Petroleum Geoche111Istry in Exploration The Norwegian Sheh

Download as pdf or txt
Download as pdf or txt
You are on page 1of 324

Petroleum Geoche111istry

in Exploration of
the Norwegian SheH
Conference Organizer
Elisabeth Holter (Norwegian Petroleum Society)

Conference Committee
Svein S. Eggen (Statoil)
Rolf Magne Larsen (Statoil)
Bruce M. Thomas (Norske Shell)
Anthony K. Yeats (BP Petroleum Development)

Editorial Committee
Anthony G. Dore (Conoco Norway)
Svein S. Eggen (Statoil)
Philip C. Home (BP Petroleum Development)
Rolf Magne Larsen (Statoil)
Bruce M. Thomas (Norske Shell)
Petroleun1 Geochemistry
in Exploration of
the Nonvegian SheH
Proceedings of a Norwegian Petroleum Society (NPF)
conference Organic Geochemistry in Exploration of
the Norwegian Shelf held in Stavanger, 22-24 October 1984

Edited by
B. M. Thomas et a1.

Published by
Graham & Trotman Ltd
for the Norwegian Petroleum Society
First published in 1985 by
Graham & Trotman Limited Graham & Trotman Inc
Sterling House 13 Park Avenue
66 Wilton Road Gaithersburg
London SWlV lDE MD20877
UK USA

British Library Cataloguing in Publication Data


Petroleum geochemistry in exploration of the
Norwegian Shelf: proceedings of a Norwegian
Petroleum Society (NPF) conference held in
Stavanger, 22-24 October, 1984.
1. Petroleum - Geology - North Sea 2. Geochemical
prospecting - North Sea 3. Petroleum industry and
trade - Norway
I. Thomas, B.M. et ai. II. Norsk Petroleumsforening
622'.18282'0916336 TN874.N78

ISBN-13: 978-94-010-8364-5 e-ISBN-13: 978-94-009-4199-1


DOl: 10.1007/978-94-009-4199-1

© Norwegian Petroleum Society


Softcover reprint of the hardcover 1st edition 1985

This publication is protected by


international copyright law. All rights
reserved. No part of this publication
may be reproduced, stored in a
retrieval system, or transmitted in any
form or by any means, electronic,
mechanical, photocopying, recording
or otherwise, without the prior
permission of the publishers.

Typeset in Great Britain by Spire Print Services, Salisbury, UK


FOREWORD
This volume is the record of a three day symposium entitled "Organic Geochemistry in Exploration of the
Norwegian Shelf", which was sponsored by the Norwegian Petroleum Society (Norsk Petroleumsforening)
and held at the Rogalands Regional College, Stavanger on 22-24 October 1984. Twenty-nine papers were
presented, and all but one are published in full herein.
The aim of the conference was to focus on the application of geochemical methods to the current and highly
active exploration of the Norwegian offshore. Emphasis was on practical interpretation and case histories rather
than laboratory methods and techniques, and a strong attendance was sought among geologists and seismic
interpreters active in exploration in Norway and Northwest Europe generally. On all counts the symposium
was a great success with a total of 213 participants registered.
In his opening address Mr Egil Bergsager, director of the Norwegian Petroleum Directorate, observed that
during the 1970s petroleum geochemistry emerged from being a somewhat academic pursuit into a practical
aid in exploration for hydrocarbons. This first stage, when many of the basic methods were developed, has
now led in the 1980s to an expansion into applications in regional geological studies, including mathematical
modelling of thermal history, hydrocarbon migration and basin development.
The papers which follow clearly reflect these trends. They also indicate the high level of geochemical research
in Norwegian exploration, and a remarkable convergence of ideas and methods among the various companies
and workers. The use of Rock-Eval pyrolysis data in routine source rock analysis is now almost universal.
Mathematical modelling of thermal history and hydrocarbon generation is seen in many forms, most of which
owe much to the early work of Karweil (1955) and Lopatin (1971). Regional mapping of thermal maturity on
key source horizons is now routine, and some papers have followed with a detailed consideration of possible
hydrocarbon migration paths.
The complex nature of the principal Upper Jurassic source interval in the North Sea was emphasized by
several speakers, and the concept of organic facies variations within it is now firmly established. The use of
GC-MS techniques in detailed oil-oil, and oil-source correlation studies seems to be progressing from an
earlier experimental stage to one of practical applications in the exploration sphere. Several papers on marine
surface geochemical and so-called "sniffing" techniques indicate increased sophistication in methods and
interpretation, but, as before, still provoke healthy debate.
The symposium was strongly supported by companies active in exploration in Norway, both in terms of
papers and delegates. The organizers are particularly grateful for the high level of technical disclosure
represented in many of the contributions, which clearly indicate the strong emphasis given to geochemical
methods in modern prospect and basin evaluation.
Many thanks also to the anonymous helpers involved in refereeing of papers for this volume, and to the
Rogalands Regional College for making available to the NPF such a superb venue for the symposium.

B. M. Thomas
For the Conference and Editorial Committees

REFERENCES
Karweil, J. 1955. Die Metamorphose der Kohlen vom Standpunkt der physikalischen Chemie.
Dt. Ceol. Ces. Z. 107,132-139.
Lopatin, N.v. 1971. Temperature and geologic time as factors in coalification (in Russian).
Aluid. Nauk SSSR, Izv. Ser. Ceol., 3,95-106.
CONTENTS
FOREWORD (B. M. Thomas) V
REGIONAL STUDIES

1. Organic Facies and hydrocarbon distributions in the Norwegian North Sea


B. M. Thomas, P. M0ller-Pedersen, M. F. Whitaker and N. D. Shaw 3

2. Correlation of the offshore sequences referred to the Kimmeridge Clay


Formation - relevance to the Norwegian sector
A. G. Dore, ]. Vollset and G. P. Hamar 27

3. Organic geochemistry in exploration of the northern North Sea


]. D. Field 39

4. Oil-generating coals
S. Thompson, B. S. Cooper, R. ]. Morley and P. C. Barnard 59

5. Geochemistry of Greater Ekofisk crude oils


W. B. Hughes, A. G. Holba, D. E. Miller and]. S. Richardson 75

6. Correlation of northern North Sea oils: the different facies of their Jurassic
source
M. A. Northam 93

7. Oil-oil and oil-source rock correlation, northern North Sea


L. Schou, S. Eggen and M. Schoell 101

MODELLING

8. Modelling of petroleum genesis and migration with a bidimensional computer


model in the Frigg sector, Viking Graben
P. Ungerer, A. Chiarelli and]. L. Oudin 121

9. Heat flow, geothermal gradients and maturation modelling on the Norwegian


continental shelf using computer methods
R. H. Leadholm, T. T. Y. Ho and S. K. Sahai 131

10. Modelling burial history, temperature and maturation


P. K. Jensen, L. Holm and E. Thomsen 145

11. Basin analysis and maturation modelling onshore Denmark, a case study
from the Danish first round
S. S0rensen 153
12. Quantitative modelling of the sourcing of giant oil fields in the East Shetland
basin
]. C. Goff (abstract only) 161

CASE HISTORIES

13. The geological history and geochemical evaluation of Block 2/2,


offshore Norway
R. H. Gabrielsen, S. Ulvcben, A. Elvsborg and O. Fredrik 165

14. Organic matter quality changes in an Upper Jurassic shale sequence from
the Viking Graben
A. Y. Hue, H. Irwin and M. Schoell 179

15. Organic geochemistry of the Oseberg Field (I)


B. Dahl and G. C. Speers 185

16. Block 33/6 geochemical evaluation


B. Maragna, G. Zaro and P. Pessina 197

17. Maturity studies of the Jurassic "coal unit" in three wells from the
Haltenbanken area
J. L. Pittion and J. Gouadain 205

18. Origin of the gas - condensate of the Midgard Field at Haltenbanken


A. Elvsborg, T. Hagevang and T. Throndsen 213

SURFACE GEOCHEMISTRY

19. Surface geochemical exploration for hydrocarbons in offshore areas -


principles, methods and results
E. ]. Schiener, G. Stober and E. Faber 223

20. Geochemical exploration on the Norwegian continental shelf by analysis of


shallow cores
R. H. Emmel, M. Bjor¢y and G. van Grass 239

21. Regional geochemical analysis of the southern portion of the Norwegian


sector of the North Sea
J. L. Gevirtz, B. D. Carey Jr. and S. R. Blanco 247

22. Recent formed methane-derived carbonates from the North Sea floor
M. Hovland, M. Talbot, S. Olaussen and L. Aasberg 263
METHODS

23. Integrated kerogen typing in the recognition and quantitative assessment


of petroleum source rocks
S. R. Larter 269

24. Hydrous pyrolysis of asphaltenes


N. Telmes, G. C. Speers, A. Steen and A. G. Douglas 287

25. Aromatic hydrocarbon variations in North Sea wells


P. B. Hall, L. Schou and M. Bjorl/ly 293

26. Quantification of steranes and triterpanes by gas chromatographic -


mass spectrometric analysis
B. Dahl, G. C. Speers, A. Steen, N. Teln<es and]. E. Johansen 303

27. Characterization of sedimentary organic matter using nuclear magnetic


resonance and pyrolysis techniques
H. Solli, L. Schou, ]. Kranet, T. Skjetne and P. Leplat 309

28. Study of the diagenesis, catagenesis and metagenesis of coals and dispersed
organic matter with a high-temperature modified Rock-Eval
P. Leplat and]. Paulet 319

29. Analysis of source rocks, reservoir rocks and cap rocks by combined thermal
extraction and pyrolysis - gas chromatography
M. Bjorl/ly, H. Solli, K. Hall and P. Leplat 327
Part I
Regional Studies
1 Organic facies and hydrocarbon
distributions in the Norwegian
North Sea
B. M. Thom.as, P. Moller-Pedersen, M. F. Whitaker and N. D. Shaw
A/S Norske Shell, Stavanger, Norway

Significant hydrocarbon accumulations have been found in North Sea reservoirs ranging in age from Devonian
(Buchan Field) to Eocene (Frigg area). In Norwegian waters, the oldest reservoirs are at present Rhaeto-Liassic
Statfjord Formation sands. Source rock development in the northern North Sea is restricted almost entirely to the
Jurassic. Rich oil-prone source rocks occur mainly in the Humber Group, and particularly within the Kimmeridge
Clay Equivalent (Upper Jurassic 'Hot Shale'). In basinal settings within the Viking and Central Grabens,
Kimmeridge source rock sections may be over 1000 m thick. Net source rock thicknesses exceeding 400 mare
documented in several wells.
North Sea oils appear to be sourced from these Humber Group anoxic shales, although the Liassic Dunlin Group
may also contribute in parts of the northern Viking Graben. Biodegraded oils are found in fields where reservoir
temperatures are less than 75°C, and some possible low maturity crudes occur in the Egersund Sub-basin. Mature
Jurassic 'kitchen' areas coincide closely with the Viking and Central Grabens and deeper parts of the East Shetland
Basin, although a small oil kitchen also occurs in the Egersund Sub-basin.
Rock Eval pyrolysis and solvent extract studies of core material from the Kimmeridge Clay Equivalent suggest
that the principal zone of oil generation for 'Type II' source rocks lies between VR/E = 0.7 and 1. 0*, with the peak
possibly at VR/E = O.S.
Organic facies variations occur within the Kimmeridge Clay Equivalent. In thick basinal sections source rocks
are rich and oil-prone, although the inertinite content may be locally high in turbidite sequences. The Kimmeridge
Clay Equivalent on the basin margins and over some structural highs was deposited under predominantly
dysaerobic conditions, often with some terrestrial influx. Source rocks from these settings are leaner and of mixed
organic type. However, such facies variations are thought to be, at best, only a secondary influence on the
distribution of apparent oil- and gas-prone areas in the North Sea. Other factors, such as thermal maturity of
kitchen areas, timing relationships, migration paths and charge competition are considered to be more significant.
Hydrocarbon distributions and migration paths in the North Sea are complex. In some areas, e.g. Central
Graben, hydrocarbon migration patterns seem to be dominantly vertical, whilst in other areas significant lateral
migration is apparent, e.g. in Troll Field. On many occasions migration has occurred to a stratigraphically lower
level. Geochemical data from deep, mature kitchen areas suggest that hydrocarbon charge vastly exceeds the
available volume in adjacent traps. Many structures appear to be filled to spill-point, and in some areas long chains
of fields are related to each other and a common 'kitchen' as a continuous fill-and-overspill sequence.

INTRODUCTION basin evaluation techniques) as new and effective tools


in modern petroleum exploration. The concept of the
Since the mid-1960s, the development of the North 'generative basin' has rapidly come into widespread
Sea a8 one of the world's major hydrocarbon- use in the exploration industry (Demaison, 1984).
producing provinces has been paralleled by the Hydrocarbons are known to originate from distinctive
emergence of applied organic geochemistry (and allied source beds where they are thermally mature within
mappable 'kitchen' areas. Their subsequent migration
paths are largely determined by the stratigraphic
*Abbreviatiolls: framework and basin geometry, which defines the
'drainage area' of the trapping structure.
VR ~ vitrinite reflectance, measured directly on teleocollinite, The North Sea, with its large data base, rich and
VR/E ~ vitrinite reflectance equivalent, either measured with
another maturity parameter (e.g. spore discolouration) or distinctive source horizons, and numerous hydrocar-
estimated by thermal modelling techniques (e.g. Lopatin bon accumulations, provides an ideal calibration area
method). for such methods, which may then be applied to the
Petrolellln Geochemistry in Exploration o.fthe Nonvegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 3-26.
4 Petroleum Geochemistry in Exploration of the Norwegian Shelf

\

~\

60000' EAST 60"00'


SHETLAND

~
PLATFORM

STORD
BAS I N

-J
HORDA
PLATFORM
59°00'

......--'Fi:'ADEN
~OUND
WITCH SPUR
GROUND
GRABEN

'"""~

FORTH
\
\
APPROACHES

BAS IN

...

MID
NORTH SEA
HIGH

LEGEND
.SIUT OIAP I R

Fig. 1. Tectonic elements and well locations. For regional cross-sections, see Fig. 15.

appraisal of remaining prospects or be used as an TECTONIC FRAMEWORK


analogue for the evaluation of other basins.
The following is a summary of some recent investig- The tectonic framework of the northern North Sea
ations of the hydrocarbon habitat of the Norwegian (Fig. 1) is dominated by the north-trending Viking
sector of the North Sea, with particular emphasis on Graben, which intersects in the south with the
the northern part of the Central Graben and the Vik- southeasterly-oriented Central Graben and the wes-
ing Graben. terly Witch Ground Graben. The thick, mainly-
Hydrocarbon Distributions in the North Sea 5

Mesozoic fill of the North Sea Graben system is over- LITHOLOGY AND MAJOR
lain by gently downwarped Tertiary sediments, com- AGE OIL/GAS OCCURRENCES METRES
prising a total sedimentary section of up to 9 km.
Development of the North Sea rift began in the PLEISTOCENE :=:=:=:=:=:=
600-
Late Permian (Ziegler, 1981, 1982). The primitive PLIOCENE ~~=-~~:=:===
graben system was accentuated by intermittent > MIOCENE :-~:-~-~. 1500
a:
'Kimmerian' tectonic events which occurred between < OLiGOC.

the late Triassic and the early Cretaceous. These i=


a: 300-
movements resulted in deep axial graben systems W
I-
EOCENE 900
bounded by rotated fault block complexes. Such PALEOC.
200-
characteristics are most evident in the Viking Graben, 600

as they are partially obscured by salt flowage and 600-


UPPER
diapirism in the Central Graben and Southern North I-' 1200
W
Sea. a:
()
Triassic sediments within the rift system are almost LOWER 0-1200
entirely non-marine clastics, largely 'red-beds', but by
late Liassic time fully marine conditions were estab- iii UPPER 0-700
lished in the North Sea area and have persisted, with ~ MIDDLE
a: 0-300
minor interruptions, to the present. Rift movements .,
:l LOWER 0-500
gradually diminished during the late Cretaceous, and
ceased altogether after the early Tertiary Laramide
tectonic phase, to be replaced by broad regional
TRIASSIC 0-2000
down warping.
BREHT, ALWYN,
STATFJORD, BERYl
GUllFAKS

HYDROCARBON OCCURRENCES 0-100?


PERMIAN
Principal reservoir intervals and the occurrence of o -1000?
DEVONIAN
hydrocarbons in the northern North Sea are summar-
ized in Figs. 2 and 3. Significant hydrocarbon accumu- CALEDONIAN BASEMENT
lations have been found in reservoirs ranging in age Fig. 2. Stratigraphic occurrence of hydrocarbons, Viking
from Devonian (Buchan Field) to Eocene (Frigg area). Graben (modified after Ziegler, 1980).
In Norwegian waters, the oldest hydrocarbon-bearing
reservoirs are Rhaeto-Liassic Statfjord sands of the LITHOLOGY AND MAJOR
AGE
Statfjord and Gullfaks Fields, and in a recent discov- OIL/GAS OCCURRENCES METRES
ery in 34/4-4. The geographical distribution of North PLEISTOCENE
Sea fields is shown in Fig. 4.
In the Central Graben, the major oil fields occur in PLIOCENE =====:======

t
1000-
unique fractured Danian-Maastrichtian chalk reser- >
a: MIOCENE ---~-: ~
3300
voirs, the largest being Ekofisk (1.0 billion barrels < OLIGOC. =:-:-:-:-:-:- COD SANDS
i=
a:
::::;:. =¥¥
~~ ~ COD. FORTIES
recoverable) which was the first commercial oil field in
Norway. Gas-condensate occurs in Paleocene sand-
W
I-
r--:~-~
~;~~:~~:IC 0- 600
stones in the Cod Field, but this reservoir sequence is 1---+--1 DAN IAN CHALK 0-250
most productive in UK waters where it hosts such oil '--
EKOFtSK. fDDA
WEST EKOftSK
fields as Forties (1.8 billion barrels recoverable), MAASTR. ~ ELDFISK. TOR . .
ALBUSKJELL. HOD
500-
Maureen and Montrose. To date, Jurassic reservoirs UPPER
VALHALl
TOMMELITEN
1100

of the Central Graben have yielded only a few small


discoveries in Norwegian waters, the most significant 0-700

being the Ula Field (150 million barrels recoverable),


0-1200
where oil occurs in marine sandstones of Oxfor- ~P5_~"~N~
PIPER, CLAYMORE
dian-Kimmeridgian age. ULA. FULMAR
0-500
In the Sleipner area of the southern Viking Graben,

*
gas-condensate occurs widely in Lower-Middle TRIASSIC
TRIASSIC SAJoWS
JOSEPHINE 0- 1200
Jurassic sandstones of the Vestland Group, and in the
'Gamma' structure, also in overlying Paleocene sands. ~ ZECHSTEIN ZECHsmN
0-1000
In contrast, in adjacent UK waters, oil is found in a ~ ROTLlEG.87 AUK, ARGYll
0-500
string of fields (Brae, Tiffany, Toni, Thelma) where
Upper Jurassic fault-scarp sandstones are productive. L. CARBONIF ~ OLO REO SANOS
0-600+

The Gudrun Field, in Norwegian waters, has oil in DEVONIAN~) • BUCHAN o 600+
Upper Jurassic turbidites (presumably distal equival- CALEDONIAN BASEMENT
ents of the Brae sands) and gas in Middle Jurassic
sands of the Vestland Group. Fig. 3. Stratigraphic occurrence of hydrocarbons, Central
In the central Viking Graben, known accumulations Graben (modified after Ziegler, 1980).
are largely in the Tertiary, where Eocene sandstones
are gas-productive in the Frigg area; the Paleocene Jurassic contains oil at Heimdal and Idun, and in UK
contains gas at Heimdal and oil at Balder. Both fields Bruce and Beryl.
Paleocene and Eocene reservoirs are believed to be In the northern Viking Graben and on its adjacent
deep-water sand fans shed from westerly sources on flanks, Middle Jurassic paralic sands of the Brent
the East Shetland Platform. The Lower-Middle Group are widely productive and host major
6
Petroleum Geochemistry in Exploration of the Norwegian Shelf

2°00' 4°00' 62000'


0°00' ~~----_,
62°00'r-------~--~--~--------~------

61°00'

60°00'

.
,,

< .'
I
59°00' I 59°00'

f
0 1GUDRUN

OB•Attel
~ :
:~AGN' ~
~
PIPER TIFFANY,

~
• TONI..
m~
OSlEIPHER" ""
~THEL"",,~ 16'
6ALMOR"'l~ \ \ 1
-..l , ",MAUREE N
~'u ....BEL I
"' ANDREW
,
1 .,
58°00'

57°00'
57°00'

~
• OIL ACCUMULATION
o GAS ACCUMULATION

Fig. 4. Major fields, northern North Sea.

accumulations in Norwegian waters at Statfjord, Gull- sandstones of the Vestland Group in the Egersund
faks and Oseberg. Oxfordian-Kimmeridgian 'Humber Sub-basin.
Group sands' are gas- and oil-bearing in the Troll
Field, and contain oil at Brage. Rhaeto-Liassic Statf-
jord sands are also oil-productive at Statfjord and SOURCE ROCKS
Gullfaks, as well as in the Brent Field in UK waters. A
small gas discovery has been made in Lower Cretace- Geochemical profiles of North Sea wells indicate that
ous submarine fan sands at Agat. Three small oil the occurrence of significant source rock intervals is
accumulations have also been found in Middle Jurassic restricted almost entirely to the Jurassic section (see
Hydrocarbon Distributions in the N01'1h Sea 7

CROMER KNOLL GROUP

T.O.C.=O.4 10 3.3 '1.

x x
w w
0 o
lE lE
z z
w
w
"a:
0
0
400 "oa:o
> >
J: J:

200

50 100 150 50 100 150


OXYGEN INDEX OXYGEN INDEX

HEATHER FORMATION DUNLIN GROUP

T.O.C.=O.6 to 10.9 % T.O.C. =0.9 to 6.9 %

x x
w w
0 0
lE lE
z z
w w

"
0
a:
0
"a:
0
0
> >

J: J:
@

. .\ .. A
© •••

~.
....
• \
~ ... •
• •
• • • • • •~ .(
50 100 150 50 100 150
OXYGEN INDEX OXYGEN INDEX

BRENT GROUP ORGANIC FACIES


TYPE I

T.O.C.=O.B to 37.6 %

x
w
0
lE x
w
Z 0
w lE
"0
a:
z
w
0
>
J: "a:
0
0
400

>
J:

® ®
MIXED FACIES

200
VITRINITIC FACIES

.~ ~ ~TYPEIII
•• -INERTINIT,IC FACIES TYPE IV

50 100 150 50 100 150


OXYGEN INDEX OXYGEN INDEX

Fig. 5. Organic facies, Jurassic and Lower Cretaceous units,


8 Petroleum Geochemistry in Exploration of the Norwegian She(f

Figs. 2, 3 and 5). Source rocks within the meridge kerogen (high hydrogen index, sapropelic
Rhaeto-Liassic Statfjord Formation and Middle facies; see Fig. 5(b», although it is evident that there
Jurassic Brent Group-coals and carbonaceous is significant variation in quality within most units
shales-are mainly gas-prone (Fig. 5(e». Marine including the Kimmeridge equivalent (see Core Calib-
Liassic (especially Toarcian) shales are rich oil-prone ration Study, below).
source rocks over extensive areas of continental
Europe and in Great Britain. North of the Mid-North
Sea High, however, equivalent marine shales of the SEDIMENTARY AND ORGANIC FACIES
Dunlin Group (Drake Formation) are mainly organi- OF THE HUMBER GROUP
cally lean and inertinitic, except in the greater
Oseberg area where this interval has some potential The Humber Group was deposited under marine con-
for oil generation, see Fig. 5(d). ditions in Bathonian to Ryasanian times in the deepen-
The best source rocks in terms of richness, thick- ing North Sea rift system (see Fig. 6). Broadly, three
ness and type, occur within the Humber Group (Bath- lithostratigraphic units are recognized within the
onian-Ryazanian). The richest intervals fall within group: an upper, highly radioactive black shale which
the Kimmeridge Clay Equivalent (Draupne, Farsund was deposited under anoxic conditions (Kimmeridge
and MandaI Formations of proposed new stratigraphic Clay Equivalent); a lower, less radioactive marine
nomenclature (see Vollset and Dore, 1984); Upper shale deposited under more oxic conditions (Heather
Jurassic 'Hot Shale' of some authors (referred to Formation); and, on the basin margins, a coastal sandy
herein as 'Kimmeridge'), where organic content is fre- sequence laterally equivalent to parts of both shale
quently in the range 5-12% total organic carbon units, which is the reservoir in the Troll Field. In
(TOC) and the organic type is predominantly oil- some areas, the Kimmeridge may contain fault scarp
prone. The underlying Heather Formation ('Cold sands or basinally-equivalent turbidites, reflecting
Shale' of some authors) is often much leaner (2-4% 'Late Kimmerian' tectonism. On high blocks, where
TOC) with an organic type which is either dominantly sections are usually thin, the Kimmeridge/Heather
gas-prone or inertinitic. However, in some areas the Formation lithological distinction is clear and an
Heather Formation may contain richer, oil-prone unconformity may separate the two units (Kirk, 1980).
source rocks and the Kimmeridge Clay Equivalent Deeper in the basin, where a thicker section is pre-
may also include lean intervals of relatively poor qual- served, the relationship is often gradational since
ity. The concept of 'organic facies'-lateral variation sedimentation was largely continuous.
in source rock type and potential within a single When viewed in terms of chronostratigraphy, the
stratigraphic unit-is therefore relevant within the Kimmeridge Clay Equivalent can be seen in its thick
.Jurassic of the North Sea, and will be discussed in basinal expression (e.g. in the Gudrun area) to span a
more detail later. Within a thick Upper Jurassic period as broad as from Oxfordian to Ryazanian,
marine shale section in the Egersund Sub-basin, only whilst on the basin margin it may be thin and limited
the Tau Formation (Kimmeridgian-Volgian) contains mainly to the Ryazanian stage (e.g. in the Troll area).
rich, oil-prone source rocks. In an intermediate structural position in the basin
The thickness of the Humber Group (and its source (e.g. in the Oseberg area), Heather Formation of
intervals) varies dramatically, reflecting deposition in Oxfordian age can be seen to be age-equivalent to
a tectonically active graben system. In the Central coastal sands in the Troll area and organic Kim-
Graben, basinal sections have been documented con- meridge shales of the basin deep. The lithostrati-
sisting of over 1000 m of almost continuous source graphic units of the Humber Group may thus be consi-
rock. In the southern Viking Graben, thick basinal dered as time-transgressive facies variants in a single
'black shales' are interbedded with oil-bearing turbi- marine sedimentary system, as illustrated in simp-
dite sands. On active Jurassic highs or on the flanks of lified model form in Fig. 6(a).
the basin, the Humber Group may be thin, incomplete In these terms, the geological history of the North
or totally eroded. Since the majority of wells in the Sea rift during Humber Group times may be better
North Sea have been drilled on these structural highs, understood, as follows:
they provide a potentially misleading and often pes- In the Bathonian, the Brent-Vestland delta was
simistic record of source rock occurrence which cannot drowned during the Humber Group marine transgres-
be extrapolated into adjacent un drilled basinal areas. sion. A widespread basal Heather 'hot shale' suggests
In the axial parts of the Viking and Central Grabens, that at least dysaerobic, and possibly anaerobic, con-
well control on Jurassic 'sinks' is often poor or incom- ditions were soon established in the marine basin, pre-
plete yet, without doubt, it is from within these deep sumably due to stagnant salt-water stratification.
mature kitchen areas that the majority of North Sea With further transgression (except in the most basinal
oil and gas reserves originate. settings) the water column became increasingly
Variations in organic quality within the Jurassic and aerobic, leading to a widespread ventilation of the
Lower Cretaceous of the North Sea area are illus- basin in the late Callovian-early Oxfordian. This
trated in Fig. 5, where Rock Eval pyrolysis data event marks the linking of the southern and northern
(Espitalie et al., 1977) have been plotted on a 'van parts of the North Sea across the Mid North Sea High,
Krevelen'-type diagram. The evolutionary paths of and ended faunal distinctions between the two areas
the organic types of Tissot et al. (1974) are indicated, which are apparent up to this time.
together with the inertinitic Type IV of Harwood During the Callovian, Oxfordian and Kimmeridgian,
(1977), and areas on the hydrogen index/oxygen index the basin appears to have been largely oxygenated
plot have been assigned to various 'organic facies' except in the deepest grabens, where anoxia had been
according to their chemical and visual characteristics established during the early stages of transgression.
(see Fig. 5(f). The low quality (inertintic facies) of Even in basinal settings, conditions at this time seem
Cromer Knoll kerogen is immediately apparent (see to have varied along the axis of the Viking Graben.
Fig. 5(a» and contrasts strongly with oil-prone Kim- Fully anaerobic conditions are apparent in the Gudrun
Hydrocarbon Distributions in the North Sea 9

STRUCTURAL BASINAL TRANSITION FLANK


SETTING eg. Gudrun area eg. Oseberg area eg. Troll area
AGE

130
VALANGINIAN CROMER KNOLL GROUP
RYAZIINIAN
140
KIMMERIDGE CLAY EQUIVALENT

150
VOLGIAN
V- NO RECORD

160
KIMMERIDGIAN

OXFORDIAN f-----
f---------,
HEATHER FORMATION
®
. BASAL
CALLOVIAN
------
.. "HUMBER GROUP
~
HOT SHALE_ _ _ _ .....l
BATHONIAN SANDS"
170 ~

BAJOCIAN
BRENT GROUP AND EQUIVALENTS
180
my
LITHOSTRATIGRAPHY

AGE

130,--------r-----------------------------------------------------.
VALANGINIAN MARINE AEROBIC

RYAZANIAN
140
MARINE ANAEROBIC
VOLGIAN
NO RECORD
150 +-------~

SHALLOW
@
MARINE SANDS

BAJOCIAN
PARALIC SANDS
180 l========L______________________________________________________
my
ENVIRONMENTS

AGE

130
VALANGINIAN
INERTINITIC FACIES (m 1llZ:)
RYAZANIAN
140 IIIm
SAPROPELIC FACIES cn)
VOLGIAN ~

150 NO RECORD

160
OXFORDIAN
MIXED RESERVOIR
©
CALLOVIAN
FACIES FACIES
BATHONIAN (1I/m)
170

BAJOCIAN
RESERVOIR FACIES
IBO
my
ORGANIC TYPE
Fig. 6. Schematic facies model of the Humber Group, Viking Graben.
10 Petrolenm Geochemistry in Exploration of the Norwegian Shelf

area as early as in the Oxfordian, whilst in the north effects could at times be partly the result of such
the basin was largely aerobic until Volgian- Ryazanian 'noise' in the data.
times. This may reflect the broader geometry of the For example, on the basin margins or in wells
Viking Graben in the north, and the opportunity for located on structural highs, the Kimmeridge interval
access to oxygenated waters from the juvenile Atlan- is usually thin, and geochemical data may often sug-
tic-Arctic seas. gest that the unit is present in a mixed or inertinitic
As rifting activity intensified and the basin facies. However, it is under such conditions that cav-
deepened during the Kimmeridgian-Volgian, condi- ing effects will be greatest, and since the overlying
tions became more widely anaerobic and were almost Cromer Knoll Group is organically lean and inertinitic,
universal in the basin during the Ryazanian, which some dilution of apparent Kimmeridge quality is
corresponds to the peak of anoxia in many areas. Fol- almost inevitable. Hand-picking often reveals the
lowing the Ryazanian, rapid deepening of the basin higher quality of in situ Kimmeridge black shales.
and opening of broader seaways to the Nor- In a few wells, oil-based muds were used. In addi-
wegian-Greenland sea and also to Tethys (via the Pol- tion to precluding any solvent extraction analysis of
ish Trough) resulted in freer water circulation and the cuttings from these wells, even pyrolysis data are
re-establishment of aerobic conditions, which are adversely affected by the presence of such excess
marked by the Cromer Knoll Group. quantities of free hydrocarbons (Clementz, 1979), and
Source potential and the concept of organic facies Rock Eval data from untreated samples may exhibit
may be simply related to this Humber Group sedimen- an abnormally high hydrogen index (Sz/TOC). This
tary model as follows (see Fig. 6(c». Conditions for effect may also occur in uncontaminated samples of
preservation of organic matter are ideal in an very rich source rocks (such as the Kimmeridge),
anaerobic environment, therefore black shales of the where they are currently at peak oil generation. In
Kimmeridge facies are organically rich and form an this condition a large excess of free hydrocarbons is
oil-prone sapropelic facies (Type II of Tissot et al. present in the source rock and leads to spurious
(1974». Under aerobic conditions, organic matter, results. A more reliable hydrogen index is obtained
particularly that of marine origin, is quickly oxidized if sample material is first extracted with organic
and destroyed, leaving only inert remnants and solvent.
more-resistant material of largely terrestrial origin. Another cause of unreliable Rock Eval data is the
This kerogen is hydrogen-poor (Type III to Type presence of carbonates, especially siderite, which
IV) and represents an inertinitic facies, as seen results in an abnormally high oxygen index due to
in the Cromer Knoll Group and parts of the Heather decomposition of the carbonates during pyrolysis. The
Formation. solution is a simple acid decarbonation of calcareous
The Heather facies is highly variable in character, samples before Rock Eval analysis.
and represents an intermediate condition in the basin Despite the hundreds of exploration wells in the
where terrestrial input was limited but conditions North Sea, only a handful have been located in a truly
within the water column varied from aerobic to mildly basinal setting and been drilled deeply into a genera-
anaerobic (i.e. dysaerobic), see Fig. 6(b). A dys- tive Jurassic sequence. Such key wells provide critical
aerobic environment reflects the condition where the data for calibration of charge models, and allow direct
oxic/anoxic interface lies close to the sedimentary observation of hydrocarbon generation without the
surface, providing a variable and imperfect condition need for laboratory stimulation. In the search for con-
for organic preservation. Burrowing organisms, which trol in the mature kitchen areas of the Central and
are absent in a truly anoxic environment, may occa- Viking Grabens, it became evident that existing
sionally be active, and partial oxidation of organic cuttings-based geochemical data are often unreliable,
matter will occur (Demaison and Moore, 1980). The apparently due to the widespread use of turbo-drilling
result is a 'mixed' organic facies where Type II (sap- techniques in theRe deep Rynclinal wells. The effects of
ropelic) and degraded marine material is combined 'bit metamorphism' (Taylor, 1983) include elevation of
with Type IV (inertinite), frequently with an admix- apparent thermal maturities which are incompatible
ture of Type III (vitrinitic) material. A similar facies with current burial depths and temperatures, and a
is also recognized within the Kimmeridge Clay Equi- comparable reduction in pyrolysis-yield consistent
valent along the basin flanks and on some high blocks with a general 'burnt-out' character. Under these con-
where dysaerobic conditions often prevailed (e.g. ditions palynological age determinations may also be
Troll area). impossible.

CORE CALffiRATION STUDY Core calibration set


Geochemical data from cuttings In view of the inherent uncertainties in all cuttings-
based geochemical data, a calibration study of Kim-
The concept of a range of organic facies within indi- meridge Clay Equivalent source rocks was conducted,
vidual Jurassic source units of the North Sea area has and was based almost entirely on core (conventional
been proposed on several occasions (Dypvik et al., and sidewall) and outcrop sample material.
1979; Demaison & Moore, 1980; Barnard & Cooper, The aims of the study were:
1981) and has been described in some detail by
Robertson Research International (1979, 1981). (1) to investigate more precisely the organic charac-
Detailed regional mapping of organic facies is more ter of Kimmeridge source rocks;
difficult, however, as the majority of available (2) to document possible variations in organic facies of
geochemical data is based on cuttings material, with the Kimmeridge interval using reliable sample
attendant problems of poor sample quality, caving and material; and
contamination. Subtle variations attributed to facies (3) by choosing material from a wide range of thermal
Hydrocarbon Distrilmtions in the North Sea 11

KIMMERIDGE CLAY EQUIVALENT


CALIBRATION SET

900 TYPE I
DEPOSITIONAL
~
800 X Flank
.a. Transition
• Basinal
• Basinal
with turbidites
700 TYPE IT
Anaerobic
Environment

600

+)( 500
w
C
Z

Z
w
<!l 400 "TYPE IT/m"
oa: Dysaerobic
c
> Environment
:r.
with
~
« 300 Terrestrial
>
w Influx
:.:
()
o
a: 200

100
TYPE m
Aerobic
Environment

TYPE Dr
2.0 1.5 10 08 06 0.5 04 0.3

... VITRINITE REFLECTANCE

Fig. 7. Organic facies and maturation pathways, Kimmeridge Clay Equivalent. For extract data, points 1-5, see Fig. 8.

maturities, to study the generation of hydrocar- vitrinite reflectance (using a logarithmic scale) pro-
bons within the Kimmeridge source rocks. vides the required discrimination (see Fig. 7) and also
overcomes the inherent unreliability of oxygen index
The calibration set covers a range of thermal and atomic Oje ratio in van Krevelen-type diagrams.
maturities from VRjE = 0.3 to 1.0, and comes from a Samples in the calibration set are categorized accord-
variety of structural positions in the basin. All sam- ing to their depositional setting in the basin-'flank',
ples are dark grey-black, organically-rich shales, often 'transition', 'basinal', and 'basinal with turbidites'
finely laminated and sparsely fossiliferous, consistent (refer also to Fig. 6).
with deposition in a broadly anoxic environment. For immature samples, the hydrogen index of
Source potential and organic type were routinely Kimmeridge kerogen varies from less than 100 to
determined by 'Rock EvaI' pyrolysis. Since spot more than 700 (see Fig. 7). There is little evidence of
checks indicated the presence of some minor carbon- segregation into discrete organic types as proposed by
ate, many samples were rerun after decarbonation. Tissot et al. (1974); rather, a continuum exists be-
Samples with a thermal maturity above VRjE = 0.7 tween 'Type II' and 'Type IV'. Kimmeridge source
were also re-analysed after solvent extraction. Dupli- rocks from deep basinal wells are organically rich
cate samples were prepared using standard palynolog- (5-12% TOe) and mostly have very high hydrogen
ical methods, and were examined microscopically indices (500-700) which are characteristic of a predo-
under both white and ultraviolet light. minantly sapropelic facies. Microscopic examination
When plotted on a standard van Krevelen-type confirms that such samples consist almost entirely of
plot (hydrogen index versus oxygen index), see highly fluorescent, structureless sapropelic organic
Fig. 5(b), the calibration set exhibits a wide range in matter (SOM) of probable algal/bacterial origin. Simi-
apparent organic types, but offers little distinction lar intervals also occur in some wells on the basin
between immature source rocks of rather poor prim- margin, particularly in the Ryazanian, when the sap-
ary quality and post-mature source rocks whose past ropelic facies extended across the whole basin. Sam-
potential was probably high, since both have low hy- ples from the flanks of the basin, however, are often
drogen indices. A cross-plot of hydrogen index versus leaner (2-5% TOe) and have mainly moderate hyd-
12 Petroleum Geochemistry in Exploration of the Norwegian She(f

rogen indices (200-400) which are suggestive of a pre- quences where thin but rich hemipelagic shales are
dominantly dysaerobic environment, probably with interbedded with more massive sand bodies. Under
some terrestrial influx. Microscopic examination con- these conditions, geochemical data from cuttings mat-
firms a 'mixed' organic facies where structureless erial may also be pessimistic, due to dilution effects.
organic matter can often be replaced by spores and Pyrolysis data may be coupled with indications of
inertinite as the dominant constituents. In addition, source potential from wireline log character to provide
structureless organic matter deposited under such an estimate of net source thickness, which can then
dysaerobic conditions often appears to be partially form the basis of a quantitative estimate of hydrocar-
oxidized and degraded, is only weakly fluorescent and bon charge (Nederlof, 1979), see below.
is probably of reduced generative potential. At least
some of this hydrogen-poor structureless material is
thought to be of terrestrial origin. A REFINEMENT OF THE 'OIL WINDOW'
Some Kimmeridge samples from a deep basinal set- CONCEPT
ting in the southern Viking Graben have very low
hydrogen indices (ca. 100), although thermal The concept of the 'oil window' (Landes, 1967; Pusey,
maturities are moderate (see Fig. 7). This material is 1973) began with the definition of the oil 'floor' in
from black hemipelagic shale within a thick turbidite terms of coal rank. White (1915) observed that in the
sequence, and microscopic examination reveals the Appalachian area oil no longer occured if associated
presence of abundant inertinite and oxidized SOM, coals were above about 65% fixed carbon, and thus
which was apparently carried into the basin along with formulated his 'carbon ratio theory'. The oil floor is
coarse clastic material from source areas to the west. essentially a cracking phenomenon, representing the
Hence, organic-rich shales of low source potential may lower (deepest) limit of thermal conditions suitable for
also occur in the deep basin, and are often interbedded the preservation of liquid hydrocarbons, and is
with other more oil-prone source rocks (see Fig. 6). reported by various authors to lie between VR = 1.2
The calibration set thus confirms a wide range of and 1.35.
organic facies within the Kimmeridge source interval, The onset of kerogen transformation to oil (the
and provides a detailed correlation between optical upper limit of the oil window) was originally deter-
and chemical typing methods. Above about mined by solvent extraction techniques (Philippi,
VR/E = 0.7, hydrogen indices decline due to matura- 1965) and is thought to lie between VR = 0.5 and 0.7,
tion effects, as will be discussed later. depending to some extent on organic type. This
broadly-defined oil window (VR = 0.5 to 1.35) there-
fore represents a generalized observation from many
Wireline log expression
source rocks and in numerous basins. It embraces
With their unusual physical, as well as chemical, prop- upper zones of generation and oil expUlsion, and a zone
erties, rich source rocks have a distinctive expression beneath it where oil may exist after generation
on wireline logs (Meyer and Nederlof, 1984). Because because thermal conditions (defined by the oil floor)
of their high organic content, source rocks (and coals) are not sufficiently severe to cause large-scale crack-
have a low density (Schmoker, 1979) and a correspond- ing to gas. This lower zone of oil preservation is
ingly high sonic transit time. As source rocks become characterized by high levels of extractable hydrocar-
mature, their resistivity often increases dramatically bons, but is not a site of major oil generation.
due to the presence of free hydrocarbons during the The Kimmeridge core calibration set therefore
main phase of oil generation (Meissner, 1984). At high offers an opportunity to define more precisely the
maturities, resistivity again declines as the remaining 'principal zone of oil generation' for a specific source
oil is progressively cracked to gas and expelled. rock. Such a calibration will allow more-precise use of
The most widely observed petrophysical property of thermal maturity maps in defining oil and gas kitchen
rich marine source rocks is their high radioactivity, as areas in the North Sea.
indicated on gamma-ray logs (BjS'/rlykke et al., 1975; Rock Eval pyrolysis data, being semi-quantitative,
Schmoker, 1981). The abnormal levels of radioactivity provide a potentially precise indicator of the genera-
are largely due to the presence of uranium, which is tive process, as the relative pyrolysis yield (i.e. hyd-
preferentially enriched in organic matter deposited rogen index) of a source rock is seen to decline dramat-
under stagnant, anoxic conditions. Studies made by ically during the generation of oil (see Fig. 7). Taking
the authors of Kimmeridge data indicate that there is the maturation path of Type II kerogen, little decline
often a close relation between relative gamma-ray is seen in hydrogen index from its initial immature
count and total organic carbon content, but that abso- values of 600-700 until at least VR/E = 0.7. A major
lute values of radioactivity on wireline logs may not be reduction in pyrolysis yield, presumably due to gener-
a direct indication of source potential. This arises ation and expUlsion of hydrocarbons, occurs between
because gamma-ray log response is influenced by both VR/E = 0.75 and 1.0, with the probable peak at
borehole size and mud weight (Schlumberger, 1979, around VR/E = 0.8. Electrical resistivity measure-
p. 17), and in many deep, basinal wells the natural ments of mature Kimmeridge source rocks also are
radioactivity of the Jurassic section is suppressed by highest at VR/E = 0.8 (Goff, 1983), tending to con-
the high mud weights used to combat high formation firm this value for peak oil generation. By
pressures. In addition, the ability of uranium to com- VR/E = 1.0, Kimmeridge source rocks have declined
plex with organic matter under anoxic conditions also to a 'Pseudo-vitrinitic' composition and appear to be
varies with organic type, such that the most oil-prone post-mature for oil generation. It is likely that by this
material may not have the highest gamma-ray stage active gas/condensate generation is already
response (Schlumberger, 1982). occurring.
Wireline logs are a poor indicator of source potential A composite generative profile of Kimmeridge
where source beds are too thin to be resolved by log- source rock extracts confirms the above model. Gas-
ging tools. This occurs particularly in turbidite se- chromatograms of solvent extracts from the calibra-
Hydrocarbon Distributions in the North Sea 13

tion set indicate immaturity in samples of


VR/E = 0.51 and 0.56 (Figs. 7 and 8, points 1 and 2). A
KIMMERIDGE CLAY EQUIVALENT sample at VR/E = 0.73 is largely 'oil-like' with some
slight indications of immaturity (point 3). Extracts of
source rocks at VR/E = 0.77 (point 4) and 1.0
Ph
IMMATURE
VR/E ,0'51 (point 5) are fully mature and are closely comparable
H.I. '666 with many North Sea light crudes.

OIL GEOCHEMISTRY AND OIL-SOURCE


CORRELATION
A number of typical oils from the Norwegian North
Sea are illustrated in Fig. 9. The majority of North
Sea crudes are medium to light (30-40 a API) and
paraffinic, with a low sulphur content (less than 0.5%),
IMMATURE
e.g. Brage, 30/6-5, Statfjord, etc. Another group of
medium-to-heavy (24-28 a API) naphthenic oils,
VR/E,0·56
which are deficient in normal alkanes, are believed to
H.I.' 585
.11 be biodegraded, e.g. Troll, Balder, Frigg and Gull-
I
Ph
faks. Some relatively heavy (28-29 a API) oils from
® the Egersund Sub-basin are not obviously biodeg-
raded, e.g. Brisling, and may be low-maturity crudes.
The lack of light ends and a high isoprenoid and sul-
phur content support this interpretation. The large
paraffinic group of oils fits generally into the 'marine'
type of Tissot and Roucache (1980), although a moder-
ate wax content suggests a contribution of terrestrial
plant material. Low pristane/phytane ratios are
EARLY MATURE consistent with source rocks deposited in anoxic
VR/E ,0·73 conditions.
H.I.' 578 In general, the API gravity of North Sea oils
increases with reservoir depth, and the trend is even
@ stronger for API versus formation temperature (see
Fig. 10). The biodegraded oils are restricted to shal-
low reservoirs with present-day temperatures below
75 aC. A few oils which plot away from the dominant
trend appear to be mildly transformed, perhaps by
water-washing.
As was demonstrated earlier (Figs. 8 and 9) solvent
extracts from thermally-mature shales of the sap-
OIL-GENERATIVE ropelic facies of the Kimmeridge Clay Equivalent
, 20 VRiE,O'77
compare closely in terms of normal alkane distribution
i: H.I.' 264 with many North Sea paraffinic crudes. This observa-
II tion is generally supported by more-detailed correla-
I I tions using mass spectrometric techniques. Additional
25 @ confirmation of the genetic link between Kimmeridge
source rocks and North Sea oils is provided by artifi-
cial maturation experiments, where immature source
rocks are heated under controlled laboratory condi-
tions (see Fig. 11). A grossly-immature Kimmeridge
source rock (Type II, hydrogen index = 666) has an
extract GC-profile quite unlike any North Sea oil.
After heating for 6 days at 330 aC (Lijmbach et at.,
GAS-MATURE
I I 1983), the resultant 'synthetic crude' is comparable
I I
VR/E,1·00 with many naturally occurring North Sea oils.
1
20 H.I.' 116 Carbon isotopes are widely regarded as a powerful
oil-source correlation tool, and have been used with
! II II ® apparent success in a number of basins (see Hunt,
I
1 25
I
1979). North Sea oils vary quite widely in carbon
isotope values, from approximately -27 to -31 per
mil, which may in isolation be taken as an indication of
more than one source. However, carbon isotope data
from solvent extracts of Kimmeridge cores in the
calibration set cover a range from -26.8 to -31.1 per
Fig. 8. Composite generative profile, Kimmeridge Clay mil, and most of this regional range has also been
Equivalent (Type II kerogen). Refer also to Fig. 7. observed in a single well section. Similar variations
are described by Reitsema (1983) fqr Kimmeridge
kerogens in the Brae area. In the Ekofisk area of the
14 Petroleum Geochemistry in Exploration of the Norwegian Shelf

API GRAVITY

10 20 30 40 50 60

NORTH SEA OILS


60

------U -------
BIODEGRAOED
30/6-5
390 API
p, -28'2%0 w
80 •
LOW MATURITY ~ ..
-75°C

a: OILS •• .::-'.
'0 ::J
~ • •
I•
ffi
"-
100
••• ttl• PRIMARY
OILS

,..
~ MILDLY
w
l- TRANSFORMED ••• e.

IJI.!.UL II:
(5
120

~•
>
a: •
w
~ 140

a: •
I

BRAGE 160
---
Fig. 10. API gravity versus reservoir temperature, North
.
37 0 API
-28'7~00 Sea oils.
'0
Central Graben, the carbon isotope values of analysed
Ph oils range from about -27 to -30 per mil, yet they all

~ I I.,
seem to be derived from the same kitchen area, where
~ l~ i !~ lUm ~
Upper Jurassic shales are the only apparent source
rocks.
If such a diverse stratigraphic unit as the Kim-
meridge Clay Equivalent is considered to be a 'single'
P,
source, a precise correlation between the carbon
BRISLING isotope values of an oil (representing the average of
I"h
i
280 API the charge from an entire kitchen area) and a few spot
-29'1 '100
source rock samples (possibly not even from the
I
I
I III' mature kitchen area) may be largely fortuitous. It also
follows that a good apparent correlation will not be
1

25

,
convincing unless it is accompanied by other indepen-
dent lines of evidence.

I~IJ .I ,HUll
I The hazards in oil-source correlation are amplified
-
I
I A in the Oseberg area where the Liassic Dunlin Group is
lilt,
",,"
WI
1" also a potential oil-source. Available extract data indi-
cate that Dunlin shales are generally similar in charac-
ter to overlying Kimmeridge source rocks, both being
TROLL of marine-sapropelic type. Therefore, it s.eems likely
that a Dunlin-sourced oil may be difficult to distin-
guish from one of Kimmeridge origin, bearing in
mind the regional variations in source rock facies and
oil types which are apparent in the North Sea.

TEMPERATURE DATA AND


GEOTHERMAL GRADIENTS
Average geothermal gradients to the top of the
Jurassic have been computed from temperature data
in over 150 North Sea wells (see Fig. 12). Estimates of
true formation temperatures were made from wireline
GULLFAKS
log data using a Horner-type plot, and have been cali-
brated with temperatures from drill stem tests where
290 API
available.
-29'4%0
Geothermal gradients within the Viking Graben are
relatively low, ranging from 3.0 to 3.5 °C/100 m,
whilst higher gradients occur along the flanks and on
basement highs. A pronounced thermal anomaly in the
Troll Field (5.0 °C/I00 m) is attributed largely to the
insultating effects of this giant gas accumulation. In
contrast with the Viking Graben, the axis of the Cen-
tral Graben is associated with high average gradients
Fig. 9. Selected North Sea oils. (more than 3.75 °C/IOO m), although the data set may
be biased since most wells in this area were drilled on
salt-induced features.
Hydrocarbon Distributions in the North Sea 15

Ph
IMMATURE
EXTRACT
---
SYNTHETIC
p, VR/E~O·51 CRUDE
I H.I. ~ 666 20

I
I
LABORATORY
MATURATION

• 30

1111.I.lll l I

Fig. 11. Laboratory maturation experiment, Kimmeridge Clay Equivalent (type II kerogen).

THERMAL MATURITY TRENDS ration patterns seem to be almost entirely vertical,


and fields are coincident with mature kitchen areas,
Reliable vitrinite reflectance measurements are only whilst in others large-scale lateral migration is appar-
obtained from properly identified teleocollinite, one of ent. On many occasions, migration has occurred to a
several species of vitrinite which may be distinguished stratigraphically lower level. Some parts of the basin
microscopically (Buiskool-Toxopeus, 1983). For this appear to be predominantly gas-prone, e.g. the Frigg
reason the direct measurement of thermal maturity is and Sleipner areas; others are prolific oil provinces,
difficult in many North Sea wells. Reflectance data e.g. Ekofisk area and Tampen Spur.
from the Cretaceous and Tertiary are often spurious, Geochemical studies were coupled with wireline log
and detailed maturity profiles of entire well sections indications of source rock thickness in the Humber
(from which geothermal history may be recon- Group for a number of key basinal wells (see Fig. 1),
structed) are almost impossible to obtain. Vitrinite is as follows:
virtually absent within the rich Kimmeridge source
interval (a common feature of many Type II marine Well 30/4-1 (Oseberg Kitchen)-200 m net source rock
source rocks), and reliable measurements in many Well 24/9-1 (Balder Kitchen) -420+ m net source rock
wells are obtained only from a few coaly intervals in Well 15/3-1 (Sleipner Kitchen) -580 m net source rock
the Middle to Lower Jurassic section. Well 2/11-1 (Ekofisk Kitchen) -570+ m net source rock
In these studies, the thermal modelling method of
Lopatin (1971) was used to estimate maturity levels These data from the deep mature kitchen areas sug-
from burial history and geothermal gradients. Temp- gest that hydrocarbon charge vastly exceeds the
erature d~ta were carefully evaluated, since the available volume in adjacent traps. Many structures
method of Lopatin is very sensitive to this parameter, appear to be filled to spill-point and, in some areas,
and the results of modelling were calibrated in wells long chains of fields can be seen to be related to each
with reliable measurements to vitrinite reflectance. other and a common kitchen area as a continuous fill-
Present geothermal gradients were used (see and-overspill sequence (Gussow, 1954).
Fig. 12), since they usually provide a good correlation
between measured and estimated vitrinite reflec- Northern Viking Graben (60°30' to 62°00'N)
tance.
Using seismic structure maps and well data, a reg- Agat Kitchen
ional thermal maturity map of the Northern North The most northerly occurrence of hydrocarbons in the
Sea was prepared for the top of the Kimmeridge North Sea is at 'Agat' (Myrland et al., 1981), a small
source rock interval (i.e. top Humber Group), see gas-condensate discovery in Lower Cretaceous sub-
Fig. 13. Mature 'kitchen' areas coincide closely with marine fan sands on the eastern flank of the Viking
the Viking and Central Grabens, and deeper parts of Graben (see Fig. 4). Drainage patterns indicate that
the East Shetland Basin. In the east, source rocks of the source of the gas is the deep Agat Kitchen about
the Volgian-Kimmeridgian Tau Formation reach 20 km to the west, where presumed Humber Group
early oil maturity in the Egersund Sub-basin. Indi- source rocks are currently gas-mature (see Fig. 13).
vidual kitchen areas have been named to facilitate
further discussion of hydrocarbon generation and Troll Kitchen
migration. The giant Troll gas and oil field (Brekke et al., 1981)
and two smaller gas-condensate disceveries in block
35/8 are believed to be directly linked to the large
HYDROCARBON MIGRATION AND gas-mature Troll Kitchen (see Figs. 13 and 14). The
ENTRAPMENT thermal origin of Troll gas is confirmed by carbon
isotopic data (delta 13C = -44.5 per mil). The Troll
The location of known hydrocarbon occurrences with Kitchen is a deep asymmetrical syncline, with a broad
respect to kitchen areas and stratigraphy provides an southeasterly flank (see Fig. 15). Migration of hyd-
indication of likely hydrocarbon migration routes, rocarbons from mature Jurassic source rocks is there-
which may then be confirmed by detailed mapping of fore focused preferentially east and southeast towards
drainage areas on regional seal horizons. Hydrocarbon the Troll area, some 40 km distant. Although Humber
distributions in the North Sea are complex, since Group source rocks are currently gas-mature in the
reservoirs occur at several levels within the Mesozoic undrilled central kitchen area, a significant prior oil
and Tertiary section. In some areas hydrocarbon mig- charge is indicated by a biodegraded heavy-oil rim in
16 Petrolenm Geochemistry in Exploration of the Norwegian She(f

\
\r;J75
\
,
40

57°00'
\
MOST WELLS
DRILLED ON
SALT DOMES.
TEMPERATURE
GRADIENTS IN
BASINAL AREAS
MAY THEREFORE
BE LOWER
\
,,
,
AVERAGE GRADIENTS TO TOP JURASSIC ,
56°00' ~-----'---_ _-'-_ _ _ _--J._,:,\_ _ _.......-:-----J 56°00'
0000' 2°00' 4°00'

Fig. 12. Geothermal gradients (OC/lOO m), Norwegian North Sea.

the Troll structure and evidence of oil-staining within a biodegraded naphthenic residuum, possibly repres-
the present Troll gas cap. It is estimated that oil gen- enting the early trapped oil, which has been diluted by
eration from presumed Kimmeridge Clay Equivalent a fresh paraffinic charge. Westerly migration of hydro-
source rocks deep in the Troll Kitchen would have carbons out of the Troll Kitchen may be limited not
occurred mainly in the early Tertiary, whilst on its only by basin geometry, but also by a major northeast
flanks the same sequence is currently still oil-mature trending fault system (see Fig. 14), and accumulations
(see Fig. 15). This is consistent with the character of on the Tampen Spur are thought to be mainly related
Troll oil (see Fig. 9), which appears to be comprised of to another group of hydrocarbon kitchens.
h Sea
Hydrocarbon Distributions in the Nort 17

CONTOURS: VITRI NITE


REFL ECTA NCE

ber Group.
Fig. 13. Thermal maturity, top Hum

Troll Field. The


Oseberg Kitchen (east flank) into the western compartment of the be charged by
appe ars to
and spill relation- Huldra gas-condensate field
Hydrocarbon migration patt erns , and overspilI
Ose berg area . The Oseberg both the Oseberg and Troll Kitchens Field.
ships are complex in the may occur into the main Ose berg
) is char ged directly from Il area occur in
oil field (Larsen et ai., 1981 to the west. AlI hydrocarbons in the Ose berg -Tro control sug-
Kitc hen only 10-1 5 km linal weIl
the Oseberg
to have fed the Jurassic reservoirs. Deep sync
Overspill of oil from Oseberg app ears Oseberg Kitchen
, whic h could also overspilI gests that the Humber Group in the meridge Clay
e to the east rone Kim
Brage stru ctur consists of relatively thin oil-p
18 Petroleum Geochemistry in Krp/omtion of the Norwegian Shelf'

\. ~1t
MURCHISON':l •
\

J \
,
J ~

J
l,.1\\ 14.4
R

34/7

STRUCTURE
STAT FJORD J \ ~J 35/8
,
t

BRENT
I • J
'J\

t
, +

NORTH

\ -J
J
ALWYN

f /,
+
t
\
t0J
+~J

,
: I:({;WHI LD

+
y' ."'AGE
'OSEBERG
J
10
I
15 20km
!

LEGEND
• OIL ACCUMULATION P PALEOCENE RESERVOIR
<fiJ GAS ACCUM ULATION JURASSIC RESERVOIR
U MIGRATION PATH R TRIASSIC RESERVOIR
JIf OVERSPILL

Fig. 14. Upper Jurassic hydrocarbon kitchens and migration paths, northern Viking Graben.

