 600+F bottoms that contained more than 1% sulfur were examined.
51
-------
This was the lowest level of total sulfur, for which a sulfur GLC
profile could be obtained without distortion from the background
noise on our equipment.
The sulfur GLC profiles ("fingerprints") are classified into six types,
three of which are shown in Figure 7, Type A, which is distinguished
by two broad partially resolved peaks of equal magnitude, is character-
istic for 6 crude oil sources, Nos. 15, 16, 20, 30, 43 and 69. Type B,
which consists of 8 different sources, Nos. 1?, 28, 29, 46, 59, 60, 66
and 74, is characterized by a sharp rise to a large, broad partially-
resolved center peak followed by a smaller broad peak. The Type C
profile is found for 13 crude oil sources and contains several re-
solved peaks preceeding the Type B peaks. These 13 Type C profiles
are for Nos. 10, 18, 19, 26, 27, 31, 41, 64, 65, 67, 68, 71 and 80.
Eight of the 600 F bottoms show only a broad background envelope with
no distinctive peaks or highlights, Type D. The last two types con-
sist of 5 profiles with a broad large peak and 8 profiles with a large
broad peak with two sharp characteristic peaks. The latter 8 include
5 South Florida Basin crude oils and 3 California oils, which are low
in saturates and high in sulfur.
As with the hydrocarbon GLC profile, this sulfur profile exhibits
sufficient variety to serve as an initial screening parameter for high
and medium level sulfur residues. The attractive feature of this
parameter is the fact that the profile is unaffected by weathering,
which apparently destroys n-paraffins by biological degradation.
This alteration by weathering reduces the usefulness of the hydro-
carbon GLC profile as a screening parameter.
52
-------
A. MISSISSIPPI EAST YELLOW CREEK,
EPA NO. 15
TYPE A
MISSISSIPPI NANCY PLACID
EPA NO. 17
TYPE B
C.
KUWAIT MAGWA-AHMADI
E PA NO. 64
TYPE C
MINUTES
FIGURE 7 - TYPICAL GAS CHROMATOGRAPHIC SULFUR FINGERPRINTS
53
-------
Silica Gel Fractionations
Where sufficient differences exist in the values for the previous
parameters, unequivocal identifications are feasible. In other cases,
the diagnostic power of these measured parameters fails to permit any-
definitive conclusions as to source. Further characterization of the
600+F bottoms was made by fractionation of each into three portions:
saturates, aromatics, and asphaltics. The amount in each portion was
measured and additional measurements were then made on each fraction.
The data is summarized in Table 10.
The distribution of the 600 F bottoms among the three silica gel
fractions (Table 10) varies from 19 to 83 weight per cent for the
saturates, from 6 to 6? weight per cent for the aromatics, and from
0.3 to 31.4 weight per cent for the asphaltics. The per cent recovery
varied from 73 to 102$. Although experience indicates that the
recovery should exceed &$%, there will be sources which may contain
polar compounds such as naphthenic and tar acids that are retained on
the silica gel. High boiling waxes (saturates) are also retained and
are recovered from heated columns. A low recovery indicates conclusions
should be carefully drawn and the sample should be examined more fully.
Carbon Isotopic Composition of Silica Gel Fractions
The carbon isotopic composition was measured on each fraction from the
silica gel separations. The carbon isotopic composition of the 600+F
bottoms is compared with that of each fraction in Table 10. The
saturate fraction is normally equal to or isotopically lighter, i.e.,
a more negative number, than the bottoms by about one per mil. The
aromatic and asphaltic fractions are isotopically heavier, i.e., less
negative values, than the bottoms.
54
-------
TABLE 10
VALUES FOR CARBON ISOTOPIC COMPOSITION ON CRUDE OIL BOTTOMS AND THEIR SILICA GEL FRACTIONS
CRUGE
OIL
NO.
1
2
:-4
^
5
6
7
8
•^
10
11
12
13
14
1 5
16
17
18
1 "-,
20
21
22
23
24
25
26
27
28
29
30
31
•-j.,2
33
34
TOTAL CisUCfc
DEL C-13
-30.2
-30.2
-79.5
-24.9
-29,3
-30.4
-22 .4
-22.6
-22. a
-23.7
-26.9
-26.8
— 26.6
-^7.5
-25.3
-25.5
-24.9
-25.1
-25.0
-26.5
-27.9
-28.1
-28.5
-29.3
-27.8
-26,3
-27.2
-25.9
-26.6
-26.6
-26.7
-27.2
-27. ti
-28.3
SATURATES
if. #
57.0
57.3
67.3
67.2
40.9
50.1
60. U
27.9
13.2
40,3
50, 3
67.7
54.2
93.1
26,2
36.8
62.2
32,7
52.7
3S.3
77.2
81.0
54.6
o4.4
61.9
39.4
33.0
52.5
48.3
36.7
38.7
o2.2
60.8
51.0
DEL c-:
-31.7
-31.8
-31.0
-26.0
-30.3
-31.3
-23.2
-24.6
-23,6
-30.1
-28.3
-28.3
-27.9
-29.1
-25.0
-25.5
-25.1
-25.2
-25.0
— 26,4
-28.5
-28.3
-28.7
-29.4
-29.2
-27.5
-27.9
-26,1
-26.8
-27.1
-27.8
-28.3
-29.1
-29. 1
AROMATICS
«T. %
37.3
.38.6
29.5
29.0
51.5
42.2
33.1
57.7
62.5
49.9
41.2
28.9
38.7
6.3
64.1
53.6
35.7
5a.5
39.5
53.9
20.4
17.2
39.3
32.1
34.4
49.8
57.6
42.3
45.9
55.2
54.6
32.5
33.6
40.7
OEL C-
-29.3
-30,5
-29.3
-24.7
-29.2
-30.6
-22.8
-23,4
-22,5
-29.1
-26.9
-27.3
-27,3
—
-25.1
-25.1
-24.2
-25.1
-25.0
-25.1
-27.8
-27,6
-27.9
-28.5
-27.9
-26.8
-26.9
-25.6
-26.2
-26.8
-27.2
-27.1
-27.6
-27.6
ASPHALTICS
rr. %
5.7
4.1
3.1
3.8
7,6
7.7
6.9
14.4
24.3
9. a
8.5
3.4
7,0
0.7
9.7
9.6
2.1
8.8
7.8
9.3
2.4
1.9
6.1
3.6
3.7
10.8
9.5
5.1
5.7
8.1
6.7
5.3
5.6
8.3
DEL C-13
-29.6
-29.5
-29.2
-24.2
-29.3
-29.9
-21.6
-22.7
-22.3
-28.3
-27.0
-27.2
-27.2
—
-23.5
-25.2
-24.3
-25.6
-24.9
-25.0
-26.7
-28.1
-23.1
-28.6
-28.5
-25.8
-27.3
-25.8
-24.7
-26.8
-27.2
-26.4
-27.0
-26.7
-------
TABIE 10 (COMB.)
CRUDE
OIL
NO,
SATURATES
AHOMATICS
TOTAL C.iUCE
DEL €-13
ASPHALTICS
WT. «
DEL C-13
:>5
36
37
«+1
42
4^
44
45
4o
47
51
52
53
5^
5 5
5*j
o7
5 a
59
60
ol
62
63
64
65
66
67
6 a
69
70
71
72
73
-27.6
-27.3
-28.3
-25.1
-23.9
-25.5
-26.3
-26.9
-27.1
-28.4
-25.8
-26.2
-26.8
-26.7
-26.6
-27.5
-27.4
-26.6
-26.0
-26.2
-25.3
-27.4
-27.6
-26.9
-26.6
-26.3
-26.fi.
-26.1
-26.9
-29.5
-27.1
-25.3
-27.1
67.2
63.5
76.7
50,3
81.0
42.6
53.3
15.7
38.0
68.5
76.1
25.4
24.9
26.4
27.3
3U.5
61.9
64.1
51.3
59.9
60.6
67.8
36.7
45.3
39.5
43.2
48.5
42.9
35.4
32.4
45.6
54.6
43.9
46.2
-28.4
-28.3
-28.9
-25.S
-25.5
-26.2
-27.2
-28.2
-27.3
-29.4
-28.6
-26,1
-27.9
-27.9
-28.4
-27.2
-27,6
-26.9
-26.6
-26.0
-27.3
-25.2
-28.1
-28.3
-27.1
-26.8
-26.6
-26.7
-26.3
-26.9
-30.0
-27.3
-25.8
-26.9
WT. %
28.5
32.8
20.7
37.3
17.5
47.7
3fc.3
62. Z
48. S
28.1
20.2
66.9
60.6
65.8
62.9
61.2
34.2
2B.6
42.8
35.5
35.9
26.1
48.7
46.4
53.6
48.7
44.2
44.6
52.0
48.6
42. a
36.9
38.6
41.6
DEL C-13
-26.5
-26.8
-28.5
-24.6
-23.7
-25.1
-26.0
-27.9
-27.7
-28.8
-27.2
-26.4
-26.9
-27.3
-27.3
-27.1
-27.2
-26.7
-26.1
-25.9
-25.7
-24.7
-27.3
-27.3
-26.8
-26.1
-25.9
-26.1
-26.3
-26.7
-Z&.9
-26.8
-24.6
-26.9
WT. 1
4,3
3.7
2.6
12.4
1.5
9.7
8.4
22.1
12.7
3.5
3.7
7.7
14.5
7.8
9.8
8.3
3.9
7.3
5.8
4.6
3.4
6.1
14.5
S.4
6.8
8.1
7.2
12.5
12.6
19.1
11.6
8.5
17.5
12.2
DEL C-13
-25.2
-23.9
-27.5
-21. fl
-24.8
-25.9
-26,0
-27.3
-27.6
-27.3
-25.5
-26.3
-27.1
-27,2
-26.4
-26.6
-26,5
-27.1
-26.0
-25.3
-25.1
-24.0
-24.1
-24.1
-26.7
-26.2
-26.2
-26.3
-26.3
-27.0
-28.3
-26.9
-24.6
-27.3
-------
TABIE 10 (CONTD.)
vr«
CRUDE
U1L
NO.
SATURATES
AROMATICS
ASPHALTICS
TOTAL C-
55.7
61.3
62.1
ao.2
73.5
30.3
35.2
-26.2
-22.7
-22.8
-25.2
-27.3
-26.7
-26.7
-26.5
-28.6
-27. S
-27.6
WT. 1
DEL C-13
47.4
64.1
65.1
9.8
36.9
il.6
33.0
17.1
19.2
53.5
53.3
-25.5
-22.4
-22.1
-24.0
-26.3
-26.1
-26.1
-25.9
-23.1
-27.1
-25.9
WT. I
31.4
21.3
20.5
0.7
7.4
7.1
5.0
2.6
2.4
16.2
11.5
0£L C-13
-26.0
-21.7
-21.9
-24.5
-26,1
-25.8
-26.0
-26.0
-27.4
-27.3
-25.3
-------
These observations are expected based on the hypothesis of hydrocarbon
formation. The mechanism of light hydrocarbon formation by thermal
12 ^12
cracking of organic material favors rupture of the C - C bond over
the C12 - C1^ bond (by about 8% in simple molecules). The remaining
residual organic material will be isotopically heavier with respect to
the original. This mechanism implies a .dehydrogenation resulting in
double bond formation in the parent molecules; as this process continues
polyenes may form, which can aromatize or polymerize. This would lead
to aromatics and asphaltene-type compounds, respectively, that are
isotopically heavier than the non-aromatics (saturates). The validity
Q/l
of this hypothesis is confirmed in part by Silverman and Epstein for
27
aromatics and by Silverman for asphaltenes.
Examination of the carbon isotope data on the silica gel fractions
600 F bottoms show that the saturates are isotopically lighter than
600 F bottoms (A = 0.6 or greater) for 41 of 77 sources.
Normal Paraffins and Odd-Even Predominance Curves
The n-paraffins are isolated from the saturate fraction by urea
adduction. When urea crystallizes from solution the n-paraffins are
trapped within the crystal lattice, the dimensions of which accommodate
selectively the linear molecules but exclude the branched and cyclic
paraffin molecules. The relative n-paraffin distribution is obtained
by normalization of the measured weight per cent n-paraffins from C
through C^g. The n-paraffin distribution is utilized to calculate the
odd-even predominance or OEP, which is a measure of deviation from an
unbiased distribution of n-paraffins of odd and of even carbon numbers
28
as a function of carbon number. The OEP values are computed accord-
ing to the following equation:
-------
OEP
i+2
&C
(-1)1+1
i+2
(7)
where C^ is the relative weight per cent of a n-paraffin containing i
carbons per molecule. The computed OEP value is assigned to the
n-paraffin containing i+2 carbon atoms in the molecule. The OEP value
is unity for an unbiased distribution, greater than unity if odd-
carbon numbered n-paraffins predominate and less than unity if even-
carbon numbered n-paraffins predominate. The OEP values are plotted
as a function of carbon number from C.. - to C _. The shape of the
resulting c,urve, i.e., the location and number of maxima and minima,
is characteristic of the crude oil genesis source and therefore, of
the crude oil itself. The shape of the curve varies among unrelated
crude oils and is similar for related samples. Typical OEP curves are
illustrated in Figure 8. The curves in Figure 6A are for two crude
oils from Nigeria, Nos. 35 and 36, and two from the Southern Alaskan
coast, Nos. 3 and 4. Four offshore Louisiana, Nos. 78, 79, 80, and 81,
and eight western shore Persian Gulf sources, Nos. 59, 60, 64, 65, 66,
67, 68, and 69 give similar curves in each case, Figure 6B and 6C.
This suggests a common geologic origin for each group of crude oils.
However, the similarity prevents application of this parameter to
distinguish among the sources within each group. Distinctive OEP
curves characterize related oils in other oil producing areas of the
world and these curves are significantly different from those examples
in Figure 8. On this basis the OEP curves could contribute to
pollution source identification.
59
-------
1.1 h
15
20
25
30
35
40
n-PARAFFIN CARBON NUMBER
FIGURE 8 - TYPICAL ODD-EVEN PREDOMINANCE (OEP) CURVES
60
-------
SECTION VIII
COMPARISON TECHNIQUES
The utilization of a multiparameter oil pollution source identifica-
tion system requires the completion of three tasks: collection of
pollutant samples and samples representative of potential sources,
measurement of the identification parameters on each sample, and
comparison of the values of the measured parameters of the pollutant
sample with those of the potential source samples to identify the
pollution source. The procedures for sample handling and parameter
measurement have already been described. Methods used in the com-
parison of parameter values are described in this section.
The potential sources can be in the form of a data library, which
consists of the measured values of the identification parameters for a
large number of oils. Alternatively, the potential sources can be
samples from several suspect sources in the vicinity of the spill at
the time of the spill. In either case, the correct identification is
possible only when the "real" source is among the suspect sources.
The present lack of an officially sanctioned oil pollution source
identification system makes establishment of a data library premature.
In cases where a small number of suspect source samples exists, it is
possible to compare sample parameters without using a data handling
system. In these cases, the bias of the individual analyst may
influence the identification.
Two computer-based comparison systems were applied to the identifica-
tion of an "unknown" as being equivalent to one of the sources in our
data library. Computerized systems have the advantages that analyst
bias is eliminated and that comparisons of a large number of oils are
61
-------
9
readily made. The first system is based on Qwick-Qwery (QQ) , a
computer data handling system, which is applied to match, in sequence,
each of the parameters for the unknown with those in the data library.
Those sources that match on 7 of 7, 6 of 7, etc., of the parameters
are printed out. A match of any one of the parameters is scored if
the library data fall within an assigned - value of the unknown. In
order for an oil to be identified as the oil pollution source, the
pollution sample and the potential source should match in all of the
measured parameters.
The second computer-based comparison system is based on a statistical
procedure that incorporates the fundamentals of multivariate normal
analysis. A detailed description of the procedure is given in
Appendix N. Using this procedure, each of the oils in the library is
compared with the unknown and the "closeness" of match of the identifi-
cation parameters is calculated in terms of the discriminant variate for
each potential source. The computer output ranks the potential sources
according to increasing value of the discriminant variate. The
discriminant variate will be zero if the identification parameters for
the unknown perfectly match those of a library source. The expected
value of the discriminant variate for two oils from the same source
but subject to measurement variations is equal to the number of
parameters tested, i.e., a value of 6 is expected, if 6 parameters are
applied to identification.
A decision problem arises using either of the computer-based comparison
systems when less than a perfect match is attained between an unknown
and a library oil. Using the Qwick-Qwery system when no or more than
one perfect match occurs the data of each of the prime suspect sources
is examined and a decision made on the basis of the data and the
geographical location of the source. Although Qwick-Qwery reduces the
62
-------
number of suspects to the most likely sources, the final identification
is still subject to individual bias.
Using the statistical comparison system a decision must be made as to
what value of the discriminant variate eliminates an oil from consid-
eration as the pollution source. An approximation to this decision
can be achieved by estimation of the discriminant variate value when
the number of separate samples from each potential source is equal to
the number of identification parameters matched. If the unknown and a
potential source are the same based on 6 parameters, the probability
is 95$ that the value of the discriminant variate will be 20 or less.
With 5 parameters the 95$ probability value of the discriminant
variate increases to 40. These values were used in evaluation of this
system to identify oils.
It is recommended that both Qwick-Qwery and the statistical procedure
be applied to select the most likely sources. The measured parameters
for these sources are compared visually to asstme that no potential
source is removed from consideration because of a measurement error.
When the possibility of such an error has been eliminated, the identi-
fication can be made using the statistical procedure.
These two comparison systems were evaluated in their ability to
identify an unknown with five combinations of the measured parameters.
These combinations were: (A) sulfur, nitrogen, vanadium and nickel
content and the hydrocarbon GLC profile, (B) the five parameters in A
plus carbon isotopic composition, (C) the six parameters of B plus
sulfur isotopic composition, (D) the six parameters of B plus the
weight per cent saturates, aromatics and asphaltics and the carbon
isotopic composition on these three fractions, and (E) the twelve
parameters of D plus the sulfur isotopic composition. Neither the
odd-even predominance ratio nor the sulfur GI£ profile parameters have
63
-------
digital values associated with them and for this reason they were
omitted from the evaluation studies.
Each of the five parameter combinations was evaluated with five
"unknown" oils using both comparison systems. The five unknowns
selected were:
a. Unknown 1 Chosen to resemble oil No. 54
b. Unknown 2 Chosen to resemble oil No. 10
c. Unknown 3 Measured values for oil No. 30
d. Unknown 4 Measured values for oil No. 65
e. Unknown 5 Average value of each parameter
among all library oils
The values of the identification parameters for each of the five
unknowns are listed in column 2 of Tables 11-15. The Qwick-Qwery -
values of allowed deviation are given in column 3. Potential crude
oil sources and their parameters that are the most similar to the
unknowns are listed in the remaining columns in the upper half of each
table. The lower halves summarize the results of the two comparison
systems for each of the five parameter combinations, i.e., the number
of matched parameters for each potential source for Qwick-Qwery and
the value of the discriminant variate for the statistical procedure.
The results of the evaluation can be summarized by the statement that
each comparison system matched the unknowns 1 to 4 with the correct oil
and that with one exception no other oils in the data library are
matched close enough for a mistake in identification to occur. The
one exception occurs with unknown 1, which although identified
correctly as oil No. 54, is not significantly different from No. 53
(Table ll). For unknown 5> the average oil, none of the oils in the
data library can be seriously considered as a potential source.
