AUTHORS

Cari L. Johnson � Department of Geologicaland Environmental Sciences, Building 320, StanfordUniversity, Stanford, California, 94305-2115; currentaddress: Department of Geology and Geophysics,University of Utah, 135 South 1460 East, BrowningBuilding, Salt Lake City, Utah, 84112-0011;cjohnson@mines.utah.edu

Cari Johnson is an assistant professor at theUniversity of Utah. She received geology degreesfrom Carleton College (B.A., 1996) and StanfordUniversity (Ph.D., 2002), where she also completedpostdoctoral research on sequence stratigraphy ofthe San Joaquin basin. Her dissertation focused onthe sedimentary record of Late Mezosoic extensionin the China–Mongolia border region. She continuesto conduct research in basin analysis, sedimentationand tectonics, and petroleum geology in east-centralAsia and western North America.

Todd J. Greene � Department of Geological andEnvironmental Sciences, Building 320, StanfordUniversity, Stanford, California, 94305-2115; currentaddress: Anadarko Petroleum Corporation, 1201Lake Robbins Drive, The Woodlands, Texas, 77380

Todd Greene attained a B.S. degree in earthsciences from the University of California at SantaCruz (1994) and a Ph.D. in geological sciences atStanford University (2000). His dissertation focusedon tectonics, sedimentology, organic geochemistry,and petroleum systems of the Turpan-Hami basin ofnorthwestern China. He is currently employed byAnadarko Petroleum in Houston, Texas, where he ispart of a regional studies team investigating basinsand play types in the greater Rocky Mountains.

David A. Zinniker � Department of Geologicaland Environmental Sciences, Building 320, StanfordUniversity, Stanford, California, 94305-2115

David A. Zinniker is a Ph.D. candidate in the Depart-ment of Geological and Environmental Sciences atStanford University. His research focuses on molec-ular fossils of plants and algae and their bearingupon ecology, evolution, depositional systems, andpetroleum geology. His future projects include usingmolecular and macromolecular markers to studycurrent ecological processes and events deep ingeologic time.

J. Michael Moldowan � Department of Geolog-ical and Environmental Sciences, Building 320,Stanford University, Stanford, California, 94305-2115

J. Michael Moldowan attained a B.S. degree inchemistry from Wayne State University, 1968, anda Ph.D. in chemistry from the University of Michi-gan in 1972. Following a postdoctoral fellowship inmarine natural products with Professor Carl Djerassi

Geochemical characteristics andcorrelation of oil and nonmarinesource rocks from MongoliaCari L. Johnson, Todd J. Greene, David A. Zinniker,J. Michael Moldowan, Marc S. Hendrix, andAlan R. Carroll

ABSTRACT

New bulk and molecular organic geochemical analyses of source rock

and oil samples from Mongolia indicate the presence of lacustrine-

sourced petroleum systems in this frontier region. Samples of po-

tential source rocks include carbonate, coal, and lacustrine-mudstone

lithologies that range from Paleozoic to Mesozoic in age, and rep-

resent a variety of tectonic settings and depositional environments.

Rock-Eval and total organic carbon data from these samples reflect

generally high-quality source rocks, including both oil- and gas-prone

kerogen types, mainly in the early stages of generation. Bulk geo-

chemical and biomarker data indicate that Lower Cretaceous lacus-

trine mudstone found in core from the Zuunbayan field is the most

likely source facies for the East Gobi basin of southeastern Mon-

golia. Oil and selected source rock samples from the Zuunbayan

and Tsagan Els fields (both in the East Gobi basin) demonstrate geo-

chemical characteristics consistent with nonmarine source environ-

ments and indicate strong evidence for algal input into fresh- to

brackish-water source facies, including elevated concentrations of

unusual hexacyclic and heptacyclic polyprenoids. Despite similar-

ities between Zuunbayan and Tsagan Els oil samples, biomarker

parameters suggest higher algal input in facies sourcing Zuunbayan

oil compared to Tsagan Els oil. Tsagan Els oil samples are also gen-

erated by distinctly more mature source rocks than oil from the

Zuunbayan field, based on sterane and hopane isomerization ratios.

INTRODUCTION

Central and eastern China contain several petroleum-bearing, late

Mesozoic rift basins (e.g., Songliao and Erlian; Figure 1, inset). Var-

ious studies address these Asian lacustrine systems from regional

stratigraphic and source-geochemical perspectives (Yang, 1985; Fu

Copyright #2003. The American Association of Petroleum Geologists. All rights reserved.

Manuscript received January 18, 2002; provisional acceptance August 20, 2002; revised manuscriptreceived November 25, 2002; final acceptance December 17, 2002.

AAPG Bulletin, v. 87, no. 5 (May 2003), pp. 817–846 817

818 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

and Sheng, 1992; Gao et al., 1997; Lin et al., 2001). Comparatively

little is known about related Mesozoic rift basins in Mongolia (spe-

cifically, the Tamtsag and East Gobi basins; Figure 1), where pe-

troleum systems remain poorly understood and underexplored

(Penttila, 1994; Sladen and Traynor, 2000).

Penttila (1994) estimated as much as 3–6 billion BOE recoverable

hydrocarbon resources in Mongolia, although reserve calculations are

difficult to constrain in this poorly known petroleum province. Petro-

leum production in Mongolia is currently limited to the East Gobi

and Tamtsag basins (Figure 1), the latter hosting discoveries in 2001

totaling an estimated 1.5 billion bbl of oil in place (Soco Inter-

national, 2001). Although historic discoveries in Mongolia generally

have been modest, the potential for larger hydrocarbon accumula-

tions exists by analogy to similar basins in China. These include the

largest producing field in China, Daqing (production �1.0 million

bbl/day; United States Department of Energy, 2001), in addition to

smaller accumulations in Erlian basin (e.g., Aershan field with esti-

mated 100 million bbl reserves; Sladen and Traynor, 2000).

The East Gobi basin shares several characteristics with late

Mesozoic rift basins in China. These similarities include the age and

depositional style of nonmarine basin fill, and northeast-southwest

orientation of rift structures (Liu, 1986; Watson et al., 1987). Rifting

began during the Late Jurassic and continued through the Early Cre-

taceous, with widespread deposition of fluvial-lacustrine strata and

periodic bimodal volcanic activity (Traynor and Sladen, 1995; Johnson

et al., 2001). Middle Cretaceous contraction and strike-slip faulting

inverted the East Gobi basin along its margin, forming a regional

angular unconformity overlapped by an Upper Cretaceous postrift

sequence (Figure 2) (Graham et al., 2001). The total original thick-

ness of synrift strata is not known because of erosion during the

basin-inversion event, but more than 2.5 km of sedimentary and

volcanic graben fill is still preserved in the subsurface of the East

Gobi, based on proprietary seismic and well-log data (Johnson, 2002)

and correlative outcrop studies (Graham et al., 2001).

Synrift strata form source, reservoir, and seal units in south-

eastern Mongolia, mainly in inversion-related structural traps (Fig-

ure 2). In the East Gobi basin, the main source rocks appear to be

lacustrine shales of the Lower Cretaceous Tsagantsav Formation

(synrift sequence 3 of Graham et al., 2001), based on previous geo-

chemical studies of source rocks from southeastern Mongolia (Yama-

moto et al., 1998). This formation is widely distributed in eastern

Mongolia and correlates to Lower Cretaceous lacustrine shale in the

Nilga and Tamtsag basins as well (Badamgarav et al., 1995; Neves

et al., 2000). The overlying Zuunbayan Group (synrift sequence 4,

Figure 2) is generally more sand rich, although fine-grained lacus-

trine units are also present (Jerzykiewicz and Russell, 1991). Pe-

troleum reservoirs mainly occur in fluvial to perilacustrine (deltaic)

sandstone of the Lower Cretaceous strata (Johnson et al., 2000)

and are commonly limited in quality by abundant lithic grains derived

from synrift volcanic activity, in addition to associated porosity-

reducing zeolite cements. Particularly in Mongolia, outcropping synrift

at Stanford University, he joined Chevron’s BiomarkerGroup in 1974. Moldowan joined the Department ofGeological and Environmental Sciences of StanfordUniversity as professor (research) in 1993.

Marc S. Hendrix � Department of Geology,University of Montana, Missoula, Montana, 59812

Marc S. Hendrix received geology degrees fromWittenberg University (B.A., 1985), the Universityof Wisconsin, Madison (M.S., 1987), and StanfordUniversity (Ph.D., 1992). He completed postdoctoralresearch at Stanford in 1994 and since has been aprofessor of geology at the University of Montana,Missoula. His research interests include sedimentarybasins and paleoclimate studies, particularly inwestern North America and central Asia.

Alan R. Carroll � Department of Geology andGeophysics, University of Wisconsin, Madison, 1215W. Dayton St., Madison, Wisconsin, 53706

Alan Carroll conducts research on large lake basinsin Asia and the western United States, focusing ontheir tectonic setting, sequence stratigraphy, andpetroleum potential. He worked for three years asan exploration geologist for Sohio, and five years forExxon Production Research. He is currently an asso-ciate professor at the University of Wisconsin.

ACKNOWLEDGEMENTS

We thank D. Badamgarav, G. Badarch, R. Barsbold,D. Janchiv, and our other colleagues at the MongolianAcademy of Sciences, Institute of Geology and Min-eral Resources, the Petroleum Authority of Mongolia,and the Mongolian Paleontological Institute for theirscientific and logistic support. Funding for this studywas provided by Roc Oil, the Stanford GraduateFellowship, the International Research and ExchangeBoard, the Stanford-Mongolia Industrial Affiliatesprogram, and by National Science Foundation grantsEAR-9708207 and EAR-961455 to S. Graham and M.Hendrix, respectively. S. Graham was a principaladvisor on this and related projects in Mongolia andoffered much insight throughout this study. Col-leagues at various labs assisted with sample analysis,including K. K. Bissada (Houston Advanced ResearchCenter), H. Hada (Micropaleoconsultants), andZhengzheng Chen, D. Mucciarone, and F. Fago atStanford University. Additional bulk geochemicalanalyses were completed at ExxonMobil UpstreamResearch (formerly EPRC). We thank P. Albrecht andP. Adam for discussion and supporting data onpolycyclized polyprenoids and E. Chang (StanfordUniversity) for the Chinese oil samples. J. Amory,M. Beck, L. Lamb, R. Lenegan, D. Sjostrom, E. Sobel,and L. Webb provided additional assistance in thefield. Reviews by M. Fowler, K. Peters, and Wan Yanggreatly improved this manuscript.

Johnson et al. 819

Figu

re1

.Reg

iona

lmap

show

ing

gene

ralo

utlin

esof

Mes

ozoi

cse

dim

enta

ryba

sins

ofM

ongo

lia(s

eein

setm

aps

for

loca

tion

ofne

arby

Chi

nese

basi

ns,a

ndde

tail

ofth

eZu

unba

yan

area

).Ba

sin

outli

nes

and

nam

esre

vise

dfr

omPe

nttil

a(1

994)

and

Stei

nsho

uer

etal

.(19

99).

Sour

cero

cksa

mpl

elo

calit

ies

and

Zuun

baya

n/Ts

agan

Els

oilf

ield

sar

esh

own.

See

Tabl

e1

for

addi

tiona

lsa

mpl

ean

dlo

calit

yin

form

atio

n.

820 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

fluvial-lacustrine facies have only recently been described

in detail (Graham et al., 2001), and there remains almost

no English-language published literature discussing sub-

surface data (Johnson, 2002).

The organic geochemistry database for Mongolian

oil and source rocks is also limited. Sladen and Traynor

(2000) reported Rock-Eval and total organic carbon

(TOC) data from Paleozoic–Mesozoic source rocks, as

well as oil data suggesting geochemical oil-source cor-

relation to low-maturity Lower Cretaceous lacustrine

mudstones in southeastern Mongolia, including com-

bustible oil shales. Palynological evidence for green

algae (including Botryococcus), and the presence of

b-carotane in source rock extracts indicate algal input in

slightly saline to freshwater lakes with anoxic bottom

waters throughout the synrift sequences (Neves et al.,

2000; Sladen and Traynor, 2000). Yamamoto et al. (1993,

1998) also reported geochemical evidence for bloom-

ing autotrophic prokaryotes and thermal stratification

of Cretaceous lakes with anoxic lake-bottom conditions

in the East Gobi and Nilga basins. Similarly, dinoflag-

ellate blooms signaling periods of high nutrient flux

were common in lakes of the Tamtsag basin (Neves

et al., 2000) and in China (Wang Renhou et al., 1996;

Wan et al., 1997) during the Early Cretaceous. Most of

these interpretations are based on spore/pollen analyses,

Rock-Eval data, and whole-rock geochemistry, lacking

detailed biomarker data.