Equivalent «100 m) overlying a thick and mainly found. This charge is consistent with the highly-
gas-prone Heather Formation. Observed hydrocarbon mature Oseberg Kitchen, and is unlikely to originate
distributions and types on the flanks of the Oseberg from the less-mature kitchens to the west. Overspill
Kitchen are consistent with a present-day charge from from North Alwyn has charged the giant Brent struc-
upper Humber Group source rocks which are in a ture (Bowen, 1975), which may also benefit from wes-
late-oil and gas-condensate stage of generation (see terly drainage out of the Ninian oil kitchen (see
Fig. 13). A contribution from Liassic Dunlin Group Fig. 13). The gas cap and high GOR oils in Brent
oil-prone source rocks which occur in this area cannot (unusual amongst oilfields of the East Shetland Basin)
be excluded, but a major gas charge from Brent and appear again to reflect a significant charge contribu-
Statfjord coals is not apparent, otherwise widespread tion from the Oseberg Kitchen. The North
displacement of the oil charge would have occurred. Alwyn-Brent overspill chain continues back into
Oseberg Kitchen (west flank)
:.r orwegian waters to the Statfjord Field (Kirk, 1980).
It is significant that the giant Brent and Statfjord
Hydrocarbons appear to have migrated out of the fields lie in an area where Jurassic source rocks are
Oseberg Kitchen into a number of fields in UK waters immature, and that these structures must rely
on the west flank of the Viking Graben (see Fig. 14). entirely on lateral migration and overspill for their
Several accumulations in the Alwyn complex are charge. North of Statfjord, a chain of small oil fields
well-situated to be charged by the Oseberg Kitchen, centred on Murchison appears to be charged from the
although the oil-mature Alwyn and Ninian Kitchens to northerly World's End Kitchen (see Fig. 14).
the west also drain into the same structural trend. Structural studies indicate that the recent 34/4-4 oil
Confirmation of a major charge from the east is seen in discovery is on the northern extent of a very large
North Alwyn, where high GOR oil and a gas cap are rotated J urassic-Triassic fault block which lies mainly
Hydrocarbon Distributions in the North Sea 19

NW

U.K NORWAY

34/7
*
TROll
SE

JU
JML
R-D

L-________________________________ ~~ _____________________________________________________________L8km

EIS
• * * • NIW
W E

BERYL U.K. N06WAY FRIGG NE FRIGG IDUN

L-__________________________________________________ ~~ __________________________"8km ®


W

*
E
BRAE U.K. NORWAY GUDRUN

lEG EN D
[2] TERTIARY (UNDIFFERENTIATED)

~ LOWER EOCENE

El
o
PALEOCENE
CRETACEOUS
~ UPPER JURASSIC
~ MIDDLE AND LOWER JURASSIC
IT] TRIASSIC

~ TRIASSIC-DEVONIAN
8 PERMIAN-DEVONIAN
[Q3
@
PERMIAN SALT

~ __ ~1L ________________________________________ ~8km


D BASEMENT

W

FUlMAR U.K. NORWAY

EKOFISK
E

Fig. 15. Regional cross-sections, North Sea (Section 4, after Ziegler, 1982). For locations, see Fig. 1.
20 Petroleum Geochemistry in Exploration of the Norwegian Shelf

2°00'

60"30'

BRESSAY~P


25/2 -5

1
J

,P BERYL
~P J

CRAWFORD

5 9 " C ( ) ' ' - - - - - - - ; : ! = - - - - - - - : - ' : ; c , - , ; - - - - - - 2 : ' : : - O c - , - - - - - - - - - - - - ' 59°00'

LEGEND
• OIL ACCUMULATION E EOCENE RESERVOIR
ell GAS ACCUMULATION P PALEOCENE RESERVOIR
J JURASSIC RESERVOIR

Fig. 16. Upper Jurassic hydrocarbon Kitchens and major fields, Central Graben.

in block 34/7. Hydrocarbon charge into this structure 15). The small, subsidiary Beryl Kitchen is also evi-
can be envisaged by multiple routes out ofthe World's dent in UK waters.
End and Pancake Kitchens, and by overspill from the
North Alwyn-Brent-Statfjord chain. Overspill from Frigg Kitchen
the 34/7 structure * will feed the Gullfaks oil field The deep Frigg Kitchen, defined at Jurassic levels
(Skarpnes et al., 1982), which is on the structural cul- (see Fig. 13), underlies an area where gas fields are
mination of the Tampen Spur, although other kitchens predominant and accumulations occur mainly in Terti-
to the east and south may also contribute. ary (Eocene) deepwater sands.
The thermal maturity of Humber Group source
Central Viking Graben (59°00' to 60 30'N) 0
rocks in the Frigg Kitchen is mainly above
Two major kitchen areas are recognized in the central VR/E = 1.0, where oil generation is largely complete
(assuming suitable source type) and gas generation is
Viking Graben area: the Frigg Kitchen, centred on the
predominant, which is confirmed by the Eocene Frigg
Frigg gas field, and the Balder Kitchen, to the
immediate west of the Balder oil field (see Figs. 13 and Gas Field and its satellites (see Figs. 15 and 16). Oil has
also clearly been generated in the Frigg Kitchen, as
indicated by the heavy (biodegraded) oil at Frigg and
'Recently named the Snorre Field. East Frigg (Heritier et al., 1980). Light oil was reco-
Hydrocarbon Distribut?:ons in the North Sea 21

vered from Jurassic reservoirs on the eastern flank of Kitchen (see Figs. 13 and 16). The effective hydrocar-
the Frigg Kitchen at Idun and 25/2-5, where Humber bon charge from this area is predominantly oil, as is
Group source rocks are still within the oil window. indicated by the Paleocene Balder oil field (Sarg and
In the greater Frigg area, vertical migration of Skjold, 1982) and several Jurassic discoveries in UK
hydrocarbons from Jurassic source rocks has clearly waters. Oil in the Jurassic at Heimdal may be from
been dominant, possibly facilitated by microfracturing either the Balder or Frigg Kitchens, although gas in
of overlying abnormally pressured Cretaceous shales Paleocene sands suggests that there is a significant
(Goff, 1983). The Eocene umbrella closure of the grea- contribution from the Frigg gas kitchen.
ter Frigg structure is largely coincident with the
Beryl Kitchen
gas-mature Jurassic Frigg Kitchen. In the west, both
vertical and lateral migration paths out of the Frigg The small Beryl Kitchen, entirely within UK waters,
Kitchen are indicated by a number of Paleocene heavy appears to have charged a cluster of Jurassic and
oil fields (e.g. Bressay) which occur in UK waters on Paleocene oil fields in the Beryl-Bruce area (see
the East Shetland Platform. Figs. 13 and 16).
Apart from maturity considerations, trap timing
may also playa role in the apparent gas-prone nature Southern Viking Graben (58° to 59°N)
of the Frigg area. Since many of the shallow produc-
The southern Viking Graben differs significantly in
tive structures were not sealed until at least late
structure and stratigraphy from its central and north-
Eocene times, thermal modelling studies suggest
ern counterparts, and these differences have direct
that these traps largely postdate the early oil charge
implications on hydrocarbon sources, migration paths
from the Frigg Kitchen. Their content therefore and distributions. The southern Viking Graben is nar-
reflects the later (and current) charge where gas row and deep, and at Triassic and Jurassic levels is
predominates.
highly asymmetrical in structural profile (see regional
cross-section 3, Fig. 15). At Middle Jurassic levels
Balder Kitchen
structure takes the form of a half-graben, with the
A thick sequence of Humber Group source rocks is deepest part of the basin in the west adjacent to the
mature for oil and gas generation within the Balder basin-bounding fault system. At top Jurassic level, the

TOP HUMBER GROUP BASE HUMBER GROUP


f
f
I
+

BALDER'

@I"'" @I"'"

DAGNY r\)

l""""
TIFFANY 58°30' TIFFANY

TONI TONI
f!)PdJ THELMA

,
BALMORAL BALMORAL
,
-
\

'\MAUREEN
\

ANDREW ANDREW'
MABEL ,
\
1°00' 2°0C'

• OIL ACCUMULATION VITRINITE REFL ECTANCE CONTOURS.


® GAS ACCUMULATION FOR CROSS- SECTION 3. SEE FIG. 15.

10 20 30 40 5O'm
I ! !
I

Fig. 17. Thermal maturity, top and base Humber Group, southern Viking Graben.
22 Petroleum Geochemistry in Exploration of the Norwegian Shelf

UPPER JURASSIC MIDDLE JURASSIC


2'00

5B"4~

EASTERLY
OIL CHARGE
NOT TRAPPED
DUE TO
DISPLACEMENT
58°30
BY COMPETING 58 0 30'
MIDDLE JURASSIC
GAS CHARGE

o ~ IOkm
~

• Oil ACCUMULATION ill OVERSPH.L


e GAS ACCUMULATION P PALEOCENE RESERVOIR
{7 MIGRATION PATH J JURASSIC RESERVOIR

Fig. 18. Hydrocarbon kitchens and migration paths for the Upper and Middle Jurassic, southern Viking Graben.

basin is broadly symmetrical with the axis approxi- Group source sequence reflecting the structural com-
mately following the UKjN orway median line. plexity already discussed, and certain fields can be
In the UK sector, several oil fields (Brae group, attributed to shallow 'Upper Jurassic' drainage pat-
Tiffany, Toni, Thelma) are found in Jurassic fault terns, whilst others are related to deep 'Middle
scarp sands along a major western fault system Jurassic' systems (see Fig. 18).
(Harms et al., 1981), see Figs. 17 and 18. Distally- Westerly migration out of the Upper Jurassic oil
equivalent turbidites are also oil-productive in UK kitchen is confirmed by the Brae-type fields in UK
well 16j8-1 and in Norwegian waters at Gudrun. The waters which lie immediately adjacent to rich, mature
majority of hydrocarbons found in Norway, however, Kimmeridge source rocks (Reitsema, 1983). Oil-
are gas-condensate, commonly in Middle Jurassic bearing Upper Jurassic turbidite sands at Gudrun lie
reservoirs along the eastern flank of the basin, but within a currently-generative source sequence.
also in Paleocene sands in the Sleipner area (Pegrum Although a degree of easterly migration out of the
and Ljones, 1984). This asymmetrical distribution of Upper Jurassic syncline is implied by the Gudrun oil
hydrocarbons along the flanks of the Viking Graben, sands, comparable oil accumulations to those in the
with British fields being largely oil-productive and UK sector have not been found in Norwegian waters.
those in Norway gas-prone, is a direct consequence of As an easterly oil charge out of the generative Upper
the structural asymmetry of the basin at Middle Jurassic section is considered highly likely, the
Jurassic and older levels. absence of oil fields on the east flank of the basin is
The deep Gudrun discovery well 15j3-1 penetrated a attributed to displacement by a competitive gas
Humber Group turbidite sequence over 1000 m thick, charge from a deep 'Middle Jurassic' kitchen (see
of which 580 m is considered net source rock. Figs. 18 and 19).
Geochemical studies indicate that much of this is rich Thermal maturity patterns at the base of the
and oil-prone. Seismic data suggest that the Humber Humber Group reflect the structural asymmetry of
Group may exceed 2000 m in thickness adjacent to the the basin at Middle Jurassic levels (see Fig. 17). A
western basin-bounding fault system. As a result, major gas kitchen occurs in the west, largely in the
upper levels of the Humber Group are currently oil- UK sector, and migration paths are mainly east and
mature, whilst deeper in the section, similar source southeast. A major element of the Middle Jurassic gas
rocks are already gas-generative (see Fig. 17). Migra- charge is focused on a chain of Jurassic fault-traps in
tion paths also differ with depth within the Humber Norwegian waters leading through Dagny to the
Hydrocarbon Distributions in the North Sea 23

®
PALEOCENE BRAE AREA SLEIPNER AREA

• •
HEAVY OIL

o Mainly Shales
( Cretaceous - Tertiary) D Reservoirs
ElK imrneridge Clay 5l Palaeozoic/Basement
DIAGRAMMATIC ONLY: NOT TO SCALE Source Rocks .... Oil Migration Path
~ Heather Fonnation 0{> Gas Migration Path

Fig. 19. Schematic cross-section, southern Viking Graben, showing kitchen areas and migration paths.

Sleipner complex (Larsen and Jaarvik, 1981; Fager- primary pattern of the North Sea rift system is heav-
land, 1984). All of these structures contain gas- ily modified with diapiric salt tectonics. As in the Vik-
condensate (see Fig. 18). Gas is also found in the ing Graben, the principal source unit is the Humber
Paleocene at Gamma and Theta, which is attributed to Group, which exceeds 1000 m in thickness in basinal
vertical communication with the Jurassic migration areas. In the Norwegian sector, fields are commonly
route, possibly via a major fault along the western in fractured Danian-Maastrichtian chalk reservoirs
margin of the Gamma feature (Fagerland, 1984). (van den Bark and Thomas, 1980; Watts, 1983) which
In the Gudrun area, Middle Jurassic migration have been arched over deep intrusive Zechstein salt
paths are apparently radial, resulting in gas in deep bodies (see Fig. 15). Paleocene sands are productive,
reservoirs in UK 16/3a-l, UK 16/8-1 and Gudrun. The particularly in UK waters (Carman and Young, 1981),
occurrence of high-GOR oil in North Brae (in contrast and some hydrocarbons are also found in Jurassic and
with Central and South Brae) is possible evidence of a older reservoirs (Bailey et al., 1981). Mapping ther-
contribution from deeper highly mature source rocks, mal maturity at top Humber Group level (see Fig. 13),
due to an element of easterly dip at Middle Jurassic almost every field in the Central Graben is contained
levels in this part of the kitchen. Similarly, Toni and within a fairway defined by the VR/E = 0.6 contour,
Thelma, with Upper Jurassic reservoirs containing i.e. fields appear to be restricted closely to locations
high GOR oils, lie in another area where the Middle within or above the currently-generative Jurassic
Jurassic basin has a small western flank. The basin (see Fig. 20). Since in the Norwegian Sector
Maureen, Mabel and Andrew oil fields in Paleocene fields lie mainly in Cretaceous reservoirs, a strong
sands are thought to be related to other migration element of vertical migration is implied with little
paths out of the southerly Andrew Kitchen (see apparent lateral migration. Assuming an oil-floor of
Fig. 13). VR/E = 1.2, the central trend of the Cod-Albusk-
A contribution to hydrocarbon accumulations in the jell-Ekofisk gas kitchen is clearly defined. Almost
southern Viking Graben from pre-Humber Group every gas field in the Central Graben is either within
source rocks, particularly coals of the Vestland Group or immediately adjacent to the limits of a gas-mature
and Statfjord Formation, cannot be excluded. How- Jurassic kitchen, again confirming the dominance of
ever, the wet gas accumulations of the Sleipner area vertical migration systems in this area.
are most compatible with an origin from post-mature
oil-prone source rocks, such as are found in the lower Egersund Sub-basin
Humber Group in this area.
Three small oil discoveries have been made in the
Central Graben (56° to 58°N) Egersund Sub-basin: Bream, Brisling and 18/10-1 (see
Fig. 4). Reservoir sands are within the Middle
The Central Graben is structurally complex, since the Jurassic Vestland Group, and trapping structures are
24 Petroleum Geochemistry in Exploration of the Norwegian Shelf

1°00' 2°00' 3°00'


58°00' r----~--------....-r__-""'------....,..---------_, 58°00'

-
FORTIES e P

\
\
\
\

.
\
P LOMOND
\
\ o
'6'
\
OIL KITCHEN P/J\
\ J
\
\p

57 0 00'
\
\ 0>01

~~\

K. ."'~\\'"
OIL KITCHEN

OIL KITCHEN

KITCHEN
• J

-K
K
(> K
JOANNE .K \ K

JOSEPHINE. ALBUSKJEL" TOR, ,K


J 1I'K 'K
J • K , EKOFISK
FUlMAR.J \@
P J \
e~lYDE
r
• \
\@K K
~ ElDFISK
AUK \
10
,
15 20km
!
\VAlHAll' ~",J>." ..... x
\ ~x;;'J>.~'f.
CONTOUR VALUES VRIE J.\ HOD.K x ...... Q~'"
TOP HUMBER GROUP \
\
'r/
\
56°00' .......- __ --:=::-________-:--;-________~___;_--\~-----~ 56000'
10 00' 2°00' 3°00'
LEGEND
• OIL ACCUMULATION K CRETACEOUS RESERVOIR
~ GAS ACCUMULATION J JURASSIC RESERVOIR
01 OLIGOCENE RESERVOIR 11 TRIASSIC RESERVOIR
PALEOCENE RESERVOIR Pe PERMIAN RESERVOIR

Fig. 20, Upper Jurassic hydrocarbon kitchens and major fields, Central Graben.

halokinetic anticlinal features in an area of major salt apparent oil- and gas-prone areas in the North Sea
diapirism. Rich, oil-prone source rocks occur in the since, in general, Kimmeridge source rocks are pres-
Volgian-Kimmeridgian Tau Formation, which is early ent in their richest, most oil-prone facies in deep,
mature for oil generation (VR/E = 0.6 to 0.8) over mature kitchen areas of the Central and Viking Gra-
much of the Egersund Sub-basin (see Fig. 13). The bens. Other factors, such as thermal maturity of
three oil-bearing structures lie immediately adjacent kitchen areas, timing relationships, migration paths
to an arm of this Bream oil kitchen. and charge competition are considered to be more sig-
nificant. The presence of excess charge in several
areas has resulted in structures which are full to
CONCLUSIONS spill-point, and groups of fields can often be related to
one another and a common kitchen as a continuous
Hydrocarbon occurrences in the North Sea graben fill-and-overspill sequence.
system can largely be explained in terms of the dis-
tribution and thermal maturity of source rocks within
the Kimmeridge Clay Equivalent (Oxfordian-Ryaza- ACKNOWLEDGEME NTS
nian). Other source units, e.g. Dunlin Group, Brent
Group and Heather Formation, may also make signif- The authors wish to acknowledge the assistance of
icant charge contributions locally. Although variations Statoil, Norsk Hydro, Amoco Norway, Conoco Nor-
in organic facies occur within the dominant Kim- way, Phillips Norway and the Norwegian Petroleum
meridge source interval, these are thought to be, at Directorate, who supplied oils and core material for
best, only a secondary influence on the distribution of analysis. Most of the oil and extract analysis were the
Hydrocarbon Distributions in the North Sea 25

work of geochemists at Koninklijke/Shell Exploratie Field complex. In: Norweqian Symp. on Exploratioll (NSE-81) ,
en Produktie Laboratorium, the Netherlands; Rock Bergen. Norwegian Petroleum Society.
Larsen, V., Aasheim, S. M. and Masset. J. M. 1981. 30j6-Alpha
Eval analyses were performed by Robertson structure: a field case study in the silver block. In: Norweqian
Research International, Wales and IKU, Norway. SYllip. un Exploration (NSE-S1) , Bergen. Norwegian Petroleum
Partners in production licence 054 kindly approved Societv.
release of confidential data from the Troll Field. For Lijmbach, G. W. M., van del' Veen, F. M. and Engelhardt, E. D.
1983. Geochemical characterisation of crude oils and source rocks
helpful discussions and encouragement we thankJ. R. using field ionisation mass spectrometry. In: Bjorf/y, M. et al.
Freake, R. J. Murris, P. A. Ziegler, J. W. A. (eels.), r1dvances in Orqa.nic Geochemistry 19111. John Wiley and
Bodenhausen, J. M. A. Buiskool-Toxopeus and E. Tal- Sons, New York, 880 pp.
leraas. This paper is published with the permission of Lopatin, N. V. 1971. Temperature and geologic time as factors in
A/S Norske Shell, Forus, Norway. Coalification. Akad. Nauk. SSSR Izv. Ser. Geol., 3, 95-106 (in
Russian).
Meissner, F. F. 1984. Petroleum geology of the Bakken Formation,
Williston Basin, North Dakota and Montana. In: Demaison, G.
REFERENCES and Murris, R. J. (eds.), PetroleulII Geuchelllistl'Y and Basill
Evalnation. AAPG Memoir, 35, 159-179.
Bailey, C. C., Price, 1. and Spencer, A. M. 1981. The Ula oil field, Meyer, B. L., and N ederlof, M. H. 1984. The identification of source
Block 7/12, Norway. In: Norwegian Symp. on Exploration rocks on wireline logs by density/resistivity cross plots. AAPG
(NSE-81), Bergen. Norwegian Petroleum Society. Bull., 68(2), 121-129.
Barnard, P. C. and Cooper, B. S. 1981. Oils ami source rocks of the Myrland, R., Messel, K. and Raestad, N. 1981. Gas discovery in
North Sea area. In: Illing, L. V. and Hobson, G. D. (eds.), Pet- 35/3: an exploration history. In: Norwegian SYllip. all E:rplora-
rolemn Geology of the Continental Shelf of Northwest Europe. tion (NSE-81) , Bergen. Norwegian Petroleum Society.
Institute of Petroleum, London, 521 pp. Nederlof, M. H. 1979. The use of habitat of oil models in exploration
Bjorlykke, K., Dypvik, H. and Finstad, K. G. 1975. The Kim- prospect appraisal. Proc. 10th Wurld Petroleulli Congr., Buchar-
meridge Shale, its composition and radioactivity. In: Finstad, K. est, September 1979.
G. and Selley, R. C. (co-ordinators), Proc. Jura.oSsic Northern Pegrum, R. M. and Ljones, T. E. 1984. 15/9 Gamma gas field off-
North Sea SYll1p. 1975. Norwegian Petroleum Society. shore .'Iorway, new trap type for North Sea Basin \\~th regional
Bowen, J. M. 1975. The Brent oil-field. In: Woodland, A. W. (eel.), structural implications. AAPG Bltil., 68(7), 874-902.
PetroleulII and the COlltinental Shelf of North West Europe, Phillippi, G. T. 1965. On the depth, time and mechanism of pet-
Vol. 1, Geology. Applied Science Publishers, Barking, 501 pp. roleum generation. Geochim. Cnsillochilli. Acta, 29, 1021-1049.
Brekke, T., Pegrum, R. M. and Watts, P. B. 1981. First exploration Pusey, W. C. 1973. The ESR-kerogen method, a new technique of
results in Block 31/2, offshore Norway. In: Norwegian Symp. Oil estimating the organic maturity of sedimentary rocks. petroleum
Exploration (NSE -81), Bergen . .'Iorwegian Petroleum Society. Tillles, 12 January.
Buiskool-Toxopeus, J. M. A. 1983. Selection criteria for the use of Reitsema, R. H. 1983. Geochemistry of the North anel South Brae
vitrinite reflectance as a maturity tool in petroleum exploration. areas, Korth Sea. In: Brooks, J. (ed.), Petroleum Geochemistry
In: Brooks, J. (ed.), Petro leu In Geochelliistry and E.rplomtion of and Exploration of Europe. Blackwell Scientific Publications,
Europe. Blackwell Scientific Publications, Oxford, 379 pp. Oxford, 379 pp.
Carman, G. J. and Young, R. 1981. Reservoir geology ofthe Forties Robertson Research International Limited 1979. Southern Offshore
oil-field. In: Illing, L. V. and Hobson, G. D. (eds.), PetroleulII 01 orway: the stratigraphy and petroleum geochemistry of the
Geology of the Co-ntinental Shelf of Northwest Europe. Institute Jurassic to Tertiary sediments. Unpublished multi-client report.
of Petroleum, London, 521 pp. Robertson Research International Limited 1981. Southern Offshore
Clementz, D. M. 1979. Effect of oil and bitumen saturation on Norway Phase II Study: the stratigraphy and petroleum
source rock pyrolysis. AAPG Rull., 63(12), 2227-2232. geochemistry of the Jurassic to Tertiary sediments. U npublishecl
Demaison, G. 1984. The generative basin concept. In: Demaison, G. multi-client report.
and Murris, R. J. (eds.) Petroleum Geochemistry and Basill Sarg, J. F. and Skjold, L. J. 1982. Stratigraphic traps in Paleocene
Development. AAPG Memoi1', 35, 1-14. sands in the Balder area, North Sea. In: Halbouty, M. T. (ed.),
Demaison, G. J. and .'vIoore, G. T. 1980. Anoxic environments and The Deliberate Search for the Subtle Trap. AAPG Menwil', 32,
oil source bed genesis. AAPG Rull., 64(8), 1179-1209. 197-206.
D.I'pvik, H., Ruesli'ttten, H. G. and Thronclsen, T. 1979. Composi- Schlumberger 1979. Log interpretation charts, 1979 Edition.
tion of organic matter from North Atlantic Kimmeridgian Shales. Schlumberger 1982. Natural gamma ray spectrometry, essentials of
AAPG Bull., 63(12), 2222-2226. N. G. S. interpretation.
Espitalie, J., Laporte, J. L., Madec, M., Marquis, F., Leplat, P. and Schmoker, J. W. 1979. Determination of organic content of
Boutefeu, A. 1977. Methode rapide de caracterisation des roches Appalachian Devonian shales from formation-density logs. AAPG
meres, de leur potential petrolier et de leur degre d'evolution. Bull., 63(9), 1504-1537.
Rev. Illst. Fr-. Petrol. 32,23-42. Schmoker, J. W. 1981. Determination of organic-matter content of
Fagerland, N. 1984. A rift-related stratigraphic trap. J. Petr-ol. Appalachian Devonian shales from gamma-ray logs. AAPG Bull.,
Geol., 7, 227-236. 65(7), 1285-1298.
Goff, J. C. 1983. Hydrocarbon generation and migration from Skarpnes, 0., Briseid, E. and Milton, D. 1. 1982. The 34/10 Delta
Jurassic source rocks in the East Shetland Basin and Viking Gra- prospect of the K orwegian North Sea: exploration study of an
ben of the northern North Sea. J. Geol. Soc. London, 140, unconformity trap. In: Halbouty, M. T. (ed.), The Del-iberale
445-474. Search for- the Subtle Trap. AAPG Memoir, 32, 207-216.
Gussow, W. C. 1954. Differential entrapment of oil and gas; a fun- Taylor, J. C. M. 1983. Bit-metamorphism, illustrated by lithological
damental principle. AAPG Bull., 38(5), 816-853. data from German North Sea wells. Geol. JI1)nbouw 62, 211-219.
Harms, ,J. C., Tackenberg, P., Pickles, E. and Pollock, R. E. 1981. Tissot, B. P. and Roucache, ,J. 1980. Principles of a genetic classifi-
The Brae oil field area. In: Illing, L. V. and Hobson, G. D. (eds.), cation of crude oils. In: Miall, A. D. (ed.), Facts and Principles of
PetroleulI! Geology of the COlltinental ShelfofNorthwest Europe. World Pelroleum Occul'rwce. Call. Soc. Petr-ol. Geol. Mellioir, 6,
I nstitute of Petroleum, London, 521 pp. 1003 pp.
Harwood, R. J. 1977. Oil and gas generation by laboratory pyrolysis Tissot, B., Durand, B., Espitalie, J. anel Combaz, A. 1974. Influ-
of kerogen. AAPG Bull., 61(12), 2082-2102. ence of the nature and diagenesis of organic matter in formation
Heritier, - F. E., Lossel, P. and Wathne, E. 1980. Frigg of petroleum. AAPG Bull., 58, 499-506.
Field-large submarine-fan trap in Lower Eocene rocks of the van den Bark, E. and Thomas, O. D. 1980. Ekofisk: first of the giant
Viking Graben, North Sea. In: Halbouty, M. T. (ed.), Giant Oil oil fields in Western Europe. In: Halbouty, M. T. (ed.), Giant Oil
alld Gas Fields of the Decade 1968-1978. AAPG Memoir, 30, and Gas Fields of the Decade 19611-1!1711. AAPG Memoir, 30,
59-79. 195-224.
Hunt, J. M. 1979. PetTaleum Geochelilistry and Geology. W. H. Vollset,J. and Dore, A. G. (eds)(1984). ArevisedTriassicandJurassic
Freeman and Company, San Francisco, 617 pp. lithostratigraphic nomenclature for the Norwegian North Sea.
Kirk, R. H. 1980. Statfjord-a North Sea giant. in: Halbouty, M. NPD Hill!. 3.
T. (ed.), Giant Oil and Ga~fields of the Decade 1968-1978. AAPG Watts, N. L. 1983. Microfractures in chalks of Albuskjell field,
Jfeilloir, 30, 95-116. Norwegian sector, North Sea: possible origin anel distribution.
Landes, K. K. 1967. Eometamorphism and oil and gas in time and AAPG Ruli., 67(2), 201-234.
space. AAPa Bull., 51(6), 828-841. White, D. 1915. Some relations in origin between coal and pet-
Larsen, R. Yl. and Jaarvik, L. J. 1981. The geology of the Sleipner roleum. Wash. Acad. Sci. J., 5(6), 189-212.
26 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Ziegler, P. A. 19S0. Northwest European Basin: Geology and hyd- Geology q(the CUlllillelllal Shelf of" ,Vortli west Eumpe. Institute
rocarbon provinces. In: :\1iall, A. D. (eel.), Facts ami Plillciples uf of Petroleum, London, 521 pp.
Wodd PelmleulII OccU1~·ellce. Call. Soc. Petml. Geul. :fJemoir, 6, Ziegler, P. A. 1982. Geological Allao of' We.,tem alld Gelllmi
1003 pp. Europe. Shell Internationale Petroleum :Vlaatschappij B. V.,
Ziegler, P. A. 1981. Evolution of sedimentary basisn in N orth-West 130 pp. + enclosures.
Europe. In: Illing, L. V. and Hobson, G. D. (eds.), PetroleuiJI
2 Correlation of the offshore
sequences referred to the
Kimmeridge Clay
Formation-relevance to
the Norwegian sector
A. G. Dare, * J. Vollsett and G. P. Harnar+
*Conoco Norway Inc., Stavanger, Norway
tNorwegian Petroleum Directorate, Stavanger, Norway
+statoil, Stavanger, Norway

The lithologies referred to the Kimmeridge Clay Formation are acknowledged to include the most important
petroleum source rocks known in the North Sea. However, the correlations and historical review presented herein
demonstrate that widespread use of the term 'Kimmeridge Clay Formation' is a misleading piece of stratigraphic
nomenclature which stems from conflicting definitions of the unit both offshore and onshore. This has led to
striking differences in the way the unit is perceived in different basins of the North Sea area.
The base of a unit is regarded as a fundamental criterion in stratigraphy, and yet several quite different
definitions for the base of the Kimmeridge Clay Formation exist in the literature. In eastern England the base is
placed at a chronostratigraphic marker (the Oxfordian-Kimmeridgian boundary) even when present within a
homogeneous claystone sequence. Immediately offshore this entire claystone sequence is included in the
Kimmeridge Clay and an older, but more Iithostratigraphically significant, horizon (the contact with Corallian
limestones) is taken as the base. In total contrast, the base in the northern North Sea is viewed as the contact
between bituminous claystones and underlying claystones of more normal character. The basic criteria on which
the northern Kimmeridge Clay Formation was defined (anomalously high organic content and associated log
responses) were never documented in the offshore type and reference sections of the southern UK North Sea.
Similarly, bituminous shale developments are not typical of the Kimmeridge Clay as defined in the Netherlands
sector.
Adoption of a local formation nomenclature for the Norwegian sector is urged as a means to more-precise
correlation. A scheme proposed by the Norwegian Lithostratigraphic Nomenclature Committee is outlined. This
classification is based on the recognition that the economically important organic shales are discrete and readily
mappable units, but that they are not always synchronous or continuous between basins. A review of the role of the
Kimmeridge Clay Formation in other national sectors is also suggested.

INTRODUCTION previous attempts at offshore lithostratigraphic classi-


fication. A decision was therefore taken to replace
In the recent work of the Norwegian Jurassic Litho- the existing nomenclature with a number of locally-
stratigraphic Nomenclature Committee (1980-1984) defined units, and these proposals were published by
much time was devoted to assessing the place of the Vollset and Dore (1984).
Upper Jurassic-lowermost Cretaceous Kimmeridge Unfortunately, the term 'Kimmeridge Clay Forma-
Clay Formation in the Norwegian Lithostratigraphic tion' is strongly entrenched in the literature and par-
scheme. During the course of this work it became lance of North Sea explorationists, in all involved
clear that the term was no longer easy to apply in the national areas, with the possible exception of the Dan-
Norwegian sector. The unit previously defined as the ish sector. Normally this consideration alone would be
Kimmeridge Clay Formation was considered to grounds for extreme caution in modifying the nomen-
be too broad, to contain too many facies types, and to clature. However, it is precisely this entrenchment
be of limited use in correlation or in the understanding which leads to misunderstanding and problems of cor-
of basin evolution. These problems were compounded relation; although the term is widely used, there is no
by imprecise and variable definitions of the unit in consensus as to its meaning. The problem is essen-
Petroleum Geochemistry in Exploration of the Nonvegian She(f © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 27-37.
28 Petroleum Geochemistry in Exploration of the Norwegian Shelf

tially a linguistic one. To illustrate this point, some The Ampthill Clay extends as far south as Oxford-
popularly held beliefs (not necessarily published, but shire. In the southern part of its outcrop there is a
frequently heard) on the Kimmeridge Clay Formation non-sequence between the Ampthill and Kimmeridge
are listed below. Clays, commonly expressed by a phosphate nodule
The Kimmeridge Clay is variously thought to com- band (Arkell, 1933). Elsewhere, however such as on
prise the following. Humberside, there is very little or no hiatus. Key sec-
tions at South Ferriby (Kelly and Rawson, 1983,
(i) All shales of late Jurassic age, regardless of log
p. 66) and Worlaby (Richardson, 1979, p. 13) show the
character (particUlarly in the Norway-UK Central Ampthill Clay-Kimmeridge Clay transition to be one
Graben).
of continuous deposition, with no perceptible lithologi-
(ii) A sequence of bituminous shales distinguished
cal difference above and below the 'contact'. In fact,
by very high gamma-ray readings and anomalously the facies of the basal Kimmeridge Clay and the
low seismic velocities (widely accepted in areas north
underlying Ampthill Clay-dark, calcareous, shelly,
of the Mid-North Sea High). non-bituminous claystone-are everywhere remark-
(iii) A sequence of shales defined by the particular ably similar. In boreholes the 'contact' has almost
class of kerogen found in its organic fraction. invariably been determined on palaeontological
(iv) Shales of Kimmeridgian age. Surprisingly, grounds alone (e.g. Arkell, 1933, p. 418; Wright, 1980,
terms such as 'Kimmeridgian shale' still occur in the
p. 74). The base of the Kimmeridge Clay Formation
literature, where clearly referring to sequences of occurs at the base of the lowermost (baylei) zone ofthe
much broader chronostratigraphic range, or even to Kimmeridgian stage, but this marker has little signifi-
non-Kimmeridgian sequences. This probably results cance in lithostratigraphic terms. Brookfield (1973,
from the fact that the Kimmeridge Clay Formation of
p. 141) writes: 'Despite contrary opinions ... there is
the English mainland is equated with the Kimmerid-
no major lithological or faunal break between the
gian stage sensu anglico (see the following section).
Oxfordian and Kimmeridgian stages in Britain'.
Brookfield (1973), in referring to onshore British The Kimmeridge and Ampthill Clays thin north-
stratigraphy, wrote: 'The lithostratigraphic terminol- wards towards the Market Weighton High (Fig. 1).
ogy of the Oxfordian and Kimmeridgian is truly a North of the high in the Yorkshire Basin boreholes
mess' (p. 137). In this brief account we attempt to have proved a thick Kimmeridge Clay sequence and
show that some of these problems have been transfer- limited exposures indicate, for the uppermost zones at
red to the offshore realm. In examining these problems least, strong similarities to the Dorset type area
we will expand upon our proposal to reclassify locally (Cope, 1974). The Kimmeridge Clay was long consi-
the Upper Jurassic-lowermost Cretaceous shale dered to rest directly on Corallian beds in this area,
units, which include the richest source rocks known in but recent boreholes have proven the existence of
the North Sea, and will draw attention to some of the intervening Ampthill Clay. Once again, however, this
conceptual problems that exist in Norway and in other identification appears to be based solely on late Oxfor-
national sectors. dian datings and not on any lithostratigraphic crite-
rion (I.G.S., 1974).

THE 'SPREAD' OF THE KIMMERIDGE Southern North Sea, UK sector


CLAY FORMATION FROM ITS TYPE AREA The first application of the name Kimmeridge Clay
Formation to the offshore realm was by Rhys (1974) in
Onshore England the recommendations of a joint Oil Industry-I.G.S.
The name 'Kimmeridge Clay' has been in use since the committee on North Sea nomenclature. The formation
dawn of British stratigraphy in the early 19th Century was defined as the uppermost unit of the Humber
(e.g. Fitton, 1836). The type area for the Kimmeridge Group, a primarily claystone sequence of Oxfor-
Clay Formation is Dorset on the south coast of Eng- dian-Kimmeridgian age found in the Southern North
land (e.g. Arkell, 1947), where the unit consists of Sea Basin. Extension of the name offshore was
dark claystones with numerous limestone/dolomite intended to express correspondence with the deposits
stringers and phosphatic horizons. The lowermost documented in Norfolk, Lincolnshire and Yorkshire,
part of the unit is developed in a calcareous, shelly and hence with the Dorset type area. The type well for
facies but higher in the sequence bituminous clays and the offshore Kimmeridge Clay was 47/15-1 (Phillips)
oil shales are frequent. The Kimmeridge Clay of the with 47/9-1 (Gulf) as a reference section. Both wells
English mainland is essentially synonymous with the are approximately 65 km east of the Humber estuary.
Kimmeridgian stage sensu anglico (e.g. Arkell, 1933; Well 48/22-2 (Burmah) also serves to illustrate the
Cox and Gallois, 1981; Cope, 1980, fig. 14), or Kim- Kimmeridge Clay Formation and lies about 65 km
meridgian-early Middle Volgian in Boreal terminol- ENE of the Wash (Fig. 1).
ogy. The Humber Group as defined by Rhys (1974) also
Over much of its northeastern outcrop in England includes the Corallian and the underlying Oxford Clay
(Fig. 1) the Kimmeridge Clay overlies more clays- Formation. In the type well the Kimmeridge Clay
tones of Late Oxfordian age, the Ampthill Clay. The Formation is only 34 m thick, and is shown to rest
latter unit in turn overlies, and is in part equivalent directly on the Corallian. The comparatively minor
to, limestones and calcareous sandstones of the Coral- thickness of Kimmeridge Clay may indicate contem-
lian Formation. The relationship between the Kim- poraneous thinning towards an offshore extension of
meridge Clay and Ampthill Clay is generally obscure the Market Weighton High (Fig. 1). However, it is
due to lack of exposure, and is mainly documented in also clear from the wells that there is a major uncon-
early literature, summarized by Arkell (1933). It is formity between the Kimmeridge Clay and overlying
alluded to here because the contact between these two Volgian-Lower Cretaceous sediments (Rhys, 1974,
units is of significance offshore. figs. 6 and 7). It seems likely that, as on Humberside
Correlation of the Kimmeridge Clay Formation 29

6° W go E

LEGEND

E2l BASEMENT OR INTRA-


BASINAL HIGH

ENGLISH KIMMERIDGE
~ CLA Y OUTCROP

ENGLISH KIMMERIDGE
~ CLA Y SUBCROP

INTERNA TIONAL
SECTOR BOUNDARY
a 100 200
I ! !
kms

58 Q

56°

54°

52°
52°

1
KIMMERIDGE BAY, DORSET
50°

3° W 0° 3°
Fig. 1. Generalized Jurassic tectonic setting of the North Sea Area, showing distribution of the Kimmeridge Clay on the UK
mainland (from Cox and Gallois, 1981) and position of southern type and reference wells referred to in the text.
30 Petroleum Geochemistry in Exploration of the Norwegian Shelf

1l
II
m CORALLIAN I SPILSBY SST. AND EQUIVS.
II
~-

0
~ (")
o 1:,:,:1:,:,:
~ 0
-<
o 1',1,1 ,:,
'',,'' 0"0
orn
It""""""'" :'~ :'
~ ~
1,',1 11,',1, ~ '1'1 I
~ 1,',1[1,',', :1:':': :':1:':':1: 1'1',' I ',' ~, QO~ :Or
Z ',',']1]1,1, I:':':':':':':':':':':' I',',' I II
',' 1',1,',',',1,',
',1]1,1,1,1,1,',
;0;-
ll'- (f)Z
I,',',',',', ''',,,'''
> ,',',',1,1]',11,',']',1,' 1;1] =:1»

,'',,'',,'',,'1','',,1,'',',,'',,
Z ,1,1,',',',1,',11,',',',1,' ,I,','
,'I'll
1,1, ~ C"
"':..
::to
-(f)
~ :':':':':':' 11',',',',',',',',11,',1,',1,' I'll,' 11'1 :0.
I eo'" m"
~eo
.."
o
:0
O~rZ~[ AMPTHILL
CLAY KIMMERIDGE CLAY ["SP"EETON CLA Y "'.0
~
s::
>
I
~
(5 CORALLIAN SPILSBY SANDSTONE (f)
Z
,:,:,:,r':l: I(,:,::f. J T
til,
i::r >Z

~
.."
',',I,',']'r F "I',',',',J,J .,.)
> --+-.0\>
11',r;r,',',',',' '"yo ~O
Z:o
o ''1',','1',',
IIIIIIIIIIIIII~
m ,',',',',',~
11'1 1111'1 1
,',',',',',',','11
11,1,1,',1,',1,1 ~~III,IIIIIIIII
. 1',',',',1,1,' eo -~
c::t
(f)
:':':1:':1:' I ""t'I'J"""I:if----+·,,)
1'1'1"1 111 ',',',' ';1:':':1:':':': ~ '" ~m
;";(f)
~
>
-<
1l
°m~[ KIMMERIDGE CLAY (Sensu Rhys 1974) I ~ON CLAY
m
(f)
KIMMERIDGE CLAY (Sensu N.A.M. & R.G.D. 1980) Z
II Om
c,- ..,,~
• Z ..,,::t
,o·!> (f)m
~;: ::t:O
Or
:g~ :0>
3 m 0Z
"UPPER KIMMERIDGE CLAY·
(f)

0
l> >
Z
0.;:
~-
<> _.
~ n "iii
:o::t
I/>::r
-
<> <> >0
•~ I/>tII m
mz
~~ z~
",II'
(")l>OO :0
r-Z:D" ~ >
l>l> m
<m;:Z
r
enll r ;:
6g~l>
z-<:o 0
Z m
:,:,:,:,:,:,:,:,:,:,:,::,:1:':':':1
ITI('")
en zz
!7 " 'II,
m
.,)
:: :' :!I: 111
i 0_ ~o
':':':':1:':1:':';1:':':':':':':':':';'1 I 0< :0:0
1,'1,1,',,',',1',',"1'1'1',')'111,', .;.~ >::::

D
::' ::': :'l:I:, ',',',1,',','1',',1,',',1,',",1,',',',',',' rm
II
I/>
.,)
1:
1 ""
~<>
~ ,I 1IIII1I,' 1' ,' ,' I' " ',',',',',', I, ~ ,' ,'I , eo-
:~ '"I ',',1,1,',1 1","'11111'11, II1I
~QD :0-
HAUGESUND FARSUND MANDAL
(= Kimm. clay »z
rabFISKl Sensu Olstad 19B3) tII
m
en Z

0
l>
Z
TAU
*"
....en
0
Z
m
en

*
~
SAND- EGERSUND B0RGLUM UNIT
;: -NES
l>
II
G'l
Z r-
l>
r ~
,I,'1,1,, :::Z
en
l>
Z
m
....en ':':':':':':':':':':':':':':'::':':\11'"
',',',',',',',',',',',',',',',',',',',',',', , I i
I I1 I
~ 1,1 :1:'
" 1 ,','
' 1
,I,'
' "

,'1'I,1,',
11::,:1
11~ ,','
;0;'
-z
" ,','I III'
I' I'" Zo
" 1
I I
"

I',', I~:':1:1':,'',:
0 0 ',1 1'1"'1',','1',',',',',',',',',',',',',',', I , "
Z ',1 1',1,',',1,',1 1',',111,1,',1:1,':1,1,1,1,'1'" I :1 I ,',I ":0
': I''1:':,' ,',I
0 m
m en :1:':':':':';':':';1:1:';':'::':':1:';':':':':':'11 I ,,~

<
m
r
111'1111,1,11~1"II""'IIII"IIII:'I"I"""II,I"lir,
I'
I I ill
' 'I" I:': I:':' :o::t
>(J)
I m
tII
0 HEATHER DRAUPNE (=KIMMERIDGE CLAY (Sensu Dee9an + Scull, 1977)

0
"m
;: ~>

....Z *
en ..
Z
....0 en
en r.... OXFORDIAN I KIMMERIDGIAN I VOLGIAN RYAZANIAN
I en
0 ....
:;: 0
;z Z
m LOWER
en UPPER JURASSIC CRETACEOUS

Fig. 2. Correlation chart for the Kimmeridge Clay and associated sequences in the North Sea. Unit nomenclature is indicated
at formation level.
Correlation o/the Kimmeridge Clay Formation 31

(e.g. Richardson, 1979) all but the lowermost zones of Where the unit is bounded by argillaceous sequences
the Kimmeridge Clay have been removed by erosion. no obvious distinction can be made.
The Ampthill Clay is not delineated offshore, but In essence, therefore, the basic criteria on which
Rhys (1974, p. 7) writes: 'The Kimmeridge Clay For- the northern 'Kimmeridge Clay Formation' was
mation in the type well section probably includes at its defined were never documented in the southern type
base a few feet of shale which on a chronostrati- and reference sections.
graphic basis would belong to the Ampthill Clay, but
which cannot be identified as such without detailed
paleontological evidence' and 'Elsewhere the Southern North Sea, Netherlands sector
Humber Group consists entirely of mudstones or
shales, equivalent to the land area sequence of Oxford, The publication of a comprehensive Netherlands
Ampthill and Kimmeridge Clays, which are indistin- lithostratigraphic scheme by Nederlandse Aardolie
guishable on the available lithological evidence Maatschappij and Rijks Geologische Dienst (1980)
(italics added). further extended the Kimmeridge Clay Formation
The inclusion which has occurred of offshore cor- eastwards into the Netherlands sector (see Figs. 1 and
relatives of the Ampthill Clay in the Kimmeridge Clay 2), again with an increased age range (Oxfor-
makes sense in lithostratigraphic terms; offshore, as dian-?Ryazanian) compared with the UK sector. The
onshore, there is no obvious lithological criterion to unit is described as medium to dark olive-grey, gener-
separate the two units. Nevertheless, in using the ally silty shales with numerous thin dolomite streaks.
name Kimmeridge Clay Formation offshore the It is not defined by anomalous gamma-ray and sonic
stratigraphic range of the unit has been broadened, log responses, and an examination of the reference
and its definition altered, compared to the 'chrono- well F3-3 (N.A.M. and R.G.D., 1980, enc. 22) shows
stratigraphically defined' onshore unit (see Fig. 2). overall similarity with the formation in the UK sector
of the southern North Sea, albeit a much thicker
section. The base of the Kimmeridge Clay sensu
N.A.M. and R.G.D. is within the Late Oxfordian and
Northern North Sea, UK and Norway sectors it is clear that, as in the UK sector shales which are
Extension of the nomenclature to this area was car- ti~e-equivalent and lithologically si:nilar to the Amp-
ried out by Deegan and Scull (1977), who described thlll Clay have been absorbed into the Kimmeridge
well reference sections some 500 km to the north of Clay Formation (see Fig. 2).
the offshore type well. The Kimmeridge Clay Forma- In the southern part of the Dutch Central Graben
tion sensu Deegan and Scull occupies a much broader the formation is divided by a marginal marine sand
chronostratigraphic range (Oxfordian-Ryazanian) (Pu~zle Hole Formation) into Lower and Upper Kim-
than the southern formation. It is defined by its high merIdge Clay Members. The lower unit is of Late
organic content, high gamma-ray readings (greater Oxfordian age, while the upper unit is of Kimmerid-
than 100 API units) and anomalously low sonic vel- gian-?Ryazanian age. These units do not, however
ocities. Deegan and Scull remarked that the formation relate in any obvious way to the Upper and Lowe;
in the southern North Sea is considerably less divisions of the type Kimmeridge Clay, which corres-
radioactive (less than 60 API units) but 'still has the pond to the Upper and Lower Kimmeridgian stage
typical low velocity'. sensu anpl~co (Arkell, 1947, p. 67; Cope, 1980,
Examination of the southern Kimmeridge Clay type p. 77). ThIS IS a potential source of further confusion.
and reference sections confirms that the gamma-ray Interestingly, N.A.M. and R.G.D. (1980) do not
log shows a normal shale response, both in an absolute extend the Humber Group across the median line into
sense (i.e. when corrected for borehole and mud the Netherlands sector, instead referring the Kim-
density effects) and in a relative sense (Le. compared meridge Clay Formation to the locally-defined Scruff
with other shales in the sequence). Furthermore, the Group (p. 38). This perhaps reflects the inadequacy of
velocities of the southern unit are not anomalously low the attenuated type and reference sections in the
for a claystone unit. The sonic log of Cretaceous type southern UK sector in providing a definition of the
well 48/22-2 (Rhys, 1974) shows that the interval Humber Group (and its subdivisions) which can be
transit time gradient of the formation is more-or-Iess confidently extended to other offshore areas.
on trend with that of the overlying Speeton Clay For-
mation of the Lower Cretaceous (see Fig. 3(a». Simi- Summary
lar results are obtained by comparing the Kimmeridge 1. The Kimmeridge Clay in the Yorkshire Basin and
Clay sensu Rhys with other argillaceous units (the East Midlands Shelf can usually only be distinguished
Oxford Clay and Lias) in Jurassic type well 47/15-1. from the lithologically similar Ampthill Clay on
Interval transit times for the Kimmeridge Clay For- palaeontological grounds-Leo by the presence of
mation sensu Rhys, when compared with normal Kimmeridgian fossils.
compaction trends for northern North Sea shales 2. In the Southern North Sea Basin there is simi-
(e.g. Chiarelli and Duffaud, 1980), appear, in fact, to larly no obvious stratigraphic marker at Base Kim-
be somewhat 'fast' (see Fig. 3(b». This is perhaps not meridgian, but here 'Ampthill Clay equivalents' are
surprising in view of the position of the type and included in the Kimmeridge Clay.
reference wells close to the Market Weighton High 3. The 'hot' organic shales of the northern
and Dowsing Fault Zone (Fig. 1), where burial and Kimmeridge Clay Formation sensu Deegan and Scull
subsequent uplift may be expected to have occurred. are lithologically distinct from the formation in the
The southern Kimmeridge Clay Formation only southern North Sea. Neither the formation sensu
appears 'slow' when bounded by units of completely Rhys in the southern UK sector, nor the formation
different lithology; e.g. where overlain by the Spilsby sensu N.A.M. and R.G.D. in the Netherlands sector
Sandstone (Volgian-Ryazanian) or Red Chalk are distinguished by particularly high organic content:
(Albian), and underlain by Corallian limestones. 4. Succeeding offshore extensions of the unit have
32 Petroleurn Geochernistry in Exploration of the Norwegian She!f

GAMMA RAY INTERVAL TRANSIT TIME INTERV AL TRANSIT


(A.P.r. UNITS) (I-'
SECS/FT ) TIME (I'
SECS/FT )
o 120 240 140 90 200
I
100
I
SO
I

100m

00 ARROW OENOTES
~ OVERALL TRANSIT 200m
TIME GRAOIENT
IN THE SHALE
SE~UENCE i
0
">,...
300m
~

SPEETON n
0
CLAY ~ 400m
"'0
>
SOm FORMATION n
... W'
(5
z ~ SOOm
... %
"mZ
~ f\ 600m

S.NORTH
SEA VALUES 700m
OOm (FROM RHYS,
1974)

SPILSBY
SANDSTONE ) 800m

FORMATION
% 900m

KIMMERIDGE
SOm CLAY FORMATION 1000m

(sensu Rhys, 1974)


1100m

a.) Adapted from Rhys (1974) b.) Adapted from


Chiarelli & Duffaud (1980)

Fig. 3. Velocity characteristics of the Kimmeridge Clay in the southern North Sea. (a) GR-Sonic log of Cretaceous type well
48/22-2 (Rhys, 1974) showing similarity of Kimmeridge Clay velocity gradient with overlying Speeton Clay. (b) Comparison of
southern Kimmeridge Clay transit times with normally compacted Jurassic shales in northern North Sea.

therefore resulted in significantly different defini- cal in understanding the lithostratigraphic modifica-
tions, while maintaining the same name. tions which have subsequently been proposed.
Shale sequences deposited in a basin with anoxic
bottom conditions are usually easily recognizable in
RELEVANCE OF ANOXIC the subsurface, from cuttings and on wireline logs
ENVIRONMENTS TO UPPER such as gamma-ray, sonic, density and resistivity. In
JURASSIC-LOWERMOST CRETACEOUS marine shales there is generally a good correlation
LITHOSTRATIGRAPHY between gamma-ray response and total organic car-
bon content, particularly when the gamma-ray is cor-
Because of its importance to source rock studies, re- rected for borehole and mud density effects. This is
cognition of anoxic environments is regarded as an due to the tendency of uranium to form complexes
important basis for stratigraphic classification in the with organic material under reducing conditions (e.g.
Northern North Sea. Dominance of organic-rich oil- Serra et al., 1980). Usually high gamma ray response
shale lithology was the essential criterion in the corresponds with depressed density/velocity values
Deegan and Scull (1977) definition of the Kimmeridge and high resistivities, the highest resistivities occur-
Clay Formation. The character, distribution and reg- ring in thermally mature sections. Note, however,
ional significance of anoxic intervals is therefore criti- that none of these log responses is a direct reflection of
Correlation of the Kimmeridge Clay Formation 33

occurrence of major anoxic shale developments was


probably a function of more localized control. Studies
of the northern North Sea area show that timing and
duration of the organic shale sequences can vary from
basin to basin (Rawson and Riley, 1982; Hamar et al.,
1983). Restriction of circulation within a partiCUlar
structural province probably related to its overall
bathymetric and geometric configuration, to its posi-
tion on the NW European shelf, and to local tectonic
effects. Detailed palynofacies studies also support this
observation (pers. comm., M. F. Whitaker). This is
nowhere more obvious than in the Norwegian sector.
In the Norwegian Central Graben 'hot' shales of pre-
dominantly Ryazanian age occur (MandaI Formation
= of Hamar et al., 1983). To the northeast in the Nor-
wegian-Danish Basin a quite separate event depo-
= sited 'hot' shales in the Kimmeridgian-Volgian (Tau
Formation of Olsen and Strass, 1982). There is no
suggestion of a diachronous communication between
these units across the Southern Vestland Arch. To the
north in the Viking Graben the organic shales occupy a
broader chronostratigraphic range, from Oxfordian to
Ryazanian (Draupne Formation of Vollset and Dore,
1984) although 'warm' shales of lesser organic carbon
content frequently typify the uppermost and lower-
most parts of the sequence. Correlation of these units
= TETHYS
is shown in Fig. 2, and is addressed in more detail in
the following section.
There are some hints at synchronicity of certain
anoxic events; in particular, the onset of radioactive
Fig. 4. Schematic palaeogeography of NW Europe-Green- shales at the base of the Tau Formation probably has
land in late Jurassic to Ryazanian times. Shaded areas rep- regional significance (see next section). In general,
resent basement or intra-basinal highs. Arrows indicate however, the anoxic intervals of the Northern North
connections of North Sea basin complex to major marine Sea are best recognized as locally developed litho-
systems. LM = Laurentian Massif, FSS = Fennoscandian
Shield, VG = Viking Graben, SVA = Southern Vestland stratigraphic units, which should only be extended
Arch, NDB = Norwegian-Danish Basin, MNSH = Mid- from basin to basin with extreme caution.
North Sea High, CG = Central Graben, RFH = Rinkllb-
ing-Finn High, SNSB = Southern North Sea Basin, LBM
= London Brabant Massif.
NOMENCLATURE AND CORRELATION IN
THE NORWEGIAN SECTOR
the quality of organic matter (kerogen type) present in
Figure 2 summarizes the current lithostratigraphic
the shale. The term 'hot shale' is used herein as a
scheme for the Oxfordian-Ryazanian rocks of the
shorthand to denote highly radioactive shales, which
Norwegian sector, and compares it with other areas
by implication are of high organic richness.
considered in this account. A fuller and more descrip-
Models for the formation of anoxic shale sequences tive lithostratigraphy for the Norwegian sector is
were discussed extensively by Demaison and Moore given in Vollset and Dore (1984). It is only discussed
(1980). In one such model they demonstrate from here insofar as it relates to the central issue, which is
studies of the Black and Baltic Seas how anoxic condi-
the desirability of establishing a local nomenclature
tions are able to develop in a 'silled basin' setting,
system for Norwegian waters.
during times of worldwide transgression (p. 1204).
Tyson et al. (1979) apply a similar model to explain
Norwegian Central Graben
facies assemblages in the type Kimmeridge Clay.
They suggest that the major transgressive episodes of The Upper Jurassic of this area was not considered in
the Late Jurassic, which occurred over a shelf with detail by Deegan and Scull (1977), who concentrated
limited connection to the Tethyan Ocean in southern mainly on type and reference sections in the Viking
Europe (see Fig. 4), resulted in a stratified water col- Graben and East Shetland Basin. Consequently, there
umn with a stagnant, frequently anaerobic bottom has been some uncertainty in identifying the base of
layer. Deposition of anoxic or more 'normal' marine their 'Kimmeridge Clay Formation'. In some of the
shales depended upon cyclical fluctuations within the older well reports the contact is placed at the base of
water column. Seasonal convection of nutrients from what is now called the Farsund Formation. This unit,
the lower anoxic layers caused coccolith blooms, which is 200 m thick or more in the graben centre, is a
depositing the thin carbonate stringers which are typ- Kimmeridgian to Volgian shale sequence of moder-
ical of the type Kimmeridge Clay and some of the ately high organic content (see Fig. 2).
offshore shale sequences. More commonly, the 'Kimmeridge Clay' has been
These models provide an overall regional identified with the overlying 'hot' shale sequence of
framework within which anoxic bottom conditions, late Volgian-Ryazanian age, now known as the Man-
and consequent preservation of organic material, can dai Formation. This unit (type well 7/12-3A) is up to
occur. Given this tendency, however, the preferential 70 m thick and thins over structural highs. It is fre-
LEGEND
LJJ Buemen! or

J
',;; In!ra-ballnal

high
,1 00m
~ Land VERT.
~ LOG
SCALE

R '" Ryuanlan

Known IIml!

Known IIml!
of Tau Fm.

Known limit
of Mandai Fm.