64
-------
TABLE 11
COMPARISON OF UNKNOWN 1 WITH POTENTIAL SOURCES
Allowed
Crude Oil Numbers
vn
Parameters Unknown
1. Sulfur, Wt. %
2. Nitrogen, Wt. %
3. Vanadium, Wt. ppm
4. Nickel Wt. ppm
5. HC GLC Profile
6. 813 C - Total
7. Saturates, Wt. %
8. £-3c - Saturates
9. Aromatics, Wt. %
10. S13C - Aromatics
11. Asphaltics, Wt. %
12. 813C - Asphaltics
13. §34s
4.90
0.20
55.7
63.5
1
-27.0
26.5
-28.0
63.0
-27.0
9.5
-27.0
-14.0
1 Deviation (±) 54
0.24
0.10
5.0
5.0
1
0.5
2.0
0.5
2.0
0.5
2.0
0.5
1.5
4.83
0.24
58.9
66.9
1
-26.7
27.3
-28.5
62.9
-27.4
9.8
-26.5
-14.4
53
4.65
0.20
52.4
60.2
2
-26.8
26.4
-28.0
65.8
-27.4
7.8
-27.3
-14.2
55
4.07
0.16
43.2
38.5
3
-26.6
30.5
-27.3
61.2
-27.2
8.3
-26.7
-13.5
52
4.39
0.24
82.9
39.8
3
-26.2
24.9
-28.0
60.6
-27.0
14.5
-27.2
-14.3
51
5.18
o.n
92.0
25.2
1
-25.8
25.4
-26.2
66.9
-26.5
7.7
-26.4
-13.2
43
1.37
0.21
57.1
31.5
6
-25.5
42.6
-26.3
47.7
-25.2
9.7
-26.0
2.5
15
4.91
0.11
15.6
14.1
2
-25.3
26.2
-25.1
64.1
-25.2
9.7
-25.6
-7.1
27
3.80
0.20
73.3
23.8
0
-27.3
33.0
-28.0
57.6
-27.0
9.5
-27. 4
-8.7
Number of
Qwick Qwery Comparison
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
Parameters
5
6
7
12
13
5
6
7
12
13
4
5
6
10
11
Number
1
2
3
6
7
of Matched Parameters
1
1
2
5
6
2
2
3
5
6
3
3
3
4
4
3
3
3
6
6
2
3
3
7
7
Number of
Statistical Comparison
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
Parameters
5
6
7
12
13
11
12
13
20
21
22
22
22
36
36
Discriminant
332
334
334
362
362
367
374
374
408
408
Variate
612
628
629
690
691
Value
7854
7879
8181
8451
8753
7804
7836
7883
7989
8035
582
583
614
656
687
-------
ON
ON
TABLE 12
COMPARISON OF UNKNOWN 2 WITH POTENTIAL SOURCES
Allowed Crude Oil Numbers
Parameters Unknown
1. Sulfur,, Wt. %
2. Nitrogen, Wt. %
3. Vanadium, Wt. ppm
4. Nickel, Wt. pm
5. HC GLC Profile
6. g!3 C - Total
7. Saturates, Wt. %
8. S^C - Saturates
9. Aromatics, Wt. %
10. g^C - Aromatics
11. Asphaltics, Wt. %
12.- 813C - Asphaltics
13. 8^S
2.94
0.19
12.0
8.0
2
-28.4
40.0
-29.9
50.0
-28.8
10.0
-28.1
-4.0
2 Deviation (±)
0.24
0.10
5.0
5.0
1
0.5
2.0
0.5
2.0
0.5
2.0
0.5
1.5
10
2.90
0.18
11.2
7.5
1
-28.7
40.3
-30.2
49.9
-29.2
9.8
-28.4
-4.3
16
3.91
0.22
14.4
15.7
1
-25.5
36.8
-25.6
53.6
-25.2
9.6
-25.5
-8.2
19
2.97
0.10
12.4
7.1
2
-25.0
52.7
-25.1
39.5
-25.1
7.8
-25.0
-9.6
29
2.39
0.09
10.0
2.8
1
-26.6
48.3
-26.9
45.9
-26.3
5.7
-24.8
-
5
1.30
0.26
18.8
12.3
6
-29.3
40.9
-30.4
51.5
-29.3
7.6
-29.4
-3.1
46
1.21
0.23
39.0
9.0
10
-27.1
38.6
-27.4
48.8
-27.8
12.7
-27.7
8.7
65
2.11
0.09
14.0
3.8
6
-26.3
48.5
-26.7
44.2
-26.0
7.3
-26.3
-8.3
67
2.95
0.14
25.6
7.0
1
-26.6
42.9
-26.8
44.6
-26.2
12.5
-26.4
-2.7
Number of
Qwick Qwery Comparison Parameters
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
5
6
7
12
13
5
6
7
12
13
3
3
3
4
4
Number
5
5
5
5
5
of Matched Parameters
3
3
3
3
3
3
3
4
7
8
2
2
2
5
5
4
4
4
4
4
4
4
5
4
5
Number of
Statistical Comparison
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
Parameters
5
6
7
12
13
9
10
11
14
15
224
317
337
111
796
Discriminant
23
151
186
942
977
264
300
-
694
—
Variate
1993
2002
2003
2035
2036
Value
4240
4258
4437
4351
4530
410
459
479
809
830
399
435
437
693
695
-------
COMPARISON OF UNKNOWN 3 WITH POTENTIAL SOURCES
Parameters
1. Sulfur, Wt. %
2. Nitrogen, Wt. %
3. Vanadium, Wt. ppm
4. Nickel, Wt. ppm
5. HC GLC Profile
6. 513 c - Total
7. Saturates, Wt. %
8. S^c - Saturates
Unknown 3
Allowed
Deviation (±)
Crude Oil Numbers
9. Aromatics, Wt. %
10. §13 C - Aromatics
11. Asphaltics, Wt. %
12. 813C - Asnhaltics
13. 834S
Qwick Qwery Comparison
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
Statistical Comparison
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
3.90
0.23
83.3
24.8
5
-26.6
36.7
-27.2
55.5
-26.9
8.1
-26.9
-5.4
Number of
Parameters
5
6
7
12
13
Number of
Parameters
5
6
7
12
13
0.24
0.10
5.0
5.0
1
0.5
2.0
0.5
2.0
0.5
2.0
0.5
1.5
30
3.90
0.23
83.3
24.8
5
-26.6
36.7
-27.2
55.2
-26.9
8.1
-26.9
-5.4
52
4.39
0.24
82.9
39.8
3
-26.2
24.9
-28.0
60.6
-27.0
14.5
-27.2
-14.3
27
3.80
0.20
73.3
23.8
0
-27.2
33.0
-28.0
57.6
-27.0
9.5
-27.4
-8.7
69
3.82
0.18
63.4
23.2
0
-26.9
32.4
-27.0
48.6
-26.8
19.1
-27.1
-9.6
31
3.77
0.13
47.9
15.0
0
-26.7
38.7
-27.9
54.6
-27.3
6.7
-27.3
-9.6
Number of Matched Parameters
5
6
7
12
13
2
3
3
5
5
3
3
3
6
6
Discriminant Variate
0
0
0
0
0
69
71
158
285
373
148
152
164
182
194
3
4
4
7
7
Value
323
324
344
512
532
2
3
3
8
8
1190
1190
1210
1206
1226
-------
ON
CO
TABLE
COMPARISON OF UNKNOWN 4
Allowed
14
WITH POTENTIAL SOURCES
Crude Oil Numbers
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Parameters
Sulfur, Wt. %
Nitrogen, Wt. %
Vanadium, Wt. ppm
Nickel, Wt. pm
HC GLC Profile
§13 c - Total
Saturates, wt. %
g^-3c - Saturates
Aromatics, Wt. %
§ ^C - Aromatics
Asphaltics, Wt. %
5 ^C - Asphaltics
834s
Qwick Qwery Comparison
IT
B.
C.
D.
E.
Parameters 1-5
Parameters 1-6
Parameters 1-6+13
Parameters 1-12
Parameters 1-13
Statistical Comparison
A.
B.
C.
D.
E.
Parameters 1-5
Parameters 1-6
Parameters 1-6+13
Parameters 1-12
Parameters 1-13
Unknown 4 Deviation (±)
2.86
0.12
23.9
8.4
4
-26.6
43.2
-26.9
48.7
-26.2
8.1
-26.3
-14.9
Number of
Parameters
5
6
7
12
13
Number of
Parameters
5
6
7
12
13
0
0
5
5
1
0
2
0
2
0
2
0
1
.24
.10
.0
.0
.0
.5
.0
.5
.0
.5
.0
.5
.5
65
2.
0.
23.
8.
4
-26.
43.
-26.
48.
-26.
8.
-26.
-14.
5
6
7
12
13
86
12
9
4
6
2
9
7
2
1
3
9
67
2.95
0.14
25.6
7.0
1
-26.6
42.9
-26.8
44.6
-26.2
12.5
-26.4
-2.7
Number
4
5
5
9
9
18
4.08
0.13
19.3
14.1
2
-25-1
32.7
-25.3
58.5
-25.2
8.8
-25.7
-6.0
64
3.03
0.19
49.0
12.2
2
-26.9
39.5
-27.2
53.6
-26.9
6.8
-26.8
-10.0
55
4.07
0.16
43.2
38.5
3
-26.6
30.5
-27.3
61.2
-27.2
8.3
-26.7
-13.5
44
1.70
0.17
85.5
10.7
4
-26.3
53.3
-27.3
38.3
-26.1
8.4
-26.1
6.6
of Matched Parameters
2
2
2
3
3
3
4
4
7
7
2
3
4
6
7
3
4
4
8
8
Discriminant Variate Value
0
0
0
0
0
47
47
211
83
248
356
380
468
631
719
544
545
572
594
620
684
684
685
1015
1018
1979
1980
2493
2193
2705
-------
TABLE 15
COMPARISON OF UNKNOWN 5 WITH POTENTIAL SOURCES
o
Parameters
1. Sulfur, Wt. %
2. Nitrogen, Wt. %
3. Vanadium, Wt. ppm
4. Nickel, Wt. pian
5. HC GLC Profile
6. §13 C - Total
7. Saturates, Wt. %
8. 5 C - Saturates
9. Aromatics, Wt. %
10. §-*-3c - Aromatics
11. Asphalt ics, Wt. %
12. S13G - Asphalt ics
13. 834s
Qwick Qwery Comparison
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
Statistical Comparison
A. Parameters 1-5
B. Parameters 1-6
C. Parameters 1-6+13
D. Parameters 1-12
E. Parameters 1-13
Unknown 5
1.78
0.18
34.8
18.6
4
-26.7
50;4
-27.4
41.3
-26.6
8.3
-26.2
-1.7
Number of
Parameters
5
6
7
12
13
Number of
Parameters
5
6
7
12
13
Allowed
Deviation (±)
0.24
0.10
5.0
5.0
1
0.5
2.0
0.5
2.0
0.5
2.0
0.5
1.5
Crude Oil Numbers
43
1.37
0.21
57.1
31.5
6
-25.5
42.6
-26.3
47.7
-25.2
9.7
-26.0
2.5
1
1
1
3
3
44
1.70
0.17
85.5
10.7
4
-26.3
53.3
-27.3
38.3
-26.1
8.4
-26.1
6.6
Number of
3
4
4
8
8
67
2.95
0.14
25-6
7.0
1
-26.6
42.9
-26.8
44.6
-26.2
12.5
-26.4
-2.7
65
2.86
0.12
23.9
8.4
4
-26.6
43.2
-26.9
48.7
-26.2
8.1
-26.3
-14.9
11
0.26
0.08
1.07
0.80
10
-26.9
50.3
-28.4
41.2
-27.0
8.5
-27.1
-1.8
Matched Parameters
1
2
3
4
5
Discriminant Varlate
453
468
488
607
627
928
930
1007
951
1027
687
688
689
779
780
2
3
3
7
7
Value
662
662
855
Ilk
96?
1
2
3
6
7
286, 787
286,788
286,788
286,810
286,810
-------
The successful distinction of five Mississippi crude oils, Nos. 15-19,
from one another was suggested as a rigorous test of any matching
scheme (page 40-41). The statistical procedure based on the four
elemental parameters achieved the unique distinction as demonstrated
by the values of the discriminant variate when each oil source in turn
is treated as the unknown (Table 16).
The statistical procedure was also evaluated on the basis of the
efficiency of four, five and six parameters to distinguish among the
crude oils in a 50 source library. Each of the oils were used, in
turn, as an unknown to determine those sources that were not identified
uniquely. Sources were considered non-unique when the value of the
discriminant variate was less than 40 for four and five parameters and
less than 20 for six parameters. The results are summarized in
Table 1? for the three parameter combinations. The numerator in the
fraction is the number of sources that are not uniquely identified
from among the 50 sources. The efficiency of the sulfur isotopic
composition as a diagnostic parameter is demonstrated by the fact that
only four sources are not distinguished from all other sources. These
four are two pair of crude oil sources, Nos. 59-60 and 62-63 from
adjacent fields in Abu Dhabi and Israel. Each pair is distinguished
from the other as well as from all other sources. The paired oils are
indistinguishable and each pair should be considered a single source.
For practical purposes this source library contains 48 unique sources
based on these six identification parameters and 50 crude oil samples.
70
-------
TABLE 16
COMPARISON OF MISSISSIPPI CRUDE OIL SOURCES
Crude Oil Numbers
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Parameters
Sulfur, Wt, %
Nitrogen, Wt. %
Vanadium, Wt. ppm
Nickel, Wt." pirn
HC GLC Profile
813C - Total
Saturates, Wt. %
813C - Saturates
Aromatics, Wt. %
13
5 C - Aromatics
Asphaltics, Wt. %
§13 c - Astshaltics
»34s
Unknown Oil
15
16
17
18
19
15
4.91
0.11
15.6
14.1
2
-25.3
26.2
-25.1
64.1
-25.2
9.7
-25.6
-7.1
16
3.91
0.22
14.4
15.7
1
-25.5
36.8
-25.6
53.6
-25.2
9.6
-25.3
-8.2
17
1.65
0.06
<0.2
2.0
3
-24.9
62.2
-25.2
35.7
-24.3
2.1
-24.4
-9.4
Discriminant Variate
0
181
2609
159
596
224
0
2059
145
239
872, 290
740,381
0
1,335,119
545,145
18
4.08
0.13
19.3
14.1
2
-25.1
32.7
-25.3
58.5
-25.2
8.8
-25.7
-6.0
Value*
166
103
2266
0
386
19
2.97
0.10
12.4
7.1
2
-25.0
52.7
-25.1
39.5
-25.1
7. a
-25.0
-9.6
1060
350
1390
651
0
-"-Discriminant variate determined using the four parameters: sulfur, nickel, vanadium and
nickel contents.
-------
TAB IE 17
IDENTIFICATION EFFICIENCY OF SULFUE ISOTOPIC COMPOSITION
Parameters 4 5 6
s
N
V
Ni
MB
S
N
V
Ni
r&r
o 0
S
N
V
Ni
613C
Unidentified Sources 16/50 13/50 4/50
As pointed out in Appendix N, several assumptions and approximations
are embodied in the statistical procedure because only one sample of
each source was available. The validity of the assumptions and the
consequence of the approximations could be established by measurement
of the variations in the values of the proposed identification
parameters that occur in the normal course of oil production from
commercial crude oil sources.
72
-------
SECTION IX
WEATHERING STUDIES
The demonstration of the feasibility to identify a crude oil based on
a comparison of measured values for selected parameters has been made
on crude oils which were not subjected to environmental conditions at
an oil spill site.
The identification system takes evaporative weathering into account by
measuring the parameters of the crude oil after it has been topped to
600 F to remove the volatile components. In an actual oil spill,
however, in addition to evaporation, the characteristics of a crude
oil could also change through bacterial degradation, photolysis,
oxidation, and water solubility of some components. The influence of
weathering on the identification parameters was studied by weathering
experiments under simulated-ocean spill conditions.
Weathering Procedure
The weathering experiments were conducted in 2 x 4 foot galvanized
metal tubs located on a building roof in the Phillips Research Center.
The interior surfaces of the tubs were coated with an epoxy paint to
eliminate any influence from the zinc galvanizing and to reduce
corrosion. Approximately eight gallons of synthetic sea water
(Instant Ocean from Aquarium Systems, Wickliffe, Ohio 44092) were
placed in each tub. One end of the tub was filled with quartz river
sand to give a small area above the water lev/el to simulate a beach.
To provide a marine coastal microbial culture, one quart of tidal
flats mud from the Sorrento Valley estuary (California) was mixed into
the sea water and allowed to settle overnight. From 500 to 800 ml of
crude oil was poured onto the water surface. Initially the oil
covered 1/3 to 1/2 of the water surface and after a week of outdoor
73 '
-------
exposure the water was completely covered with oil, but not uniformly.
Air was bubbled through the water continuously to simulate wave action.
Two crude oils, Nos. 83 and 84, were weathered under simulated ocean
conditions at ambient atmospheric conditions from August 10 to
September 29, 1972. Crude oil No. 83 from Monagas, Venezuela formed a
crusty, viscous coating on the water surface after the fourth day.
After 49 days, the water surface was nearly free of oilj the oil
residue had either sunk to the bottom or stuck to the walls of the
tank. Crude oil No. 84 from Ecuador covered the water surface com-
pletely after several days with a fluid residue. The thickness of the
oil layer depended upon the wind conditions, a thicker layer being
produced at the downwind side of the tank.
Samples of the oil residue remaining on the water surface were taken
after 1, 4> 8, 21 and 49 days. Each sample was refrigerated to
solidify the oil and any water was removed by decantation. The oils
were dissolved in toluene, and then topped by vacuum distillation to
produce the 600 F bottoms for measurement of the parameters. The
values obtained are reported in Tables 18 and 19.
The values for the sulfur, nitrogen, nickel and vanadium concentrations
as well as the carbon and sulfur isotopic compositions are within the
range of values expected from variations in sampling, distillation and
measurement and, therefore, were not effected by these weathering con-
ditions. The GIL profiles and the values from the silica gel separa-
tion are affected and thus are of decreased value in oil source
identification. With each oil, the asphaltics content nearly doubles
during the course of the weathering. The loss of the lower boiling
components, i.e., saturates and aromatics, is shown by changes in the
front end of the GLC profiles. Initially, for the Ecuador crude oil
74
-------
TABLE 18
EFFECT OF WEATHERING CRUDE OIL NO. 83, MONAGAS
PIPELINE CRUDE OIL, VENEZUELA
Days Weathered
Parameter
Sulfur, Wt. %
Nitrogen, Wt. %
Nickel, ppm
Vanadium, ppm
Saturates, Wt. %
Aromatic s, Wt. %
Asphaltics, Wt. %
a) 600+F Residue
b) Saturates
c) Aromatics
d) Asphaltics
C^H-o O/
PITM* /CO
GLC Profile
: 0
2,
0
70
232
30
53
16
-27
-27
-27
-27
7
10
.4
.48
.3
.5
.2
.0
.8
.1
.3
.6
1
2.4
0
84
256
29
50
19
-27
-27
-27
-26
6
10
.53
.9
.9
.2
.1
.6
.3
.8
.5
4
2.
0.
81
244
28.
48.
22.
-27.
-27.
-26.
-27.
7.
10
5
57
7
9
4
9
6
8
4
2
8
2
0
86
254
27
44
27
-28
-27
-2?
-27
10
.5
.65
.6
.7
.7
.5
.5
.2
.4
-
21
2.
0.
75
249
25.
43.
30.
-27.
-27.
-27.
-27.
10
5
59
8
8
4
4
2
2
2
49
2.6
0.54
76
241
26.8
45.3
27.8
-27.1
-27.6
-27.3
-27.3
7.4
10
75
-------
Parameter
TABLE 19
EFFECT OF WEATHERING CRUDE OIL NO. 84,
ECUADOR COMPOSITE CRUDE
Days Weathered
Sulfur, Wt. %
Nitrogen, Wt. %
Nickel, ppra
Vanadium, ppm
Saturates, Wt. %
Aromatics, Wt. %
Asphaltics, Wt. %
i!3p o/
6 LPDB> /°°
a) 600+F Residue
b) Saturates
c) Aromatics
d) Asphaltics
6 SCDM* °/
GLC Profile
0
1.2
0.28
29
73
35.2
53.3
11.5
-25.8
-27.6
-25.9
-25.3
1
1.0
0.25
36
69
45.9
42.5
11.6
-26.1
-27.8
-26.0
-25.5
_k__
1.1
0.28
39
79
40.3
47.8
11.9
-26.1
-27.5
-25.4
-25.6
8
1.1
0.32
37
70
43.2
42.3
14.5
-26.3
-27.4
-25.4
-25.4
21
1.2
0.33
37
85
39.9
41.0
19.1
-26.1
-27.3
-25.3
-25.9
49
1.3
0.35
36
83
32.1
47.6
20.3
-26.2
-27.3
-25.6
-25.7
1.6
6
- 2.0
9
76
-------
the GLC profile on the 600 F bottoms contains large peaks for the
normal paraffins C-j, - C^g. After 49 days of weathering the C . and
C,r peaks are reduced to a small fraction of their initial size and
the C^ - C.^ peaks are reduced significantly as is the envelope under
the paraffin peaks through C_5 (Figure 9). The heavier Monagas
Venezuelan crude oil also loses the lighter ends during weathering
(Figure 10). The area between the two curves represents the material
lost during weathering. The OEP curves for the Ecuador crude oil
samples did not change significantly as a function of weathering time.