The purpose of this study is twofold. First, we pres-

ent the results of reconnaissance sampling of a range

of source rocks from throughout Mongolia. Bulk geo-

chemical analyses of this sample suite demonstrate the

presence of a variety of potential source rocks, but strong-

ly suggest that known occurrences of oil in southeastern

Mongolia are sourced by Lower Cretaceous lacustrine

mudstone. We investigate this correlation further in the

second part of the study, by examining detailed source-

rock-to-oil relationships in the Tsagan Els and Zuun-

bayan fields of the East Gobi basin.

METHODS

Our database includes oil and source rock samples col-

lected from 1992 to 2000 during fieldwork in Mongo-

lia (Table 1). Sample preparation and analyses were

performed at several laboratories, although the major-

ity of molecular data presented here were collected at

Stanford University’s Molecular Organic Geochemis-

try Laboratory. Source rock samples were evaluated by

Rock-Eval and TOC analyses using standard proce-

dures (Table 2). For selected source rock samples, sol-

uble bitumens were extracted from rock powders using

a Soxhlet apparatus and a mixture of methanol and tol-

uene solvents. Samples collected in 1992–1993 (sample

numbers beginning with 92 or 93, Table 1), were ex-

tracted and analyzed using standard (nC12 and higher)

gas chromatography at ExxonMobil Upstream Research

(formerly EPRC). Subsequently collected samples were

extracted and analyzed by gas chromatography with

flame ionization detection (GC-FID) at Stanford Uni-

versity using a Hewlett-Packard 5890A gas chromato-

graph. Other conventional organic geochemical analyses

(stable carbon isotope, sulfur, wax, etc.) were completed

at the Houston Advanced Research Center and at Stan-

ford University.

All reported rock extracts and oil samples were

separated by high-performance liquid chromatography

(HPLC) into saturate and aromatic fractions at Stan-

ford University, following procedures outlined by Peters

and Moldowan (1993). Saturate cuts were further pre-

pared using molecular sieves (silicalite) to remove all

of the n-alkanes and increase the signal of more di-

agnostic biomarkers. Saturate and aromatic fractions

were analyzed on a Hewlett-Packard 5890 Series II-Trio

1 VG Masslab gas chromatograph–mass spectrometer

(GCMS) system. A Hewlett-Packard 5890 Series II-

Micromass Autospec Q hybrid GCMS system was used

for further sterane analyses using MRM-GCMS (meta-

stable reaction monitoring), in addition to scan runs for

analysis of hexacyclic and heptacyclic cyclic polyprenoids.

SOURCE ROCK SAMPLE DESCRIPTION

More than 75 potential source rocks ranging in age

from Riphean–Cambrian to Cretaceous, were sampled

between 1992 and 1999. The source rocks are mainly

from outcrops in Mongolia spanning some 800,000

km2 of sampling area (Figure 1), but also include four

subsurface core samples (ZB core, Table 1). Lithologies

include shale, calcareous mudstone, coal, coaly mudstone,

and carbonate (Table 2). Ages are mainly based on pub-

lished and unpublished mapping by Mongolian agencies

Johnson et al. 821

Figure 2. Overview of Mesozoic tectonostratigraphic units in southeastern Mongolia. Sequence-stratigraphic nomenclature of Traynor and Sladen(1995), Graham et al. (2001), and Johnson (2002) are shown, along with interpreted tectonic settings and local formation names. The total thicknessof preserved synrift deposits in the subsurface is approximately 2.5–3 km as estimated from seismic reflection profiles and velocity models.

(e.g., Yanshin, 1989; Badarch, 1999, personal commu-

nication), in addition to spore/pollen analyses, vertebrate

and invertebrate faunal assemblages, and 40Ar/39Ar data

where available (Jerzykiewicz and Russell, 1991; Gra-

ham et al., 2001).

Our source rock sample suite encompasses a range

of tectonostratigraphic units representing the geologi-

cally diverse history of Mongolia. The oldest samples

are Riphean–Cambrian carbonates from northern Mon-

golia (MO and WH localities; Figure 1), which formed

on carbonate platforms of the south-facing Siberian pas-

sive margin (megasequence 1 of Traynor and Sladen,

1995). These rocks predate assembly of the present-

day Mongolian basement, which occurred mainly during

the Paleozoic through amalgamation of several volcan-

ic arcs and related basins (Lamb and Badarch, 1997).

Carboniferous–Permian coal and coaly mudstone sam-

ples (TT, SJ localities) immediately postdate these

Paleozoic collisions and represent the beginning of non-

marine deposition in central Mongolia (Traynor and

Sladen, 1995; Amory, 1996).

Triassic and Lower Jurassic samples from western,

central, and southern Mongolia (samples NU, JL, CM,

DZ, and SO) formed in lakes and swamps of foreland

and intermontane basins associated with continuing col-

lisions along the margins of central Asia (megasequence

3 of Traynor and Sladen, 1995; Hendrix et al., 2001;

Sjostrom et al., 2001). In the East Gobi basin, a Lower

to Middle Jurassic prerift section is represented by the

Khamarkhavoor Formation (Figure 2), which includes

thin coal seams and palustrine shale (Graham et al.,

2001). Samples from this unit were not analyzed in the

present study, but future analyses may address the po-

tential for a Jurassic, prerift source in the East Gobi basin.

This period of early Mesozoic contractile tectonism

ended with widespread uplift and erosion during the

Middle Jurassic, followed by Late Jurassic–Early Cre-

taceous extension in eastern and southeastern Mongo-

lia. Upper Mesozoic coal and lacustrine shale samples

from the East Gobi and related basins are part of this

intracontinental rift sequence (BT, NL, SH, TH, ZB,

and UB samples). SH samples are from the type Shin-

hudag section outcropping in the Nilga basin, north of

the East Gobi (Figure 1). The locality has more than

100 m of weathered marl and laminated mudstone,

and is part of the Zuunbayan Group, a late synrift se-

quence (Lower Cretaceous; Badamgarav et al., 1995).

TH samples are from the synrift section deposited on

822 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Table 1. List of Sample Abbreviations, Locality Names, and Location*

Sample Locality Latitude (N) Longitude (E)

Source Rocks97-SH-(02–13) Shinhudag 44.71 107.93

98-TH-(314–316) Tavan Har 44.03 109.37

ZB-(1716–1978) Zuunbayan core (SWZB-310b well) 44.25 109.55

92-UB-(4–15) Ulaan Bataar 47.46 107.18

93-NL-(3) Nilga basin 44.80 106.95

94-BT-(1–3) Bayan Tolgoy 45.70 101.54

92-JL-(6–9) Jargalant 47.30 92.37

92-CM-(3–22) Chandaman 45.22 97.53

92-DZ-(2–6) Dzinst 44.34 99.13

93-SO-(1–209) Sayan Obo 45.73 105.18

92-NU-(48–52) Noyon Uul 43.14 101.55

93-TT-(201–302) Tsogt Tsitsee 43.62 105.47

92-SJ-(36a) Shin Jeanst 44.22 99.25

93-MO-(2–4) Moron 50.55 101.18

93-WH-(3–4) West Hobsgul 50.55 101.18

Oil SamplesTE-(A1-25) Tsagan Els field 44.00 109.75

ZB-(310b; 163251a–b) Zuunbayan field 44.25 109.90

ER-(1846) Erlian basin (Saihantala sag) 42.25 110.50

DQ-(2198) Daqing (Yushuling field) 45.75 125.50

*Sample abbreviations based on last two digits of sampling year, two-letter locality abbreviation, and sample number. Locality names are based on local towns,landmarks, or formation names. Sample locations are in decimal degrees latitude and longitude.

the Tavan Har basement structure, about 50 km south-

west of the Zuunbayan field. The section includes

about 300 m of oil shale overlain by volcanic ash and

basalt units, and can be tied to the Tsagantsav Forma-

tion (Lower Cretaceous) by proprietary regional seis-

mic reflection profiles. UB samples were all collected

from an open-pit coal mine about 15 km east of Ulaan

Bataar. Exposed strata in the mine are reportedly of

Albian–Aptian age (D. Badamgarav, 1992, personal

communication) and include coal, organic-rich clay-

stone, and sandstone. Abundant climbing ripples and

soft-sediment deformation in the sandstone beds sug-

gest they were deposited rapidly, possibly as crevasse

splay deposits in a fluvial overbank or swamp setting

(Farrell, 2001). BT samples were collected from an open-

pit coal mine in the centralGobi basin.This coal is mapped

as Cretaceous (Yanshin, 1989), although no published

age data are available. The NL samples are from an ex-

humed oil field that is presently being quarried for

tar sand in the Nilga basin, about 200 km southeast

of Ulaan Bataar. Source rock samples from this basin

include laminated, organic-rich shale.

ZB core samples are taken from conventional 6-cm-

diameter core of the ZB-310b well in the southwest

Zuunbayan field. Over 600 m of Lower Cretaceous

fluvial-lacustrine facies comprise the Zuunbayan core

(Graham et al., 2001). A 125-m-thick section of the

core (unit 2, Figure 3) consists of finely laminated

mudstone and micrite, dolomitic breccia, and calcar-

eous siltstone (described in Johnson, 2002). These fine-

grained units are interbedded with grainstone and thin,

normally graded sandstone beds interpreted as distal

lacustrine turbidites (Grabowski and Pevear, 1985).

This lacustrine sequence has been mapped in both the

Zuunbayan and Tsagen Els subbasins based on well-log

and seismic reflection character (A. Hall, 2000, person-

al communication). Evidence for anoxic conditions at

the lake bottom includes framboidal (biogenic) pyrite,

high %TOC, dominance of light-colored amorphous

organic constituents, carbonate precipitation, and a pau-

city of trace fossils relative to the rest of the core (John-

son, 2002). Abundant fish remain, possible algal cysts,

and conchostrachan and ostracod valves attest to better

oxygenation in the upper parts of the water column. The

lower half of the lacustrine section (depths 605–550 m)

is dominated by laminated micrite and dolomitic breccia

units (thought to be subaqueously precipitated; Wolf-

bauer and Surdam, 1974; Johnson, 2002), and generally

lacks trace and invertebrate fossils. The upper section

(544–468-m depth) contains more of these fossils

(mainly gastropods, ostracods, and conchostrachans)

in poorly laminated calcareous siltstone, having local

traction-current structures such as thin, normally grad-

ed, rippled beds. The lower section is interpreted as an

anoxic lake-bottom sequence in a stratified lake, which

includes samples ZB-1978 and ZB-1933 (Powell, 1986).

The upper half, which includes sample ZB-1740, appears

to have better oxygenation resulting from more circula-

tion at the lake bottom, representing a nonstratified lake

or possibly more marginal facies (Powell, 1986). Sample

numbers represent depth below KB (4 m) in feet (con-

verted to meters on Figure 3).

SOURCE ROCK BULK GEOCHEMISTRY

Rock-Eval pyrolysis indicates a suite of generally high-

quality source rocks (Table 2), excluding the Riphaen–

Cambrian strata. TOC ranges from 1.5 to 15% for shale

and 40 to 75% for coal and coaly mudstone samples.

More than 75% of the samples have S1 and S2 values

greater than 0.5 and 10, respectively, also indicating

good to very good quality source rocks (Peters, 1986).