Open marine
59° 1=-
-
=1 anaerObic
to mildly
claystones
t<:::J Sandstones

58°

Fig. 5. Known distribution of organic shale intervals (Upper Jurassic-lowermost Cretaceous) in the Norwegian sector, with
specimen log profiles. Gamma ray scale: 0-200 A.P.I. units. Sonic log scale 140-40 jJsjft.
Correlation of the Kimmeridge Clay Formation 35

quently typified by an unusually high proportion of the MandaI Formation in the Central Graben (see
algal kerogen. Its approximate distribution area is Fig. 2).
shown in Fig. 5. The unit was named the 'Kimmeridge
Clay Formation' by Ofstad (1983), informally redesig- Viking Graben and East Shetland Basin
nated the MandaI Formation by Hamar et al. (1983)
and formally defined by Vollset and Dare (1984). It The predominantly anaerobic Draupne Formation
appears to correlate southwards with highly radioac- (type well 30;6-5) is present throughout the Viking
tive shales in the upper part of 'U nit J4' in the Danish Graben, East Shetland Basin and over part of the
part of the Central Graben (Michelsen and Andersen, Horda Platform (see Fig. 5). It essentially corres-
1983, fig. 4) but is not seen further south in the pub- ponds to the northern 'Kimmeridge Clay Formation'
lished wells of the Dutch Sector (N.A.M. and R.G.D. as defined by Deegan and Scull (1977). It is of Oxfor-
dian to Ryazanian age, approximately time-equivalent
1980, Enc!. 2).
The MandaI Formation appears to represent a pre- to the Farsund-Mandal sequence of the Central Gra-
dominantly Ryazanian anoxic event peculiar to the ben, and overlies the predominantly open marine
Central Graben, which is not matched by similar log shales of the Heather Formation. The
responses elsewhere. It is possible that the establish- Draupne-Heather contact is diachronous between the
ment of open marine circulation between Tethys and Oxfordian and Kimmeridgian, probably in part a
the proto-Atlantic rift system in latest Ryazanian result of anoxic bottom conditions reaching progres-
times was preceded by the incoming of slightly more sively higher parts of the basin complex during the
oxygenated waters in the Norwegian-Danish Basin course of the Late Oxfordian-Middle Kimmeridgian
and East Shetland Basin (see Fig. 4). The central, transgressive episodes.
more-protected position of the Central Graben may, in The thickness of the Draupne Formation is
contrast, have allowed progressive stagnation to occur extremely variable, due to onlap of the basin margins,
but probably reaches over 1000 m in axial parts of the
as transgression continued, resulting in the anoxic
yiking Graben. As defined in Vollset and Dore (1984),
maximum of the Ryazanian.
It encompasses a fairly wide range of lithotypes.
Organic content varies (albeit against a generally high
Norwegian-Danish Basin 'background') with the 'hottest' shales of all frequently
The Tau Formation-'hot' shales of Kimmeridgian to occurring in the middle (Kimmeridgian-Volgian) por-
early Volgian age-is widely distributed in the Eger- tion of the sequence. A tripartite subdivision of the
sund and Fiskebank Sub-Basins of the Nor- Draupne Formation on this basis may eventually be
wegian-Danish Basin (see Fig. 5) and is usually about desirable. Minor sandstone and siltstone interbeds
50 m thick. The type well for the formation is 9/4-3. It probably turbiditic, are frequently present. The;
overlies a sequence of silty marine shales of Callo- occasionally dominate the sequence, reducing overall
vian-Kimmeridgian age, the Egersund Formation. gamma-ray response. Larger, more massive sand
The Tau Formation was formerly equated with the bodies are found enclosed within the Draupne shales
B0rglum Unit, a dark (but less radioactive) close to the basin margins and intra-basinal highs of
homogeneous marine shale unit occurring in the Dan- the UK sector; for example, the Magnus sand (De'Ath
ish sector (Deegan and Scull, 1977, p. 22; Michelsen, and Schuyleman, 1981). The 'Kimmeridge Clay For-
1978, pp. 21-22). Hamar et al. (1983) acknowledged mation' ofthe Outer Moray Firth basin (Turner et al.,
the distinctive character of the Tau unit by separating 1984), also Oxfordian-Ryazanian, has a similarly var-
it out as a member of the 'B0rglum Formation', but ied lithology with a major intercalated sand, the
the unit is now considered widespread enough to merit Claymore Sandstone Member. The sequence in this
formation status (Vollset and Dore, 1984; see also area appears to match well with the newly defined
Olsen and Strass, 1982). The Tau Formation and its Draupne Formation. However, the further question of
bounding silty lithologies grade into the more correlation with the Inner Moray Firth and North-
homogeneous B0rglum Unit in the southeastern part of East Scotland is not addressed here.
the Fiskebank Sub-Basin (see Figs. 2 and 5).
Rawson and Riley (1982) pointed out that the mid-
dle Kimmeridgian appearance of 'hot shales' at the DISCUSSION
base of the Tau Formation corresponds to the first
appearance of oil-shale facies in the eudoxus zone over Organic shale sequences of approximately Late
the entire English Kimmeridge Clay outcrop. This Jurassic age are a notable example of the 'persistence
event can also be recognized in the Norwegian Central of facies' concept-the deposition of very similar
Graben, where 'hot' streaks are not infrequent in the lithologies over immense areas at certain points in the
basal part of the Farsund Formation around the geological record (e.g. Ager, 1981). Widespread dis-
Southern Vestland Arch, and in the Viking Graben tribution of the organic shales can be explained by an
where it is often observed as an upward transition to overall sea level rise combined with a configuration
'hotter' shales within the Draupne sequence. This favourable for restricted marine circulation in the
event, which clearly represents a widespread middle megastructure of northern Europe (see earlier sec-
Kimmeridgian transgression, has potential use as a tions; also Ziegler, 1982, enc. 20). Roughly contem-
lithostratigraphic boundary in areas other than the poraneous deposits of this type exist northwards along
Norwegian-Danish Basin (see Fig. 2). the Norwegian shelf into the Arctic regions and are
The Tau Formation is overlain by a silt- found as far afield as And~ya, Troms0flaket, Sval-
stone-claystone unit ofVolgian age, the Sauda Forma- bard, East Greenland and Western Siberia (see Table
tion. The succeeding finer grained deposits of the 1).
Flekkefjord Formation represent a major Ryazanian Although in the past there has been a tendency to
transgression (Rawson and Riley, 1982), and corres- express the conceptual unity of such intervals in com-
pond in time with the much more radioactive shales of mon lithostratigraphic terms (Arkell (1933, p. 613)
36 Petroleum Geochemistry in Exploration of the Norwegian Shelf

TABLE 1. Organic shale intervals roughly contemporaneous with 'Kimmeridge


Clay Formation' in the Arctic region

NalYle of
Location lithostratigraphic unit Age Authority

And0ya (N. Norway) Ratj0nna Member ?Kimmeridgian Dalland (1979)


(of Dragneset Fm) to Volgian
Troms0flaket Under discussion Kimmeridgian to Westre (1948)
Hauterivian and unpub!. well data
East Greenland Hareelv Formation Oxfordian to e.g. Surlyk (1975)
and Bjernberg Kimmeridgian
Formation
Svalbard Agardhfjellet Member Callovian to e.g. Steel and
(of Janusfjellet Fm) Volgian Worsley (1984)
Western Siberia 'Bazhenov Deposit' Various intervals e.g. B.E.I.C.P.
and others within Callovian and Robertson
to Ryazanian Research (1980)

noted 'Kimmeridge Clay' in the Moscow region), this pp. 7-8). Formation nomenclature is not intended as a
is not in accord with modern lithostratigraphic princi- vehicle for broad unifying concepts on a continental
ples, since it presupposes a connection which has not scale. This is the realm of 'event stratigraphy', bio-
been proven. A prime criterion for identifying a litho- stratigraphy and seismic stratigraphy.
stratigraphic unit is continuity (or original continuity)
with the type area (Hedberg, 1976). This is frequently
impossible to demonstrate in a new area, and this is
why local nomenclature systems evolve. Later,
detailed correlation work between areas may make it CONCLUSIONS
desirable that certain units should be unified under a
single name. However, in identifying the Kimmeridge
Clay Formation over wide offshore areas this correla- 1. There are significant differences in the way the
tion has been taken for granted. It has been assumed 'Kimmeridge Clay Formation' is defined-both in
that lithological similarity and continuity are terms of lithotype and boundaries-between the
self-evident, whereas in reality they have not been onshore type area, the southern North Sea and the
demonstrated. northern North Sea. This has made regional correla-
We have remarked that the defining lithological tion difficult and led to lack of communication between
characteristics of the Kimmeridge Clay in the north- individual workers.
ern North Sea do not correspond with those of the 2. Anoxic shale units ofthe northern North Sea are
offshore type area. Even accepting that a fairly wide frequently readily mappable units which should be
lateral variation in lithology may be acceptable within classed as formations; however, they are not necessar-
a formation, there remains a fundamental problem ily synchronous and are of limited lateral continuity.
concerning the base of the unit. The recognition of a The 'hot shale' definition of the Kimmeridge Clay
firmly defined base is becoming increasingly impor- Formation provided by Deegan and Scull (1977) has
tant in modern stratigraphy (e.g. Ager, 1981, led to major problems of correlation, particularly in
pp. 67-79). In the case of the Kimmeridge Clay For- the Norwegian sector. Adoption of local nomenclature
mation there is no correspondence between the base of is urged for these intervals. This, at least, provides a
the unit onshore, where it is essentially a chrono- lithostratigraphic scheme which is locally workable.
stratigraphic marker separating similar. Oxfordian and Extension of these units between adjacent structural
Kimmeridgian claystones, and offshore, where Oxfor- provinces may well be possible, but should be carried
dian claystones have been included in the Kimmeridge out with caution. The concept of a unified 'Upper
Clay (see the second section). Weare thus left with no Jurassic transgressive unit', containing anoxic shale
clear directive about the base of the unit; there is no intervals as well as 'normal' shales and other
'golden spike' for the Kimmeridge Clay Formation. lithologies, could be expressed at group or supergroup
It is this lack of definition which is partly responsi- level if desired.
ible for the progressively broader and more disparate 3. Use of the term Kimmeridge Clay Formation in
view of the Kimmeridge Clay formation we find as we the southern UK and Netherlands sectors is ques-
move further afield in the North Sea area. tioned, and a review suggested. It may at least be
The reclassified units of the Norwegian sector are desirable to rename the 'Upper' and 'Lower Kim-
therefore advocated in preference to a diverse Kim- meridge Clay Members' of the Netherlands to avoid
meridge Clay Formation of limited lithostratigraphic confusion with the Upper and Lower divisions of the
value. This has the advantage that correlation of type Kimmeridge Clay.
sequences and events can be discussed free of the false 4. The Upper Jurassic-lowermost Cretaceous shale
sense of equivalence which the name Kimmeridge sequences of the North Sea in general are important
Clay Formation imposes. It should be remembered enough to merit a thorough review. Many current
that lithostratigraphic units are 'localized in occurr- problems of regional correlation and nomenclature
ence' and are defined 'to provide an observable could be resolved by a Correlation Group including
framework of rock for reference in subsequent corre- representatives from each of the national sectors
lation and interpretation' (Holland et al., 1978, involved.
Correlation of the Kimmeridge Clay Formation 37

REFERENCES mid-Cretaceous sections, as demonstrated on a field meeting,


18-20 May 1979. Proc. Geol. Ass., 94(1), 65-73.
Michelsen, O. 1978. Stratigraphy and distribution of Jurassic
Ager, D. V. 1981. The Nature of the Stratigraphical Record (2nd deposits of the Norwegian-Danish Basin. Dan. Geol. Unders.
Edn). Macmillan Press. Ser. B, 2.
Arkell, W. J. 1933. The Jurassic System in Great Britain. Claren- Michelsen, O. and Andersen, C. 1983. Mesozoic structural and
don Press, Oxford. sedimentary development of the Danish Central Graben. In: Pet-
Arkell, W. J. 1947. The geology of the country around Weymouth, mlcum Geology of the South-eastern North Sea and Adjacent
Swanage, Corfe and Lulworth. Mem. Geol. Surv. G.B. Onshore Areas. Geol. Mzinh, The Hague.
B.E.loC.P. (Bureau d'Etudes Industrielles et de Cooperation de Nederlandse Aardolie Maatschappij B.V. and Rijks Geologische
I'Institut Fran~ais du Petrole) and Robertson Research Interna- Dienst (N.A.M. and R.G.D.) 1980. Stratigraphic Nomenclature
tional Ltd 1980. The Petroleum Industry in the U.S.S.R. Non- of the Netherlands. Verlog K. Nea. Geol. Mijnb. Genoot.
exclusive industry report. Ofstad, K. 1983. Geology of the southernmost part of the Nor-
Brookfield, M. 1973. Palaeogeography of the Upper Oxfordian and wegian section of the Central Trough. N.P.D. Paper No. 32.
Lower Kimmeridgian (Jurassic) in Britain. Palaeogeogr., Olsen, R. C. and Strass, 1. F. 1982. The Norwegian-Danish Basin.
Palaeoclimatol., Palaeoecol., 14, 137-167. N.P.D. Paper No. 31.
Chiarelli, A. and Duffaud, F. 1980. Pressure origin and distribution Rawson, P. F. and Riley, L. A. 1982. Latest Jurassic-Early Cre-
of Jurassic of Viking Basin (United Kingdom-Norway). Bull. taceous events and the 'Late Cimmerian Unconformity' in the
Am. Ass. Petrol. Geol., 64(8). 1245-1266. North Sea area. Bull. Am. Ass. Petml. Geol., 66(12), 2628-2648.
Cope, J. C. W. 1974. New information on the Kimmeridge Clay of Rhys, G. H. (compiler) 1974. A proposed standard lithostrati-
Yorkshire. Proc. Geol. Ass. 85(2), 211-221. graphic nomenclature for the southern North Sea and an outline
Cope, J. C. W. 1980. Kimmeridgian correlation chart. In: Cope, J. structural nomenclature for the whole of the (UK) North Sea.
C. W. et al. (ed.), A Cm-relation of Jurassic Rocks in the British Rep. Inst. Geol. Sci., No. 74/8.
Isles. Part 2. Geol. Soc. Land. Spec. Rep. 15. Richardson, G. 1979. The Mesozoic stratigraphy of two boreholes
Cox, B. M. and Gallois, R. W. 1981. The stratigraphy of the Kim- near Worlaby, Humberside. Bull. Geol. Surv. G.B. 58.
meridge Clay of the Dorset type area, and its correlation with Serra, 0., Baldwin, J. and Quirein, J. 1980. Theory, interpretation
some other Kimmeridgian sequences. Rep. Inst. Geol. Sci., 80/4. and practical application of natural Gamma Ray Spectroscopy.
Dalland, A. 1979. The sedimentary sequence of Andflya, Northern SPWLA, B1st A. Log. Symp. Trans.-Paper Q.
Norway. Depositional and Structural History. In: Norwegian Sea Steel, J. R. and Worsley, D. 1984. Svalbard's post-Caledonian
Symp. Norwegian Petroleum Society. Tromsfl. strata-an atlas of sedimentological patterns and paleogeo-
De'Ath, N. G. and Schuyleman, S. F. 1980. The geology of the graphic evoluion. In: Spencer, A. M. et al. (eds.), Petroleum Geol-
Magnus oil field. In: Illing, L. V. and Hobson, G. D. (eds.), Pet- ogy qfthe North European Mar:qin. Graham and Trotman, Lon-
roleUin Geology of the Continental Shelf of N. W. Europe. Insti- don, for Norwegian Petroleum Society, 436 pp.
tute of Petroleum, London, pp. 342-351. Surlyk, F. 1975. Block faulting and associated marine sedimentation
Deegan, C. E. and Scull, B. J., (compilers) 1977. A proposed stan- at the Jurassic-Cretaceous boundary, East Greenland. Jurassic
dard lithostratigraphic nomenclature for the Central and North- Northern North Sea Symp. Norwegian Petroleum Soc.,
ern )/orth Sea. Rep. Inst. Geol. Sci. 77/25; Bull. No1'1)). Petrol. Stavanger.
Direct., 1. H.M.S.O., London. Turner, C. C., Richards, P. C., Swallow, J. L. and Grimshaw, S. P.
Demaison, G. J. and Moore, G. T. 1980. Anoxic environments and 1984. Upper Jurassic stratigraphy and sedimentary facies in the
oil source bed genesis. AAPG Bull .. 64(8), 1179-1209. Central Moray Firth Basin, North Sea. Mar. Petrol. Geol., I,
Fitton, W. H. 1836. Observations on some of the strata between the 105-117.
Chalk and the Oxford Oolite in the south-east of England. Trans. Tyson, R. V., Wilson, R. C. 1. and Downie, C. 1979. A stratified
Geol. Soc. Land., 2(3), 103-388. water column environment model for the type Kimmeridge Clay.
Hamar, G. P., Fjreran, T. and Hesjedal, A. 1983. Jurassic stratig- Nature, 277 (5695), 377-380.
raphy and tectonics of the south-eastern )/ orwegian offshore. In: Vollset, J. and Dore, A. G. (eds.) 1984. A revised Triassic and
Petroleum Geology of the South-eastern North Sea and Adjacent Jurassic lithostratigraphic nomenclature for the Norwegian
Onshore Areas. Geol. en. Mzjnb, The Hague. North Sea. N.P.D. Bull. No. S.
Hedberg, H. D. (ed.) 1976. IntC1'1wtional Stratigraphic Guide: Westre, S. 1984. The Askeladden gas find-Troms 1. In: Spencer,
Guide to Stratigraphic Classificntion, Terminology and Proce- A. M. et al. (eds.), Petrolewn Geology of the North Europenl1
dure. International Subcommission on Stratigraphic Classifica- Jlargin. Graham and Trotman, London, for Norwegian Pet-
tion of LU.G.S. Commission on Stratigraphy. Wiley-Intersci- roleum Society, 436pp.
ence, New York. Wright, J. K. 1980. Oxfordian correlation chart. In: Cope, J. C. W.
Holland, C. H. et al. 1978. A guide to stratigraphical procedure. et al. (eds), A Correlation of Jurassic Rocks in the British Isles.
Geol. Soc. Land. Spec. Rep. No. 10. Part 2. Geol. Soc. Lond. Spec. Rep. 15.
loG.S. (Institute of Geological Sciences) 1974. Annual Report for Ziegler, P. A. 1982. Geological Atlas of Western and Central
1973. London. E-u)"upe. Shell Internasjonale Petroleum Maatschappij B. V., The
Kelly, S. R. A. and Rawson, P. F. 1983. Some late Jurassic to Hague.
3 Organic geochemistry
in exploration of the
Northern North Sea
J. D. Field
Tenneco United Kingdom Inc., 1 Castle Yard, Richmond, Surrey, UK

A subsurface geochemical study has been conducted in the Northern North Sea with the following objectives:
(1) to identify potential hydrocarbon-source horizons;
(2) to determine their regional maturity, and hence the most likely type of hydrocarbons expelled into carrier
beds and
(3) to identify secondary migration routes and to evaluate the probable gravity of accumulated hydrocarbons.
In the study, a combination of sedimentary facies, organic facies, regional structure maps and isopachs, and
modelling techniques were used to evaluate the distribution and quality of mainly Jurassic source horizons. The
Upper Jurassic Kimmeridge Clay Formation is the most important oil source horizon on the Tampen Spur and in
the Northern North Sea graben system. Shales of the Heather Formation and Dunlin Group, and coals of the Brent
Group and Statfjord Formation also have source potential, but mainly for gas. The state of maturation at top Brent
Group and top Humber Group levels and likely hydrocarbon products from these horizons were assessed with well
data and burial modelling in undrilled areas. The timing of expUlsion and migration routes of hydrocarbons were
determined based upon present and reconstructed structural configurations on the Tampen Spur (Alwyn-Statfjord
trend, Brent Field), in the South Viking Graben (Brae, Sleipner, Thelma, etc., complexes) and in the North Viking
Graben (Huldra, Oseberg, Brage, Troll fields). The overall source-migration-accumulation model can be tested
against both known fields and barren structures, and results in a predictive ability for estimating hydrocarbons in
untested prospects.

INTRODUCTION SOURCE POTENTIAL


A vital function of the petroleum geologist is to predict The Kimmeridge Clay Formation and its lateral equi-
the type, or gravity, of hydrocarbons accumulated in valents have been identified as the major source
an exploration area of interest. The probabilities of a intervals in the Northern North Sea (Fuller, 1975;
structure containing oil or gas can have considerable Barnard and Cooper, 1981; Goff, 1983). This formation
influence on the economic perception of the prospect. is the source ofthe majority of oil and gas fields in the
In the Northern North Sea, the distribution of oil and study area, including Brae (Reitsema, 1983), Brent
gas fields and their related source rocks is becoming (Bowen, 1975), Magnus (De'Ath and Schuyleman,
increasingly well-documented. This area therefore 1981), Ninian (Albright et al., 1980), Piper (Maher,
provides a natural laboratory to identify the subsur- 1981), Statfjord (Kirk, 1980) and Troll (Brekke et al.,
face processes which take place during the generation, 1981). Exceptions include the Frigg complex, where
migration and accumulation of hydrocarbons. The Middle and Lower Jurassic strata have been sug-
models developed are being applied in frontier basins gested as possible oil sources (Heritier et al., 1980),
with similar geology, particularly those bordering the and the Sleipner complex, where Middle Jurassic coals
North Atlantic rift zone in North-West Europe. and basal Upper Jurassic shales of the Heather For-
The areas of study are outlined in Fig. 1. A regional mation may have sourced gas and condensate, respec-
survey has been conducted over the area bounded by tively (Larsen and Jaarvik, 1981). Brent Group and
58° and 62°N, and 0° and 6°E. Two sub-regional sur- Statfjord Formation coals are reported to be possible
veys in the North and South Viking Grabens are also contributors of gas to the Brent and Troll fields. Mid-
described. Data presented in this paper are drawn dle Jurassic shales are a possible source of oil in the
from published material, in-house analyses of released Alwyn area (Oudin, 1976), while unspecified Jurassi<.:
well data, and modelling studies. sources are invoked for Heimdal gas and condensate
Petroleum Geochemistry in Exploration of the Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 39-57.
40 Petroleum Geochemistry in Exploration of the Norwegian Shelf

V
1--++-+-+-I--+-+-+--+--if-f
!