The decrease in concentration of both saturates and aromatics indicates
that an evaporative mechanism is causing most of the changes rather
than microbial action. Bacterial degradation is expected to affect
only the n-paraffins.
The enrichment of the heavier end of the 600 F bottoms (Ecuador) at the
expense of the lighter ends during weathering is illustrated by the
data in Table 20. The lighter normal paraffins C,, - C,/ decreased
from a total of 9.2 per cent of the total paraffins (C - C_g) to
about 1.0 per cent during the 49 days of weathering. The Venezuelan
crude oil contains too small an amount of n-paraffins to be measured
by the techniques employed. This fact itself is sufficient to dis-
tinguish this crude oil from the majority of those in our survey.
The infrared spectra of the weathered samples after distillation
indicated a progressive small increase in hydroxyl, -OH, and carbonyl,
"C = 0, absorptions with the duration of weathering. Oxidation of
crude oil residues is expected under our weathering conditions and the
observed infrared spectra confirm this. The variations in the spectra
of the two unweathered 600+F bottoms are insufficient to distinguish
between them.
77
-------
A. UNWEATHERED
3. WEATHERED 8 DAYS
00
C. WEATHERED 21 DAYS
D. WEATHERED 49 DAYS
FIGURE 9 - EFFECT OF WEATHERING ON GLC PROFILE, 600^ BOTTOMS OF ECUADOR CRUDE OIL
-------
VENZUELA CRUDE OIL
UPPER CURVE'UNWEATHERED
LOWER CURVE : WEATHERED 21 DAYS
C30
841
C20
651
C15
519
C13
°F
-K.
FIGURE 10 - EFFECT OF WEATHERING ON GLC PROFILE (600 F BOTTOMS)
-------
_lk_
2.0
1.6
0.8
0.2
0.05
0.04
11.
3.1
3.0
2.0
1.3
0.3
0.1
16_
4.1
4.1
3.3
3.0
1.8
0.9
1L.
5.9
5.8
5.0
5.4
5.1
3.8
18_
6.0
6.0
5.3
6.0
6.3
5.5
22_
2.4
2.5
2.8
2.5
3.1
4.2
3k.
1.5
1.4
1.7
1.4
2.1
3.0
3JL
0.8
0.8
0.9
0.7
1.1
1.8
2I_
0.6
0.7
0.7
0.6
0.9
1.7
TABLE 20
EFFECT OF WEATHERING ON RELATIVE NORMAL PARAFFIN CONTENT
Relative Normal Paraffin Content. Wt. %
Carbon Number
Days Weathered
0
1
4
8
21
49
#
The normal paraffins that are recovered by urea adduction of the
saturates fraction are measured by gas liquid chromatography. The
relative concentrations are calculated by normalization of paraffins
in the C^ - C»g range to 100%.
This weathering study demonstrates that the four elemental contents:
sulfur, nitrogen, vanadium and nickel, and the carbon and sulfur
isotopic compositions are stable to weathering when the measurements
are made on a 600 F bottoms. The other parameters are affected by
weathering and their application to source identification is limited
by the extent of weathering.
80
-------
SECTION X
SUMMARY DISCUSSION
A comprehensive multiparameter oil pollution source identification
system utilizing 15 diagnostic parameters is presented. The results
of measuring these 15 parameters on eighty crudes is* presented. The
effect of weathering crude oils in a simulated ocean environment is
described. Two computerized procedures for comparing the parameters
of an unknown oil with the parameters of the eighty crude oils are
also presented.
A prime objective of any passive tagging system is the elimination of
weathering as a variable in the identification of pollutants and their
source. This can be accomplished by either measuring parameters that
are unaffected by weathering or by subjecting both weathered and un-
weathered samples to a procedure designed to minimize or eliminate
any changes in samples caused by weathering. If subjected to oil
spill conditions, the properties of crude oils will begin to change
immediately due to evaporation and dissolution of lighter components.
The elemental analysis parameters are likely to change during weather-
ing because components containing heteroelements, S, N, V, and Ni, will
tend to be associated with the heavier components, thus their effective
concentration will increase as weathering progresses. Analogous
effects will also be noted with other parameters. Other changes may
be observed due to microbial action, photolysis, and oxidation effects
that occur during extended weathering.
The/measurement of the parameters on the 600 F bottoms prepared from
the crude oil samples represents a realistic approach to eliminating
weathering effects. Both weathered and unweathered oils are reduced
81
-------
to a common base on which measured parameters are directly comparable.
The adoption of 600 F bottoms (rather than some other temperature) as
the material on which to measure parameters represents an arbitrary
choice. As noted in Section IV and IX, some changes in the GLC
profiles, silica gel fractions, and OEP curves do occur during
weathering. These changes might be minimized by measurement of
parameters on bottoms having a higher initial boiling point such as
the 650 F bottoms recommended by the Institute of Petroleum. Such
a change in procedure should be carefully considered to assess the
influence of hydrocarbon cracking at the higher temperature involved.
Although the GLC profile is excluded as one of the quantitative
identification parameters because of weathering, the GLC profile can
serve as a preliminary screening measurement to establish whether an
unknown is derived from a refined product or a crude oil. The GLC
profile variation for the 600 F bottoms is sufficient to eliminate
some potential sources, even with weathered samples. For example,
those profiles exhibiting strong n-paraffin peaks are distinguished
from those having small or no n-paraffin peaks, i.e., Section VII,
Figure 6, page 47. The loss of the lighter n-paraffin by evaporation
does not invalidate the distinction.
The sulfur, nitrogen, vanadium, and nickel content parameters have a
sufficiently wide range of values to be useful in distinguishing
among oil sources. No significant changes in values are noted in
weathering studies. In many instances, these four parameters alone
are sufficient to distinguish among many potential oil sources.
These parameters are easily measured in most laboratories.
The carbon and sulfur isotopic composition parameters strengthen the
identification system, since each exhibits a range of values and
82
-------
remains unchanged during weathering. This measurement requires an
isotope-ratio mass spectrometer facility including the combustion-
purification trains to prepare samples for measurement.
It is possible to distinguish among most oil sources on the basis of
these seven parameters. In cases where further corroboration is
required, the sulfur GLC profile and fractions by silica gel separa-
tion for additional measurements can be obtained.
The sulfur GLC profiles and OEP curves of crude oils each represent
parameters that have a high potential for being able to distinguish
among sources. At this point, however, the utilization of these data
requires manual comparison and is thus time consuming and subject to
human error. Transformation of the data into a form compatible with
computerized analysis appears justified. Improved handling of the
hydrocarbon GLC profile data may permit their quantitative applica-
tion.
The use of V/Ni ratios as an identification has been suggested. '
In addition, S/N, V/S, V/N and other ratios might be considered. In
some situations, a ratio, such as the V/Ni ratio, may provide some
identification potential since change occurring in crude oils during
weathering may affect the individual concentrations without changing
ratios. In these cases, the ratio is independent of the concentra-
tions and represents a useful identification parameter.
The use of all the reported values based on the silica gel separations,
even in the absence of weathering, does not improve the identification
potential enough to justify the effort expended. The large variations
noted in the amount of asphaltic fraction are of potential value.
Additional work to obtain this fraction, or a comparable fraction, in
83
-------
a less time consuming process might be justified. Although the CEP
curves are potentially useful, studies to obtain the needed informa-
tion with less effort are warranted.
Three different comparison techniques have been used in evaluating the
identification techniques. These are a direct manual comparison and
the computer-based Qwick-Qwery and statistical comparison procedures.
Each of these techniques has merit for particular applications. The
best system will utilize a combination of these approaches.
Practical Evaluation of the Proposed System
The proposed system was tested on "real" samples that were obtained
from the Oil Spill Branch, Office of Research and Monitoring,
Environmental Protection Agency, Edison, New Jersey. Twelve crude
oils and two weathered residues were received. The weathered
residues were known to have originated from among the twelve crude
oils. Nine parameters were measured on each crude oil source sample
but only four parameters could be measured on the weathered residues
because of the small samples obtained (2-3 ml). All the measurements
were made on 600 F bottoms. The data are presented in Table 21. The
crude oils are numbers 85 through 96, inclusive, and the weathered
residues are numbers 99 and 100. Direct comparison of the data
indicates that oils 86 and 93 are nearly identical as are oils 94 and
95. Oils 85 and 89 are similar but the agreement is not as good.
The weathered (16 days) residue No. 99 matches oils 86 and 93 based
on the sulfur content alone. The vanadium and nickel contents agree
also. No further identification parameters are necessary to match
this weathered residue with the two low-sulfur crude oils in this
12-source library. The weathered (2 days) residue No. 100 may have
originated from oil 96 based on direct comparison of the sulfur
84
-------
TABLE 21. SUMMARY OF VALUES ON EPA CRUDE OILS AND WEATHERED SAMPLES
03
VA
EPA
Crude
Oil
Number
85
86
87
88
89
90
91
92
93
94
95
96
600 F
Residue
Wt. %
71.1
60.4
73.4
68.2
67.9
77.1
78.1
73.1
68.6
61.1
59.7
79.8
99 Weathered
Residue
100 Weathered
Sulfur
Wt. %
1.73
0.2?
1.27
2.01
1.60
1.66
1.30
1.45x
0.24
1.19
1.18
2.04
0.31
2.03
Nitrogen
Wt. %
0.26
0.17
0.26
0.25
0.25
0.24
0.28
0.25
0.17
0.07
0.07
0.29
-
Vanadium
Wppm
167
1.0
87.7
55.3
147
198
90.8
129
0.9
0.4
0.3
250
,(1)
200^
Nickel
Wppm
20.4
7.0
40.7
16.4
17.4
17.3
23.4
16.6
6.3
0.7
1.0
26.5
9(D
25^
GLC
Profile
3
1
8
2
6
0
8
0
1
3
3
8
9
9
A13r
o t
o/oo
-26.6
-26.6
-26.7
-26.6
-26.6
-26.5
-27.6
-26.2
-26.5
-25.8
-25.8
-26.5
-26.6
-26.8
Silica
Gel Fractions
Saturates Aromatics
31.2 60.9
39.3
34.0
46.0
38.4
41.3
33.7
32.2
34.6
41.2
42.6
28.6
-
57.3
59.6
48.3
54.6
53.9
59.0
61.7
59.0
54.7
54.0
62.8
-
. Wt. % _
Asphaltic
8.0
3.4
6.4
5.5
7.0
4.8
7.3
6.1
6.4
4.2
3.4
8.6
-
Residue
sample did not permit highest accuracy.
-------
vanadium and nickel contents. On the basis of the data oils 85, 89
and 90 are also potential sources. The small amount of weathered
residues that were available prevented the vanadium and nickel
measurements with the normal precision and accuracy. The uncertainty
of these values reduces the confidence in the match.
The agreement in the carbon isotopic composition between the weathered
residues and the potential sources confirm the tentative assignments.
Nine of the 12 crude oil sources have equivalent carbon isotopic
values. The values for oil Nos. 94 and 95 agree with one another and
differ from the others. The value for oil 91 suggests that it is from
a different geologic origin than the others.
The statistical comparison method was also applied to identification
of these weathered residues. Each of the twelve sources was treated
as an unknown to test for equivalency of sources. Two pair of oil
sources, Nos. 86-93 and 94-95, were found to be equivalent and another
pair 85-89 are probably equivalent. Therefore, there are only 10
distinctly different sources and possibly only 9.
The weathered residue No. 99 was matched statistically with the source
pair 86-93. The value of the discriminant variate was 266 for this
source. This value is larger than normally considered for a match.
However, with the knowledge that the unknown was derived from this
library, it is concluded that the match is the best possible. The
next closest match, oil pair 94-95, had a discriminant variate of 3198.
Positive identification of oil 100 was not feasible based on the
discriminant variate values:
86
-------
Source Discriminant
Oil Number Variate
96 182
85-89 331
90 405
Although the most likely source, No. 96, is the same as selected by
manual comparison of the data, the statistical procedure does not allow
a firm conclusion to be reached.
The Environmental Protection Agency confirmed our identification that
the weathered oil 99 was derived from source oil 93, which is identical
to 86. The weathered oil 100 was derived from source oil 90, which is
one of the three potential sources selected. Those two examples
demonstrate the application of the proposed identification system to
a real situation in which several suspect sources are nearly equivalent
and small unknown samples are available. In both cases the number of
potential sources was reduced to the point where other circumstantial
evidence is needed.
87
-------
SECTION XI
ACKNOWLEDGMENTS
The results and conclusions reported herein are possible only through
the assistance and cooperation of professional associates and fellow
employees. Their guidance, advice and interest from conception to
completion of the study is deeply appreciated and is hereby acknowledged,
The special assistance of Dr. T. V. lorns who performed the experi-
mental studies on the measurement of sulfur isotopic composition and
contributed to the preparation of the final report through extensive
discussions and written drafts is hereby gratefully acknowledged.
A number of crude oil samples were obtained from the Bartlesville
Energy Research Center, Bureau of Mines, U.S. Department of Interior.
Their interest and willingness in making these sources available to
the project is appreciated.
The support of the project by the Oil Spill Branch, Office of Research
and Monitoring, Environmental Protection Agency and the help provided
by Dr. J. LaFornara and Mr. B. Hornstein, Project Officers, is
acknowledged with appreciation.
89
-------
SECTION XII
REFERENCES
£•> 20 (48), 51 (Nov. 27, 1972).
2. "Oil Tagging System Study", Melpar, Federal Water Pollution
Control Administration, Contract No. 14-12-500, May, 1970.
3. Berkowitz, A. E., Lahut, J. A., and Meiklejohn, W. H., The Oil
Daily. November 14, 1972, p. 32 and Chan. Ene. News. £0 #47:2
(Nov. 20, 1972).
4. Kreider, R. E., "Identification of Oil Leaks and Spills", Proc.
Joint Conf . on Prevention and Control of Oil Spills, June 15-17,
1971, Washington, D.C., pp. 119-124.
5. Adlard, E. R., "A Review of the Methods for the Identification of
Persistent Hydrocarbon Pollutants on Seas and Beaches", J. Inst.
Petroleum, j>8 (560), 63-74 (March 1972).
6, Erdman, J. G. , and Morris, D, A., "Crude Oil Correlations",
Presented at the Geological Society of America Meeting,
November 13-15, 1972, Minneapolis.
7. Oil and Gas J., 21 (2), 50-1 (Jan. 8, 1973).
8. Oil and Gas J., 21 (5), 106 (Jan. 29, 1973).
9. "U. S. Energy Outlook: An Initial Appraisal 1971-1985" by
National Petroleum Council, Committee on U. S. Energy Outlook,
Chrmn - John G. McLean, July 15, 1971, p. 27-8.
10. Agazzi, E. J., Burtner, D. C., Crittenden, D. J., and Patterson,
D. R., "Determination of Trace Metals in Oil by Sulfur Incinera-
tion and Spectrophotometric Measurements", Anal. Chem., 21, 332-5
(1963).
11. Anon. "Analytical Methods for the Identification of the Source of
Pollution by Oil of the Seas, Rivers and Beaches", J. Inst. Petrol.
J£, (548), 107-117 (1970).
91
-------
12. Scalan, R. S. and Morgan, T. D., Isotope Ratio Mass Spectrometer
Instrumentation and Application to Organic Matter Contained in
Recent Sediments", Intl. J. Mass Spectre, and Ion Phys. 4* 267-281
(1970).
13. Craig, H., "Geochemistry of the Stable Carbon Isotopes", Geochjffli.
et Cosmochim. Acta, 3_, 53-92 (1953).
14. Adlard, E. R., Ceaser, L. F., and Matthews, P. H. D., "Identifica-
tion of Hydrocarbon Pollutants on Seas and Beaches by Gas
Chromatography", Anal. Chem. 44, 64-73 (1972).
15. Smith, C. L, and Maclntyre, W. G., "Initial Aging of Fuel 'Oil
Films on Sea Water", Proc. Joint Conf. on Prevention and Control
of Oil Spills, June 15-17, 1971, Washington, D. C., pp. 457-461.
16. Whisman, M. L. and Cotton, F. 0., "Butaines Data Promises Help in
Identifying Petroleum-Spill Sources", Oil and Gas J., 69 (52),
111-3 (Dec. 27, 1971).
17. Nelson, W. L., "How Much Metals in Crude Oils", Oil and Gas J.,
20 (82), 48-50 (Aug. 7, 1972).
18. Bryan, D. E. and coworkers "Development of Nuclear Analytical
Techniques for Oil Slick Identification (Phase I)", U. S. Atomic
Energy Comm. Report GA-9889 (Jan. 21, 1970); (Phase IIA),
GULF-RT-A-10684, (June 11, 1971).
19. Eckelmann, W. R., Broecker, W. S., Whitlock, D. W, and Alsup,
J. R., "Implications of Carbon Isotopic Compositions of Total
Organic Carbon of Some Recent Sediments and Ancient Oils", Bull.
Amer. Assoc. Petrol. Geol. 4ji, 699-704 (1962).
20. Thode, H. G., "Sulfur Isotope Geochemistry", in Studies in
Analytical Geochemistry. Dennis M. Shaw, edt. Univ. of Toronto
Press, Toronto, Canada, 1963, pp. 25-41.
21. Thode, H. G., Monster, J., and Dunford, H. B., "Sulfur Isotope
Abundances in Petroleum and Associated Materials", Bull. Amer.
Assoc. Petrol. Geol.. 42, 2619-2641 (1958).
92
-------
22. Monster, J.,"Homogeneity of Sulfur and Carbon Isotope Ratios
34S/32S and 13C/12C in Petroleum", Bull. Amer. Assoc. Petrol.
Geol. 56, 941-949 (1972).
23. Manowitz, B. and Tucker, W., "Determination of Sulfur Isotope
Eatios in the Atmospheric Diagnostics Program at BNL", Trans.
Amer. Nuc. Spc.. 12, 487-8 (1969).
24. Rail, H. T., Thompson, C. J., Coleman, H. J., and Hopkins,
R. L., "Sulfur Compounds in Crude Oil", U. S. Bureau of Mines,
Bulletin 659 (1972).
25. Brody, S. S. and Chaney, "Flame Photometric Detector", J. E. J.
Gas Chromatog. £., 42-46 (1966).
26. Silverman, S. R. and Epstein, S., "Carbon Isotopic Compositions of
Petroleum and Sedimentary Organic Materials", Bull. Amer. Assoc.
Petrol. Geol. /£, 998-1012 (1958).
27. Silverman, S. R., in "Isotopic and Cosmic Chemistry", H. Craig,
S. L. Miller, and G. J. Wassenburg, eds., North-Holland Publishing
Co., Amsterdam, 1964, p. 92.
28. Scalan, R. S. and Smith, J. E., "Improved Measure of Odd-Even
Predominance in Normal Alkanes of Sediment Extracts and Petroleum"
Geochim. et Cosmochim. Acta. 3_4_, 611-20 (1970).
29. Trademark of Consolidated Analysis Centers, Inc., (C.A.C.I.).
93
-------
SECTION XIII
APPENDICES
Title
A. Preparation and Determination of
600 F Bottoms from Crude Oil
B. Determination of Total Nitrogen
in 600 F Bottoms from Crude Oil
by Micro Dumas-Gas Chromatography
C. Measurement of Hydrocarbon Gas
Chromatographic (GLC) Profile of
Crude Oils and Their Residue
D. Isotope Ratio Mass Spectrometer
Instrumentation
E. Determination of Total Organic
Sulfur in Oils. Hydrogenation-
Gravimetric Silver Sulfide Method
F. Preparation of Sulfur Dioxide Gas
from Silver Sulfide Samples for
Isotope Ratio Mass Spectrometry
G. Oxygen Flask Combustion-Gravimetric
Method for Determination of Sulfur
in Organic Compounds
H. Preparation of Silver Sulfide from
Barium Sulfate. Hydriodic Acid
Reduction Method
I. Operating Conditions for Sulfur
GLC Profile Measurement
Phillips Petroleum
Method
7128-AH-l
7234-AN
7304-AG
7302-AF
7303-AZ
6511-AF-l
7301-AZ
Page No,
97
102
109
113
117
12$
132
135
139
95
-------
APPENDICES (Cont'd.)