Source rock maturities are mainly immature to mid-

dle oil window, although the Triassic–Jurassic NU

and Permian TT samples approach peak oil window to

overmature based on Tmax, PI, and Ro values (Tables 2,

3; Peters and Moldowan, 1993). HI-versus-OI values

(Figure 4) indicate dominantly type I and type II (oil-

prone) kerogen, but also include type III (gas-prone)

kerogen in some of the lower Mesozoic coal samples

(Tissot et al., 1974). Visual maceral analysis confirms a

range of oil-prone liptinite (from algal and waxy-plant

debris) in the Triassic–Jurassic oil shales (NU sam-

ples), whereas gas-prone vitrinite and inertinite (type

IV nongenerative kerogen of Demaison et al., 1983)

dominate the coal samples (JL, CM, and DZ samples;

Table 3). Gas chromatographic data were collected

from whole-rock extracts for selected Mesozoic sam-

ples (Table 4). Maximum n-alkane peaks are mainly

between nC15–nC25 (Figure 5), show a rapid falloff of

higher n-alkanes (>nC29), and a slight to pronounced

odd-over-even preference, particularly in the Lower

Cretaceous samples.

The source rocks form five main groups reflecting

distinct ages, lithofacies, geographic distribution, kero-

gen type, and maturity (Table 5). Lower Cretaceous

rocks from central and southern Mongolia (group 1)

include TOC values of 1.5–30% for mudstone and

oil shale and greater than 45% for coaly samples. HI-

versus-OI ratios indicate variable kerogen type in group

1 (Figure 4) (Peters, 1986; Tissot et al., 1974). SH and

Johnson et al. 823

824 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Table 2. Rock-Eval plus TOC Data for Mongolian Source Rock Samples*

Age Sample Lithology Lab %TOC S1 S2 S3 PI Tmax OI HI S2/S3 S1/TOC

Late Cretaceous 97-SH-02 shale 1 9.83 0.76 63.04 5.49 0.01 444 56 641 11.48 7.73

97-SH-03 shale 1 5.72 0.40 25.36 4.39 0.02 433 77 443 5.78 6.99

97-SH-04 shale 1 6.70 0.58 40.12 4.25 0.01 439 63 599 9.44 8.66

97-SH-05 shale 1 7.84 0.90 43.56 5.55 0.02 441 71 556 7.85 11.48

97-SH-07 shale 1 4.60 0.41 25.74 3.91 0.02 437 85 560 6.58 8.91

97-SH-09 shale 1 8.39 1.17 57.78 4.16 0.02 443 50 689 13.89 13.95

97-SH-11 shale 1 6.56 0.93 45.04 3.32 0.02 438 51 687 13.57 14.18

97-SH-12 shale 1 14.60 1.50 132.40 6.10 0.01 446 42 907 21.70 10.27

97-SH-13 shale 1 7.10 0.53 56.14 3.75 0.01 440 53 791 14.97 7.46

98-TH-314 shale 2 1.54 0.32 4.98 0.08 0.06 437 5 323 62.25 20.78

98-TH-316 shale 2 2.30 0.51 10.09 0.20 0.05 433 8 438 50.45 22.21

ZB-1716 (core) mudstone 3 1.80 0.25 7.57 0.40 0.03 438 22 421 18.93 14.00

ZB-1740 (core) mudstone 3 2.03 0.35 9.13 0.48 0.04 439 24 450 19.02 17.00

ZB-1933 (core) mudstone 3 4.23 1.54 24.67 0.71 0.06 434 17 583 34.75 36.00

ZB-1978 (core) mudstone 3 2.00 0.31 8.46 0.44 0.04 438 22 423 19.23 16.00

92-UB-4 coal 4 51.77 0.70 54.92 15.43 0.01 419 30 106 3.56 1.35

92-UB-10 coaly mudstone 4 44.77 2.44 122.35 9.54 0.02 426 21 273 12.82 5.45

92-UB-12 coal 4 54.27 0.82 74.97 15.27 0.01 411 28 138 4.91 1.51

92-UB-15 carbonate 4 10.47 0.28 21.72 2.64 0.01 425 25 207 8.23 2.67

93-NL-3 shale 5 2.65 0.11 7.96 1.87 0.01 434 71 300 4.26 4.15

94-BT-1 coal 5 59.22 2.53 196.40 9.05 0.01 428 15 332 21.70 4.27

94-BT-2 coal 5 56.75 3.15 236.80 1.78 0.01 425 3 417 133.03 5.55

94-BT-3 shale 5 29.94 2.89 144.90 2.12 0.02 432 7 484 68.35 9.65

Lower–Middle

Jurassic 92-JL-6 coaly mudstone 4 48.89 1.01 35.56 40.45 0.03 422 83 73 0.88 2.07

92-JL-9 coaly mudstone 4 44.43 0.83 17.25 36.51 0.05 422 82 39 0.47 1.87

92-CM-3 coal 4 64.54 0.87 32.23 22.46 0.03 429 35 50 1.43 1.35

92-CM-10 coal 4 61.27 1.47 41.59 21.02 0.03 427 34 68 1.98 2.40

92-CM-12 calcareous mudstone 4 33.23 1.86 38.29 13.20 0.05 428 40 115 2.90 5.60

92-CM-17 coal 4 55.49 2.85 44.62 14.54 0.06 428 26 80 3.07 5.14

92-CM-20 coal 4 59.85 1.51 39.64 21.23 0.04 427 35 66 1.87 2.52

92-CM-22 calcareous siltstone 4 11.03 0.75 17.44 2.68 0.04 437 24 158 6.51 6.80

92-DZ-2 calcareous siltstone 4 2.46 0.16 2.95 0.63 0.05 436 26 120 4.68 6.50

92-DZ-6 coal 4 52.47 0.74 32.43 23.75 0.02 432 45 62 1.37 1.41

93-SO-1 shale 5 13.23 1.02 54.19 1.60 0.02 429 12 410 33.87 7.71

93-SO-2 shale 5 14.10 1.18 60.27 2.27 0.02 429 16 427 26.55 8.37

93-SO-3 shale 5 22.64 2.26 97.17 2.31 0.02 427 10 429 42.06 9.98

93-SO-201 coaly mudstone 5 49.65 1.99 209.35 5.87 0.01 421 12 422 35.66 4.01

93-SO-207 coal 5 63.24 1.52 192.66 8.28 0.01 423 13 305 23.27 2.40

93-SO-208 coaly mudstone 5 43.42 2.38 251.17 2.50 0.01 434 6 578 100.47 5.48

93-SO-209 coal 5 55.41 2.56 238.79 3.48 0.01 429 6 431 68.62 4.62

Triassic–Jurassic 92-NU-48B shale 4 3.36 0.13 18.10 0.28 0.01 443 8 539 64.64 3.87

92-NU-49 shale 4 2.92 0.24 18.06 0.34 0.01 444 12 618 53.12 8.22

92-NU-50 shale 4 3.01 0.23 19.52 0.26 0.01 444 9 649 75.08 7.64

92-NU-51 shale 4 2.68 0.22 16.01 0.31 0.01 441 12 597 51.65 8.21

92-NU-52 shale 4 3.48 0.20 17.28 0.35 0.01 441 10 497 49.37 5.75

NL samples from the Nilga basin have relatively high

HI and OI values, whereas other source rocks from

central and southern Mongolia have lower OI values

and variable HI values ranging from type I and II (ZB,

TH, and BT samples) to near type III (gas-prone) ker-

ogen in the UB coal samples (which also generally show

a predominance of vitrinite macerals, Table 3). Tmax

values in group 1 range from 411 to 444 (Table 2),

indicating immature to early oil window maturity levels.

PI values are consistent with this interpretation, ranging

from 0.01 to 0.02 for most group 1 samples and slightly

higher (0.03–0.06) for ZB and TH samples from the

East Gobi basin. Similarly, a high carbon preference in-

dex for SH and TH outcrop samples is likely the result

of lower thermal maturities in these outcrops along the

basin margins (Figure 6) (Peters and Moldowan, 1993).

Lower Mesozoic samples from central and western

Mongolia consist of shale, coaly mudstone, and coal li-

thologies having TOC values ranging from 2 to 65%.

These Triassic to Lower Jurassic samples form two

groups distinguished by kerogen type and maturity

(Table 5): Lower to Middle Jurassic JL, CM, and DZ

samples (group 2) contain predominantly vitrinite

and intertinite macerals, and plot within the type II

and type III (liquid to slightly gas-prone) range on a

modified Van Krevelen diagram (Figure 4). Triassic to

Lower Jurassic NU and SO samples (group 3) are liquid-

source prone based on maceral analysis (‘‘plant tissue’’

Johnson et al. 825

Age Sample Lithology Lab %TOC S1 S2 S3 PI Tmax OI HI S2/S3 S1/TOC

Permian 93-TT-201 coal 4 69.1 6.31 173.6 1.44 0.04 453 2.1 251.3 121 0.09

93-TT-202 coal 4 62.6 5.61 145.7 1.1 0.04 447 1.8 232.8 132 0.09

93-TT-204 coaly mudstone 4 20.9 1.5 51.1 0.86 0.03 452 4.1 244.4 59 0.07

93-TT-205 coal 4 66.6 3 143.8 1.32 0.02 446 2.0 216.0 109 0.05

93-TT-206 coal 4 72.7 5.71 173.9 0.86 0.03 452 1.2 239.2 202 0.08

93-TT-207 coal 4 68.2 7.07 179.8 1.01 0.04 452 1.5 263.7 178 0.10

93-TT-208 coal 4 73.2 5.53 182.1 1.27 0.03 450 1.7 248.8 143 0.08

93-TT-209 coal 4 68.7 7.31 179.6 0.91 0.04 453 1.3 261.5 197 0.11

93-TT-210 coal 4 66.9 5.38 147.4 0.53 0.04 448 0.8 220.4 278 0.08

93-TT-211 coal 4 72.3 5.56 182.9 0.74 0.03 451 1.0 252.9 247 0.08

93-TT-212 coal 4 74.6 4.63 175.9 1.23 0.03 454 1.6 235.6 143 0.06

93-TT-213 coal 4 67.4 6.79 111.6 0.96 0.06 447 1.4 165.7 116 0.10

93-TT-214 coal 4 70.3 5.75 128.4 1.5 0.04 453 2.1 182.8 86 0.08

93-TT-215 coal 4 64.6 5.15 153.3 0.86 0.03 448 1.3 237.3 178 0.08

93-TT-216 coal 4 76.4 7.42 207.7 0.52 0.04 449 0.7 271.8 399 0.10

93-TT-217 coal 4 59.2 4.2 152.9 0.7 0.03 448 1.2 258.2 218 0.07

93-TT-218 coal 4 70.0 4.88 143.6 1.33 0.03 446 1.9 205.1 108 0.07

93-TT-219 coal 4 62.4 5.2 108.0 1.09 0.05 447 1.7 173.2 99 0.08

93-TT-220 coal 4 64.0 4.43 104.9 1.05 0.04 450 1.6 163.8 100 0.07

93-TT-301 coal 4 56.7 3.74 110.1 0.71 0.03 463 1.3 194.3 155 0.07

93-TT-302 coal 4 65.7 5.03 145.8 0.55 0.03 455 0.8 221.9 265 0.08

Devonian–

Carboniferous 92-SJ-36A shale 5 1.68 0.00 0.19 0.47 0.00 441 28 11 0.40 0.00

Cambrian 93-WH-3D carbonate 4 0.75 0.02 0.76 0.12 0.03 434 16 101 6.33 2.67

93-WH-4B carbonate 4 1.19 0.03 2.43 0.09 0.01 429 8 204 27.00 2.52

Riphean 93-MO-2F carbonate 4 0.65 0.01 0.34 0.07 0.03 438 11 52 4.86 1.54

93-MO-3G carbonate 4 0.58 0.01 0.72 0.13 0.01 435 22 124 5.54 1.72

*See Peters and Moldowan (1993) for discussion of analytical procedures. Locations are shown in Table 1. Laboratories are denoted as follows: 1 = Petrobras; 2 =Houston Advanced Research Center; 3 = Humble; 4 = ExxonMobil Upstream Research; 5 = Core Laboratories. TOC = weight percent organic carbon; S1, S2 =milligrams of hydrocarbons/gram of rock; S3 = milligrams of carbon dioxide/gram of rock; T max = temperature (jC). HI = S2 � 100/TOC; OI = S3 � 100/TOC; PI =S1/(S1 + S2); S1/TOC = S1 � 100/TOC.