1\
1203 5 I

~--- ---

" -- --~~~

SCOTlAND 1 ,~

3 4
I, 56]6

t-

~ 34 35 : 36 37 38 139 ; 1 55031 5504 15505 1_ 1 5~!-j-,iJ_6l--+_~+1


/'rt--t---t--+--t--+-lc--+-++-+-+-+-+--
1 1

i
H t-t- ~ A I . -f-- 1-
t-
~ I--::~
, t-H
f';+-I-- ~ 11
' \ \, A '" B "
\-\++-+-+~ ++-+--+-1-+-+++-+-++ ++--+-j-+-+-I/I-+--+- j I ~1
~~~-++~~-++~~-++I-+-~~-+'~~ C D
: I

Fig. 1. Northern North Sea, location of study areas.

(Heritier et al., 1980). The relative source potential of Figure 2 demonstrates the average potential of each
strata in the Northern North Sea is discussed below. horizon, and local enhancement of source quality is
Palaeozoic source rocks are predicted to have virtu- shown in a lighter tone. Total organic carbon (TOC)
ally no economic potential in the study area. Car- content is expressed in weight percent, and the hyd-
boniferous strata. are absent north of the Outer Moray rogen index values are representative of immature
Firth, and the Permian Kupferschiefer and overlying rocks.
lagoonal carbonates are restricted to the South Viking
Graben south of 60 N (Ziegler, 1982). In the Mesozoic
0
Statfjord Formation
and Cenozoic section, major source horizons are
developed in the Jurassic System (Fig. 2). Shales of The Statfjord Formation consists of fluvial sand-
the Humber Group, which is subdivided into the stones, siltsones, claystones and coals overlain by
Kimmeridge Clay and Heather Formations, and Dun- marginal-marine sandstones and shales. The forma-
lin Group, and coals of the Brent Group and Statfjord tion has poor average source potential with TOC less
Formation are apparent as potential source horizons. than 1%. However, more than lO m of coal is
OTganic Geochemistry in NOTth Sea Exploration 41

IORGANIC CARBONI HYDROGEN INDEX


LlTHOSTRAT.
110 % ~ mgHC/gmOM
60~

HORDALAND
GROUP

ROGALAND
GROUP

SHETLAND
GROUP

CROMER
KNOLL GROUP

KIMMERIDGE
CLAY FMN.

HEATHER
f::
V:-:-:·· :-:{:::l
v-:·:-:
FORMATION r·:-:-:-: I.'':::):
1:-:-:-:-:
1;.•••••••• ..::::::1
.. ~

BRENT COALS
GROUP
1:0::7: 8:0:~
r.~.-:-

1-:::::-:
:::::::-:::::-:::1
DUNLIN ... '-"-,
":::::-:-:::-:::1
'~'""~:"";.,,

GROUP

STATFJORD COALS
FORMATION

CORMORANT
FORMATION
( NO VERTICAL SCALE )
Fig. 2. Northern North Sea, source potential of Mesozoic and Cenozoic strata.

developed in the Raude and Eiriksson Members on the Jurassic marine transgression, which reached its max-
Horda Platforrri (Fig. 2) and it is a proposed secondary imum extent in Toarcian-Aalenian times with deposi-
source of gas for the Troll accumulation (Brekke et al., tion of the Drake Formation (Skarpnes et al., 1980)
1981). see Fig. 3. The source quality of the Dunlin Group,
and particularly the Drake Formation, varies across
Dunlin Group the study area. On the Tampen Spur, the Dunlin
Group has only limited source potential with an aver-
The siltstones, shales and offshore-bar sandstones of age TOe of 1.5% characterized by inertinitic kerogen
the Dunlin Group represent deposits of the Early (type IV). In the North Viking Graben area, however,
42 Petroleurn Geochernistry in Exploration of the Norwegian Shelf

59" + ,
1>-
"' <:{
:.cIS

-- /
::J,cr:

,I~ ../
~ .\ /
I
LOWER JURASSIC
ERODED IN
'K. MID JURASSIC

I \
58° + I \+ + + + 58°

\ \ 2~
I
~km
I
( PALAEOGEOGRAPHY AND ISOPACHS
MODIFIED AFTER SKARPNES ET AL 1980)

o
LEGEND

ASSOC. KEROGEN ASSOC. KEROGEN


FACIES TYPE FACIES TYPE

CJ
---
EJ
MARGINAL MARINE OPEN MARINE
.' .. ' . m+Ill II +ill
-"- SANDS & SIL TS SHALES & SILTS

SHALLOW MARINE DUNLIN GROUP


~
.-.- SILTS & SHALES
m ISOPACHS

Fig. 3. Northern North Sea, Dunlin Group, isopachs, palaeogeography and kerogen types at time of maximum transgression.
Organic Geochemistry in North Sea Exploration 43

average TOC ranges up to 3% with amorphous sap- tinues into the South Viking Graben, where the
ropel, exinite and vitrinite more prevalent (types II Heather may be a partial source of Sleipner conden-
and III), resulting in hydrogen index values up to 200 sate and gas.
(Ungerer et al., 1984), see Fig. 2. The limited data
available indicate that organic facies are related to the Kimmeridge Clay Formation
sedimentary facies belts outlined in Fig. 3. Margi-
nal-marine and shallow-marine sands, silts and The Kimmeridge Clay Formation and its lateral equi-
shales are dominated by inertinite and vitrinite (types valents have an average TOC of 6% and hydrogen
IV-III), and the open-marine shales in the basin index values ranging up to 600. As indicated in Fig. 2,
centre are characterized by vitrinite, exinite and optimum source quality is generally developed in the
amorphous sapropelic kerogens (types III-II). The upper part of the formation. The shales are dominated
basin-margin shales have only minor gas source by oil-prone, amorphous sapropel and exinite, but
potential. The basin-centre shales have gas and oil exhibit considerable facies-related, lateral variations
source potential in the M~re Basin, the North Viking as discussed by Barnard et al. (1981), Barnard and
and Sogn grabens, and possibly in the undrilled Stord Cooper (1981) and Cooper and Barnard (1984). The
Basin. richest source rocks were deposited in the graben
centre and restricted sub-basins in the East Shetland
Brent Group Basin, far removed from terrestrially-derived, higher
plant debris. On the basin margins and contemporane-
The Brent Group was deposited during the major ous highs, agitated oxygenated conditions prevailed,
Middle Jurassic regressive pulse associated with the which favoured the preservation of inertinite and vit-
volcanic updoming of the Outer Moray Firth (Eynon, rinite. As already stated, the Kimmeridge Clay For-
1981). The average source potential of these domin- mation is believed to be the prime source of the major-
antly arenaceous, deltaic deposits is low, but associ- ity of hydrocarbon accumulations in the Northern
ated delta-top coals and coal-bearing shales have con- North Sea. Oil-to-source correlations by Fuller (1975),
siderable source potential. The coal-bearing sequences Oudin (1976), Grantham et al. (1980), Van den Bark
exhibit TOC up to 45 per cent (Oudin, 1976) and high and Thomas (1980), Bissada (1983) and Reitsema
hydrogen index values in the range of 200-250 (1983) confirm this observation.
(Durand and Paratte, 1983; Ungerer et al., 1984). It
has yet to be demonstrated that coals have contri- Cretaceous and Tertiary
buted significantly to an oil field in the area of study,
but these values indicate a considerable generative Average TOC less than 1%, dominated by inertinite,
capability for liquid hydrocarbons and gas. and hydrogen index values less than 50 (see Fig. 2),
Oudin (1976) reports that the heavy, alkane- indicate that only poor to fair source potential is pres-
deficient oil of UK 3/15-2 exhibits good correlation ent in these systems. Locally minor gas- and oil-prone
with extracts of Middle Jurassic carbonaceous beds kerogen is developed, but the source potential of this
based on mass-spectrometry and gas-chromatography section is minor in comparison to the underlying
techniques. Heritier et al. (1980) indicate, based upon Jurassic.
pristane/nC 17 : phytane/nC 18 observations, that the
naphthenic oil of East Frigg originates from Lower
and Middle Jurassic source rocks. However, in both MATURATION
cases the oils are altered, thus the correlations may be
debatable. Goff (1983) argues for Frigg gas originating Maturation data based upon spore colouration and vit-
from sapropelic Upper Jurassic source horizons in the rinite reflectance studies are available from a number
Kimmeridge Clay Formation, based on stable carbon of structural provinces in the Northern North Sea
isotope data. By association, it might be inferred that (Goff, 1983; Pearson et al., 1983; Bissada, 1983). These
Frigg oil is also, at least partially, sourced from the studies generally indicate that maturity gradients
Kimmeridge Clay Formation.
vary between structural provinces. The most marked
difference is between the Viking Graben and its flank-
Heather Formation ing highs: the Vestland Arch, the Horda Platform, the
The Heather Formation and its lateral equivalents East Shetland Platform and the Tampen Spur. Table 1
have an average TOC of 2-2.5% and hydrogen index summarizes the differences in the maturation gra-
values ranging up to 200. It is characterized by vitri- dients for these two classes of structural province.
nite, with minor exinite and amorphous sapropel, and The greater depths of burial required in the graben
is capable of sourcing gas and associated light liquids. axis to achieve a certain level of maturity compared
Larsen and Jaarvik (1981) consider the Heather For- with the flanking highs reflect its lower average
mation may have sourced condensate in the Sleipner geothermal gradient. Figure 4 summarizes the mean
complex. The Heather Formation exhibits strong local geothermal gradients in the Northern North Sea mod-
areal variations in source potential. Several ified after the work of Carstens and Finstad (1981) and
restricted-circulation sub-basins appear to have
developed in the East Shetland Basin as the structural TABLE 1. Maturation gradients in the Viking Graben
highs developed in the late Callovian to early Kim- and on its flanking highs
meridgian. As indicated in Fig. 2, source quality in
these basins is enhanced with TOC up to 5% and hyd- Viking Graben Graben Margin
rogen index values up to 300. In the Outer Moray Ro (m) highs (m)
Firth, Bissada (1983) reports an average TOC of3.3%,
with amorphous kerogen dominant. Woody kerogens, Onset oil generation 0.55 3600 3100
Peak oil generation 0.75 4200 3600
however, are preferentially developed in the south- Floor of oil generation 1.30 4900 4400
eastern corner of UK Quadrant 15, a trend which con-
44 Petroleum Geochemistry in Exploration of the Norwegian Shelf


62o r-------~~~~

+ 60'"

+ +

+ + +
o 2~
I


LEGEND

Mod. after: Carstens & Finstad, 1981', and Evans & Coleman. 1974

Fig. 4. Northern North Sea, mean geothermal gradients.


Organic Geochemistry in North Sea Exploration 45

MILLIONS OF YEARS B P

163 145 98 65 55 25 0 MILLIONS OF YEARS B P


JUR. E. CAEl. L. CREl. PAL EOC.-OL. MIO.-QUAT.
163 '045 98 65 55 25
10° C JUR. E. CREl. L. CAET. PAL EOC.-OL MIO.-QUAT

~
~ 1-;'\

!
IMMATURE
II) ::la:
Ii! 3 >- IMMATURE

~ ..~
~ 3

.
~
~

.i!i
~
x
~ 3925m Ro 0.65> '. >-
~
3750m Ao
100 0 C
OIL WINDOW

1500 C OIL'WKJOW
04850m

5 45° API OIL PRESERVATION


517~ ________ _
4775m

OVER MATURE
OVER MATURE

GEOTHERMAl GRADIENT 156SOm BASE ~'AP' PRESERVATION GEOTHERMAL GAAOIENT .


32!i'C lOOm" BV EXTRAPOlATIONI 28'C 100m'
150D C
200 0 C

A B
Fig. 5. Northern North Sea, maturation modelling profIles. (A) Blocks 34/9-35/2; (B) Blocks 30/2-30/5.

Evans and Coleman (1974). In contrast to the Central MIGRATION


North Sea, the lowest geothermal gradients are
located along the graben axis and the highest gra- Primary migration of hydrocarbons from the predo-
dients on the flanking highs. These elevated gradients minantly fine-grained source rocks to coarser-grained
probably result from a combination of connate water carrier and reservoir rocks is now generally consi-
flowing from the basinal areas to the flanking highs, dered to take place in a separate hydrocarbon phase
and higher heat conductivities on the flanking highs. (Jones, 1981). The limited solubility of oil-gravity hyd-
In the deepest sections of the graben system, only rocarbons in water and the reduced availability of
limited well data are available and maturation model- movable water in the hydrocarbon-generating zone,
ling using the Lopatin Method, as outlined by Waples argue against transportation in solution (Durand,
(1980), has been utilized. Two model profiles are pre- 1983). In biphasic flow, to overcome capillary pres-
sented for the graben axis in the area of blocks 34/9 sures of the normally water-wet shales, a continuous
and 35/2, and blocks 30/2 and 30/5 (see Fig. 5). In each oil-wet phase of oil (Meissner, 1978), or kerogen
instance, the model simulates burying a 300 m sec- (McAuliffe, 1980) is believed to be developed (Dickey,
tion of Upper Jurassic strata using the present geo- 1976). Expulsion of the hydrocarbon phase is aided by
thermal gradient. Allowance for decompaction was microfracturing, achieved by aquathermal pressuring
incorporated using the porosity-depth relationships of (Barker, 1972) and expansion of the organic phase
Sclater and Christie (1980), but water depths were upon hydrocarbon generation (Momper, 1978). Solu-
held constant. Depths and timing of each maturity tion in water of light aromatics and methane, and dif-
stage are noted on the graphs. fusion of low molecular weight n-alkanes are two
Depth-structure maps were combined with mat- alternative primary migration mechanisms for
uration data in each structural province to derive lower-weight hydrocarbons (Tissot, 1984; Leyth-
level-of-maturation maps for the tops of the Brent and aeuser et al., 1983). Whatever the primary-migration
Humber Groups (see Figs. 6 and 7). Within the Middle mechanism, the process is aided by interdigitation or
and Lower Jurassic gas-zone maturities (Ro greater juxtaposition of source and carrier beds, and is
than 1.3) are achieved in the Viking and Sogn graben impeded by a thick poorly drained source horizon.
systems and in the Mflre Basin (see Fig. 6). Oil- Secondary migration of oil and gas is generally
window maturity levels are predicted on the flanks of believed to occur as a discrete hydrocarbon phase
the graben system, on the Tampen Spur, and in the (Tissot and Welte, 1978). Flow of gas-bearing forma-
Stord Basin. In the Upper Jurassic, gas-zone maturity tion waters may be an important migration mechan-
levels are relatively more restricted within the graben ism for light hydrocarbons (Maximov, 1981), but on
system and in the Mflre Basin (see Fig. 7). Oil genera- release from solution the hydrocarbons act as a dis-
tion is taking place within the graben system, along its crete phase. The main processes acting to migrate the
flanks, and on the Tampen Spur. The northwestern hydrocarbon phase are buoyant rise and hydro-
part of the Stord Basin may be of sufficient depth for dynamic flow, while its movement is resisted by capil-
early oil generation, if the high geothermal gradients lary pressures. Secondary migration routes due to
of the Horda Platform and Utsira High extend across buoyancy can be derived from structure maps com-
the area. bined with sand-distribution maps of the probable
46 Petroleum Geochemistry in Exploration o/the Norwegian She{f

61'

EAST

60' SHETLAND

PLATFORM

59'

. 0 Kms 50
. ~ F*3 . .

58"N 58"N
0" 4" 5" 6"E

Fig. 6. Northern North Sea, level of maturation, top Brent Group"

conduit. Modification of these routes due to hydro- sures centred on the graben, which was divided into a
dynamic flow can be assessed through the con- number of pressure compartments. Barriers between
struction of potentiometric surface maps and subse- the compartments do not always follow obvious tec-
quent resolution of the buoyancy and hydrodynamic tonic discontinuities. Detailed potentiometric mapping
force vectors (Dahlberg, 1982). For regional mapping, within each pressure compartment is required to
however, construction of detailed potentiometric define the hydrodynamics of each area. Secondary
maps is limited by the presence of flow barriers not migration routes on Figs. 9, 10, 13, and 14, below, are
detected by the sparse pressure data available. based on buoyancy principles only. The flow of fluids
Chiarelli and Duffaud (1980) presented an equi- from overpressured graben areas to flanking highs is
potentiometric map of the Jurassic in the Northern anticipated to complement and reinforce the buoyancy
North Sea. They identified a zone of abnormal pres- vector.
Organic Geochemistry in North Sea Exploration 47

• I ' •

61'

EAST

60' SHETLAND

PLATFORM

59'

:.:.:.:o· ·u· JURASSIC THIN TO ABSENT


...... C:J IMMATURE - EARLY MATURE
.. G";:'I OIL WINDOW ( Ro .55 - 1.3 )
......... ~ GASZONE(Ro 1.3+,):'
. . . . . . . . . 0 . . . Km~ ..
50.·.·
'~.'

_______ J
Fig. 7. Northern North Sea, level of maturation, top Humber Group.

PREDICTION OF HYDROCARBON (110-850 scf/barrel). This results from the Upper


ACCUMULATION PROPERTIES Jurassic source envelope being predominantly in the
middle maturation zone and deeper buried, potential
Tampen Spur gas-source rocks, such as Dunlin shales, being domi-
nated by inertinite. The eastern margin of the Tampen
The properties of accumulated hydrocarbons result Spur has relatively higher gas-oil ratios of 65-500
from the combination of kerogen types, maturation (365-2800), and free gas caps are developed on the
states, and migration pathways in the drainage area. North Alwyn and Brent fields.
On the Tampen Spur, fields west of the Alwyn-Stat- In the Alwyn-Statfjord structural trend, the
fjord trend have low gas contents with no fr,ee gas caps gas-oil ratio of each accumulation is proportional to its
and gas-oil ratios in the range of 20-150 m3 gas/m 3 oil proximity to the deepest part of the graben in Quad-
48 Petroleum Geochemistry in Exploration of the Norwegian Shelf

rant 29, east of the North Alwyn field. Migration of tively (Tissot and Welte, 1978; Brooks et al., 1984 in
gas from the North Viking Graben toward the west is Cornford, 1984). Assuming a downdip source volume
facilitated by the low eastward dip of Jurassic strata of 7.5 km 3 of seam coal and conservative generation
and the decrease in pore pressure across the graben parameters, the potential for gas generation in the
margin. In the Brent reservoirs of the Brent field, the graben exceeds volumes discovered in accumulations
gas cap and oil leg are not in equilibrium. Studies indi- by up to one order of magnitude. Additional source
cate that only 1 million years would be required to potential is present in the overlying Heather shales.
equilibrate the phases, and strongly suggest that gas The Sleipner complex structures are full to spill-point,
is migrating into the structure at present (Bath et al., as might be expected since the coals are interbedded
1980). It is inferred that gas is migrating northwards with sands resulting in a high expUlsion and migration
from the North Alwyn area into the Brent Field, and efficiency. The asymmetry of the graben at Middle
that this gas is subsequently displacing oil northwards Jurassic levels (see Figs. 9 and 12) results in efficient
towards the Statfjord Field. migration of hydrocarbons released from the Sleipner
Oil in fields supplied from the southern flank of the coals and Heather shales eastwards into the Sleipner
M0re Basin show different properties (such as lower Complex. Little or no gas travels westward from the
gas-oil ratios for comparable depths) from those graben depocentre and the movement of gas vertically
farther south on the Tampen Spur, suggesting differ- upwards is restricted by the overlying shales having
ences in source-rock composition between the two higher pore pressures than the Middle Jurassic sand-
areas. Extensive migration of gas generated in the stones.
M0re Basin (Figs. 6 and 7) onto the Tampen Spur is Within the Upper Jurassic, the source-rock shales
restricted by the northward termination of Brent and reservoir sands are interbedded, resulting in a
Group sandstones, which form the major secondary- close relationship between oil type and source-rock
migration conduit. The Upper Jurassic reservoir of the kerogen type, and maturation. Reitsema (1983) has
Magnus Field is locally sourced from the enveloping detailed the kerogen types and hydrocarbon matching
Kimmeridge Clay Formation, which lies close to peak in the Brae area. The depths to which oil is preserved
oil-generation maturity levels. These reservoir sands in this area, considerably greater than 4000 m, is a
are isolated from gas-productive shales down dip to the function of this close source-reservoir coupling.
north. The synclinal axis of the Upper Jurassic is displaced
to the east of the Middle Jurassic axis, resulting in
significant volumes of Upper Jurassic sediments
South Viking Graben
draining to the west (see Figs. 9, 10 and 12). The vol-
Figure 8 summarizes the hydrocarbon properties of ume of source rocks downdip of the accumulations is
the major accumulations in the South Viking Graben. more than sufficient to fill graben-margin structures
The most apparent distribution pattern is the concen- to their effective spill-points, as indicated by accumu-
tration of gas fields in the east and oil fields in the west lations outside the graben such as the West Brae
of the graben. In the Sleipner complex, up to Field.
300 X 10 9 m3 of gas is recoverable from Middle It is concluded that the gross distribution of hyd-
Jurassic and Paleocene reservoirs. In the west of the rocarbon properties can be explained by the preferred
graben, Upper and Middle Jurassic reservoirs contain secondary migration routes in the Middle and Upper
up to 100 X 10 6 m3 of oil and 50 x 10 9 m3 of gas. This Jurassic fluid systems. These systems may not be
distribution of hydrocarbon gravities can be largely entirely separate, as they both have pore fluid over-
explained as a function of different secondary- pressuring to approximately 1.28 g/cm 3 • Thin, prob-
migration routes in the Middle and Upper Jurassic ably discontinuous sands exhibit higher pressures.
fluid systems. The lighter gravities of North Brae crude compared to
In the South Viking Graben, oil generation South and Central Brae (see Fig. 8) have been attri-
(Ro 0.55-1.3) occurs between approximately 3500 m buted to longer migration routes into the former
and 4900 m, with peak generation in type II kerogen (Reitsema, 1983). It is suggested here that the con-
(R 0 0.75) occurring at 4100 m (Pearson et al., 1983). centration oflighter-hydrocarbon accumulations in the
The rich oil-prone Upper Jurassic source rocks have north of the South Viking Graben is associated with
been in the oil window from Miocene times onwards, the syncline between East Brae and Gudrun, in which
peak oil generation occurring approximately 10 million the base of the Upper Jurassic reaches a maximum
years ago. Gas generation from Sleipner Formation depth for the area of approximately 6000 m (see
coals can be expected to take place at RoO. 7 and Fig. 8).
above, reaching a maximum at perhaps Ro 1.5
(Durand and Paratte, 1983), corresponding to depths North Viking Graben
of 4100 m and below 5000 m, respectively. The zones
of maturation at top Brent Group and top Humber As previously noted, the Humber Group is the prime
Group levels are mapped in Figs. 9 and 10, together potential source-rock horizon in the North Viking
with secondary-migration pathways. Graben. However, Brent Group coals with high
Sleipner Field gas is considered to have predomin- extract abilities and hydrogen indexes, have good gas
antly originated from Brent Group (Sleipner Forma- and light-liquid potential (see Fig. 2). Maturation
tion) coals (Larsen and Jaarvik, 1981). The isopach of states in the graben have been estimated from limited
these coals is presented in Fig. 11, and demonstrates well data and maturation modelling (see Fig. 5). This
that more than 30 m of seam coal is developed in the modelling results in an oil window between approxi-
axis of the graben. As noted previously, this coal has a mately 3600 m and 4800 m, and peak oil generation at
high hydrogen content and thus a high hydrocarbon- 4000-4200 m. Oil generation commenced in the
generating potential. Gas potential at an Roof 1. 5 var- Paleocene, and gas generation in the Miocene (see
ies between 1.2 and 14 x 10 9 m3 gas/km3 of 1% TOC Fig. 5).
source rock, for type III and type II kerogens respec- The areal distribution of these maturation zones has
Organic Geochemistry in North Sea E xploration 49

' · OO'E
59" 00 ' N r::-=--- -- - --,---y- - -- r - - - -- -----.
----

2 "00'E --I
59"00' N
16 , 15
.I
I

,
.I
.I
_ U.J_ I
BRAE ( W)
, /f' .' 35- 49° API
~<;./ GOR 530-1660 I,
\..2~' "
PAL. - DEli
23-32° API MILLER
GOR 46-17000

I
.I
,I

.f
{

, I
{
{
\


\

"OO'E
-- --- ---
o Kftl S .
58 ' 00' N '--_ _ _ _ _ _ __
10
_ _ _ _ __ _ __ _ _ _--L_ _...J
58' 00'N
2 " 00' E
N.B. ' GOR rn 3/rn 3

Fig. 8. South Viking Graben, hydrocarbon properties in major accumulations.


"OO' E 2 ' OO' E

16'.. .. ...
59'OO' N 59' OO' N
. . ,

. ..

... .. . .. ... .. . , .. . .

••••
.. ,

. .. .
.....
, ....
.

-
58 ' OO' N 58 ' OO'N
' "OO' E 2 " OO'E
KEY
Ro
.2 - .55 IMMATURE-EARLY MATURE OIL FIELD
.55 - .75 MIDDLE MATURE GAS + CONOENSATE FIELD
PEAK OIL GENERATION
~ .75 - 1.3 LATE MATURE -r- SYNCLINAL Axes
~ 1.3+ POST MATURE --......- MIGRATION PATHWAY

Fig. 9. South Viking Graben, top Brent Group, level of maturation and migration pathways,
1· 00' E 2 · 00'E

r<1~6~"·•.j •T•.:S:: ~"-,,,-,0~allll~


59· 00' N

59· 00'N

. · . ...
. . . . . .... . . . . .....
. ...... . ... .. .

.. ..., .... .. ..
. . .....
. · . ... . .. .
. .. . . .
· . ., . ......
. .. . . . .

, ...... -

58" 00'N 58 " 00' N


l "OO' E 2 " 00'E
Ro KEY

o
r-7'I
.2 - .55 IMMATURE-EARLY MATURE OIL FIELD
C&:..:..J .55 - ,75 MIDDLE MATURE
GAS + CONDENSATE FIELD
PEAK OIL GENERA nON

------
.75 - 1,3 LATE MATURE SYNCLINAL AXES
1,3 .. POST MATURE MIGRATION PATHWAY

Fig. 10. South Viking Graben, top Humber Group, level of maturation and migration pathways,
52 Petroleum Geochemistry in Exploration of the Norwegian Shelf

--- -,
'"OO'E
59"00' N r--------~--y"T'"'""":"""'""T"""T"'7"1":"~:o:"':"':":~~7.":'"~':":":'1
16

oE-3 Kms.
e--"""

58° 00' N L - - -_ _ _....!.-....l!....-..I..:....::..:=c::....:....:..=:..:...:..=:'-'-'-'-:....:....:......:....:....:....:....:....:..:....:....:......:....:....:.."'-'-'-':...:...:....:..:....:..I 58° 00' N


'"OO'E 2"00'E

Fig. 11. South Viking Graben, Brent Group, coal isopach.


Organic Geochemistry in North Sea Exploration 53

WEST EAST
U.K. NORWAY
SOUTH VIKING GRABEN SLEIPNER TERRACE
o~~.~.~.~
..~.~.~~~~~~~~~~.~.~~~.~.~
..~.~.~.~.~ ..~.~.~
..=.~.~.~.~.~.~.~~o
. . . . . . . . . . . ..
· .................. . , . . . . . . . . . . . . . . . . . . . . . . ..
..
. . . . ....... , ...... .
,
. . . . . . .. . . ...... . . . . . . . . . . . . . . . . . . . . . . . . ............ .. .
·..., ., ..., .., , ... ... ., , .. .
.......
. , .
, .. , ..
. . . .. .. .. . . . . , .... . . . . . .
,
. .
· . . . . . . . . . .. . . . .... ....... . . . .. .......
. .
. .. . . . .,. ,. ...
, , .... .
. .
. . . . . . . . ..... · .......... . . . . . . . ..
. . . .
. . . . . . . . .. ....... . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . .. .. . . . . . . . . . . . ., .. ., .....

<.: .< .: .:. :..::.: ...... .


TER.TIARY . .. :-:.... . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ' . . . . .
. . . . . . . . . . . .
. . , ........................................
...... , . ............... .
. . . . ..
2 . , . . . . . . . ..
. ..
. . . . .
. . . . . . . · ........ .. .. . . . . . . . . . . . . . . . . . . . . . ..... .
. . . . . . . . . . . . . . . . . · . .. ........... . . . . .
. . . . . . . . . . . .. .
.. . . . .
• • r • • • • • • • • , • • • • • •

. , .... , ... , . . . , . , . . . , .....


.

0··
.. Immature /:::::::1 Middle Mature Wttl Late Mature f:: ~ o Post Mature

Fig. 12. South Viking Graben, E-W cross section at 58°30'N.

been mapped at top Middle and top Upper Jurassic potential gas-source areas to the immediate west and
horizons. Using regional isopachs and the maturation possibly to the north, if sourced by spillage from the
depths calculated above, isopachs of oil-mature and Huldra area. The areas to the west of Oseberg are
gas-mature Humber Group sediments are outlined on probably sourcing gas into the structure, as is indi-
Figs. 13 and 14, respectively. When combined with cated by the presence of gas caps on the westernmost
probable secondary-migration routes, it is possible to fault blocks.
estimate the relative proportions of oil or gas entering The drainage basins of Huldra and Oseberg can be
a particular structure or accumulation. The distribu- treated quantitatively to calculate the volumes of oil
tion of oil and gas reserves in the study area can be and gas generated and the entrapment efficiency.
largely accounted for by these maps. Table 2 presents approximate volumes of Humber and
Significant oil reserves are present in Troll, Brage Brent Group source rocks in the maturity zones of the
and Oseberg Fields and in the North and South Beta drainage areas of these structures, based on Figs. 13
prospects in 30/3 and 30/6. Oil in the Troll accumula- and 14. A representative ultimate yield of oil per unit
tion has been largely sourced from the Horda Slope, volume (km 3) of 1% TOC source rock characterised by
north of the study area (see Fig. 7), with possible type II kerogen is taken as 6 x 10 6 m3 oil (Goff, 1983;
additional contributions through spillage of the Brage Sluijk and Nederlof, 1984; Brooks et al., 1984, in
Field. Brage is possibly sourced from oil migrating Cornford, 1984). For gas an equivalent yield of
through North and South Beta prospects. An alterna- 14 x 10 9 m3 gas is adopted. Brent Group coals, with
tive source of oil for Brage is from spillage of oil from relatively high hydrogen indexes, are estimated to
the southeastern closure of Oseberg, due to displace- have an ultimate yield of 275 x 10 9 m3 gas/km3 seam
ment of early oil by later-stage gas. Another alterna- coal. Representative TOC values as indicated in Fig. 2
tive is that oil migrates from the basin areas to the were utilized in the generation calculations.
southeast (see Fig. 13). Oseberg oil is probably The efficiency of entrapment can be calculated from
derived both from the Bergen Sub-basin to the north- the ratio C)f volumes of trapped oil and gas to those
east, and from the Viking Graben to the west. generated. The entrapment efficiency of Oseberg oil
Huldra, Troll and Oseberg Fields contain significant and gas is approximately 3%; Huldra gas has a lower
reserves of free gas. Gas in Troll has migrated from value of approximately 1%. These values compare
the Sogn Graben southwards up the Horda Slope into with the 20-30% efficiences noted by Goff (1983) for
Troll, which lies on the culmination of the Horda Plat- fields on the Tampen Spur. The overall low values of
form (see Fig. 7). Huldra gas and condensate can be efficiency suggest restricted drainage of oil from the
derived in significant quantities from basins to the source envelope, escape or retention of oil and gas on
west, north and east. This field represents the first migration routes, and escape of gas and oil from the
major structure on the migration routes from these accumulations through the caprock or by spillage at
source basins, and any early-accumulated oil has been, the closure. Inefficient primary and secondary migra-
at least partially, displaced southwards. Oseberg has tion from the Viking Graben and spillage from the
54 Petroleum Geochemistry in Explomtion of the N o1Wegian She(f

60'30' 60'30'

::::.: ::~ilFi~L~ ::::: ...... .


.:::: ::
(~r': G:A~·~· ~oN·D~N:s~ TE :Fui~~ :::
. . . . . . . :.: ....... : ·iSOPACHS Oil' GENERATING' .
. :: ~10~O-:- ... HUMBER GROUP (m) .::
:::~ SYNCLINA'l ·AXIS:··
·.·.·:·.·.~SECONciARY MIGRATION.'
. . . . . . . . . . . ·.PATHWAYS· ...

Fig, 13, North Viking Graben, Humber Group, isopach of oil-prone source rocks.

TABLE 2. North Viking Graben: drainage area


volumes TABLE 3. North Viking Graben: gas-oil ratios
(m 3/m 3 )
Source rock volume (km.1)
Source volume Source generation
Oil-prone Gas-prone Gas-prone
(m 8 gas-prone rock/m" (m 3 gas generated/m 3
Humber Gp HumberGp Brent Gp oil generated)
coal oil-prone rock)
shale shale
Oseberg 0.35 0.4 X 103
190 55 0.5 1.2 103
Oseberg Huldra 1.0 X
Huldra 150 120 1.1
----
Organic Geochemistry in North Sea Exploration 55

--~~4~-'~-----1
3'00'
'.-:-:-:-:':-:-:-. :-:- .
61'OO'N
..•.•.•.•.•. • ................ •. ",., i

I
\ .'

60'30'

::,:'::.,,,,,,, "", ..
-1000- ISOPACHS GAS GENERA TlNG
.. HUMB~R GROUP (m) .
':~:SYNCLINAL AXIS .....
. . . . , ... , ...... , ..
'.' .~.SECONDARY MIGRATION· . .
. . . . . . . . . '. PATH:-'!AYS . . . . .

'.: ·6 .. Kms 10 ..
I
160'oo'N
3'00' 3'40'E -.-JI
60'OO'N

Fig. 14. North Viking Graben, Humber Group, isopach of gas-prone source rocks,

structures eastwards onto the Horda Platform are minations result in a useful predictive capability in
conceived to be the prime reasons for the low entrap- untested areas.
ment efficiency,
The gas-oil ratios of the drainage areas in units of
rock and hydrocarbon volume are presented in Table CONCLUSIONS
3, For source productivity, it appears that ratios of
gas generation to oil generation greater than 1 (xl0 3) Potential source horizons are identified in shales of the
may lead to gas-dominated accumulations such as Kimmeridge Clay Formation, Heather Formation
Huldra. Ratios less than 0,5 (xl0 3) may result in oil- and Dunlin Group, and coals of the Brent Group and
dominated accumulations such as Oseberg, Although Statfjord Formation, The Kimmeridge Clay Forma-
the models require further refinement, particularly tion is widely developed and is the most important
with respect to migration efficiencies, such deter- source horizon on the Tampen Spur and in the graben
56 Petroleilln Geochemistry in Exploration of the Norwegian Shelf

system. Heather Formation shales may act as a sec- Chiarelli, A. amI Duffaud, F. 1980. Pressure origin and distribution in
ondary source in restricted basins and in the graben Jurassic of Viking Basin (United Kingdom-N orway) . .4..4.PG Hull.,
64, 1245-1250.
system where, due to their depth of burial, they have
Cooper, B. S. amI Barnard, P. C. 1984. Source rocks and oils of the
considerable gas potential. Brent Group coals have Central and Northern North Sea. In: Demaison, G. and Murris, R.
gas potential in the graben system and have probably J. (eds.), PetroleulII Geochrllli,slt,l/ al/d Basill Euaillatioll. AAPG
formed the source for the Sleipner gas in the South J!e III oil', 35, 303-814.
Cornford, C. 1984. Source rocks and hydrocarbons of the North Sea.
Viking Graben. Dunlin Group shales, although widely In: Glennie, K. W. (ed.), [llt/'Or/llelioll 10 Pclmlellill Gcolo.'lY oflhe
dominated by inertinite, apparently have enhanced oil Notth Sea. Blackwell Scientific, Oxford, pp. 171-204.
and gas potential in the graben depocentres and poss- Dahlberg, K 1982. Applie(i Hydl'UdYII(lIl1ic8 ill Petrulelllll
ibly in the Stord Basin. Exploratiol/. Springer-Verlag, New York.
The gross hydrocarbon properties of many accumu- De'Ath, N. and Schuyleman, S. F. 1981. The geology ofthe Magnus
Oilfield. In: Illing, L. V. and Hobson, G. D. (eds.), PetrulclIlII
lations can be explained by the maturation states of Geolo.'l!l(~fthe COl/til/elltal ShelfolNol'th- West Eum])e. Instituteof
their source areas, but these must be carefully defined Petroleum, London, pp. 342-351.
in terms of secondary-migration pathways. Examples Dickey, P. A. 1976. 'Possible primary migration of oil from source
are drawn from the Tampen Spur and South and rock in the oil phase. AAPG Bull., 59, 3:i7-345.
North Viking Grabens. In the latter area, quantitative Durant\, B. 1983. Present trends in organic geochemistry in research
on migration of hydrocarbons. In: Bjor0y. :VI. (e(l.), Adv(lIIces ill
calculations based upon sourcing from Humber Group Olpallie Geochel/li8ll'Y 1981. Wiley, London, pp. 117-128.
shales and Brent Group coals indicate minimum Durand, B. ancl Paratte, :VI. 1983. Oil potential of coals: a geochemical
entrapment efficiencies of 1-5%. Higher efficiencies approach. In: Brooks, J. (eel.). Pelmlelul/ Geoe-heolistl'!! alld
would be recorded if a more restricted volume of the Exploratioll off')ul'Ope. Geol. Soc. LOII(I. Spec. PuM. 12,255-265.
(Blackwell Scientific. Oxford.)
source envelope were being drained, or if the addi- Evans, T. R. and Coleman, N. C. 1974. :'>Iorth Sea geothermal
tional reserves in updip associated structures were gradients. Sature 247, 28-30.
included. Gas-oil ratios of the drainage areas in units Eynon, G. 1981. Basin development and sedimentation in the Middle
of rock and hydrocarbon volume can be compared Jurassicofthe Northern North Sea. In: Illing, L. V. amI Hobson, G.
with known accumulations. The predictive models D. (cds.), PetroleulII Geolu91l uf' the COl/tillel/tal Shelf' qf
Xorth-West Europe. Institute of Petroleum, London, pp. 196-204.
developed can be utilized in frontier exploration areas, Fuller. J. G. C. 1\1. 1975. Jurassic source-rock potential and
particularly in those with similar geological histories, hydrocarbon correlation, ,,"orth Sea. In: Ju !'assi,. .\'orlhe/'ll ,Vol'th
such as offshore Northern Norway. Sea Symp. 1975. Norwegian Petroleum Society, pp. 11/1-18.
Goff, J. C. 1983. Hydrocarbon generation and migration from Jurassic
source rocks in the East Shetland Basin and Viking Graben of the
l\orthcl'll North Sea. J. Geol. 80c. LOlld. 140,445-474.
ACKNOWLEDGEME NTS Grantham, P. J., Posthuma, J. and De Groot, K. 1980. Variation am\
significance ofthe C27 and C2H triterpane content of aN orth Sea core
The author wishes to thank Tenneco Oil Company for and various ,,"orth Sea crude oils. In: Douglas, A. G. and :VIaxwell,J.
R. (eds.), Adv((//('e,~ ill (h.qanic Geuchell/istry, 1979. Pergamon
permission to publish this paper. I am indebted to D. Press, Oxford, pp. 29-:38.
J. Griffiths for editing and C. S. Smeeton for typing Heritier, F. E., Lossel, P. and Wathne, E. 1980. Frigg Field-large
the text, and R. Salter and staff for draughting the submarine-fan trap in Lower Eocene rocks of the Viking Graben,
figures. I also thank colleagues in Tenneco for their North Sea. In: Halbouty, YI. T. (eel.), Gialll Oil (lInl GasFieids ofthe
technical contributions. fJecnde 1.968-1978. AAPG Jieowir, 30,59-79.
Jones, R. W. 1981. Some mass balance and geological constraints on
migration mechanisms. AAPG Bull., 65(1), 103-122.
Kirk, R. H. 1980. Statfjord Field-a North Sea giant. In: Halbouty,
REFERENCES 1\1. T. (eeL), Giant Oil I/I/(I Gas Fields of the Decade 1968-1978 .
.4..4.PG .IIelllOir, 30, 95-l1fl.
Albright, W. A., Turner, W. L. and Williamson, K. R. 1980. Ninian Larsen, R. :VI. and J aarvik, L ..J. 1981. The geology of the Sleipner
Field, U. Sector, North Sea. In: Halbouty, M. T. (ed.), Giallt Oil Field complex. In: .\'ol'l{'eqiul/ ".limp. 01/ E.rplomtioll (NSE-81).
al/d Gus Fields of the Decade 1968-1.978. AAPG .1Iel/lOir. 30, NorwPgian Petroleum Society, pp. l\SE/15, 1-81.
173-19:3. Lcythaeuser, D .. Schaefel', R. G. and Pooch, H. 1983. Diffusion of
Barker, C. 1972. Aquathermal pressuring-role of temperature in light hydrocarbons in subsurface sedimentary rocks. AAPG Bull ..
development of abnormal-pressure zones. AAPG Hull., 56, 67, 889-895.
2068-2071. Maher, C. E.1981. The Pipel'uilfield. In: Illing, L. V. ant) Hobson, G.
Barnard, P. C. and Cooper, B. S. 1981. Oils and source rocb of the D. (eds.), Petro/e(ull Geologll qf' the COl/tinenial Shelf of
North Sea area. In: Illing, L. V. and Hobson, G. D. (ells.), N orlh- Wcst Europe. Institute of Petroleum, London, pp. 358-370.
PetrolewlI Geulog!! of the COlllillclltal She(fofNorth-Wcst EUJ'Ope. :'IIaximov, S. P. 1981. Oil and gas occurrences on the East-European
Institute of Petroleum, London. pp. 169-175. platform as relate() to its tectonic history. J. Pelml. GeoZ., 3,
Barnard, P. C., Collins, A. G. and Cooper, B. S. 1981. Identification 445-453.
and distribution of kerogen facies in a source rock horizon McAuliffe, C. D. 1980. Oil and gas migration: chemical and physical
-examples from the North Sea Basin. In: Brooks, J. (ed.), Organic constraints. AAPG Siudies ill Geoluqy. No. 10, pp. 89-107.
.lin/umti(m Studies alld Fossil Fuels E;rplumtiol/. Academic _vleissner, F. F. 1978. Petroleum geology of the Bakken Formation
Press, London, pp. 271-282. Williston Basin, North Dakota and Montana. rn: Demaison, G. and
Bath, P. G., Fowler, W. :'>I. and Russel, M. P. 1980. The Brent Field, a Murris, R. J. (eds.). Pel I'Uleu III Geochemistry a lUI Ba"ill
reservoir engineering review . Ell t. Offshore COIlf. alld E:l'h ibitioll, Evaluatioll . .4.APG .Hel/lOir, 35, 159-179.
London. EUR 164, pp. 179-192. Momper, ,J. A. 1978. Oil migration limitations suggested by geological
Bissada, K. 1983. Petroleum Generation in Mesozoic Sediments ofthe and geochemical considerations . .4..4.PG Cuurse Notes Selies No.8.
Moray Firth Basin, British North Sea Area. In: Bjor0.\', M. (eel.), pp. , 1-60.
Adv(!IIcesi II a rgani cGeoche mio$tr,l/1981. Wiley, Lomlon, pp. 7-15. Oudin, ,J. L. 1970. Etude geochimique du bassin de la Mer du Nord.
Bowen, J. :VI. 1975. The Brent oil-field. In: Woodland, A. W. (ed.), Bnl/. Ceilire Rerh. Pall, 10, 889-:{58.
Petroleum and the COlltinelltal Shell ~r Notth-West Ell rope, Pearson,:VI. J., Watkins, D., Pittion, J. L., Caston, D. and Small, J. S.
Vol. 1., Geology. Applied Science Publishers,
London, 1983. Aspects of burial diagenesis, organic maturation and
pp. B53-361. palaeothermal history of an area in the South Viking Graben, North
Brekke, T., Pegrum, R. M. and Watts, P. B. 1981. First exploration Sea. In: Brooks, J. (er1.),Pelroleuul Geuchelllistryaw/ Exploration
results in Block 31/2. In: Norwegiall Syll/p. Oil Kxplumliol/ of' Europe. Geol. Sut. Loud. Spec. Pub!. 12. 161-173. (Blackwell
('ySE-8J). Norwegian Petroleum Society, pp. NSE/16, 1-84. Scientific, Oxford.)
Carstens, H. and Finstarl, K. G. 1981. Geothermal gradients ofthe Reitsema, R. H. 1983. Geochemistry of North and South Brae Areas,
Northern North Sea Basin 59-62°N. In: Illing, L. V. and Hobson, North Sea. In: Brooks, .J. (ed.), Petl'OlfUll1 Geochemist)'!! awl
G. D. (eds.), Peirolelwi Geolo.'lY ol the COlltillellt((1 She(l of E:rplomtioll qlf')umpe. Geol. Soc. Low/. Spec. Publ. 12,203-212,
N otlh- West EUI'ope. Institute of Petroleum, London, pp. 152-161. (Blackwell Scientific, Oxford.)
Organic Geochem.istry in North Sea Exploration 57

Sclater, J. G. and Christie, P. A. F. 1980. Continental stretching: an Ungerer, P., Bessis, F., Chenet, P. Y., Durand, B., Nogaret, E.,
explanation of the post mid-Cretaceous subsidence of the Central Chiarelli, A., Oudin, J. L. and Perrin, J. F. 1984. Geological and
North Sea Basin. Geophys. Res., 85,3711-3739. geochemical models in oil exploration; principles and practical
Skarpnes, 0., Hamar, G. P., Jakobsson, K. H. and Ormaasen, D. E. examples. In: Demaison, G. and Murris, R. J. (eds.), Petroleum
1980. Regional Jurassic setting of the North Sea, north of the Geochemistry and Basin Evaluation. AAPG Memoir, 35, 53-77.
Central Highs. In: The Sedimentatian of the North Sea Reservoir vanden Bark, E. and Thomas, O. D. 1980. Ekofisk: first of the giant oil
Rocks. Norwegian Petroleum Society, pp. XIII, 1-8. fields in Western Europe. In: Halbouty, M. T. (ed.), Giant Oil and
Sluijk, D. and N ederlof, M. H. 1984. Worldwide geological experience Gas Fields of the Decade 1968-1978. AAPGMemoir, 30, 195-224.
as a systematic basis for prospect appraisal. In: Demaison, G. and Waples, D. 1980. Time and temperature in petroleum formation;
Murris, R. J. (eds.),· Petroleum Geochemistry and Basin application of Lopatin's method to petroleum exploration. AAPG
Evaluation. AAPG Memoir, 35, 15-26. Bull., 64, 916-926.
Tissot, B. P. 1984. Recent advances in petroleum geochemistry Ziegler, P. A. 1982. Geological Atlas ofWestern and Central Europe.
applied to hydrocarbon exploration. AAPG Bull., 68, 545-563. Elsevier, Amsterdam.
Tissot, B. P. and Welte, D. H. 1978. Petroleum Formation and
OcCUrrence. Springer-Verlag, Berlin.
4 Oil-generating coals
s. Thompson*, B. S. Cooped, R. J. Morley* and P.C. Bamard*
* Robertson Research International Ltd, Ty'n-y-Coed, Llanrhos,
Llandudno, Gwynedd LL30 lSA, UK
tBSC Associates, The Beeches, Byfield Road, Chipping Warden,
Nr. Banbury, Oxford OX17 lLD, UK.

Coals are oil source rocks in many of the Tertiary basins of Southeast Asia. The precursors of these hydrogen-rich
and oxygen-poor coals are coastal plain peats which have mainly developed in an everwet and tropical climate. In
these environments water flow and reworking can concentrate liptinitic kerogen in preference to vitrinitic
kerogen. The distribution, petrography and chemistry of the coaly Miocene source rocks present in the Kutai Basin
are described. Studies of Tertiary and contemporaneous coastal plain deposits can be applied to trends in source
rock distribution in offshore Norway. Comments on the environment of deposition of coaly sediments in the basins
of the Norwegian Sea are discussed with reference to their possible oil and/or gas sourcing potential. The
Triassic-Jurassic coals of the Haltenbanken area may become more oil-prone towards the delta margins, and facies
mapping could aid oil exploration in this area.

INTRODUCTION of Indonesia. In Kalimantan, everwet peat deposits


extend over several thousand square miles, and their
The most productive oil source rocks of Indonesia environment and development, including the more
are drift coals and associated shales deposited within restricted areas likely to form very good quality oil
tidal dominated coastal plain environments during the source rocks, are a suitable analogue for discussion.
Tertiary. Present analogues for environments in Triassic-Jurassic coals are relatively common in the
which such sediments accumulate are to be seen in Haltenbanken area of northern offshore Norway, and
many parts of Southeast Asia where the processes of an area of delta plain deposits containing thick coals
the sedimentation of organic matter have been has been described (RI3nnevik et al., 1983); it is prob-
studied. These studies of Tertiary and contemporane- able that coal-bearing deposits of this age and facies
ous coastal plain deposits can be used as a guide to are extensive in the Norwegian Atlantic offshore nro-
trends in source rock distribution in offshore Norway vince. In this paper we try to compare the coals of
where coastal plain deposits occur in the Triassic and this area with those of the more extensively examined
Jurassic. Indonesian swamp environments. For the purpose of
The possibility that coals may act as oil source demonstration, pyrolysis data from several Indonesian
rocks has been widely discussed over the last few coals are presented, and the oil potential of such coals
years (Thomas, 1982; Durand and Paratte, 1983), is discussed.
those of Indonesia having been evaluated by Roe and
Polita (1979) and Durand and Oudin (1979). It is doubt-
ful that significant quantities of oil are generated from PRESENT DAY COASTAL PLAIN
vitrinite-rich coals; Boudou et al. (1984) suggest that PEAT-FORMING ENVIRONMENTS IN
when kerogen is rich in oxygen, carbon dioxide and SOUTHEAST ASIA
water will be evolved in preference to hydrocarbons.
However, it is clear that many coals, such as some of The distribution of peats in Southeast Asia is control-
the liptinite-rich examples described by Teichmuller led essentially by climate (se~ Fig. 1). Areas which
and Durand (1983), are hydrogen-rich and oxygen-poor yield peats greater than 1 m in thickness are more-or-
at relatively low levels of thermal maturity. less wholly confined to everwet regions or areas
Hydrogen-rich coals frequently occur in the subsur- exhibiting minor seasonality of climate (Morley, 1981).
face in Indonesian basins, and accumulate as a result Outside these areas, humic organic matter is broken
of everwet and tropical conditions which have per- down by desiccation and oxidation during dry seasons,
sisted through the Tertiary to the present over much and true peats do not form. Peat formation occurs
Petl'uleum Geochemi8try in Exploration ~(the Norwegian Shelf© Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 59-73.
Fig. 1. Distribution of lowland peats in the Sunda region of Southeast Asia (after Morley, 1981).
Oil-generating Coals 61

over the upper coastal plain as far as the intertidal Oligocene or Miocene coastal plain deposits of these
zone, where they are either bounded by beach ridges basins frequently comprise interbedded shales, sand-
or give way to mangrove swamp. They are domed stones and coals. The coals and shales are often
(ombrogenous), their water supply being maintained good-quality oil source rocks, and not infrequently the
by rainfall only, and may reach up to 20 m in thickness coals have better oil generating capacity than the
(Anderson, 1964). They bear a raised water table and shales. Although the samples described in this paper
may be compared with blanket bogs of temperate come from several of these basins, for the purposes of
areas. descriptive convenience, and because of the availabil-
The flora of Southeast Asian peat swamps consists ity of published information, we have concentrated on
predominantly of trees. The vegetation is strongly the Kutai Basin, which underlies the present
zoned and may be highly diverse on thin peats, but Mahakam Delta of Kalimantan.
species-poor over thick parts (Fig. 2). Many of the The basic geochemical data of the Mahakam Delta
trees, such as members of the angiosperm family have been reported by Combaz and de Matharel
Dipterocarpaceae, and the conifer A,qathis (1978), Durand and Oudin (1979) and Verdier et al.
(Araucariaceae) produce highly resinous wood. As (1980). A correlation of the oils to the coaly kerogens
peats accumulate, bacterial activity occurs throughout in the Kutai Basin (Mahakam Delta) in general, and
the peat, resulting in the breakdown of most woody the Balikpapan Group in the Handil field in particular,
material to a water-soluble semi-liquid humic acid gel. has been published by Schoell et al. (1983).
Peat accumulation within prograding coastal plain
environments has a profound effect on the manner of
sediment accumulation. Peat accumulation takes place DISTRmUTION OF BALIKP APAN GROUP
much more rapidly than levee development, and hence SOURCE ROCKS
any depressions within the upper coastal plain rapidly
become filled. Lakes, lagoons or crevasse splay The Balikpapan Group, which is of Middle Miocene,
deposits within interdistributary bays are thus rare. but locally also Late Miocene, age, represents a phase
The peats are, however, highly unstable and are easily of delta progradation which has continued, with
eroded by the scouring action of tidal channels and transgressive interruptions, to the present day
shifting water courses in the upper coastal plain. In Mahakam Delta phase of deposition (Oudin and
strongly tidal deltas, autochthonous peats are rarely Picard, 1982). Sediments have been deposited from
preserved as sediments. Once eroded, humic acid gels one dominant eastward-flovving river system into a
go into solution, whereas water-insoluble fragments, subsiding basin with its north-south axis parallel and
such as plant cuticle, plant waxes, resins and humified close to the delta front, the axis migrating eastwards
woody fragments become part of the suspension or with time after an early basin inversion. A varied
traction load. Particulate organic matter is deposited suite of facies representing upper delta plain to delta
in intertidal or inner shelf shallow water depths as front environments with occasional transgressive
streams lose energy when they reach the sea, whereas limestones is present. The distribution of sediments
the soluble humic acid component will generally be and their contained organic matter has been strongly
carried to greater water depths before precipitation influenced by the interaction of tides and rivers which
(as the precursor to amorphous or unstructured vitri- has concentrated the deposition of organic material
nite) or oxidation. The process of separation of soluble within tidal flats and interdistributary areas of the
and insoluble organic matter results in the accumula- lower coastal plain, and has resulted in the deposition
tion of liptinite-enriched allochthonous (drift) peats of drift coals and black shales. The Balikpapan Group
and carbonaceous shales in estuarine or shallow source rock sequence at its richest includes some
marine environments along the coastal margin. This 175 m (cumUlative) of coal and nearly 1750 m (cumula-
process can currently be observed in the modern tive) of shale. The source rock sequence is in a zone
Mahakam Delta, where allochthonous peats are depo- trending north-south parallel to the lithofacies
sited adjacent to beach ridges (Allen et al., 1979). trends, and probably thins at the northern and south-
These liptinite-enriched peats may be further ern margins of the main area of deltaic sedimentation.
enriched by fresh liptinitic material derived from veg- There are also rapid variations both to the east and to
etation growing behind the coastal plain, of which the west of the main north-south source rock zone,
resin-producing trees may form a large component. the shales becoming thinner and vitrinitic, and the
Peats preserved within upper coastal plain deposits coals becoming thinner and inertinitic (see Fig. 4).
tend to retain their initial high proportion of humic The coaly sequences of the Balikpapan Group of the
acid gels, and are more likely to form vitrinite-rich Kutai Basin extend some 230 km NE-SW and about
lignites during burial. Such sediments are potentially 180 km NW-SE underneath the Mahakam Delta.
gas source, rather than oil source, rocks. However, However, the area where coals (and shales) of good oil
unusually high rates of water flow may flush out humic source quality occur in this group, extends about
acid gels from these in sit7t peats, and in this way also 175 km NE-SW and about 55 km NW-SE, and is
give rise to enrichment in liptinitic kerogen. broadly equivalent to the lower delta plain back
swamp and intertidal areas. In other basins where a
seasonal influence is more marked, black (and often
TERTIARY COASTAL PLAIN DEPOSITS algal-rich) shales deposited in semi-permanent lakes
IN INDONESIA and lagoons, occur within the coal-bearing sequence.

Most of the sedimentary basins of western Indonesia


contain coastal plain deposits which have significant EXAMPLES OF INDONESIAN COALS
oil source potential. Of these basins, the foremost oil
producers are the Central and South Sumatra, Sunda, Coal samples have been selected from the principal oil
Northwest Java, and Kutai basins (see Fig. 3). The source rock formations of the Kutai Basin (three
~

MIXED SWAMP FOREST PADANG FOREST PADANG


FOREST
(ADVANCED ""0
~
STAGE) C1
~

;:i
'"
o
8
~
'"'
40 '"~.
'"
36 J
(f) Sj'
W t.>:l
0:: 24
I- ~
W
c
::2: a::
18
iio·
w ;;:s
.~
z 12
> s.
a:: '"<-
o-
l- 6
I j
<.9

;::;
W 0- PEAT
I if!
~
'-.
I
'"
~
I- 6
a... MINERAL SUBSTRATE
W
o
Fig. 2. Idealized cross-section of a domed peat swamp. Mixed swamp forest-high, species-rich forests on thin (1-3 m) peats. Padang
forest-low, species-poor forests un thick (> 3 m) peats.
PACIFIC OCEAN

,
SOUTH CHINA SEA

2
~
~
~
~.
INOIAN OCEAN
C1
~
&;"
/J
@
.'
()
o

LEGEND

Sedlmentarv Basin C?D

Fig. 3. Sedimentary basins of the Indonesian area.

<::>
<:.:>
64 Petroleum Geochemistry in Exploration of the Norwegian Shelf

116°E 117'E 118°E 119°E


1°N.---------------------------------.-------------------------------~r_------------------------------_,1°N

EQUATOR
0° 0'

1'Sr-------------------~0~E:~~~t=~~~~~tF~-------_t--------------------------~

2oSL-______________ ~ ________________L __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _L __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___J2'5

116 Q E 117°E llBoE "goE

SCALE 1 : 2 000 000

LEGEND

Back delta plain . Prodelta . . . . . ~


Mid delta plain with very good 'Duree rock' . . ... rewA) Marine shell. .. .. ................ [0
Delta plain with good source rocks . . . . . . . . . . ~
Delta front.

Fig. 4. Di8tribution of Miocene source rocks in the Kutai Basin.


Oil-generatill.f1 Coals 65

wells), Northwest Java Basin (two wells) and Sunda and interpreted kerogen composition data are pre-
Basin (one well) of western Indonesia. A key to the sented in Table 3.
samples, which comprise 12 coals and two mixed
lithologies, from Oligocene and Middle Miocene forma-
tions, is presented in Table 1. Sufficient published METHODS
work exists for some comparisons to be made between
the coals of the Haltenbanken area and the Indonesian Organic carbon analysis was conducted on a Leco car-
coals discussed here. The chemical data for the bon analyser, and is inclusive of ash content. An
samples are presented in Table 2, and the maturity IFP-Fina 'Rock-Eval' apparatus was used for

TABLE 1. Key to samples

Sall/ple code Illdonesiall Basin Depth (ji) FOl'lllation Age

B-1 Kutai 3000 Balikpapan (Gp) Middle Miocene


B-2 Kutai 4780 Balikpapan (Gp) Yliddle Miocene
B-3 Kutai 7710 Balikpapan (Gp) Middle Miocene
H-l Kutai 5950 Balikpapan (Gp) Middle ;VIiocene
H-2 Kutai 8230 Balikpapan (Gp) Middle Miocene
H-3 Kutai 8660 Balikpapan (Gp) Middle Miocene
P-l Kutai 3800 Balikpapan (Gp) Middle :I1:iocene
P-2 Kutai 8800 Balikpapan (Gp) Middle Miocene
C-l Northwest Java 7260 Lower Cibulakan Oligocene
C-2 Northwest Java 7410 Lower Cibulakan Oligocene
C-3 Northwest Java 7520 Lower Cibulakan Oligocene
8-1 Northwest ,Java S560 Lower Cibulakan Oligocene
8-2 Northwest Java 8920 Lower Cibulakan Oligocene
Y-l 8unda 6420 Talang Akar Oligocene

TABLE 2. Chemical data

Rock-Evai pyrolysis data

To/rLlol'yanic Tmax Potential yield


S(tll/pie code Lithology alia! ysed carboll ('Yo) (OC) HI 01 Production iwle;r (8) (ppm)

B-1 Coal 51.7 415 246 21 0.02 127300


B-2 Coal 62.8 425 263 10 0.03 165400
B-3 Brn-blk claystone 12.S 434 360 12 0.08 46280
+ 10% Coal
H-l Coal 58.6 422 347 10 0.12 203650
H-2 Coal 69.6 442 354 5 0.06 247050
H-3 Coal 72.1 437 355 21 0.11 256280
P-l Coal + 20% 21.8 417 348 8 0.08 75870
brn-blk shale
P-2 Coal 43.1 427 293 5 0.07 126490
C-l Coal 66.1 444 400 13 0.09 265000
C-2 Coal 51.9 445 446 21 0.04 232000
C-3 Coal 26.9 443 389 1 0.09 104700
8-1 Coal 63.6 449 351 6 0.08 223800
8-2 Coal 37.6 447 305 9 0.07 114 750
Y-l Coal 49.7 439 461 12 0.02 229600

TABLE 3. Maturity and kerogen composition data

Interpreted kemgen cUlllpu,r'i'iulI

Lithology Vitrillite Spore colour Wa.,y (I + II)


Sall/ple code analysed rejlecti oity 'ill de;)' Illtertillite Vit"illite Algal saprupel kerogell

B-1 Coal 0.42(21) 2.5 60 15 0 25


B-2 Coal 0.45(22) 3.5-4.0 55 15 5 25
B-2 Claystone, 0.65(23)b 5.0 a 50 35 0 15
brn-blk +
10% coal
H-l Coal 0.43(") 4.5-5.0 40 15 20 25
H-2 Coal 0.40(a) 5.0-5.5" 35 5 0 60
H-3 Coal 0.49(28)" 5.0-5.5 e 20 25 10 45
P-l Coal + 20% 0.52(22) 3.5 a 40 25 15 20
brn-blk shale
P-2 Coal 0.62(23) 5.0-5.5" 30 35 15 20
C-l Coal 0.59(30) 4.0-4.50 30 20 0 50
C-2 Coal 0.60(80) 4.5-5.0 a 10 40 0 50
C-3 Coal 0.63(29) 4.5-5.0 c 25 25 0 50
8-1 Coal 0.63(22) 6.0 a 25 I) I) (i5
8-2 Coal 0.65(22) G.O" 80 15 0 55
Y-l Coal 0.48(") 4.0" 30 15 5 50

"Estimated from gradient. b Measured on coal. cMeawl'ed un kerogen of mixed cuttings.


66 Petroleum Geochemistry in Exploration of the Norwegian Shelf

pyrolysis analysis using 20 mg aliquots of crushed relatively high contents of clays. Calculation of
sample. Vitrinite reflectivity was measured on a Leitz kerogen composition with respect to the hydrogen and
Orthoplan microscope against glass standards. Spore oxygen indices of theoretical type I, II, III and IV
colour was measured under transmitted light on unox- kerogen end members (Barnard et al., 1981) indicates
idized kerogen fractions using a standard palynologi- that vitrinite is subordinate to oil-prone (type I and II)
cal microscope with incident u. v. light attachments. kerogens. The calculated liptinitic (I and II) compo-
nent of these coals ranges between 45% and 65% and
the vitrinite contents are generally below 15%. How-
PETROGRAPHY OF THE ANALYSED ever, incident u. v. light microscopy does not reveal
SAMPLES anything like 45% of fluorescent components, although
yellow fluorescent exsudatinite and orange-yellow
Under incident light, the coals appear to be composed fluorescent algal, spore and cuticle debris are com-
mainly of vitrinite, although layers of liptinitic mater- mon. Again, this shows that the groundmass contains
ial are not uncommon. The areas of coal with lower a mixed vitrinite and liptinite composition. The lipti-
reflectance and lower relief than semi-fusinite, and nite of such kerogen compositions either in coals or
higher reflectance than liptinite (e.g. sporinite) are disseminated in shales, acts as an oil source (Snowdon
usually given the generic name vitrinite, and their and Powell, 1982).
reflectance is measured. However, the term vitrinite
may include up to six or more sub-type (maceral
types) of humic kerogen, all with different CHARACTERISTICS OF THE OILS
morphology, reflectance, and chemical composition GENERATED FROM COALS AND
(Stach, 1975). ASSOCIATED SHALES
Apart from normal vitrinite, two other types are
frequent or even dominant. The reflectance of The oils of the Indonesian basins discussed here are
vitrinite in the samples analysed is often relatively often waxy and associated with condensates and gas.
high, particularly at low levels of maturity, with They have been described by Sutton (1979), Roe and
respect to the spore colours. Benedict et aZ. (1968) Polito (1979), Seifert and Moldowan (1981) and Schoell
described anomalously high reflecting vitrinite in et al. (1983). Those generated at a relatively low level
Appalachian coals and called it pseudo-vitrinite. They of thermal maturity have very high pristane contents
consider that it results from alteration, probably by and a marked odd-over-even preference in the gener-
partial oxidation, either before or during the peat ally pronounced C27 to C33 waxy component (see
stage of coalification. Thompson and Benedict (1974) Fig. 5). At higher levels of maturity these features
noted that this material did not behave like vitrinite in tend to become less pronounced or absent. In most of
terms of coking properties, being more inert. The the basins, the carbon isotope ratios of the C 15 +
morphology of pseudo-vitrinite is similar to that of the fractions of these oils are around -30%(, PDB,
high reflecting vitrinite noted in the analysed samples. suggestive of derivation from higher plant kerogen. In
In many cases the vitrinite ofthe Indonesian samples the few Indonesian basins in which Tertiary coastal
resembles the coal maceral desmocollinite, described plain deposits have accumulated under the influence of
by Teichmiiller in Stach (1975), which has a relatively a more marked seasonal climate, organic-rich
low reflectance. Electron micrographic studies by lacustrine and lagoonal shales are developed as source
Taylor (1966) have suggested that this maceral is vit- rocks in the mid-coastal plain. They yield oils with
rinite impregnated with submicroscopic liptinite, poss- carbon isotope ratios of around -22%(, PDB, which
ibly resinite, intercalations. In some cases the Sutton (1979) attributed to salt marsh biota.
groundmass of the 'vitrinite' exhibits fluorescence, Analysis of the cyclic hydrocarbons by gas
and the inclusion of amorphous type II (liptinitic) chromatography-mass spectrometry reveals a
kerogen in a non-fluorescent matrix derived from number of wax and resin-derived components in oils
humic acid gel (the precursor of vitrinite s.s.) would from Indonesian basins (Grantham et aZ., 1983). The
seem a probable explanation. presence of these hydrocarbons, which are considered
Hence, although apparently vitrinite-rich, the to be the maturation products of resins, is best
pyrolysis analysis of these coals in terms of types I, II detected using mass fragmentography. The mass
and III kerogens may indicate a different composition. fragmentograms of (a) an Indonesian oil generated
from a coaly source rock, (b) an Indonesian oil gener-
ated from a non-coaly source rock and (c) a coaly
CHEMICAL CHARACTERISTICS OF THE source rock, are presented in Figs. 6, 7 and 8, respec-
ANALYSED SAMPLES tively. The mass spectra of these resin-derived hyd-
rocarbons have base peaks at mle 163, and sub-
Chemically, the coals of coastal plain environments of dominant peaks at mle 259, 217 and 191. Hence the
Indonesia are different from the Carboniferous and presence of these hydrocarbons is clearly shown by
Tertiary coals of North-West Europe, in that the bulk the mle 163 fragmentogram, which in their absence
of their kerogen, which has low relief and is grey only detects the hopane series, as in Fig. 7.
under reflected light, is hydrogen-rich and oxygen-
poor. These properties are clearly demonstrated by
the pyrolysis data (see Table 2) in that the hydrogen OIL SOURCE ROCKS IN THE NORWEGIAN
indices of these coals are often above 300, and the OFFSHORE AREA
oxygen indices are usually below 15. By comparison,
vitrinitic Carboniferous coals of North America and It has previously been demonstrated by a number of
Europe at similar levels of maturity may be expected workers (Cooper and Barnard, 1984; Goff, 1984) that
to have hydrogen indices of about 150, and oxygen one of the most important oil source rocks in the Cen-
indices of about 35. These Tertiary coals also contain tral and Northern North Sea is the Late
Pristane

2
{;s
29 27
~
I'
C")
c
~
'"

I !

I
l'
il II !
~ IiI I I
I . i
I
Jl~c.LcJl,~, I
JlJ,j
LI~vol 'llLc~ 'L Joe) ,
l~'
~il~L~J JLJUlu wJ~~v~l~","~li,lJ.llt~
0
Fig. 5. Gas chromatogram of a waxy oil (34 API gravity) from Indonesia.

a>
-'l
68 Petroleum Geochemistry in Exploration of the Norwegian She(f

SCANS 694 TO 1794

100.€1
R
R

191 C31
,......, ±
191.191
a.see

v......,phlp,..J.r\w~_.....~.;A-. ........

69.5

217 217.217
::t Ia.see

34. ~,

259 259. 2'5~

~~
± e. 5e~

I I
188.13

163 163.16:
± 0.S01

800 1000 1200 140~3 1600 SCAN


1 ~:: 20 16:4(1 2'0:013 2:3;28 26:40 TIME

Fig. 6. Mass fragmentograms (mje 191, 217, 259, 163) of alkanes of an Indonesian oil generated from a coaly source rock. Key: R =
resin-derived cyclic alkane; 0 = Oleanane; C30 = C30 hopane; C31 = 22S and 22R C31 homohopane.

Jurassic-Early Cretaceous Kimmeridge Clay Forma- and on Svalbard immature through to post-mature
tion. Oil source rocks of equivalent age have been very good quality Early and Middle Triassic black
described in the subsurface in Haltenbanken (Hollan- shales have been well-documented (M(lrk and Bjor(ly,
der, 1984) and in Troms(l (Westre, 1984). Similar aged 1984).
source rocks have also been postUlated in Trrenaban- In the context of this paper, however, the occurr-
ken (Larsen and Skarpnes, 1984) and in the undrilled ence of coal and carbonaceous shale sequences in the
M(lre Basin (Hamar and Hjelle, 1984). Other source Norwegian offshore is of greaest interest. The car-
rocks are known in the areas fringing the Norwegian bonaceous Early and Middle Jurassic sequences have
Sea, including post-mature Late Permian black shales been known for many years in the North Sea Basin,
which outcrop in East Greenland (Surlyk et al., 1984) and on occasions laterally discontinuous drift coals
Oil-qenerating Coals 69

SCAt~S 724 TO 1824

11313.13

191 191.191
± 0.51313

4.4

217 217.217
± 121.500

3.1

I I
259
~I II ±
259.259
13.51313

I~H_
~.9

163 163.163
± 0.500

80e 101313 1200 14130 160121 180121 SCAN


13:213 16:40 20:00 23:20 26:40 30:00 Tlr1E

Fig. 7. Mass fragmentograms (l11/e 191.217.259,163) of alkanes of an Indonesian oil generated from a non-coaly source rock. Key: Cao ~ Cao
hopane; C31 ~ 228 and 22R C31 homohopane.

have been found to contain significant enrichments of ciently buried to generate major amounts of oil. By
liptinitic material. The contribution of these source contrast the Late-Triassic-Early Jurassic coaly se-
rocks to discovered hydrocarbon accumulations in this quence is both thick (greater than 500 m) and deeply
area is not thought to be very great, although several buried (greater than 3000 m).
of the larger gas with heavy oil accumulations in the A postulated reconstruction of the Late
Norwegian sector have not yet been attributed to the Triassic-Early Jurassic basin is shown in Fig. 9. On
usual Kimmeridge Clay Formation source rock. the eastern margins of the basin along the coast of
In the more northerly drilled areas of Haltenbanken Norway red sandstones and shales with several inter-
and Troms~, the Kimmeridge Clay Formation equi- vals of halite were deposited during the Triassic. In
valent appears to be regionally rather thin and insuffi- the Rhaetic, the climate appears to have dramatically
70 Petroleum Geochemistry in Exploration of the Norwegian She!f

SCANS 635 TO 1935

100.0

o
191 191.191
± 0.500

26.1

217 217.217
± 0.500

13.1

2:59 259.259
± 0.500

97.3

163 163.163
:i: a.see

1000 1200 1400 SCAN


16.40 20.00 23.20 Tll1E

Fig. 8. Mass fragmentograms (mje 191, 217, 259, 163) of alkanes of a Miocene coal from Indonesia. Key: R = resin-derived cyclic alkane; 0 =
Oleanane; C30 = C30 hopane; C 31 = 22S and 22R C31 homohopane.

changed (Jacobsen and van Veen, 1984), and this margins at the same time as a major marine trans-
change may be associated with the progressive gression began. The marine margins of the coastal
northward drift of the continental mass. At this time plain environments are not recognized in the latest
Central and Southern Norway are reported to have Triassic sediments in the North Sea Basin or the west
been at a latitude of about 30 N with a subtropical
0 of Scotland. In these areas, only non-marine fluvial
climate, while the northern province of Svalbard was and upper delta deposits are seen. Their coastal plain
in a temperate regime at a latitude of about 45°N. and marine equivalents are presumed to be present in
The change in climate from an arid to an apparently the subsurface in the M0re Basin and in the Faeroes
very wet climate in tbe Norian-Rhaetian, resulted in Trough.
the establishment of significant fluvial and deltaic sedi- From Late Triassic through to Cretaceous time, the
ment transport and depositional systems on the basin area presently covered by the Norwegian Sea was a
Oil-generating Coals 71

DELTAIC-

MARINE
J;l I Barents ~ _ - -
---- ~-~~
- - - ---~TEMPERATE/ MARINE

~ SEDIMENT SOURCE
.,.,--
;/DELTAIC-
I MARGINAL (--\S-~
IMARINE

\ {:!}
(".J-
GREENLAND ?
MARINE .J
?
~

MARINE
/--
/

(i
r! ~
North Sea

Fig. 9. Palaeolatitudes and palaeoclimates of the Norwegian area during the Triassic and Jurassic.
72 Petroleum Geochelllistry in Exploration of the Norwegian Shelf'

rather narrow marine embayment in which strong Barnard, P. C., Collins, A. G. and Cooper, B. S. 1981. Identification
tidal influences are expected to have occurred. By and distribution of kerogen facies in a source rock hori-
zon-examples from the North Sea Basin. In: Brooks, J. (ed.),
contrast, the Barents Sea and Svalbard areas were on Organic Maturation Studies and Fossil Fuel EJ:ploratioll.
the margins of a large oceanic system where tidal Academic Press, London, pp. 271-282.
effects are likely to have been limited. In the Halten- Benedict, L. G., Thompson, R. R., Shigo, J. J. III and Aikman, R.
banken area the delta top coaly sediments with rooted P. 1968. Pseudo vitrinite in Appalachian coking coals. Fuel, 47,
125-143.
coals, shales and sands are seen. The delta front is
Boudou, J. P., Pelet, R. and Letolle, R. 1984. A model of the
expected to have been influenced by relatively strong diagenetic evolution of coaly sedimentary organic matter.
tides, leading to enrichment by winnowing of drifted Geochim. Cosmochilll. Acta, 48, 1357-1362.
liptinitic debris. This enriched source rock zone may Combaz, A. and De Matharel, M. 1978. Organic sedimentation and
correspond to the facies change described by Gowers genesis of petroleum in Mahakam Delta Borneo. AAPG Rull.,
62, 1684-1695.
and Lunde (1984) from seismic reflections, which Cooper, B. S., anel Barnard, P. C. 1984. Source rock and oils of the
occurs to the north and west of Tramabanken and to Central and Northern North Sea. In: Uemaison, G. and Murris,
the west and south-west of Haltenbanken. Early R. J. (eels.), Petroleum Geochelllistry and Basia Evaluation.
Jurassic deltaic systems have been demonstrated in AAPG ;V1ellloir, 35, 303-315.
Durand, B. and Omlin. J. L. 1~79. Exemple de migration des hy-
the Troms(l area but here with a more northerly cli-
drocarbul'es dam; une ,erie deltaique: Ie delta de la Mahakam.
mate and weaker tidal effect similar oil-prone enrich- Kalimantan, Indonesie. 11)th World PetroleulII C(J1/ql' .. Pmc.,
ments may not occur. The sequence of Svalbard while Vol. 2, pp. 3-11.
deltaic to shallow marine is apparently only gas-prone Duran(l, B. an(l Paratta, :\II. 198:i. Oil potential of coals. In: Brooks,
(M(lrk and Bjor(lY, 1984). ,J. (ed.) Petl'oleuill Geucilell(isli'!f alld E,rplol'(ltioll lit' EUl'Ope.
Blackwell Scientific, Oxforrl, pp. 285-292.
A further inference of the model we have proposed Goff, J. C. 1984. Hydrocarbon generation and migration frum
is that gas-prone, rooted coals may be present on the Jurassic source rocks in the East Shetland Basin and Viking Gra-
delta top while oil-prone, drifted coals may be present ben of the Northern North Sea. In: Demaison, G. and Murris, It
at the delta margins. If the deltaic system is buried J. (eels.), Petroleulil Geochemistry and Rasia Eva.!uatioll.
sufficiently deeply to generate hydrocarbons, then AAPG Memoil', 35, 273-303.
Gowers, M. B. and Lunde, G. 1984. The geological history of
structures having drainage areas within the delta top Trrenabanken. In: Spencer, A. M. et al. (eds.), PetroleulII Geol-
may contain gas and condensate while structures oqy ofthe N ul'th European :~[(Lrqill. :'IIorwegian Petroleum Soci-
draining the delta margins could contain oil. If this ety, pp. 237-251.
model is correct then in an area such as Haltenbanken Grantham, P. J., Posthuma, J. ami Baak, A. 1983. Triterpanes of
Far-eastern crude oils. In: Bjor0Y, M. et (LI. (eds.), Advances in
it may be suggested that the better oil prospects are Orqallic Geochelilistry 1981. Wiley, Chichester, pp. fi75-683.
deeper in the basin. Hamar, G. P. and Hjelle, K. 1984. Tectonic framework of the M0re
Basin and the Northern North Sea. In: Spencer, A. M. et al.
(eds.), PetruleulII Geolllqy uf the North EUl'Openn :V1arqill.
CONCLUSIONS Norwegian Petroleum Society, pp. 349-359.
Hollander, N. 1984. Geohistory and hydrocarhon evaluation of the
Haltenbank area. In: Spencer, A. M. et al. (eels.). Petroleolll
It is apparent that coals deposited in particular Geolugy of the N ol'fh European :'v1arqill. Norwegian Petroleum
sedimentary environments may be good quality oil Society, pp. 3~3-388.
source rocks. A process of liptini~e enrichment is Jacobsen, V. and van Veen P. 1984. The Triassic offshore Norway
North of 62 eN. In: Spencer, A. M. et al. (eds.), Petrolelll;1
necessary in order for coal to become hydrogen-rich Geology of the North European :Hargill. Norwegian Petroleum
and oxygen-poor. Reworking of coastal plain peats to Society, pp. 317-329.
form drift deposits in tidal flat or lagoonal environ- Larsen, R. M. and Skarpnes, O. S. 1984. Regional interpretation
ments along the coastal margin is one mechanism by ami hydrocarbon potential of the Trrenabanken area. In: Spencer
which this can occur. Such processes can be observed et al. (eds.), Petroleulli Geology of the North European :VIar-
gin. Norwegian Petroleum Society, pp. 217-236.
in present Indonesian deltas, and are considered to be :\IIorley, R. J. 1981. Development and vegetation dynamics of a
responsible for the formation of coaly oil source rocks lowland ombrogenous peat swamp in Kalimantan Tengah,
of the Indonesian basins. It is probable that the sub- Indonesia. f. Bio,qeo.q., 8, 383-404.
tropical coastal swamp environments of the M0rk, A. and Bjor0Y, M. 1984. Mesozoic source rocks on Svalbard.
In: Spencer, A. M. etal. (eels.),PetroICUlIi Geolugy oftheNorth
Late Triassic/Early Jurassic offshore Central and European .11arglli. Norwegian Petroleum Society, pp. 371-383.
Northern Norway also contain intertidal or inner shelf Oudin, J. L. and Picard, P. F. 1982. Genesis of hydrocarbons in the
areas rich in reworked humic kerogen, which may :Vlahakam delta and the relationship between their distribution
similarly have good oil source quality. Facies mapping and the over-pressure zones. llIriOllesiall Petrol. A.Bs., 11th A.
COIIV. 1982, Proc., pp. 181-202.
could aid oil exploration in these areas. Roe, G. D. and Polito, L. J. 1979. Source rocks for oils in the
Ardjuna sub-basin of the Northwest Java basin, Indonesia. Proc.
Selilillar Oil Genemtioll and l'vlaturation of HydrocariJolis ill
ACKNOWLEDGEMENTS Sedimentar!! Basills, Manila. 1977. UNjESCAP, ceop Techn.
Publ. 6, Bangkok, pp. 180-194.
R0nnevik, H., Eggen, S. and Vallset, J, 19R;j. Exploration of the
We would like to thank the Directors of Robertson Norwegian shelf. In: Brooks. J. (eel.), Petroleulll Geochemistry
Research International Limited, and Pertamina, the and E:rplomtioll of Europe. Blackwell Scientific, Oxford.
Indonesian national oil company, for permission to pp. 71-93.
publish this paper. Schoell, M., Teschner, M., Wehner, H., Durancl, B. and Oudin, J.
L. 1983. Maturity related biomarker and stable isotope variations
and their application to oilf source rock correlation in the Mahakam
delta, Kalimantan. In: Bjor0y, !Vl. e/ al. (eds.), Advances hi
REFERENCES Organic Geochemist,.y 1981. Wiley, Chichester, pp. 156-163.
Seifert, W. K. and Moldowan, J. :vI. 1981. Paleoreconstruction by
Allen, G. P., Laurier, D. and Thouvenin, J. 1979. Etude sedimen- biological markers. Geochilli. CO.'"loehilll. Act(t, 45, 783-794.
tologique du Delta de la Mahakam. Notes et Mem01:res No. 15. Snowdon, L. R. and Powell, T. G. 1982. Immature oil and conden-
Compagnie Fran~aise des Petroles, Paris, pp. 1-156. sate-modification of hydrocarbon generation for terreetrial
Anderson, J. A. R. 1964. The structure and development of the organic matter. AAl'G Ball., 66, 775-788.
peat-swamps of Sarawak and Brunei. f. Tmp. Geogl., 18, Stach, E. 1975. Te,etbuok <deoal Petroloqy. Borntraeger, Berlin,
7-16. 42R pp.
Oil-generating Coals 73

Surlyk, F., Piasecki, S., Rolle, F., Stemmerik, L., Thomsen, E. Thomas, B. M. 1982. Land-plant source rocks for oil and their sig-
and Wrang, P., 1984. The Permian Basin of East Greenland. In: nificance in Australian basins. AJ'EA Journal, 22(1), 164-17il.
Spencer, A. M. et al. (eos.), Petroleum Geology of the North Thompson, R. R. and Benedict, L. G. 1974. Vitrinite reflectance as
European .'vfargin. Norwegian Petroleum Society, pp. 303-317. an indicator of coal metamorphism for cokemaking. Geol. Soc.
Sutton, C. 1979. Depositional environments and their relation to Alii. Spec. Paper No. 15i!, pp. 95-108.
chemical composition of Java Sea crude oils. Proc. Seminar' on Veroier, A. C., Oki, T. and Suardy, A. 1980. Geology of the Handil
Generation and Maturation of Hydrocarbons in Sedimentary field (East Kalimantan, Indonesia). In: Halbouty, M. T. (ed.),
Rosins, Manila, 1977. UNjESCAP, CCOP Techn. Publ. 6, Giant Oil and Gas Fields of the Decade 1968-1978. AA.PG
Bangkok, pp. 163-179. Mell/oir, 30, 399-421.
Taylor, G. H. 1966. The electron microscopy of vitrinites. CoalS ci., Westre, S. 1984. The Askeladdan gas find- Troms 1. In: Spencer,
Adv. Chem., Ser. 55, Washington, DC, pp. 274-283. A. M. et 01. (eds.), Petroleum Geology of the North European
Teichmiiller, M. and Durand, B. 1983. Fluorescence microscopical Mamin. Norwegian Petroleum Society, pp. 33-41.
rank studies on liptinites and vitrinites in peat and coals, and
comparison with results of Rock-Eval pyrolysis. Int. J. Cool
Geol., 2, 197-230.
5 Geochemistry of greater
Ekofisk crude oils
W. B. Hughes, A. G. Holba, D. E.Miller and J. S. Richardson
Phillips Petroleum Company, Bartlesville, Oklahoma 74004, USA

Detailed analyses of bulk properties and of molecular composition by gas chromatography and gas
chromatography-mass spectrometry of 30 oils from the greater Ekofisk area including Hod, Valhall, Eldfisk,
Edda, Ekofisk, West Ekofisk, Albuskjell, Northwest Tor, and Tor fields show the following.
The major differences among the oils are due to maturity, which decreases systematically by field along a
NW-SE trend as follows: Albuskjell > Northwest Tor, Tor, West Ekofisk > Ekofisk, Edda, Eldfisk > Valhall,
Hod. These maturity variations could be due to differences in the degree of thermal stress experienced by the
source rocks, to thermal alteration in the reservoirs, or to some combination of these two factors.
Minor, but definite, source differences exist which are not distributed systematically by field. The oils fall in the
following order of increasing terrestrial input and oxicity: Eldfisk (oil 27) < Eldfisk (oil 26) < Ekofisk, Tor,
Northwest Tor < Eldfisk, West Ekofisk, Hod, Edda (oil 17) < Albuskjell, Valhall, Ekofisk (oil 12), Edda (oil 18).
These variations are the result of locally different source rock subfacies within the Upper Jurassic. Within the
Ekofisk, Valhall, Edda, West Ekofisk and Eldfisk structures the Danian and/or Cretaceous reservoirs appear to be
differentially filled from these source subfacies.

INTRODUCTION SAMPLES AND METHODS OF ANALYSIS


The discovery of the Ekofisk Field in the Norwegian The 30 greater Ekofisk crude oils included in the study
sector of the North Sea has been cited as one of the are listed in Table 1 and located by well and field of
major finds of the decade 1968-1978, and the first of origin on the index map in Fig. 1.
the giant oil fields in Western Europe (Van den Bark After precipitation of asphaltenes with 20 x volumes
and Thomas, 1980). Since Ekofisk, many additional of n-pentane, and filtration, the oils were separated by
fields, including a number of giants, have been dis- medium-pressure liquid chromatography on the com-
covered in the North Sea, with present estimated re- bination of a cyanophase pre-column and an activated
coverable reserves of greater than 40 billion barrels of silica main column. The saturate and aromatic frac-
oil and 130 trillion ft 3 of gas (Dolan, 1983). tions were eluted with n-hexane from the main col-
Organic geochemistry has clarified the source of umn, and the polar fraction was backflushed with di-
Ekofisk oils (Van den Bark and Thomas, 1980), the chloromethane from the pre-column. The aromatic frac-
source in broad terms of central and northern North tion was analyzed directly by gas chromatography
Sea oils (Oudin, 1976; Barnard and Cooper, 1981, using a flame photometric detector (GC-FPD) and by
1984), and the source/maturity relationships among gas chromatography-mass spectrometry (GC-MS).
oils from the major grabens (Grantham et al., 1980; The total saturate fraction was analyzed by gas
Mackenzie et al., 1983). The economic and historical chromatography using a flame ionization detector
significance of the Ekofisk area merits a comprehen- (GC-FID) and, following removal of n-alkanes by
sive geochemical study of greater Ekofisk crude oils. molecular sieve adduction, by GC-MS. The n-alkanes
The objective of this study was to infer the source liberated by dissolution of the molecular sieves with
(organic input and depositional environment) and hydrofluoric acid were analyzed by GC-FID.
thermal maturity relationships of the oils. This was GC-FID analysis of the total saturate fraction was
accomplished by obtaining detailed geochemical ana- made on a 30 m SE-54 fused silica capillary column
lyses, including bulk properties, saturated and aroma- heated from 120°C to 320°C at 4°C/min. The n-alkane
tic molecular fossils and carbon isotopes, of 30 greater fraction GC-FID analysis was made on the same col-
Ekofisk crude oils. umn heated from 150°C to 320°C at 8°C/min.
Petmleum Geochemist"y in Exploration oflhe Norwegian Shei( © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 75-92.
76 Petroleum Geochemist/~ll ill Exploration of the No/'wegian Shelf

116 2/4 2/5 shales generally act as seals for the greater Ekofisk
reservOIrs.
Rapid subsidence, coupled with a eustatic rise in sea
1Of) level, resulted in the deposition of organic-rich mud-
A19 N.W. TOR
stones during the late Jurassic. Reduced circulation
within the Central Graben, caused by differential sub-
sidence resulted in anoxic conditions at the sedi-
ment-water interface with resulting preservation and
<J) burial of the organic matter (Van den Bark and
w. KOFISK
Thomas, 1980; MacKenzie et aI., 1983a). These
organic-rich Upper Jurassic shales are generally
\ 1/9 2/8
\ ~o5 believed to be the source of most crude oils in the
\
\ ~EDDA northern and central North Sea (Barnard and Cooper,
\
1981, 1984).
~
\ '2
8-14
\
8-11 8-9
,\1-
0
EKOFISK\ , -6
A-11 "
\0 3, A-2
C>1'~ A 21A RESULTS
"f-\-vJ- 217 ELDFISK
\
, Bulk properties
\

25 \ , The geochemical properties we define as bulk proper-


ties are listed in Table 2 for the greater Ekofisk oils.
o 5 10 \ 2/10 The crudes are medium to high API gravity and low in
"-+-~==? \ 2/11 HOD
MILES \
\ , c::::> sulfur nitrogen. As observed for many crude oils
\
(Tissot and Welte, 1978), there is an inverse correlation
between sulfur and API gravity (see Fig. 2). In addi-
Fig. 1. Location of crude oils analyzed by well and field of
tion, this correlation follows a general geographi.cal
origin. trend by field with West Ekofisk/Albuskjell havmg
the highest API gravities and lowest sulfur contents
and Valhall/Hod having the lowest API gravities and
highest sulfur contents (see Fig. 2)*. The nitrogen
GC-FfD analysis of the aromatic fraction was made content shows a linear correlation with the sulfur con-
on a 60 m DB-1 fused silica capillary column heated tent (R 2 = 0.73).
from 148°C to 225°C at 5°C/min. followed by heating
to 280°C at lOoC/min. .40
GC-MS analysis of the non-adduct portion of the + n
saturate fraction was made using a 60 m DB-5 fused 35
silica capillary column heated from 150°C to 310°C at
5°C/min. The column was coupled directly to the .30
source of a Hewlett-Packard mass-selective detector *
operated in the selected ion mode. The aromatic frac-
tion GC-MS analysis was made using a 90 m SE-54
*...
~
25

+
+
+
+ *
fused silica capillary column heated from 40°C to a: .20 .p Q>

3lOoC at 6°C/min. The column was connected via a "u.-' 0 +


0>
heated (300°C) line to a Kratos MS25 mass spectrome- "'" 15
t&,
ter operated in the scan mode (50-500 a.m.u.).
The carbon isotopic compositions of the total crude
. 10 '"
oils were determined using a Craig combustion system I>
.0 5 r¢
and a MM903 isotope ratio mass spectrometer. 0 0
Multivariate factor analysis was accomplished with 0.00
ARTHUR 81, a statistical software package obtained 25 30 35
GRAVITY (%APll
40 45 50 55

from Infometrix, Inc.


Fig. 2. Variation in gravity with wt% sulfur for greater
Ekofisk crude oils (see Table 1 for field symbol key).

GEOLOGICAL SETTING OF THE The oils are very low in asphaltenes, with the not-
GREATER EKOFISK AREA able exception of the Eldfisk Jurassic oil. The other
chemical class compositional data (saturate, aromatic,
The eight fields included in the study lie in a general and polar fractions) are plotted on the triangular dia-
NW-SE geographic trend within the Central Graben gram in Fig. 3. A trend with field is evident with the
in the southern part of the Norwegian sector of the Albuskjell oils highest in saturates and lowest in
North Sea (see Fig. 1). Most of the oils are produced polars, and the Hod oil next to the lowest in saturates
from either the Ekofisk Formation of Danian age or and highest in polm's. The Eldfisk Jurassic oil is signif-
the Tor Formation of Maastrichtian age (see Table 1 icantlv different from the other Eldfisk oils and similar
for reservoir formations and depths). Three oils are in its 'chemical class composition to the Hod oil.
from the Hod Formation of Campanian-Turonian age.
The remaining oil is from the Jurassic in the Eldfisk
field. The Ekofisk, Tor, and Hod reservoirs are frac-
tured chalks located on structural highs, which often 'The Northwest Tor oil is very similar to the Tor oils, and is plotted
overlie salt diapirs. Overpressured Paleocene age with them in this and all subsequent figures.
Geochemistry of Greater Ekofisk Crudes 77

TABLE 1. Greater Ekofisk crude oils

Sample Field
numbeT sYlllbol Field Well Depth (m)" Formation

1 0 Tor 2/4-8A 3194.3-3203.4 Tor


2 0 Tor 2/4-8A 3014.5-3041.9 Ekofisk
3 0 Tor 2/5-1 3154.1-3190.6 Tor
4 0 Tor 2/5-1 3224.8-3233.9 Tor
5 0 Tor 2/4-7 3304.0-3328.4 Tor
6 0 Northwest Tor 2/4-10 3304.0-3322.3 Tor
7 0 Albuskjell 1/6-1 3270.5-3279.6 Tor
8 0 Albuskjell 2/4-F16 3283.9-3302.S Tor
9 0 Albuskjell 1/6A-19 3266.8-3285.7 Tor
10 0 West Ekofisk 2/4-5 3124.2-3151.6 Ekofisk
11 0 West Ekofisk 2/4-5 3270.5-3310.1 Tor
12 0 Ekofisk 2/4-2 3169.0-3199.5 Ekofisk
13 0 Ekofisk 2/4-3 3218.7-3224.8 Ekofisk
14 0 Ekofisk 2/4-3 3127.2-3200.4 Ekofisk
15 0 Ekofisk 2/4-4A 3163.8-3203.4 Tor
16 0 Ekofisk 2/4-4A 3041.9-3075.4 Ekofisk
17 [:0 Eclcla 2/7-4 3105.9-3113.5 Ekofisk
18 [:0 Eclcla 2/7-5 3209.5-3218.7 Tor
19 t Elclfisk 2/7B-12 NA Ekofisk
20 + Elclfisk 2/7B-14 NA Tor
21 + Elclfisk 2/7B-6 2933.1-2991.0 Tor
22 Eldfisk 2/7B-11 NA Ekofisk + Tor
23 + Elclfisk 2/7B-9 NA Ekofisk + Tor
24 Eldfisk 2/7A-11 2812.4-2863.6 Ekofisk
25 + Eidfisk 2/7A-2 NA Tor
26 + Eldfisk 2/7A-21A 2816.0-2844.7 Hod
27 + Eldfisk 2/7-3 3502.2-3517.4 (Jurassic)
28 Valhall 2/8-A5 2790.1 Hod
29 Valhall 2/8-A2 3293.1-3304.0 Tor
30 XX Hod 2/11-6A NA Hod

NA = )I at available. aTrue vertical depth.

TABLE 2. Bulk properties of greater Ekofisk crude oils

Class cOIl/positioll (wt%)


Sail/pie Gravity Su(lia Nitmgen
number ("AP!) (wt%) (wt%) Sa.turate Aromatic Polar Asphaltene ol:Jc TOTa (%,)
39.9 0.12 0.05 70.4 19.4 9.2 0.9 -28.0
2 39.5 0.13 0.05 70.0 19.7 9.0 1.0 -27.8
3 39.4 0.12 0.04 68.7 18.8 11.6 0.8 - 27.9
4 39.2 0.12 O.OS 70.0 17.9 11.1 1.0 -27.S
5 31.0 0.11 0.05 71.5 18.5 9.7 0.3 -27.3
6 38.2 0.15 0.06 64.6 22.6 12.1 0.6 -28.1
7 44.4 0.05 0.04 78.6 16.2 4.9 0.3 -28.3
S 48.6 O.O~ 0.01 86.S 12.0 1.2 0.0 -28.2
9 50.3 0.02 0.01 89.0 9.8 1.2 0.0 -28.3
10 42.8 0.07 0.02 58.2 35.8 5.5 0.5 -28.2
11 44.6 0.05 0.04 73.9 18.4 7.2 0.5 -2S.2
12 33.5 0.21 0.13 57.7 23.R 16.9 l.6 -27.3
13 36.5 0.20 0.15 5G.3 24.5 18.5 0.7 -27.4
14 36.8 0.20 0.13 5u.7 22.9 20.6 0.8 -27.4
15 36.9 0.17 0.08 58.4 20.6 19.6 1.4 -27.4
16 38.0 0.15 0.07 54.9 23.8 17.2 0.3 -27.6
17 36.2 0.11 0.6 67.4 19.7 11.7 1.1 -27.4
IS 39.0 0.12 0.07 68.3 20.5 11.4 KA -2S.7
19 88.7 0.17 0.15 59.5 24.2 15.5 0.3 -28.0
20 31.4 0.22 0.22 58.9 23.3 17.5 0.2 -27.6
21 :06.7 0.20 0.14 57.4 22.~ 17.S 1.9 -27.2
22 37.0 0.18 0.13 58.1 24.4 17.2 0.3 -27.2
2:0 33.0 0.25 0.14 57.2 24.9 16.9 l.0 -27.4
24 35.3 0.23 0.18 56.7 23.0 20.2 0.1 -27.2
25 :03.2 0.20 0.12 59.7 22.7 17.2 0.4 -27.2
26 36.6 0.20 0.12 62.0 20.0 17.4 0.6 -27.2
27 28.3 O.ml 0.29 42.4 21.2 22.9 1:1.6 -26.8
28 36.6 0.25 0.17 57.5 22.5 19.6 0.4 -27.3
29 35.5 0.29 0.13 54.5 2il.4 21.8 0.:3 -27.3
30 31.5 0.38 0.18 49.1 25.1 24.7 1.1 -27.1

NA = Kot available. aTotal crude relative to PDB.


78 Petroleum Geochemistry in Exploration of the Norwegian Shelf

SATURATES
100
o TOR [DO A.
f:.
o ALBUSkJ[lL + [LOnSI(
o W. [I(OflSI( '1"LHALl
;I(
o EKOf'ISK D HOD

POLARS AROMA TICS


100 L-----*---+.-----i:~--+--~ 100

Fig. 3. Distribution by wt% of saturate, aromatic and polar fractions for greater Ekofisk crude oils (see Table 1 for field
symbol key).

The pristane/phytane ratios for the oils (see Table 3)


TOA determined by GC-FID analysis of the saturate frac-
AlBUSKJELL
tion fall in the very narrow range of 1.25-1.50 (exclud-
ing the Eldfisk Jurassic oil (pristane/phytane = 0.98)
W. EKOFISK
which appears anomalous compared with the
EKOFISK Danian-Cretaceous oils). The pristane/phytane ratios
EDOA
do not appear to follow any trend with field of origin,
implying that the redox potential of the depositional
ELDFISK
environments were similar for the source rocks of all
VALHAlL the oils. Values of <2 for the pristane/phytane ratio
HOD
fall within the range associated with anoxic conditions
-(Didyk et al., 1978).
-29.0 -28.0 -27.0 -26.0

8'3 C TOTAL CAUDE


Steranes
Fig. 4. Variation in <5 13C of total crude with field of origin
for greater Ekofisk crude oils. Molecular properties based on the sterane family of
chemical fossils were determined by GC-MS analysis
of the molecular sieve non-adduct portion of the satu-
The t5 13C values for the total crudes span the rela- rate fraction. The ions monitored for the steranes
tively narrow range of -26.8 to -28.7%0 (see Table were m/z 217, 218, and 259.
2). A plot of <5 13C values by field (Fig. 4) indicates that, Representative m/z 217 sterane mass fragmento-
in general, the oils become isotopically lighter in going grams for the greater Ekofisk oils are shown in Fig. 5.
from Hod to Albuskjell, although there is considerable The thermal maturities of the oils are reflected in the
overlap of the ranges for one field with another. relative proportions of 24-ethylcholestane isomers
(Mackenzie et al., 1983b). The C 29 biological sterol pre-
Molecular properties cursor has the 5a, 14a, 17a-20R stereochemistry
which is inherited by the C29 sterane (peak 8 in Fig. 5)
The molecular properties of the greater Ekofisk oils formed during diagenesis. This isomer is converted to
are various ratios of the relative concentrations of dis- the 5a, 14a, 17a-20S configuration (peak 6 in Fig. 5)
crete molecules as determined by GC or GC-MS during catagenesis. Both 20R and 20S 5a, 14a, 17a
analysis. For convenience, the molecular properties isomers are converted to the more stable 5a, 14/1, 17/1