Phillips Petroleum
Title Method Page No.
J. Silica Gel Separation of 600+F 7204-AZ 140
Bottoms from Crude Oil
K. Isolation of n-Paraffins from 7214-AZ 146
Saturate Fractions by Urea
Adduction
L. Determination of the Relative 7305-AG 148
Weight Distribution of
n-Paraffins Urea-Adducted from
Crude Oil Fractions - C^ to
C - Boiling Range
Jo
M. Published Sulfur Isotope Data - 160
N. Statistical Procedure - 165
96
-------
APPENDIX A
PHILLIPS PETROLEUM COMPANY - METHOD 7128-AH-l*
PREPARATION AND DETERMINATION OF 600+F BOTTOMS FROM CRUDE OIL
I. SCOPE
This method is for the preparation and determination of 600+F bottoms
from crude oil. These bottoms are characterized by measurements of selected
properties useful, in oil pollution source identification.
II. OUTLINE OF METHOD
A sample of crude oil which has been equilibrated at room temperature
is topped under vacuum in a distillation apparatus to an end point of 214 F
(101 C) and 0.15 mm Hg pressure. The residual oil is weighed and the per cent
bottoms calculated. The 600"*"? bottoms are collected for additional measurements.
III. APPARATUS
(a) Centrifuge. A Sorvall Superspeed RC2-B modified for operation at tempera-
tures above ambient, made by Ivan Sorvall, Inc., Newtown, Connecticut, or
equivalent. Teflon-lined centrifuge tubes,.50 ml, with caps.
(b) Pan Balance.
(c) Distillation Unit, Figure 1,
1. Round bottom distillation kettle, 500 ml, two-necked.
2. Thermometer, -10 to 250 C.
3. Delivery Tube.
4. Receiver.
5. Dewar Flask.
(d) Vacuum System. Figure 1.
1. Mega vac Pump, 35 liters per minute.
2. High Vacuum Surge Tank.
3. Controlled Vacuum Surge Tank.
Issued: December 15, 1971
Revised: June 8, 1972
97
-------
V Method 7128-AH-l
Page 2
4. Solenoid Valve, Normally Open.
5. Compound Gauge, 30 psi to 30 inch vacuum.
6. McLeod Gauge.
7. Mercury Regulator Switch,
8. Relay Switch.
(e) Variac.
(f) Heating Mantle. Glascol Hemisphere.
IV. REAGENTS AND MATERIALS
(a) Sodium Sulfate. Anhydrous.
(b) Water Soluble Vacuum Grease. Mix 30 g glycerin, 30 g d-mannitol, 30 g
d-sorbitol and 10 g of water. Heat and stir at a low temperature until
the solution is clear. Cool rapidly with beating to induce fine crystal
formation and to yield a smooth cream.
(c) Benzene.
(d) Acetone.
(e) Dichloromethane.
(f) Liquid Nitrogen.
(g) Boiling Sticks.
(h) Glass Wool.
V. PROCEDURE
(a) Sample Preparation.
The crude oil sample should be equilibrated at room temperature and
well mixed before sampling. Pour about 35 ml of crude oil sample into each
of four Teflon-lined centrifuge tubes containing about 1 gram of anhydrous
sodium sulfate. Adjust the weight of each filled centrifuge tube and cap
so that they all weigh the same (±0.1 g). This may be done by adding or
removing sample,. Place caps on tubes. The. tubes must be balanced in order
to be used in the high speed centrifuge.. Centrifuge the sample at 20,000
rpm for 90 minutes at 30 C.
98
-------
APPENDIX A Method 7128_AH-1
Page 3
Transfer the centrifuged crude oil sample (4 tubes) into a tared
distillation kettle. During the transfer avoid adding any of the material
settled out at the bottom of the centrifuge tube. Reweigh the distilla-
tion kettle and add a boiling stick.
(b) Distillation.
!• Preparation of Vacuum System. Using Figure 1 as a guide, adjust the
following valves as described;
(a) Rotate the regulator until contact is made between the mercury
column and the wire electrode.
(b) Activate the relay to close the solenoid valve.
(c) Open Valve 3, open Valve 4, and open Valve 2.
(d) Close Valve 1 and close Valve 5.
(e) Turn on vacuum pump.
2, Preparation of Distillation Unit. Connect the receiver, delivery tube
and distillation kettle together as shown in Figure 1 using water-
soluble grease on all glass joints. Cpen stopcock of the receiver
and open Valve 5 so that a flow of nitrogen may displace the air in
the system and be vented out the thermometer well opening. After a
couple of minutes of purging, add liquid nitrogen to the Dewar flask
and continue the nitrogen purge for 3 more minutes. Close Valve 5 and
insert the thermometer into the well using water-soluble grease on the
joint. Activate the relay switch to pull a partial vacuum on^the
system through the control vacuum surge tank. Place the heating
mantle under the kettle and insulate the kettle and neck with glass
wool. Leave an opening so that the inside of the kettle may be seen
with the aid of a flashlight. Set the variac at 50 and reduce the
pressure inside the system slowly by activating the relay. As the
temperature increases and the pressure decreases, boiling will increase.
Do not allow the sample to "bump" over into the receiver by reducing
the pressure too rapidly. If necessary, nitrogen may be admitted into
the system through Valve 5 to increase the pressure. Do not let boiling
stop, however, or bumping is apt to occur on start up. When the pressure
is reduced sufficiently, the McLeod gauge may be used to measure the
pressure. Near the end point it is necessary to open Valve 1 to obtain
a "hard" vacuum. The end point is reached at 214 F (101 C) and 0.15 mm
Hg when boiling virtually ceases and no further distillate is collected.
When the end point is reached, close Valve 1 and activate the
relay. Admit nitrogen into the system slowly by opening Valve 5. When
the system is at atmospheric pressure, remove the thermometer and remove
the distillation kettle from the delivery tube. Wipe off the stopcock
grease and weigh the distillation kettle.
99
-------
APPENDIX A Method 7128-AH-l
Page 4
Transfer 1-2 grams of the 600 F bottoms to a small vial and seal
under nitrogen. Store these vials at 40 F. Transfer the remaining
600 F bottoms to a clean wide-mouthed sample jar.
Glassware is cleaned by rinsing with dichloromethane, benzene,
water; and acetone and drying in an oven.
VI. CAICUIATION
Weight per cent bottoms (600+F) = residue x 100
0 ^ grams sample
100
-------
PHILLIPS PETROLEUM COMPANY— METHOD 7128 —AH
PRESSURE GAUGES
COMPOUND - 30 PSI TO 3O IN VAC
MC LEOD - 5 TO 0.01 MM
o
I—*
CONTROL VAC. SURGE TANK
'"d
0
U
H
FIGURE 1
SCHEMATIC DIAGRAM OF APPARATUS
-------
PHILLIPS PETROLEUM COMPANY - METHOD ?234-AN *
DETERMINATION OF TOTAL NITROGEN IN 600+F BOTTOMS
FROM CRUDE OIL BY MICRO DUMAS-GAS CHROMATOGRAPHY
I. SCOPE
This method is for determination of 0.03 per cent and higher concentra~
tions of total nitrogen in 600"*"F bottoms from crude oil.
II. OUTLINE OF METHOD
A weighed sample is placed in a pyrolysis-combustion tube and is
oxidized in an atmosphere of helium and oxygen. Water and carbon dioxide are
removed by trapping at liquid nitrogen temperature. Nitrogen oxides are reduced
to nitrogen with hot copper. The nitrogen is collected on molecular sieve at
liquid nitrogen temperature and is determined quantitatively by gas chromatography
using thermal conductivity detectors.
III. APPARATUS
1. Micro Dumas - Gas Chromatography Apparatus. A schematic diagram is shown in
Figure 1. The basic components of this apparatus are:
(a) Combustion Tube. Quartz, 30 in, long- x 1/2 in. O.D. with 2 mm wall
thickness.
(°) Furnace* Electric, 900 C, for combustion tube.
Pyrex, heavy-wall, 22 in. long x 1/2 in. O.D. Filled with
copper and heated to $50 C by electric furnace.
(d) Trap for C02 and HgO, Glass, 25 ml.
(e) Nitrogen Trap. Attached to Perkin-Elmer valve. See Figure 2.
Manometer. Dual, 30-in. , mercury-filled.
Constant Temperature.. Bath . E. H. Sargent Co. S-84810 except bath made
of stainless steel and insulated. General Electric Company silicone
fluid SF 96 (100) used as bath liquid,
Column. 13X Molecular Sieve.
(i) V-F Converter, Vidar 240.
(j) Electronic Counter. Hewlett Packard 5512A.
(k) Bridge^ Power Supply. Applied Automation Model 11.
*Issueds December 20, 1972.
-------
APPENDIX E Method 7234-AN
Page 2
(l) Thermistors. Fenwal G-112j Fenwal Electronics, Inc., Framingham, Mass.
W Pressure Regulator. Sub-AtmosjghgriR.. Moore Model 43 j Moore Products Co.,
Spring House, Pa.
2. Auxiliary Apparatus
(a) Combustion Boats. Platinum, 39 am long.
(b) Balance. Six-place, fast-weighing.
(c) Dewar Flasks.
(d) Pipets . Pasteur, disposable.
(e) Spatula. Micro.
I?. REAGENTS AND MATERIALS
1. Acetanilide. N.B.S. No. 141 bj National Bureau of Standards, Washington, D. C.
2. Cuprin. A specially sized and treated metallic copper. Fisher Scientific Co.
13-109-10.
3* Cuprox. A specially sized and treated copper oxide. Fisher Scientific Co,
13-109-15.
4. Helium.
5. Liquid Nitrogen.
6. Molecular SieTe. 131, screen 12/30 5 Coast Engineering Co., Redondo Beach,
California.
7. Oxygen. Ultra high purity! Matheson Co., Inc., Joliet, Illinois.
8. Quartz Wool.
9- Silver Gauze. Approximately 80 mesh.
V. PROCEDURE
1. Preparation of Apparatus
Set up the apparatus as shown in Figure 1.
place.
103
-------
APPENDIX B Method 7234-AN
Page 3
(b) Add silver gauze and Cuprox to the quartz combustion tube as shown
below.
-11 in.-»
(c) Fill the pyrex U-tube with Cuprin so that all of the Cuprin is in the
jfurnace zone when the tube is in place. Use quartz wool plugs to hold
the filling in place.
(d) Prepare and attach the nitrogen trap as shown in Figure 2, The trap
is made of 1/4-in. stainless steel tubing.
(c) Add 13X molecular sieve to 5 ft. of 1/4-in. copper tubing and 6 in. of
charcoal to the outlet end. Use silver gauze to hold the material in
place. This column is shown in Figure 1 as 10.
2. Determination ofBlank
After the apparatus is set up as shown in Figure 1, adjust the column
helium flow as shown. Adjust the combustion helium flow to a fast rate (~100 cc/
minute). After 10 minutes, place Dewar flasks filled with liquid nitrogen under
traps 21, 22, and 6. After 15 minutes, place Dewar flask filled with liquid
nitrogen under trap 8. Wait 2 minutes and adjust the combustion helium flow to
1 cc/minute. Adjust the oxygen flow to 1 cc/minute. Light the gas burne.r and
slowly move the burner down the combustion tube to the electric furnace.
(Approximate burn time is 20 minutes.) Turn off the burner, turn off the oxygen,
and increase the combustion helium flow to about 100 cc/minute. Sweep the system
for 15 minutes. Turn trap 8 to the inject position. Adjust inlet column helium
pressure to 20 pslg with valve 19 and indicate by marking on the manometer so
that this pressure setting may be repeated. Adjust the outlet column helium
pressure with valve 15 so that the outlet pressure is 5 mm below atmospheric
pressure. Indicate on the manometer by marking so that this pressure setting
may be repeated. Adjust the bridge circuit so that a very slow count is
indicated on the electronic counter. Remove the Dewar flask containing liquid
nitrogen from nitrogen trap 8 and replace with a Dewar flask containing water
at room temperature. Record the number of counts produced by the nitrogen peak.
Turn the trap to the collect position. Determine blanks periodically in order
to get a representative value.
NOTE: Blank determinations should represent less than 100 micrograms
of nitrogen.
104
-------
APPENDIX D Method 7234-AN
4
3. Preparation of Calibration Curve
Weigh 1, 3, 5, 10, 20 mg (6-place balance) of acetanllide (NK)in
separate platinum boats. Place a piece of dry ice on top of the combustion
tube in the area where the sample boat is to be placed. This cools the tube
to eliminate evaporation of the sample during the sweep time. Place the sample
boat in the combustion tube and sweep the system for 15 minutes with helium at
100 cc/minute to. remove air. Place a Dewar flask containing liquid nitrogen
under nitrogen trap 8. Wait 2 minutes and adjust the combustion helium flow to
1 cc/minute. Adjust the oxygen flow to 1 cc/minute. Remove the dry ice from
the tube. Light the burner and slowly move the burner toward the sample. The
sample is volatilized and oxidized. Since excess oxygen is removed by the
copper U-tube, the amount of oxygen added to the combustion system is regulated
depending on the amount and type of sample being analyzed. If samples are
volatilized too rapidly, incomplete combustion products are formed on the outlet
end of the combustion tube. After sample combustion is complete (ash may be
present in the boat), turn the oxygen flow off. Increase the comoustion helium
flow to 100 cc per minute and sweep for 15 minutes. Inject and read the number
of counts produced by the nitrogen peak as described in 2. Determination of
Blank.
Calculate a factor (F) for each of the standard samples from the
following formula.
_ ^ mg acetanilide. NBS x per cent nitrogen in acetanilide
= (counts - blank counts) 100
Plot F vs net counts for each standard. Draw a curve through the points which
best fits the points.
4. Analysj^of Sample
Weigh about 0.5 gram of sample in a tared platinum boat. Place the
boat in the cooled combustion tube and proceed to sweep, combust, sweep, inject
and read the number of counts produced by the sample as described in Sections 2
and 3 above. At the end of the day, remove trap 6 and discard the water and C02<,
NOTE: The combustion tube and U-tube usually last 2 days under constant
use.
VI. CALCULATIONS
Calculate the nitrogen concentration in the sample as follows:
Nitrogen, Wt. % = JQ~^
F . Factor read from calibration curve corresponding to net counts of
sample.
N = Net counts of sample.
W = Grams of sample.
105
-------
APPENDIX B
Method 7234-AN
Page 5
VII. PRECISION
A statistical evaluation of data from duplicate determinations on
twelve samples showed;
Range, Degrees of Standard
Wt. % N Freedom Deviation Repeatability
0.03 to 0.7 12 0.019 % N 0.061 % N
Duplicate results should be considered suspect if they differ by
more than the repeatability value shown (95 per cent confidence level).
106
-------
PHILLIPS PETROLEUM COMPANY
METHOD 7234—AN
COLUMN
HE
NOTES: ^^^ -_
(A) ITEMS (V) AND Q_o) ARE PLACED IN A CONSTANT
TEMPERATURE BATH AT 71C.
(B) CONNECTING TUBING IS 1/8 IN. STAINLESS
STEEL.
LEGEND
O
ROTAMETER
Z\ COMBUSTION TUBE
~3\ GAS BURNER
FURNACE, 900C
s FURNACE, 5SOC, FOR U-TUBE
Vj TRAP FOR CO2 , H2O
PERKIN—ELMER VALVE
M3J NITROGEN TRAP
Csj THERMAL COND. DETECTOR
no) MOLE SIEVE COLUMN
m) BRIDGE, POWER SUPPLY
(fa) V. F. CONVERTER, COUNTER
(ra) STRIP RECORDER
n~4) SURGE TANK
my SUB—ATM. CONTROL VALVE
ney PRESSURE REG., e PSIG
(yn OUTLET PRESSURE MANOMETER
nm INLET PRESSURE MANOMETER
(T5) PRESSURE REG., 2O PSIG
(20) PRESS. REG. VALVE, 3O PSIG
(2y PURIFICATION TRAP
(22) PURIFICATION TRAP
tJ
FIGURE 1
SCHEMATIC FLOW DIAGRAM
-------
FIGURE 2
NITROGEN TRAP
/r~ -\
1 1
t :;
1 1 1
1 I
\
PERKIN—ELMER VALVE
u
EPOXY
SEALS
IMMERSION
LEVEL OF LIQUID
NITROGEN WHEN
COLLECTING
1 2/ 5 SS
BALL JOINTS
SILVER GAUZE
I 3X MOLECULAR SIEVE
108
-------
APPENDIX C
PHILLIPS PETROLEUM COMPANY - METHOD 7304-AG*
MEASUREMENT OF HYDROCABBON GAS CHROMATOGRAPHIC (GLC) PROFILE OF
CRUDE OILS AND THEIR RESIDUES
I, SCOPE
This procedure is for obtaining chromatograms of crude oils and
crude oil residues for use in source identification studies.
II, OUTLINE OF METHOD
The sample (prepared according to Phillips Method 7128-AH-l) is
placed in a glass sample holder and introduced into a' vaporization chamber
through a specially-designed injection system. The sample is flash vaporized
and carried into the gas-liquid partition chromatography column with helium
carrier gas. The hydrocarbon components are separated in the approximate
order of their boiling points using temperature programmed gas chromatography.
III. APPARATUS
(a) Chromatograph. Any chromatograph which is equipped to take 1/4-in. diameter
columns, has temperature-programming capability to 350 C, and is equipped
with a flame ionization detector. A Perkin-Elmer Model 880 was used in
the development of this method.
(b) Strip-Chart Recorder. 1-mv range and 2-second pen speed.
(c) Injection System. Designed by Phillips Petroleum Company. Includes
injection system, vaporizer chamber, sample holder, and adapter and
supports for installation on a Perkin-Elmer Model 880 gas chromatograph.
The over-all sketch of this equipment is shown in Figure 1. Detailed
drawings of this injection system are available from Phillips Petroleum
Company under Drawing No. RS2-4053 (Note l).
NOTE. 1; Basic design for above apparatus was adapted from
~~Ramsdale, S. J., Wilkinson, R. E., J. Inst. Petroleum .5Jt,
No. 539, 326 (1968).
(d) Temperature Controller, Any controlling pygcm|ter which will control the
temperature of the vaporizer chamber at 538 C - 5 C.
IV. REAGENTS AND MATERIALS
(a) Helium.,
(b) Hydrogen.
Issueds January 1?> 1973
109
-------
APPENDIX C Method 7304-AG
Page 2
(c) Cylinder Air.
Chromosorb P. 40-60 Mesh, available from Johns Manville Company, Celite
Division, New York, New York.
(Q) Tubing. 1/4-inch O.D., stainless steel.
(f) Poly-m-Phenoxylene (PPE-201. Available from Varian Aerograph, Walnut
Creek, California.
(g) Chloroform. Reagent Grade.
V. PRELIMINARY PREPARATIONS
(a) 5 Wt. % PPE-20 on Chromosorb P. Weigh 1 g of PPE-20 into a 100-ml
beaker and add 80 ml of chloroform. Stir the mixture until completely
dissolved and add the mixture to 19 g of Chromosorb P (40-60 mesh) in
an 11-cm evaporating dish. Carefully evaporate the chloroform using
low heat on a hot plate. Continue evaporation until the mixture appears
dry.
(b) Column. Clean a 10-cm section of 1/4-inch O.D. stainless steel tubing
and a 5-foot section of stainless steel tubing with a 100-volume wash of
acetone and air-dry the tubing. Fill both lengths of tubing with the
prepared packing (Section V,(a) and plug the ends of the tubing with
glass wool. Connect the 10-cm pre-column section to the 5-foot analyzer
section. Mount the column in the chromatograph so the pre-column
section is at the inlet end.
(c) Column Conditioning. With the exit end of the column disconnected from
the detector the column is conditioned as follows:
1. Establish a flow rate of 30 cc/min. of helium carrier gas through
the column.
2. Program the column oven temperature from ambient to 300°C @ 6°/min.
and hold at upper limit for two hours.
3. Increase column oven temperature to 380°C and maintain for 24 hours.
4. Cool column to ambient temperature and remove 5-foot analyzer
section from pre-column section. Turn 5-foot section end for end
and reconnect to pre-column and connect to detector and column is
ready for use.