Table 2. Continued

Table 3) and a modified Van Krevelen plot (type I highly

liquid-prone field, Figure 4). Groups 2 and 3 are also

distinguishable by maturity differences. Although Tmax

values range from 421 to 437 in both groups, JL, CM,

and DZ samples (group 2) have higher PI values than

group 3 NU and SO samples (0.02–0.06 versus 0.01–

0.02, respectively), indicating slightly higher maturity

(early to peak oil window) in group 2. The lower Meso-

zoic samples are also distinguished by high pristane/

phytane ratios in the group 2 samples (Figure 6) relative

to all other source rocks, indicating higher maturity and/

or more oxic environments (Peters and Moldowan, 1993).

826 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Figure 3. Logged coresection showing three ZBcore source rock samples.Note that samples ZB-1933 and ZB-1978 arefrom microlaminatedmicrite units interpretedas anoxic deposits in astratified lake, versussample ZB-1740, which isinterpreted as part of abetter oxygenated system(either more marginalfacies or a nonstratifiedlake; Johnson, 2002).Core sample numbersrepresent depth in feetbelow kelly bushing (KB)(+4 m), converted tometers-depth in the corelog.

Permian coal from central Mongolia (group 4) is

a type I (Figure 4), highly liquid-prone source (TOC

generally 50–70%). Maturity indicated by Tmax (441–

463) and PI values (0.02–0.06) is in the peak oil window

range. By comparison, pre-Permian carbonate and shale

samples from central and northern Mongolia (group 5)

have low TOC values (0.58–1.68), show a wider range

of kerogen types (Figure 4) (types I–IV of Tissot et al.,

1974; Demaison et al., 1983), and are also likely mature

(Tmax 429–441), although PI values are low.

This survey of bulk geochemical data indicates

several additional prospective source rock groups in

Mongolia, including several pre-Cretaceous units that

could source undiscovered hydrocarbon accumula-

tions. Lower Cretaceous organic-rich shale is widely

distributed in and around the East Gobi basin both in

outcrop and in the subsurface, whereas older (Paleozoic–

lower Mesozoic) source facies are mainly limited to

central and western Mongolia (Figure 1; Graham et

al., 2001). One exception is the Lower to Middle

Jurassic Khamarkhavoor Formation, which crops out

only rarely in the East Gobi basin and may be anal-

ogous to lower Mesozoic lacustrine oil shale from

Noyon Uul in southwestern Mongolia (Table 5, group

3) (Hendrix et al., 1997; Graham et al., 2001). The

Khamarkhavoor Formation was not sampled in this

study, but it may constitute a prerift source (Figure 2).

Lower Cretaceous source rocks of group 1 (Table 5)

are generally interpreted as the main source of oil in

the East Gobi basin (Yamamoto et al., 1993, 1998;

Traynor and Sladen, 1995; Sladen and Traynor, 2000).

Detailed biomarker analyses of select oil and source

rock samples completed in the second part of this study

confirm and further characterize this inferred correlation.

OIL GEOCHEMISTRY

Six Mongolian oil samples (including extracts from

the ZB tar sand) were analyzed for bulk and molecular

organic geochemistry. The Mongolian samples are from

the Tsagan Els (TE samples) and Zuunbayan (ZB sam-

ples) fields, which are located on separate structures

(about 25 km apart) in the Zuunbayan subbasin in

southeastern Mongolia (Figure 1, inset). Like most

nonmarine, Mesozoic-sourced oil in the region, oil

samples from the East Gobi basin tend to be waxy

(�20–30% paraffinic), having high pour points

(10–30jC; Table 6; Chen et al., 1994; Traynor and

Sladen, 1995; Dou et al., 1998; A. Hall, 2000, personal

communication). Two oil samples from lacustrine-

sourced Early Cretaceous basins in China (Erlian and

Johnson et al. 827

Figure 4. HI versus OI (modified VanKrevelen) plot indicating hydrocarbon-generative types of source rocks (Peters,1986). Type I = highly oil prone, type II =oil prone, type III = gas prone. See Table2 for data.

828 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Tab

le3

.Vi

sual

Ker

ogen

Sum

mar

yw

ith%

Org

anic

Con

stitu

ents

Sam

ple

Age

Am

orph

ous

Type

3

(Fin

ely

Dis

sem

inat

ed)

Trile

teSp

ores

and/

orPo

llen

Plan

tTi

ssue

-

Mem

bran

ous

Deb

ris

Fung

al

Deb

ris

Vitr

inite

(Ang

ular

-Str

uctu

red)

Iner

tinite

(Inc

lude

sPy

rite

)A

vera

ge%

Ro

92-U

B-4

Low

erC

reta

ceou

s8

6231

0.39

92-U

B-10

Low

erC

reta

ceou

s8

867

170.

36

92-U

B-12

Low

erC

reta

ceou

s8

6231

0.35

92-U

B-15

Low

erC

reta

ceou

s33

633

617

60.

37

92-J

L-6

Low

erto

Mid

dle

Jura

ssic

1080

100.

41

92-J

L-9

Low

erto

Mid

dle

Jura

ssic

973

180.

43

92-C

M-3

Low

erto

Mid

dle

Jura

ssic

77

5729

0.77

92-C

M-1

0Lo

wer

toM

iddl

eJu

rass

ic7

757

290.

76

92-C

M-1

2Lo

wer

toM

iddl

eJu

rass

ic8

867

170.

62

92-C

M-1

7Lo

wer

toM

iddl

eJu

rass

ic6

650

380.

74

92-C

M-2

0Lo

wer

toM

iddl

eJu

rass

ic7

5340

0.72

92-C

M-2

2Lo

wer

toM

iddl

eJu

rass

ic42

1121

215

0.79

92-D

Z-2

Low

erto

Mid

dle

Jura

ssic

633

4712

0.93

92-D

Z-6

Low

erto

Mid

dle

Jura

ssic

862

310.

65

92-N

U-4

8bTr

iass

ic–

Jura

ssic

8911

0.86

92-N

U-4

9Tr

iass

ic–

Jura

ssic

8010

100.

65

92-N

U-5

0Tr

iass

ic–

Jura

ssic

973

99

0.93

92-N

U-5

1Tr

iass

ic–

Jura

ssic

678

817

0.69

92-N

U-5

2Tr

iass

ic–

Jura

ssic

8010

100.

78

92-S

J-36

aD

evon

ian

–C

arbo

nife

rous

5714

295.

11

Daqing/Songliao) were included in the GCMS analyses

for comparison (Figure 1). The Erlian basin sample is

from the Saihantala sag, and the Daqing sample is from

the Yushuling oil field (Table 1). Both oils are believed

to have originated from Lower Cretaceous lacustrine

units analogous to those in the East Gobi basin (Yang

et al., 1985; Dou et al., 1998).

Petroleum Source Facies Indicators

All oil samples show similar GC patterns, with the

presence of high-molecular-weight n-alkanes indicating

algal or higher plant input (Peters and Moldowan, 1993),

although the n-alkanes drop off in abundance after nC23

(Figures 7, 8). Whole oil d13C (versus PDB [Peedee

Johnson et al. 829

Table 4. Ratios from Pyrolysis-GC Analyses for Selected Source Rock and Oil Samples*

Sample CPI (1) CPI (2) Pr/Ph Pr/nC17 Ph/nC18

nC17

Anomaly

Presence of

b-carotane?

Lower Cretaceous core samples ZB-1978 1.75 2.07 0.96 0.58 0.71 1.11 y?

ZB-1740 2.07 2.23 0.54 0.59 1.36 1.01 n

ZB-1933 1.18 1.22 0.87 0.34 0.40 1.03 y?

Other Lower Cretaceous samples 97-SH-4 1.18 1.70 0.89 1.31 0.62 0.69 y?

97-SH-9 1.38 2.08 0.25 0.83 8.94 2.42 y

97-SH-13 1.95 2.06 0.16 1.30 11.50 1.42 y?

97-TH-316 3.11 4.11 0.29 0.40 0.92 1.18 y

92-UB-4 2.82 2.64 1.96 1.88 0.76 0.95 y

92-UB-10 3.73 3.56 1.25 1.36 0.95 1.08 n

92-UB-12 4.93 4.55 1.99 3.89 1.19 0.84 n

92-UB-15 4.89 4.11 0.51 0.49 1.23 1.33 n

Jurassic samples 92-JL-9 1.38 1.70 3.00 1.03 0.22 0.96 n

92-CM-3 1.41 1.40 10.29 7.22 0.61 1.02 n

92-CM-10 1.40 1.29 9.97 3.63 0.31 1.01 n

92-CM-12 1.44 1.30 9.38 4.26 0.36 0.98 n

92-CM-17 1.43 1.23 8.88 2.57 0.25 1.00 n

92-CM-20 1.36 1.09 7.86 1.58 0.18 1.03 n

92-CM-22 1.30 1.45 4.25 1.27 0.30 1.07 n

92-DZ-2 1.07 1.17 4.65 2.26 0.39 0.93 n

92-DZ-6 1.30 0.75 7.43 0.67 0.07 1.01 n

Triassic–Lower Jurassic samples 92-NU-48B 1.29 1.65 1.36 0.54 0.35 0.99 y

92-NU-49 1.59 2.56 0.74 0.51 0.59 0.93 y

92-NU-50 1.46 2.83 0.97 0.40 0.43 1.06 y

92-NU-51 1.24 1.75 0.85 1.04 0.97 1.01 y

92-NU-52 1.10 1.01 0.80 0.47 0.55 1.01 y

Oil samples ZB-310b 1.11 1.15 1.10 0.19 0.16 1.03 y

ZB oil 163251A 1.20 1.25 0.86 0.43 0.48 1.04 y

ZB tar 163251B 1.13 1.18 1.03 0.46 0.44 1.03 y

TE-25 1.11 1.07 1.17 0.16 0.13 0.98 y

TE-A1 1.11 1.15 1.10 0.19 0.16 1.03 y

TE-A2 1.09 1.05 1.11 0.17 0.16 1.09 y

ER-1846 1.18 1.20 0.79 0.32 0.43 1.05 n

DQ-2198 1.10 1.09 1.15 0.27 0.25 1.04 n

*Ratio Calculation SourceCPI(1) [(nC(25 + 27 + 29 + 31 + 33))/(nC(26 + 28 + 30 + 32 + 34)) +

(nC(25 + 27 + 29 + 31 + 33))/(nC(24 + 26 + 28 + 30 + 32))]/2 Bray and Evans (1961)CPI (2) [2nC29/(nC28 + nC30)] Bray and Evans (1961)nC17 anomaly [2nC17/(nC16 + nC18)] Yamamoto et al. (1998)b-Carotane identification based on GC data only (see Appendix for GCMS results)

belemnite]) isotope measurements of three oil samples

from the two Mongolian fields yield similar ratios near

�31x(Table 7), suggesting that a similar source rock

for each of the Mongolian oil groups. Recycling of 13C-

depleted carbon by organisms living in stratified lakes

may cause these unusually negative values (Scholl et al.,

1994). Distinctions between the oil groups are indicated

by differences in pristane/phytane ratios,which are slightly

greater than 1 (1.1–1.7) for the Tsagan Els and Daqing oil

samples and slightly lower (0.8–1.1) for the Zuunbayan

and Erlian oil samples (Table 4). Oil from the Zuun-

bayan field also tends to be less waxy, having a lower

pour point than those at Tsagan Els (Table 6), suggesting

some variation in facies and/or maturity of source rocks.

Conventional biomarker ratios (Appendix) are

consistent with GC data, which suggest similar lacus-

trine source rock facies for the Mongolian oil accu-

mulations. Prokaryote-derived hopanes are the most

abundant biomarker group and are detected as major

peaks even on sterane transitions (Figure 9), whereas

830 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Figure 5. Example gaschromatograms for ex-tracted whole-rock sam-ples ZB-1978 (a: Zuunba-yan core) and 92-NU-48b(b: Noyon Uul) showingtypical fingerprints forMesozoic source rocksanalyzed in this study.