have been divided into two general categories: source isomers (peaks 7 in Fig. 5) with increasing maturation
input/depositional environment indicators (Table 3) (Petrov et al., 1976; Mackenzie et al., 1980, 1982;
and maturity indicators (Table 4). No doubt some Seifert and Moldowan, 1981; van Graas et al., 1982).
overlap of these two categories is possible. In the dis- The 20S to 20R + 20S ratio of 5a, 14a, 17a, 24-
cussions which follow, the molecular properties will be ethylcholestane isomers ranges from 0.25 to 0.52 (see
grouped by chemical family. Table 4) for the oils and decreases fairly systemati-
cally from the Tor to the Hod field (Fig. 6(A». The
n-Alkanes/ isoprenoids most mature oils (Tor, Albuskjell and oilll from West
All of the oils have n-alkane distributions biased Ekofisk) fall in or near the proposed equilibrium range
toward low molecular weight (maximum at C I6 or C17) of 0.50-0.55 (Mackenzie et al., 1980, 1981a). The ratio
and an average odd-even predominance (Scalan and of 5a, 14/1, 17/1 to 5a, 14a, 17a plus 5a, 14/1, 17/1.
Smith, 1970) close to unity. There is no evidence for (20R + 20S) ethylcholestane isomers range from 0.33
biodegradation of the oils. to 0.63 (see Table 4) and also show a reasonably sys-
Geochemistry of Greater Ekojlsk Crudes 79

TABLE 3. Molecular properties of greater Ekofisk crude oils: organic input and depositional
environment indicators

Sample Prislane DBT/ Diastemne/ Norhopane/ ;268/ CnPP Aquatic/


number phytane phen. stemne hoparle iSS C28 PP total

1 1.40 0.06 1.31 0.43 0.39 1.28 0.69


2 1.40 0.08 1.32 0.48 0.49 1.27 0.69
3 1.40 0.08 1.44 0.47 0.44 1.33 0.71
4 1.50 0.10 1.31 0.44 0.46 1.31 0.68
5 1.50 0.06 1.19 0.48 0.41 1.14 0.68
6 1.36 0.09 1.19 0.44 0.48 1.21 0.70
7 1.3il NA 1.32 0.49 NA 1.09 0.68
8 1.35 0.08 1.28 0.49 0.59 1.26 0.67
9 1.54 0.06 1.41 0.53 0.62 1.20 0.68
10 1.44 0.10 0.91 0..5.5 0.57 1.12 0.67
11 1.44 0.11 1.41 0.51 0.78 1.42 0.70
12 1.47 0.10 0.37 0.46 0.63 0.92 0.63
13 1.39 0.14 0.55 0.53 0.60 1.28 0.70
14 1.::\7 0.12 0.54 0.51 NA 1.20 0.68
15 1.45 0.11 0.66 0.51 0.52 1.26 0.70
16 1.38 NA 0.64 0.53 KA 1.27 0.69
17 1.48 0.18 0.73 0.50 0.67 1.39 0.70
18 1.52 0.19 0.52 0.43 0.58 1.13 0.66
19 1.42 0.16 0.38 0.48 0.47 1.01 0.61
20 1.41 0.20 0.41 0.48 0.60 1.22 0.69
21 1.30 0.20 0.45 0.56 0.52 1.29 0.65
22 1.39 0.17 0.46 0.48 0.53 1.28 0.69
23 1.32 0.20 0.40 0.50 0.56 1.23 0.68
24 1.33 0.21 0.37 0.48 0.61 1.01 0.61
25 1.33 0.20 0.42 0.48 0.51 1.12 0.64
26 1.25 0.19 0.57 0.61 0.48 1.21 0.70
27 0.98 0.25 0.44 0.1l3 0.47 1.37 0.71
28 1.44 0.26 0.17 0.44 0.67 0.86 0.56
29 1.:39 0.27 0.22 0.52 0.69 1.02 0.63
30 1.34 0.25 0.40 0.62 0.67 1.11 0.67

Pristane/phytane-GC-FID peak areas from total saturates fraction.


BDT/phen-GC-:vIS areas of dibenzothiophenes (m/z 184)/phenanthrene (//I/z 178).
Diasterane/sterane-GC-MS (I11/Z 217) areas of peaks (1 + 2)/(3 + 4 + 5) as in Fig. 5.
Norhopane/hopane-GC-MS (I11/Z 191) areas of peaks 27/29 as in Fig. 11.
26S/28S-GC-MS (111/ 231) areas of peaks 35/37 as in Fig. 14.
C27PP/C28PP-GC-MS (m/z 218) areas of peaks 9/10 as in Fig. 8(A).
Aquatic/total-GC- (m /z 218) areas of peaks (9 + 10)/(9 + 10 + 11) as in Fig. 8(A).
See also Table 8 for peak identification.
NA = Not available.

tematic increase with field of origin (see Fig. 6(B)) early diagenesis may be obtained from the m/z 217
from Tor/Albuskjell (most mature) to Valhall/Hod mass fragmentogram. The conversion of stano Is to
(least mature). Based on intensities from the m/z 217 diasterenes, which are ultimately reduced to dia-
mass fragmentograms, the equilibrium value for the steranes, in sediments is believed to be catalyzed by
14P, 17p/14a, 17a + 14P, 17P ratio appears to be acidic sites on certain types of clays (Rubenstein et al.,
around 0.63. This is slightly lower than the range of 1975; Sieskind et al., 1979). Thus, the ratio of dia-
0.65-0.75 given by Mackenzie et al. (1980, 1981a). steranes to regular steranes in an oil reflects the activ-
However, this is not surprising since Mackenzie et al. ity of unbuffered acidic clays during diagenesis of the
used the m/z 218 trace to obtain the 14{3,17{3 peak source rock organic matter. This ratio was determined
areas, which would give larger numbers for these for the oils using the C28 diasteranes (peaks 1 and 2 in
isomers, hence higher 14{3, 17{3/14a, 17a + 14P, 17P Fig. 5) and C28 steranes (peaks 3-5 in Fig. 5). The
ratios. ratios are given in Table 3 and plotted by field in
Several significant conclusions about the maturities Fig. 7. The diasteranes make up from 15% (oil 28 from
of the greater Ekofisk oils may be drawn from the Valhall) to 59% (oil 3 from Tor) of the total steranes.
sterane isomer distributions. First, there is a major The diasterane/sterane ratio increases very systema-
N-S and minor W-E trend from more to less mature. tically by field from Valhall to Tor; the Hod oil appears
Second, the Tor and Albuskjell oils and oil 11 from to have an anomalously high ratio relative to the Val-
West Ekofisk are of advanced maturity, having hall oils. Again, a significant difference between the
reached essentially the thermodynamic equilibrium two West Ekofisk oils is observed. The Ekofisk oils
distribution of C29 isomers. Third, the West Ekofisk have values similar to the Eldfisk oils, but signific-
oils are significantly more mature than the Ekofisk antly different from West Ekofisk.
oils. Fourth, the range of maturities spanned by the The relative abundances of C27 , C28 , and C29 steranes
Eldfisk and Ekofisk oils are very similar. Fifth, rela- in oils reflects the carbon number distribution of the
tive to oil generation, the oils range from early to late sterols present at the time of deposition and thus
stage oil window maturity. reflects the type of organic matter input to the source
Information about the chemical environment during rocks of the oils (Huang and Meinschein, 1979). The
00
o

AL8USKJELL OIL EKOFISK OIL


2/4-F16 2/4-3

1
7
II 1 4
II II 5 h 8 '0
2 I I
3 6 '"
q-
I 2 I
3 4 >- I £.
>- 11\ f0- ~
f0-
Ul ;:,
Ul Z
Z W
W fo-
CCl
fo- Z '"o
Z C">

~
;::;.
~

TIME - - - - - - - -
~
TIME
;;:
t>:l
-[
~~.
HOD OIL .~
2/11-6A s:.
'">'
S~
8 :::1
ELDFISK OIL I ~
5
2178-11 I 5"
C/)
;::;-
'"
-c::;;
7
7 6 I
II I
>- 5 8 >-
f0- f0-
I I
Ul Ul
Z Z
W W
fo- fo-
Z Z

TIME TIME - - - - - - - - - - - -

Fig. 5. Representative //lIz 217 mass fragmentograms for greater Ekofisk crude oils (see Table 8 for peak identification key).
Geochemistry of Greater Ekofisk Crudes 81
(A) (E)
TOR +--+ TOR

ALBUSKJELL AlBUSKJEll

W. EKOFISK W. EKOFISK

EKOFISK I I EKOFISK

EDOA EODA

ELDFISK ElDFISK I II I I I

VALHALL ++ VALHALL

HOD HOD

.20 .30 .40 .50 .60 30 35 40 45 .50 55 .60 ,65 70 75 80

20S
20R + 208
14 a 1 7a +1413 1 713

Fig. 6. (A) Variation in 20S/20R + 20S 5a, 14'a, 17a-24-ethylcholestane ratio with field of origin for greater Ekofisk crude oils.
(E) Variation in fJfJ/ aa + fJfJ (20R + 20S) 24-ethylcholestane ratio with field of origin for greater Ekofisk crude oil.

sterane carbon number distributions were determined diagram in Fig. 9. Very similar distributions are
from the relative intensities of C Z7 (peaks 9), C28 found for all the oils, and there is no trend with field of
(peaks 10), and Cz9 (peaks 11) 5a, 14P, 17P (20R + 208) origin.
steranes in the m/z 218 mass fragmentogram (a typical The ratio of C27 and C28 to the total of C27 , C28 , and
m/z 218 trace is shown in Fig. 8(A)). The sterane car- C29 PP steranes generally reflects the ratio of organic
bon number distributions are plotted on the triangular matter derived from aquatic organisms to the sum of
aquatic- and terrestrially-derived organic matter. The
TABLE 4. Molecular properties of greater Ekofisk aquatic/total ratios for the greater Ekofisk oils (see
crude oils: maturity indicators Table 3) range from 0.60 to 0.72, suggesting a predo-
minantly aquatic input with a significant terrestrial
Sample 20S fJfJ/ C!!O+!!l/ contribution.
number (20R + 20S) (fJfJ + IXIX) Ts/Tm Ci6 +27 + i8 MPI-l However, a recent example indicates that oils from
1 0.50 0.63 2.49 1.21 0.76
pre-Ordovician carbonate sources, which predate the
2 0.52 0.60 2.60 1.53 0.78 appearance of extensive land vegetation, have high
3 0.50 0.63 2.99 1.47 0.77 C29 sterane contents (McKirdy et al., 1983). In the
4 0.52 0.63 3.07 1.96 0.74 present case, significant terrestrial input is supported
5 0.52 0.63 2.33 1.11 0.75 by our observation of conifer diterpenoids in three of
6 0.50 0.63 2.69 1.17 0.64 the oils (see below) and by reports of land plant-
7 0.48 0.61 1.88 NA NA derived kerogen in the presumed Upper Jurassic (Kim-
8 0.48 0.63 2.61 2.54 0.72 meridge Clay Formation and equivalents) source
9 0.49 0.61 3.06 2.54 0.72 rocks (Van den Bark and Thomas, 1980; Barnard and
10 0.45 0.59 1.95 1.07 0.74 Cooper, 1984).
11 0.52 0.63 2.98 1.53 0.74 The sterane carbon number distribution was also
12 0.33 0.46 1.09 0.55 0.62 determined from the m/z 259 mass fragmentogram
13 0.35 0.48 1.54 0.42 0.60 (Fig. 8(B)) using the relative intensities of the C27
14 0.33 0.45 1.50 0.48 0.62
15 0.42 0.51 1.69 0.49 0.63 (peaks 12 and 13), C28 (peaks 14 and 15), and C29 (peaks
16 0.39 0.50 1.43 NA NA 16 and 17) diasteranes. The diasterane carbon number
17 0.40 0.56 2.25 0.69 0.49 distributions are also plotted on the triangular dia-
18 0.41 0.59 2.08 0.59 0.49 gram in Fig. 9 for comparison with the PP steranes. The
19 0.38 0.49 1.12 0.36 0.57 distributions are quite SImilar, and the six oils which
20 0.34 0.43 1.29 0.34 0.57 fall outside the main cluster (toward the C29 vertex)
21 0.38 0.46 1.01 0.28 0.56
22 0.35 0.47 1.53 0.42 0.59
23 0.34 0.46 1.19 0.31 0.58
24 0.38 0.49 1.12 0.37 TOR
0.54
25 0.36 0.49 1.19 0.27 0.59
ALBUSKJElL
26 0.43 0.54 1.07 0.39 0.57
27 0.35 0.52 1.02 0.16 0.58 W. EKOFISK

28 0.27 0.36 0.96 0.21 0.54


EKOFISK
29 0.26 0.38 0.98 0.22 0.55 II

30 0.25 0.33 0.93 0.15 0.56 EODA

ElDFISK
20S/20R + 20S-GC-MS (m/z 217) areas of peaks 8/(6 + 8) as in
Fig. 5. VAlHAll
fJfJ/fJfJ + IXIl-GC-MS (m/z 217) areas of peaks 7/(6 + 7 + 8) as in
Fig. 5. HOD
Ts/Tm-GC-MS (m/z 191) areas of peaKs 24/26 as in Fig. 11.
Czo + Zl/CZ6 + 27 + 28-GC-MS (m/z 231) areas of peaks (33 + 34)/ 0.0 1.0 1 •
(35 + 36 + 37 + 38 + 39) as in Fig. 14.
MPI-I-GC-MS (m/z 178 and 192) peak areas where MPI-1 = C 28 0lASTERANES

1.5(3 - +2-methylphenanthrenes)/(phenanthrene + 9- C 28 STERANES


+ 1-methylphenanthrenes).
Also see Table 8 for peak identifications. Fig. 7. Variation in C 28 diasterane/C 28 sterane ratio with
NA = Not available. field of origin for greater Ekofisk crude oils.
82 Petroleurn Geochernistry in Exploration of the Norwegian Shelf

EKOFISK OIL
2/4-3

11
10
I II

>-
t-
(/)
Z
W
t-
Z

TIME ------~

(B) 12

13 16

14
15 EKOFISK OIL
17 2/4-3

>-
t-
(/)
Z
W
t-
Z

TIM E - - - - - - - - - <...

Fig. 8. (A) rn/z 218 mass fragmentogram for 2/4-3 Ekofisk oil (see Table 8 for peak identification key). (B) rn/z 259 mass
fragmentogram for 2/4-3 Ekofisk oil (see Table 8 for peak identification key).

C27
100
+ 148,17B STEAANES (mlz 218)
o DIASrERANES (m/z 259)

C29 C28
100 '--------;';-------;ci-------;';-------;-;--------" 100

Fig. 9. Carbon number (C 27 , C28 , C29 ) distributions of 15 a,14p,17P, (20R + 208) steranes (+) and of 13P,17 a-20 (R + 8)
diasteranes (0).
Geochemistry of Greater Ekofisk Crudes 83

.80 trend observed in the 20R/20R + 20S and f3f3/1l1l + f3f3


sterane maturity indicators.
75 Several other terpane molecular ratios show similar
geographic trends as Ts/Tm with field of origin (see
.70 Table 5). These are the ratio of the C23-C Z6 tricyclic
"P,17P
STERANES
diterpenoids (peaks 18-21 in Fig. 11) to hopane (peak
.65 29), the ratio of the unknown C30 pentacyclic triter-
0* penoid (peak B) to hopane, and the ratio of the
+ +
.60 unknown CZ9 pentacyclic triterpenoid (peak Z) to
norhopane (peak 27). In all cases, these ratios gener-
55 * ally decrease from north to south, as does Ts/Tm. In
contrast to Ts/Tm, however, Albuskjell shows some-
,5 o-IL--~-----<~-~----+---=-----:<
what higher values than Tor and West Ekofisk in the
50 55 .60 65 .70 .75 .80
tricyclicjhopane and compound B/hopane ratios. Also
DIASTERANES
in contrast to Ts/Tm, the values for the other three
Fig. 10. Relationship between aquatic/total ratio ratios are similar for the Edda and Ekofisk oils. In
(C Z7 + CZslC 2S + Czs+ CZ9) calculated from 5a, 14{3, 17{3 general, however, the ratios of compounds Band Z
(20R + 208) steranes and 13{3, 17a (20R + 208) diasteranes. and of the C23-C Z6 tricyclics to hopane behave very
similarly to Ts/Tm; these three ratios versus Ts/Tm
have linear correlation coefficients (R~ of 0.59-0.77.
are the same in both cases. These samples are oil 12 Contrary to the other molecular terpane ratios (see
from Ekofisk, oils 19, 24, and 25 from Eldfisk, and oils Tables 4 and 5), the ratio of norhopane (peak 27) to
28 and 29 from Valhall (see Table 1). hopane (peak 29) remains relatively constant from
The aquatic/total sterane ratios calculated from the field to field (see Table 3 and Fig. I2(B». A similar
f3f3 sterane and diasterane distributions are compared constancy by field was observed in the carbon number
graphically in Fig. 10; the solid line in the figure rep- distribution of steranes of the same configurations.
resents a 1 : 1 equivalence. As this plot shows, the The bisnorhopane (peak Y) to hopane (peak 29) ratio
two ratios give essentially the same result. This indi- (see Table 5) shows an intermediate behavior. No sys-
cates no preferential conversion by carbon number of tematic geographic trend by field of origin is observed;
sterols to diasteranes during diagenesis. most of the oils have values in the range 0.07-0.23,
and the ratio for the Hod oil is in the same range as
Terpanes those of the Tor oils. The Albuskjell and West Ekofisk
The representative rn/z 191 mass fragmentograms oils show significantly higher values (0.29-0.44). Only
shown in Fig. 11 contain significant amounts of the a poor linear correlation is found with Ts/Tm
(R z = 0.19). However, the most mature oils based on
hopane series of pentacyclic triterpenoids (peaks 24,
26, 27, 29, 31 + 32) commonly found in crude oils and the Ts/Tm ratio tend to have the highest bisnorhopane
rock extracts. Present in variable amounts are a series /hopane ratios.
of Czr Cz6 and Czg-C 30 tricyclic terpenoids (peaks
18-23 and 25). Of low abundance in all the oils are Aromatics
moretane (peak 30) and normoretane (peak 28). In The aromatic fractions of the greater Ekofisk crude
addition, several other compounds detected in the rn/z oils were analyzed by GC-FPD and by GC-MS in the
191 mass fragmentogram are of interest. These are full scan mode. The fingerprint ions used in this study
bisnorhopane (peak Y), previously reported (Gran- were m/z 178, 184, 192, 198 and 231.
tham et aZ., 1979) in North Sea oils; a 24 carbon tetra- Representative gas chromatograms of the aromatic
cyclic terpenoid (peak X) whose structure has fraction using a flame photometric (sulfur-selective)
recently been established (Rinaldi, 1984); two 30 car- detector are shown in Fig. 13. The major organosulfur
bon pentacyclic triterpenoids (peaks A and B) of components are benzothiophenes (peaks 40); diben-
unknown structure (molecular ions m/z 412, base zothiophene (peak 41); and a series of methyl- (peaks
peaks rn/z 191); and a 29 carbon pentacyclic 42-44), dimethyl- (peaks 45), and trimethyl- (peaks
triterpenoid (peak Z) of unknown structure (molecular 46) dibenzothiophenes. In going from Hod to Eldfisk
ion m/z 398, base peak m./z 191). to Ekofisk there is a disappearance of the ben-
Overall, the m/z 191 mass fragmentograms for the zothiophenes, a decrease in dibenzothiophene relative
oils are fairly similar. Features in common 'include a to methyl derivatives, and a redistribution of isomers
predominance of hopanes over tricyclics, hopane > among the methyl-, dimethyl-, and trimethyldiben-
norhopane, and low concentrations of moretanes. zothiophenes. The last change was quantitated from
However, definite trends exist in the rn/z 191 finger- the rn/z 198 mass fragmentogram, and is reflected in
print which are related to the geographical location of the ratio of the I-methyl isomer to the sum of the 1-,
the field of origin of the oil. Although the range of 2-, 3-, and 4-methyl isomers (MDBTR in Table 5). It
values within some of the fields is fairly large, the was observed previously (Hughes, 1984) that oils with
ratio of 18 11- to 17 Il-trisnorhopane (Ts/Tm) increases the distribution of methyldibenzothiophene isomers
in going north from Hod to Albuskjell and Tor (see found in the Hod oil, i.e. peak 42 > peak 44 > peak 43,
Table 4 and Fig. 12(A». The Ts/Tm ratios for West are from carbonate or calcareous shale sources or
Ekofisk and Edda are higher than for Ekofisk. There sources deposited in highly reducing environments. It
is also a considerable difference in this ratio for the was also found that redistribution of the methyldiben-
two West Ekofisk oils. The Ts/Tm ratio has been zothiophenes to the pattern in the Albuskjell oil, i.e.
proposed as a maturity indicator (Seifert and Mol- peak 42 > > peak 43 > peak 44, results from advanced
dowan, 1978) with higher values indicating greater maturation of such oils. Changes in the distribution of
maturity. The major N-S and minor W-E geographic dibenzothiophenes have also been observed by Radke
trend toward higher maturity in Ts/Tm is the same et aZ. (1982a).
84 Petroleum Geochemistry in Exploration of the Norwegia.n Shelf

24

29

ALBUSKJELL OIL 2/4-F 16


23 A

>
.,...z
w
... "';"Jou.•.JvIWWIWW~V"''''''''
!

29

27

EKOFISK OIL 2/4-3


24
18

27 29
I I

24
I
ELDFISK OIL 217-11
26
I Z 31

..
> I I,
~ 18
I
Z
...w
!
TIME

27 29
I I

HOD OIL 2/11-SA

31
z II

TlME--------

Fig. 11. Representative m!z 191 mass fragmentograms of non-adduct saturate fractions from greater Ekofisk crude oils (see
Table 8 for peak identification key),
Geochemistry of Greater Ekofisk Crudes 85

(A) (E)
TOR TOR

ALBUSKJElL ALBUSKJELL

WEST EKOFISK WEST EKOFISK

EKOFISK EKOFISK

EDDA EDOA

ELDFISK ELDFISK

VALHALL VALHALL

HOD HOD

1.0 1.52.0 2.53_03.5 .3 .7 .9


TSITM NORHOPANE/HOPANE

Fig. 12. (Al Variation in 18a,21fJ (Ts) to 17a,21fJ (Tm) 22,29,30-trisnorhopane ratio with field of origin for greater Ekofisk
crude oils. (E) Variation in 17a,21fJ-30-norhopane to 17a,21fJ-hopane ratio with field of origin for greater Ekofisk crude oils.

The relative distribution of phenanthrene and its 1-, The ratio of C20 + C21 to C2fi + C27 + C28 triaromatic
2-,3-, and 9-methyl derivatives has been proposed as a steranes from the m/z 231 mass fragmentograms
maturity indicator, termed the methylphenanthrene (peaks 33 + 34/35-39 in Fig. 14) decreases systemati-
index (MPI-1), by Radke et al. (1982a). The MPI-1 cally from Albuskjell to Hod (see Table 4). This molecu-
values for the greater Ekofisk oils (see Table 4) are lar ratio is similar to the C201C 28 + Czo aromatic
highest in the northern (Tor, Albuskjell) and West sterane ratio used by Mackenzie et al. (1981b, c) as a
Ekofisk oils; intermediate in Ekofisk; and lowest in maturity index operative in the early to late oil win-
Eldfisk, Valhall, and Hod. The Edda and Northwest dow range; both indicators increase with maturity.
Tor oils have anomalouslv low values. The MPI-1 indi- Based on the C2o-2dC26-28 ratio, Ekofisk oils are more
cator does not differenti~te Valhall and Hod oils from mature than Eldfisk oils and West Ekofisk and Edda
Eldfisk as do the sterane and triterpane maturity indi- oils are more mature than Ekofisk oils. This maturity
cators in Table 4. trend matches that observed for the 20R/20R + 20S

TABLE 5. Selected molecular properties of greater Ekofisk crude oils

Sal/lple RiwoThopane/ 19 Pentacyclic/ TI"derpan€ B / Tricyclics/


Intmher .11DRTR hopane Norhopane iwpanc hopane

1 0.10 0.23 1.12 0.39 1.48


2 0.09 0.17 1.02 0.37 0.86
3 0.13 0.16 0.96 0.37 1.41
4 0.10 0.13 0.98 0.:35 1.25
5 0.11 0.23 1.04 0.41 1.42
6 0.12 0.16 0.94 0.31 1.27
7 NA 0.44 1.02 0.38 I\A
8 0.10 0.43 0.91 0.52 1.58
9 0.11 I\A 0.90 0.59 2.28
10 0.13 0.29 0.54 0.39 0.89
11 0.09 O.:H 0.95 0.41 1.85
12 0.15 0.11 0.37 0.19 NA
13 0.17 0.09 0.47 0.11 0.66
14 0.16 0.08 0.44 0.11 0.38
15 0.16 0.08 0.49 0.12 0.61
16 NA 0.09 0.46 0.12 0.85
17 0.18 0.12 0.47 0.13 0.85
18 0.18 0.10 0.45 0.13 0.25
19 0.18 NA 0.51 0.19 NA
20 0.19 0.08 0.:16 0.08 0.17
21 0.21 0.19 0.52 0.20 NA
22 0.19 0.07 0.38 0.08 0.17
23 0.20 0.07 0.34 0.07 0.14
24 0.19 NA 0.51 0.19 NA
25 0.20 NA 0.56 0.20 NA
26 0.20 NA 0.67 0.30 NA
27 0.26 0.12 0.31 0.04 0.27
28 0.23 0.07 0.33 0.07 0.08
29 0.25 0.10 0.32 0.06 0.08
30 0.25 0.17 0.2il 0.05 0.19

MDBTR-GC-MS (I11/Z 19R) areas of peaks 1-methyldibenzothiophene/(4 + 2 + 3 + 1-methyldibenzothio-


phenes).
Bisnorhopane/hopane-GC-:vrS (m/z 191) areas of peaks Y/29 as in Fig. 11.
29 Pentacyclic/norhopane-GC-MS (III /z 191) areas of peaks Z/27 as in Fig. 11.
Triterpane B/hopane-GC-MS (mjz 191) areas of peaks B/29 as in Fig. 11.
Tricyclics/hopane-GC-;vrs (III /z 191) areas of peaks (18 + 19 + 20 + 21)/29 as in Fig. 11.
Also see Table 8 for peak identifications.
NA = Not available.
86 Petroleum Geochemistry in Exploration of the Norwegian Shelf

42
42

ALBUSKJELL OIL EKOFISK OIL


116-1 214-2

41
41 3 ..1L , ~5
3

1
40 ~
144
I 4
1111, I. J .11,11. 1. JI

TIME------ TIME

42

42

ELDFISK OIL HOD OIL


217-11 2111-6A
41
3
\44
/

45
41
.ll
3

,.!L
~
,1.1.
I
TIME TIME

Fig. 13. Representative GC-FPD chromatograms of aromatic fractions from greater Ekofisk crude oils (see Table 8 for peak
identification key).

and f3f3/aa + f3f3 sterane ratios and the Ts/Tm tris- tion of these compounds in the other oils, which are
norhopane ratio. more mature than the Valhall/Hod oils, or to some
The relative distribution of C2o C2R triaromatic combination of source and maturity.
steranes (peaks 35-39 in Fig. 14) is similar for all the The ratio of dibenzothiophene to phenanthrene is
oils, suggesting a similar source organic input (Mack- thought to be an indicator of source rock lithology
enzie et ai., 1983a). Some variation in internal ratios is (Holba, unpubl. data). The low values for the greater
observed, however. For example, the ratio of the C26 , Ekofisk oils (see Table 3) are in the range for shaly
20S (peak 35) to the C28 , 20S (peak 37) triaromatic source rocks. The dibenzothiophene/phenanthrene
steranes (see Table 3) varies from 0.39 (Tor, sample 1) ratio increases systematically from Albuskjell and Tor
to 0.78 (West Ekofisk, sample 11). The variation is not to Valhall and Hod, being similar in this regard to the
systematic by field, or with maturity indicators such maturity indicators. This ratio correlated highly
as f3f3/aa + f3f3 steranes and Ts/Tm, or with other (R 2 = 0.89) with the methyldibenzothiophene isomer
source indicators, such as the ratio of Cn/C 29 f3f3 ratio (see Table 5).
steranes (peaks 9 and 11, respectively, in Fig. 8(A)).
It appears that different factors may control the car-
bon number distribution of triaromatic and saturated
DISCUSSION
steranes.
The Valhall and Hod oils have significant levels of
Source versus maturity effects
the conifer-derived (Thomas, 1969) diterpenoid retene
and, in low concentrations, its precursors tetrahyd- A classical problem in organic geochemistry is to dis-
roretene and simonellite. The absence of these diter- entangle the effects of source (which we take to
penoids in the other greater Ekof'isk oils could be due include both organic input and depositional environ-
to a source (organic input) difference, to the destruc- ment) and maturity on bulk and molecular properties
Geochemistry of Greater Ekofisk Crudes 87

ALBUSKJELL OIL
33
I 2/4-F16 EKOFISK OIL
34 2/4-2
I

34
I
36
I

,,11Lo
37
>
to
Ii
z
III
to
Z

TIME---- TlME---

jl3 36
I ,36

,
34
ELDFISK OIL
33
HOD OIL
217B-9 I 2/11-6A
35 3n8
I

35
37
138 , 34
39
I
I I

>
to
>-
I-
Ii
z
III
'"
Z
W
to I-
! Z

TlME----

Fig. 14. Representative mjz 231 mass fragmentograms of aromatic fractions from greater Ekofisk crude oils (see Table 8
for peak identification key).

of crude oils and rock extracts. So far our geochemical most to the first principal component are either matur-
analysis of greater Ekofisk crude oils has yielded the ity indicators (C 2o_2JC 26-28 triaromatic steranes,
general observation that bulk properties (see Table 2) Ts/Tm hopanes, 20S/20S + 20R steranes, and
and molecular maturity indicators (see Table 4) show a PP/rxrx + PP steranes) or ones found to correlate highly
systematic variation by field (NW-SE) and that with maturity indicators (remaining properties in
molecular source indicators (see Table 3) generally Table 6). The fact that two properties are important in
show very little or no systematic change (exceptions the same eigenvector means that they are correlated.
are the dibenzothiophene/ phenanthrene and dia- The properties contributing the most to the second
sterane/ sterane ratios). This observation suggests that principal component are source-related (organic input
maturity, and not source, is the major factor for the or depositional environment) indicators.
differences in geochemical properties of the oils. Based on the principal component analysis of all the
To assist in interpreting the large amount of geochemical data, we conclude that the differences in
geochemical data in this study and in determining bulk and molecular properties among the oils are due
what are the most important factors in controlling the predominantly to maturity (51% of total variance),
variation among the samples we have used principal and, to a significantly lesser extent, to source (13% of
component (eigenvector) analysis (Harper et al., total variance).
1984). This statistical technique reduces the dimen- A cross-plot (see Fig. 15) of the first and second prin-
sionality of the data by extracting the minimum cipal components shows the distribution of the oils
number of factors (eigenvectors) which best span the with respect to the major factors of maturity and
original data. The original data in this case consisted source. The samples are distributed along the first
of all the bulk and molecular properties listed in Tables principal component in the order (with increasing
2-5 plus nine additional molecular ratios. maturity): Hod, Eldfisk oil 27, Valhall < Eldfisk,
Two eigenvectors were found to account for 64.4% Ekofisk, Edda < Tor, Northwest Tor, West Ekofisk,
of the total variance. Most (31.1 %) of the remaining Albuskjell oil 7 < Albuskjell. The distribution of oils
variance was accounted for by nine additional eigen- by field along the maturity principal component cor-
vectors with individual contributions of 1.3-7.1%. responds very well to the NW -SE trend previously
Table 6 llsts the percentage of the total variance observed in maturity indicators (see Table 4 and
accounted for by each of the first two principal compo- Figs. 6 and 12(A».
nents (eigenvectors) and the individual contribution of The samples are distributed along the second prin-
the most important molecular or bulk properties to cipal component in the order (with increasing terres-
each eigenvector. The properties contributing the trial input and oxicity): Eldfisk oil 27 < Eldfisk oil 26 <
88 Petroleum Geochemistry in Exploration of the Norwegian Shelf

TABLE 6. Factor loadings in first and second


principal components from multivariate analysis of
geochemical data EKOF1$K
r 12
z

~
w
Feature Variance (%) z 19
,~
EDDA
24
0
First Principal Component (51.4% of total variance) U
10
WEST
EKOFISK

Polars (%) 5.04 ~ 2a~1

0
17 11
23 22
Diasteranej sterane 5.03
C20 + 2t!C26 + 27 + 2S 4.90
4.87 ~ 12
Triterpane Bjhopane 0
Z
0
Gl 3 TOR

Saturates (%) 4.85 U


W
N.W.
'OR

MDBTR 4.82 W

20/28 R 4.82 26
ELDflSK
Sulfur (%) 4.81
Ts/Tm 4.80 27 ELOFI$K
20S/(20R + 20S) 4.75
PMPP + aa) 4.55 FIRST PRINCIPAL COMPONENT

Nitrogen (%) 4.29


DBT /phenanthrene 4.29
Fig. 15. Plot of first versus second principal components
Tricyclics/hopane 4.28 resulting from factor analysis of greater Ekofisk geochem-
API gravity 4.27 ical data.
29 pentacyclicjnorhopane 4.01

Second Principal Component (13.0% of total variance) by maturity-as observed in the greater Ekofisk oils.
Aquatic/total (mjz 218) 13.41 This means that these molecular ratios can be used for
Aquatic/total (m/z 259) 12.44 oil/oil and oil/source rock correlations only for oils or
C27PP/C2SPP steranes (m/z 218) 11.37 rock extracts of comparable maturity or in cases
C27/C 28 Diasteranes (m/z 259) 10.47
Pristane/phytane 6.94 where compensation for the effect of maturity can be
26S/28S (m/z 231) 6.35 made.
Asphaltenes (%) 5.16 A similar duality of source and maturity control
probably applies to the Ts/Tm indicator. It is interest-
20/28R-GC-MS (mjz 231) areas of peaks 33/39 as in Fig. 14. ing that the aromatic maturity indicator MPI-1, does
Aquatic/total (m/z 259)-GC-MS areas of peaks
(12 + 13 + 14 + 15)/(12 + 13 + 14 + 15 + 16 + 17) as in Fig. 8(B). not have a high loading to the first (maturity) principal
C27jC 28 Diasteranes (mjz 259)-GC-MS areas of peaks (12 + 13)j component.
(14 + 15) as in Fig. 8(B). The principal component analysis indicates that var-
All other variables defined in footnotes to Tables 3, 4, and 5. iations in the bulk properties, API gravity, percen-
Also see Table 8 for peak identifications.
tage sulfur, percentage nitrogen, percentage polars,
and percentage saturates are primarily the result of
maturity differences among the oils.
Ekofisk, Tor, Northwest Tor < Eldfisk, West Eko- The b13C of the total crude does not appear among
fisk, Hod, Edda oil 17 < Albuskjell, Valhall, Ekofisk oil the important features in the maturity principal com-
12, Edda oil 18. ponent (see Table 6). Inspection of the correlation
Both our 'classical' analysis and the principal com- matrix for all the data shows that bl~ of the total
ponent analysis of the data indicate that maturity is crude has the highest correlation with the maturity-
the most important factor controlling the geochemical related features API gravity (R 2 = 0.62), percentage
features of the greater Ekofisk oils. Thus, this group sulfur (R 2 = 0.64) and 14P,17p/14a,17a + 14P,17P
of oils represents a case in which source effects are sterane ratio (R 2 = 0.44) and a low correlation with
reasonably constant, permitting the effect of maturity the source-related features m/z 218 aquatic/total ratio
alone to be assessed. (R 2 = 0.02) and pristane/phytane ratio (R 2 = 0.28).
The diasterane/sterane ratio (see Table 3), which is
taken to be a depositional environment indicator of
lithology, appears to be, for the greater Ekofisk oils, 1.6
primarily maturity controlled. This conclusion is sub-
stantiated by the high loading of this ratio in the first 1.' o S
(maturity) principal component (see Table 6) and by 00 8
o
the smooth increase in the ratio with the pp/aa + PP
1.2

C29 sterane ratio (see Fig. 16). The deviation of the 1.0

Hod oil from the general trend may be due to source


(high clastic/terrestrial input?). For similar reasons,
C 2 eOIASTERANES

C 28 STERANES
..
the variations in the dibenzothiophene/phenanthrene .6
ratio (see Table 3), the methyldibenzothiophene
isomer ratio and the triterpane-derived ratios in Table .4 =
5 are largely due to maturity differences among the
oils. .2

The diasterane/sterane, dibenzothiophene/phenan- 0.0 +--+---+--+--+--+--+--+----<


threne, and methyldibenzothiophene isomer molecular .3 .4 .5
17{3
.6 .7

ratios discussed above are examples of geochemical


14~,

14 a, 17a + 14{3,17fJ
parameters which respond to source, lithology and
maturity. Probably the most important control is Fig. 16. Variation in C2S diasterane/ sterane ratio with
source lithology, which determines the initial values PP/ ClCl + PP (20R + 208) 24-ethylscholestane ratio for
for the ratios. Subsequently, the ratios are modified greater Ekofisk crude oils (see Table 1 for field symbol key.
Geochemistry of Greater Ekofisk Crudes 89

Disregarding the outlying points, namely oil 18 from The distribution along the source component axis in
Edda and oil 27 from Eldfisk, Fig. 4 indicates a Fig. 15 also shows that oil 12 falls significantly outside
NW-SE, hence maturity, trend, with the oils becom- the cluster ofthe other Ekofisk oils, and that the two
ing isotopically lighter with increasing maturity by Valhall oils, the two Edda oils and, to a lesser extent,
approximately 1%0. This trenais opposite to conven- the two West Ekofisk oils, are separated. In each
tional ideas that oils become isotopically heavier with case, the pairs of oils from Valhall, Edda, and West
maturation (Silverman, 1967; Sofer, 1984). Ekofisk are from different reservoir formations.
A possible explanation is thermal alteration (crack- Within Eldfisk, oils 19 and 24 are separated from the
ing) of the oil, either before expulsion in the source main group; it is interesting that these two oils are the
rock or after accumulation in the reservoir. Such a only unmixed Ekofisk Formation productions. A
cracking process would tend toward the formation of a plausible interpretation is that the Ekofisk, Valhall,
low molecular weight, condensate-like fraction and a Edda, West Ekofisk, and Eldfisk structures have
high molecular weight, coke-like fraction, with the received oil from different source subfacies of the Late
former isotopically lighter and the latter isotopically Jurassic and that the Danian and/or Cretaceous reser-
heavier on average than the original whole oil. The voirs may be differently filled from these source sub-
lower molecular weight fraction would be preferen- facies.
tially recovered during production, hence the mea- Inspection of the molecular ratios used in this study
sured whole-oil 013C value would decrease as this pro- shows that they fall into two general categories (see
cess progresses, i.e. as the oils become more mature. Table 7): those based on homologous molecular pairs
A number of the molecular indicators involving and those based on isomeric molecular pairs. Based on
members of homologous series are maturity indepen- the results of this study, the following generalizations
dent and appear as significant features in the second can be made. Over the range of maturities exemplified
(source) principal component. In this group are those by the greater Ekofisk oils, which we estimate to be
ratios based on the carbon number distribution in the approximately equivalent to 0.55-1.00 ± 0.05 vitri-
saturated steranes and diasteranes and the triaroma- nite reflectance, molecular ratios based on homologous
tic steranes. The carbon number distribution of PP pairs are unaffected by maturity, and may be used
20(8 + R) steranes has been cited as reliable (un- unequivocally as source indicators and as oil/oil and
affected by maturity) source indicator (Mackenzie et al. , oil/rock correlation parameters (examples of these
1982). Of interest is the occurrence of pristane/ ratios are listed in column A of Table 7). Over the
phytane in this group of indicators. Our data show that range of maturities of the greater Ekofisk oils,
over the range of maturities found in the greater molecular ratios based on isomeric pairs change sys-
Ekofisk oils, pristane/phytane is unaffected. It tematically with maturity (examples of these ratios
is also interesting that the principal component are listed in column B of Table 7). Those isomeric
analysis shows the percentage asphaltenes in the oils ratios where initial value is not source dependent,
to be source controlled (see Table 6). such as pp/aa + PP steranes, may be used universally
The distribution of the oils along the second princi- as maturity indicators. Those isomeric ratios whose
pal component gives important information about the initial value varies with source may be used as com-
source of the oils. The Jurassic oil from Eldfisk (sam- parative maturity indices only when source is con-
ple 27) falls at one extreme of the source component stant.
(see Fig. 15). Its geochemical characteristics include
the highest aquatic/total ratio, lowest pristane/ Bisnorhopane
phytane ratio and high sulfur contentflow gravity for
its maturity. We conclude that oil 27 is a locally Since the first report of 17a(H), 18a(H), 21P(H)-
sourced oil from a subfacies of the Upper Jurassic 28,30-bisnorhopane (Seifert and Moldowan, 1978) in
deposited in a marine and very anoxic environment. the Monterey shale, the factors controlling its rather
Oil 26 from Eldfisk falls significantly outside the other enigmatic occurrence in oils and the biological source
Eldfisk oils and near the Eldfisk Jurassic oil. This oil is of this triterpane have been of interest. Grantham et
from the deepest of the Cretaceous reservoirs (Hod al. (1980) have previously reported the presence of
Formation) and may be a mixture of the Jurassic type bisnorhopane in North Sea oils. They also found the
oil and the type represented by the main group of ratio of bisnorhopane to hopane tended to increase
Eldfisk oils. Excluding these two oils, the second with increasing sulfur content. In this study, the
(source) principal component indicates that the Tor, opposite is true; bisnorhopanejhopane generally
Northwest Tor, and Ekofisk oils are from the most increases with decreasing sulfur content (see Fig. 17).
marine and anoxic source relative to the other oils. The two samples with the greatest deviation from the

TABLE 7. Ratios based on homologous and isomeric molecular pairs


A B
Homologous Molecular Pairs Isomeric Molecular Pairs

Pristane 5a,14a,17a,20(R + Sl-24-ethylcholestane


Phytane 5a,14p,17p,20(R + S)-24-ethylcholestane
5a,14p,17P,20R(or 20Sl-cholestane 5a,14a,17 a,20R( or 20Sl-24-methylcholestane
5a, 14P, 17P,20R(or 20Sl-24-methylcholestane 13P ,17a,20it(or 2OS)-24-methyldiacholestane
17a(H),21P(H)-30-norhopane 17a(Hl ,21P(H)-22,29,30-trisnorhopane
17a(Hl,21P(Hl-hopane 18a(Hl-21P(Hl-22,29,30-trisnorhopane
C23 tricyclic terpane 1-methyldibenzothiophene
C24 tricyclic terpane 4-methyldibenzothiophene
..
90 Petroleum Geochemistry in Exploration of the Norwegian Shelf

.50
(A) 1.00 ~.~- .......
.40 ELDFI8K-OIL 21

+ -t HOD .80
co.
....
~

c:Q

.30
SULFUR (WTt.)
co.
.... '<t . 60
~

~- +
.20 co.
'<t l:!
~ .... .40
~

tl'
.10
'<t
+
0.00 +--<_-+---+-~~-+-+---+---+---+-+--<_-+--I-+-+-<
0.00 .10 .20 .30 .40 .50
0.00
BI$NORHOPANE
0.00 .20 .40 .60 .80 1.00
HOPANE
Ts/Tm
Fig. 17. Variation in wt% sulfur with biBnorhopane/hopane
ratio for greater Ekofisk crude oils.
(B) 1.00

general trend are the Eldfisk Jurassic oil (sample 27)


.80
and the Hod oil. We have already seen that the Eldfisk ~
oil is sourced by a different subfacies than the other ~

oils. Why the Hod oil deviates from the trend is not co.c:Q
.... '<t .60
clear. We have also already established that sulfur
~
~

co.- t1+
content is controlled by maturity. '<t ....
We believe that our data and that of Grantham et al. lj'"
.40
(1980) can be explained by the following argument. '<t
The major control on bisnorhopane/hopane ratio is .20
source. However, among samples from a similar •
source, and perhaps when bisnorhopane/hopane is ini-
tially rather low, maturity is the controlling factor and 0.00
0.00 .20 .40 .60 .80 1.00
an inverse relationship exists between this ratio and
sulfur content. This relationship is not direct but in- MPI-l
direct, i.e. they are both responding to a common fac-
tor-thermal stress. Moreover, the relationship need
not be linear, since the decrease in sulfur and increase
(e) 1.00 .. -=.. ~--~~----~

in bisnorhopane/hopane ratio can respond to thermal •
stress at different rates. Our data show the process co.
....
.80
responsible for increasing the bisnorhopane/hopane
ratio has a higher activation energy than the process cri
~ '<t
.60
of desulfurization of the oil.
~

~- +
If this hypothesis is correct, then the decrease in
bisnorhopane/hopane ratio with decreasing sulfur con- ::
co.~
.... .40
tent observed by Grantham et al. (1980) is source- tl
rather than maturity-controlled. Also, in this '<t
20
scenario, the bisnorhopane/hopane ratio in the Hod oil
and possibly the Eldfisk Jurassic oil are source-
controlled. Since the Hod oil is very similar to the
.20 .40 .60 .80 1.00
other oils in source indicators (see Table 3), it is not
clear why the bisnorhopane/hopane ratio should be in
such high (relative) concentration in this sample. Evi-
dently, whatever is controlling the bisnorhopane con-
Fig. 18. (A) Variation in Ts/Tm ratio with pp/aa + PP
tent is not reflected in other commonly measured (20R + 208) 24-ethylcholestane ratio for greater Ekofisk
geochemical properties. crude oils. (B) Variation in MPI-l with pp/aa + PP
(20R + 208) 24-ethylcholestane ratio for greater Ekofisk
Comparison of maturity indicators crude oils. (e) Variation in e20-2de26-28 triaromatic sterane
ratio with pp/aa + PP (20R + 208) 24-ethylcholestane ratio
Intercomparison of the various thermal indicators for greater Ekofisk crude oils.
used in this study can give information about their
sensitivity to thermal stress. To make these compari-
sons we have plotted Ts/Tm, MPI-1, and C20-2dC26-28
triaromatic sterane ratios versus PP/rxrx + PP sterane very anomalous MPI-l values. To remove the effects
ratio. Since we are interested in assessing maturity of differing ranges in which each indicator is opera-
effects only, we have eliminated those samples which tive, the indicators were scaled by subtracting the
appear to have a strong source component. These lowest measured value from each observation and
samples are the Eldfisk Jurassic oil (sample 27) and dividing the result by the observed range. The result-
the Eldfisk oil which appears to contain a significant ing plots are shown in Fig. 18(A)-(C). The solid
amount of the Jurassic-type oil (sample 26). The two curves were calculated by a polynomial regression
Edda oils were also eliminated as they appear to have analysis of the data.
Geochemistry of Greater Ekofisk Crudes 91

1.00 TABLE 8. Peak identification key for figures


/ Peak Compound(s)
/
." .80
.... I
~
/ III/Z 217 GC-MS sterane trace (Fig. 5)
." I
~ ~ .60
1 13p,17a,20S-24-methyl-diacholestane
I 2 13p,17a,20R-24-methyl-diacholestane
~ +
oi\j 3 5a, 14a, 17a,20S-24-methyl-cholestane
;! .... - Ts/Tm 4 14p,17P(20S + 20R)-24-methyl-cholestanes
~ .40
C20-2l 5 5a, 14a, 17a,20R-24-methyl-cholestane
---- C 26 - 28 6 5a, 14a, 17a,20S-24-ethyl-cholestane
........ MPI-l
7 14P,17P,(20S + 20R)-24-ethyl-cholestanes
.20 8 5a, 14a, 17a,20R-24-ethyl-cholestane

m/z 218 GC-MS sterane trace (Fig. 8(A»


o. 0 O~-'---:'-:--L..--:-'-::--,---:-'-::-.......--,-:-::-.......--:-,
0.00 .20 .40 .60 .80 1.00 9 14p,17P(20S + 20R)-cholestanes
MATURITY INDICATOR 10 14P, 17P(20S + 20R)-24-methyl-cholestanes
11 14p,17P(20S + 20R)-24-ethyl-cholestanes
Fig. 19. Comparison of regression curves for Ts/Tm ratio,
MPI-l, and C20-21/C26-28 triaromatic sterane ratio with In /z 259 GC-MS diasterane trace (Fig. S(B»
PP/ exex + PP (20R + 208) 24-ethylcholestane sterane ratio for 12 13p,17a,20S-diacholestane
greater Ekofisk crude oils. 13 13p,17a,20R-diacholestane
14 13P, 17a,20S-24-methyl-diacholestanes
15 13p,17a,20R-24-methyl-diachloestanes
16 13p,17a,20S-24-ethyl-diacholestane
The general shape of the three plots is similar, and 17 13p, 17a,20R-24-ethyl-diacholestane
indicates that the aromatic sterane and triterpane
indicators are less sensitive to thermal stress (this In /z 191 GC-MS terpane trace (Fig. 11)
assumes that heating rates were not drastically differ- 18 23 carbon tricyclic terpenoid
ent for the sources of the oils). The best-fit curves for 19 24 carbon tricyclic terpenoid
20 25 carbon tricyclic terpenoids
all three indicators are plotted versus the fJfJ/aa + fJfJ 21 26 carbon tricyclic terpenoids
sterane indicator on the same axes in Fig. 19. Relative 22 28 carbon tricyclic terpenoids
to sterane isomerization and for heating rates compar- 23 29 carbon tricyclic terpenoids
able with the greater Ekofisk area, sensitivity to 24 lSa(H),21P(H)-22,29,30-trisnorhopane (Ts)
25 30 carbon tricyclic terpenoids
thermal stress increases in the order: C20- 2dC 26--28 26 17a(H),21P(H)-22,29-30-trisnohopane (Tm)
triaromatic sterane ratio < Ts/Tm = Mpr-1. The 27 17a(H),21P(H)-30-norhopane
rates of the processes responsible for the changes in 2S 17P(H),21a(H)-30-nomoretane
these thermal indicators decrease in the same order. 29 17a(H),21P(H)-hopane
30 17P(H),21a(H)-moretane
31 17a(H),21P(H),22(S + R)30-homohopanes
32 17a(H),21P(H),22(S + R)-30,31-bishomohopanes
CONCLUSIONS A 30 carbon pentacyclic triterpenoid (unidentified)
B 30 carbon pentacyclic triterpenoid (unidentified)
The results of bulk property, carbon isotope, and X 24 carbon tetracyclic terpenoid
Y 17a(H),lSa(H),21P(H)-28,30-bisnorhopane
saturated and aromatic molecular fossil analyses of 30 Z 29 carbon pentacyclic triterpenoid (unidentified)
oils from the greater Ekofisk area of the Norwegian
sector of the North Sea suggest the following. //I /z 231 GC-MS triaromatic sterane trace (Fig. 14)
The major differences among the oils are due to 33 20 carbon triaromatic sterane
maturity, which decreases systematically by field 34 21 carbon triaromatic sterane
along a NW-SE trend as follows: Albuskjell > 35 26 carbon (20S) triaromatic sterane
36 26 carbon (20R) + 27 carbon (20S) triaromatic steranes
Northwest Tor, Tor, West Ekofisk > Ekofisk, Edda, 37 28 carbon (20S) triaromatic sterane
Eldfisk > Valhall, Hod. Whether the maturity varia- 38 27 carbon (20R) triaromatic sterane
tions are due to differences in the degree of thermal 39 28 carbon (20R) triaromatic sterane
stress experienced by the source rocks, to thermal
Thiophenic sulfur GC-FPD trace (Fig. 13)
alteration in the reservoirs, or to some combination of
these two factors is not known. 40 C2 + C3-benzothiophenes
41 dibenzothiophene
Minor, but definite, source differences exist which 42 4-methyldibenzothiophene
are not distributed systematically by field. The oils fall 43 2- + 3-methyldibenzothiophene
in the following order of increasing terrestrial input 44 1-methyldibenzothiophene
and oxicity: Eldfisk (oil 27) < Eldfisk (oil 26) < Ekof- 45 dimethy ldibenzothiophenes
isk, Tor, West Tor < Eldfisk, West Ekofisk, Hod, 46 trimethyldibenzothiophenes
Edda (oil 17) < Albuskjell, Valhall, Ekofisk (oil 12),
Edda (oil 18). These variations are the result oflocally These include the pristane/phytane and norhopane/
different source rock subfacies within the Upper hopane ratios and the ratios of steranes of different
Jurassic. Within the Ekofisk, Valhall, Edda, West carbon numbers with the same stereo-
Ekofisk, and Eldfisk structures the Danian and/or chemistry. Molecular ratios based on isomeric pairs
Cretaceous reservoirs appear to be differentially filled change systematically with maturity, and may be
from these source subfacies. used as comparative maturity indicators only when
Over the maturity range of the greater Ekofisk oils source is constant or the initial value is not source-
(early to late oil window; approximately 0.55-1.00 Ro) dependent. These include, in the former category, the
molecular ratios based on homologous pairs are rela- Ts/Tm hopane ratio and, in the latter category, the C29
tively unaffected by maturity and may be used un- sterane isomer ratios. Also included is the source-
equivocally as source indicators and in correlations. related indicator, the diasteranefsterane ratio.
92 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Comparison of maturity indicators shows that France-IV. Laboratory thermal alteration studies. Geochim.
within the suite of greater Ekof'isk oils they have the Cosmochim. Acta, 45, 2369-2376.
Mackenzie, A. S., Brassell, S. C., Eglington, G. and Maxwell, J. R.
following sensitivity to thermal stress: sterane isomer- 1982. Chemical fossils: the geological fate of steroids. Science,
ization > methylphenanthrene index = Ts/Tm hopane 217,491-504.
ratio > C26-21/C26-28 triaromatic sterane ratio. Mackenzie, A. S., Maxwell, J. R., Coleman, M. L. and Deegan, C.
E. 1983a. Biological marker and isotope studies of North Sea
crude oils and sediments. Proc. 11th World Petroleum Congr.
vol. 2, pp. 45-56.
ACKNOWLEDGEMENTS Mackenzie, A. S., Ren-Wei, L., Maxwell, J. R., Moldowan, J. M.
and Seifert, W. K. 1983b. Molecular measurement of thermal
maturation of Cretaceous shales from the overthrust belt, Wyom-
The authors wish to thank the following individuals for ing, USA. In: Bjorlly, M. (ed.), Advances in Organic Geochemis-
helpful discussions or assistance in obtaining informa- try 1981. pp. 496-503.
tion for this study: D. C. Boatwright, M. A. Sloan, J. McKirdy, D. M., Aldridge, A. K. and Ypma, P. J. M. 1983. A
S. Shveima, and M. E. Smith, Exploration and Pro- geochemical comparison of some crude oils from pre-Ordovician
carbonate rocks. In: Bjor0Y, M. (ed.), Advances in Organic
duction Group, Phillips Petroleum Company; W. G. Geochemistry 1981. Wiley, pp. 99-107.
Lyon, Research and Development, Phillips Petroleum Oudin, J. L. 1976. Geochemical study ofthe North Sea Basin. Bull.
Company; and J. Brewster, Phillips Petroleum Centre Rech. Pau-SNPA, 10,339-358.
Company Norway. The experimental assistance of Petrov, A. A., PustiPnikova, S. D., Abrjutina, N. N. and
J. A. Lytle, J. S. Spears, and T. Smith is also Kagramonova, G. E. 1976. Petroleum steranes and triterpanes.
Neftekhimiya, 16, 411-427.
acknowledged. Radke, M:, Welte, D. H. and Willsch, H. 1982a. Geochemical study
on a well in the western Canada basin: relation of the aromatic
distribution pattern to maturity of organic matter. Geochim.
Cosmochim. Acta, 46, 1-10.
REFERENCES Radke, M., Willsch, H., Leythhaeuser, D. and Teichmiiller, M.
1982b. Aromatic components of coal: relation of distribution pat-
Barnard, P. C. and Cooper, B. S. 1981. In Illing, L. V. and Hobson, tern to rank. Geochim. Cosmochim. Acta, 46, 1831-1848.
G. D. (eds.), Petroleum Geology of the Continental Shelf of North Rubenstein, I., Sieskind, O. and Albrecht, P. 1975. Rearranged
West Europe, Institute of Petroleum, London, pp. 169-175. sterenes in a shale: occurrence and simulated formation. J. Chem.
Barnard, P. and Cooper, B. S. 1984. Source rocks and oils of the Soc. Perkin Trans., I, 1833-1836.
central and northern North Sea. In Demaison, G. and Murris, R. Rinaldi, G. G. L. 1984. Novel series oftetracyclic terpanes in crude
J. (eds.), Petroleum Geochemistry and Basin Evaluation. AAPG oils and rock extracts. Am. Chem. Soc. Abstr. Papers, 187th
Memoir, 35, pp. 303-314. Natn. Mig; GEOC-57.
Didyk, B. M., Simoneit, B. R. T., Brassell, S. C. and Eglington, G. Scalan, R. S. and Smith, J. S. 1970. An improved measure of the
1978. Organic geochemical indicators of paleoenvironmental con- odd -even predominance in the normal alkanes of sediment
ditions of sedimentation. Nature, 272, 216-222. extracts and petroleum. Geochim. Cosmochim. Acta, 34,
Dolan, P. 1983. North Sea oil province: a failed rift basin. Oil Gas 611-620.
J., 185-196 (31 January). Seifert, W. K. and Moldowan, J. M. 1978. Applications of steranes,
Grantham, P. J., Posthuma, J. and de Groot, K. 1980. Variation and terpanes, and monoaromatics to the maturation, migration, and
significance of the C27 and C28 triterpane content of a North Sea source of crude oils. Geochim. Cosmochim. Acta, 42, 77-95.
core and various North Sea crude oils. In Douglas, A. G. and Seifert, W. K. and Moldowan, J. M. 1981. Paleoreconstruction by
Maxwell, J. R. (eds.), Advances in Organic Geochemistry 1979. biological markers. Geochirn. Cosmochim. Acta, 45, 783-795.
Pergamon Press, Oxford, pp. 29-48. Seifert, W. K., Carison, R. M. K. and Moldowan, J. M. 1981.
Harper, A. M., Meuzelaar, H. L. C. and Given, P. H. 1984. Correla- Geomimetric synthesis, structure assignment, and geochemical
tions between pyrolysis mass spectra of Rocky Mountain coals correlation application of monoaromatized petroleum steroids. In:
and conventional coal parameters. Fuel, 63, 793-802. BjorllY, M. (ed.), Advances in Organic Geochemistry 1981.
Huang, W-Y. and Meinschein, W. G. 1979. Sterols as ecological pp. 710-724.
indicators. Geochim. Cosmochim. Acta, 43,739-745. Sieskind, 0., Joly, G. and Albrecht, P. 1979. Simulation of the
Hughes, W. B. 1984. The use ofthiophenic organosulfur compounds geochemical transformations of sterols: superacid effect of clay
in characterizing crude oils derived from carbonate versus silicic- minerals. Geochim. Cosmochim. Acta, 43, 1675-1679.
lastic sources. In: Palacas, J. G. (ed.), Petroleum Geochemistry Silverman, S. R. 1967. Carbon isotopic evidence for the role oflipids
and Source-Rock Potential of Carbonate Rocks. AAPG Studies in petroleum. J. Am. Oil Chemists Soc., 44,691-695.
in Geology 18. Sofer, Z. 1984. Stable carbon isotope compositions of crude oils:
Mackenzie, A. S., Patience, R. L., Maxwell, J. R., Vandenbroucke, applications to source depositional environments and petroleum
M. and durand, B. 1980. Molecular parameters of maturation in alterations. AAPG Bull., 68,31-49.
the Toarcian shales, Paris Basin, France-I.-Changes in the con- Thomas, B. R. 1969. Kauri resins-modern and fossil. In: Eglington,
figurations of a cyclic isoprenoid alkanes, steranes and triter- G. and Murphy, M. T. J. (eds.), Organic Geochemistry.
panes. Geochim. Cosmochim. Acta. 44, 1709-1723. Springer-Verlag, New York, pp. 599-619.
Mackenzie, A. S. and Maxwell, J. R. 1981a. Assessment of thermal Tissot, B. and Welte, D. H. 1978. Petroleum Formation and
maturation in sedimentary rocks by molecular measurements. In: Occurrence. Springer-Verlag, New York, 538 pp.
Brooks, J. (ed.), Organic Maturation Studies and Fossil Fuel van Graas, G., Baas, J. M. A., van de Graaf, B. and de Leeuw, J.
Exploration, Academic Press, London, pp. 239-254. 1982. Theoretical organic geochemistry. I. The thermodynamic
Mackenzie, A. S., Hoffmann, C. F. and Maxwell, J. R. 1981b. stability of several cholestane isomers calculated by molecular
Molecular parameters of maturation in the Toarcian shales, Paris mechanics. Geochim. Cosrnochirn. Acta, 46, 2399-2402.
Basin, France-III. Changes in aromatic steroid hydrocarbons. Van den Bark, E. and Thomas, O. D. 1980. Ekofisk: first of the
Geochim. Cosmochim. Acta, 45, 1345-1355. giant oil fields in western Europe. in: Halbouty, M. T. (ed.),
Mackenzie, A. S., Lewis, C. A. and Maxwell, J. R. 1981c. Molecular Giant Oil and Gas Fields of the Decade 1968-1978. AAPG
parameters of maturation in the Toarcian shales, Paris Basin, Memoir, 30, 195-224.
6 Correlation of Northern North
Sea oils: the different facies of
their Jurassic source
Mark A. Northam, Mobil Research and Development Corporation, Dallas
Research Laboratory, Dallas, Texas 75881, USA

Fifty-four crude oils from the Norwegian Shelf and the UK sector of the North Sea have been compared
geochemically to determine their genetic relationships. All are considered to have been sourced from the Jurassic
Kimmeridge Clay. Although most compositional differences among the oils are the result of post-accumulation
processes, some of the parameters reflect organic facies changes in the source. These facies changes are a result of
the varying depositional environments present in the Viking and Central Grabens during the Jurassic. Evidence
for the facies changes can be found in their stable carbon isotope profiles. Correlation of North Sea crude oils allows
identification of Kimmeridge-sourced sub-families, which may be useful in delineating drainage areas and
migration pathways. Comparison of the shape of the isotope profiles within one of these oil sub-families is helpful in
explaining the maturity trend of oils in the Beryl Field, and may aid in determining timing of migration of
hydrocarbons.

INTRODUCTION analyzed at Mobil over the past years. Our conven-


tional geochemical data lead to the same conclusions
The North Sea has been of great interest to geochem- arrived at by other groups. The oils were also
ists because it forms a natural laboratory in which to examined using a technique called isotope profiling, a
study the processes of generation and migration in a parameter which is sensitive to depositional environ-
large family of oils. The source of these oils is gener- ment as well as thermal maturity. Results of our study
ally accepted to be the organic-rich shales of the Kim- of these oils show that there are, indeed, organic
meridge Clay Formation. Because the oils are reser- facies differences which can be recognized in the
voired at different levels in fields which are spread Jurassic source rocks.
over a relatively large area, there is a strong likeli-
hood that depositional environment, organic facies,
and thermal maturity changes are present in their NORTH SEA OILS-GEOCHEMICAL DATA
source rocks. The wide range of compositional varia-
tions which are present in the North Sea oil set have, The 54 oils in this study were obtained from reservoirs
however, been attributed mainly to thermal maturity at depths between 6000 and 13 700 ft (1830-4175 m).
differences in their otherwise homogeneous source All have been subjected to compositional analyses and
rocks. oil-oil correlation techniques, which are standard
Cornford et al. (1983) came to this conclusion for a geochemical procedures and need no special explana-
large suite of UK North Sea oils. They compared tion in a paper of this scope. Results show that the oils
maturity-dependent biomarker ratios determined for have ranges of API gravities, sulfur contents, and
the oils and extracts of source rocks at known matur- hydrocarbon compositions which fit within the
ity levels in order to explain the origin of the different ranges published by Cornford et al. (1983), who
oil types. As expected, post-generation processes such showed that these parameters vary with and are
as biodegradation and asphaltene precipitation were dependent upon the thermal maturity of North Sea
found to be responsible for the formation of some of oils.
the heavy oils. They also found that certain biomarker Biomarker analysis, the determination of the rela-
ratios were somewhat useful in correlating oils with tive distribution of chemical fossils present in oils and
regional facies of the source rock. rock extracts, is generally considered to be a sensitive
A suite of oils from 12 fields in the Viking and Cen- indicator of the type of organic material which was
tral Grabens of the North Sea (see Fig. 1) have been deposited and incorporated into the source rock
Petroleum Geochemistry in Exploration of the Nmwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 93-99.
94 Petmleum Geochemistry in Exploration of the Norwegian She(f

o 50 100 MI. ment of separate source facies. Although these values


80 180 KM. span the range published by Cornford et al. (1983),
the Pr/Ph values do not show any trend with matur-
ity. The Pr/Ph data suggest that the source rock
depositional environments vary from slightly reducing
to slightly oxidizing. There is no obvious relationship
between the stable isotope ratio and Pr/Ph ratios for
80't---+-l-+--+-t----t1i1"/
these oils, since the former is dependent entirely upon
, I
BERYL
the nature of the organic input and the latter is depen-
I dent primarily upon the oxidation state of the deposi-
tional environment. Since neither of these two para-
meters is sensitive to thermal exposure, it is difficult
to use standard geochemical analyses to describe in
detail the variations which exist in the source of North
Sea oils without an extensive study of the rock sam-
ples themselves.

~.'"
57't--+-+--+-r-t---+-r-r /
\
.. '
rEKOFISK" /

\./
.)-

56·+--+-+-C=t-:..::.=-....j-- CARBON ISOTOPE PROFILES


Fig. 1. Northern North Sea oil fields. A carbon isotope profile of a crude oil is constructed by
measuring the stable carbon isotope ratio (oi3C) of the
whole oil, 11 distillation cuts, and the over 300 DC
kerogen. The families of compounds most frequently residuum of the distillation. The distillation cuts are
used for biomarker analysis are the steranes and ter- taken at 25 DC intervals from 50°C to 300 DC. This
panes, because of their resistance to degradation. The range represents hydrocarbons with volatilities be-
distribution of biogenic steranes (20R-5IX, 14IX,17IX-C 27 , tween those of n-C 5 and n-C 17 • The 013C values of the
-C 28 , and -C 29 steranes) in the oils is dependent almost cuts and residuum are plotted versus the maximum
entirely upon the input to the source rock of sterols distillation temperature of the cut on a diagram as
from indigenous organisms. Oils generated in the illustrated in Fig. 4. The shape of the profile and its
same source inherit the same biomarker fingerprints. relationship to the whole oil 013C values are important
The mass chromatograms ofsteranes (m/e = 217) and for oil-oil correlation purposes.
terpanes (m/e = 191) are very similar for all of the Carbon isotope profiles have been utilized in inter-
North Sea oils, excepting maturity related differ- nal reports at Mobil Research and Development Cor-
ences. That the oils in this study fall into a tight group poration for more than 20 years. Over 800 oils from
on a plot of regular sterane distributions (Fig. (2a» worldwide sources have been profiled, mainly for
leads to the conclusion that the oils were generated in oil-oil correlation studies. Isotope profiles are only
related source rocks and that the contribution of warranted in special cases. They have been found to
biological material to the Kimmeridge Clay was be most useful when used to detect subtle differences
uniform over the entire area of this study. in large families of oils generated in an apparently
The relative thermal maturities of oils from a given homogeneous source rock, as is the case in the North
family can be determined by measuring the extent of Sea. Their interpretation can be complicated, how-
the isomerization of several thermally sensitive ever, because the shape of a given profile depends on
molecules (Seifert and Moldowan, 1981; MacKenzie et several factors and can be changed by any process
al., 1982). For this study, two transformations of the which alters the composition of an oil. Isotope profile
C29 sterane were graphically compared after the shapes vary greatly for different oils, just as
method of Seifert and Moldowan (1981). The first is biomarker fingerprints vary due to differences in the
the transformation from the biogenic IXIX 20R isomer make-up of the source kerogens. Biodegradation and
(see above) to the more thermally stable /l/l isomer water-washing, when severe enough, can render an
(5IX,14/l,17/l (20R)-C Z9 sterane). The second is the isotope profile· useless for correlation purposes. It
transformation from the biogenic IXIX 20R isomer to is helpful, therefore, to understand as much about a
the more thermally stable IXIX 20S isomer. The plot of family of oils as possible before interpreting isotope
IXIX 20S/ IXIX 20R versus /l/l 20R/ IXIX 20R (20S/20R profiles.
versus /l/l/ IXIX) for this data set (Fig. 2(b» shows the Thermal exposure is another of the natural proces-
broad range of maturities covered by the oils from the ses which can alter the composition of a crude oil. Oil
12 represented fields. composition changes with increasing maturity of the
Not all the variations in the composition of these oils source rock during generation, and with thermal
can be ascribed to maturation effects. For instance, exposure after expUlsion. Several chemical paramet-
the whole oil stable carbon isotope ratios of this sam- ers can be used to determine maturity levels of oils,
ple set range from -27.4 to -30.2 per mil versus but cannot make the distinction between thermal
PDB. The distribution of isotope ratios, shown in exposure before or after expulsion from the source
Fig. 3(a), shows that most of the oils group around a rock. The shape of an isotope profile is sensitive to the
value of - 29 per mil. The nearly 3 per mil spread in effects of thermally induced change in oils and can
values suggests differences in the isotope ratios of the often be used to distinguish between increasing
source rock kerogen and, hence, in the original organic maturity in the source and thermal exposure after
input. Similarly, the variation in the ratio of pristane expulsion forfarnilies of oils. Making this distinction is
to phytane (Pr/Ph) for this sample set (see Fig. 3(b» important when timing of generation is considered for
suggests subtle changes in the depositional environ- exploration purposes.
Correlation of Northern North Sea Oils 95

(a)

PLANKTON

TERRESTRIAL

75 50 25
PERCENT C28 STERANE

(b) 1.5,..-------------,---------.-------------,,--------,

ENTIRE NORTH SEA OIL SET

o 0
o
1.0
o 0 o

0.5 20R =5a, 14a, 17a (20R)-C29 STERANE


20S=5a, 14a, 17a (20S)-C29 STERANE
aa =20R
(3(3 =Sa,14(3 17(3(20R)-C29 STERANE

0.5 1.0 1.5 2.0


(3(3/a a
Fig. 2. (a) Distribution of steranes and (b) relative maturity of 41 North Sea oils.
96 Petroleum Geochemistry in Exploration of the Norwegian Sheil

10~----------------------------------~ 11~--------------------------------~
b.
NORTH SEA OILS 10 NORTH SEA OILS
N=57

III III

..
W W •
U 7 U
Z Z
W W 7
0: 0:
:J :J
U
~
.
U
o • o
o ~

.
0:
W

:IE 3
..
a:
w •
:IE
:J ~ 3
Z

-31 -30 -29 -28 -27 -,. .75 1.00 1.25 1.50 1.75 2.00 2.25
CARBON ISOTOPE RATIO (%0 YS POB) Pr/Ph
(a) (b)
Fig. 3. Histogram of (a) aloC Values and (b) Pr/Ph ratios.

Distinguishing the difference between thermal fractionation occurs during the cracking process. The
exposure in the source rock and that occurring after volatile hydrocarbons which result from the cracking
expulsion is possible because the isotope profiles are at first isotopically lighter than the residuum. As
behave differently in each case. The precursor carbon the cracking progresses, the isotopically light carbon
pool (kerogen) in a source rock, which is in the oil is removed from the residuum pool and added to the
generation window, is much larger than that of the volatile pool. Thus while the whole oil b13C value
product (oil). The isotope ratio of the kerogen is remains constant, the volatile fractions become isotop-
changed little by converting some of it to oil, since ically lighter and the residuum becomes isotopically
there is minimal isotopic fractionation between heavier. The resulting effect upon the shape of the
kerogen and whole oil. Similarly, the isotope ratio of isotope profile is opposite to that of increased maturity
the oil generated remains relatively constant with in the source. The magnitude of the change in shape of
increasing maturity. The shape of the isotope profile, the isotope profile depends in part upon the maturity
however, changes during the process. The aloC value of the oil when expelled from the source rock.
of the residuum, which is very similar to the whole oil,
changes very little. During early generation, the
isotope ratios of the distillation cuts can be quite dif-
ferent from that ofthe residuum, either more positive, RESULTS
more negative, or both, depending on kerogen type.
The isotope profile can be quite steep or zig-zagged. Isotope profiles for oils of similar maturity levels (as
As maturity increases, the profile flattens out. determined by sterane rearrangement, discussed
Towards the end of the oil generation window, the above) from Statfjord, Beryl, and Ekofisk Field wells
volatile fractions are increasingly enriched in heavy show that there are distinguishable subfamilies over
carbon with each successive generation step. The vol- the geographic range of this study (see Fig. 4). Both
atile fractions become isotopically heavier than the the shape of the profiles and the residuum b13C values
residuum, creating a steep profile sloping downward. are different for the oils. The Ekofisk oil is isotopi-
An illustration of this case will be presented in the cally heaviest and the profile is relatively flat. The
discussion of the Beryl Field (see Fig. 8(b), below). profiles for the Statfjord and Beryl oils have more
Isotope profiles of oil that is cracked due to con- character to them. Although the biomarker finger-
tinued thermal exposure after expulsion behave in a prints of these oils are nearly identical, this plot shows
markedly different way. This case is a closed system definite source facies variation between the Central
with a relatively small precursor carbon pool. Isotopic and Viking Graben oils. Also, there is evidence of two
different oils reservoired in the Statfjord Field.
Isotope profiles for all of the Statfjord oils, which
were obtained from drill stem tests, are shown in
-26 Fig. 5. Although the shapes of the profiles are nearly
E''(OFISK OIL
identical for the two oil types, they are displaced by
-27 approximately 1.5 per mil. This displacement is a
,:>--------------"-------------------- source-related characteristic. The isotopically heavier
-28
oil occurs in Brent, Dunlin, and Statfjord sand reser-
..
c -2 • voirs, the lighter only in Statfjord reservoirs. These
data suggest several conclusions. The oils reservoired
!:? -30 in the Statfjord sands are not in communication
.0
because distinct accumulations can be recognized. The
-31 two oil types were generated in different source
facies, and therefore probably migrated from separate
-32
drainage areas. Although the drainage area for each
-33
type has not been identified, it is important to know
0 SO 75 100125 150 175 200225250275300 300+ that more than one is involved in any effort to identify
DISTILLATION TEMPERATURE IN OC
migration fairways for nearby fields.
Fig. 4. Isotope profiles of oils from Ekofisk, Beryl, and Other fields in the vicinity of Statfjord are the Hut-
Statfjord Fields. ton and Brent to the SW, and the Thistle to the NW.
Correlation of Northern North Sea Oils 97

-20 appears that these oils may have been derived from
the same source facies as the Statfjord oils, but have
been thermally altered after expulsion from that
-27
source. This information may be useful in conjunction
STATFJORD FIELD OILS with structure maps and temperature gradient data
-28
for mapping the drainage area of Statfjord, Hutton,
Brent, and Thistle Fields.
As stated above, the isotope profile for the oil from
the Ekofisk Field in the Central Graben is markedly
different, both in shape and olac values, from those for
the Statfjord Field. Going generally north from Eko-
-30 fisk to Beryl, there is a regular progression from the
Ekofisk to Statfjord type of profile (see Fig. 7) indicat-
ing a systematic change in the source rocks as one
-31
moves from the Viking Graben into the Central Gra-
50 100 150 200 250 300 300.
ben. Interestingly, the representative from the Mon-
DISTILLATION TEMPERATURE IN ·C trose Field is a high gravity oil. It has an isotope pro-
file characteristic of an oil generated near the end of
Fig. 5. Isotope profiles of Statfjord Field oils. the oil window. Yet its rearranged steranes indicate
that it is less mature than the others in this group. The
-25,--------------------,
steranes occur in low abundance in the Montrose oil. It
appears that this is a high maturity oil, depleted in
-26
THISTLE, HUTTON, AND BRENT FIELD OILS. high molecular weight hydrocarbons, which has
-27
extracted biomarker compounds from less-mature
rock somewhere along its migration pathway.
-28
Finally, the Beryl Field presents an interesting
-29 case. Eleven oils covering a broad range of maturities
(see Fig. 8(a» were studied. The oils were taken from
-30
reservoirs between 9000 and 11 000 ft (2700-3300 m)
-31 deep. There is no consistent correlation between
maturity and reservoir depth. Although stacked
-32
reservoirs in one well have oils whose maturities
-33 increase with depth, the difference in depth is not suf-
-34 L-_-'---L.---'---'----'-----"_L-L--"--L.-'-_----"----..l
ficient to invoke thermal alteration in the reservoir. In
o 50 75 100125150 175200 225250275 300 300+ another case, the shallowest oil in the well is the most
DISTILLATION TEMPERATURE IN °c
mature.
Fig. 6. Isotope profiles of Thistle, Hutton, and Brent Field Seven successive maturity levels have been num-
oils. bered, with 1 as the least mature, for comparison of
their isotope profiles, shown in Fig. 8(b). It is appar-
ent that there is a systematic change in shape from the
steep immature profiles to the flatter mature profiles.
-28 MON There is very little change in the Jl3C value of the
residua. This is an excellent example of successive
-27
episodes of generation and expUlsion from an increas-
-28 ingly more mature source rock within one drainage
ID
Q area. The timing of generation can be determined in a
-29
: relative sense. It may be possible to determine path-
Cj.

~ -30 /'
ways of migration for these oils by integrating the
.0
-31
MAU
timing data with paleo-reconstructed structure maps
OilS FROM FIELDS SOUTH OF BERYL:
of the Beryl Field.
MAU - MAUREEN
-32 F - FORTIES
MON - MONTROSE
A - ARGYLL
-33 e - EKOFISK
CONCL USIONS
-34 "--_-"----'---'----'-----"_L-L--"----'---'----'-_ _L----"
o 50 75 100125 150175200225250275300 300
DISTILLATION TEMPERATURE IN DC Several overall conclusions can be drawn from this
study. Biomarker analyses show that the North Sea
Fig. 7. Isotope profiles of oils from fields south of Beryl. oils have been generated in one common source.
Isotope profiles indicate that different organic facies
Isotope profiles for oils in these fields are very similar of this source can be recognized by identifying varia-
to the major Statfjord oil type in shape, but the Ol3C tions in the isotopic composition of the oils. This
values for the residua are approximately 1 per mil information can be used for determining the continuity
heavier (see Fig. 6). These profiles have all the of reservoir sands and for identifying accumulations
characteristics that would be expected of a Statfjord which have migrated from different drainage areas.
oil which has received thermal exposure after expul- Isotope profiles are also useful for determining the
sion. In fact, in a laboratory experiment, one of the relative thermal maturities of oils in the subfamilies of
Statfjord oils was heated in a bomb for 15 d at 350 DC the North Sea, and whether the maturity was due to
to see the effect on the shape of its isotope profile. The pre- or post-expulsion thermal exposure. The
profile after heating was nearly identical to those from geochemical interpretations, when combined with
the Hutton field (R. M. Squires, pers. comm.). It geological data, can be helpful in determining timing
98 Petroleum Geochemistry in Exploration of the Norwegian She(f

1.5r--------------,---------------.---------------.----------------,
BERYL FIELD OILS
N=9

(a)

1.0

a:
o('\I
"o
C/)

('\I

0.5 20R =5a , 14a, 17a (20R)-C29 STERANE


20S=5a, 14a, 17a (20S)-C29 STERANE
aa =20R
/3/3 =5a,14/3 17/:?(20R)-C29 STERANE

0.5 1.0 1.5 2.0


/3/31 a a
(AFTER SEIFERT AND MOLDOWAN,1981)

-25

(b)
-26 BERYL FIELD OILS
7
-27

-28
m
0
Q. 4
-29 6
.&
> 3
U
...
I? -30
2
.0
RELATIVE MATURITY:
-31
1 - LEAST MATURE
-32 1 7 - MOST MATURE

-33

-34
0 50 75 100 125 150 175 200225 250275 300 300+
DISTILLATION TEMPERATURE IN %C

Fig. 8. (a) Relative maturity of nine Beryl Field oils. (b) Isotope profiles of Beryl Field Oils.
Correlation of Northe1'n North Sea Oils 99

of migration and the history of structural develop- REFERENCES


ment.
Cornford, C., Morrow, J. A., Turrington, A., Miles, J. A. and
Brooks, J. 1983. Some geological controls on oil composition in
the UK North Sea. In: Brooks, J. (ed.), Petroleum Geochemis-
ACKNOWLEDGEMENTS try and Exploration of Europe. Blackwell Scientific, Oxford,
pp. 175-194.
R. M. Squires of Mobil Research and Development MacKenzie, A. S., Lamb, N. A. and Maxwell, J. R. 1982. Steroid
Corporation constructed the isotope profiles used in hydrocarbons and the thermal history of sediments. Nature, 295,
this report. His efforts and discussions are greatly 223-226.
Seifert, W. K. and Moldowan, J. M. 1981. Paleoreconstruction by
appreciated. W. L. Orr, also of MRDC, added valu- biological markers. Geochim. Cosmochim. Acta, 45, 783-794.
able insight to the interpretation of the isotope data. I
thank the management of Mobil Research and
Development Corporation for releasing this work for
outside publication.
7 Oil-oil and oil-source rock
correlation, Northern North Sea
L. Schou*, s. Eggent and M. Schoell+
*Continental Shelf Institute (IKU), trondheim, Norway
tStatoil, Stavanger, Norway
:j:Chevron Oil Field Research Company, La Habra, California, USA

Nine oils from seven different discoveries in the Northern North Sea area and 10 Jurassic shale samples from three
wells in the same area have been analysed and correlated using molecular marker and isotopic properties. The oils
were divided into two groups, the main difference between the groups being the maturity. Highest maturity is
encountered in the two samples from blocks 30/4 and 35/8. The most likely source kerogens for the oils are different
maturity stages of the Draupne and Heather Formations such as in 30/7-7, possibly with some input from the
Drake Formation, represented by the maturity stage in 30/3-1.

INTRODUCTION gas window (deeper than 5 km) (Rj,}nnevik et al.,


1983).
Nine oil samples from seven different discoveries in Potential source rocks are found in several forma-
the Northern North Sea area and ten Jurassic shale tions within the Jurassic. Both the Upper Jurassic
samples from three wells in the same area (Fig. 1, (Draupne and Heather Formations), and the Lower
Table 1) have been geochemically analysed and com- Jurassic (Drake and Burton/Amundsen Formations)
pared. The main objective was to study the relation- have the characteristics of potentially rich source
ship between oils trapped in different structures and rocks for both oil and gas.
potential source rocks in the same area. Based on organic richness and kerogen type, the
Draupne Formation is by far the 'richest' source rock
in the area. In terms of migration efficiency, the
Lower Jurassic and intra-Brent shales may be more
GEOLOGICAL SETTING effective source rocks due to close juxtaposition with
The main tectonic framework of the area studied is carrier/reservoir beds.
shown in Fig. l.
The Northern North Sea Basin is an area that
underwent rifting and crustal stretching during the SAMPLES AND EXPERIMENTAL
'Kimmerian' tectonic episode. Crustal thinning during METHODS
Jurassic to early Cretaceous times resulted in strong
thermal subsidence in the Cretaceous. The depocentre Oil samples
during the Upper Jurassic to Cretaceous was centred Oil samples were obtained from the operators for the
along an axial graben trend (see Fig. 2). different discoveries that were to be studied. The
The Northern North Sea area is a prolific oil- samples were stored in glass bottles for a short time
producing basin. The combination of rich and mature before they were sent for analysis.
source rocks, large structures and excellent reservoir
quality have resulted in oil or gas being found in many
Rock samples
structures in the area.
The main source rocks in the area are believed to be A major problem in correlation studies is to obtain
Jurassic organic-rich shales (Rj,}nnevik et al., 1983). representative samples from a source rock interval.
Shales of Jurassic age are thermally immature on the An attempt was made to solve this problem by picking
Horda Platform, whereas in the graben area oil matur- a large number of samples throughout the formations
ity is reached at approximately 3.5 km. The Jurassic studied, and then mixing them before extraction. One
in the central part of the graben is at present in the analysed rock sample in this study therefore repre-
Petroleum Geochemist,y in Exploration qfthe Norwegian She!f © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 101-117.
102 Petroleum Geochemistry in Exploration of the Norwegian Shelf

~ Block 30/7
~ Block 30/3
illIII1 Block 35/8
® Oil Samples
59 N 59 N
0 E

Fig. 1. Main structural elements, top Jurassic.

sents the average over a several hundred metre thick at 4 °C for 24 h, then filtered to separate the asphal-
interval (see Table 1). tene fractions. The filtrates and the EOM from the
extracts were separated into a saturated, an aromatic
Experimental methods and a non-hydrocarbon fraction using a MPLC system
A flow diagram (Fig. 3) illustrates the different with n-hexane as eluant (Radke et al., 1980).
analydcal techniques which were employed. Gas chromatographic (GC) analyses: all of the GC
analyses were performed on polar glass capillary col-
Total organic carbon (TO C) analyses were run on umns (OV-1 or OV-101). Hydrogen was used as the
samples pretreated with HCI for carbonate removal. carrier gas for the whole oils and the saturated hyd-
The determinations were made on a Leco CR12 car- rocarbons, while helium was used for the aromatic hyd-
bon analyser. rocarbons. Medium length columns (25 and 30 m) were
Rock-Eval pyrolysis: crushed samples (100 mg) used for the whole oils and the saturated compounds
were analysed on a Rock-Eval pyrolyser. while a 100 m column with a specially designed back-
Extractable organic matter (EOM) and chromato- flushing system was used for the separation of the
graphic separation: approximately 50 g of powdered aromatic hydrocarbons. The analyses were performed
rock was extracted by flowblending for 3 min using on three different GCs, an HP 5730 (whole oils), a
dichloromethane (DCM) as solvent. Aliquots of the oil Varian 3700 (aromatic hydrocarbons) and a Carlo
samples were diluted with n-pentane (40 mljmg), left Erba Fractovap 2150 (saturated hydrocarbons). The
Northern North Sea Source Rock Correlation 103

TABLE 1. Source rock and oil samples analysed

Figure
Sample No. Code Well Depth interval (m) Formation

Source rocks
M-9138(31) 30/3-1 3873-3950 Drake
M-9139(32) 30/3-1 4128-4227 Burton/ Amundsen
M-9028(U) 30/7-7 3900-3925 Draupne
M-9029(12) 30/7-7 3925-4200 Heather
M-9030(13) 30/7-7 4200-4500 Heather
:.1-9031(4) 30/7-7 4500-4730 Drake
M-9032(5) 30/7-7 4790-4910 Burton/ Amundsen
M-9235(21) 35/8-1 3180-3350 Draupne
M-9136(22) 35/8-1 3350-3500 Heather
M-9137(23) 35/8-1 3700-3800 Drake

Oils
M-49 31/2-2 1512 Upper Jurassic
M-50 30/4-2 3870 Middle Jurassic
M-51 31/2-5 1503 Upper Jurassic
M-52 31/4-5 2135 Middle Jurassic
M-53 35/8-1 3565 Middle Jurassic
M-54 30/6-4 2630 Middle Jurassic
M-55 34/10-2 3355 Middle Jurassic
M-56 30/6-3 2430 Middle Jurassic
M-57 30/3-2 2870 Middle Jurassic

temperature programs applied were 80°C (2 min) to 35/8-1 plot in the transition area between
260°C at 4°C/min for the hydrocarbon fractions and immature/mature (Fig. 4). The 30/3-1 samples plot in
-50°C (2 min) to 260 °C at 4°C/min for the whole the oil window together with samples from 30/7-7.
oils. The data processing for all the GC analyses was The high maturity of the samples from both 30/3-1
performed on a VG Multichrom system. and 30/7-7 makes their kerogen classification dubious.
GC-MS analysis: the analyses were performed on a However, type II or mixed type II/III can be inferred
VG 70-70H GC-MS-DS equipped with a Varian 3700 from Fig. 4. The deepest sample from 35/8-1 (Lower
GC. Helium was used as the carrier gas, but apart Jurassic) appears to be the only pure type III kerogen
from this the other GC conditions were similar to (see Tables 1 and 2). The elemental analysis and
the single GC analyses. Multiple ion detection (MID) pyrolysis data are in good agreement.
was applied for the saturated steranes and terpanes, Isotope data on kerogens and extracts as well as
and the ratios from the mass chromatograms were elemental analyses are plotted in Fig. 5 and summar-
calculated from peak heights in the appropriate ized in Table 3. It is evident that no dramatic change
chromatogram. in isotopic composition takes place with increasing
Isotope analysis: the methods for isotopic analysis depth of burial. Neither is there a relationship bet-
are as described in Schoell et al. (1983). The internal ween geographic position and isotopic composition of
standard NBS 22 lubricating oil (013C) was run eight the kerogens. Only the samples from 30/7-7 show an
times during the period of analysis, and revealed a increase in negative numbers versus depth. Six out of
value of -29.73 ± 0.05. 10 kerogens have an almost identical isotopic composi-
tion, with a mean value of -25.4±0.1% (three
samples from 30/7-7 and all samples from 35/8-1).
This could be an effect of the way the samples were
RESULTS AND DISCUSSION averaged before analysis.
The parallel change in the isotopic composition of
Source rocks the kerogens and the extracts suggests that most of
Data on the source rock intervals studied are pre- the extracts are indigenous. The two lowermost
sented in Tables 1 and 2. Samples have been picked samples of 30/7-7 are strongly depleted in heavy
from wells where the Jurassic formations are within C-isotopes in the NSO and aromatic fractions (Fig. 5),
the oil window (3-5 km) at present. However, few although this may be a result of incomplete compound
wells have been drilled in the deeper part of the gra- class separation.
ben, thus limiting the choice of suitable locations in The general distribution of the terpane and sterane
this area. biomarkers confirms that there are only relatively
The organic content of the studied source rocks var- small variations between the various formations in the
ies from 1.6% to 4.32%, the richest being in the three wells. The main differences are thought to be
Draupne and Heather Formations of wells 30/7-7 and due to differences in maturity. The steranes in the
35/8-1. The Lower Jurassic shales have less than 2% Draupne and Heather Formations show a very similar
organic carbon content. It should be borne in mind distribution; the main difference seen is the maturity-
that these TOC values represent avemges, and as related increase in 14P ,17p-steranes and in %20S
such they fit well with the natural spread in values 14 a,17 a-steranes with increasing burial depth. A
published by Ronnevik et al. (1983). similar picture is seen in the terpane distribution
The kerogens were separated out and analysed, (shown in Fig. 6). The higher maturity of 30/7-7 is
both for elemental and for isotopic composition. In a seen both in the absolute amounts of C30 hopane and
modified van Krevelen diagram the samples from in the maturity ratios. The Tm/Ts ratio (Seifert and
104 Petroleum Geochemistry in Exploration of the Norwegian Shelf

5 E
~~ __~______~ 62N

bO N

KEY ISOCHRON MAP


____ < 2.0 !lillllt.~ l.O - .1,5M1t BASE OF CRETACEOUS
i:=J 2 ,O ~ 1:!I '.e. _ .l..~- 4.C1
o Conto UIS based 0 n two way return ti me
$01) 1:. .

~ 4.0 elltc. .
7 .5-:1..0 "8' , _
(interval 0.5 seconds)

Fig. 2. Isochron map, base Cretaceous.

TABLE 2. Results of elemental analyses of kerogens

Figure code Sample no. Well Depth (m) e (%) H (%) 0(%) Ash (%) H/e Ole
11 M-9028 30/7-7 3900-3925 26.86 1.31 2.17 40.0 0.585 0.061
12 M-9029 30/7-7 3925-4200 30.41 1.50 2.38 42.3 0.592 0.059
13 M-9030 30/7-7 4200-4500 27.78 1.50 2.48 42.0 0.648 0.077
14 M-9031 30/7-7 4500-4730 27.88 1.11 3.11 42.6 0.478 0.084
15 M-9032 30/7-7 4790-4910 29.37 1.23 3.34 41.8 0.502 0.085
21 M-9135 35/8-1 3180-3350 33.14 2.49 4.36 37.7 0.902 0.099
22 M-9136 35/8-1 3350-3500 24.08 1.85 3.24 45.4 0.922 0.101
23 M-9137 35/8-1 3700-3800 15.20 1.00 3.40 52.6 0.789 0.168
31 M-9138 30/3-1 3875-3950 29.28 1.77 3.46 42.3 0.725 0.089
32 M-9139 30/3-1 4128-4227 34.94 1.88 4.71 37.5 0.646 0.101
Northern North Sea Source Rock Correlation 105

I COMPOSITE CUTTINGS
I I CRUDE OILS I
t
TOTAL ORGANIC I WHOLE OIL GC I
CARBON
EXTRACTION
- DCM
ROCK-EVAL
PYROL YSIS

I
C- AND H-ISOTOPES
ON KEROGEN
.. BITUMEN/OIL


CHROMATOGRAPHIC
SEPARATION, MPLC

t t t
I
l ,
RESIN (NSO)
I I AROMATIC HC'S

t
I SATURATED HC'S

t
I ,
ASPHALTENES

I C- AND H-ISOTOPES I I GC
I GC l C- AND H-ISOTOPES J
• t
l ,
GC-MS, TIC
J GC·MS, MID
m/z 177, 191, 205,

I C- AND H·ISOTOPES I 217,218,259

t
C- AND H-ISOTOPES

Fig. 3. Flow diagram, analytical procedure.

TABLE 3. Results of isotopic analysis of kerogens and extracts

SHC AHC NSO Kerogen

Figure code Sample no. Depth (m) Well 0 13(: oD 0 13(: oD 0 13(: oD 0 13(:- oDb

11 M-9028 3900-3925 30/7-7 -27.7 -113 -26.0 -116 -25.8 -110 -25.4 -105
12 M-9029 3925-4200 30/7-7 -27.7 -114 -26.0 -109 -25.6 .'-105 -24.4 -106
13 M-9030 4200-4500 30/7-7 -27.9 -120 -26.4 -105 -26.1 -104 -24.6 -105
14 M-9031 4500-4730 30/7-7 -28.3 -109 -27.6 -103 -27.9 -105 -25.3 -103
15 M-9032 4790-4910 30/7-7 -28.4 -117 -27.4 -105 -27.7 -103 -25.5 -130
21 M-9135 3180-3350 35/8-1 -28.4 -138 -26.8 -124 -26.7 -124 -25.6 -119
22 M-9136 3350-3500 35/8-1 -28.0 -131 -26.6 -116 -28.8 -125 -25.5 -123
23 M-9137 3700-3800 35/8-1 -27.9 -121 -26.3 -118 -26.4 -112 -25.3
31 M-9138 3873-3950 30/3-1 -28.7 -124 -27.2 -121 -119 -26.3 -lll
32 M-9139 4128-4227 30/3-1 -28.5 -123 -27.3 -115 -27.5 -112 -27.0 -106

_Decarbonated.
bDemineralized.
106 Petroleum Geochemistry in Exploration of'the Norwegian Shelf

1.5

1.0

Well 35/8 - 1

0.5

Overmature Mature Immature

0.1 0.2 0.3

Atomic OIC
Fig. 4. Elemental analysis of kerogen (Van Krevelen diagram).

H/e

0.5 1.0 -28 -26 -24


3000

'1111111111111111I11I1::n:111I11I11I11I1I111I111111111 35/8 - 1
E
'111111111111191371111111111111111
.c 9028 ~
~ 4000
Cl
..--_ _ _3_0/-,3 - 1 1
Wells

1 35/8 -1
30/3 - 1
30/7 - 7
30/7 - 7

5000 / I "-
SHe NSO AeH
Fig. 5. Variation ofH/C and b13C in kerogen and compound classes with depth. Sample numbers (9135, etc.) refer to Table 2.
Northern North Sea Source Rock Correlation 107

DRAUPNE FM. HEATHER FM.


HOT SHALE

Well 35/8 -1
Well 35/8 -1 3350 - 3500m
3180 - 3350m

Well 30/7 -7 Well 30/7 -- 7


3900 - 3925m 3925 - 4200m

We1l30!7 -7
4200 - 4500m

~-------- -------

Fig. 6. Terpane distributions in Draupne and Heather Formations (for key to peak identification, see Table 6).

Moldowan, 1978) has a very low value in the top two from the lower concentration of terpanes in 30/7-7.
samples in 30/7-7, compared with the other three This concentration difference is probably caused by
samples. These uppermost two samples also contain a the difference in maturity of the two wells.
high relative amount of an unknown compound (X), The Drake and Burton/Amundsen samples in the
that is very often seen in samples where the general three wells show the same increase in the sterane
concentration of terpanes is low. maturity parameters with increasing depth, as was
The difference between the two Heather samples in seen for the Draupne and Heather samples. The main
30/7-7 may be an effect of the composite samples that difference seems to be a higher input of C29 steranes
were used. From the terpane distribution, the upper- in the 30/7-7 sample. In the terpane distribution the
most sample of the two seems to be a mixture of the picture is somewhat more complex (Fig. 7). In 30/7-7,
Draupne one above and the one below. This could be the absolute content of hopanes in both the Drake and
due to caving of Draupne material into the uppermost the Burton/Amundsen Formations is increased by a
Heather interval, or it may reflect a more-or-Iess con- factor of 10 compared with the Heather samples in this
tinuous change in character from Draupne down well. The Burton/Amundsen sample in 30/3-1 is dif-
through the Heather. ferent from the others, in that the lower molecular
The lowermost Heather sample in 30/7-7 is similar weight tricyclic terpanes are more abundant in
to the Draupne and Heather samples in 35/8-1, apart 30/3-1.
108 Petroleum Geochemistry in Exploration of the Norwegian Shelf

DRAKE FM. BURTON/AMUNDSEN FM.

Well 30/3 - 1
3873 - 3950m Well 30/3-1
4128 - 4227m

Well 30/7 - 7 Well 30/7 - 7


4500 - 4730m 4790 - 4910m

I
(:.J~~l ,I".~_,_ _ _
Fig. 7. Terpane distributions in Drake and Burton/Amundsen Formations (for key to peak identification, see Table 6).

Thus, most of the maturity ratios show fairly good Two liquid samples from an oil/condensate field (35/8)
depth relationships, the only ratio that does not is the and a gas/condensate field (30/4 and 30/7) situated on
Tm/Ts ratio, which has no clear trend. This may be the flanks of the Viking Trough were also analysed, as
due to variations in organic matter input. It is appar- was one sample from the Tampen Spur area (34/10).
ent that the Drake and Burton/Amundsen samples The oils were all from Jurassic reservoirs at depths
from 30/7-7 and the Burton/Amundsen sample in ranging from 1500 m to nearly 4000 m (see Table 1). A
30/3-1 are significantly different from the other decrease in density (API gravity) related to depth is
samples. evident in Fig. 8. The two deepest samples may be
In summary, all of the intervals, apart from the classified as condensates.
deepest Drake in 35/8-1, may have potential for The isotopic data show that within the six shallow-
generating oil. The Draupne and Heather intervals in est oil samples there is a trend towards more negative
35/8-1 are immature, while all the other samples are C- and D-isotope values with increasing depth (Fig. 8
from within the oil window. and Table 4). The two deepest samples (from 30/4-2
and 35/8-1) seem to fall on a different trend.
The gross composition, determined by chromato-
Oils
graphic separation, is presented in Table 5. All of the
The majority of the oils analysed were taken from oil oils analysed have more than 60% light components.
and oil/gas discoveries on the Horda Platform (Fig. 1). Three of the oils (31/2-5, 30/6-4, 30/3-2) have high
Northern North Sea Source Rock Correlation 109