VI. PROCEDURE
Set the supply gases for optimum flame detector operation, the
carrier gas flow at 30 cc/min. at ambient temperature, the controlling
pyrometer to 538 C, the initial oven temperatures to 150°C and the temperature
110
-------
APP1SHEIX C Method 7304-AG
Page 3
program rate control to 6°C/min. The instrument has equilibrated when the
recorder baseline is stable.
Place the sample in the glass sample holder, using a small glass
rod drawn to a medium point, by just dipping the tip of the rod in the
sample and then rotating the tip around the inside wall of the sample
holder (Note 2).
NOTE 2; The viscosity of the sample determines the amount
of sample clinging to the rod and repeatable sample
sizes are difficult to achieve.
In Figure 1, open and close valves A and B in succession, allowing
the sample to fall into the vaporization chamber without loss of sample
vapors or carrier gas from the system. As soon as the sample holder drops into
the vaporization chamber, initiate the temperature program controller and
the separation of components over the temperature range of 150 to 350°C at
a program rate of 6 /min. Record the chromatogram on the strip chart recorder.
Retrieve the glass sample holder containing an enclosed iron core from the
vaporization chamber with a magnet.
The pre-column section retains the high molecular weight material
and should be changed after every 20 runs.
VII. DISCUSSION
Duplicate analyses, carried out several weeks apart, showed the
method to be repeatable over extended periods of time.
The hydrocarbon GLC profiles for crude oil residues of known
source are visually compared with samples of unknown source. Conclusions
as to identity are based on the similarity or differences observed.
Ill
-------
APPENDIX C
INSULATED
VAPORIZATION
CHAMBER
CARRIER GAS
INLET
INLET TO
CHROMATOGRAPH
SUPPORT LEG
FIGURE 1
INJECTION SYSTEM ADAPTED
FOR PERKIN-ELMER 880 CHROMATOGRAPH
112
-------
APPENDIX D
ISOTOPE RATIO MASS SPECTROMETER INSTRUMENTATION
The isotope ratio mass spectrometer in this laboratory was
built by Avco Corporation in Tulsa, Oklahoma. Two fundamental design
characteristics make possible extremely precise isotope ratio measure-
ments (1): (a) a sample inlet system that permits rapid, alternate
introduction of sample gas and reference gas and (b) a dual collection
system which permits electrical measurement of ratio signals from the
simultaneous collection of major and minor isotopes. The mass analyzer
is first order direction focusing with 35 cm radius of curvature, 90°
angle of deflection and 12° beta axis focusing. The source and analyzer
are evacuated with a 500 liter/second cold cathode sputter-ion pump
—9
which attains 10 7 torr in a clean system. The high voltage power
supply is a Fluke Model 410B. The electromagnet is programmable from
one to ten kilogauss and adjustable to within 0.5 gauss. The low
voltage, high current-type magnet power supply is stable to one part in
100,000 for long term operation. The readout system includes dual
vibrating reed electrometers and null ratio difference Kelvin Varley
divider circuit, Gary Model 3125700, and a Sargent Model MR recorder.
Carbon isotope ratio measurements can be made precisely and
conveniently by converting the carbon to carbon dioxide. One advantage
of this is that carbon dioxide can be cleanly pumped out of the system
in a short time so that "memory" effects are small. Phillips' geo-
chemistry research program also requires isotope ratio measurements for
sulfur, which is most conveniently run in the form of sulfur dioxide.
Unfortunately, sulfur dioxide has a tendency to adhere to several sub-
stances which may exist within the mass spectrometer and inlet system
and cause pumping difficulty. Furthermore, we have found that the
initial instability of the output signal when sulfur dioxide is intro-
duced to a clean mass spectrometer gradually improves with time. This
113
-------
indicates that the filament becomes "conditioned" in an atmosphere of
sulfur dioxide.
Alterations were made in the sample handling region, outlined
by dashed lines in Figure 1, so that carbon and sulfur isotope measure-
ments can be readily made with the same mass spectrometer. The altered
arrangement, shown in Figure 2, consists essentially of two separate
identical inlet systems, one for carbon dioxide and the other for sulfur
dioxide. When using the instrument to measure carbon isotope ratios,
valves A and C are open and valves B and D are closed. Each system
includes four single action solenoid valves, connected as shown, for
alternately introducing sample and standard to the mass spectrometer.
For each measurement, the sample is introduced into the spectrometer
three times so that six measured differences between sample and standard
are obtained. The 6 values are calculated by using the average of these
six instrumental measurements.
The precision for measuring carbon or sulfur isotope ratios
is affected by sample preparation and reproducibility of the mass
spectrometry measurement. Carbon dioxide suitable for mass spectro-
metry is prepared from organic carbon samples by a combustion method.
Sulfur dioxide is prepared by pyrolysis of silver sulfide at 1000°C
with copper oxide as the oxidant.
The over-all standard deviation for carbon isotope ratio
determinations, calculated from results of 33 independent sample
preparations and mass spectrometer measurements for a single laboratory
working standard is 0.06. The standard deviation of instrumental error
for carbon is 0.05, determined with 122 degrees of freedom by running
the same carbon dioxide sample as both sample and standard.
114
-------
RECORDER
VJt
LEAK ADJUSTING
DEVICES
MAGNET CURRENT
POWER SUPPLY
DECADE BALANCE
RESISTOR
VIBRATING REED
ELECTROMETER
NO. 1
V
VIBRATING REED
ELECTROMETER
NO, 2
•xi
w
M
X
HG
MANOMETER
FIGURE 1
ISOTOPE RATIO MASS SPECTROMETER SYSTEM
-------
GAS STANDARD IN
s
TO ION SOURCE OF MASS SPECTROMETER
o
s
r
i \
TYPICAL
c/\i nirtir\
SOLENOID
VALVE
1 f
Q
!
0
p
5
2
^_
Q
u
\r
LEAKS J
GAS
UXEZ3*- SAMPLE
IN
ION
VACUUM
PUMP
CD
CAPILLARY 1
GAS J1!
STANDARD —€ZZKZ, )
IN
j
LEAKS
LEAK
ADJUSTING
DEVICES
GAS
SAMPLE
IN
C02 INLET SYSTEM
S02 INLET SYSTEM
M
X
FIGURE 2 - DUAL SAMPLE INLET SYSTEM
-------
AFPE1IDIX E
PHILLIPS PETROLEUM COMPANY - METHOD 7302-AF*
DETERMINATION OF TOTAL ORGANIC SULFUR IN OILS.
HYDROGENATION-GRAVIMETRIC SILVER SULFIDE METHOD.
I. SCOPE
This method Is for the gravimetric determination of 0.5 to 10 per
cent total, organic sulfur in oils. The silver sulfide produced is suitable
for use in isotope ratio mass spectrometric studies. Most inorganic sulfur
including barium sulfate, if present, is measured also as organic sulfur.
11 • OUTLINE OF iMETHOD
A weighed amount of oil is vaporized in an atmosphere of humidified
hydrogen and passed over platinum gauze at 1100 C in a combustion tube.
Organic sulfur is converted to hydrogen sulfide which is absorbed in a silver
nitrate solution. The precipitated silver sulfide is filtered, washed, dried,
and measured gravimetrically. The hydrogen ion concentration is titrated as
an internal check of the results.
III. APPARATUS
(a) Hydrogenation (See Figure 1 for schematic diagram).
I- Flowmeter. For measuring hydrogen flow at 15 liters per hour, with
a sensitivity to 2 per cent change in flow.
2. Combustion Tube. Fused silica, 20 mm I.D., 60 to 75 cm long, with
24/40 inter joint on the inlet and a reduced end at the exit.
3» Tube Furnace. 9 to 12 inches long and capable of maintaining a
temperature of 1100 ± 10°C in its middle 3-inch section.
4« Pyrometer. Capable of indicating the temperature at 1100 C within -10 C.
5. Combustion Boat. Platinum or porcelain, capable of holding 3 ml of
sample„
6. Absorption Flask. 50-ml (see Figure 2).
7. Water Bubbler, (see Figure l).
8. Analytical Balance.
*,
Issued: January 10, 1973
117
-------
APPENDIX E Method 7302-AF
Page 2
(b) Gravimetric Measurements.
1. Crucibles. Gooch Type Low Form, medium porosity, fritted disc, 30-ml,
2, Drying Oven.
3« Filtering Flask. 500-ml.
(c) Titration.
1. p_H Meterand Electrodes. Capable of determining end point of acid-
base titration.
2. Magnetic Stirrer.
3. Buret.
IV. REAGENTS AND MATERIAL
All reagents are A.C.S. reagent grade or equivalent unless otherwise
noted. References to water indicate distilled or deionized water.
(a) Hydrogenation.
1. Platinum Catalyst. Platinum-rhodium (90:10), 80-mesh gauze, 3 in. x
12 in.
2. Hydrogen. 99.5 per cent purity or better. It shall meet the require-
ment of Section V,(a) Hydrogenation, (l) c, d. A high-pressure
cylinder shall be fitted with a pressure regulator to reduce the
delivery pressure to 1 to 5 psi.
3. Sulfur Absorber. Activated charcoal packed absorption tube, approxi-
mately 3/4-in. (2-cm) diameter by 10 in. (25 cm) long with tubing
connections on each end. (Note l).
NOTE 1: This item is not required if hydrogen purity
permits absorbance specifications of Section
V,(a) Hydrogenation, (l) d, to be met,
4" Silver Nitrate Solution, 0.1 N. Dissolve 16.99 g of silver nitrate in
water and dilute to one liter.
5. Nitrogen. Prepurified grade or better.
(b) Gravimetric Measurements.
1. Ammonium Hydroxide,, Concentrated.
118
-------
APPENDIX E Method 7302-AF
Page .3
(c) Titration.
1« Sodium Hydroxide. Standard 0.1 N. Standardized accurately against
standard acid.
?. PROCEDURE
(a) Hjdrogenation.
1. Preparation of Apparatus.
a. Wind the platinum-rhodium gauze into a tight roll 3 in. long and
insert into the center of the refractory tube. Assemble all
parts of the apparatus according to Figure 1 except the absorption
flask, and connect with glass or aluminum tubing and a minimum of
rubber or polyvinyl chloride tubing. Connect the hydrogen supply
and fill the water bubbler about half full of distilled water.
b. Connect the outlet of the combustion tube to a vent line through
''which combustible gases can be discharged safely. Purge air
from the system with nitrogen (Note. 2').,and- start the flow of
hydrogen. Adjust the hydrogen,..flow rate to 15 liters per hour
and heat the furnace to 1100 C.
NOTE 2: Mixtures of hydrogen and air may explode
in contact with platinum, even at ambient
temperatures. Therefore, air must always
be. purged from the system with nitrogen or ^
other inert gas, before starting the hydroge'n
flow.
c.
Place 30 ml of silver nitrate solution in a 50-ml absorption flask.
Insert the stopper with bubbler tube. Disconnect the outlet of
the refractory tube from the vent line and connect it to the inlet
of tSe absorption flask. Connect the outlet of the absorption
flask to the vent line. Continue passing hydrogen through the
apparatus at a rate of 15 liters per hour for 50 minutes.
Disconnect the absorption flask but continue the passage of
hydrogen.
ffi srs.issr sr KISS,™ :?«;•. ..
119
-------
APPENDIX E Method 7302-AF
Page 4
2. Analysis..
a, Weigh accurately 0.5 to 2.0 g of sample into a tared boat.
(Note 3)-
Be careful that the amount of sample used
does not produce more hydrogen sulfide than
the silver nitrate solution can absorb. It
may be necessary to make a preliminary
analysis of the oil to determine the maximum
amount of oil that can be used.
Insert the boat into the refractory tube. Prepare an absorption
flask and connect it to the apparatus as in Section V,(a), (l), c.
Purge the system for 5 minutes with nitrogen.
b. With the hydrogen flowing at 15 liters per hour, heat the sample
by passing current through nichrome resistance wire wrapped
around the combustion tube in the region of the boat. The sample
should be heated and volatilized at such a rate that carbon is
not formed at the outlet end of the refractory tube. Some samples
may require 8 hours for volatilization.
c. At the finish, some carbonaceous residues may remain in the boat.
Using a Meker burner at maximum heat, heat the combustion tube in
the area of the boat and in the area between the boat and the
furnace to red heat. Continue the hydrogen sweep for 20 minutes.
Purge hydrogen from the system with nitrogen and remove the
' absorption flask from the train. Pass air through the refractory
tube for 5 minutes to remove any carbon collected on the platinum
gauze. Purge the system again with nitrogen before starting
hydrogen flow for the next run (Note 4).
NOTE k' To prevent an explosion due to slight valve
leakage, the air and hydrogen' supplies must
not be connected to the system at the same
time. Physically disconnect and separate one
supply from the system before connecting the
other.
d. Save the silver nitrate solution in the absorption flask for the
gravimetric measurements and titration.
(b) Gravimetric Measurements.
1. Preparation of Apparatus.
a. Wash filtering crucible with water, ammonium hydroxide and water
using vacuum filtration. Dry in vacuum oven at 60°C for 2 hours.
120
-------
APPENDIX E Method 7302-AF
Page 5
b. Cool to room temperature in a desiccator. Determine tare weight.
2. Filtering. Washing and Drying of Silver Sulfide.
a. Filter the silver nitrate solution from Section V,(a), 2 through
the tared filtering crucible. With the aid of a policeman,
quantitatively transfer the silver sulfide into the crucible.
Save the filtrate in the filtering flask for titration. Section
V,(a), 2.
b. Use another filtering flask to collect the three 10-ml water,
two 10-ml concentrated ammonium hydroxide solution, and three
10-ml water washings of the silver sulfide. Discard the washings,
Dry the crucible in a vacuum oven at 60°C for .2 hours and cool
to room temperature in a desiccator. Determine the gross weight,
(c) Titration.
1. Titration of Filtrate.
a. Transfer filtrate into beaker and titrate with 0.1 N standardized
sodium hydroxide, using a pH meter to determine the maximum
inflection point (approximately 5.5 pH).
b. Record the volume of sodium hydroxide used.
VI. CALCULATIONS
(a) Gravimetric Measurements.
* A 12.94
Sulfur, weight per cent = j—-*
where: A = grams silver sulfide
g = grams of sample
(b) Titration.
* BH 1.603.
Sulfur, weight per cent = g
wheret B = ml of NaOH solution used
N » normality of NaOH solution
g = grams of sample
121
-------
APPENDIX E Method 7302-AF
Page 6
VII. PRECISION
For crude oils containing 1-4 per cent sulfur, the following
precision data were obtained:
Method Standard Deviation Degrees of Freedom Repeatability
Gravimetric 0.0464$ 16 0.14$
Titration , 0.109$.. 16 0.33$
ASTM D1552 0.0914$ 10 0.29$
ASTM D1552: "Standard Method of Test for Sulfur in Petroleum Products
(High Temperature Method)", 1972 ASTM Standards, Part 18.
122
-------
PHILLIPS PETROLEUM COMPANY — METHOD 73O2-AF
PYROMETER
T-BORE
STOPCOCK
PLATINUM GAUZE
VENT
ABSORPTION
FLASK
FIGURE 1
SCHEMATIC DIAGRAM OF HYDROGENATION APPARATUS
-------
APPENDIX E
u.
<
<\l
o
X
III
D.
O
O
s
D
UJ
_l
o
K
tf)
0.
J
J
0.
•24/ 40 INTERJOINT
50-ML MARK
SCALE: i: i
FIGURE 2
50-ML ABSORPTION FLASK
124
-------
APPENDIX F
PHILLIPS PETROLEUM COMPANY - METHOD 7303-AZ
PREPARATION OF SULFUR -DIOXIDE GAS FROM SILVER SULFTDE SAMPLES
FOR ISOTOPE RATIO MASS SPECTROMETRY
I. SCOPE
This method is for the preparation of sulfur dioxide gas, without
fractionation of the sulfur isotopes, from samples of silver sulfide prepared
from various sources. The gas must be pure in order for the mass spectrometer
to operate properly.
II. OUTLINE OF METHOD
A sample of silver sulfide is mixed with copper oxide and roasted
in an evacuated quartz tube at 1050 C. The sulfur dioxide is purified,
measured and transferred to an evacuated, glass sample bomb for measurement
on the isotope ratio mass spectrometer.
III. APPARATUS
(a) Vacuum Line-Combustion Apparatus. Figure 1.
(b) Tube Furnace. Capable of maintaining temperature of 1050 - 20°C.
(c) Pyrometer. Capable of measuring the temperature at 1050 C within -20 C.
(d) Sample Tube. Fused silica, 9 mm O.D., sealed at one end, 5 cm long.
(e) Magnetic Sample Carrier. Fused silica tube with Alnico V magnet sealed
in one end. Opening on other end large enough to.accept..sample tube.
(f) Analytical Balance.
Pgwar Flask. Pyrex, 1 quart (2 required).
Dewar Flask. Pyrex, 1/2-pint (4 required).
Glass Sample Bulbs. 25 cc.
IV, REAGENTS AND MATERIALS
(a) Cuoric Oxide. Powder, ACS Reagent Grade.
(b) Cnppgr Metal Turnings. Washed with acetone and dried before use.
Issued: January 12, 1973
125
-------
APPENDIX F Method 7303-AZ
Page 2
(c) Toluene Slush Bath.-95°C. Cool toluene in a dewar flask by careful
addition of liquid nitrogen until a liquid-solid slurry of even
consistency is formed. Stir while cooling.
(d) Pentane Slush Bath, -130°C. Cool n-pentane in same manner as described
above,
(e) Liquid Nitrogen.
V. PROCEDURE
(a) Preparation of Vacuum System. Turn on all pumps, gauges and utilities.
Evacuate the entire system to a pressure of 1 x 10-* torr or less.
Fill dewar flask around the trap nearest to the diffusion pump with
liquid nitrogen. Adjust furnace so that a temperature of 1050 C is
maintained inside central portion of combustion tube.
(b) Sample Preparation. Obtain tare weight of sample tube using an
analytical balance. Carefully weigh 30-40 mg of finely divided Ag_S
and 120-160 mg CuO powder into the tube. Mix the two powders by
carefully rotating and shaking the tube. Place small plug of copper
turnings in open end of sample tube.
(c) Introduction of Sample into Vacuum System.
1. Close stopcock 2,
2. Open stopcock 3.
3. Remove end cap.
4. Place sample tube in magnetic sample carrier. (If sample tube
is too loose, quartz wool may be inserted to secure tube.)
5. Place carrier with sample in cool portion of combustion tube.
6. Replace end cap.
(d) Evacuation of Combustion Tube Using Rough Vacuum.
7. Close stopcocks 1 and 4»
8. Open stopcock 5«
9. Close stopcock 3.
10, Open stopcock 2, wait 1-2 minutes.
126
-------
APPENDIX F Method
Page 3
(e) Evacuation of Combustion Tube Using Hard Vacuum^
11. Close stopcock 5«
12. Open stopcock 4» Wait until vacuum of 1 x 10~^ torr or better
is attained.
(f) Preparation of Sulfur Dioxide.
13, Place dewar flask containing liquid nitrogen on trap 1.
14. Slowly slide sample into hot zone using an external magnet.
Pressure on thermocouple gauge will rise to a maximum of about
100 microns then drop back to zero. After pressure drops (0
production ceases) wait approximately. 5 minutes*
15. Slowly slide sample holder into cool zone.
16. Close stopcock 2.
(g) Purification of Sulfur Dioxide. Part I, Removal of less volatile
impurities such as H 0,.SO , etc...
17. Place dewar flask containing liquid nitrogen on trap 2. Remove
liquid nitrogen from trap 1. Allow transfer of gases to continue
until trap 1 reaches room temperature.
18. Close stopcocks 4 and 7.
19. Remove liquid nitrogen from trap 2 and place toluene slush bath
on trap 2 immediately. (Check to make sure slush bath is of
proper consistency.)
20. Close stopcocks & and 9.
21. Open stopcock 7«
22. Place dewar containing liquid nitrogen on cold finger (trap 3).
Allow transfer to continue for 3 minutes.
23, Close stopcock 11.
24. Warm cold finger (trap 3) to room temperature. Read pressure on
manometer.
127
-------
APPENDIX F Method 7303-AZ
Page 4
(h) Purification of Sulfur Dioxide. Part II. Eeomval of more volatile
impurities such as CO , etc.