Johnson et al. 831

Tab

le5

.C

hara

cter

istic

sof

Sour

ceRo

ckG

roup

sof

Mon

golia

Gro

upA

geLo

catio

nLi

thol

ogy

Cha

ract

eris

tics

1Lo

wer

Cre

tace

ous

East

Gob

iba

sin,

Ula

anBa

taar

basi

n,

Nilg

aba

sin

(ZB,

UB,

SH,

TH,

NL

sam

ples

)

Coa

lan

dla

cust

rine

mud

ston

eTO

Cva

lues

1.5

–30

%fo

rm

udst

one

and

oil

shal

e;>

45%

for

coal

-ric

hsa

mpl

es

Vari

able

kero

gen

type

:m

ainl

yty

peI–

IIke

roge

n

(oil

pron

eto

high

lyoi

lpr

one)

;U

Bsa

mpl

esar

e

near

type

IIIan

dsh

owpr

edom

inan

ce

ofvi

trin

itean

din

ertin

ite(g

aspr

one)

CPI

(1)

vari

able

(1–

5);

high

est

inTH

/UB

sam

ples

Pr/P

hra

tios

0.25

–1.

99

Imm

atur

eto

earl

yoi

lw

indo

w

2Lo

wer

toM

iddl

eJu

rass

icW

este

rnan

dce

ntra

lM

ongo

liaC

oal

and

coal

ym

udst

one

Type

III–

IVke

roge

n(g

aspr

one

toin

ert)

(JL,

CM

,D

Zsa

mpl

es)

Pred

omin

ance

ofvi

trin

itean

din

ertin

ite

CPI

(1)

valu

es1.

0–

1.44

Hig

hPr

/Ph

ratio

s(3

–10

)

Earl

yoi

lw

indo

wto

peak

oil

win

dow

3Tr

iass

ic–

Jura

ssic

Cen

tral

and

sout

hwes

tern

Mon

golia

Lacu

stri

nem

udst

one

Type

Ike

roge

n(o

ilpr

one)

(NU

,SO

sam

ples

)Pr

edom

inan

ceof

liptin

ite(a

lgal

,w

axy

plan

tde

bris

)

CPI

(1)

valu

es1.

3–

1.6

Pr/P

hra

tios�

1(0

.75

–1.

36)

Earl

yoi

lw

indo

w

4Pe

rmia

nC

entr

alM

ongo

lia(T

Tsa

mpl

es)

Coa

lTy

peI

kero

gen

(oil

pron

e)

GC

data

n/a

Peak

oil

win

dow

5C

arbo

nife

rous

;

Riph

ean

–C

ambr

ian

Nor

ther

nan

dw

este

rnM

ongo

lia

(SG

,M

O,

WH

sam

ples

)

Car

bona

teK

erog

enty

peva

riab

le

(I–

IV;

oil

toga

spr

one

plus

iner

t)

GC

data

n/a

Mat

ure

toov

erm

atur

e(g

asw

indo

wto

nong

ener

ativ

e)

C30 24-n-propylcholestanes are absent, consistent with

a nonmarine source (Moldowan et al., 1990). Indicators of

salinity and/or water column stratification in the source

facies include the presence of b-carotane, g-carotane, and

gammacerane (Figure 9; Appendix) (Fu et al., 1990),

which are notably higher in the Zuunbayan oil than the

Tsagan Els oil.

Several biomarker parameters suggest significant

algal input to the source rock depositional environ-

ment. Ternary plots of C27, C28, and C29 regular ste-

ranes and monoaromatic steroids (Figure 10) suggest a

similar lacustrine source with algal input in Tsagan Els,

Zuunbayan, and Erlian samples (the Daqing oil plots

slightly out of this field with even higher C27 sterane

and steroids; Moldowan et al., 1985; Peters and Mol-

dowan, 1993) (Figure 10). The C26 steranes (24-

norcholestanes) and related C26 diasteranes (24-

nordiacholestanes) are likely derived from diatom

precursors (Moldowan et al., 1991; Holba et al., 1998a).

Their ratios to the nontaxa-specific C26 steranes, 27-

norcholestanes and 27-nordiacholestanes (NCR and

NDR, respectively; Holba et al., 1998a; Appendix),

are in the low range of values expected for Lower Cre-

taceous, nonmarine-sourced oil and contrast with

the higher values common in Tertiary nonmarine oil

(e.g., Holba et al., 1998b, Ritts et al., 1999). Although

C30 4-methyl steranes occur in both freshwater lacus-

trine and marine sediments, the presence of elevated

832 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Table 6. Characteristics of Oil Samples from Cretaceous-Sourced Lacustrine Basins in China and Mongolia*

Sample Locations jAPI Density (g/cm3) Viscosity at Surface (cp) Paraffin Analysis Pour Point (jC)

Tsagan Els field, East Gobi basin �30 0.872–0.878 4.7–38 average = 35.1% 36–38

Zuunbayan field, East Gobi basin �30 0.848–0.917 4.8–130 average = 20.5% 12.3–21

Jirgalangtu depression, Erlian basin n/a 0.8468–0.942 n/a 13.25% 28

Saertu pool, Daqing, Songliao basin n/a n/a 12.0–99 21–29% 22–34

*Tsagan Els and Zuunbayan (TE and ZB) data are from several samples reported in Shirokov and Kopytchenko (1983), in addition to previously unpublished industrydata (A. Hall, 2001, personal communication). Erlian basin oil data are from two oil samples described by Dou et al. (1998) in the Jirgalangtu depression. Daqingdata is from Saertu pool samples reported by Yang (1985).

Figure 6. Crossplotof pristine/phytane ratioversus carbon preferenceindex (CPI-1, see Table 4for calculation of this pa-rameter). Ellipses showgeneral source rockgroupings discussed inthe text. K1 = Lower Cre-taceous; Tr–J = Triassic–Jurassic. See Table 1 forsample location and in-formation.

4a-methyl-24-ethylcholestane suggests a strong affilia-

tion with lacustrine sediments (Murray et al., 1994). We

also note the high relative abundance of C30 dinosteranes

(4a,23,24-trimethylcholestanes) and triaromatic dino-

steroids, which is additional evidence for dinoflagellate-

algae contribution in the source facies (Figure 11a, b)

(Moldowan et al., 1996; 2001).

Zuunbayan and Tsagan Els oil samples also contain

C30 tetracyclic polyprenoids (TPP), which were recog-

nized by Holba et al. (2000) as indicators of algal-rich,

Johnson et al. 833

Figure 7. Gas chroma-tographs for whole-oilsamples ZB-310b (a) andTE-A1 (b), showing simi-lar fingerprints with high-er pristine/phytane valuesrelative to nC17 for theZuunbayan (ZB) oil.

fresh- to brackish-water lacustrine environments. The

TPP ratio (Holba et al., 2000), allied with other bio-

marker parameters, differentiates oil and source rocks

from lacustrine, marine, and mixed depositional envi-

ronments. Green algae commonly found in freshwater

environments are presumed to be the dominant source

of TPP (Holba et al., 2000). The Zuunbayan and Tsa-

gan Els oil samples plot in the Lacustrine I field, char-

acteristic of Cretaceous or younger, algal-lacustrine-

sourced oil (Figure 11c).

Although both groups of Mongolian oil (TE and ZB)

overlap on regular and aromatic steroid ternary plots,

certain parameters suggest derivation from slightly dif-

ferent source facies. The main difference in the two oil

groups appears to be varying amounts of algal-derived

biomarker indicators. In addition, the Zuunbayan sam-

ples have lower pristane/phytane ratios and lower

n-alkane concentrations than the Tsagan Els samples.

The Tsagan Els oil also contains lower but detectable

amounts of b-carotane, gammacerane, triaromatic dino-

steroids, and C30 dinosteranes (Appendix; Figure 11).

Tricyclic and tetracyclic terpane ratios, commonly used as

correlation parameters (Peters and Moldowan, 1993),

also indicate distinct oil groups between the two fields

(Figure 12). Thus, the lacustrine-sourced Mongolian

oil samples show subtle source-facies variations, with

greater algal and dinoflagellate input into the Zuunba-

yan oil source. These variations could reflect different

lacustrine source intervals with varying degrees of algal

input or lateral facies changes from offshore to mar-

ginal lacustrine environments.

Hexacyclic and Heptacyclic Polyprenoids

All of the Mongolian oil samples contain elevated con-

centrations of unusual hexacyclic and heptacyclic al-

kanes. These compounds are most abundant in the TE

834 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Figure 8. Plot comparing normalized n-alkane distribution for Mongolian and Chinese oil samples showing similar patterns, withthe presence of high-molecular weight n-alkanes indicating algal or higher plant input (Peters and Moldowan, 1993).

Table 7. Stable 13C Isotope Values for Select Oil and Source

Rock Samples

Sample Number d13C versus PDB (x)

Oil SamplesTEA-1 oil �30.70

TE-25 �30.56

SWZB-310b oil �31.21

Source Rock SamplesZB-1933 �30.96

ZB-1740 �31.19

ZB-1978 �30.61

97-SH-4 �26.25

98-TH-316 �29.69

oil, where they form some of the largest peaks on the m/z217 and TIC (total ion count) chromatograms in saturate

fractions with n-alkanes removed (branched and cyclic

fractions, Figure 9). Mass spectra from full-scan GCMS

confirms that these are six- and seven-ring (as much as

40-carbon) polyprenoids (Figure 13). These compounds

have been observed in only a few previous studies, in-

cluding the ostracod zone of the Western Canada sedi-

mentary basin (Li et al., 1996) and the Aquitaine basin of

France and northern Spain (Grosjean et al., 2000).

Cyclized polyaromatic polyprenoids presumed to be

related to these saturated compounds are also found in

aromatic fractions of the Eocene Messel shale in Ger-

many (Poinsot et al., 1995) and in western Canada (Li

et al., 1997). Although little is known about the global

occurrence of polyprenoids (Grosjean et al., 2000, 2001),

their abundance in these basins and association with

abundant four-ringed polyprenoids (TPP) appear to in-

dicate fresh- to brackish-water lacustrine source rock

environments. Interestingly, source facies in the East

Gobi, Western Canada, and Aquitaine basins are mainly

Early Cretaceous or younger, suggesting the possibility of

an age-sensitive biomarker. Li et al. (1996) reported at

least three isotopically distinct groups based on d13C

ratios of tri-, tetra-, and pentacyclic alkanes in extracts

from western Canada, suggesting derivation from dif-

ferent groups of organisms or different biosynthetic path-

ways in a single group.

Grosjean et al. (2000, 2001) offered preliminary

structural analysis of some of these isolated com-

pounds, which also occur in the East Gobi oil (Figure

13). Further isolation, structural characterization, and

confirmation of these six- to seven-ring polyprenoids

will help to elucidate their usefulness as source rock

environmental indicators (work is in progress in col-

laboration with P. Albrecht et al., L’Universite Louis

Pasteur, Strasbourg). Although their biological precur-

sor is unknown, these compounds represent a highly

specific correlation tool that may represent a new class

of biolipids (Grosjean et al, 2000, 2001). The six- to

seven-ring polyprenoid compounds have very late elu-

tion times (Figures 9, 13), which may be undetected

by some GCMS and MRM-GCMS time-temperature

protocols not optimized for their detection. Our pre-

liminary analyses of two oil samples from the Erlian

and Songliao basins did not reveal these compounds,

but further organic geochemical studies of other Early

Cretaceous lacustrine basins in the China–Mongolia

border zone specifically targeted toward these poly-

cyclic polyprenoids may reveal additional information

about their distribution, origin, and chemical structure.

Maturity Indicators

Biomarker parameters also suggest differences in matu-

rity between the Zuunbayan and Tsagan Els oil samples.

Several sterane and terpane isomerization parameters

(including C29 sterane aaa20S/20R, C31 homohopane

(S/R), Ts/Tm, and moretane indices) consistently in-

dicate that TE oil is more mature than the ZB oil (Fig-

ure 14). Likewise, the Erlian and Daqing samples ap-

pear to originate from more mature source rocks than

the Zuunbayan oil, with Erlian showing similar group-

ings to the Tsagan Els samples and Daqing generally

plotting on its own. These differences suggest at least

two oil-generating source rock maturities, reflecting

either different source rock intervals or two periods of

oil generation. The Zuunbayan field lies adjacent to a

major strike-slip fault active since at least the Late

Cretaceous, whereas the Tsagan Els field occupies a

midbasin position several kilometers away from this

fault. Thus, higher maturity in the Tsagan Els oil could

be the result of deeper burial away from this structure.