APlo 013C (%0) 00 (%0)

20 30 40 50 60 70 -30 -28 -26 -140 -130 -120 -110


/
/
1500 L0==t'i-o-'D

2000
~
E [ill
2500
....c-
.:::.
()'r----{]
O.>
Cl fr--O----D
3000

D------------fr
3500
0-0
0 SHC
4000
0
[).
AHC
NSO " ~

Fig. 8. Variation of gravity of oils and isotopic composition of compound classes with depth. Sample numbers (49, etc.) refer
to Table 4. I and II are proposed oil groups.

TABLE 4. Reservoir data and isotopic composition of the oil fractions


SAT ARO HET

Sample 110. Depth (m) API (degrees) GOR (1I/3/ m 3) 0 13 on o13G oD oUG oD

M-51 1503 27.6 66 -28.8 -129 -27.8 -124 -27.8 -124


M-49 1512 24.0 59 -29.0 -120 -28.0 -ll6 -27.9 -131
M-52 2135 35.5 82 -28.0 -126 -28.2 -127 -27.9 -124
M-56 2430 54.6 3230 -28.2 -27.9
M-54 2630 32.2 100 -29.6 -135 -28.3 -127 -27.8 -134
M-57 2~70 38.9 84 -29.6 -138 -28.7 -128 -28.3 -143
M-55 3355 43.8 200 -29.4 -122 -28.0 -141 -27.1 -126
M-53 3565 44.4 3953 -27.5 -122 -26.6 -ll7 -25.9 -122
M-50 3870 42.7 4715 -28.2 -ll5 -26.9 -ll5 -26.3 -122

TABLE 5. Gross composition

SatHG AroHG HG NSO Asph. Light HG


Sample Amount of
no. oil (II/g) (mg) (%) (11109) (%) (mg) (%) (11109) (%) (mg) (%) (lng) (%)

M-49 178.7 48.1 27 18.7 10 66.8 37 5.0 3 0.2 0.1 106.7 60


M-50 219.1 49.1 22 9.8 4 57.9 26 3.9 2 0.9 0.4 156.4 71
M-Sl 202.2 37.1 18 19.3 10 56.4 28 13.9 7 3.2 1.6 128.7 64
M-S2 205.2 44.4 22 16.3 8 60.7 30 9.2 4 1.5 0.7 134.0 65
M-S3 202.4 45.8 23 12.9 6 58.7 29 3.8 2 0.3 0.2 139.6 70
M-54 195.4 42.5 22 19.3 10 61.8 32 12.4 6 3.5 1.8 177.7 60
M-55 208.6 41.0 20 11.1 5 32.1 25 1.9 1 0.01 0.00 154.6 74
M-56 215.3 20.5 10 7.2 3 27.1 13 0.8 0.3 0.1 0.05 187.3 87
M-57 219.6 49.6 23 14.4 7 64.0 29 7.9 4 3.8 1.7 143.9 66

asphaltene content (more than 1%). The hydrocarbon tive of the three different types of oil. GC analyses of
content varies from 17% to 37%. the saturated hydrocarbon fractions distinguish only
Gas chromatographic separation of the oils indicates two of the groups, since the C15+ hydrocarbons have
that there are significant differences in some of the the same general distribution in all samples apart from
samples. Whole oil gas chromatograms (nine oils) the most heavily biodegraded one. Aromatic hydro-
reveal that the general distribution of hydrocarbons is carbons exhibit only small variations betwen the oils.
similar for seven of the oils (represented by the oil GC-MS analyses show that the sterane distribution
from 30/4 in Fig. 9). The exceptions are the two oils is very similar in all the oils, although some differ-
from 31/2, which seem to have undergone different ences are seen within the terpanes (see Fig. 10). Vary-
degrees of biodegradation: one shows a significant loss ing amounts of 28,30-bisnorhopane and tricyclic ter-
of light compounds, while the other has lost all panes are seen, in addition to more maturity-related
n-alkanes. Figure 9 shows the GC traces representa- variations. From these data there seems to be two
110 Petroleum Geochemistry in Exploration of the Norwegian Shelf


• • 31/2 - 2

• •
• • •
• * •
Pr

l~·~\W~1~j, [L~ ,;I '~e' JJ,~jJ1jLl;: . .


c
c
l"
l !.~
30/4 - 2

··
• •
A •

I
I
I


.. • •

~,1 1\ ~JAlw.l.JJJdJ~iJ
Ii··.
,

I I

I ** * * Pr
' •• •

j .UJLiii ,
31/2 - 5

.j1r II,
1

• n-alkanes
* isoprenoids
C cyclic compounds
A aromatic compounds
Fig. 9. Whole oil gas chromatograms.

main groups of oils, which is in good agreement with of 14{3,17{3 C29 steranes is shown in Fig. 11. A fairly
the isotopic data. The oils from 30/4 and 35/8 form a good linear increase occurs with depth for source rock
different group (group II) from the rest of the oils values up to 70-75%. Cornford et al. (1983) state that
(group I), see Fig. 8 and Table 4. the upper part of the curve with the biggest changes
in the values is in the early mature range, while the
deeper source rocks are moderately mature. All of the
Maturity of oils and source rocks oils fall within a narrow range with values between
In an attempt to compare the maturity of the oils with 70% and 80%. These data would suggest that source
source rock maturity, plots of some maturity ratios rocks of maturities similar to those below approxi-
against depth for both oils and source rock samples mately 3800~4000 m have sourced these oils.
were made. The plot representing the relative amount A similar relationship was seen for the relative
Nonhern North Sea Source Rock Correlation 111

E
mil 191 M 49 mIl 191 M 55

31/2 - 2

E E
mil 191 M51 mil 191 M57

31/2 - 5 30/3 - 2

<:

,~
E A

I
m/z 191 M 52 mIl 191 M 50

31/4 - 5 <: EI 30/4 - 2


". B !I·
;;: i.
<: " I III, II
. )' ,I·}.:ill
'_',,,,,...,'l,J\. 1'1

"'J~" 1H J
""-JV~rJ,J( K
--~-~-

E m/z 191 M 54 E
mfl 191 M 53

,-~_~_~_J_A_B_nZ_~_D__, -L-, = ,-K~ L ,~~~d_~,jlJJIJ:tL~_


30/6 - 4
35/8 - 1

Fig. 10. Terpane distribution in oil samples (for key to peak identification, see Table 6).

TABLE 6. Mass chromatograms representing terpanes (m/z 191)


A Ts, 18a(H)-trisnorneohopane C27H46 K 17a(H)-trishomohopane (22S,22R)
B Tm, 17a(H)-trisnorhopane CZ7H 46 L 17a(H)-tetrakishomohopane (228,22R)
C 17a(H)-norhopane CwH50 M 17a(H)-pentakishomohopane (22S,22R)
D 17p(H)-normoretane CwH50 Z bisnorhopane
E 17a(H)-hopane C30H5Z X unknown triterpane
F 17p(H)-moretane C3oH52 p tricyclic terpane
G 17a(H)-homohopane (228) C31 H 54 Q tricyclic terpane
H 17a(H)-homohopane (22R) C31H 54 R tricyclic terpane (17R,178)
+ unknown triterpane (gammecerane?) 8 tetracyclic terpane
I 17P(H)-homomoretane T tricyclic terpane (17R,17S)
J 17a(H)-bishomohopane (228,22R)

Letters refer to Figs. 6, 7 and 10.


112 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Depth
(m) 50 60 70 80
1000.-----~----------~--------~-----------L---

2000
52

56

54

57

3000

35/8 - 1

}:IM 55

53
.... •
.r:::: !:? IDR

I
0
o M 50
MK][oR *Ii.
4000
H III
Jlli!1 A ][

I
H

DR I
IB/A I
5000

* Oils
IT Source rocks
Fig. 11. Depth plot of percentage fJfJ C29 steranes. Sample numbers (49, etc.) and formations refer to Table 1.
Northern North Sea Source Rock Correlation 113

Depth
MPI1
(m) 0.3 0.5 0.7 0.9
1000------L-----L---~-----L----J-----L---~-------

51 49
• •

2000 52

51i

54

57

3000
35/8 - 1

I:'· I ]
55


53

I
,....
~
CI
~ l[OR
CI
M
j:i i 50.
MK 1@R ]IT
H

H
4000
J!!/A I 1II

DR
I
5000

• Oils
IT Source rocks
Fig. 12. Depth plot of methyl phenanthrene index, MPI 1. Sample numbers (49, etc.) and formations refer to Table 1.
114 Petrolcum Geochemistry in Exploration of the Norwegian Shelf

-100-r----------------=-~~----------------------------~

~ Oils

o Source rocks

0
-120
CD 35/8 - 1
'<l

CD 30/3 - 1
CD 30/7 - 7

-140
0
-30 -28 -26 -24

o13 C(%o)
Fig. 13. Comparison of oils and extracts in a 813C/8D cross-plot. ARC = Aromatic hydrocarbons.

amount of 208 steranes, but for this parameter the than the 31/2-2 oil. The least mature oils appear to be
trend is less clear. The oil samples have values in a the three oils of intermediate depth, from a depth of
fairly narrow range, with the two deepest samples 2500-3000 m. The deepest three oils have a maturity
(from 30/4-2 and 35/8-1), and possibly the samples similar to source rocks from below 3600 m.
from 34/10-2 and 30/3-2, being slightly more mature These somewhat contradictory conclusions from the
than the rest. saturated and aromatic maturity parameters may be
The shallowest two oils, from 31/2-2 and 31/2-5, explained in two ways. Either the aromatic ratio gives
have undergone different degrees of biodegradation. the correct order of maturity of the oil samples-the
Based on the sterane ratios the maturity of the two sterane maturity ratios are so close to equilibrium that
samples is approximately the same, and the distribu- the variations may not be significant-or the differ-
tion of biomarkers in general is quite similar in both ence between the two sets of data may reflect a mixed
oils. A difference can be seen in the terpane distribu- input from source rocks of varying maturity.
tion, in that the abundance of tricyclic terpanes is
higher in the 31/2-2 oil and the Tm/Ts ratio is smaller Correlation of oils and source rocks
in the same well, suggesting that the oil from 31/2-2 is
of slightly higher maturity than that from 31/2-5. The A genetic relationship between a kerogen and an oil
relative amounts of 28,30-bisnorhopane are the same. may exist when both are isotopically similar. This is
An explanation of the difference in the terpane dis- the case only if oils and kerogens and their extracts
tribution may be that the same source has generated have not experienced severe maturation. In most
both oils, but at slightly different maturity levels. The cases, the aromatic fraction is least affected by secon-
oil of lowest maturity probably underwent biodegra- dary processes (Lewan, 1982; Schoell, in press). Fig-
dation first, and a more mature originally non- ure 13 compares the results of carbon and hydrogen
degraded oil subsequently filled the reservoir in well isotopic analyses of the aromatic fraction both from
31/2-2. This observation parallels the conclusion of the oils and rock extracts. The extracts reveal a con-
Cornford et al. (1983) that North Sea oils generally are siderable variation in the isotopic concentration, such
'a mixture of products generated from a single source that the three wells are distinguishable. The isotopic
rock horizon buried to progressively deeper levels variations in extracts from 30/7-7 are significantly dif-
towards the basin centre'. ferent from both groups of oils, but those from wells
Similar plots of maturity-related ratios from the 35/8-1 and 30/3-1 are similar to the isotopic variations
aromatic hydrocarbon distribution show a somewhat of oil group II (30/4 and 35/8) but are distinct from
different picture. The methyl phenanthrene index group 1. A similar result is obtained when carbon
(MPI-l) used by Radke et al. (1980) gives a very good isotope concentrations of kerogens and the aromatic
linear relationship for the source rocks with depth, fractions of the oils are compared. Assuming that oils
with MPI-l values between 0.3 and 0.9 (see Fig. 12). could be genetically related to kerogens if the aroma-
This corresponds to a vitrinite reflectance range of tic fraction ofthe oil is not depleted by more than 1.5%
approximately 0.5-1.0% Ro. The ratio indicates that of the heavy isotope, several oils and kerogens can be
the shallowest oils are the most mature, the heavily genetically linked. Kerogens from the lower Jurassic
biodegraded oil in 31/2-5 being slightly less mature in 30/3-1 are potential sources for the oils of group I,
Northern North Sea Source Rock Correlation 115

Depth BISNORHOPANE/HOPANE
(m) 0.1 0.2
1000~----~----~----~----~----~-------------------

2000

.56

3000

4000

5000

Fig. 14. Depth plot of the relative content of bisnorhopane in oil samples. Sample numbers (49, etc.) refer to Table 1.

and various other kerogens are possible candidates for Horn Basin, where an isotopic shift in the aromatic
group II. and NSO-fraction ofthe oils was attributed to thermal
The genetic significance of the grouping of the oils is alteration (Chung et aI., 1981).
equivocal. The isotopic differences between the two The oils from 35/8-1 and 30/4-2 are the deepest of
groups could be due to the primary isotopic differ- all the oils studied, and are characterized by the high-
ences in the source rocks. The clear relationship bet- est gas/oil ratios. This may suggest that these oils
ween the isotope data and depth and gravity suggests have undergone considerable thermal alteration. Thus
that the differences found may be due to thermal mat- the isotopic concentrations of these oils may not be
uration of the oils in the reservoirs. This could mean representative of their source. Since the causes of the
that the oils of group II formed from oils of group I isotopic shift are not known, any further inference
through thermal alteration. An almost identical would be speculative.
change in isotopic composition was observed in the Big Assuming that all of the oils were derived from an
116 Petroleum Geochemistry in Exploration of the Norwegian Shelf

isotopically uniform source, then this source rock are indigenous to the kerogens, and thus no great
would be expected to have an isotopic concentration input of migrated hydrocarbons would be expected.
more negative than the oils (and the source rocks) in The main question is, what formation (or forma-
this study. Upper Jurassic sections in this area have tions) is the source for these oils? It is fairly well-
been seen to have isotopic concentrations that change documented (Cornford et al., 1983) that the main
gradually from -30%0 wac) in the uppermost part source for most North Sea oils is the Upper Jurassic
(Draupne Formation) to less negative values in the Draupne Formation, with additional inputs from other
Heather Formation. If it is assumed that the sections underlying Jurassic Formations. The oils in this study
with the lowest values are relatively thin, then the are all believed to be sourced from one type of source
sample averaging may have caused a loss of informa- rock, the main difference between the oils being in
tion. maturity. If we assume that the most mature oils
An alternative explanation for the differences in the (group II oils) were expelled at higher maturity levels
characteristics of the oils would be that the same type of the source rock, the relative content of bis-
of source has generated all the oils, but at different norhopane is simple to explain. Based on the isotopic
stages of maturity. data this would mean than the 30/3-1 Drake sample
Based on the distribution ofthe sterane biomarkers, represents an interval that is most similar to the
the most likely source for the oils are the Draupne and highly mature oils (group II). For the less mature oils
Heather Formations in 30/7-7 and the Drake Forma- it is necessary to look for source rocks with more
tion in 30/3-1. The Drake and Burton/Amundsen negative carbon isotope values. In this respect none of
intervals in 3bj7-7 are not likely to have generated any the Upper Jurassic samples analysed can be excluded,
of these oils based on the sterane distributions. How- due to the lack of detailed information which unfortu-
ever, the Burton/Amundsen interval in 30/3-1 cannot nately resulted from the composite sampling method
be excluded as a source, while in well 35/8-1 it is more used.
difficult to correlate due to the low maturity levels.
The terpane distributions of the source rocks sug-
gest that the Draupne and Upper Heather intervals in
30/7-7, and possibly the Drake in 30/3-1, are most CONCLUSIONS
similar to the group II oils. The most characteristic
features are the low Tm/Ts ratio and the relatively
1. The oils can be divided into two groups, one con-
high abundance of the component X. The main differ-
taining the deepest oils from 35/8-1 and 30/4-2 (group
ence between the oils and these source rocks is the
II) and the other group containing the rest of the oils
higher content of tricyclic terpanes in the latter. The
(group I). The grouping of the oils is likely to be
group I oil samples show a terpane distribution fairly
related to differences in maturity, rather than to dif-
similar to the lowest Heather, the Drake and Burton/ ferences in the organic matter.
Amundsen intervals in 30/7-7 and the Draupne and
2. The isotope values obtained for Draupne samples
Heather in 35/8-1. in this study are less negative than those normally
The main difference between these source rock
encountered in this formation. This is probably due to
samples and the group I oils is the presence of bis- loss of significant information through the use of com-
norhopane in all the group I oils, and the total lack of posite samples. The source rocks for the oils are
this compound in the source rocks. This could mean expected to have b13C values of -30%0 to -29%0.
that none of the source rocks analysed has generated 3. Excluding the isotopic data, the most likely
any of the group I oils, or the oils may have extracted source kerogens for the most mature (group II) oils
the bisnorhopane during migration through sediments seem to be represented by the Draupne and Heather
with a high content of this compound. It may, how- Formations in well 30/7-7, and possibly the Drake
ever, be a maturity-related effect. According to Corn- Formation in well 30/3-1.
ford et al. (1983) the relative content of bisnorhopane 4. The less mature group I oils are most probably
decreases with increasing depth and maturity of sourced from the same type of kerogen, but at a lower
kerogens. Thus, an earlier mature stage of the 30/7-7 maturity level.
and 30/3-1 kerogens may have contained bis- 5. The Drake and Burton/Amundsen Formations in
norhopane, and all the group I oils could have been well 30/7-7 are not believed to have contributed sig-
generated at lower maturity than the group II oils. A nificantly to any of the oils.
depth plot of the relative content of bisnorhopane 6. This study also confirms that the abundance of
shows a good relationship for all the oils (see Fig. 14), bisnorhopane in North Sea oils can be related to
indicating that the shallowest oils could have been depth, and thus probably to maturity, indicating that
generated from the same but less mature source rocks care should be taken when this compound is used as a
than the deeper oils. source-related parameter.
Huc et at. (1985) have presented a possible reverse
trend for the bisnorhopane concentration. Over a
fairly thin Upper Jurassic section they found an
increase in the bisnorhopane content with increasing
depth. The maturity of this section was, however, ACKNOWLEDGEMENTS
very uniform, and the effect of increased bisnorhopane
content is probably due to changes in organic matter, We would like to acknowledge the following oil com-
rather than maturity variations. panies for giving permission to present these data: BP
The Burton/Amundsen interval in 30/3-1 is very Norge A/S; Gulf Oil corporation, Norway; Norsk
different from all the oils due to the presence of a very Hydro a.s.; A/S Norske Shell and Statoil.
high abundance of tri- and tetracyclic terpanes. This In addition, we would like to thank the technical
could be due to migrated hydrocarbons in this sample, staff at IKU for the analytical work, and also Tone
although the isotopic data suggest that the extracts Aas who made all the figures.
Northern North Sea Source Rock Correlation 117

REFERENCES 'Stable Isotopes in Hydrocarbon Exploration', Stavanger,


October 1982.
Radke, M., Willsch, H. and Welte, D. H. 1980. Preparative hyd-
Chung, H. M., Brand, S. W. and Crizzle, P. L. 1981. Carbon isotope rocarbon group type determination by automated MPLC.
geochemistry of Palaeozoic oils from the Big Horn Basin. Analyt. Chem., 52, 406-411.
Geochim. Cosmochim. Acta, 45, 1803-1815. Rilnnevik, H., Eggen, S. and Vollset, J. 1983. Exploration of the
Cornford, C., Morrow, J. A., Turrington, A., Miles, J. A. and Norwegian Shelf. In: Brooks, J. (ed.), Petroleum Geochemistry
Brooks, J. 1983. Some geological controls on oil composition in the and Exploration of Europe. Geol. Soc. Spec. Publ. 12, 71-93.
U.K. North Sea. In: Brooks. J. (ed.), Petroleum Geochemistry (Blackwell Scientific, Oxford.)
and Exploration of Europe. Ceol. Soc. Spec. Publ. 12, 35-50. Schoeli, M., Faber, E. and Coleman, M. L. 1983. Carbon and hyd-
(Blackwell Scientific, Oxford) rogen isotopic compositions of NBS 22 and NBS 21 stable isotope
Huc, A., Irwin, H. and Schoell, M. 1985. Organic matter quality reference materials: an inter-laboratory comparison. Org.
changes in an Upper Jurassic Shale section from the Viking Geochem., 5, 3-6.
Graben, (this volume). Schoeli, M. in press. Paper submitted to Geol. Jb.
Lewan, M. D. 1982. Effects of thermal maturation on stable organic Seifert, W. K. and Moldowan, J.M. 1978. Applications of steranes,
carbon isotopes as determined by hydrous pyrolysis of Woodford terpanes and monoaromatics to the maturation, migration and
Shale. Paper presented at Norwegian Petrol. Soc. Mtg 6409 source of crude oils. Geochim. Cosmochim. Acta, 42, 77-95.
Part II
Modelling
8 Modelling of petroleum genesis
and migration with a
bidimensional computer model in
the Frigg sector, Viking Graben
P. Ungerer*, A. Chiarellit and J. L. Oudin+
*IFP, France
tSNEA(P)-ELF, France
+CFP-TOTAL, France

Expulsion of oil from source rocks and subsequent migration towards traps are assumed to take place as a discrete
phase distinct from water. The main cause of expulsion is compaction of sediments during burial, while secondary
migration is ruled by buoyancy. They are described by a diphasic and bidimensional computer model based on
Darcy's law and the relative permeability concept. The extent of formation of oil at depth is provided by a kinetic
model, using the knowledge of basin thermal history and distribution of organic sedimentary facies.
The example presented is a typical tilted block structure of the Frigg sector. According to this diphasic model,
hydrocarbon expulsion began around 70 Ma from the Brent coals, and a few million years later from the Kim-
meridge Clay and Heather Formations, leading readily to an accumulation in the Brent sandstones at the top of the
block. This accumulation developed up to the present time, and partial rlysmigration occurred through the caprocks
at the top of the block.
Although gas migration should be computed by a triphasic model, it has been attempted to simulate it by the
same diphasic model, using low density and viscosity values for the hydrocarbon phase. The model then found an
extensive dysmigration, leading additionally to an accumulation in the Paleocene sands, as is actually found. The
heavy oil ring in the Frigg field is thought to be the result of a migration in the gas phase, the dissolved oil being
condensed during the upwards mi!,'Tation of the gas because of the important pressure drop, and later being
biodegraded.

INTRODUCTION assess the quality of the source rocks, which are of


Jurassic age. Among these the richest are: the Kim-
The final aim of migration modelling-which implies meridge Clay Formation, which contains organic mat-
necessarily oil generation modelling-is to give pre- ter of generally good quality; the Heather Formation,
dictions about the location and volume of oil accumula- generally less rich but thicker; and Brent coals, which
tions in a given basin. However, work is still in pro- may produce oil (Durand et al., 1983), with their low
gress, and the present step is to simulate the migration yield compensated by a high organic content.
processes in well-documented cases with numerous The geological features of the North Sea are well-
control parameters: pressures, volumes of proven documented in terms of the stratigraphy, the
accumulations, and so on. Once the model has proved sedimentology and the tectonic history. The present
to be satisfactory in such a case, its predictions may structure of the Viking Graben is the result of a rifting
become reliable for other locations in the basin, pro- period that occurred mainly during Middle and Late
vided a minimum of data is available. In any case, the Jurassic. The rifting declined progressively during
use of the migration model needs some geological and Late Cretaceous and was followed by regional subsi-
geochemical data from neighbouring wells. If the bur- dence until the present time. The Triassic and Jurassic
ial history, the lithology of the series or the richness of sediments were extensively faulted during the rifting
the source rocks are unknown, it is impossible to period, which results in tilted block structures, as
quantify the amount of hydrocarbons generated, and seen on E- W cross-sections (Fig. 1). The bottom part
any study of migration is impossible. of these blocks is commonly 5000-8000 m deep in the
The Viking Graben, North Sea, is a typical area central part of the graben, whereas their top is often
where the processes of oil genesis and migration may truncated by the Cimmerian unconformity, due to a
be reasonably reconstructed. Most of the wells have phase of erosion or non-deposition in the Late
been drilled since 1970: thus a lot of reliable geochemi- Jurassic. The Jurassic and older sediments are capped
cal data has been collected. These data enable us to by a continuous Cretaceous and Tertiary series, com-
Pell'Oleum Geochelllistl~11 ill Exp/()mti()/l orthe NonJlcgian Shelr © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 121-129.
122 Petroleum Geochemistry in Exploration of the Norwegian Shelf

wsw ENE
MIOCENE TO PLEISTOCENE - -
OLIGOCENE --
MIOOLE & UPPER EOCENE __ --~--=----=-=~=-=~-=~=~=-~==-=--= ~----- --- ---=-= ---=-=-=-----------=--:--=-:--=-====-~~.=- - .'
LOW. EOCENE - PALEOCENE __

UPPER CRETACEOUS -

LOWER CRETACEOUS -
KIMMERIDGE CLAY -
CALLOVO-OXFORDIAN _ 5
BRENT-
STATFJORD -
o 1 2 3 4 5km
TRIAS _

Fig. 1. Schematic section of part of the Viking Graben in the North Sea. The framed zone is the one selected for application of
the model. It is bounded by the two largest faults, which define a tilted block below the major unconformity.

monly 2500-4000 m thick in the central part of the Although the first two monophasic processes have
graben. Since the thickness of eroded sediments at the been proposed by various authors in the past, current
top of the blocks is quite limited, the burial of sedi- evidence, and data (Durand, 1981) strongly favour the
ments may be reconstructed with good accuracy. third, biphasic process, which has consequently been
The Frigg sector, which has been selected for the adopted for the present migration model.
application of the model, is representative of the hyd- The main driving force for expUlsion is considered to
rocarbon deposits of the Viking Graben: oil and gas be the compaction of the source rock under the sedi-
are found in Lower and Middle Jurassic sandstones ment load. In the first stage of maturation, the oil
(Statfjord and Brent) at the top of the tilted blocks, saturation within the pore system of the source rock is
and in the Paleocene sands where the accumulation of low, so that the oil or gas phase is discontinuous, and
gas is particularly large. As described by Chiarelli and migration is prevented by capillary forces. Later,
Duffaud (1980) and Carstens and Dypvik (1981), provided the petroleum potential is sufficiently high,
abnormal fluid pressures commonly occur in the the hydrocarbon phase becomes continuous and expul-
Jurassic, whereas the Paleocene is at hydrostatic sion can start. Once the hydrocarbons have reached a
pressure (Heritier et al., 1979). permeable bed, buoyancy becomes the most important
Coupled to the organic matter maturation data, driving force.
these features are the principal control parameters for This scheme is valid if the hydrocarbon phase is
the model. It could be claimed that such a determinis- unique, which is probably the case for most source
tic model has numerous uncontrolled parameters, such rocks within the oil-generating zone (Ungerer et al.,
as typical permeabilities, kinetic parameters, vis- 1981). However, it may not be the case at shallower
cosities, etc. These parameters could always be depth, where the decrease in temperature and pres-
adjusted to account for the available control paramet- sure can cause separation into a gas phase and an oil
ers. Actually, we keep exactly the same set of uncon- phase.
trolled parameters when switching from one basin to
another. Of course, the geochemical and geological
data differ. Thus a satisfactory match with field
observations is really significant. GENERATION AND MIGRATION
MODELLING

PHYSICAL AND CHEMICAL PROCESSES Based on the above considerations, the modelling of
INVOLVED migration should include the computation of:

It is generally accepted that the generation of pet- 1. oil and gas generation;
roleum results from maturation of the sedimentary 2. compaction;
organic matter, as controlled by temperature and 3. tri-phasic fluid flow and
time. There is no evidence that pressure plays an 4. phase behaviour of the hydrocarbon mixtures.
important role (Tissot and Welte, 1978, p. 183). The
These four processes are interrelated and, as a con-
modelling of maturation has been described by several sequence, the migration model should integrate all
authors: Huck and Karweil (1955), Tissot (1969),
four into a unique system. It should also be at least
Lopatin and Bostick (1973), Tissot and Espitalie two-dimensional, since lateral migration of petroleum
(1975), among others.
is generally needed for its accumulation in traps.
On the other hand, migration of oil, and especially This problem is even more complex than the
expUlsion from the source rocks, is still not well- triphasic and three-dimensional computer models that
known, although a wide range of mechanisms have have been developed in reservoir engineering. Effec-
been proposed: tively, the porous medium is not static, but changing
1. migration in aqueous solution; as a result of burial and compaction, and the process
2. migration by diffusion and takes place on a million year time-scale. This high
3. migration of hydrocarbons and water in separate complexity prevents us from accounting for the pro-
phases. cess 4, above, in the present migration model. There-
Modelling of Pelrolenlll Genesis/Jvligmholl 123

fore, the hydrocarbon mixtures are treated as mono- the Frigg area. Several wells on the cross-section or in
phasic. the vicinity serve as control points for the stratigra-
The generation of oil is computed from the kinetic phy and the sedimentology of the various units.
model published by Tissot and Espitalie (1975), with The geometry and the lithology of the sediments is
the difference that the present model makes no provi- not very well ascertained in the deepest parts (i.e.
sion for secondary degradation of hydrocarbons (dis- under the unconformity in the deepest parts of the
mutation into gas and carbonaceous residue). Gas tilted blocks). Indeed, the seismic resolution is poor
formation can also be simulated, modifying it by a and there are no control wells in synclinal locations. In
coefficient which is assumed to account for the secon- this study, the role of the Paleozoic sediments, which
dary cracking yield. have never been reached by drilling in the area, and of
Hydrocarbon expulsion is influenced by pressure in the Triassic sediments, which never show notable
the compaction process. Thus it is not possible simply properties, either as reservoirs or as source rocks,
to use porosity versus depth relationships, as this were neglected.
method does not allow for possible undercompaction. In the application of the present model, one is
Following Smith (1971), compaction is represented in limited to an 8 x 20 elements grid, and this prevents
the present model by the law of effective stress, representation of the whole cross-section of Fig. 1.
whereby effective stress is considered to be a function Consequently, a zone limited by two major faults
of porosity. The function differs from one lithology to under the unconformity, which includes the deepest
another, reflecting the fact that different rocks com- source rocks as well as possible traps in the Brent
pact in different ways. reservoirs at the top of the block, was considered.
Darcy's law, extended to polyphasic flows through The minor faults inside the block are assumed not to
the relative permeability concept, has been adopted act as impermeable barriers to fluid migration, and
for the model. According to Matheron (1967), applica- consequently the Statfjord and Brent sandstones
tion of Darcy's law to very slow monophasic flows is were represented as continuous beds in the grid (see
perfectly justified. Its extension to polyphasic flows Fig. 2).
by the relative permeability concept has, on the other The selection of boundary conditions has been made
hand, no theoretical justification and it must be con- as follows:
sidered as a pragmatic solution only. This formulation 1. at the bottom of the grid, impermeable conditions
accounts for the buoyancy of oil versus water. (no fluid flow);
Pressures in the hydrocarbon and water phases are 2. laterally, it was assumed that the faults located at
assumed to be linked by the capillarity equation. The the boundaries of the block form impermeable barriers
intrinsic permeability of the porous medium is com- under the unconformity (no fluid flow);
puted from the porosity and the average grain size of 3. above the unconformity, it is assumed that water
the rock. Source rocks or permeable beds through flow through the lateral boundaries is limited by
which fluids pass in the course of secondary migration hydraulic resistance which decreases upward. Effec-
are treated similarly, taking anisotropy into account. tively, the relatively permeable units of the Upper
Furthermore, in the source rock kerogen is converted Cretaceous appear on seismic cross-sections to be in
into oil, ultimately leading to an increased hydrocar- connection with hydrostatic zones. For instance, this
bon pressure within its pores (Du Rouchet, 1981). lateral resistance was assumed to be negligible for the
When pressure (P) increases in a porous rock, rup- Paleocene sands boundaries, which enabled represen-
ture occurs when one of the effective stress compo- tation of their hydrostatic regime.
nents becomes negative. This is evidenced either by
creep (diapirism), or by the opening of fissures. These Six classes of lithology have been used for the elab-
fissures, through which fluids may flow, close as soon oration of the grid: clay, sand, sandy silt, silty clay,
as pressure drops. To account for this process, it is marl and coal, with the lithological and geochemical
assumed here that if pressure exceeds the minimum characteristics shown in Table 1.
stress S 3, rock permeability increases in proportion to
the square of the difference (P - S:J. TABLE 1. Lithological and geochemical characteristics
All of the basic equations of the model are detailed of the six classes of lithology
in the Appendix.
The data required by the model are mainly the .vIeall oryallic :v!eall ;11;/;01
geometry of the cross-section, the age and lithology of WTiJu11 cOlllellt uil po/elltial
the stratigraphic levels, the subsidence and tempera- Formation Litholu!!y (wi %) (lIIyl!! OC)
ture history and, finally, the initial petroleum poten-
Eocene-Paleocene Sand 0 0
tial of the various sediments. The results consist of a Silty sand 0.5 210
reconstruction of compaction (porosity, pressure), of Silty clay 0.7 210
oil generation (maturation level) and of migration (oil Clay 1
saturations in source rocks and reservoirs). Senonian Marl 0.7 70
Clay 0.7 70
Turonian Marl 1.5 70
APPLICATION OF THE MODEL TO A Cenomanian Marl 1.5 70
TILTED BLOCK STRUCTURE OF THE Lower Cretaceous Clay 1.5 70
FRIGG SECTOR Upper } Kimmeridge Clay 7 310
,Jurassic Heather Clay 6.5 210
Geological and geochemical data Middle Brent Sand 0 0
The E-W cross-section considered (see Fig. 1) has Jurassic Coal 30 220
been derived from the interpretation of a seismic pro- Lower } Dunlin Clay 3.:> 210
file typical of the central part of the Viking Graben in Jurassic Statfjord Sand 0 0
124 Petroleum Geochemistry in Exploration of the Norwegian Shelf

OIL MIGRATION
a.: o 0
a:i
(/)
1000 1000
a::
L:3
>- 2000 2000
z
o
::J 3000 3000
-' K KIMMERIDGE CLAY
::i: B BRENT SANDS
10 4000 4000
CD ot
4
t
5km t

o 0
]
I

1000 1000 Ii:


UJ
0

2000 2000

3000 3000

4000 4000

5000 5000

l
6000 [I::wJ COAL 6000
c:::J CLAY
C=:J 0-5 % EB MARL
HCSATURATIONS OIl 5-10 % c=J SAND,SANDSTONE
_ HC ACCUMULATION IN SAND OR SANDSTONE rnIIIl 10-20 % E::::J SILTY CLAY
- - TOP OFTHE OIL WINDOW II!lllIIIIlIIl ',;320 % E3 SILTY SAND

Fig. 2. Representation of the computed oil saturations for two successive stages of the cross section. The first stage corres-
ponds to the beginning of oil accumulation (Maastrichtian, 65 Ma B.P.). At the present stage, a pool has formed in the permeable
levels of the Lower Jurassic, and dysmigration has occurred in the Upper Jurassic and Lower Cretaceous at the top of the
block. The top of the oil formation zone corresponds to a 10% conversion ratio of the potential.

It may be noted that a relatively high oil potential bottom part of the block, 110-100 Ma ago. At this
has been assigned to Brent coals. An elementary time, the Dunlin had still a very low hydrocarbon satu-
analysis of them does, in fact, indicate that they are ration (5%) and therefore a low relative permeability
richer in hY9rogen than the kerogens which serve to oil (see Fig. 5, below) that did not permit expUlsion.
as a reference for type III of Tissot et at. (1974). In The Brent coals, however, had a high saturation by the
addition, the oil potentials assigned to the Kim- same time, as a consequence of their good potential,
meridge Clay, and especially to the Heather Forma- high organic carbon content and low porosity. Thus
tion, are lower than for good quality, marine organic they started to expel hydrocarbons into the Brent
matter, which is suggested by their elementary sandstone, although in very small amounts since only
analysis and pyrolysis results in adjacent wells. a small part of the block was sufficiently mature. In
Paleotemperatures have been derived from burial these sandstones of high intrinsic permeability, secon-
reconstruction, assuming a geothermal gradient of dary migration is found to occur with low residual
30 °C/km and a surface temperature of 5°C. saturations (1-2%) which is not possible before signif-
icant amounts of hydrocarbons have been expelled. As
a consequence, the formation of an oil pool is found to
Computed history of oil formation and migration
start only when the expulsion is important, i.e. 75-65
Before considering the results ofthe model in a particu- Ma ago. At this time, the Heather source rocks
lar case such as the block used here, it should be noted entered the oil formation zone, but the Kimmeridge
that they can be very sensitive to the input data. Since Clay did not. However, the oil saturation of the
these data are uncertain, all the figures (timing of the Heather «20%) was not sufficient for oil expUlsion.
events, volumes accumulated, pressure, etc.) must be The saturations in the cross-section, as found 65 Ma
regarded as estimations only. ago, are represented by Fig. 2. Four classes of satu-
As computed by the model, the zone of important oil rations are represented: 0-5%, 5-10%, 10-20% and
formation (10% conversion of the oil potential) corres- 20% or more.
ponds to a depth of about 3000 m. This result is in Subsequently, the Kimmeridge Clay, Heather and
good' agreement with present maturation measure- Dunlin source rocks in their turn, supplied the reser-
ments. The first source rocks that reached the matura- voirs: for instance, the Heather was found to expel at
tion zone were the Dunlin shales and Brent coals in the 55-50 Ma in the deepest parts of the section. Because
Modelling of Petroleum Genesis/Migration 125

FLUID PRESSURE DISTRIBUTION


o o

1000 1000

2000 2000

3000 3000

4000 4000

o o
.5
:I:

1000 1000 ~
o

L&J 2000 2000


~
~
I- 3000 3000
Z
L&J
en
L&J 4000 4000
et:
a..
5000 5000

6000 6000
[I::.J COAL
c::J '0-5 MPo c::::J CLAY
[IT] 5-10MPo EEl MARL
EXCESS PRESSURE VERSUS ( DID 10-20 MP, c:::J SAND,SANDSTONE
HYDROSTATIC PRESSURE rT1TT1TTITm "'0 '0
K KIMMERIDGE CLAY Wlllll!illJ c' -30 MPo EJ SILTY CLAY
B BRENT SANDS _;;'30MPo E3 SILTY SANO

Fig. 3. Representation of the computed excess pressures versus hydrostatic, for the same stages as Fig. 2. The area of greatest
excess pressures corresponds to the Heather and Kimmeridge Clay Formations in the deepest part of the block. The Upper
Cretaceous and Tertiary sediments are quasi-hydrostatic.

of their lower organic carbon content, the Kimmeridge


Computed history of fluid pressure and compaction
and Heather source rocks are found to have a lower
expulsion efficiency than the Brent coals. However, As soon as the Upper Jurassic caprock was deposited,
their greater thickness compensates for this, and they it provided a barrier to fluid migration which resulted
are the most productive source rocks ofthe block if the in undercompaction and high pressures in the Brent
whole geological history is considered. and Statfjord sandstones, as predicted by the model
Dysmigration is also found to have occurred from 135 Ma to the present time. The pressure regime
through the Upper Jurassic and Lower Cretaceous is illustrated by Fig. 3, which shows the computed
caprocks at the top of the block from about 55 Ma to excess pressures versus hydrostatic for the successive
the present time. This was due to the high pressures stages (65 Ma and the present time). Five classes of
in the Lower Jurassic sandstones, as discussed below. excess pressure have been established: 0-5 MPa,
As a consequence, the hydrocarbon saturation 5-10 MPa, 10-20 MPa, 20-30 MPa and 30 MPa or
increases in the Lower Cretaceous units, just above more. Since water flows from areas of high excess
the oil deposit, as seen in Fig. 2. This result matches pressure towards areas of lower excess pressure, an
the oil shows that have been encountered in these idea of the directions of water flow can be obtained.
units during drilling. The order of magnitUde of the excess pressure com-
The calculated quantities of accumulated hydrocar- puted by the model at the top of the Brent sandstone
bons in the Brent, extrapolated from the two- is quite comparable with the one reported by pressure
dimensional model to the three-dimensional real tests. In addition, the hydrostatic regime of the
world, are of the same order of magnitude as those Paleocene sands is also reproduced. The pressure pro-
observed. In the case presented, the ratio of accumu- file is thus very similar to the one reported by Cars-
lated to generated hydrocarbons was found to be tens and Dypvik (1981): the excess pressure increases
0-10%, depending on the stage considered in the his- progressively from the hydrostatic Upper Cretaceous
tory of the pool. The major part of the hydrocarbons to the overpressured Lower Jurassic, but does not
generated is thus found to have remained in the source reach, or even approach, the geostatic pressure.
and carrier rocks or to have undergone dysmigra- The greatest excess pressures are found in the
tion-the latter being far from negligible. Jurassic shales, in a synclinal location. As a consequ-
ence, the fluids are considered to be expelled down-
126 Petroleum Geochemistry in Exploration o/the Norwegian She(f

wards into the Brent sandstones, which accounts for accumulation is consistent with tRe order of mag-
the contribution of the Heather and Kimmeridge nitude of the Frigg gas deposit. The fluid pressures
source rocks to the accumulation of oil in the Brent. are found to be slightly different from the previous
This pattern of fluid flow can be explained as follows: oil-phase run, but remain roughly consistent with the
the fluids from the syncline are more easily expelled field data.
laterally eastward through the Brent sandstone and
then vertically through the thin caprock at the top of Interpretation and discussion
the block, than straight across the very thick shale
Although the results above account satisfactorily for
unit in the western part of the block.
the pressure data and the volume of oil or gas accumu-
lations, it should not be forgotten that they are subject
Gas migration
to important uncertainities.
As mentioned earlier, gas migration should be treated Firstly, the model must simplify the cross-section
simultaneously with oil migration by a triphasic oil/ under study. Its bidimensional character and the
gas/water model, which is not possible with the pre- limited number of grid units prevent a full description
sent diphasic model. However, one can try to compute of the actual process at whole basin scale. In the case
the movement of gas approximately, choosing gas-like of Frigg, for instance, it is most probable that several
values for the hydrocarbon phase. The hydrocarbon tilted blocks have contributed to the filling of the gas
potential of the organic matter was lowered by 50%, deposit in the Paleocene sands. Besides, the model is
which is assumed to account for the secondary crack- only diphasic, and allows only two separate schemes
ing yield (gas formed/oil cracked). as seen above: oil migration or gas migration.
Due to its low viscosity, the gas is found to move Secondly, the input data of the model generally suf-
much more easily than oil. Additionally, the lower fer from uncertainties that may greatly modify the
density of the gas favours a rapid increase of satura- results. This is generally the case in the deep synclinal
tion, and thus a higher expUlsion efficiency. It also areas, where few wells are usually drilled. The Frigg
favours dysmigration, as a consequence of the impor- area is no exception to this rule: the geometry and the
tant parameter buoyancy. lithology of the Jurassic are speculative in the deepest
These trends are exemplified by Fig. 4, which parts of the blocks, the palaeotemperatures are
shows the gas saturations computed by the model as roughly estimated, and so on.
described above, at the present time. The major dif- However, these results may be used as a guidelme
ference from the previous run (Fig. 2) is that the dys- to reconstruct the history of hydrocarbon generation
migration is far more extensive. Although a deposit is and migration. Until the burial of the main source
still found in the Brent sandstone, the major part of rocks exceeded 4500 m, i.e. 65 Ma B.P., they did not
the gas is found to have undergone vertical migration reach the metagenesis zone and little gas was gener-
from the top of the block, building an accumulation in ated. The reconstruction provided by the standard run
the Paleocene sands from 35 Ma B.P. to the present (oil phase) of the model may be considered as rep-
time. The total volume of gas accumulated is much resentative of the real processes at these times.
larger than the previously computed volume of oil Later, increasing amounts of gas were produced.
accumulated. When extrapolated to the three- Although the oil/gas mixture is most probably mono-
dimensional real world, the volume of the Paleocene phasic in the high pressure zone, the density and vis-

GAS MIGRATION 0!
4
I
5km
I

0 0

1000 1000

LLJ 2000 2000


::IE
~ ]
~ 3000 3000 :x:
Z Ii:
LLJ w
en 0
LLJ 4000 4000
a::
0-

5000 5000

II3J COAL
6000 c::::::J CLAY 6000
EEl MARL
c=J SANO,SANOSTONE
c::::::J 0-5 % E=::J SILTY CLAY

K KIMMERIOGE CLAY
1ITIJ 5-/0 %
COMPUTEOGAS SATURATION DI!IIl /0-20 %
E3 SILTY SANO

IlIIIIW 20-/00 % _ GAS ACCUMULATIONS IN SANOS OR SANOSTONES


B BRENT SANOS
Fig. 4. Representation of the computed gas satgurations at the present time, assuming gas-like density and viscosity values
for the hydrocarbon phase. The petroleum potentials have been lowered by 50% in order to account for the secondar~ cr~cking
yield. The model finds an extensive dysmigration through the Cretaceous sediments, and a subsequent accumulatIOn In the
Palaeocene sands.
;'\;lodelling of Petroleum Genesis/Migration 127

cosity of the hydrocarbon phase are much lower than are the depth of the oil generation zone, the pressure
for a conventional oil. In the present-day cross- regime and the approximate size of hydrocarbon
section, a major part of the 80urce 8equence is in the accumulations.
gas generation zone. It is not surprising, then, that It is found by the model that successively, the Brent
the present real state is nearer to the 'gas phase' run coals and the Heather and Kimmeridge Clay Forma-
than the 'oil pha8e' run. Indeed, the Paleocene gas tions expelled oil towards the Brent reservoir, leading
field is reproduced only by the 'ga8 phase' run. to an oil accumulation at the top of the block between
It could be claimed that the oil phase run should 75 and 65 Ma D.P.
account for an oil accumulation in the Paleocene, since The present situation is best reproduced by the
a heavy oil ring is reported under the Frigg gas model if gas-like density and viscosity values are
(Heritier et ai., 1979). In the opinion of the authors, selected for the hydrocarbon phase. The model then
this oil has migrated in gaseous solution, during the predicts an extensive dysmigration through the
upward migration of the gas from the Brent accumula- Upper Jurassic and Lower Cretaceous caprocks, and
tion through the Cretaceous units. reproduces the gas accumulation in the Paleocene
In this context, Nogaret (1983) proved that an sands. The heavy oil ring beneath the gas field is
initial hydrocarbon mixture, which is monophasic in thought to be the result of the retrograde condensation
the overpressured zone, separates out into an oil phase of the oil from the upward migrating gas phase,
and a gas phase when the pressure and temperature followed by biodegradation.
decrease. Upwards, successively lighter oils con- The satisfactory reconstruction of the block con-
dense. Their amount is far from negligible: assuming sidered is, however, very sensitive to such data as the
an initial oil : gas ratio of 1 : 3 in weight, 40% of the geothermal gradient, the petroleum potential of the
initial oil is condensed between 3000 m and 2500 m, source rocks, and so on. For this reason, it is thought
and an additional 7% between 2500 m and 2000 m. that this approach can lead only to rough quantitative
These figures would be slightly different in the case of estimates of the accumulations in undrilled areas.
Frigg, but they prove that the process is quantita- However, it helps the explorationist to understand the
tively important. Further biodegradation of the con- process of oil accumulation and its key parameters. It
densed oil explains the present heavy character of the gives a consistent frame for the reconstruction of oil
Paleocene oil, as mentioned by Heritier et ai. (1979), formation and migration through geological time, and
who found that it contained virtuallv no normal enables comparison of different prospects. Thus it is a
alkanes. • very useful tool for hydrocarbon exploration.

Sensitivity of the model to geological and


geochemical data
APPENDIX: OIL FORMATION AND
A knowledge of the model sensitivity is very useful for MIGRATION MODEL *
two main purposes.
(1) It gives an estimate of the possible variations in Hydrocarbon formation
results that are caused by the uncertainty of the data. The degradation of organic matter is described in the
This provides a range of results concerning matura- Tissot and Espitalie model (1975) by a series of six
tions, pressures, accumulations, etc. If the control parallel chemical reactions obeying first-order kinetics
parameters fall within this range, the model may be and the Arrhenius law:
considered as satisfactory.
6
(2) It makes it possible to assess the relative
influence of the parameters. This can serve as a guide Q =
k
I
~ 1
(qkO - qk),
for further exploration, searching for favourable loca-
tions in the area by investigation of the key paramet-
ers for oil accumulation.
In the case of Frigg, the influence of several para- where:
meters has been investigated. For instance, the
geothermal gradient was increased and both gener- Q = quantity of hydrocarbons formed (mg/g Corg

ation and migration of oil were found to have occurred initial),


earlier in the geological history, and accumulations were qkO = initial oil potential relative to reaction
k (mg/ g C org initial),
found to be larger. Other important parameters are:
qk = residual oil potential relative to reaction
petroleum potential of the source rocks, presence of
k (mg/ g Cor. initial),
permeable beds interbedded with source rock units,
Ak = reaction coefficient (S-1),
and maximum depth of the basement. The values at
E k = activation energy (cal/mole),
control points have been found to fall within the range
R = ideal gas constant (R = 2 caljmole K),
caused by possible uncertainties, which proves the
t = time (s) and
reliability of the model. T = temperature (K).

The values of parameters qkO, Ak and E k are


CONCLUSION fixed for each type of organic matter whose degrada-

The present oil formation and migration model is still


limited by computational difficulties, since it cannot *The equations used in the model have been taken from Tissot and
represent the whole basin nor the oil/gas thermo- Espitalie (1975) for oil generation, and from Marie (1972) and
dynamic equilibrium. However, it has proved effi- Scheidegger (1960) for the fluid flow calculations. The authors were
cient in predicting the main control parameters, which also inspired by Smith (1971) for the description of compaction.
128 Petroleum Geochemistry in Exploration of the Norwegian Shelf

I
tion is to be described, on the basis of the classifica-
tion of Tissot et aZ. (1974).

Representation of fluid flows (water and oil)


The formulation of fluid flows in porous environments ~
is based on Darcy's law adapted to the diphasic case U') OIL
by the use of relative permeabilities: ~ 08
:::;
I
ii5
<t
For water: w
~ 0.6
Ve = -KK re Peg dH w
gra e; c..
lie ~
For the hydrocarbon phase, assumed to be
Ei 04
-'
w
unique: cr
0.2
V" = - KK ,·It Pltg grad. H ,,;
Ii/;
O~--~--~--~~~--~
where:
o 20 40 60 80 100
---WATER SATURATION ( % ) -
K = intrinsic permeability (m'J,
V", V h filtration velocity of water and
= Fig. 5. Form of relative permeability curves used for mod-
hydrocarbons (m/ s), elling. For low saturation values, relative permeability is
lin lih = dynamic viscosities of water and very low, but not nil.
hydrocarbons (Pa.s),
K,." K", = relative permeabilities of water and
hydrocarbons (dimensionless), The value of Pc is computed from a mean radius of
Pc> Ph = densities of water and hvdrocarbons access to pores R such that capillary pressure in the
(kg/m'I), • clays is approximately 2 MPa.
= acceleration due to gravity (=9.81 m/ S2),
Representation of compaction
He =(P/Peg)-Z= head (water),
The equation of effective stress is used to express
Hh =(P,J Phg) - Z = head (hydrocarbons), stresses between the solid matrix and fluids:
P, Ph = pressure of water and oil (Pa) and
Z = depth (m). S,,=a+P,
where:
Characteristic intrinsic permeabilities of the porous So = total stress (Pa),
environment are calculated according to porosity by a = effective stress (Pa) and
the Kozeny-Carman formula, introducing an aniso- P = pressure (Pa).
tropy coefficient:
As it was not possible to do otherwise, only the
vertical component of stress was considered. This
component is given by the following relationship,
which expresses the weight of oveJ'lying sediments:
K,,= AK,.,
where:
K,., K" = horizontal and vertical intrinsic S,' = J'- pg dz,
permeabilities (m'J, II

¢ = porosity (dimensionless), where:


S (I = specific area of rock matrix (m2/m') and z = depth of point considered (m),
A = anisotropy coefficient, less than unity p = total density of rock (kg/ m ' ) and
(dimensionless). g = acceleration due to gTavity.

Once the permeability coefficients for each element Total stress, Sq, is then termed geostatic load. To
of the grid have been calculated, those applicable to describe sediment compaction it is assumed, following
each boundary are derived by interpolation. Smith (En1), that effective stress, (J, is a function of
Relative permeabilities K,.,. and K,h aJ'e assumed to porosity:
depend on oil saturation by curves similar to those
used for modelling fluid movement in reservoirs (see
Fig. 5).
The index k signifies that the function S k differs
Capillarity equation from one lithology to another (see Fig. 6). The S k
functions were determined from porosity measure-
ments 011 well logs in series with a hydrostatic
Ph - P = P,.,
system of fluid pressure, as it is then easy to calculate
where: the effective stress corresponding to each measure-
Ph = pressure in the oil phase (Pa). ment.
P = pressure in the water phase (Pa) and The concept of compaction il'l'eyersibility is
Pc = capillary pressure (Pa). introduced by linking a, 8 k and ¢ as soon as values
Modelling of Petroleum Genesis/Migration 129

hydrostatic pressure and overburden by sediments of

I
60

r
uniform 2000 kg/m 3 density, free of vertical displace-
ments and (3) at the right and left limits of the grid,
w flow limitation by hydraulic resistance, varying with
a::
I! 4000 ~
examples treated, free of vertical displacements.
50 I! '"a::w The number of grid units in the upward direction
I! "-
increases with evolution. An extra row is added when
II '-'
rI!
~
,CLAY ~ the depth of the top of the last grid level exceeds the
'"
0
initial thickness of the following level. The number of
-040
"- It a::
0
?r: grid units is constant breadthwise. Using the
::IE
I~AfAf?L 3000

,. ~ equations set forth earlier, water and hydrocarbon


'"
'"
w I. \ w flow velocities can be calculated at each moment; these
a::
~ 30
'\ !ci:
::IE velocities are used in each iteration to produce the
w \\ ~~COAL x
> equations of mass balance of water, hydrocarbons and
H
0
a::
~Gmax
......w
"-
I
2000
"-
~
solid matrix for each element of the grid, which in turn
'\ :x: enables computation of the pressures, porosities,
20 ~ saturations and node depths through a finite-
~ w
:\ 0
difference method.
.
<.!>
:z
~\ C5

:,: \ .
~ :z
10
'\ \ 1000 ~
w
a:: REFERENCES
.;.~a " ",".......
a::
0
'-'

: .........._-...
:........ "- Carstens, H. and Dypvik, H. 1981. Abnormal formation pressure
0
'l'min I and shale porosity. AAPG Bull., 65,344-350.
0 0.1 0.2 0.3 0.4 0.5 0.6 0,7 Chiarelli, A. and Duffaud, F. 1980. Pressure origin and distribution
POROSITY-- in Jurassic of Viking basin (United Kingdom-Norway). AAPG
Bull., 64(8), 1245-1266.
Fig. 6. Effective stress versus porosity curves used in Durand, B. 1981. Present trends in organic geochemistry in
studying a North Sea block. If effective stress declines, the research on migration of hydrocarbons. Paper presented at the
curve followed (a) is different from the initial one and joins it 10th Int. Organic Geochem. Congr., Bergen, September 1981.
at the point of maximum compaction reached (I/Jrnin' umax ). Durand, B., Parratte, M. and Bertrand, P. 1983. Le potentiel en
huile des charbons: une approche geochimique. Revue IFP, 38(6),
713-721.
Du Rouchet, J. 1981. Stress fields, a key to oil migration. AAPG
of 4> stop decreasing continuously in a grid unit and Bull., 65, 74-85.
tend to increase above the 4>min value then reached, Heritier, F. E., Lossel, P. and Wathne, E. 1979. Frigg
using an empirical relationship which is easy to field-Large submarine fan trap in Lower Eocene rocks of North
calculate. This relationship expresses the fact that Sea, Viking Graben. AAPG Bull., 63(11), 1999-2020.
once values of 4> increase, they no longer follow the Huck, G. and Karweil, J. 1955. Physikalishe Problem del'
Inkohlung. Rrennst.-Chem., 36, 1-11.
initial (J = Sk curves but, rather, much steeper ones Latil (1975). Recuperation assistee. Technip, 275 pp.
which meet the (J = S k curves at a tangent at the Lopatin, N. V. and Bostick, N. H. 1973. Geology in coal
point where the 4>min value is reached (see Fig. 6). catagenesis. In: Nature a/Organic Matter in Recent and Ancient
The density for the water (p,) and that for oil (Ph) Sediments. Symp. Nauka, Moscow, pp. 79-90 (in Russian).
have been assumed to be constant. MarIe 1972. Les ecoulements polyphasiques en milieux poreux.
COUTS de production, tome IV. Technip.
The density and viscosity values used correspond to Matheron, G. 1967. E/.tjments pour'une The01'ie des Milieux poreux.
a low viscosity oil containing dissolved gas: Masson, Paris, 164 pp.
Nogaret, E. 1983. Solubilite des hydrocarbures dans les gaz com-
primes; application a la migration du petrole dans les bassins
Ph = 750 kg/m 3 (density) and sedimentaires. l'hewis, Ecole des Mines, Paris (May).
I1h = ae h/ T (dynamic viscosity according to Andrade Scheidegger. 1960. The Physics of flow through Porous Media. Uni-
formula in Latil (1975)). versity Press, Toronto.
Smith, J. K 1971. The dynamics of shale compaction and evolution
T is the temperature, a and b are selected so as to of pore fluid pressure. Math. Geol., 3(3),239-269.
Tissot, B. P. 1969. Premieres donnees sur les mecanismes et la
obtain a viscosity of 10 x 10 3 Pa s at 40°C and cinetique de la formation du petrole dans les sediments. Simula-
1.15 x 10- 3 Pa s at 160°C. tion d'un schema reactionnel sur ordinateur. Revue IPP,
Water viscosity according to temperature is calcu- XXIV(4), 470-501.
lated by the standard Bingham formula (Latil, 1975). Tissot, B. and Espitalie, J. 1975. Uevolution thermique de la
matiere organique des sediments: applications d'une simulation
mathematique. Rev. Inst. Fr. Petrol., 30, 743-777.
Algorithm and boundary conditions Tissot, B. and Welte, D. H. 1978. Petroleum Pormation and Occur-
rence. Springer-Verlag, Berlin.
The geological section under investigation from the Tissot, B., Durand, B., Espitalie, J. and Combaz, A. 1974. Influ-
beginning of subsidence up to the present time is ence of nature and diagenesis of organic matter in formation of
represented by a finite-element grid with conditions petroleum. Am. Ass. Petrol. Geol. Bull., 58(3),499-506.
at the limits as follows: (1) at the base of the grid, Ungerer, P., Behar, E. and Discamps, D. 1981. Tentative calcula-
which it is assumed represents an impermeable bed- tion of the overall volume expansion of organic matter during
hydrocarbon genesis from geochemistry data. Implications for
rock affected by subsidence, zero flow and displace- primary migration. Paper presented at the 10th Int. Organic
ment specified; (2) at the upper boundary of the grid, Geochem. Congr., Bergen, September 1981.
9 Heat flow, geothermal gradients
and maturation modelling on the
Norwegian continental shelf
using computer methods
R. H. Leadholm, * Thomas T. Y. Hot and S. K. Sahait
*Exploration Department, Conoco Norway Inc., Stavanger, Norway
tRegional Research, Conoco Exploration Research Division, Ponca City,
Oklahoma, USA

Two computer programs have been developed to model maturation parameters. The first program (referred to as
K-MOD) relates thermal conductivity to interval velocity and then computes either geothermal gradient or heat
flow, depending on which is known. The program can be run using interval velocities obtained from either well or
seismic data. The second program (referred to as M-MOD) is a maturation modelling program based on
time-temperature, theoretical kinetic equations. M-MOD requires burial history, interval geothermal gradients
and unconformity data as input. The program then computes present-day equivalent vitrinite reflectance (Re)
values for up to 21 horizons along with palaeo-temperature, palaeo-depth and palaeo-R, values for designated
time-stratigraphic horizons.
Sensitivity studies were conducted, and show that reasonable interval velocity errors do not greatly affect
M-MOD Re output. The studies also show that both programs are sensitive to errors in heat flow estimates.
Using these programs, heat flow, geothermal gradient and estimated Ro (Re) maps were constructed for the
North Sea, Haltenbanken and Troms areas. Heat flow was first mapped using K-MOD and corrected temperature
data from available wells. These data were extrapolated and mapped over each basin. Interval geothermal
gradients were then computed for selected seismic shotpoints, again using K-MOD. Finally, M-MOD was run and
present-day Re values were mapped to delimit oil and gas windows.
Approximately 100 seismic and well data points were modelled in each basin. Before running the bulk of the
data, cross-plots were constructed to ensure data reliability. A good correlation was established between
laboratory-measured Ro values and M-MOD-computed Re values. For the North Sea, the Upper Jurassic 'hot shale'
was modelled, am! R, maps show it to be oil-mature on the flanks of the Viking and Central Grabens, becoming
gas/condensate-mature towards the axes. Both the 'hot shale' and the lower .Jurassic 'Coaly Unit' were modelled in
Haltenbanken, ",ith resulting Re maps indicating a mature oil and a mature gas window, respectively. In Trams,
the top and base of the 'hot shale' were modelled, and a mature oil window was delimited on the north anr! west
flanks of the Hammerfest Basin.
Results from the three basins suggest that utilization of the K-MOD and M-MOD programs together with
interval velocity information provides a rapid means to map maturation trends over large basin-wide areas. The
programs are especially useful in basins with limited well data.

INTRODUCTION thermodynamic and geochemical data. For maturation


estimates they used borehole-based time and temper-
In recent years the search for oil and gas has become ature data to calculate maturation level following the
more risky and more selective because of political method of Lopatin (1971). This approach is adequate
instability in producing-countries, tighter exploration for a mature basin with sufficient well control. How-
budgets, and the high cost of exploring for elusive and ever, it is difficult or impossible to use it in a frontier
remote plays. Thi::; problem is augmented by basin where access to well samples is limited or non-
decreases in oil and gas prices. To cope with these existent. For this reason, it is obvious that a method
business challenges it is necessary to develop new based on surface geophysical data is needed to gener-
methods, such as computerized basin analysis, to find ate the time and temperature input required for mod-
suitable prospects quickly and precisely. elling maturation in an unexplored basin.
One of these computerized techniques was proposed This paper discusses a seismic-geochemical techni-
by Welte and Yiikler (1981), who published a model- que for estimating time/temperature data and offers a
ling method that integrated geological, hydrodynamic, brief treatment of a kinetic method developed by Con-
Petmielllll Geochemistry ill E:rp/omtioll ofthe NO)'u'egian She(f © Nonve!,rian Petroleum Society (Graham & Trotman, 1985) pp. 131-143
132 Petroleum Geochemistry in E.r;plomtion aft he Nm'wegian Shelf"

oco for modelling organic maturation. In Norway, this

./
9
technique has been applied to the regional maturation
mapping of the North Sea, Haltenbanken and Troms
areas.
,

./'
,
0
0 8

,..
0
CD
THEORETICAL BACKGROUND 7

Heat flow
E
0
iii
... .
6
Heat flow is the amount of heat that flows out of the 0
E ~Q
Earth in a unit time, and it has a rather uniform verti- ~ ....~
cal distribution in a geological column. Heat flow can- ~ x
.....,'"
5
not be measured directly. It is always calculated from '0

..
:::J
measured thermal conductivity (k) of the rocks and 'tJ
c:
ij

the geothermal gradient (dtjdZ) by using the expres- 0


() 4
*'
sion iii
heat flow (Q) = K (dTjdZ). (1) E
CD
.t: 3
~
The average heat flow for the Earth is about 1.5 Modified from Go . . .t ai, 1975
heat flow units (HFUs), where 1 HFU represents 1
J.lcal of heat that flows out of a l-cm square in 1 s. 2 ~--~--~--~----~--~--~~
4
Generally, the heat flow for an area can be directlv
obtained or can be extrapolated from the measur~­ Velocity, km/sec
ments found in published information that covers Fig. 1. Thermal conductivity versus sonic velocity. :.'.1od-
worldwide heat flow distribution (e.g. Lee, 1963; Lee ified from Goss et al. (1975).
and Uyeda, 1965; Jessop et ai., 1976; Cermak and
Ryback, 1979). The heat flow distribution is controlled
by basin history. It i" low in ba"ins associated with it is necessary to understand how thermal conductiv-
Precambrian shields «1.0 HFU) and with Paleozoic ity changes 'with depth. This is accomplished by
or older orogenies (e.g. Anadarko-Ardmore Basin understanding the effect of various geological factors
with a heat flow of <1.2 HFU). However, it is some- on the thermal conductivity of common rocks.
what higher in hasins with Mesozoic or Tertiary There are several physical parameters that affect
orogenies (e.g. West Natuna Basin, South China Sea, the thermal conductivity of rocks. The most important
with a heat flow of 1.9 HFU) and even higher in basins of these for the present purposes are porosity, density
with Tertiary volcanism (e.g. >8 HFL' in South and fluid saturation. Pressure and temperature also
Sumatra). - affect thermal conductivity to some extent. For
The heat flow in an area i" also affected by local example, there is a general decrease in conductivity
geological factors such as differential rates of with an increase in temperature. Also, the thermal
sedimentation and erosion, intrusions, hydrothermal conductivity increases with an increase in pressure.
activity, radioactivity, etc. For example, faster The factors that affect the thermal conductivity of a
sedimentation decreases the heat flow more than rock similarlv affect the velocitv of seismic waves in
slower sedimentation. Conversely, faster uplift and the rock. It has been shown pr~viously (e.g. Gross et
erosion increase the heat flow more than do 810wer al., 1975; Houbolt and Wells, 1980) that thermal con-
uplift and erosion. Volcanic intrusions can drasticallv ductivity (K) is proportional to sonic velocity (see
alter the geothermal regime (although temporarily), Fig. 1). This study reviews a computer program
thereby greatly affecting organic maturation in the (referred to as K-MOD) that is based on a similar rela-
vicinity of the intrusive body. tionship. The program solves the ba8ic heat flow equa-
Sometimes, heat flow in an area may be determined tion for either heat flow or geothermal gradient,
from the corrected bottom-hole temperatme mea- depending on which is known. The geothermal gra-
smements in a nearby well. The corrected tempera- dient data are then input together with burial history,
tures are used to calculate the geothermal gradient unconformity, intrusion and surface temperature data
from the expression into a maturation modelling program (referred to as
M-IVIOO). The M-:VIOD program is based on
(clTjclZ) = [T,.(Z) - TJjZ. (2) time-temperature theoretical Arrhenius-type kinetic
equations. The general form below refers to a cross-
where T/Z) is the corrected at depth Z, and T, is the plot:
mean annual sm-face temperatme. The thermal con-
ductivity of the geological column to depth Z i8 deter- In Rp; = 11/JII e i,,'INTi/ldt + C, (8)
mined by the procedure discussed in the following sec-
tion. The heat flow is then calculated from equation
(1). T(t) = Til + I't,
where R E is the equivalent vitrinite reflectance value,
Thermal conductivity illis the slope, E is activation energy, c is intercept of
The implication of the heat flow equation (1) is that the the regression line, R is the ideal gas constant, and T
temperature gradient changes with variations in the is subsurface temperature (in K), which is a function of
thermal conductivity of the rocks. Our main interest is heating time (t). Heating time (t) is calculated from the
in subsurface temperatures, because they control the difference in geological ages between the top and hase
formation and preservation of oil and gas. Therefore, of a stratigraphic unit.
Maturation Modelling on the Norwegian Shelf 133

This program differs from the Lopatin method EFFECT OF VELOCITY ERROR ON CALCULATED Re VALUE:
CONSTANT THICKNESS OF SEDIMENTS

(Lopatin, 1971; Waples, 1980) in two important


respects. First, it calculates maturity in terms of
equivalent vitrinite reflectance, R" rather than a (+)20
time-temperature index, TTL Second, the method is
CONTROL VALUE' R e ,"O.85%

based on integration of the time-temperature kinetic


equations rather than a graphical summation of
,:
time-temperature data. The use of these equations II: (-)40 (-)20 (+120 {+)40

treats maturation rate as a variable, rather than ffi'T


OO'~--+-----~~----~~----~----~~-­

assuming a constant doubling for every 10 DC as in the


Lopatin method. In this respect it is similar to the
work of Tissot and Espitalie (1975). The M-MOD
program is also capable of modelling uplift and ero- H20 (A)
sion.
The greatest advantage in using the K-MOD and
M-MOD programs to map maturation is that they can EFFECT OF VELOCITY ERROR ON CALCULATED REVALUE:

both be run on data obtained directly from seismic VARIABLE THICKNESS OF SEDIMENTS

sections. A seismic-geochemical analysis of remote


basins is thus possible. For example, an early version (+)20

of this program predicted the first oil discovery by CONTROL VALUE: RE ,,0.85%

the Norwegian industry north of latitude 62 N at D

Tyrihans (Ho and Sahai, 1982). Any available well


data can be tied in to confirm results. A further advan- (")40

tage of the K-MOD program is that, because it relates


interval velocity to thermal conductivity, the effects of
overpressure on geothermal gradient are also taken
into consideration.
H20 (B)

Fig. 2. Effect of velocity error on calculated R 0 values. (A)


APPLICATION Stratigraphic sequence thicknesses are held constant. This
is representative of a case using well data to pick tops and
Heat flow, geothermal gradient and maturation Re erroneous sonic logs or check shot data to calculate interval
(equivalent Ro) maps were constructed for the North velocity. (B) Variable thickness of sediments. This is rep-
Sea, Haltenbanken and Troms areas of the Norwegian resentative of a case using DIX corrected stacking vel-
continental shelf using the K-MOD and M-MOD prog- ocities. Note that velocity error essentially cancels out
thickness error.
rams. Both seismic and well data were utilized, and
results showed a good correlation with existing
geochemical data. More than 400 seismic and well data supports a similar variation of heat flow in this region
points were modelled. Before running the bulk of the (Toth et al., 1981).
data, program sensitivity to variations in interval vel- Figure 4 is based on a M-MOD run for a specific data
ocity and heat flow was first checked. point where all parameters, except heat flow, are held
constant. Varying the heat flow from 1.1 to 1.8 HFU
Program sensitivity moves the predicted R 0 for a specific source rock
interval in this example all the way from the margi-
Errors in interval velocity determination affect both nally mature window (0.6-0.8% Rr), completely
K-MOD and M-MOD output. Therefore, sensitivity through the bulk generation window (0.8-1.2% Rr)
stUdies, such as shown in Fig. 2, were conducted for a and into the late generation window (> 1.2% R 0)' This
test point where the actual Roof the source horizon is high sensitivity suggests that it is important to esti-
0.85%. Figure 2(A) shows one representative example mate and map heat flow as accurately as possible
where well data were used to pick interval thicknesses before modelling maturation.
and erroneous sonic data were used to calculate vel-
ocities. A 20% increase in interval velocity causes a Area of study, data base, modelled horizons
significant (14%) decrease in predicted Ro. However,
Fig. 2(B) illustrates that if we vary the interval vel- Figure 5 delimits the areas studied in the North Sea,
ocities and use the same erroneous velocities to calcu- Haltenbanken and Troms regions. More than 400
late interval thicknesses, a 20% velocity input error seismic and well data points were modelled and the
causes less than a 7% error in predicted R 0 values. distribution of this data is detailed in Fig. 6. The
This latter example is representative of a case where study began in the North Sea, where there are
erroneous DIX-corrected stacking velocities are used, numerous wells and where geochemical data are avail-
and suggests that in remote areas stacking velocities able to check modelling results. It then moved north
are acceptable for K-MOD input because interval vel- to the more-sparsely explored basins, where there is
ocity errors effectively cancel out thickness errors. essentially only seismic data with which to work. The
Figure 3 is a representative profile of heat flow cal- data points for modelling were drawn from a seismic
culated from well data in the Viking Graben. It shows base of 38000 km (23 612 miles).
that in the Viking Graben heat flow varies from In all three areas the computed maturation state of
1.2 HFU on the East Shetlands Platform to more than the Upper Jurassic 'hot shale' was mapped (see
1. 7 HFU on the Utsira High. These variations are Fig. 7). This is an organic-rich, radioactive marine
interpreted to be primarily due to timing of tectonic shale whose prolific source potential for both oil and
movements. Previous work reported in the literature gas is well-documented in the literature (Demaison,
134 Petroleum Geochemistry in Exploration of the Nmwegian Shelf

1.2 1.3 1.7 1.8


;~
"11

MSS EAST SHETLANDS MSS


1000 PLATFORM
1000
VIKING GRABEN UTSIRA HIGH

2000 2000

STABLE
3000 OEVONIAN
3000
PLATFORM

4000 4000

APPROXIMATE SCALE
5000 NOT KNOWN
5000

Fig. 3. Variance in heat flow across Viking Graben. Values where calculated from corrected bottom hole temperatures and
K-MOD .

.5 CONTROL VALUES

HEAT FLOW " 1.35 HFU


.6 -J---+
1.1 HFU Re = .85

.8 + __-\_ ~ 1.3 HFU

1.5 HFU
%Re OIL WINDOW

1.2 -J----~.- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 1.7 HFU

1.5

1.0 1.5 2.0


HEAT FLOW
(HFU)

Fig. 4. Sensitivity of M-MOD to heat flow error. Varying the heat flow from 1.1 to 1.8 HFU moves the calculated Ro value
completely through the oil window.

1984; Pearson et al., 1983; RJ'innevik et al., 1983). The Assumptions


name of this unit is presently being revised by the
Norwegian Lithostratigraphic Nomenclature Com- Geothermal gradient and heat flow are considered
mittee. It has been known variously as the 'hot shale' , constant through time. Goff (1983) has pointed out
the 'Kimmeridge Clay', and, most recently, as the that such an assumption introduces an optimistic error
Draupne Formation or MandaI Formation (Vollset and into maturation estimates. However, the present data
Dore, 1984). For simplification, this source horizon is cross-plots (see Fig. 8) show a good correlation bet-
refered to only as the 'hot shale' herein. In addition to ween laboratory measured Ro values and M-MOD
the 'hot shale' the Lower Jurassic 'coaly unit' was modelled values, thus demonstrating that this error, if
mapped as a gas-prone horizon in the Haltenbanken present, is not significant. The data were not examined
area (see Fig. 7). statistically.
Maturation Modelling on the Norwegian Shelf 135

TROMS
Surface temperatures were estimated using a mean
annual water bottom temperature chart published by
Evans and Coleman (1974). An attempt was made to
use extrapolated surface temperatures based on well
data; however, this method resulted in M-MOD output
that did not correlate with available geochemical data.
Recently, several authors have considered the
effects of fluid flow on the observed vertical variations
of goethermal gradients and heat flows in well-bores
HAL TENBANKEN
(Carstens and Finstad, 1981; Oxburgh, 1984). In this
paper, it has been assumed that fluids are static and

.'" '"B
average heat flow values calculated from the well data
have been used .

METHODOLOGY
To summarize the methodology (see Fig. 9): the first
\ goal was to estimate and map heat flow using either
, published values and/or well data run through
200 MIlES

K-MOD. Mter the heat flow maps were constructed,


Fig. 5. Location plot showing areas of study. interval geothermal gradients were determined for

NORTH SEA HALTENBANKEN


l"E

.
2CE
r--~---=+--.c...r'----,62'
'\'t-
~'\ .......
...
3°E

....
"
4"E
••

.
6506 • 6507· • 6508
y.. .~~ •• " • • ••• •
• -.
'

211 ~ .·34 •• ~~ ~
· .. t· •• •
•• • ••••
i.
....
I· ••

.~
• •
• •
••
.
,-----=-t--_r---~-65a N

\. .. . .•
~

\
61'
• • •• • ••
••••
••• •• •
~
3· \ . 30· 31 •• • •
•• •• • ••• • • ••
• ••
• • • • •••••
, 6406 6407
60' 6408

••••01 (B)

• 25 26
7'. 8' •
••
9 /
• SEISMIC DATA o
-------,-I,i50kms

.
f-I
• WELL DATA

.. I. •
o 3 omlles

TROMS
./
'. •• •
..
5 9' N

--Ie •
.., •
•• /•
~16~15 16
• •
17.

• I •• • ••
~
t \ •
• • - 5 8' N


22
'\'.
\ • ..7 8 •• ••
• •••
• •• • • •
• •

(A)
\
~\\
.Ie
. •
••
57° N
....
, • tfj

-:-(C~)-t----+------t71' N
1 • 2 • 3 E 4 • 20"E 22° E
0 ,50kms 50kms
• WELL • SEISMIC
I
0 i25mlles DATA DATA
• SEISMIC DATA
• WELL DATA
1------0','
30miles

Fig. 6. Distribution of modelled points: (A) North Sea, (B) Haltenbanken and (C) Troms. A total of 277 seismic data points and
132 wells were used in the study.
136 Petroleum Geochemistry in Exploration of the Norwegian She((

structural contours. In addition, burial and matura-


tion history plots, and data cross-plots to check accu-
racy against laboratory measurements and seismic
"HOT SHALE"
displays were constructed using the M-MOD data.

NORTH SEA
In the North Sea, 120 wells and 119 seismic data
points were modelled. Forty-seven of the well data
points lie in UK waters (see Fig. 6(A».
"COAL Y UNIT" K-MOD modelling of well data shows that heat flow
in the North Sea varies in the approximate range
1.2-1.8 HFU (see Fig. 10(A». Heat flow is lowest
over the stable Precambrian Norwegian Shield and
Fig. 7. Representative stratigraphic column. The 'hot over the East Shetlands Platform. The Utsira High is
shale' was modelled in the North Sea. Haltenbanken and
Troms areas. The 'coaly unit' was modelled only in Halten-
hot (greater than 1. 7 HFU), reflecting Mesozoic-Ter-
banken. tiary tectonic activity. The Viking and Central Gra-
bens have heat flow values similar to the worldwide
average of 1.4-1.5 HFU. Toth et al. (1981) showed
each data point using either well data and/or seismic similar regional variations in the UK sector of the
data and the program K-MOD once again. The geoth- Central Graben. Geothermal gradients computed from
ermal gradIents were then input together with burial both well and seismic data vary from approximately 30
history and unconformity data into the program to 45 °C/km , increasing along the same regional
M-MOD. Program output permitted the construction trends as the heat flow (see Fig. lO(B». Local
of basin-wide maturation maps for each modelled hori- anomalies are attributed to variations in lithology (salt
zon. Maps are interpolated along Base Cretaceous intrusions, etc.).

NORTH SEA HALTENBANKEN


(56 3/4 0 - 60 0 )

- - - COALY UNIT TREND


HOT SHALE
1000 1000 "0· ; TREND

'\'\
• measured C
W • measured
C .. M-MOD (seis) III .. M-MOO (wells)
W ..:

"W,\. '
III c M-MOD (wells) w 'ii' oM-MOO (seis)
..:
w
.~ If)

If) tP l ::0 -0 ••
:: ~'t:~ ...J
0" ..
0 W . ~\"c
. "'"
3000 ~~c 3000
...J III
W .~'" 'fl If)
III D i" ~

\"o~i'.f
""~;. 0::
If) W
0::
d~"
~
L _
f-
w
f-
w
G ~L

c. W
::; • 0,
::;
,\,
"
5000 5000

"
"
(A) (e)

0.1 1.0 10 0.1 1.0

MEASURED & COMPUTED R 0 VALUES MEASURED & COMPUTED Ro VALUES

NORTH SEA TROMS


(60 0 -62°)

1000 1000 • measured


~
oM-MOD
C ~ • M-MOD (wells) C

'\~co
w
\
W oM-MOD (seis)
III III
..: ..:
w
n' w
• 'i"
.
If) If)
"
::
0
...J 3000
'(,
oCt
::
0
...J 3000
.
6:, -
~~"

'.
d'
W W
III III

:~'.
0'
If) If)
0:: 0::
W ~§ w
f- f-
w
::; '~'%', W
::;
ecoC.

5000 5000

(B) (D)
I
0.1 1.0 10 0.1 1.0 10

MEASURED & COMPUTED Ro VALUES MEASURED & COMPUTED R 0 VALUES

Fig. 8. Data cross-plots of measured and computed R 0 values versus depth. (A) North Sea example shows good correlation of
laboratory values and computed values. The computed data were divided by a constant (1.08) to achieve this fit. (B) North Sea
example shows that similar results are achieved using either well or seismic data to run K-MOD and M-MOD. (C) Haltenban-
ken example shows overall good correlation. Divergence in trends below 8000 m is attributed to variance in thermal conductiv-
ity with depth. (D) The limited measured data in Troms shows a fair correlation v,rith computed values.
Maturation Modelling on the Norwegian Shelf 137

BASIN MATURATION ANALYSIS Figures 8(A) and (B) are cross-plots constructed to
demonstrate quality control of M-MOD data output.
R 0 values obtained from laboratory measurements are

"I
plotted, along with M-MOD data against depth. Fig-
ure 8(A) is based on the abundant well data south of
HUT FLOW M.f.PS
1/ latitude 60 oN. The M-MOD data were divided by a
constant (1.08) to achieve this good fit. The data were
A-. not adjusted in any of the other areas. Figure 8(B)
compares Revalues based on well data with Revalues
based on seismic data for the area between 60 oN and
62 oN and demonstrates that the data sources are
~ INTERVAL
GEOTHERMAL GRADIENTS
J compatible.
In the North Sea, Re (equivalent R;J maps were
only constructed for the top 'hot shale' horizon. Figure
11 is a sketch of the complete map and Fig. 12 is a

, detail of that map. The map shows that the 'hot shale'
is oil-mature along the flanks of the Viking and Cen-
tral Grabens becoming more thermally mature
MATURATION MAPS

towards the axes. All of the major platforms and


highs, as well as the Stord Basin and Norwegian-Dan-
ish Basin are immature at this level. These results
thus confirm previous published work by R~nnevik et
al. (1983) and Goff (1983). Cornford et al. (1983)
Fig. 9. This flowchart shows systematic approach to map- showed from geochemical observations that the onset
ping and documenting basin maturation.

COMPUTED HEAT FLOWS COMPUTED GEOTHERMAL GRADIENTS

Fig. 10. North Sea heat flow and geothermal gradient maps. (A) Heat flow varies from 1.2 to 1.8 HFU. Values are lowest on
the East Shetlands Platform, somewhat higher in the Viking Graben and are highest on the Utsira High. (B) Geothermal
gradients range from 25 to 45 °C/km. In general, values follow similar regional trends to the heat flow map.
138 Petroleum Geochemistry in Explomtion of the Norwegian Shelf

TOP "HOT SHALE" MA TURA TJON MAP


•WELL DATA
•SEIS DATA
~HOT SHALE
~ ABSENT
oI SCALE 50kms
!
I
o 30mlles

Fig. 11. North Sea maturation map for the top of the 'hot
shale' horizon. The map shows that the formation is oil-
mature on the flanks of the Viking and Central Grabens
becoming thermally more mature towards the axes.
Matumtion Modelling on the Norwegian She(f 139

Fig. 12. Detail of North Sea maturation map. Note that some of the large fields (Troll, Statfjord, etc.) lie outside the mature
windows for their respective hydrocarbon types.

of oil generation in the 'hot shale' occurs at 170-200 of ation and migration. For example, Fig. 14(A) shows
in this area. The temperatures predicted by K-MOD that at the location modelled (see Fig. 13(A», oil gen-
are compatible with these findings. eration in the 'hot shale' occurred from early Miocene
The present detailed mapping (see Fig. 12), indicates to middle Pliocene times. RMnevik et al. (1983)
that some of the large fields lie outside the peak gen- showed similar results in their study of the Viking
eration windows for their respective hydrocarbon Graben. In order to gain a better understanding of the
types. This implies that significant migration has timing of migration, it is necessary to extrapolate
occurred or an alternative source horizon exists. Troll these data and construct regional palaeo-R, maps.
is the only large field that lies completely within an
immature area. Fields in the East Shetlands Basin are
predominantly oil-prone, while some fields on the east HALTENBANKEN
flank of the Viking Graben contain large amounts of
gas. The maturation map for the 'hot shale' suggests After determining that M-MOD Re values agree with
that the East Shetlands Basin oil fields are sourced laboratory-measured R 0 values in areas with dense
from within adjacent individual fault compartments, well control (see Figs. SeA) and (B», the study was
while the fields on the east flank of the Viking Graben extended north to more frontier basins. In Halten-
drain more thermally mature shales in the graben banken 98 seismic data points and six wells were
centre. modelled (see Fig. 6(B».
Using M-MOD data it is possible to construct a The heat flow map for Haltenbanken is based on
geoseismic display to give a cross-sectional view of the limited well data and the understanding of local tec-
basin maturation. Figure 13(A) is a sketch of a com- tonic history (see Fig. 15). The heat flow varies from
plete line display and Fig. 13(B) is a detail of that 1.1 to 1. 7 HFU, with the highest values found to the
sketch. The lines are coded to show present hydrocar- west. This is opposite to the trend expected from
bon windows for the 'hot shale' horizon. The displays sedimentary thickness, and is attributed to
are constructed by using M-MOD to model depths to Mesozoic-Tertiary volcanic activity in the V0ring
nominated R, values at selected shotpoints along the Basin.
line. These depths are converted to two-way times A good correlation between M-MOD R, data and
and the windows are then plotted. Across the top of measured Ro values was established (see Fig. 8(C».
the line the present R, values for the top of the 'hot Two horizons were modelled, and the cross-plot shows
shale' are annotated. that below 3000 m, the 'Coaly Unit' matures with a
The M-MOD program can also generate a burial and less steep gradient than the 'hot shale'. This is attri-
maturation history display for each data point model- buted to variations in the thermal conductivity of the
led (see Fig. 14). These displays show both the pres- intervening sediments with depth.
ent depth to various significant Re values and the Figures 16(A) and (B) are sketch maps of the ma-
geological age at which these values were reached. turation state for both the top 'hot shale' and the top
The plots also show the burial history of the source Coaly Unit'. The Haltenbanken 'hot shale' maturation
horizon. Perhaps the most important use for this type map shows that the formation is oil-mature over
of display is to study the timing of hydrocarbon gener- nearly half of the mapped area. On the Tr0ndelag Plat-
140 Petroleum Geochemistry in Exploration of the Norwegian Shelf

SEISMIC LINE SHOWING OIL WINDOW

NW SE
33/6113417 1
34/8

.72

2-WAV TIME .82 1.71


(sec)

----l

4.0

[SASE CRETACEousl

(A)
o 20 KMS
~I-------------rl~!
o 10 MILES

NW DET AIL OF SEISMIC DISPLAY


SE
TWO
WAY
TIME RE 1.19
(seconds) I

1.0

2.0 COMPUTED % Ro (RE)

.6

.8 --

3.0
1.2 -

(8)
4.0 o
o
-
2 3 kms

Fig. 13. Seismic display showing cross-sectional view of maturation in the North Sea. Location ofline is annotated on Fig. 12.
(A) Sketch of entire line; (B) detail.
Maturation Modelling on the Norwegian Shelf 141

NORTH SEA TROMS


(60 0 .62°)

1000
1000

RE 2000
2000
0.60%

0.80%
3000

1.2% 1.20'\.

(e)
4000

HALTENBANKEN
BURIAL & MATURATION HISTORY

1000

2000

3000
0.80%

: 1. 1 O'*'

I
4000
Fig. 14. Burial and maturation histories plotted for: (A)
North Sea, (B) Haltenbanken and (C) Troms. The plots
(B) demonstrate that sedimentation rate is a controlling factor
5000 t-----::c...."'.c;:;ss"',C,--,--;;;CR;;:;ET;-;:AC;:;::EO;;;-U;;CS-'-;P;;:;AL"Eo;;O;:]C"'OL'"'G"""'M'"'O-.;:1 in oil generation timing.
AGE

form the 'hot shale' is immature (Re < 0.5), but mov- The Lower Jurassic 'Coaly Unit' maturation map
ing out into the Trl'1ndelag and VI'1ring Basins the for- delimits a large region of peak gas generation poten-
mation rapidly becomes thermal gas-mature to over- tial. This area almost directly underlies the 'hot shale'
mature (R e > 1.5%). This is attributed to both depth peak oil generation window. As a result, additional
of burial and the high heat flows associated with high GOR oil discoveries, similar to the results of
Mesozoic-Tertiary volcanics in these basins. wells drilled in the Tyrihans Field, are anticipated in
Haltenbanken.
Figure 14(B) is an example of a burial and matura-
tion history modelled for the location indicated on the
sketch maps (Fig. 16). It shows that at this position
the 'hot shale' reached peak oil generation in the
Pliocene. This suggests that for this area even the
most recent structures are prospective. Once again,
regional palaeo-R e maps are needed before general-
ized assumptions are made.

TROMS
Further north in the Troms areas, 80 seismic data
points and six wells were modelled (see Fig. 6(C».
The limited laboratory Ro data available show a fair
correlation with M-MOD Re output (see Fig. 8(D». In
Troms, maturation maps were generated for both the
top and the base of the 'hot shale' (see Fig. 17), which
is oil-mature, particularly along the western and
northern margins of the Hammerfest Basin. Although
no commercial oil has yet been discovered in Troms,
o 50kms several wells in the basin have encountered oil shows,
-------r"
o
f-I

i 3 oml1es and there is still hope for finding commercial oil in as


yet unlicensed acreage.
Fig. 15. Heat flow map for Haltenbanken based on well Figure 14 compares burial and maturation history
data and local tectonic history. plots from areas where the 'hot shale' lies at similar
142 Petroleum Geochemistry in Exploration of the Norwegian Shelf

%Re

D<.5
0·5-.6
ITl .6-.8
II .8-1.2 WI\?~~WI
11 1.2- 1.5

o %fie
Ea <: 1.0
[] .8-1.0

EI 1.0-1.2
WII:~gwI1I1.2-2.0
llii 2.0-3.0
0>3.0
(A) (B)
o 50 KMS • well data
~I--------------~i' • seismic data
o 30 MILES

Fig. 16. Maturation maps for the Haltenbanken area. (A) Top 'hot shale' map showing mature oil window; (B) top 'Coaly Unit'
map showing mature gas window.

TOP "HOT SHALE" M BASE "HOT SHALE" MATURATION MAP

------~--~------4-----------~12"N ------+------------+-----------....... 72" N

%RE

0 -=·6
1m .6-.8 7P N
~~~~--------+-----------~71·N
• •8-1.2