-------
APPENDIX F Method 7303-AZ
Page 5
44. Open stopcocks 8 and 4.
45. Return to procedure (b) for continuation.
VI. CALCULATIONS
Data recorded includes:
1. Wt. Ag2S used.
2. Wt. CuO used.
3. Pressure S02 + more volatile materials.
4. Pressure SO .
The per cent yield of S02 is calculated. The Theoretical yield of SO
(pressure in mm) is calculated from the known weight of silver sulfidi
(Note 1).
V-S^T^J Q.n < pressure SO (mm) .__.
lie Id SO-. % = 77 11: "<-"" 2 — 7—r x 100
2> theoretical pressure (mm)
NOTE 1: The calculation of theoretical pressure of SCu is
lengthy and will only be summarized. The volume of
the cold finger and manometer can be measured using
standard techniques. The volume will depend on the
pressure of the system, and will have an equation of
the form V «= V -f aP. Assuming that SO- behaves as an
ideal gas at tne pressures encountered, the pressure of
SO^ gas expected from a given weight of Ag~S can be
calculated. Since these calculations involve the
solution of a second degree equation, it is easiest
to utilize a computer to generate a table of theoreti-
cal pressures of SO vs. weight of Ag S. Such a table
has been generated and should be used in the above
calculation. CAUTION; If the appropriate weight of
Ag9S is not in the table or if the volume of the
appropriate portion of the vacuum line has been
changed since the calibration was made, consult
technical supervisor before proceeding.
129
-------
APPENDIX F Method 7303-AZ
Page 6
VII. PRECISION
An evaluation of the data from 82 consecutive conversions of
different Ag?S samples showed:
Maximum yield - 101$
Minimum yield - 79$
Average yield - 93«5$
Standard Deviation - 4»48
SO^ samples obtained in less than 90 per cent yield should be
considered suspect.
130
-------
V-0
PHILLIPS PETROLEUM COMPANY — METHOD 73OS—/
TO COLD TRAP
AND MERCURY
DIFFUSION PUMP
(
\
U
MAGNET
SAMPLE
TUBE
\
c c:
MAGNETIC SAMPLE
CARRIER
TO MERCURY
MANOMETER
's—'
COLD—FINGER
(TRAP 3)
\
TO ION GAUGE
TO THERMOCOUPLE
GAUGE
n
HIGH VACUUM MANIFOLD
FURNACE
**~S
4
H-[
U
TO VACUUM
PUMP
TRAP 2
SAMPLE
BULB
^ ^^
±
TO VACUUM
PUMP
i
TRAP t
g
M
FIGURE 1
VACUUM CINE-COMBUSTION APPARATUS
-------
APPENDIX G
*
PHILLIPS PETROLEUM COMPANY - METHOD 6511-AF-l
OXYGEN FLASK COMBUSTION - GRAVIMETRIC METHOD FOR DETERMINATION
OF SULFUR IN ORGANIC COMPOUNDS
I. SCOPE
This method is for the determination of 5 to 50 per cent total
sulfur in organic compounds which can be burned in a Schoniger oxygen
combustion flask. It is applicable to solid samples and to liquids of low
volatility.
II. OUTLINE OF METHOD
A 25- to 50-milligram sample is ignited in a combustion flask con-
taining oxygen. The sulfur, as sulfate in the flask washings, is determined
gravimetrically as barium sulfate.
III. APPARATUS
(a) Analytical Balance. Accurate to -0.1 mg.
(b) Filtering Crucibles. 30-ml, porous porcelain, fine porosity.
(c) Muffle Furnace. Capable of operating at 800 C.
(d) Oven. Electric, operated at 110 C.
(e) Sample Wrappers. Filter paper cut to a shape and size to facilitate
wrapping, folding and igniting samples. Arthur H. Thomas Co. No, 6471-F
or equivalent.
(f ) Sehb'niger Combustion Unit , With electric ignition. Arthur H. Thomas Co.
No. 6471-H or equivalent.
(g) Schoniger Combustion Flask. 1000-ml, borosilicate glass, with glass
hooks and 29/2*9 inter joint. Arthur H. Thomas Co. No. 6471-H10 or
equivalent. (Also available in 500-ml and 2000-ml sizes.)
(h) Stopper for Combustion Flask. Borosilicate glass, with platinum gauze
basket and electrical contact. Arthur H. Thomas Co. No. 6471-H15 or
equivalent .
*Issued: April 1, 1965 as VO-65R
Revised: December 21, 1972
132
-------
APPENDIX G Method 6fll-AF-l
Page 2
IV. REAGENTS AM) MATERIALS
All reagents are ACS reagent grade unless otherwise noted,
References to water indicate distilled or deionized water.
(a) Barium Chloride Solution,. Dissolve 100 g of Bad '2H00 in water and
dilute to 1 liter.
(b) Hydrogen Peroxide. 30 per cent solution.
(c) Hydr och lor ic_ Ac id. Concentrated, sp. gr. 1.18.
(d) Oxv_gen. Commercial grade cylinder oxygen is satisfactory.
(e) Sodium Carbonate Solution. 2 per cent aqueous,
V. PROCEDURE
(a) Combustion. Weigh 25 to 50 mg (to nearest 0.1 mg) on a paper sample
wrapper. Fold the paper and attach it to the platinum cradle of the
flask stopper. Add 10 ml of 2 per cent sodium carbonate solution to the
combustion flask. Purge the flask with oxygen from an oxygen cylinder
for 20 to 30 seconds. Moisten the glass stopper with water and immediately
insert the stopper and sample into the flask. Attach the flask to the
ignition unit and turn the plastic shield end for end to place the flask
behind the shield in an inverted position. Press the firing button to
ignite the sample.
After combustion is complete, remove the flask from the ignition
unit Shake the flask at intervals and let stand (or cool in "e^
until vapor absorption is complete. Remove ^he stopper from the iiasK
attain a volume of 250-300 ml.
r S T-i r Add 2 ml of concentrated HC1 and 2 drops of 30
^???5K=^^--^-j^r>^S^
^
200 ml on * hot plate. Ad£^ ^ f U ^^ ^ & fine
solution and add 10 ml of barium chlor i^t^n and for 2
sla
stream or drcpvd.se.
minutes thereafter.
?T£ T^r:,r=p^r^-Plate - ^ *
cool for at least 1 hour before filtering.
fa«t -Hniiid through a tared (±0.1 mg) fine-porosity
Filter the supernatant liquid througn ^ precipitate with
porcelain filtering crucible. Wash tne D« ib- ^^ free of chloride
water, first by decantation and then in ^ and ignite
Dry the crucible and contents in an oven at u.
133
-------
APPENDIX G Method 6511-AF^l
Page 3
in a muffle furnace at 800 C for 1 hour. Allow the crucible to cool to
room temperature in a desiccator, and weigh to io.l mg.
VI. CALCULATIONS
Calculate the sulfur content of the sample as follows:
Sulfur, wt. %** 1?ffi B
where: B = grams of BaSO. precipitate
if
W = grams of sample
VII. PRECISION
Statistical evaluation of data from duplicate analyses on 94
samples by three different analysts showed:
Range s
Wt/J_S_
5 to 10
10 to 20
20 to 50
Degrees of
Freedom
22
41
31
Standard
Deviation
0.1325?
0.153
0.296
Repeatability
0.39*
0.44
0.86
Duplicate determinations should be considered suspect if they differ
by more than the repeatability shown (95 per cent confidence level).
REFERENCES
1. Schoniger, W., Mikroehim, Acta., (1956), 869-
2. Lysyj, I. and Zarembo, J. E., Anal. Ghent. 20, 428 (1958).
3. Martin, A. J, and Deveraux, H., Anal. Chem. 21, 1932 (1959).
134
-------
APPENDIX H
PHILLIPS PETROLEUM COMPANY - METHOD 7301-AZ*
PREPARATION OF SILVER SULFIDE FROM BARIUM SIIT.FATF.
HYDRIODIC ACID REDUCTION METHOD
I. SCOPE
This method is for the preparation, without fractionation of the
sulfur isotopes, of silver sulfide from barium sulfate. The silver sulfide
produced is suitable for use in isotope ratio mass spectrometric studies.
II. OUTLINE OF METHOD
A weighed sample of barium sulfate is added to a solution containing
hydricdic acid, hypophosphorous acid and hydrochloric acid. The hydrogen
sulfide produced is washed with water and absorbed in a cadmium acetate
solution. The cadmium sulfide produced is converted to silver sulfide and
determined gravimetrically.
III. APPARATUS
(a) Flask. Boiling, 250 ml, round-bottomed, with gas inlet tube.
(b) Condenser.
(c) Vacuum Trap,. Plain, 6 in., 2 needed.
(d) Centrifuge Tube. 100-ml.
(e) Analytical Balance.
(f) Heating Mantle.
(g) Pasteur Pipette. Disposable.
IV, REAGENTS AND MATERIALS
All reagents are ACS reagent grade unless otherwise noted. References
to water indicate distilled or deionized water.
(a) Reducing Solution. Mix 500 ml (850 g) hydriodic acid (d = 1.7), 816 ml
hydrochloric acid, concentrated, and 245 ml hypophosphorous acid, H PO
(50 per cent) in a large Erlenmeyer flask. Boil solution for 45 minutes
to remove any sulfur present by expelling as H2S (use hood). (Note I)
Issued: January 10, 1973
135
-------
APPENDIX H Method 7301-AZ
Page 2
NOTE 1; An alternate reducing solution giving equally
satisfactory results is made by mixing 56? g
sodium iodide, 245 g sodium hypophosphite,
NaHJP02'H_0, and 180 ml water. Heat mixture
until a clear solution is obtained , then add
1310 ml hydrochloric acid, concentrated. Stir
for 10 minutes, cool to room temperature, and
filter mixture. Discard precipitate. Boil
solution for 45 minutes to remove any sulfur
present.
(b) Cadmium Acetate Solution. 0.1 N. Dissolve 26.65 g Cd(C H 0 L-2H 0 in
water and dilute to one liter. * * * & *
(c) Silver Nitrate Solution. 0.1 N. Dissolve 16.99 g of AgNO~ in water and
dilute to one liter.
Nitrogen. Prepurified grade or better.
V. PROCEDURE
(a) Preparation of Apparatus.
1. Assemble apparatus as shown in Figure 1. Mount trap 1 in reverse
manner as shown, fill trap 2 with distilled water.
2. Add 100 ml of reducing solution and several boiling beads. Adjust
nitrogen to give moderate flow and heat solution to boiling. Boil
30 minutes. Add cadmium acetate solution to centrifuge tube and con-
tinue boiling for 15 minutes. If no yellow precipitate is formed in
the cadmium acetate solution, the system is ready for use. If yellow
precipitate forms, continue boiling until all sulfur has been expelled
as evidenced by lack of yellow precipitate formation when fresh cadmium
acetate is placed in centrifuge tube. Cool reducing solution to room
temperature .
(b) Barium Sulf ate Reduction.
1. Carefully weigh 25-200 rng BaSO into a small boat. Wash BaSO into
flask containing reducing solution using 1-10 ml distilled water.
2. Adjust nitrogen flow, make certain centrifuge tube contains about
75 ml of colorless cadmium acetate solution. Boil solution for about
4 hours or until all of the sulfur has been collected as cadmium
sulfide. Add acetone to cadmium acetate to reduce foaming if necessary.
136
-------
APPENDIX H Method 7301-AZ
3
(c) Conversion to Silver SuTfide.
1. Fill centrifuge tube to 100 ml mark with water. Centrifuge for 5-10
minutes. Decant supernatant liquid. Add 50 ml distilled water and
mix thoroughly. Add 50 ml silver nitrate solution, mix thoroughly
and let stand 1 hour.
2. Wash filtering crucible with water, ammonium hydroxide and water using
vacuum filtration. Dry in vacuum oven at 60°C for 2 hours. Cool to
room temperature and determine tare weight.
3. Filter silver nitrate-silver sulfide mixture through the tared filter-
ing crucible. Wash precipitate with ammonium hydroxide and water.
Dry in vacuum oven at 60 C for 2 hours. Cool to room temperature and
determine weight of silver sulfide.
4. Save silver sulfide if needed for further studies. Wash filtering
crucible with chromic acid to clean.
VI. CALCULATIONS
(a) Theoretical Yield
weight Ag S (theoretical) -= weight BaSO^ x 1.062
(b) Per Cent Yield
v. , . t<\ weight Ag S recovered x 10Q
Held \7>) - weight Ag*s (theoretical)
REFERENCE
1. H. G. Thode, J. Monster, and H. B. Dwiford, Geochimica et Cosmochimica
Acta, 21, 159 (1961).
137
-------
APPENDIX H
N
f
O
(0
I
111
2
Ou
O
O
D
W
O
b
a.
in
a.
j
j
z
0.
JLL
TRAP 1 TRAP
-CONDENSER
24/40 INTERJOINT
250-ML FLAS
NITROGEN INUET
FIGURE 1
APPARATUS FOR BaSO4 REDUCTION
133
-------
APPENDIX I
OPERATING CONDITIONS FOR SULFUR GLC PROFILE MEASUREMENT
Chromatograph: Tracor 550 with glass injection system. Flame
photometric detector mounted on heated block.
Column: 100 feet x 0.01 inch Stainless Steel coated with Dexsil 300.
Detector: Tracor Model 100 AT Melpar flame photometric detector
used with 394 nm sulfur-selective filter.
Gas Flows: Hydrogen 150 ml/min
Oxygen 20 ml/min
Air 100 ml/min
Carrier Gas: Helium at 30 psig to give column flow of 7 ml/min.
Column Temperature: 100°C initial, programmed at 5°C/min to 330°C and
held for 16 minutes at 330°C.
Injection point: 350°C
Column outlet: 330°C
Detector base: 330°C
Chart speed: 1 inch per minute
Sample size: 2 microliters of a 9 weight percent solution of
600+F bottoms in cyclohexane
Cooling period: 16 minutes
Equilibration period: 16 minutes
139
-------
APPENDIX J
PHILLIPS PETROLEUM COMPANY - METHOD 7204-AZ*
SILICA GEL SEPARATION OF 6QO+F BOTTOMS FROM CRUDE OIL
I. SCOPE
This method is for the separation of 600 F bottoms from crude oil into
saturate,, aromatic , and asphaltic fractions by siliqa gel liquid-solid elution
chromat agraphy ,
II. OUTLINE OF MTHOD
A sample of 600 F bottoms in crude oil is placed on a silica gel coltran.
The column is eluted successively with n-pentane, dichloromethane , and a methyl
alcphol-dichloromethane mixture. The column effluents from each solvent are
collected and the solvent removed by evaporation to recover the three fractions.
The amount ©f each fraction recovered is weighed and the per cent distribution of
the bottoms among the saturate, aromatic , and asphaltic fractions is calculated.
HI- APPARATUS
Balance . Analytical, 5 place.
ib) Chroinatographic Column and Solvent Reservoir, Figure 1« The chromatographic
system is assembled from components available from Fischer & Porter Co., Lab-
Crest Division, County Line Road, Warminster, Pa., 18974* Catalog numbers are
designated "where applicable.
1. Chromatographic column with threaded glass and Teflon needle valve and
Teflon drip tip, 1/2" I.D. x 24" length. Catalog No. 274-472.
2, Glass coupling, aluminum, glass-to-glass, 1/2" I.D, Catalog No. 687-
004-0012.
3. Strip neoprene asbestos for use with couplings. Catalog No. 688-366-0012.
4. Interface gasket for threaded metal coupling for 1/2" I.D. pipe, Teflon
machined envelope with silicone rubber filler. Catalog No. 691-006.
5. Solvent reservoir, 300 ml capacity, graduated, with glass pipe, 1/2"
I.D., exit and 28/15 female ball joint on inlet. Use glass pipe Joint,
Catalog No. 670-000-1206, 1/2" I.D. x 6" length, to fabricate the
solvent reservoir as shown in, Figure 1. ,
(c) Bottles. 2 02s., wide mouth.
. 2 dram.
Roto-Vae Evaporator. Sargent-Welch Scientific Co., Dallas, Texas. Catalog
No, S-31211.
Issued2 March 8, 1972
140
-------
J
(f) Flask. 500 ml, 24/40, round-bottom.
(g) Soxhlet Extraction Apparatus. Fisher Scientific Co., Pittsburgh, Pa.
Catalog No. 9-556C. I.D. of glass extraction tube - 50 mm.
(h) Syringe and Needle. 2 cc, 6 inch needle.
(i) Water Bath. Organisation Assoc. N-Evap Model 10, Worchester, Mass., is
satisfactory.
(j) Funnel. Buchner type, Coors, Porcelain. O.D. 6? mm, plate diameter 56 mm.
Fisher Scientific Co., Pittsburgh, Pa. Catalog No. 10-356B.
(k) Flask. Filtering, graduated, with Side Tube, 500 ml. Fisher Scientific Co.
Catalog No. 10-181E.
IV. REAGENTS AID MATERIAIS
(a) Methyl Alcohol. Merck and Co., Inc., Rahway, N.J.
(b) n-Pentane. (Distilled in glass) Use as received from Burdick and Jackson
Laboratories, .Inc., Muskegon, Michigan.
(c) Dichloromethane. Distilled. Add 25 g of calcium sulfate to 4 kg of Eastman
Kodak dichloromethane to remove water. Mix thoroughly. Filter through
Whatman 40 filter paper and distill. Discard the first 100 ml of distillate.
Collect the remaining distillate and store it in bottles having Teflon-lined
caps. The following distillation apparatus is satisfactory.
1. Distillation column-vacuum jacket®! with an integral liquid fraction
take off, ground ball seat and vertical coil condenser. One meter
length, 2 cm I.D., packed with pyrex Helix rings.
2. Distillation kettle - 3 liter.
3. Reflux ratio - 1 volume reflux to 5 volumes take off.
(d) Silica Gel. Activated, W. R. Grace Co., Grade 923, 100-200 mesh. Curtin
Scientific" Co., Box 1546* Houston, Texas 77001.
Purify the silica gel before use as follows. Using the So xh let
If
«-
141
-------
Method 7204-AZ
APPENDIX J Page 3
warm silica gel occasionally until a free-flowing powder is obtained.
Activate the silica gel at 800 F (42? C) overnight. Store the activated
material in a bottle having a Teflon-lined screw cap.
(e) Hitrogen Regulator and Cylinder.
(f) Glass Wool.
V. PROCEDURE
(a) Column Preparation.. Assemble the column and reservoir as shown in Figure 1.
Place a small piece of glass wool at the bottom of the column. Rinse the
column and reservoir with dichloromethane. Discard the washings. Rinse the
column and reservoir with n-pentane and discard the washings. Add 100 ml of
n-pentane to the column with the stopcock closed. Add 60 ml of silica gel to
the column and shake in order to make a slurry of silica gel in n-pentane.
Allow the silica gel to settle and proceed to Section V (b) during this period.
(b) SampleTransfer and Saturate Fraction. Into a tared 2 dram vial add about
0.25 g of the 600*7 bottoms sample and weigh. Add 2 ml of n-pentane to the
vial to dissolve the sample. "Transfer the resulting solution to the top of
the column by means of a needle and syringe. Repeat both the addition of
n-pentane to the vial and the transfer of the solution to the column until the
sample is completely transferred to the top of the silica gel column. In many
cases the sample is not completely soluble in n-pentane and an insoluble
residue remains in the vial. Save this residue for Section ¥ (c). Apply
5 psig nitrogen pressure to the top of the solvent reservoir. Open the stop-
cock and collect the n-pentane in a clean, dry 500 mi round-bottom flask until
the n-pentane level is 1 inch above the top of the silica gel. Do not let the
silica gel go dry. The flow rate of all the solvents is about 10 ml per minute.
Add 200 ml of n-pentane to the solvent reservoir. Apply 5 psig nitrogen
pressure. Continue to collect the column effluent in the 500 ml flask until
the n-pentane level is 1 inch above the top of the silica gel. Proceed to
Section ? (c) to obtain the aromatic fraction.