Alternatively, as discussed previously, the differences

in both maturity and facies parameters between Zuun-

bayan and Tsagan Els oil groups suggest that they could

originate from different lacustrine intervals with sepa-

rate thermal histories.

OIL – SOURCE ROCK CORRELATION

Biomarker analyses were completed to facilitate oil–

source rock correlation of selected source rocks from

the East Gobi region (SH, TH, and ZB samples; Figure

1). Lower Cretaceous samples were selected as the main

focus of this study based on their widespread distribu-

tion in the East Gobi basin, bulk geochemical param-

eters indicating a possible oil-source correlation (Figure

4; Table 5), and previous work (Yang et al., 1985; Ya-

mamoto et al., 1993; 1998; Traynor and Sladen, 1995;

Dou et al., 1998; Sladen and Traynor, 2000) suggesting

that synrift lacustrine facies are the main source units

in late Mesozoic basins of China and Mongolia.

SH and TH samples are too immature (e.g., Figure

14b) for meaningful correlation tests using most bio-

marker parameters. However, SH samples are unlikely

to be related to the Mongolian oils because of their

significantly less negative d13C isotope values (Table 7).

Pyrolysis experiments were attempted to model be-

havior of these rocks at increased maturity, but these

failed to produce a convincing correlation. Based on

d13C isotope ratios, samples from the Zuunbayan core

Johnson et al. 835

836 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

(Table 7) provided a much more convincing link with

the Zuunbayan and Tsagan Els oil groups.

Two of the three Zuunbayan core samples (ZB-1740,

1978) are relatively immature (Figure 14), whereas sam-

ple ZB-1933 consistently plots close to the Zuunbayan oil

using a variety of independent sterane and terpane isom-

erization ratios (Figure 14; Appendix, ratios B, C, D, G,

O, and P). Because sample ZB-1933 is bracketed by the

other two samples in depth (Table 2), it is not expected

to have an anomalously high maturity level. Sample

Johnson et al. 837

Figure 10. Ternary plots of regularsteranes and monoaromatic steroids forMongolian oil and selected source rocks.(a) Percentages of C27-C28-C29 aaa20Rsteranes measured by GCMS analysis ofm/z 217 peak heights of the saturatefraction. (b) Percentage of monoaromaticsteroids calculated by measuring peakheights of all six isomeric compounds ofC27, C28, and C29 MA steroids on GCMSm/z 253 analyses of the aromatic fraction(as described in Peters and Moldowan,1993). Note tight grouping of oil sampleson both diagrams, with more scatteramong the Chinese oils and Mongoliansource rocks. See Appendix for data.

Figure 9. Example GCMS mass chromatograms for oil samples ZB-310b and TE-A2, showing m/z 217, 191, and TIC. All graphs shown atsame time interval (60–100 min). Note predominance of late-eluting peaks (polyprenoids), particularly in the TE (Tsagan Els) sample. Dots onm/z 217 graphs denote regular sterane isomers identified as peak numbers 6–17 (from left to right) by Peters and Moldowan (1993, p. 85).

838 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Figure 11. (a, b) Dinosterane anddinosteroid index crossplots show-ing distinct groups of TE versus ZBoil samples (see Appendix for ratiocalculations). Note that ZB-1740core sample plots closer to theTE oils, implying less algal inputcompared to the ZB oils and othercore samples. (c) Crossplot of tet-racyclic polyprenoid ratio (TPP ofHolba et al., 2000) versus biomar-ker indicators for nonmarine dino-flagellate algae (20R-4amethyl/20R-24ethyl index), showing fields ofHolba et al. (2000). The Mongolianoil and core source rock sampleshave similarly high TPP ratios, andplot in the Lacustrine I field charac-teristic of oils derived from Creta-ceous or younger, algal-rich sourcerocks (Holba et al., 2000). SeeAppendix for ratio calculations.

ZB-1933 also has a lower Tmax value (indicating more

labile carbon), twice %TOC, and much larger S1, S2, and

production index values compared to the other two ZB

core samples (Table 2). We interpret this as evidence

for minor contamination by oil migration in this sample.

Two samples from the laminated mudstone facies

of the Zuunbayan core (ZB-1978 and ZB-1933) contain

higher biomarker ratios reflecting dinosterol derivatives

and indicate more dinoflagellate input relative to sam-

ple ZB-1740 (Figure 11a, b). Tsagan Els oil samples plot

closer to the ZB-1740 core sample as well, suggesting a

possible correlation between the TE oil and facies with

less algal input. Based on tetracyclic polyprenoid ratios

versus an algal-terrigenous sterane ratio, both oil groups

and the ZB core samples plot together in the ‘‘Lacus-

trine I’’ field, which is characteristic of algal-rich, Cre-

taceous or younger samples (Figure 11c, Holba et al.,

2000). Perhaps the best evidence for rock-oil correla-

tion between the Zuunbayan lacustrine core facies and

the Tsagan Els/Zuunbayan oil groups is the presence

of unusual six- and seven-ring polycyclic polyprenoid

compounds in both core and oil samples. The cyclized

alkanes were not found in any other source rock samples

and likely represent a highly specific correlation tool.

CONCLUSIONS

Outcrop samples from Mongolia reveal numerous high-

quality potential source rocks ranging from Paleozoic to

Mesozoic in age. The extent to which pre-Cretaceous

rocks may source unrecognized petroleum systems in

central and western Mongolia is poorly understood.

Our data indicate that currently exploited hydrocarbon

accumulations in the East Gobi basin most likely orig-

inated from Lower Cretaceous lacustrine facies, as re-

ported for neighboring basins in China (Yang, 1985;

Dou, 1997). Stratigraphic evidence indicates long-

lived lacustrine systems throughout the rifting phase

(more than 30 m.y.) in the China–Mongolia border

region (Lin et al., 2001; Sladen and Traynor, 2000).

Based on core samples from the Zuunbayan field, these

lakes contained fresh- to brackish-water bodies that re-

peatedly expanded and contracted during rifting (Gra-

ham et al., 2001). Lake margin shifts may also have

been linked to periods of thermal stratification and an-

oxic lake-bottom conditions alternating with periods of

better oxygenation and mixing of the water column,

which could account for source facies variation indi-

cated by geochemical parameters discussed in this study.

Bulk geochemistry, isotope, and biomarker data for

oil from the East Gobi basin support the link to Lower

Cretaceous lacustrine source facies. The oil samples show

a predominance of hopanes over steranes. In addition,

terpane and sterane parameters (including a lack of C30

4-desmethyl steranes and lower abundances of C28

[relative to C27 and C29] regular steranes and mono-

aromatic steroids) suggest a link to a terrigenous source.

Hypersaline indicators, suchashighgammacerane/hopane

and C34 or C35 homohopane/hopane ratios, are not pres-

ent. Dinoflagellate-derived dinosteroid and aromatic dino-

steroid parameters also point to significant algal influence,

Johnson et al. 839

Figure 12. Crossplotof tricyclic and tetracyclicterpane parametersshowing distinct groupsof TE versus ZB oil, withZB core samples plottingin between. Note outlyingSH and TH samples sug-gesting poor correlationwith TE and ZB samples.

as well as bacteria-derived hopanes and minor higher

plant input. In addition to standard biomarker param-

eters, oil and source rocks from the East Gobi basin

contain unusually high concentrations of six- and seven-

ring cyclized alkanes. These polyprenoid compounds oc-

cur in only a few nonmarine basins worldwide, where

they are thought to indicate fresh- to brackish-water la-

custrine deposition (Li et al., 1996; Grosjean et al., 2000).

840 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Figure 13. (a) Mass chromatogram from GCMS-MSscan run of polyprenoid compounds of oil sample TE-A1.Regular polyprenoids (no demethylation on ring system)are signified by two numbers: the number of carbonatoms/number of rings (e.g., 34/6 = 34 carbons, 6 rings).Structures missing methyl groups on their ring systemare additionally signified by a ‘‘�.’’ (b, c) Mass spectra forpeaks 1 and 2 (labeled in [a], respectively), showingmolecular ions of each peak. Structures were identifiedby Grosjean et al. (2000), for the same compounds in anoil sample from Spain.

Although the oil samples we analyzed appear to

originate from a similar nonmarine source rock type, they

form two distinct groups based on both maturity and

source parameters. Oil from the Zuunbayan field has

higher concentrations of b-carotane and dinoflagellate-

derived dinosteranes, indicating more algal input to

its source facies compared to that from Tsagan Els oil.

Tsagan Els oil correlates well with ZB-1740, a core

sample thought to have formed in more oxygenated

water conditions than the laminated micrite (ZB-1933

and ZB-1978), which correlate well with the Zuunba-

yan oil. Thus, subtle facies differences resulting from

different water chemistry resulted in distinct oil groups

as defined by biomarker parameters. These facies dis-

tinctions may reflect separate source units, or lateral

facies variations in a single source unit.

Tsagan Els oil was generated at higher maturity than

the Zuunbayan oil, as indicated by sterane and hopane

isomerization parameters. Burial history in the East Gobi

subbasins is obscured by activity along the Zuunbayan

fault, a major basin-partitioning strike-slip fault which

is nonetheless poorly constrained in terms of timing and

magnitude of offset (Graham et al., 2001; Johnson, 2002).

Independent thermal histories between the Zuunba-

yan and Tsagan Els fields may result from differential

burial of a given source unit along this fault (including

lateral facies transitions), or could reflect entirely unique

source units having similar geochemical signatures.

Johnson et al. 841

Figure 14. Biomarker param-eter crossplots highlightingmaturity differences and oilgroupings. See Appendix forratio calculations. (a) Crossplotof C29 isomerization ratiosshowing increasing maturityfrom ZB core samples, to ZB oil,to TE oil and the Chinese oil.(b) Crossplot of terpane matu-rity parameters indicating dis-tinct oil and source rock groups.Note that one of the ZB coresamples (ZB-1933), plottingclosest to the ZB oil, is unusuallymature compared to the othercore samples and is likely con-taminated by migrated oil.