II 1.2-1.5 L(:...B:...)_20" E
_ __ __ _ _ _ __ __ _

o 50 KMS • well data


~I----------------~i'
o 30 MILES • seismic data

Fig. 17. Maturation maps for the Troms area. (A) Top 'hot shale' map showing mature oil window in the northwestern and
western portions of Hammerfest Basin; (B) base 'hot shale' map showing that a greater area lies within the oil window at this
level.
Matumtion Modelling on the Norwegian Sheif 143

depths in Troms, Haltenbanken and the North Sea. A and interpret the masses of data that were used in this
comparison of the shape of burial curves in these three stUdy.
areas suggests that sedimentation history is a major
controlling factor in timing of oil generation and mi-
gration from the 'hot shale'. As is shown in the figures, REFERENCES
time of onset of peak oil generation and migration is
earliest (Paleocene) in the Troms, latest (Pliocene) in Bjor~y, M. and Hall, P. B. 1983. A rich Middle Triassic source rock
Haltenbanken, and intermediate (early Miocene) in in the Barent8 Sea area. 15th A. Offshore Technol. Conf, Hous-
the North Sea. ton, 379-392.
Carstens, H. and Finstad, K. G. 1981. Geothermal gradients of the
Northern North Sea Ba8in, 59-62°N. In: Illing, L. V. and Hob-
son, G. D. (eds.), Petmlcum Geology of the Continental She!f of
CONCLUSIONS North-West EUl'Upe.
pp. 152-l6l.
Institute of Petroleum, London,

Cermack, V. and Rybach, L. (eds.) 1979. Ten-estrial Hea.t Flow in


Sensitivity studies have shown that reasonable inter- Europe. Springer-Verlag, New York, 328 pp.
val velocity errors will not greatly affect M-MOD Re Cornford, C., Morrow, J. A., Turrington, A., Miles, J. A. and
output. It has also been shown that heat flow maps Brooks, J. 1983. Some geological controls on oil composition in the
UK North Sea. In: Brooks, J. (eel.), Petroleuill Geochemistry a.nd
should be constructed before maturation is estimated. E,Tplomtion or Europe. Blackwell Scientific, Oxford,
Computer-estimated Re values generated for the pp. 175-194.
North Sea, Haltenbanken and Troms areas, show Demaison, G. 1984. The generation basin concept. In: Demaison, G.
good correlation with laboratory-measured R 0 values. and Mun~s, R. J. (eels.), Petroleum Geochemistry and Basin
In the North Sea, the 'hot shale' is at peak oil genera- Evaluatiun. AAPG IV!emoir, 35, 1-14.
Evan8, T. R. and Coleman, K. C. 1974 . .'Iorth Sea geothermal
tion on the flanks of the Viking and Central Grabens, gradients. Nature, 247,28-30.
becoming gas/condensate-mature towards the axes. Goff, J. C. 1983. Hydrocarbon generation and migration from
Major platform areas as well as the smaller flanking Jurassic source rocks in the East Shetlands Basin and Viking
basins are all undermature. Maturation was mapped Graben of the Northern North Sea. Geol. Soc. Lond. 140,
445-474.
on two horizons in the Haltenbanken area. The 'hot G08S, R., Coumb8, J. and Timur, A. 1975. Prediction of thermal
shale' is oil-mature over nearly half of the basin, and conductivity in rocks from other physical parameters and from
the 'coaly unit' is gas-mature over approximately the standard geophysical well logs. SPWLA 16th A. Log,qing Symp.,
same area. Additional high GOR discoveries similar to pp. 1-21.
the Tyrihans Field are expected. In the Hammerfest Ho, Thomas, T. Y. and Sahai, S. K. 1982. Estimation of organic
maturity from seismic and heat flow data (abstract). AAPG Bull.,
Basin the 'hot shale' is oil-mature along the western 66(5), 581-582.
and northern margins, and becomes rapidly over- Houholt, .J. .J. H. C. and Wells, P. R. A. 1980. Estimation of heat
mature as the formation plunges into the Troms/) flow in oil wells based on a relation between heat conductivity and
Basin. sound velocity. Geol. Mijnbouw, 59(3), 215-214.
Jessop, A. M., Hobart, M. A. anel Sclater, J. G. 1976. The Wurld
The Re maps demonstrate that utilization of K-MOD Heat Flow Data Collection-197:!. Earth PhY8ics Branch, Geo-
and M-MOD programs along with interval velocity thermal Service of Canaela, Geothermal Series 5, 125 pp.
information provides a rapid means of mapping Lee, W. H. K. 1963. Heat flow data analysis. Rev. Geophys., 1,
regional maturation trends over large basin areas. 449-479.
Cross-plots have established that the results are rela- Lee, W. H. K. and Uyeda, S. 1965. Review of heat flow data: in
terrestrial heat flow. Am. Geophys. Union Monogr. 8,87-190.
tively accurate. The method is especially useful in Lopatin, .'I. V. 1971. Temperature and geologic time as factors in
basins with limited well data. Even in basins with coalification. Akad. Nank. SSSR. Izv. Ser. Geol., 3, 95-106 (in
abundant well data this seismic-geochemical tech- Russian).
nique can be useful considering the uncertainties Oxburgh, E. R. 1984. Thermal gradients in 8edimentary basins
(abstract). AAPG Bull., 68(6), 797.
associated with direct vitrinite reflectance measure- Pearson, M. J .. Watkins, D., Pittion, J. L., Caston D. and Small J.
ment (Toxopeus, 1983). S. 1983. Aspects of burial diagenesis, organic maturation and
Additional information (such as migration and timing) palaeothermal history of an area in the South Viking Graben,
can be gained in a basin by using the burial and ther- North Sea. In Brooks, J. (eel.), Petroleum Geochemistry and
mal histories to reconstruct the maturity and struc- E.rp/oration (!( EU1·upe. Blackwell Scientific, Oxford,
pp. 161-173.
ture at a specific geological time and by extrapolat- Rl'mnevik, C. H., Eggen S. and Vollset J. 19SB. Exploration of the
ing these data to construct regional palaeo-R e maps. Korwegian Shelf. NPD-Collhibution no . .5. Also in: Brooks, J.
The M-MOD and K-MOD programs used in this (ed.), Petroleum Geochemistry and E.tplumtion of Ew-ope.
study are in a constant state of modification. Future Blackwell Scientific. Oxforel, pp. 71-94.
Tissot, B., and Espitalie, J. 1975. L'evolntion thermique de la
versions will eventually allow for the incorporation in matier organique des sediments: application8 d'une simulation
more-complex models such factors as vertical mathematique. Revne IFP, 30, 943-999.
anomalies in heat flow, basin fluid flux, decompaction, Toth, D. J., Lerche, I., Petroy, D. E., Meyer, R. J. and Kendall, C.
temporal variations in heat flow and/or geothermal G. St C. 1981. Vitrinite reflectance and the derivation of heat flow
gradients, 3-dimensional modelling, etc. Such a sys- changes \\~th time. In: Advances in Organic Geochemistry. Wiley,
New York, pp. 588-596.
tem is never complete but, hopefully, this discussion Toxopeus, J. M. A. 1983. Selection criteria for the use of vitrinite
has demonstrated that a relatively simple, reflectance as a maturity tool. In: Brooks, J. (eel.), Petm/cum
2-dimensional modelling method is now of practical Geochemistry alld Explomtioll or Europe. Blackwell Scientific,
use. Oxford, pp. 295-307.
Vollset, J. and Dore, A. G. (eds.) 1984. A revised Triassic and
Jurassic lithostratigraphy for the Norwegian North Sea. N P D
Bull., 3.
ACKNOWLEDGEMENTS Waples, D. W. 1980. Time and temperature in petroleum forma-
tion-application of Lopatin's method to petroleum exploration.
AAPG Rull., 64(6),916-926.
The authors wish to thank the entire Conoco Norway Welte, D. H. and Yiikler, M. A. 1981. Petroleum origin ann accumu-
staff, and in particular Emin Kabadayi, Peter Glagola lation in hasin evolution-a quantitative model. AAPG Rull.,
and John Adams for their efforts in helping to compile 65(8), 1387-1396.
10 Modelling burial history,
temperature and maturation
Peter K. Jensen, * Lise Holm t and Erik Thomsent
*Ris0 National Laboratory, Department of Energy Technology, DK-4000,
Roskilde, Denmark
tThe Geological Survey of Denmark, Thoravej 31, DK-2400, Copenhagen
NV, Denmark

A model that simulates burial history, temperature and maturation is described. The subsidence of a surface layer,
including compaction, is determined as a function of time by an analytical expression. Compaction of underlying
layers and erosion are also considered. Formation temperatures are calculated analytically, with conductivity
being a function of porosity. Maturity modelling is based on an empirical expression relating vitrinite reflectance,
temperature and time.
The model is tested against vitrinite reflectance profiles from the Danish part of the Central Graben and the
Danish Sub-basin. Past heat flow conditions are also evaluated, and a petroleum geological interpretation is given.