Place the flask gn the rotary^vacuum evaporator and remove the solvent
completely at about 60 C and sufficient vacuum to gently boil the solvent. The
saturates fraction is normally colorless and careful technique is required to
insure quantitative transfer of this fraction to the vial. (The aromatic and
asphaltic fractions are dark-colored and the completeness of quantitative
transfer can be observed visually.) Transfer quantitatively the saturates
fraction from the round-bottom flask to a tared vial using dichloromethane as
solvent which is added in small increments (1-2 ml) and allowed to flow down
the walls of the flask. Transfer each solvent increment to the vial by needle
and syringe before adding,the next solvent increment. A total of 9 ml of di-
chloromethane is sufficient for quantitative transfer without exceeding the
capacity of the vial. Place the vial in the water bath at 50-60°C and turn on
the nitrogen flow to evaporate the solvent. Cool, dry and weigh the vial.
The increase in weight of the vial over the tare weight represents the saturate
fraction, S grams. Store the vial containing the saturate fraction in a
capped 2 oz. wide-mouth bottle.
142
-------
APPENDIX J
**^, majority of the 600+F bottoms the sample will b
R ' > less the inle rfsidue,
Add 200 ml of dichloromethane to the solvent reservoir. Apply 5 pse
nitrogen pressure. Collect the effluent in a clean, dry 500 ml round-bottom
flask until the dichloromethane level is 1 inch above the top of the silica
gel column. Do not let the silica gel go dry. Proceed to Section V (d) to
obtain the asphaltic fraction.
Place the flask on the rotary-vacuum evaporator and remove the solvents
completely. Transfer quantitatively the aromatic fraction from the round-
bottom flask to a tared vial using dichloromethane as solvent. Place the vial
in the water bath and turn on the nitrogen flow to evaporate the solvent.
Cool, dry and weigh the vial. The increase in weight of the vial over the
tare weight represents the aromatic fraction, A grams. Store the vial con-
taining the aromatic fraction in a capped 2 oz. wide-mouth bottle.
(d) Asphaltic- Fraction. Add 200 ml of mixed solvent (75 vol. % methyl alcohol/25
vol. % dichloromethane) to the solvent reservoir.. Apply 5 psig nitrogen
pressure. Collect the effluent in a clean, dry 500_ml round-bottom flask until
all the solvent has drained from the column. (Allow the column to run dry.)
Place the flask on the rotary-vacuum evaporator, and remove the solvents.
Transfer quantitatively the asphaltic fraction from the round-bottom flask to
a tared vial using dichloromethane as solvent. Place the vial in the water
bath and turn on the nitrogen flow to evaporate the solvent. Cool, dry and
weigh the vial. The increase in weight of the vial over the tare weight
represents the asphaltic fraction, C grams.
(e) Silica Gel Removal. Silica gel may be removed from the column as follows:
With the reservoir top open to the air and the stopcock open, pull a vacuum
on the column for 2 hours. The dry silica can then be poured out of the
column into a sealable can for appropriate disposal.
143
-------
Method 7204-AZ
APPENDIX J Page 5
VI. CALCULATIONS
(a) Distribution. Normalized .
s xoo
1. Saturates, weight per cent « s + A + C
A 100
2. Aromatics, weight per cent » s -i- A + C
3. Asphaltics, weight per cent = ...... ' "
(b) Recovery.
_._. . . . , (S + A + C) 100
Silica gel recovery, weight per cent «= •* - (w « R) -
where: S «= grams in saturate fraction
A = grams in aromatic fraction
C » grams in asphaltic fraction
W = grams 600 F bottoms weighed into the sample vial
R = grams insoluble residue not transferred to the column
144
-------
APPENDIX J
,
2
)
PHILLIPS PETROLEUM COMPANY
£
v;
f^T
- METHOD 7204-AZ
f —**l , ._ _ Vf- .
^ Nitrogen
^ •** Ball Joint 28/15
-Jml'~~-v.
300 -X
inn 300 ml
•\e\f\
-lA/w n in i -ni-
e\
it
n-Pentane *""
V
? -Glass Wool ^-
nun
x /
X X
A •* ^^ Silica uel
/w\<
vs/V
XX
X X
; — r -i - T'laff Coupling
| ' J— a— stopcock
FIGORE 1. CHRCMTOGRAPHIC COUJMH AMD SOLVEHI RESERVOIB
145
-------
APPENDIX K
PHILLIPS PETROLEUM COMPANY - METHOD 7214-AZ*
ISOLATION OF n-PARAFFINS FROM SATURATE FRACTIONS BY UREA ADDUCTION
I. SCOPE
This method is for the preparation of a urea adduct to isolate the +
n-paraiffins from the saturate fraction obtained by silica gel separation of 600 F
bottoms of crude oil.
II. OUTLINE OF METHOD
A sample of the saturate fraction is dissolved in a benzene-methyl
alcohol mixture. Methyl alcohol saturated with urea is mixed with the sample.
The urea adduct precipitate which forms is allowed to stand overnight. Solvent
is removed and the urea adduct is dried. The carbon-number distribution is
measured by gas-liquid chromatpgraphy of the isolated n-paraffins (Phillips
Method 7215-AG).
III. APPARATUS
(a) Vials. Two-dram, Teflon-lined caps.
(b) Syringe and Needle. 5-cc, 23-gauge.
(c) Spatula. Micro.
(d) Pipets. Pasteur, disposable.
IV. REAGENTS AND MATERIALS
(a) Methyl Alcohol. Reagent Grade.
(b) Benzene. Reagent Grade.
(c) Benzene-Methyl Alcohol Solvent. Add 300 ml of benzene to 100 ml of methyl
alcohol and mix thoroughly.
(d) Urea. Recrystallize as follows: Saturate 200 ml of methyl alcohol with
urea. Heat the solution to about 50 C and add more urea until saturated.
Decant the hot solution into a beaker, allow to cool to room temperature.
Pour off the solution and collect the recrystallized urea.
(e) Saturated Urea in Methyl Alcohol. Add recrystallized urea to 50 ml of
methyl alcohol to form a saturated solution.
Issued: May 25, 1972
146
-------
Method 7214-AZ
APPENDIX K Page 2
V, PROCEDURE
Transfer about 50 mg of saturate fraction obtained from the silica
gel separation (Phillips Method 7204-AZ) into a clean dry 2-dram vial with the
aid of a micro spatula or a disposable pipet. Add 5 ml of benzene-methyl alcohol
solvent and cap the vial. Shake until the sample dissolves. Add 2 ml of
saturated urea in methyl alcohol solution and mix thoroughly. A precipitate
comprised of urea and the urea adduct of the n-paraffins is formed. Allow the
precipitate to stand overnight. Remove the solvent from the vial by using a
syringe with a 23 gauge needle. The inside diameter of the needle is small
enough to prevent removal of precipitate. Allow the precipitate to dry by
leaving the cap off the vial for several hours. Recap the vial containing the
dry urea adduct and save for gas-liquid chromatography.
147
-------
APPENDIX L
PHILLIPS PETROLEUM COMPANY - METHOD 7305-AG*
DETERMINATION OF RELATIVE WEIGHT DISTRIBUTION OF n-PARAFFIMS UREA
ADDUCTED FROM CRUDE OIL FRACTIONS C._ TO C-n BOILING EANGE
"•45 J50
I. SCOPE
This method determines the relative weight distribution of n-
paraffins adducted from crude oils and their fractions in the boiling range
of Cno to C0_ n-paraffins.
JO Jo
II- OUTLINE OF METHOD
The adduct sample (prepared according to Method 7214-AZ) is
dissolved in water to reduce the n-paraffins which are extracted into
cyclohexane solvent for analysis. Using n-C,o as an internal reference
time standard (the cyelohexane solvent contains x^0.25 weight per cent
n-C.n) the n-paraffins are separated on a gas-liquid partition chromatog-
rapny column. The output signal from the flame ionization detector is sent
directly to a central computer processor where data acquisition and reduc-
tion programs provide a weight per cent distribution of n-paraffins, a mol
per cent distribution of n-paraffins, smoothed weight per cent data, OEP
values (a weight ratio of odd-carbon-numbered to even-carbon-numbered
members of a homologous series), and graphical plots of all the above data
except mol per cent distribution.
III. APPARATUS
(a) Ghromatograph. Any chromatograph which is equipped to take 1/8-inch
diameter columns, has temperature programming capability to 380 C,
and is equipped with a flame ionization detector. A Perkin-Elmer
Model 880 gas chromatograph was used in the development of this method.
(b) Strip-Chart Recorder. 1-mv range and 2-second pen speed or faster.
(c~) Microliter Syringe. Model #1705, gas-tight, available from Hamilton
Company, Inc., Whittier, California.
(d) Automatic Data Acquisition and Reduction System. A minimum of an
electronic.integrator to determine the areas of the n-paraffin peaks
reduces analysis time of this method from 2.5 hours to 40 minutes.
This procedure uses an IBM 1800 computer for on-line data acquisition
and reduction. A typical computer output from this procedure is
attached to this method as Appendix I (Note l).
MOTE 1: Except for the weight per cent report >appearing on
pages 1 and 2 of Appendix I, all subsequent data in
Appendix I is generated from these data. The
computer programs for this data reduction are included
Issued: January 18, 1973
148
-------
APPENDIX L Method
Pagft :-:
'in Core Load Link Program Number 7.
IV. REAGENTS AND MATERIALS
(a) Helium.
(b) Hydrogen.
(c) Cylinder Air.
(d) Chromosorb P. 80-100 Meshs available from Johns Manville Company,
Celite Division, New York, New York.
(e) Tubing. 1/8-inch O.D., stainless steel.
(f) Poly-m-Fhenoxylene (PPE 20). Available from Varian Aerograph, Walnut
Creek, California. .
(g) Chloroform. Reagent Grade.
(h) Cyclohexane. Research Grade, available from Phillips Petroleum Company,
Special Products Division, Bartlesville, Oklahoma.
(i) n-Tetracontane_(C-40). Available from Poly Science Corporation,
Chemical Division, Evanston, Illinois.
7. PRELIMINARY PREPARATIONS
(a) 16.7 Weight Per Cent PPE 20) on Chromosorb P. Weigh 6 g of PPE 20 into
a 250 ml beaker and add 150.ml of chloroform. Stir the mixture until
completely dissolved and add the mixture to 30 g of 80-100 mesh
Chromosorb P in a 15-cm evaporating dish. Carefully evaporate, while
stirring, the chloroform using low heat on a hot plate. Continue
evaporation until the mixture appears dry.
(*>) Column. Clean a 10-foot length of 1/8-inch O.D. stainless steel tubing
with a 100 volume wash of benzene followed by a 50 volume wash of
acetone and air dry the tubing. Fill, the tubing with the prepared
packing (Section V,a) by bending the tubing into a U shape and filling
each side of the U, vibrating the tubing gently during fining. Plug
the ends of the column with glass wool and shape the column so it can
be conveniently mounted in the chromatograph.
(c) Column Conditioning. With the exit end of the column disconnected from
the detector, condition the column as follows:
1. Establish a helium carrier gas flow of 22 cc/min. through thr ••
at ambient temperature.
149
-------
APPENDIX L Method 7305-AG
Page 3
2. Program the column oven temperature from ambient to 300 C @
6 /min. and hold at upper limit for two hours.
3. Increase column oven temperature to 380 C and iraintair, for three
hours.
4. Cool oven to 320 C arad maintain column at this tercperatare for
twelve hours.
5. Cool oven to ambient temperature, disconnect column from inlet and
reconnect so that, exit end of column during conditioning period is
now connected to inlet side of chromatograph. Connect column to
the detector and column is ready for use.
VI. PROCEDURE
Set the detector supply gases for optimum flame detector operation,
the carrier gas flow to 22 cc/min. at ambient temperature, the upper limit
of temperature programming to 370 , the initial column oven temperature to
150 C, and the temperature program rate control to 12 c/min. The instrument
has equilibrated when the recorder baseline is stable.
Place the available urea-n-paraffin adduct in a 2-dram vial and
dissolve in approximately 5 ml of hot distilled water to release the
adducted hydrocarbon. Using the following table, add the indicated amount
of cyclohexane containing 0.25 weight per cent n-C.» (Note 2).
NOTE 2: The n-C.^ is an internal reference time standard
only for updating retention time changes in the
computer job definition. Relative sensitivity
values for the flame ionization detector are
assumed to be unity for all the n-paraffins
measured by this method.
Mg Adduct Cyclohexane Solvent
1 1 drop
10 5 drops
20 10 drops
Cap the vial and agitate gently by tapping twice.
150
-------
APPENDIX L Method 73Q5-AG
Page 4
Inject 2 microliters of the cyclohexane phase into the chromatograph
for all adduct samples of 10 mg or more. Inject a maximum of 5 ul into the
chromatograph for the smallest adduct samples. Immediately after sample
injection, initiate the temperature programmer and the computer data
acquisition control switches. Record the chromatogram on the strip chart
recorder. A typical chromatogram of a C..,, to C ft adduct sample is shown
in Figure 1. JJ Jo
VII, CALCULATIONS
If hand calculations are required to obtain the relative weight
distribution of the n-paraffins in the adduct samples,, peak areas of the
individual n-paraffin peaks are measured by peak height times the width of
the peak at 1/2 the peak height. Calculate the weight per cent distribution
of n-paraffins as follows:
Cj^ «t. !» =
where: C. = concentration of ith component, weight per cent
i = area of ith component
= sum of areas of all ith components in sample
VIII. PRECISION
A statistical evaluation of 12 sets of duplicate analyses
showed the following%
Ranee. Wt. % Degr^e^^f_Freedoni Standard Deviation Repeatability
0 - 1.0 69 0.06253 0.1769
1.0-10.0 210 0.2657 0.7440
151
-------
nc15
24
TIME, MINUTES
FIGURE 1
TYPICAL n-PARAFFIN GAS CHROMATOGRAM
-------
APPENDIX I
PHILLIPS PETROLEUM COMPANY - R£D DEPARTMENT
AUTOMATED INSTRUMENT ANALYSIS
DATE= 4/ 4/72 11.30 TITLE= C13-C38 ADDUCTS-GEOCHEM
INSTRUMENT NO. 16 ANALYST= COJ FEATURE NO. 6094
SAMPLE ID= 1-61F HJ-20246-1-61F GLC# 721366
SAMPLE RUN= 4/ 4/72 11.03 ANALYST.GJ
EXPECTED ACTUAL RATIO DELTA
TIME TIME
1422.00 1425.44 1.00 0.0
FILE 1 DATA SAVED AT REC NO. 1475
FILE 10 DATA SAVED AT REC NO. 1016
RUN NO,
1 OF
JOB NO. 226
CALC NO. 904
COMPONENT
TRIDECANE
TETRADECANE
PENTADECANE
HEXADECANE
HEPTADECANE
OCTADECANE
NONADECANE
EICOSANE
HENICOSANE
DOCOSANE
TRICOSANE
TETRACOSANE
PENTACOSANE
HEXACOSANE
HEPTACOSANE
OCTACOSANE
NONACOSANE
TIME
(SECS)
234.1
291.0
350.6
410.9
469.9
527.5
582.4
635.7
686.6
734.9
781.2
825.8
868.1
909.1
948.7
987.3
1024.3
AREA
(MV-SEC)
70.961
407.499
1175.242
1715.803
2353.428
2514.090
2746.255
2974.175
3306.405
3574.879
4089.936
4178.492
4419.819
4431.568
4595.598
3686.622
3300.198
RESULTS
0.126
0.721
2.080
3.037
4.165
4.450
4.861
5.264
5.852
6.327
7.239
7.395
7.823
7.843
8.134
6.525
5.841
UNITS
WEIGHTS
WEIGHTS
WEIGHTS;
t-i
P
I
O
9
U)
O
X
-------
TRIACONTANE
HENTRIACONTANE
DOTRIACONTANE
TRITRIACONTANE
TETRATRIACONTANE
PENTATRIACONTANE
HEXATRIACONTAME
HEPTATRIACONTANE
OCTATRIACONTAME
C40
TOTAL
1059.9
1094.6
11?7.9
1162.2
1193. 1
1224.1
1256.6
1292.5
1332.0
1425.4
1488.3
2338.445
1744.546
1160.015
697.50-9
408.812
258,699
172.011
J.D2.740
77.494
2291.532
19.623
58812.330
4.139
3.088
2.053
1,235
0.724
0,458
0.304
0.182
0.137
0.000
100.013
APPENDIX I - page 2
M
vn
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
270.1
326.0
389.3
437.2
449.9
496.8
508.1
564.4
607.1
618.0
670.3
707.0
719.3
755.9
765.9
PI 1.2
854.1
885. 1
896.1
925.7
936.7
965.4
975.7
1003.0
1013.0
1048.6
1083.9
1115.9
12.399
55.904
60.628
85.368
65.024
30.933
85.325
78.412
20.986
86.650
90.609
30.482
85.984
21.938
87.519
102.204
117.621
33.770
151.119
57.874
170.039
62.857
218.190
174.093
215.564
311.739
312.209
340.069
0.019
0.087
0.095
0.133
0.102
0.048
0.133
0.123
0.033
0.135
0.142
0.048
0.134
0.034
0.137
0.160
0.184
0.053
0.236
0.090
0.266
0.098
0.341
0.272
0.337
0.487
0.488
0.532
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT.
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
X
t-1
-------
APPENDIX I - Page 3
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
TOTAL
1 1 50 . 2
1181,6
1212i6
1244.6
1278.0
1317.0
1346.0
1376.4
1389.9
1508.3
1557.3
1582. 2
307.489
295.691
230.509
197.080
132.707
56.517
190.851
226.404
159.617
103.068
56.804
44.716
0.481
0.462
0.360
0.308
0.207
0.088
0.298
0.354
0.249
0.161
O.O89
0.070
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
AREA-PERCENT
5166.975
8.076
•n
vn
-------
APPEHDIX I - Page 4
1-61F
HJ-20246-1-61F
GLCS 721366
CARBON
PER MOLFCULF
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
3.0
31
32
33
34
35
36
37
38
MOLE
PERCENT
0.222
1.185
3.192
4.371
5.646
5.699
5.900
6.073
6.432
6.640
7.268
7.118
7.230
6.972
6.963
5.387
4.657
3.190
2.304
1.484
0.865
0.492
0.303
0.196
0.114
0.083
WEIGHT
PERCENT
0.126
0.721
2.080
3.036
4.165
4.449
4.860
5.263
5.851
6.326
7.238
7.394
7.822
7.842
8.133
6.524
5.840
4.138
3.087
2.053
1.234
0.723
0.458
0.304
0.182
0.137
DATE-
SMOOTHED WEIGHT
PERCENTAGES
- - -
1.987
3.023
3.867
4.443
4.877
5.325
5.848
6.436
6.999
7.423
7.703
7.799
7.495
6.688
5.557
4.320
3. 148
2.154
1.379
0.842
0.517
0.328
4/ 4/72ANAL'
OEP
1.115
1.067
1.066
1.031
1.008
1.011
1.018
1.034
1.040
1.022—
1.0221
1.046
1.086
1.093
1.084
1.068
1.033
1.006
0.98 6 J
1.010
1.012
0.952
TJ
AVERAGE OEP BETWEEN C-25 AND C-33 IS
1.047 .
-------
APPENDIX I - Page 5
1-61F
HJ-20246-1-61F
GLC# 721366
DATE-
A/72ANALYST- COJ
C
E
N
T
R
A
L
H1
^3 c
A
R
B
0
N
N
U
M
B
E
R
0.5 1.00
10--4- 4-
4- 4-
4- 4-
4 4
4- 4
15 4- 4- *
4- 4 *
4 4 *
4- 4- *
4- *
20 4- 4-*
4- 4-*
4- 4- *
4 4- *
4 4*
2 5 4- 4- *
4 4 *
4 4 *
(. 4 #
— — 4 4- *
30 — + + *
4 4- *
_w.«t So
^ * +
M__^, +sfc
35 4- 4-*
4 * 4-
h 4-
4- 4-
4 4-
40 4- 4-
4 4-
4. 4-
4 4-
4 +
45 — 4- 4-
0 . => I. 00
OEP VALUE
1.50 2.00
4 4
4 4
4- 4-
4 4
4 +
4 4-
4 4
4 4
4- 4-
4- 4-
4 4-
4 4-
4- 4-
4 4-
4- 4-
4 4-
4 4
4 4
4 4-
4- 4-
4- 4-
4- 4-
4 4-
4 4-
4 4-
4 4-
4 4
4 4-
4- 4-
4- 4-
4 4
4 4-
4- 4-
4- 4-
4- 4-
1.50 2.00
2.50
4-
4-
4-
4-
4-
4-
4-
4-
4
4-
4-
4-
4-
4
4-
4
4-
•!-
T) 4
33
M 4-
E +
S 4-
•-3 4-
X Tf
§ § *
fd 4-
"^1 O
U) M +
O Pi
Y1 g *
K 4-
O O
«S 4
^2 +
4
' 4-
4-
4-
4
85
-------
APPENDIX I - Page 6
1-61F
HJ-20246-1-61F
GLC# 721366
DATE-
4/72ANALYST- COJ
c
E
N
T
R
A
L
H1
vn
00 C
A
R
B
0
N
N
U
M
B
E
R
0.0
10 — 4-
— +
4
4*
h *
15 — 4- *
\.