842 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

TERPANES Sample Number A B C D E F G H I J K

Oil Samples TE-A-1 0.02 0.63 0.05 0.05 0.18 0.34 0.47 0.16 0.14 Y YTE-A-2 0.02 0.64 0.05 0.05 0.24 0.36 0.45 0.12 0.13 Y YTE-25 0.02 0.63 0.05 0.04 0.20 0.34 0.53 0.19 0.09 Y Y

ZB-310b 0.04 0.57 0.11 0.13 0.08 0.63 0.33 0.11 0.23 Y Y93-ZB-101 0.09 0.62 0.10 0.11 0.10 0.75 0.29 0.05 0.36 n/a Y93-ZB-103 0.08 0.64 0.10 0.12 0.02 0.76 0.30 0.07 0.29 n/a Y93-ZB-204 0.08 0.66 0.10 0.11 0.05 0.78 0.32 0.03 0.46 n/a Y

ER-1848 0.01 0.59 0.05 0.05 0.17 0.40 0.59 0.02 0.34 Y NDQ-2198 0.19 0.33 0.06 0.04 0.14 0.54 0.79 0.36 0.21 Y N

Source Rock Samples 97-SH-4 0.07 0.11 0.14 0.65 0.06 0.81 0.07 0.06 0.06 Y N97-SH-9 0.08 0.09 0.72 0.72 0.15 0.87 0.04 0.24 0.03 Y N97-SH-13 n/a n/a n/a n/a 0.26 0.86 n/a 1.00 0.06 N N97-TH-316 0.14 0.09 0.64 0.70 0.27 0.37 0.04 0.05 0.61 N N

ZB-1740 0.00 0.33 0.11 0.32 0.28 0.53 0.02 0.01 0.36 N NZB-1933 0.04 0.51 0.10 0.14 0.14 0.48 0.32 0.21 0.17 Y NZB-1978 0.00 0.29 0.15 0.28 0.27 0.54 0.15 0.02 0.28 Y N

STERANES Sample Number L M N O P Q R S T U V

Oil Samples TE-A-1 28.44 22.00 49.56 0.46 0.44 0.57 0.21 0.12 0.19 0.85 0.85TE-A-2 27.53 22.35 50.12 0.45 0.45 0.57 0.21 n/a n/a n/a n/aTE-25 30.09 26.23 43.67 0.50 0.51 0.54 0.16 n/a n/a n/a n/a

ZB-310b 34.50 24.53 40.97 0.27 0.21 0.84 0.54 0.15 0.17 0.75 0.8093-ZB-101 27.08 21.68 51.24 0.31 0.23 n/a n/a n/a n/a n/a n/a93-ZB-103 28.92 22.36 48.72 0.33 0.24 n/a n/a n/a n/a n/a n/a93-ZB-204 27.85 22.20 49.95 0.35 0.26 n/a n/a n/a n/a n/a n/a

ER-1848 34.25 14.86 50.89 0.48 0.46 n/a n/a n/a n/a n/a n/aDQ-2198 52.18 15.67 32.15 0.57 0.64 n/a n/a n/a n/a n/a n/a

Source Rock Samples 97-SH-4 17.98 23.47 58.55 0.05 0.00 0.68 0.21 n/a n/a n/a n/a97-SH-9 38.45 25.38 36.17 0.04 0.00 0.50 0.08 n/a n/a n/a n/a97-SH-13 36.85 9.75 53.41 0.00 0.00 0.00 n/a n/a n/a n/a n/a97-TH-316 60.27 13.28 26.45 0.01 0.00 0.86 0.31 n/a n/a n/a n/a

ZB-1740 32.77 13.90 53.33 0.00 0.18 0.69 0.33 0.19 0.14 0.89 0.69ZB-1933 36.30 21.71 41.99 0.24 0.20 0.88 0.60 0.12 0.13 0.72 0.77ZB-1978 32.63 12.73 54.64 0.03 0.14 0.88 0.47 0.08 0.08 0.79 0.49

AROMATICS Sample Number W X Y Z AA AB AC AD

Oil Samples TE-A-1 0.88 0.79 0.48 0.64 0.42 31.67 24.56 43.77TE-A-2 0.87 0.75 0.45 0.64 0.31 31.31 25.33 43.36TE-25 0.85 0.75 0.43 0.59 0.32 31.55 26.62 41.84

ZB-310b 0.97 0.94 0.76 0.89 0.67 29.07 24.28 46.65

ER-1848 0.93 0.87 0.59 0.80 0.54 28.20 25.66 46.14DQ-2198 0.84 0.93 0.55 0.51 0.68 39.87 30.44 29.69

Source Rock Samples ZB-1740 0.87 0.83 0.33 0.63 0.38 27.68 30.85 41.47ZB-1933 0.96 0.94 0.53 0.91 0.73 26.48 23.86 49.67ZB-1978 0.97 0.98 0.77 0.86 0.90 29.99 29.25 40.76

APPENDIX: TABLE OF SELECT BIOMARKER RATIOS (SEE KEY FOR EXPLANATION)

Johnson et al. 843

Ratio Index Source of Data Calculation

TERPANESA Gammacerane index GCMS m/z 191 gammacerane

gammacerane þ C30 hopane

B C31 homohopaneisomerization index

GCMS m/z 191 C31 ab 22SC31 ab 22S þ C31 ab 22R homohopanes

C C29, C30 moretane index GCMS m/z 191 C29 þ C30 moretaneðC29 þ C30 moretaneÞ þ ðC29 þ C30 hopaneÞ

D C29, Moretane index GCMS m/z 191 C29 moretaneC29 moretane þ C29 hopane

E C19, C23 tricyclic index GCMS m/z 191 C19 tricyclicC19 þ C23 tricyclics

F C26, C25 tricyclic index GCMS m/z 191 C26 tricyclicC26 þ C25 tricyclics

G Ts/Tm index GCMS m/z 191 TsTs þ Tm

H C21 tricyclic versus C30

hopane indexGCMS m/z 191 C21 tricyclic

C21 tricyclic þ C30 hopane

I C24 tetracyclic versus C21

tricyclic indexGCMS m/z 191 C24 tetracyclic

C24 tetra þ C21 tricyclic

J Presence of b-carotane GCMS m/z 125 yes/noK Presence of oleanane GCMS m/z 191 yes/no

STERANESL % C27 steranes GCMS 217 100�C27 regular steranes

ðC27 þ C28 þ C29Þ regular steranes

M % C28 steranes GCMS 217 100�C28 regular steranesðC27 þ C28 þ C29Þ regular steranes

N % C29 steranes GCMS 217 100�C29 regular steranesðC27 þ C28 þ C29Þ regular steranes

O C29 20S/Rstereoisomerization ratio

GCMS 217 C29aaa 20S20S þ 20R

P C29 regular steranes abb/aaastereoisomerization ratio

GCMS 217 C29 abb ð20S þ 20RÞabb ð20S þ 20RÞ þ aaa ð20S þ 20RÞ

Q C30 dinosteranes vs 3b methyl(4methyl steranes)**

MRM GCMS 414 !231 transition SUMðD1 � D4 þ 4a-methylÞðSUMðD1 � 4 þ 4a-methyl þ 3b-methylÞdinosterane

R C30 dinosterane (4th dinosterane)vs 3b methyl

MRM GCMS 414 !231 transition D4 dinosteraneðD4 þ 3b-methylÞ

S Nordiacholestane ratio (NDR) MRM GCMS 358 !217 transitionpeak areasy

24 nordiacholestanes24 nor þ 27 nordiacholestanes

T Norcholestane ratio (NCR) MRM GCMS 358 !217 transitionpeak areasy

24 norcholestanes24 nor þ 27 norcholestanes

U C30 4a-methyl/C29 24ethyl-cholestane MRM GCMS 414 !217 transitionpeak areasyy

C29aaa 20R steraneC29aaa20R sterane þ 4a-methyl dinosterane

V Tetracyclic polyprenoid ratio MRM GCMS 414 !259 transition,peak areas+ sum 27-norcholestanesz

2 � C29 tetracyclic polyprenoid peak A2 � peak A þ C27 norcholestanes

AROMATICSW Triaromatic dinosteroid

vs 4a - methyl indexGCMS m/z 245 peak areas SUMðD1 � D6 dinosteroidsÞ

SUMðD1 � D6Þ þ 4SR

X Triaromatic dinosteroidvs 3b - methyl index

GCMS m/z 245 peak areas SUMðD1 � D6 dinosteroidsÞSUMðD1 � D6Þ þ 3SR

Y All triaromatic dinosteroids GCMS m/z 245 peak areas SUMðD1 � D6 dinosteroidsÞSUMðD1 � D6Þ þ all SR=CR

Z Sixth triaromatic dinosteroidsvs 4SR

GCMS m/z 245 peak areas D6 dinosteroids4SR þ D6

AA Fifth triaromatic dinosteroidsvs 4SR

GCMS m/z 245 peak areas D5 dinosteroids3SR þ D5

Key to Appendix*

REFERENCES CITED

Amory, J., 1996, Permian sedimentation and tectonics of southernMongolia: M.S. thesis, Stanford University, 183 p.

Badamgarav, D., Y. Khand, and R. Barsbold, 1995, NonmarineCretaceous of Mongolia: the Cretaceous system in east andsouth Asia: Newsletter Special Issue 2, IGCP 350: Fukuoka,Japan, Kyushu University, 23 p.

Bray, E. E., and E. D. Evans, 1961, Distribution of n-paraffins as aclue to recognition of source beds: Geochimica et Cosmochi-mica Acta, v. 22, p. 2–15.

Chen, Z., G. Zhou, and R. Alexander, 1994, A biomarker study ofimmature crude oils from the Shengli oil field, People’sRepublic of China: Chemical Geology, v. 113, p. 117–132.

Demaison, G., A. J. J. Holck, R. W. Jones, and G. T. Moore, 1983,Predictive source bed stratigraphy; a guide to regional petro-leum occurrence: Proceedings of the 11th World PetroleumCongress 2, PD1, New York, J. Wiley and Sons, p. 17–20.

Dou, L., 1997, The Lower Cretaceous petroleum system in NEChina: Journal of Petroleum Geology, v. 20, p. 475–488.

Dou, L., Y. Zhu, T. Yang, S. Xu, and X. Ping, 1998, Origins of heavyoils in the Erlian Basin, NE China: Marine and PetroleumGeology, v. 15, p. 769–781.

Farrell, K. M., 2001, Geomorphology, facies architecture, and high-resolution, non-marine sequence stratigraphy in avulsion depos-its, Cumberland marshes, Saskatchewan: Sedimentary Geology,v. 139, p. 93–150.

Fu, J., and G. Sheng, 1992, Source and biomarker compositioncharacteristics of Chinese nonmarine crude oils, in J. K. Whelanand J. W. Farrington, eds., Organic matter; productivity, accu-mulation, and preservation in recent and ancient sediments:New York, Columbia University Press, p. 417–432.

Fu, J., G. Sheng, J. Xu, G. Eglington, A. P. Gowar, R. Jia, S. Fan,and P. Peng, 1990, Application of biological markers in theassessment of paleoenivoronments of Chinese non-marinesediments: Organic Geochemistry, v. 16, p. 769–779.

Gao, R., X. Cai, H. Xu, and Y. Ren, 1997, Sequence stratigraphyand distribution of oil and gas, Songliao basin, China, in B.Liu and S. Li, eds., Basin analysis, global sedimentary geologyand sedimentology: Proceedings of the 30th InternationalGeological Congress, v. 8, p. 26–51.

Grabowski, G., and D. Pevear, 1985, Sedimentology and petrologyof profundal lacustrine sediments, Mahogany zone of theGreen River Formation, Piceance Creek basin, NorthwestColorado: SEPM Core Workshop, v. 6, p. 386–430.

Graham, S. A., M. S. Hendrix, C. L. Johnson, D. Badamgarav, G.Badarch, J. Amory, M. Porter, R. Barsbold, L. E. Webb, andB. R. Hacker, 2001, Sedimentary record and tectonic implica-tions of late Mesozoic rifting, southeast Mongolia: GeologicalSociety of America Bulletin, v. 113, p. 1560–1579.

Grosjean, E., J. Poinsot, A. Charrie-Duhaut, S. Tabuteau, P. Adam,J. M. Trendel, P. Schaeffer, J. Connan, D. Dessort, and P.Albrecht, 2000, A heptacyclic polyprenoid hydrocarbon insediments: a clue to unprecedented biological lipids: ChemicalCommunications, v. 11, p. 923–924.

Grosjean, E., J. Poinsot, A. Charrie-Duhaut, S. Tabuteau, P. Adam,J. M. Trendel, P. Schaeffer, J. Connan, D. Dessort, and P.Albrecht, 2001, Synthesis and NMR characterization of novelhighly cyclised polyprenoid hydrocarbons from sediments:Journal of the Chemical Society, Perkin Transactions, v. 1,p. 711–719.

Hendrix, M. S., T. J. Greene, A. R. Carroll, J. M. Moldowan, andS. A. Graham, 1997, Uppermost Triassic– lowermost Jurassiclacustrine shale as a possible source rock interval for southernMongolian sedimentary basins: AAPG Annual ConventionProgram Abstracts, v. 6, p. A49.

Hendrix, M., M. Beck, G. Badarch, and S. Graham, 2001, Triassicsynorogenic sedimentation in southern Mongolia: early effectsof intracontinental deformation, in M. Hendrix and G. Davis,eds., Paleozoic and Mesozoic tectonic evolution of centralAsia: from continental assembly to intracontinental deforma-tion: Geological Society of America Memoir 194, p. 199–214.