4-
4
20 — 4-
4-
4
4-
4-
25 — 4-
1-
4
4
4-
30 — 4-
——4- *
4 *
+ .,,
35 — 4 *
4 *
K*
4*
4
40__4
4-
+
4-
4
I ....4... .4. ... + ..
WEIGHT PERCENTAGE
5.00 10.00 15.00
4 4 4
44-4-
4-44-
4-4-4
44-4-
* 4- 4- 4-
* 4 4- 4-
-J; 4 4 4
4-* 4- 4-
4- * 4- 4-
4- * 4- 4-
4*4 4
4*4- 4-
4*4 +
4*4- 4
4*4 4
4*4- 4-
4- * 4- 4-
* 4 4 4-
444
444-
4- 4.4-
444-
444
+ 44-
4-44-
444
44-4
444
44-4-
444
4 4 4
4 -r 4
+ 4 4-
20.00
4-
4
4
4
4-
4-
4
4-
+
+
4-
4-
4
*
4
4
3 +
U +
p 4-
B rt "**
3 4-
8 3 :
3 R 4
Y* S +
t 4-
§4
+
5 4-
4-
4
4
-f
4
. .4. ....... .4... , T
T)
- .. • ')
-------
APPENDIX I - Page 7
1-61F
HJ-20246-1-61F
GLCf? 721366
DATE-
4/72ANALYST- COJ
SMODTHEO WEIGHT PERCENTAGE
C
E
N
T
R
A
L
C
A
R
B
0
N
N
U
M
B
E
R
0.0 S.OO 10.00
10 + 4- +
h -f -f
4- + 4-
1- 4- 4-
4- 4 4-
15 — -I- * + +
4- * 4- 4.
4- * + 4-
4- * 4- 4-
4- *4 4-
20 — 4- + * +
4. 4- * 4.
4- 4 * 4-
1- + * 4-
4- 4- * 4.
25 4- 4- * 4
4- 4- * 4-
4- 4*4
4- 4- * 4-
4- 4- * 4-
30 4- * 4- 4
4- * + 4-
4- * 4- 4-
4- * 4 4-
4. * 4 4
35 4- * 4 4-
4- * 4- 4-
1- 4 4
4 4- 4
40 4- + 4
15.00 20.00
4- 4-
4- 4-
4- 4-
4- 4-
4- 4-
4- 4-
4 4-
4- +
4. 4-
4- 4-
4- 4-
4- +
4- 4-
4- +
4- 4-
+ 4-
4- 4-
4- 4-
4- 34-
4. 8 +
4- C 4.
4- H T? +
4- 3 +
4 9 3 +
+ 84.
4 3 B 4-
4. S H 4
4 t 4
§ +
§ 4
5 :
-------
APPENDIX M
PUBLISHED SULFUR ISOTOPE DATA
Location, Description
United States
California
Summerland 64075
Wilmington 66096
Wilmington
Kansas
Cherokee Penna.
Louisiana
Timbalier Bay (Offshore)
Montana
Woodrow PC 58-399
Oklahoma
Konawa-Dora 67074
Sho-Vel-Tum 59171
Springer Penna
Misener Penna
Texas
Ellenburger, Ordovician
Silurian Silurian
McElroy Permian
Keystone Permian
East Texas Field
Ward Estes, N. Texas
Goldsmith
Kelley Snyder
Sprayberry (Trend Area)
Headlee
Utah
Red Wash 67131
Chevron Red Wash
Green River Eocene
Reference*
.23
2.51
-7.9
+6.8
8.8
.24 +3.9
-6.4
.90 -4.8
a
a.
c
.44
1.44
.27
.1
—
—
_
_
_
_
—
_
—
.31
.2
.1
+2.5
+14.5
-.1
-5.6
7.8
+7.7
-2.4
+4.5
-5.7
.6
4.6
0
1.2
.6
+12.8
+16.0
28.2
a
a
d
d
d
d
d
d
c
c
c
c
c
c
a
a
d
^-References on Page 166, Appendix M.
160
-------
APPENDIX M lCont»d)
Location, Description
Wyoming
River Bend PC 66-39
Cowley PC 67-75
Reno PC 66-48
Amoco Reno Crude
Middle Dome PC 67-76
Beaver Creek PC 67-73
Amoco Beaver Crude
Phosphoria, Permian
Tensleep, Penna
Madison, Miss
Lakota Cretaceous
Frontier Cretaceous
Wind River Basin
Paleozoic 8,600 ft
10,000 ft
Cretaceous
Triassic
Canada
Alberta
Leduc Oil Field, Devonian D-2
Stettler "
Big Valley "
Normandville "
Fort Norman, N.W. Terr. "
Excelsior "
Stettler Devonian D-3
Big Valley "
Bashaw "
Redwater "
Leduc "
Wizard Lake "
Golden Spike "
Bonneyville, Lower Cretaceous
Lloydminster "
Abasand Quarry "
Taber "
Brooks Bantry, Blairmore
Campbell Basal Cretaceous
Whitemud "
Woodbend
Malmo Lower Cretaceous
Referende*
2.52
3.25
.77
2.58
.54
2.8
2.4
1.9
.25
.1
-6.6
+6.9
-4.5
-4.8
+5.1
-2.5
-3.3
-2.9
-5.1
-3.9
-2.2
-1.8
-7.6 to -1.7
-.6 to +7.5
-1 to ^3
+2; 5
a
a
a
a
a
a
a
d
d
d
d
d
f
f
f
f
.3
1.45
1.04
.20
.33
.67
1.77
.63
.62
.49
.27
.29
.19
3.9
3.6
5.0
1.7
2.2
.87
.59
.38
.50
+12.1
13.1
12.8
10 .1
10 *1
15.2
1018
11.7
13.2
10 < 2
12.5
15.5
15.1
+7.5
6,0
5.7
7'. 5
5.7
10.3
11.0
12.0
10.1
d
4
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
161
-------
APPENDIX M (Centid)
Location, Description %_ S 6J^S
Western Canada
Joseph Lake, Upper Cretaceous .13 -1.0
Armena " .14 -1.0
Bulwark .17 -3.0
Camrose .13 -4.1
Pembina .19 -5.2
Florence, Sask. Mississippian .51 +1.3
North Steelman .55 2.3
Carnduff .55 1.9
Forget 1.85 6.8
Stoughton 2.2 6.3
Edenuale .86 3.6
Florence 1.77 3.4
Cole-ville 2.6 4.9
Turner Valley .3 8.8
Ontario
Sombra Silurian 1.0 +9.2
West Becher Silurian .85 +8.4
Trenton Ordovician .1 +6.3
Norfolk Silurian .1 +9.7
Canada
Saskatchewan
High Prairie Field - +9.4
Stoughton - +9.0
Pierson - +3.8
South America
Vene'zuela
Lagunillas 57124 2.12 +5.2
Boscan 59190 5.53 +5.6
Lake Marecailo, Ceuta - +6.1
Columbia
Orito, Putumayo Basin - -4.2
Middle East
American Oil YSN 6796A
(Middle East) 2.5 -7.1
6796B (Middle East) 1.4 +3.2
Reference*
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
d
g
g
g
a
a
c
a
a
Iran
Agha Jari
+ .4
162
-------
APPENDIX M (Contfd)
Location, Description % 5 S34S Reference*
Northern Iraq
Ain Zalah Field, Late Cretaceous - -3.2 e
Ain Zalah Field, Middle Cretaceous - -2.8 e
Butmah Field, Late Cretaceous - -5.3 e
Late Triassic - +1.9
Alan Field, Late Triassic - +2.8 e
Demir Dagh Field, Late Cretaceous - -6.4
Kirkuk-Thurraala,-Avanah,-Baba Fields
Tertiary - -5.5 e
Late Cretaceous - -5.3 e
Middle Cretaceous - -5.5 e
Early Cretaceous - -5.9 e
Qarah Cheng Field, Early Cretaceous - -3.9 e
Bai Hassan Field, Tertiary - -7.1 e
Late Cretaceous - -7.0 e
Middle Cretaceous - -7.4 e
Early Cretaceous - -7,5 e
Jambur Field, Tertiary - -6.5 e
Early Cretaceous - -4.0 e
Pulkhana, Field, Late Cretaceous - -5.9 e
Gilabet Field, Tertiary - -3.4 e
Kuwait
Wafra 67065 3.91 -9.8 a
Neutral Zone - -8-1 c
Neutral Zone - -9.8 c
Neutral Zone - -9.7 c
Libya
Saru, Sule Basin •-? c
Saudi Arabia
Ghawar, Ain Dar 57007 1*°9 -2.1 a
Ghawar, Haradh 57005 2.14 "J'J J
Kharsaniya 57002 2.49 -1.6 J
Safaniya 67006 2.97 -8.5 a
Ghawar Blend ~ "vUa
United Arab Republic + c
EL Morgan
Far East
Sumatra - .1 c
~ "
Menas
Duri
163
-------
APPENDIX M
---REFERENCES
a, B. Manowitz and W. Tucker, "Determination of Sulfur Isotope Ratios
in the Atmospheric Diagnostics Program at BNL", Trans. Amer. Nuc.
Soc., 12, 487-8 (1969).
b. B. Manowitz, et. al., "An Isotope Ratio Method for Tracing
Atmospheric Sulfur Pollutants", Chem. Eng. Progr.. Symp. Ser..
66 (104). 163-174 (1970).
c. D. E. Bryan, V. P. Guinn, R. P. Heckelman, and H. R. Lukens,
"Development of Nuclear Analytical Techniques for Oil Slick
Identification", U. S. At. Energy Comm.. 1970. GA 9889, 134 pp.,
Jan. 1970.
d. H. G. Thode, Jan Monster, and H. B. Dunford, "Sulfur Isotope
Abundances in Petroleum and Associated Materials", Bull. Amer.
Assoc. Petrol. Geol.. 4J!, 2619-2641 (1958).
e. H. G. Thode and Jan Monster, "Sulfur Isotope Abundances and Genetic
Relations of Oil Accumulations in the Middle East Basin", Bull.
Amer. Assoc. Petrol. Geol.. j>4_, 627-637 (1970).
f. L. D. Vredenburgh and E. S. Cheney, "Sulfur and Carbon Isotopic
Investigation of Petroleum, Wind River Basin, Wyoming",
Bull. Amer. Assoc. Petrol. Geol.. 55. 1954-1975 (1971).
g. Jan Monster,"Homogeneity of Sulfur and Carbon Isotope Ratios,
34o/32~ and 13P/12P, in Petroleum", Bull. Amer. Assoc. Petrol.
^-y £2 \_j' {j J ir--.UL-.r----. I -|L.-|LL__UI-1- 1 I I --III T- —
Geol.. £6, 941-949 (1972).
164
-------
APPENDIX N
STATISTICAL PROCEDURE
Introduction
Identifying an unknown oil source falls into the category of
a classification problem. It is assumed that there are a finite number
of crude oil sources from which the unknown oil may have come and each
of these crude oil sources is characterized by a probability distribu-
tion of the measured parameters. The unknown oil is considered as a
random observation from the known population of crude oils.
The classification problem may be considered as one of
"statistical decision functions". We have eighty hypotheses: each
hypothesis is that the distribution of the measured parameters for an
unknown oil is the same as the distribution of these measured parameters
for each known oil. Our procedure must test each of these hypotheses
and accept one and reject all the others.
Procedure
The procedure classifies a multivariate x (vector of m
measured properties pertaining to the unknown oil) into population k if:
for all j f k.
In this inequality the q's are a priori probabilities and the p(x)'s
are multivariate normal density functions. For this problem the q's
are assumed equal, i.e., the probabilities are equal that an unknown
oil could have come from any one, of the oils cataloged in the source
165
-------
library. Also the probability must be unity that an unknown oil did
come from one of the sources in the library.
The type of data required for the analysis presented is
shown in Table 1, n is the number of oils in the library, m is the
number of parameters measured on each oil, and nl, n2,..., nn are the
number of replicated samples taken from each crude oil source. The
number of replicates for each oil must be greater than or equal to m
in order that each of thecovariance matrices be non-singular.
In this case the crude oil library consists of 80 (n) oils.
Each crude oil is of known source and is characterized by the measured
parameters. However, only estimates of the means (x) and variances
(a ) for these measured parameters are available for each crude oil
source. This restriction results in covariance matrices whose
estimates of covariance are zero, i.e., it requires the assumption
that the components of the vector x be uncorrelated for the oil
population. This assumption is not unrealistic if oil within a source
is homogeneous.
The procedure also assumes that the means for each oil
population are different, but that their covariance matrices are
alike. For this identification problem, this assumption can not be
made, because for sulfur, vanadium and nickel contents the variances
depend on the magnitude of these parameters. These variances can be
made homogeneous by transforming these variables, but the resulting
transformed variables will not have a normal distribution. There is
no clear cut way to get around this problem. The identification
procedure does require that the variables be distributed normal and
have homogeneous variances. If a decision making process is
166
-------
APPENDIX N
TABLE 1
DATA REQUIRED FOR CLASSIFICATION ANALYSIS
OIL MEASURED PROPERTIES AND CHARACTERISTICS OF THE OIL
(1)
X
1,1
(1)
X
1,2
(1)
X1,n1
x(2)
1,1
(2)
X1,2
*
(2)
X1,n2
(n)
1 1
{n)
(1)
x
2,1
(1)
x
2,2
*
*
X2,n1
(2)
x
2,1
(2)
\2
*
*
(2)
X2,n2
(n)
x
2,1
(n)
xo o
(1)
x • • •
3,1
(1)
x • • •
3,2
*
*
(1)
y * * *
3,n1
(2)
x * • '
3,1
(2)
X3,2
•
&
X3,n2
(n)
3,1
(n)
Xq 9
(D
X
(1)
X
m,2
*
*
(1)
x
m,m
n1 > m
(2)
x
m,1
x(2)
Xm,2
*
*
*
(2)
x _
m,rv2
n2>m
(n)
x
m,
(n)
x
m,2
(n) (n) (n) ... x(n)
\nn \nn X3,nn ^nn nn > m
167
-------
APPENDIX N
incorporated into the identification procedure that will reduce the
total number of potential crude oil sources down to three: the major
candidate and its two nearest competitors, then the ranges of the
measured parameters will be narrowed down to the point, where for all
practical purposes, their variance estimates can be assumed to be the
same. This process greatly reduces the size of the problem and the
complications involved in determining the probabilities of misclassi-
fication.
Initially the 80 crude oils were characterized by measured
parameters. It has been demonstrated that 6 (m) parameters are
sufficient to discriminate among crude oils sources given in the data
library. These six parameters measured on the 600+F bottoms are;
1. Sulfur, Wt. %
O»
2. Nitrogen, Wt. % ==x2
3. Vanadium, ppm =x~
4. Nickel, ppm =x, r~ x
5. Carbon isotopic ratio, 600+F =x_
6. Sulfur isotopic ratio, 600+F =x,
«^
Given the values for these six parameters, an unknown oil can be
classified into its most probable source and its two primary competitors
determined.
Table 2 gives the multivariate normal density function for
population i. It also shows the reduction of this function when the
covariance matrix (3r ') is diagonal.
The inequality (l) can be reduced to the expression
Pj(x) < pk(x) , for all J f k, (2)
168
-------
TABLE 2
MULTIVARIATE NORMAL DENSITY FUNCTION FOR POPULATION i
COVARIANCE MATRIX
(ff«
6X6 MATRIX
= D(i) (DIAGONAL)
REDUCED DENSITY FUNCTION WHEN COVARIANCE MATRIX IS DIAGONAL
1 0)
: (j)
1
= 6X1 VECTOR OF MEASURED PROPERTIES
= 6X1 VECTOR OF MEANS OF MEASURED PROPERTIES FOR POPULATION i.
b2 AND »jW = VARIANCES AND STANDARD DEVIATIONS OF THE SIX
MEASURED PROPERTIES FOR POPULATION i.
H
-------
APPENDIX N
since the a priori probabilities (q's) are assumed equal. Substituting
the density functions into this expression gives:
EXP
EXP
xi-^i
(3)
for all j f k.
This says to classify the vector of components measured on
the unknown oil into the population (source oil) whose density function,
when evaluated for x, is largest. This in essence is the population (k)
that minimizes the expression:
(4)
where p. is an estimate of the mean and a. (k) is an estimate of the
standard deviation for variable i of population k. This expression is
a portion of the exponential term in the density function. The failure
of the variance to be homogeneous among oil sources could conceivably
result in different conclusions being drawn from minimizing expression
(4) and maximizing the density functions, but the likelihood of this is
remote.
The procedure numerically evaluates expression (4) for each
of the eighty crude oils given in the data library. The source crude
oils are ranked from most likely to least likely that x could have
*-w
come from their distribution based on the evaluation. The procedure
170
-------
APPENDIX N
then takes the oil source that ranks first as the most probable candi-
date for x to have come from and the oil sources that rank 2 and 3 as
its primary competitors. The density functions for these 3 oil
sources are evaluated to confirm that the ranking is correct and not
influenced by the failure of variances to be homogeneous.
The assumption of equivilant covariance matrices for the
three oil populations determined in the previous step is now reasonable,
since the means of each parameter in these 3 crude oils cover a narrower
range. An estimate of the covariance matrix that is representative
of the 3 populations is obtained by pooling the covariance matrices for
the three crude oils.
This covariance matrix is then substituted into expression (3)
and the resulting inequality becomes:
6
(2ir)-
EXP
EXP
'l-,ai
(k) ^2
L)
(5)
for all j ^ k. The difference between (3) and (5) is that the standard
deviations are no longer identified with a particular crude oil source.
They are, in fact, the same for the three oil crude sources in question.
Expression (5) can be reduced farther by cancelling the constant terms
and taking the logarithm of both sides. The expression simplifies to:
Wi(d)
*rV^-
i=l\ a.
for all j ^ k
(6)
171
-------
APPENDIX N
This says, classify the vector of components x into population (crude
rw
oil source) k when the above inequality is satisfied for all j ^ k. In
other words, classify x into the population that yields a minimum value
rw
for:
(7)
This is the same criterion used to select the 3 candidate oils. Had
the covariance matrix been the same for all eighty crude oil sources,
expressions (4) and (?) would have been the same and the solution would
be greatly simplified.
Equation (6) may be further simplified:
> 0. (8)
The left hand side of this inequality is a linear function of normal
variables. Therefore, assuming the vector is a sample from crude oil
source k, then the variable
is distributed normal with mean — a. . and variance a, .,
*• K-3 KJ
where:
-------
APPENDIX N
The probability of misclassification is defined as:
Prob. {of not classifying x in population k, given that
came from population k}
= Prob {UkJ < 0; for all j f k}
= 1 - Prob {Ukj > 0; for all j ^ k}
Letting j = 1, 2 or 3 and evaluating the procedure for k = 3 this
probability statement becomes:
= Prob {of not classifying x in pop. 3 given that x came
from pop. 3}
= 1 - Prob
0,
0}
The vector
where
(11)
(12)
(13)
(14)
and the means and variances are defined as above. Since the vector
is distributed as a bivariate normal the Prob {U. > 0, U OJ =
ee
dU
31
(15)
where g
.
:p) is the bivariate normal density function.
The vector IL can be standardized to means 0, variance and correlation
p, and the value of this probability (15) can be looked up in available
bivariate normal tables. The probability of misclassification is 1
minus this value.
173
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Rec
w
A MULTIPARAMETER OIL POLLUTION SOURCE
IDENTIFICATION SYSTEM,
5, R.- ,;rr£
6. .
Miller, John W.
Phillips Petroleum Company
Research and Development Department
Bartlesville, Oklahoma 74004
15080 HDJ
68-01-0059
.
Period Cc»-c•;-/ John W. Miller
Phillips Petroleum Company
------- |