Holba, A. G., L. I. P. Dzou, W. D. Masterson, W. B. Hughes, B. J.Huizinga, M. S. Singletary, J. M. Moldowan, M. R. Mello, and E.Tegelaar, 1998a, Application of 24-norcholestanes for con-straining source age of petroleum, in B. Horsfield, M. Radke,R. G. Schaefer, and H. Wilkes, eds., Advances in organicgeochemistry 1997: Proceedings of the 18th InternationalMeeting on Organic Geochemistry; Part II: Biogeochemistry,v. 29, Oxford, New York, Pergamon, p. 1269–1283.

Holba, A. G., E. W. Tegelaar, B. J. Huizinga, J. M. Moldowan, M. S.Singletary, M. A. McCaffrey, and L. I. P. Dzou, 1998b, 24-norcholestanes as age-sensitive molecular fossils: Geology,v. 26, p. 783–786.

Holba, A. G., E. Tegelaar, L. Ellis, M. S. Singletary, and P. Albrecht,2000, Tetracyclic polyprenoids; indicators of freshwater(lacustrine) algal input: Geology, v. 28, p. 251–254.

Jerzykiewicz, T., and D. A. Russell, 1991, Late Mesozoic stratigraphyand vertebrates of the Gobi basin: Cretaceous Research, v. 12,p. 345–377.

Johnson, 2002, Sedimentary record of late Mesozoic extension,

844 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks

Ratio Index Source of data Calculation

AB %C27 monoaromatic steroidszz GCMS m/z 253(six isomers each)

100�C27 monoaromatic steroidsðC27 þ C28 þ C29 MA steroidsÞ

AC %C28 monoaromatic steroids GCMS m/z 253(six isomers each)

100�C28 monoaromatic steroidsðC27 þ C28 þ C29 MA steroidsÞ

AD %C29 monoaromatic steroids GCMS m/z 253(six isomers each)

100�C29 monoaromatic steroidsðC27 þ C28 þ C29 MA steroidsÞ

* Ratios from peak heights unless otherwise noted.** C30 4a-methyl sterane measurements include D3 (third dinosterane) peak.y see Holba et al., 1998a.yy see Holba et al., 2000.z Peters and Moldowan, 1993.zz Murray et al., 1994.

Key to Appendix. Continued

southeast Mongolia: implications for the petroleum potentialand tectonic evolution of the China–Mongolia border region:Ph.D. Thesis, Stanford University, 369 p.

Johnson, C. L., S. A. Graham, and M. S. Hendrix, 2000, Sedimentarycharacteristics and stratigraphic architecture of a Mesozoiclacustrine deltaic sequence, southern Mongolia: AAPG AnnualMeeting Program Abstracts, v. 9, p. A75.

Johnson, C. L., L. E. Webb, S. A. Graham, M. A. Hendrix, andG. Badarch, 2001, Sedimentary and structural records of lateMesozoic high-strain extension and strain partitioning, EastGobi basin, southern Mongolia, in M. Hendrix and G. A.Davis, eds., Paleozoic and Mesozoic tectonic evolution ofcentral Asia: from continental assembly to intracontinentaldeformation: Geological Society of America Memoir 194,p. 413–434.

Lamb, M. A., and G. Badarch, 1997, Paleozoic sedimentary basinsand volcanic-arc systems of southern Mongolia; new strati-graphic and sedimentologic constraints: International GeologyReview, v. 39, p. 542–576.

Li, M., C. L. Riediger, M. G. Fowler, L. R. Snowdon, and T. A.Abrajano Jr., 1996, Unusual polycyclic alkanes in LowerCretaceous ostracod sediments and related oils of the WesternCanada sedimentary basin: Organic Geochemistry, v. 25,no. 3–4, p. 199–209.

Li, M., C. L. Riediger, M. G. Fowler, and L. R. Snowdon, 1997,Unusual polycyclic aromatic hydrocarbons in Lower Creta-ceous ostracod zone sediments and related oils of the West-ern Canada sedimentary basin: Organic Geochemistry, v. 27,p. 439–448.

Lin, C., K. Eriksson, S. Li, Y. Wan, J. Ren, and Y. Zhang, 2001,Sequence architecture, depositional systems, and controls ondevelopment of lacustrine basin fills in part of the Erlian basin,Northeast China: AAPG Bulletin, v. 85, p. 2017–2043.

Liu, H., 1986, Geodynamic scenario and structural styles of Meso-zoic and Cenozoic basins in China: AAPG Bulletin, v. 70,p. 377–395.

Moldowan, J. M., W. K. Seifert, and E. J. Gallegos, 1985,Relationship between petroleum composition and depositionalenvironment of petroleum source rocks: AAPG Bulletin, v. 69,p. 1255–1268.

Moldowan, J. M., F. J. Fago, C. Y. Lee, S. R. Jacobson, D. S. Watt,N. E. Slougui, A. Jeganathan, and D. C. Young, 1990, Sedi-mentary 24-n-propylcholestanes, molecular fossils diagnosticof marine algae: Science, v. 247, p. 309–312.

Moldowan, J. M., C. Y. Lee, D. S. Watt, A. Jeganathan, N. E.Slougui, and E. J. Gallegos, 1991, Analysis and occurrence ofC26-steranes in petroleum and source rocks: Geochimica etCosmochimica Acta, v. 55, p. 1065–1081.

Moldowan, J. M., J. Dahl, S. R. Jacobson, B. J. Huizinga, F. J. Fago,R. Shetty, D. S. Watt, and K. E. Peters, 1996, Chemostrati-graphic reconstruction of biofacies; molecular evidence linkingcyst-forming dinoflagellates with pre-Triassic ancestors: Geol-ogy, v. 24, p. 159–162.

Moldowan, J. M., S. R. Jacobsen, J. Dahl, A. Al-Hajji, B. J.Huizinga, and F. J. Fago, 2001, Molecular fossils demonstratePrecambrian origin of dinoflagellates, in A. Zhuravlev and R.Riding, eds., Ecology of the Cambrian radiation: New York,Columbia University Press, p. 474–492.

Murray, A. P., R. E. Summons, C. J. Boreham, and L. M. Dowling,1994, Biomarker and n-alkane isotope profiles for Tertiary oils;relationship to source rock depositional setting: Organic Geo-chemistry, v. 22, p. 521–542.

Neves, R., D. McLean, R. H. Bate, and W. C. Penttila, 2000,Mongolia, Tamtsag basin, evidence for widespread, highquality, mature Lower Cretaceous source rock (abs): AAPGBulletin, v. 84, p. 1469.

Penttila, W. C., 1994, The recoverable oil and gas resources ofMongolia: Journal of Petroleum Geology, v. 17, p. 89–97.

Peters, K. E., 1986, Guidelines for evaluating petroleum sourcerock using programmed pyrolysis: AAPG Bulletin, v. 70,p. 318–329.

Peters, K. E., and J. M. Moldowan, 1993, The biomarker guide;interpreting molecular fossils in petroleum and ancientsediments: Englewood Cliffs, Prentice Hall, 363 p.

Poinsot, J., D. Dessort, P. Adam, G. Lacrampe, J. M. Trendel, andP. Albrect, 1995, New and rare biomarkers: saturated C33 –C43

polycycloisoprenoids, in J. O. Grimalt and C. Dorronsoro,eds., Organic geochemistry: developments and applications toenergy, climate, environment, and human history: Donostia,San Sebastian, AIGOA, p. 116–118.

Powell, T. G., 1986, Petroleum geochemistry and depositionalsetting of lacustrine source rocks: Marine and Petroleum Geol-ogy, v. 3, p. 200–219.

Ritts, B. D., A. D. Hanson, D. Zinniker, and J. M. Moldowan, 1999,Lower–Middle Jurassic nonmarine source rocks and petroleumsystems of the northern Qaidam basin, northwest China: AAPGBulletin, v. 83, p. 1980–2005.

Scholl, M., R. J. Hwang, R. M. K. Carlson, and J. E. Welton, 1994,Carbon isotopic composition of individual biomarkers in gilson-ites (Utah): Organic Geochemistry, v. 21, no. 6–7, p. 673–683.

Shirokov, V. Y., and V. N. Kopytchenko, 1983, Analysis andsummary of geological and geophysical materials on possibilityof oil and gas bearing basins of Mongolia, v. 1: Moscow, All-Union Export and Import, 291 p.

Sjostrom, D. J., M. Hendrix, D. Badamgarav, S. A. Graham, andB. K. Nelson, 2001, Sedimentology and provenance ofMesozoic nonmarine strata in western Mongolia: a record ofintracontinental deformation, in M. Hendrix and G. Davis,eds., Paleozoic and Mesozoic tectonic evolution of Asia: fromcontinental assembly to intracontinental deformation: Geolog-ical Society of America Memoir 194, p. 361–381.

Sladen, C., and J. J. Traynor, 2000, Lakes during the evolution ofMongolia: lake basins through space and time: AAPG Studiesin Geology, v. 46, p. 35–57.

Soco International, 2001, press release ‘‘Upgrade of Mongolia Project,’’http://www.socointernational.co.uk/press.php?id=17&d=2001Posted October 15, 2001, accessed November 1, 2002.

Steinshouer, D. W., J. Qiang, P. J. McCabe, and R. T. Ryder, 1999,Maps showing geology, oil and gas fields, and geologic prov-inces of the Asia Pacific region: United States GeologicalSurvey open file report OF-97-0470-F (1 CD-ROM).

Tissot, B., B. Durand, J. Espitalie, and A. Combaz, 1974, Influence ofnature and diagenesis of organic matter in formation of petro-leum: AAPG Bulletin, v. 58, p. 499–506.

Traynor, J. J., and C. Sladen, 1995, Tectonic and stratigraphicevolution of the Mongolian People’s Republic and its influenceon hydrocarbon geology and potential: Marine and PetroleumGeology, v. 12, p. 35–52.

United States Department of Energy, 2001, Country analysis brief forChina: Energy Information Association, http://www.eia.doe.gov/emeu/cabs/china.html, April Issue.

Wan, C., X. Qiao, R. Wang, and C. He, 1997, Cretaceousnonmarine microphytoplankton from the Hongqi Depressionin Hailar Basin, NE China: Acta Micropalaeontologica Sinica,v. 14, p. 405–418.

Wang, R., D. Qin, B. Wu, F. Li, C. Wan, and X. Qiao, 1996,Cretaceous dinoflagellates and other algae from Kailu basin,inner Mongolia: Acta Micropalaeontologica, v. 13, p. 303–312.

Watson, M. P., A. B. Hayward, D. N. Parkinson, and Z. M. Zhang,1987, Plate tectonic history, basin development and petroleumsource rock deposition onshore China: Marine and PetroleumGeology, v. 4, p. 205–225.

Johnson et al. 845

Wolfbauer, C. A., and R. C. Surdam, 1974, Origin of nonmarinedolomite in Eocene Lake Gosiute, Green River Basin, Wyoming:Geological Society of America Bulletin, v. 85, p. 1733–1740.

Yamamoto, M., D. Bat-Erdene, P. Ulziikhishig, M. Enomoto, Y.Kajiwara, N. Takeda, Y. Suzuki, Y. Watanabe, and T. Nakajima,1993, Preliminary report on geochemistry of Lower CretaceousDsunbayan oil shales, eastern Mongolia: Geological SurveyBulletin, Japan, v. 44, p. 685–691.

Yamamoto, M., D. Bat-Erdene, P. Ulziikhishig, Y. Watanabe, N.Imai, Y. Kajiwara, N. Takeda, and T. Nakajima, 1998, Organicgeochemistry and palynology of Lower Cretaceous Zuunbayan

oil shales, Mongolia: mineral resources in Mongolia: GeologicalSurvey Bulletin, Japan, v. 49, p. 257–274.

Yang, W., 1985, Daqing oil field, People’s Republic of China; agiant field with oil of nonmarine origin: AAPG Bulletin, v. 69,p. 1101–1111.

Yang, W., Y. Li, and R. Gao, 1985, Formation and evolution ofnonmarine petroleum in Songliao basin, China: AAPGBulletin, v. 69, p. 1112–1122.

Yanshin, A. L., 1989, Map of geologic formations of the MongolianPeople’s Republic: Moscow, Academia Nauka USSR, 1:1,500,000,2 sheets.

846 Geochemical Characteristics and Correlation of Oil and Nonmarine Source Rocks