INTRODUCTION The formation temperatures are calculated at the


base of each of the layers. The conductivity is deter-
The mathematical model which is presented in this mined from parameters characterizing the rock type
paper can be used to simulate maturity conditions of and the degree of compaction.
organic matter in sediments and has been developed A simple expression relating vitrinite reflectance
to support the evaluation of hydrocarbon potential in (maturity), time and temperature has been chosen
the Danish area. since geochemical data are currently sparse for the
Alteration of organic matter is dependent on forma- Danish area.
tion temperature and time (Tissot and Welte, 1978). A fast-working computer program has been
Therefore the temperature must be modelled through developed which determines the burial history, the
time. The formation temperatures are determined formation temperature8 and the maturity. The
from burial depth, heat conductivities, surface amount of input data may be limited.
temperatures and heat flow. The use of the model is demonstrated bv vitrinite
The mathematical formulation given here for the reflectance Rimulations for a well in the D~nish Cen-
burial history and for the temperature calculations is tral Graben and a well in the Danish Sub-basin (see
based on analytical expressions, in contrast to other Fig. 1). The simulations are then compared with
similar models that are numerical procedures. A brief actual measurements and a petroleum geological
presentation of the mathematical model is described interpretation is given for the two wells.
below.
It is assumed that the depths of the formations are
known at a given time. At a later time the depth to the
base of the surface layer iR calculated from an analyti- THE MODEL
cal expression. This new depth corresponds to the new
depth of the top of the underlying formation. The new Burial history
depth of the base of thiR layer is then calculated from The burial history may be sketched ideally as in
another analytical expression. The next to deepest Fig. 2. The sequences are divided into a number of
layer is handled similarly. units, where each unit is characterized by a certain
During uplift the compaction of the sediments is porosity-depth relationship:
considered to be unchanged; that is, the compaction of
¢(z) = ¢oe- az , (1)
a layer is equal to the compaction stage at the max-
imum burial depth. where ¢ is the porosity, z is the depth, ¢o is the Rur-
Petmielllil Geochemistry ill Kl'pioration qf'the Nmwegian She(( © Norwegian Petroleum Society (Graham & Trotman. 1985) pp_ 145-152.
146 Petroleum Geochemistry in Explomtion of the Norwegian Shelf

Fig. 1. Location of the wells Aars-1a and U-l, and the main structural elements of the study area. The shaded areas are
structural highs.

face porosity and a is a compaction coefficient. The


values of cPo and a may be different for each unit.
Geological time (see Fig. 2) is divided into time
spans each of which is characterized by a constant
sedimentation rate of the rock mass, or by a constant
erosion rate.
The development of the mathematical formulation
of the burial is treated in two stages. First the
accumulation of sediments in a surface layer is consi-
dered, then the compaction of underlying layers is
handled. Modelling of erosion is treated separately. TIME (t)
During the development of the surface layer, a con- Fig. 2. Sketch of the physical model for simulating burial
stant mass of particles or grains is added to the sur- history.
face. The grains are buried and move towards greater
depths with a certain velocity. In the following, an
expression will be deduced from which the particle (or Combination of equations (1)-(3) results in a first-
grain) velocity can be found as a function of time. order linear differential equation:
Knowing the velocity, the time necessary for a parti-
av(z) cPo ae -az
cle to reach a given depth can be determined. Thus,
the depth of the base of the surface layer can also be ---az + 1- cPoe az v(z) = 0 (4)
determined as a function of time.
In the following, a co-ordinate system is used which A solution to this equation has been found following
is fixed relative to the surface of the sediments. Bronstein and Semendjajew (1963, p. 378):
The flow of mass during the sedimentation of the (5)
surface is a physical analogue to flow of a compressible
fluid. Therefore the equation of continuity known from where Vo is the constant particle velocity at the
hydrodynamics is applicable: surface.
a(pv) = _ ap With equation (5), an important result is obtained.
For a given depth, z, the velocity of the base of the
(2)
Oz at surface layer can be calculated.
where v is the particle velocity, t is the time and pis It is seen from the equation that the velocity will
the skeletal density defined by the equation decrease with depth, because a is a positive number.
The time necessary for the surface layer to reach
P = Pm(1 - cP), (3)
depth z can be calculated from .
where Pm is the matrix density.
Equation (2) shows that the change of mass per unit
volume, fixed in space, equals the flow of mass into t = I:V(~') dz', (6)
and out of the volume. The unknown velocity, v, is to and the following equation is obtained:
be found.
The pore fluid is assumed to escape towards the
surface. t= 1 [cPo ]
z+-(e- a'-1). (7)
vo(1 - cPo) a
The right-hand side of equation (2) is zero, since the
density and thus also the porosity is constant with Unfortunately, the depth z cannot be separated on one
time for a given depth (according to equation (1». The side of the equation, and z must be found by an itera-
boundary condition to equation (2) is the mass rate of tive procedure. It is seen from equation (7) that the
rock matrix which is added to the surface. The initial depth to the base of the surface layer can be calculated
condition is the density distribution. as a function of time without subdividing this layer.
Modelling Burial History, Tempemture and 111aturation 147

The parameters v o, 1>0 and a must be known. The where Kr is the pore fluid conductivity, y = 1 - 1>,
parameters 1>0 and a have already been discussed. Vo is K s is the solid matrix conductivity, p is the quartz
calculated when a dating, T b , is known for the base of content and C is a constant.
the surface layer using equation (7): The integral in equation (12) can be evaluated
analytically, and equation (12) becomes
Vo = (1 _ 11>o)T b [z + ~o (e- aZ
- 1)], (8) 1 1
T(zz) = T(zl) + W aA (KF/A) + 1
where z is the thickness of the layer.
A + Y22) + KF/A
With respect to the compaction of layers underlying
the surface layer, a useful expression was derived by [ -In(1- Y2) + Hn ( Kr 1
arctg ( KF/A
Y2 )
+
Selater and Christie (1980). They used the mass con-
servation of the rock matrix and an exponential
decrease of porosity with depth. Only the result will In(1 - Yj) - !In(~ + Yi) - K;/A arctg (K;;A)] ,(14)
be given here.
z 1 and Zz are the initial depths to, respectively, the
top and the bottom of a unit (see Fig. 2). During bur- where A = Ks + pC - K p , Y2 = 1 - 1>0 e- a'2 and
ial, the top of the layer is moved to the depth Z3' The Yl = 1 - Yo e- aZJ •
new depth, z 4, of the base of the layer can be calcu-
lated from Vitrinite reflectance
An empirical relation evaluated by Middleton and
Falvey (1983) relating the vitrinite reflectance R 0 to
temperature T and time t has been used here:
Again, the equation is solved by an iterative
procedure.
The burial history of several layers can now be cal-
(R o>" = b I: exp [cT(t')]dt', (15)

culated. The base of the surface layer is calculated where a', band c are constants.
through time by use of equation (7) and the bases of By empirical studies, Middleton and Falvey found
the deeper layers are calculated using equation (9). that a' = 5.635, b = 2.7 x 1O- 6/Ma and c = 0.068j"C.
Compaction is considered to be an irreversible pro- The value of c is determined from a parameter of
cess. During uplift, each of the unit thicknesses is organic alteration introduced by Royden et al. (1980)
unchanged (see Fig. 2), but the porosity-depth rela- which doubles for each 10.2°C increase in temper-
tion (equation (1» for a formation is changed during ature.
uplift. That is, if the porosity-depth relation for a unit In the present study the constants a' and b were
is identical with equation (1), then the porosity-depth modified to satisfy the measured R 0 values. The
relation after uplift, h, will be modification procedure will be discussed below. It
1>h(Z) = 1> oe- a (, '''), (10) was found that a' = 13.63, b = 6.52 x 1O-9/Ma and
or c = 0.068j"C.
The integral in equation (15) is solved numerically.
1>h(Z) = 1> oe- ah e -az. (11)
In practice the porosity function 1>,,(z) may be The simulator
determined from log data, and equation (10) is used to Having outlined the equations, we now turn to the
determine the original porosity function 1>(z) before construction of the computer code.
uplift. An idealized picture of a burial history is shown in
It can be seen from equation (11) that the compac- Fig. 2. Examples of input data are given in Tables
tion coefficient a is unchanged during uplift, but the 1-4. These examples are discussed in a following sec-
surface porosity tPoe-ah is changed. tion. A flow diagram for the computer code is given in
It should be noted that the surface velocity Vo in Fig. 3.
equation (8) for a particular unit must be determined The computer code first reads the necessary input
for parameters corresponding to the situation where data for the burial history calculations (see Fig. 3).
the unit is at the surface, before uplift occurred. These are first the time span divisions, which are

Formation temperatures TABLE 1. Time-span diversions for the burial history


For a given sedimentary unit the temperature T at at the well Aars-la
the base at depth Zz is calculated according to the
expressIOn Tillie-span Present depth Erosioll
( xu!'! !fern's) (llih. KB)

t
No. (III)
'2 1
(12) 1 0-1 128
T(zz) = T(zj) +W K(z) dz, 2 1-2 220
3 2-60
4 60-7R 815
where T(z I) is the temperature at the top at depth 5 73-83 1450
z 11 W is the heat flow and K (z) is the conductivity 6 83-98 1792
of the unit. K is a function of the porosity (Robertson, 7 98-138 2194
1979): 8 1:38-150 2374
9 150-188 2500
10 188-210 3205
(13)
148 Petroleum Geochemistry in Exploration of the N orlvegian She(f

periods of constant sedimentation or erosion rate. TABLE 2. Time-span divisions for the burial history
Then the computer program reads the preserved for- at the well U-I
mation depths and the erosion depths. The surface
porosities and the compaction coefficients (¢o and a in Tillie-span Present depth Erosion
equation (1» are read for preserved as well as eroded No. (x10 6 years) (mb. surface) (III)
formations. On the basis of these data, the burial his- 1 0-2 560
tory is determined in a later subroutine. 2 2-11 1413
The data necessary for the temperature calculations 3 11-60 2153
are the thermal rock constants, the heat flux and the 4 60-75 2425
surface temperature. The thermal constants are read 5 75-144 no depn. or eros.
6 144-156 2853
for each formation. The surface temperature, as well 7 156-178 3297
as the assumed heat flux, is read as a function of time. 8 178-188 800
The formation temperatures are calculated on the 9 188-210
basis of the burial history. 10 210-213 3344

TABLE 3. Rock properties for the well Aars-Ia

Thermal COli slants


Time- Present Compaction
span smjace cuefficient Kr Ks P C
no. pOTOsity (xl0- J jm) (W m- J K-J) (W In -1 K- 1) (ratio) (xl0-'!
W 1/[-1 K-1)

1 0.62 0.594 0.60 8.1 0.65 39


2
3 0.62 0.594 0.60 8.1 0.20 125
4 0.55 0.681 0.60 4.3 0.05 0
5 0.50 0.739 0.60 3.8 0.05 0
6 0.42 0.446 0.60 3.3 0.05 0
7 0.32 0.345 0.60 8.1 0.45 -139
8 0.30 0.436 0.60 8.1 0.70 -81.8
9 0.45 0.324 0.60 8.1 0.90 -59.0
10 0.50 0.572 0.60 8.1 0.35 -188.0

KF = Fluid conductivity, Ks = Solid matrix conductivity, p = Quatz content, C = constant.

TABLE 4. Rock properties for the well U-I

Thermal constallts

Tillle- p,.esent Compaction KF Ks P C


span su,.jace coefficient (W 111- 1 K-J) (W m- I K- 1) (mlio) (xlO- 1

110. porosity ( xl0-· i jm) W 111- 1 K-1)

0.32 0.405 0.60 1.52 0.90 44.0


2 0.60 0.505 0.60 1.52 0.40 83.4
3 U.50 0.1~5 0.60 1.52 0.30 105.U
4 0.60 0.854 U.60 3.0 0.05 0.0
5
6 0.60 0.366 0.60 1.52 0.B5 13.3
7 0.60 0.274 0.60 1.52 0.35 62.0
8
9 0.60 0.367 0.60 1.52 0.85 62.0
10 0.50 0.:367 0.60 1.52 0.35 62.0

The vitrinite reflectance parameters a', band c ( in face layer is calculated using equation (7). The depths
equation (15» are read for calculation of maturity. of all underlying layers are determined from equation
When the input data are read, the burial history is (9).
computed backwards in time, beginning with present When the burial history is determined, the temper-
formation depth. New depths are calculated for each ature history can be calculated. This is done by appli-
time step ilt. For every time step it is checked cation of equation (14). The temperature is calculated
whether the time is within an erosion period or not. at the base of each of the formations, beginning with
For erosion periods, the mean erosion rate for the the surface layer and going downwards. The calcula-
particular time-span is first calculated. The new tiom; are backwards in time beginning with present
depths are then determined linearly with time. As temperatures.
mentioned in connection with equations (10) and (11), Equation (15) is now used to determine the vitrinite
the porosity-depth relationship of the formations is reflectance as a function of time at the base of each
changed during the uplift. For times within a layer. These calculations are done forwards in time
sedimentation period, the surface sedimentation vel- ending at the present time, where the simulated val-
ocity is first calculated using equation (8) (constant for ues of vitrinite reflectance can be compared with
the particular time-span). Then the depth to the sur- measured values.
Modelling Burial History, Temperature and Maturation 149

U-1 Aors -10

VI

"c
~ Q.
.,"u
0

" "e
"i!"
1000m (!)
III
'II I
c:
.,c: .x

i
U
;;
z
U
"
U ~
...J

2000m

r------~time • time + 6t

c:
u "
..
! .~ ~
-'" ~~ 7~
JOOOm
..,,,
calculate, Calrulate:
surfaoe sedimentation velocity, erosion rate,
new depth of surface layer, new fornation depths,
new depths of urderlying layers new porosity depths re- D sandstone E03 limestone B volcanics

B siltstone EI marlstone
E3 claystone § dolostone DGU 1984

os Fig. 4. Lithology, chrono- and lithostratigraphy of the wells


Aars-la and U-l. Va + R Fs = Valhall + R0dby Forma-
tions, FF = Fjerritslev Formation, HF = Haldager For-
mation.

green and Thomsen, 1982), is rather deeply buried


here, about 3000 m. In addition, core material is avail-
able from this well. The Aars area is not influenced by
halokinesis. The Mid-Cimmerian event which strongly
affected the whole North Sea region did not cause
erosion in this area. The only period with significant
erosion is assumed to be the Quaternary, in relation to
glaciation during which a few hundreds of metres of
Fig. 3. Flow-diagram of the simulator. Tertiary sediments are supposed to have been
removed.
The U-l well was chosen because it is representa-
SIMULATION EXAMPLES tive of the stratigraphic development of the Mesozoic
and Cenozoic in the Danish part of the Central Gra-
Regional setting ben. Furthermore, the stratigraphy is well-known.
The two wells investigated are situated in two differ- The Mid-Cimmerian event is here supposed to have
ent basins. The Aars-la well was drilled in the Danish caused erosion of 500 m (compacted value) ofthe Fjer-
ritslev Formation. The U-l well is mildly influenced
Sub-basin and the U-l well in the Central Graben (see
Fig. 1). The two basins are of different origin, the by a deep-seated salt pillow which evolved during the
Late Jurassic, giving rise to a slightly condensed
Danish Sub-basin being an intracratonic basin, and the
Central Graben part of a rift system. Consequently, Upper Jurassic sequence.
their subsidence histories are different (see Figs. 6 The lithologies, chrono- and lithostratigraphy of the
and 7, below). The main subsidence period of the Dan- two wells are shown in Fig. 4.
ish Sub-basin was the Triassic while the Central Gra-
ben suffered strong subsidence during the Late Input and simulation results
Jurassic and the Cenozoic, especially the Neogene and The time-spans corresponding to the drilled sequences
the Quaternary (Holm, 1983). The thermal histories of the wells U-1 and Aars-la are listed in Tables 1 and
are different, too, with higher heat flow during the 2. In the tables are shown a numbering of the time-
Jurassic in the Central Graben due to active rifting. spans, the present depth for each time-span, and the
Both basins are strongly influenced by halokinesis. erosion. The value of the erosion is the total depth of
The Aars-la well was chosen because it has been the erosion in the particular time-span. Tables 3 and 4
thoroughly investigated and because it is situated cen- show the estimated rock parameters. The present sur-
trally in the Sub-basin. The Fjerritslev Formation, face porosity is found by plotting depth versus
the most promising source rock in this area (Lind- logarithmic porosity for each lithological unit. The
150 Petroleum Geochemistry in Exploration of the Norwegian She{f

'E AARS-1A
- Temperature 200 :;: VITRINITE REFLECTANCE TREND
~ 50
- --- Heat flaw
u: E o CUTTINGS • CORES 9 SIMULATED
5
§ 40 __ L_~~1
100 ~
DEPTH
m
LlTHO-
STRA- REFLECTANCE % Ro
~20~~~~~~___ b.K.B. TIGRAPHY 0.3 0.4 0.5 0.6 0.7 OB 0.9
Vl
LL
--~~~rLs~~;-_- POST CHALK
~ ~
O!J! 1
200 100
MILL. YEARS
500
Fig. 5. Assumed surface temperatures and heat flow i

conditions.

MILL. YEARS
1000 CHALK GROUP
'\
200 100 o
1500

R00BY
FORMATION 0, •
2000

:\
VEOSTED
FORMATION
,
BREAM
FORMATION
2500
~
FORMAT10N
0

'~~
FJERRITSLEV
FORMATION
3000
o I
GASSUM
FORMATION 0 \ I
I 1

~::t-~: generotio~'~
PREMA-
IMMATURE MATURE
!TURE!

• oil --...J 3
Ro=0.6 Fig. 8. Measured vitrinite reflectance values together with
simulated results for the well Aars-Ia.
Fig. 6. Burial history corrected for compaction for the well
Aars-Ia, and time of onset of oil generation. Recent heat flow values are estimated to be about
60 and 74 mW/m2 for the wells Aars-la and U-l,
respectively. The heat flow conditions at the well U-l
MILL. YEARS may be somewhat influenced by an underlying salt
200 100 o pillow. Calculations made by Jensen (1981) indicate
that the temperature close to the bottom of the well
may increased up to 15 °C due to the presence of the
salt.
The present heat flow could be estimated for both
wells as already mentioned. In the following it is
explained how the past heat flow is estimated and how
E
.x the parameters a' and b (in equation (15» are adjusted
I to the available data. First, it can be seen that a dou-
Ii:
UJ ble logarithmic plot of measured vitrinite reflectance
2° and the calculated time-temperature integral should
fall on a straight line, if the temperature and therefore
the heat flow are estimated correctly. From the
U-1
, Onset of oil generation
straight-line plot, the constants a' and b can be deter-
3 mined. It is assumed here that the thermal paramet-
Ro =0.6
ers, the surface temperature and the burial history
are correctly determined. A plot was made for the two
Fig. 7. Burial history corrected for compaction for the well wells, aSfmming constant heat flow conditions equal to
U-I, and time of onset of oil generation.
the recent heat flow. All of the data for the Aars-la
well and low R 0 values for the U-1 well fell on a
porosities were determined from wireline log mea- straight line. It could then be concluded that the
surements. From the same plots the compaction coeffi- assumed constant heat flow condition for the Aars-la
cients were determined. Knowledge about surface well was reasonable. The heat flow in the case of the
porosities for the lithologies involved was also incor- U-l well had to be increased to about 95 mW/m2 for
porated. ages greater than 150 Ma. This is in accordance with
The thermal constants were initially determined modelling of heat flow conditions during the subsi-
from figures given by Robertson (1979) and from con- dence of the Central North Sea Basin (Sclater and
ductivity measurements in the case of the well Aars- Christie, 1980). Figure 5 shows the assumed heat flow
la (Balling, 1981). Later, the C values in equation (13) models.
were adjusted to obtain calculated recent tempera- Surface temperatures in the North Sea area have
tures equal to the measured temperatures (Balling, been evaluated by Buchardt (1978) for the Tertiary.
pers. comm.; Jensen, 1983). Cretaceous and Jurassic temperatures were assumed
Modelling Burial History, Temperature and Maturation 151

U-1 Data quality


VITRINITE REFLECTANCE TREND
° CUTTINGS • SIDEWALL - CORE 0 SIMULATED
The burial histories of the two wells investigated have
DEPTH lITHO-
been compiled by means of biostratigraphical analyses
m STRA- REFLECTANCE % Ro carried out by the Geological Survey of Denmark.
b.K.B. TIGRAPHY 0.3 0.4 0.5 0.6 0.7 0.8 0.9
These studies have been supported by lithostrati-
0 graphical correlations to wells in the close vicinity. The

0 reliability of the data presented is considered good,


I
0

500 I except for the Tertiary and the Quaternary of the


~ I Aars-la well. Here, the magnitude of original deposi-
c tion and subsequent erosion must be regarded as

\~I,
1000 uncertain.
POST CHALK
GROUP
The porosity data are based on petrophysical log
readings, mainly the acoustic sonic, supplemented by
a few sample measurements. The reliability of the
1500 I
absolute porosity values are considered to be only fair,
but the porosity trends are considered to be reliable.
i
I The temperature measurements in the well Aars-la
t- 2000 are probably within I_2°C of true formation temper-
I I
I ature. The present heat flow is estimated to lie bet-

1\
CHALK GROUP
VALHALL+R0DB
ween 50 and 80 mWjm 2. This range reflects the uncer-
2500 FORMATI~S tainty of the conductivity estimate. The temperatures
J-4 UNIT obtained in the well U-l were measured a short time

3000
J -3 UNIT
\0 after cessation of the mud circulation. The tempera-
tures have been corrected, but the accuracy is
unknown (Jensen, 1983). Also, the conductivities and
J -2 UNIT

~~::li~OL~V I
~ 0;0 the heat flow estimates are uncertain since no heat
conductivity measurements were performed.
IMMATURE Ir~~~~- MATURE From the Aars-la well, core and cuttings samples
were available for analysis of vitrinite reflectance.
Fig. 9. Measured vitrinite reflectance values together with Lignite used as a drilling mud additive restricted the
simulated results for the well U-I. use of several of the cuttings samples. However, the
rather high number of cores and the presence of
after personal communication with Buchardt. The sur- organic-rich intervals and coal in samples from the
face temperatures used in the calculations are shown Haldager and the Gassum Formations, permitted
in Fig. 5. proper control of the rank conditions. Thus, the data
The burial histories, corrected for compaction, are quality is rated high.
seen in Figs. 6 and 7 for the two wells. A simulated From the U-l well, cuttings samples were available
vitrinite reflectance equal to 0.6% R o, which is for analysis, supplemented by a single sidewall-core
regarded as a general threshold value for the onset of from the J-2 unit. According to biostratigraphical
oil generation (Thomsen et al., 1983), is marked on the studies, caving is significant in the Mesozoic intervals.
figures. Simulated and measured vitrinite reflectance This is clearly revealed in several of the samples
values are plotted as a function of depth for the two investigated in this study. Fair amounts of vitrinite
wells in Figs. 8 and 9. It is seen that the simulated have been recorded in samples from the CEN-4 and
values are close to the measured values. the CEN -5 units and from the J-2 unit. The latter data
comprise a few cuttings samples with coal fragments,
the rank of which is confirmed by measurements from
the one side-wall core from this unit. Thus, in general,
DISCUSSION the vitrinite reflectance data quality is rated fair in
this well.
The model
Results of analysis
The model is one-dimensional, and therefore its use is
Aars-la
restricted. It should be applied with caution in areas
where large lateral tectonic movements have occurred The Aars-la well is of special interest, since it repres-
and where the horizontal heat flux component is con- ents the most mature Mesozoic deposits that so far
siderable. On the other hand, a one-dimensional model have been drilled in the Danish Sub-basin. The reflec-
implies simpler equations, and analytical solutions can °
tance trend reaches about 0.73% R at a depth of about
be found. Furthermore, the computing time and the 3350 m (see Fig. 8). This is significantly higher than
amount of input data are strongly reduced. the reflectance range of 0.36-0.58% Ro for Upper
Several of the so-called 'back stripping' models have Triassic-Jurassic-Lower Cretaceous deposits at
been discussed in the literature (Perrier and Auiblier, depth intervals between 200 and 2800 m, that has
1974; Du Rouchet, 1980; Sclater and Christie, 1980; previously been reported from this area (Thomsen et
Falvey and Deighton, 1982; Middleton and Falvey, al., 1983).
1983). In these models, lithologically uniform sequ- Modelling shows that the onset of oil maturity in the
ences must be divided into a number of subunits, and Aars-la well took place during the late Cretaceous
the calculation accuracy increases with this number. (see Fig. 6). This timing is regarded as favourable
The model which is presented here works with a relative to structural development for the area in
restricted number of units. question.
152 Petroleum Geochemistry in Exploration (~f the Nurwegian Shelf

However, none of the wells drilled through these ACKNOWLEDGEMENTS


deposits, including the Aars-Ia well, has yielded com-
mercial petroleum. The lack of oil is attributed to the Thanks are given to N. Balling Laboratory of
unfavourable terrestrial type of organic matter in the Geophysics, Aarhus who collected and analysed temp-
Gassum Formation and the basal part of the Fjerrits- erature data for the well Aars-Ia.
lev Formation. However, a more favourable type of
organic matter has been encountered in the overlying
F-III and F-IV Members of the Fjerritslev Forma- REFERENCES
tion, which, under proper burial conditions, could rep-
resent potential oil source intervals. Balling, N. 1981. Principper for temperaturberegning. In:
Michelsen, O. (ed.), Kortlmgning of Potentielle Geuterilliske
U-I Reservuir i Danlliark. Geological Survey of Denmark, Series B,
1\0. 5, pp. 13-19.
Bronstein, 1. N. and Semendjajew, K. A. 1963. Taschenbuch der
Oil and gas were encountered in the U-I well. The J-4 MateTinatik. Teubner, Leibzig, 584 pp.
unit is regarded as the principal source unit by com- Buchardt, B. 1978. Oxygen isotope palaeotemperatures from the
parison with other parts of the Danish Central Graben Tertiary period in the North Sea areas. Nature, 275, 121-123.
Falvey, D. A. and Deighton, 1. 1982. Recent advances in burial and
area (Thomsen et al., 1983). The reflectance trend thermal geohistory analysis. APEA Journal, 22,65-81.
presented in this study (see Fig. 9) is an extension of a Holm, L. 1983. Subsidence history of the Jurassic sequence in the
previous preliminary study published by Michelsen Danish Central Graben. Damn. Geol. Unders., 1982, :19-51.
(1982). The higher number of samples investigated ,Jensen, P. K. 1981. Geotermiske monelberegninger ved endelige
permits a more thorough interpretation of the matur- elementers metone og analytiske metoder specielt med henblick
pa saltstrukturer. Thesis, Institute of Geophysics, Copenhagen,
ity conditions and reveals a higher maturity gradient 270 pp.
than previously recorded. The J-4 unit is within the Jensen, P. K. 1983. Formation temperatures in the Danish Central
zone of early oil generation, while the J-2 unit and the Graben. Danlll. Geol. Unders., 1982, 91-lO6.
Fjerritslev Formation are within the zone of max- Lindgreen, H. and Thomsen, E. 1982. Investigation on the Source
Rock Potential of the Dallish Onshore Area. Danish Geological
imum oil generation according to the Ro values sug- Survey, Copenhagen, 91 pp.
gested by Tissot and Welte (1978). Michelsen, O. (ed.) 1982. Geol0.9Y of the Danish Celltral Graben.
The modelling shows that the onset of the oil gener- Geological Survey of Denmark, Series B, No.8, pp. 73-86.
ation took place during the Palaeogene (see Fig. 7). Middleton, M. F. and Falvey, D. A. 1983. Maturation modelling in
This timing is regarded as favourable relative to Otway Basin, Australia. AAPG Buli., 67,271-279.
Perrier, R. and Auiblier, J. 1974. Thickness changes in sedimentary
structural development, and is in accordance with layers during compaction history, methods for quantitative
general hydrocarbon migration concepts in the study evaluation. AAPG Bull., 58,507-520.
area. Robertson, E. C. 1979. Thenllai Conductivity of Rocks. Open File
Report, 79-356, U.S. Geological Survey, 31 pp.
Rouchet, J. Du 1980. Le programme diagen, deux procedures pour
apprecier l'evolution chemique de la matiere organique. Bull.
Cent. Rech. E;rplomholl-P1'Odnctioll EllA.quitaille, 4(2),
813-831.
Royden, L., Sclater, J. E. and Von Hertsen, R. P. 1980. Continen-
tal margin subsidence and heat flow: important parameters in
CONCLUSIONS formation of petroleum hydrocarbons. AAPG Bull., 64, 173-187.
Sclater, J. G. and Christie, P. A. F. 1980. Continental stretching: an
explanation of the Post- Mid-Cretaceous subsirlence of the Cen-
It has been shown that the model can simulate satis- tral North Sea Basin. J. Geophys. Res., 85(B7) 3711-3739.
factorily the measured vitrinite reflectance trends for Thomsen, E., Lindgreen, H. and Wrang, P. 1983. Investigation on
two wells in the Danish area with quite different bur- the source rock potential of Denmark. Ill: Kaaschieter, J. P. H.
ial histories and heat flow conditions. It is, further- and Reijers, T. J. A. (eek), PetruleulII Geuloyy of the Suuihe<"i-
em North Sea and the Adjacellt Ollshote Areas. Geol. !VI(jIlOOIlU',
more, possible to estimate past heat flow conditions 62, 221-239.
for the two wells in accordance with the observed tec- Tissot, B. and Welte, H. 1978. Petrolellm Formation and Occnl'-
tonic development. renee. Springer-Verlag, Berlin, 538 pp.
11 Basin analysis and maturation
modelling, onshore Denmark;
a case study from the Danish
First Round
Steinar S¢rensen
Rogalandsforskning/Rogaland Regional College, P.O. Box 2540,
Ullandhaug, 4001 Stavanger, Norway

In the Danish Sub-basin exploration for hydrocarbons has continued for decades without any commercial success.
At present the petroleum industry is again showing an increased interest in Danish areas. The axis of the Danish
Sub-basin runs parallel to the Tornquist-Teissure Line and perpendicular to the Oslo-Rhine Graben System.
Mesozoic tectonics of the area were dominated by subsidence and halokinetic movements, in contrast to the
Cenozoic, when inversion and partial erosion took place in the eastern parts of the basin. In this study the Lower
Jurassic Fjerritslev Formation is considered to be the principal source rock, but so far only immature sections have
been encountered in the subsurface, possibly because most of the features drilled are salt structures where the
source rocks have never reached a depth greater than 1500 m. In modelling thermal history it is assumed that
source rock maturation is related exponentially to temperature and linearly to time. Temperature data for wells
were obtained from open-file information in the Danish Geological Survey. An estimate of palaeo-heat flux was
based on an interpretation of the tectonic and depositional history of the Danish Sub-basin, and is taken into
account in the model. By applying heat flux changes through time and adjusting geothermal gradients for
differences in lithological heat conductivity, porosity variations and age, the geothermal history of the Danish
Sub-basin has been modelled more precisely. It is concluded that locally-mature Lower Jurassic source rocks may
occur in the Danish Sub-basin, and have potential to generate enough oil to fill some untested structures already
identified.

INTRODUCTION 59 000 km of seismic data, although locally the data


base was more limited. In the Danish Sub-basin, some
In the Danish Sub-basin exploration for hydrocarbons 20 wells were available. Pyrolysis data and other
has continued for decades without any commercial geochemical information were available from nine
success. From an exploration viewpoint, interest in wells. A fairly detailed stratigraphy (see Fig. 1) ofthe
this area has been limited for two reasons: the pet- Danish Sub-basin is available in numerous publica-
roleum potential of the basin has been regarded as tions, mainly from the Danish Geological Survey
only fair, and the whole area was held solely by the (DGU) and, hence, formation ages are established
Dansk Undergrunds Consortium. However, blocks with some confidence.
relinquished by the DUC in 1982 were recently offered Accurate temperature information is often a prob-
for exploration to the entire industry. In April 1984, lem in thermal history analysis, but fortunately a
22 companies were awarded licences in the Danish large amount of useful data is also available from the
First Round, and at present there are more applicants DGU. This information is part of a Danish project on
for blocks in Denmark than in Norway, among them geothermal energy and geothermal reservoirs in
some of the major oil companies. onshore Denmark (Michelsen et al., 1980, 1981). In
The present study was part of a technical evaluation these publications heat flux, as well as lithological heat
prepared for a Danish First Round Application, and conductivity values, are listed. Variations in heat con-
consequently techniques and concepts, rather than ductivity are considered as a function of lithology,
specific prospects will be described. geological age (for units with basically the same lithol-
ogy) and porosity.
Based on this information, geothermal gradients
DATA BASE from the Danish Sub-basin published by Madsen
(1975, 1977) were modified and used as guidelines in
Data available for this study included 43 released this study. Heat flux values from several places in
wells from the greater Danish area, as well as onshore Denmark are published by Michelsen et al.
Petmlenm Geochelilish7! in E.rp/oration qf the Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 153-160.
154 Petroleum Geochemi.stry in Exploration of'the Norwegian Shelf'

CHRONOSTRATIGRAPHY DANISH SUBBASIN (2) WNW-ESE (110-120°) trending faults, termed


MAASTR ICHTIAN
the Hercynian Trend, outlining parts of the Fenpo-
scandian Border Zone and at an angle of about 25 to
CAMPANIAN
the Dinaric Trend.
U
SANTONIAN CHALK
(3) NNE-SSW (010-030°) trending faults, termed
CONIACIAN the Rhenian Trend and illustrated by the en echelon
TURONIAN axis of the Oslo Rift and Rhine Graben System. Many
'"
:>
0 DARK MUDSTONE MARL AND SAND of the salt walls in the German Basin follow this Trend
w CENOMANIAN
"
<f (Pratsch, 1979).
~
I- ALBIAN
w
a::
APTIAN
(4) NNW-SSE (160-170°) trending faults, termed
()
the Kattegat Trend, These faults seem to be arranged
BARREMIAN
L in a disorderly en echelon pattern in the Danish Sub-
HAUTERIVIAN VEDSTED Fm basin.
VALANGINIAN (5) ENE-WSW (040-070°) trending faults, termed
BERRIAS I RYAZAN the Skagerak Trend. This trend is represented by only
a few faults in the Skagerak area.
TITHON IVOLGIAN FREDRIKSHAVN Mb

U KIMMERIDGIAN
BREAM Fm
V B0RGLUMMb
The five structural trends listed above are also rec-
OXFORDIAN
ognized from magnetic, gravimetric, Landsat and
FLYVBJERG Mb
bathymetric data (Frost, 1977; Dikkers, 1977).
CALLOVIAN HALDAGER Fm
"iii M BATHONIAN
~
a:: SAND Mb
~ BAJOCIAN
-------- ------- DEPOSITIONAL HISTORY
AALENIAN Mb F-4

TOARCIAN Mb F - 3 In the Danish Sub-basin, Paleozoic rocks are reported


L
PLIENSBACHIAN FJERRITSLEV Fm Mb F - 2 from on Iv two wells, R9Inde-l (Rasmussen et al., 1971)
HETTANGIAN
Mb F -, and N9I~ling-l (Rasmussen et al., 1973). Hence the
- GASSUM Fm interpretation of the Paleozoic section is mainly based
RHAETIAN VINDINGFm
a::
w
~
- - - - - - ---=SKAGER~ on seismic data. From seismic lines a relatively thick
"-
:>
NORIAN ODDESUND Fm
sedimentary sequence is recognized between a base-
w
"
iii '" CARNIAN ment reflector and the base of the Mesozoic. The
'"<f
cr: IMUSCH.
LADINIAN Paleozoic section includes thick Zechstein salt
MARL, LIMESTONE, EVAPOITES,
I-
KALK ANISIAN REDBEDS deposits, which have influenced the Mesozoic and
RED SANDSTONE AND MUDSTONE
Cenozoic evolution of the basin.
BUNTER SKYTHIAN
During the Triassic, significant subsidence took
place in Denmark, particularly in th~ Danish .Sub-
Fig. 1. Stratigraphic subdivision of the Danish Sub-basin, basin. Sediments consist mainly of flUVIal to contmen-
tal deposits (Bertelsen, 1978; Michelsen et al., 1981),
with a thickness exceeding 1500 m locally (Mors-1
(1980, 1981) and are calculated fro~ the ,average well). During Keuper time the sedimentary load on
geothermal gradients and the average lIthologIcal heat
the underlying Zechstein salts reached more than
conductivities (Balling et al., 1981), 1000 m, and halokinesis was initiated. Halokinetic
movements have been occurring spasmodically since
Late Triassic times, creating extensive salt structures
STRUCTURAL OUTLINE and rim synclines in the Danish Sub-basin.
Basin subsidence continued strongly throughout the
The Danish Sub-basin is part of the greater Nor- Early Jurassic (Michelsen, 1978) declining through the
wegian-Danish Basin, and in this paper it refers to the Middle and Late Jurassic and into the Cretaceous.
north Jylland area and its adjacent waters (see Subsidence was renewed during the Tertiary, with
Fig. 2). The axis of the Danish Sub-basin runs. parallel the main depocentre in the Danish Sub-basin being
to the Tornquist-Teissure Line and perpendIcular to located just north of the Holmsland Block (Nielsen et
the trend of the Oslo Rift and Rhine Graben System, al., 1984). The northeastern part of Jylland was
Denmark, and the North Sea generally, lie close to affected by Laramide inversion during the Paleogene
the Y-shaped junction of three major plates, the Fen- and bv th~ Fennoscandian uplift during the Neogene
noscandian Shield, the North American-Greenland (Ziegl~r, 1982).
Plate and the South European-African Block (Zwart In ~3Ummary, the depositional history in the
and Dornsiepen, 1978; Stn'1mberg, 1981). Mesozoic was dominated by basin subsidence and
Five orogenic phases have affected Denmark and halokinetic movements, as opposed to the Cenozoic
have played an important role in the evolution of the when inversion and partial erosion occurred (Nielsen
area. These phases are (a) Grenvillian, (b) C~domian, et al., 1984; Pegram, 1984).
(c) Caledonian, (d) Variscan and (e) the Alpme (Sur-
lyk, 1980). From the analysis ofte,ctonic elements, five
major trends have been recog~llze~ (see belo:v), of LOWER JURASSIC FJERRITSLEV
which the first three are of maJor Importance m the FORMATION
structural and sedimentary evolution of the Danish
Sub-basin. In this studv the Lower Jurassic Fjerritslev Forma-
(1) NW-SE (130-150°) trending faults, termed the tion is regarded as the source rock, as it has a favour-
Dinaric Trend (Lindstr9lm, 1967). These trends are able kerogen composition and is buried deeply enough
dominant in the Fennoscandian Border Zone. to be mature (Lindgren and Thomsen, 1982; Thomsen
Basin Analysis and Maturation Modelling 155

LEGEND ~ .
,MAJOR FAULTS 'FAULTS STRUCTURAL HIGH '"
'"", SALT STRUCTURES
"
Fig. 2. Tectonic framework of Denmark.

et al., 1983). It is divided into four Members Toarcian source rocks of the Paris Basin (Tis sot et al.,
(Michelsen, 1978), as follows. 1971).
Member F-J (Hettangian-Pliensbachian) was depo-
The Lower Jurassic Fjerritslev Formation in the
Danish Sub-basin reaches thicknesses in excess of
sited under marine conditions and consists of deposits
grading from shallow to deep environments. The total 1000 m locally (see Fig. 3). All members of the forma-
tion have an organic carbon content which classifies
organic carbon (TOC) content is of the order of 1-2%,
them as potential source rocks. Substantial parts of
and the organic matter consists of a mixture of humic
the Formation have a favourable kerogen composition
and amorphous material.
Member F-II (Pliensbachian) was deposited in a for oil generation. The principal question which this
well-oxygenated shallow marine environment. TOe study addresses is the degree of thermal maturation.
ranges from 1% to 3%, and the organic matter consists In order to determine thermal maturity, a tempera-
ture, depth and time relationship has been established
also of a humic/amorphous mixture.
Member F-JII (Toarcian) was deposited in a deep
in order to make semi-quantitative estimates, based
on an assumed thermal history.
water marine environment at a time when sea level
changes and basin isolation resulted in a restricted
bottom water circulation. This unit is dominated by
amorphous material with a TOC of the order of 1-6%. METHODS
Member F-JV (Aalenian) was deposited locally in a
lagoonal environment with strongly reducing condi- In mapping the geothermal history of a basin, often
only two temperature data points are known, i.e. the
tions. The organic content consists of a mixture of
amorphous and humic material and TOe ranges from surface temperature (A) and a subsurface tempera-
1% to 3%. ture (B) (in this model the temperature at the present
depth of the source rock). A straight line joining point
Of the four members of the Fjerritslev Formation, A to point B provides the average geothermal gra-
Member III (of Toarcian age) is regarded as the best dient.
source rock because of its high TOC and favourable Alternatively, the temperature profile may be
kerogen composition. Member III is comparable with divided into several segments with different gradients
the Drake Shale of the Northern North Sea (Cooper (Gretener, 1981), which represent reality more closely
and Barnard, 1984), to the Posidonia Shale in the since heat conductivity varies with lithology. To esti-
Netherlands (Bodenhausen and Ott, 1981) and the mate the geothermal gradient within a lithological unit
156 Petroleum Geochemistry in Exploration of the Norwegian Shelf

"<... <;.. -- ~.
-
......

--- .....
I

" '"
LEGEND EROSIONAL AND CONTOUR INTERVAL
\ DEPOSITIONAL LIMIT \ (200m)

1
STRUCTURAL HIGH 111
' - FAULTS 111 SALT STRUCTURES
111
11

Fig. 3. Isopach map of Fjerritslev Formation.

one must know the heat flux in the area (q 0) as well as The open file information from the Danish Geological
the heat conductivity (k) of the rocks. The geothermal Survey on heat conductivity measurements as a func-
gradient within each lithological unit can be expressed tion of lithology, age and porosity, provides useful
as the quotient of the two, as follows: data for heat conductivity corrections through various
lithological sections. These measurements indicate
qo/k = gradient. (1) heat conductivity variations from as low as 1.34
The temperature calculations are derived from the W m- I K- I in parts of the Paleocene clays of the
analytical expression of Michelsen et al. (1981): Viborg-5 well to 4.20 W m- I K- 1 for the Middle-Late

T,,=To+I
n {(Z -kZ,- J [qo-IA/Zj-Z J
1
i-I
j _ l)
Jurassic sandstones of the Gassum-l well (Balling et
al., 1981). It then becomes obvious that heat conduc-
1 =1 t ) 0: 1 tivity variations have to be accounted for in subsur-
face temperature estimation.
(2) In equation (3), qo represents the heat flux through
the different units, Z and is regarded as a function of
where: time. The present modelling of heat flux is based on
Tn = temperature at the nth level, the assumption that present-day heat flux values are
To = surface temperature, probably lower than palaeo-heat flux values in the
q 0 = heat flow of the unit, area because there has been little tectonic activity in
A j = heat production within the unit, the area throughout the Neogene. In contrast, five
k i = lithological heat conductivity and earlier orogenic phases have affected the Danish area,
Zi = depth to base of unit i. and hence it is assumed that increased heat flux occur-
red during such tectonically active periods and for a
The surface temperature of the area is set at 10 DC. time after the main tectonic event (see McKenzie,
The heat production (A) is very low, with an average 1978, 1981; Richardson and Oxburgh, 1978; Royden
of 1.1 J.l.W/m 3, which is regarded as negligible for the et al., 1980; Turcotte and Ahrens, 1977).
actual temperature calculations (Balling et al., 1981). In the following example the thermal history of one
This leaves the expression of the rim synclines just north of the Grinsted Block is
qo, 1 A 2 analysed (see Fig. 2). In order to map the geothermal
T(z) = To +k - 2 T/ . (3)
history of an area of interest, the burial history also
Basin Analysis and Maturation Modelling 157


.-
TIME MILLIONS OF YEARS IBY FORMATION I

FJEA"ITllEV HAlDAOEA BREAM IIEOSTEO AGIOBY


H'ORAlh
HOD
EKO""
s~;~ -; ~a 11 ~~~
en
w
"...w "'\
~\
:Ii
~
~500

\,-
w
c

L I'~
'\ ONSET OF

I'" --
OIL GENERATION
I
Ro = 0.6
1600
-

-'.
w

1'\
=
w
~

;\

2000
~
~

2500

\
3000
', . ,.
~'...r. . . . . . . . .

3500
1/0 1/0
2~
* 1jo 110 1(0
~

TR lASSie I LOWER
JURA~~~E
UPPER I LOWER CRETACEOUS UPPER J~t EOCE;~R'I~;~CENE' MIOCENE 11
HEAT FLUX (mWni2)

r(~)i(((J~A§(ii{(/(@ir(t(tftt/t~!itt~)f?h\f?!@i!i!f!~??rf!!?!((/J!!/l :0
l!r!!fi!!!t!!!!!!!!!!!{f~Kf!!!!!!!!i!!!!.!!!!!!!!!!!!!!! ! i!i! ! ! ! ! !}! ! ti! ! t!i ti!i!i! ! ! ! ! ! ! ! ! !~tf!!!!!!!f!!!t!!/!!!!!!J!J!fl :0
Iiriii)itf!!f!!!fif]f!¢:g~tJf!!!!!!!!!!!!!.i!f!f!!!!!!!!!f!!!!!f!!]f!f!f!!~ttti}i!ttIf!!!fJf!f!f!f!f!f!f!Lrrftiriiii!!/ifi!!!iii] :8
Fig. 4. Subsidence history of the area studied.

had to be determined, and was based on an interpreta- three distinct phases-during the Early Jurassic as a
tion of relevant seismic sections (see Fig. 4). The continuation of the Triassic subsidence (Bertelsen,
well-tie in the area is good, hence the stratigraphy can 1978), during the Late Jurassic and during the Late
be reliably correlated. The section has been depth- Cretaceous-Early Tertiary.
converted by means of local velocity analysis. The In the present geothermal modelling, three differ-
lithology of the section is interpreted on the basis of ent heat flux histories were considered (see Fig. 4). A
seismic stratigraphy, and correlated with known conservative estimate of the geothermal history is
lithologies from the nearby wells. The age, lithology, represented by temperatures adjusted for lithological
porosity and heat conductivity relations have been heat conductivity, but keeping the heat flux in the
calibrated against empirical data published by the area at the present value constant with time (case I,
Danish Geological Survey (Michelsen et ai., 1980, Fig. 4). The geothermal history becomes more
1981). The burial history plot of the study area (see favourable for maturation if an increased heat flux
Fig. 4) indicates that major subsidence occurred in during the three periods of high tectonic activity is
158 Petroleum Geochemistry in Exploration of the Norwegian Shelf

VITRINITE REFLECTANCE

.2 .3 .4 .5 .6 .7 .8 .9
DEPTH
IN METERS t:; KVOLS

500 0 HYLLEBJERGET
x MORS
0 N0VlING
1000
0 RODDING

" CALIBRATED Ro
1500 ( ) POSSIBLY
UNRELIABLE

2000

2500

3000

3500

Fig. 5. Measured vitrinite reflectance versus depth for the Danish Sub-basin, and calibrated trend for the study area (see
Fig. 2).

introduced to the model, for example, from 60 to elled during the Late Paleocene, corresponding to a
70 mWjm 2 (case II) or from 60 to 80 mWjm 2 (case III, burial depth of 2750 m. The present areal distribution
Fig. 4). of the Fjerritslev Formation with R 0 > 0.60 is shown
To relate time and temperature to thermal matur- in Fig. 6. Based on the calibrated Ro versus depth
ity, the Arrhenius equation was used, which assumes relation, the R 0 level of the source rock at a present
a doubling of the reaction rate for every 10 °C rise in depth of 3200 m is estimated to be R 0 = 0.85. This is a
temperature (Tissot and Welte, 1978). Several such slightly higher maturity than is measured at a com-
models have been described by Gretener and Curtis parable depth in the Aars-1 well (Jensen et al., 1985).
(1982), Hood et al. (1975) and Waples (1980), which all This difference is probably due to the higher subsur-
assume that maturation is related to temperature face temperature in the area studied (see Fig. 3), as
exponentially and to time linearly (see also Hunt, opposed to the Aars-1 well, and to variations in geo-
1979). Following the method of Waples, TTl values thermal histories for the two areas.
were correlated with vitrinite reflectance values The modelling studies indicate that the Fjerritslev
based on in-house calibrations. Formation is currently mature for oil generation in a
It is of major importance to define the R o-level for large area, and should be capable of feeding some of
onset of oil generation in the area studied. As the structures mapped on Jylland (see Fig. 6). This is
described by Tissot (1984), the onset of oil generation confirmed by the presence of oil and gas shows in sev-
is dependent on the type of organic material, and as eral of the old exploration wells on Jylland (Gassum-l
such has a wide R o-range. Based on in-house studies of (1951) Hyllebjerget-1 (1976), Mors-1 (1967), Nj'jvling-1
the organic material present in the Lower Jurassic (1966». However, no exploration wells in the Danish
source rocks, onset of oil generation in the area is set area of the Danish Sub-basin drilled so far are
as R 0 = 0.60. The calculated R 0 versus depth relation reported to have penetrated mature source rocks.
for the area studied (using heat flux history, case II) This may be due to the fact that most of the wells (if
has been plotted, together with available measured not all) were drilled on structures which have never
vitrinite reflectance data from wells in the Danish been deeply buried.
Sub-basin (see Fig. 5). Histograms of all vitrinite
reflectance measurements were available, and the
quality of the data used is considered to be good. The CONCLUSIONS
two trends are in general agreement, and observed
discrepancies can be attributed to subtle differences in By means of thermal modelling techniques, Lower
burial and thermal history between the location mod- Jurassic source rocks are estimated to be oil-mature
elled and individual wells elsewhere in the basin onshore in Jylland. The source rocks may have been
where measurements were made. generating oil since Late Paleocene times. It seems
probable that a high percentage of the source rock
kerogen has been converted to oil.
Possible explanations for the lack of significant suc-
RESULTS cess in the area may include limited drainage into the
drilled structures, ineffective trapping mechanisms or
The timing of the onset of oil generation, as deduced lateral facies variations of the source rock. However,
from the calibrated R o-depth relation for cases I, II the author believes that hydrocarbons may be found in
and III is shown in Fig. 4. In the most likely case (II), the area, perhaps trapped below the edge of
this is estimated to have occurred at the location mod- mushroom-shaped salt structures.
Basin Analysis and Maturation Modelling 159

-.............. NORMAL FAULT

- ... - ....... NORMAL FAULT, QUESTIONABLE

_________ VALID CONTOURS

.... --. QUESTIONABLE CONTOURS

~ SALT QIAPIR

~ REFLECTION BED ABSENT OR THIN

~ LOW

C) HIGH

OGRINSlEOl

SCALE

12.5 Km 25.0 Km

CONTOUR INTERVALS 0.25 m $eC

Fig. 6. Areal distribution of mature Lower Jurassic source rocks.

ACKNOWLEDGEMENTS Mjiiller-Pedersen of AIS Norske Shell for fruitful dis-


cussions and suggestions on the manuscript.
The author would like to thank Superior Oil Company
for the permission to publish this paper. Thanks are
also due to my colleagues, K. E. Trommestad and W. REFERENCES
H. Edwards, for the seismic interpretation, to N. F.
Peterson for great help with the geochemistry and to Balling, N., Kristiansen, J. I., Breiner, N., Poulsen, K. D., Ras-
Dr O. H. J. Christie of Rogalandsforskning and P. mussen, R. and Saxov, S. 1981. Geothermal measurements and
160 Petroleum Geochemistry in Exploration of the Norwegian Shell

subsuriace temperature modelling in Denmark. GeoSkrifier Michelsen, 0., Saxov, S., Leth, J. A., Andersen, C., Balling, N.,
No. 16, Department of Geology, Aarhus University, 171 pp. Breiner, N., Holm, 1., Jensen, K., Kristiansen, J. J., Laier, T.,
Bertelsen, F. 1978. The Upper Triassic-Lower Jurassic Vinding Nygaard, E., Olsen, J. C., Poulsen, K. D., Priisholm, S., Raade,
and Gassum Formations of the Norwegian-Danish Basin. Dan- T. B., Sprensen, T. R. and Wurtz, J. 1981. Kortlregning of pot en-
marks Geol. UndeTS. Serie B, 3, 26 pp. tielle geotermiske reservoirer i Danmark. Damn. Geol. UndeTS.,
Bodenhausen, J. W. A. and Ott, W. F. 1981. Habitat ofthe Rijswijk Serie B, 5, 96 pp.
oil province, onshore :-.i'etherlands. In: Illing, L. V. and Hobson, Nielsen, O. B., SlJrensen, S., Thiede, J. and Skarb0, O. 1984. Diffe-
G. D. (eds.), PetroleulII Geology of the Continental Shelf of rential subsidence of the Cenozoic North Sea. AAPG Bull. (sub-
North-West Europe. Institute of Petroleum, London, mitted).
pp. 301-309. Pegrum, R. M. 1984. The extension of the Tornquist Zone in the
Cooper, B. S. and Barnard, P. C. 1984. Source rocks and oils of the Norwegian North Sea. NOIw. Geol. Tidsskr., 64(1), 39-68.
Central and Northern North Sea. In: Demaison, G. and Murris, Pratsch, J. C. 1979. Regional structural elements in Northwest
R. J. (eds.), PetroleulII Geochemistry aud Basin Evaluation. Germany. J. Petrol. Geol., 2(2), 159-180.
AAPG Memoir, 35, 303-314. Rasmussen, L. B., Dinesen, A., Henriksen, S. E., Bang, 1., Senes-
Dikkers, A. J. 1977. Sketch of a possible lineament pattern in tad, E., Buch, A., Christensen, O. B., Michelsen, O. and Jacob-
North-West Europe. In: Frost, R. T. C. and Dikkers, A. J. sen, F. L. 1971. Dybdeboringer R¢nde nr. 1 pa Djursland. Dan 111.
(eds.), Fault Tectonics in North-West Europe. Geol. Mijnbouw, Geol. Unders. III Rmkke, Selie no. 39, 123 pp.
56, 275-285. Rasmussen, L. B., Baartman, J. C., Henriksen, S. E., Kristof-
Frost, R. T. C. 1977. Tectonic patterns in the Danish Region (as fersen, F. N., Dinesen, A., Bang, 1., Stenestad, E., Buch, A.,
deduced from comparative analysis of magnetic, landsat, bath- Christensen, O. B., Michelsen, 0., Hansen, T. J. and Jacobsen,
ymetric and gravity lineaments). Geol. iffijnbouw, 56(4), F. L. 1973. Dybdeboringen NlJvling nr. 1 i Midtjylland. DanlYl.
351-362. Geol. Unders., III Rmkke. Selie No. ;,0, 181 pp.
Gretener, P. E. 1981. Geothermics: using temperature in hydrocar- Richardson, S. W. and Qxburgh, E. R. 1978. Heat flow, radiogenic
bon exploration. AAPG Education Conrse Note Ser., 17, 170 pp. heat production and crustal temperatures in England and Wales.
Gretener, P. E. and Curtis, C. D. 1982. Role of temperature and Geoi. Soc. Lond., 135,323-337.
time on organic metamorphism. AAPG Bull .. 66(8), 1124-1149. Royden, L., Sclater, J. G. and von Herzen, R. P. 1980. Continental
Hood, A., Gutjahr, C. C. '}1. and Heacock, R. L. 1975. Organic margin subsidence and heat flow: important parameters in forma-
metamorphism and the generation of petroleum. AAPG Bull., tion of hvdrocarbons. AAPG Bull., 64(2), 173-187.
59(6), 986-996. Str¢mberg: A. G. B. 1981. The European Caledonides and the
Hunt, J. M. 1979. Petroleum Geochemistry and Geology, W. H. Tornquist Lineament. Geol. Toren. Stockh., 103,393-410.
Freeman & Co., San Francisco, 617 pp. Surlyk, F. 1980. Geology of the European countries: Denmark.
Jensen, P. K., Holm, L. and Thomsen, E. 1985. Modelling burial Geology of the European countries. Comite National F'I'lLll(ais de
history, temperature and maturation (this volume). Geoloqie IC.N.F.G.), 26th Int. Geological Congress.
Lindgren, H. and Thomsen, E. 1982. Investigations on the source Thomsen, E., Lindgren, H. and Wrang, P. 1983. Investigations on
rock potential ofthe Danish onshore area. Dan)}!. Geol. Unders. the source rock potential of Denmark. In: Kaasschieter, J. P. H.
(June 1982) 91 pp. and Rejers, T. J. A. (eds.), Petm/eulIl Geology of the Sonthwest-
Lindstnjm, M. 1967. Om den tektoniske utvecklingen av Skfmes an NOTth Sea arid Adjacent Onshore Area8 (The Hague, 1982).
Paleozoikum. :lfedd. Dansk Geol. FOl'erling, 17, 163-165. Geol. Mijnbouw, 62,221-239.
Madsen, L. 1975. Approximate geothermal gradients in Denmark Ti~sot, B. P. 1984. Recent advances in Petroleum geochemistry
and the Danish North Sea Sector. Danlll. Geol. Undel's, 1974, applied to hydrocarbon exploration. AAPG Bul i., 68(5), 545-563.
5-16. Tissot, B. P. and Welte, D. H. 1978. Petroleum FOl'matioll alld
Madsen, L. 1977. Geotermisk Energi i Danmark-en geologisk Occul1'ence. Springer-Verlag, Berlin, 538 pp.
vurdering. Dansk Geol. FOl'en., Al'skrift for 1977, K¢benhavn, Tissot, B., Califet-Debyser, Y., Deroo, G. and Oudin, J. L. 1971.
(January 1978), 29-40. Origin and evolution of hydrocarbons in early Toarcian shales.
McKenzie, D. 1978. Some remarks on the development of sedimen- Paris Basin, France. AAPG Bull., 55(12),2177-2193.
tary basins. Earth Planet. Scie. Lett., 40, 25-32. Turcotte, D. L. and Ahren, J. L. 1977. On the thermal and subsi-
McKenzie, D. 1981. The variation of temperature ",ith time and dence history of sedimentary basins. Geophys. Res., 82(26),
hydrocarbon maturation in sedimentary basins formed bv exten- 3762-3766.
sion. Earth Planet. Sci. Lett., 55,87-98. • Waples, D. W. 1980. Time and temperature in Petroleum Forma-
.}lichelsen, O. 1978. Stratigraphy and distribution of Jurassic tion, application of Lopatin's method to petroleum exploration .
deposits of the Norwegian-Danish Basin. Dawn. Geoi. Unders., AAPG Bull., 64(6), 916-926.
Se)'ie B, 2,28 pp. Ziegler, P. A. 1982. Geological Atlas of Westerll and Ce II/ra I
Michelsen, 0., Saxov, S., Leth, J. A., Andersen, C., Balling, N., E ul'ope. Elsevier Applied Science Publishers, Barking, 130 pp.
Breiner, N., Holm, L., Jensen, K., Kristiansen, J. I., Laier, T., Zwart, H. J. and Dornsiepen, U. F. 1978. The tectonic framework
Nygaard, E., Olsen, J. C., Poulsen, K. D., Priisholm, S., Raade, of Central and Western Europe. Geol. "tljnbouw, 57, 627-654.
T. B., Silrensen, T. R. and Wurtz, J. 1980. Landsdmkkende
Geoterlflisk UndersfJqelsel'. Dell: KO)'timqnhlq og Analyse at'
Rescrvoire)·. Dansk Olie & Naturgas AjS,KlJbenhavn, lO3 pp.
12 Quantitative modelling of the
sourcing of giant oil fields in the
East Shetland Basin (Abstract)
J. C. Goff
Departrrient of Energy, Petroleum Engineering Division, London, UK

In the East Shetland Basin, Late Jurassic rifting formed tilted fault blocks and unconformity traps, sealed by
onlapping Cretaceous mudstones. Important oil-bearing reservoirs are Middle and Lower Jurassic
marginal/non-marine sheet sandstones. Mature Upper Jurassic mudstone and Middle Jurassic coal source rocks,
located in half grabens downdip of the major fields, and in the flanking western Viking Graben, generated oil in
Eocene to Miocene times.
Study of the reservoir pressures, distribution and structure defines the major Jurassic aquifers in the East
Shetland Basin. The geometrical relationship of these aquifers to the source rock kitchens defines the volumes of
source rock drained by the major oil fields.
Total organic carbon, pyrolysis and soluble extract analyses were performed on thick source rock sequences,
representative of the source rock kitchens, to determine their organic richness, oil yield and maturity. A model of oil
expulsion is developed, based on the squeezing of oil out of the mature source rocks during progressive burial
(compaction). Quantitative modelling of the volume of oil generated and expelled from the source rocks accounts
for the in-place reserves of the fields studied.

PetroleulII Geochemistl'Y in Explol'lltion of' the NOJ'wegiali Shelf' © Norwegian Petroleum Society (Graham & Trotman, 1985) p. 161.
Part III
Case Histories
13 The geological history and
geochemical evaluation of
Block 2/2, Offshore Norway
Roy H. Gabrielsen, Steinar Ulvwen, Arvid Elvsborg and Ole Fredrik
Ekern
Saga Petroleum a.s. P.b. 9. 1322 Hj9vik, Norway.

Block 2/2 is situated on the northeastern flank of the Central Graben within the 'Ula Trend', which is characterized
by the development of Upper Jurassic reservoir sandstones. The 2/2 area underwent a relatively quick subsidence
along the WNW-trending Hidra Fault Zone in late Jurassic times. The tectonic activity was influenced by
halokinesis, and the geometry along the fault zone varies significantly. Three wells have so far been drilled in block
2/2. Besides gas discoveries in Oligocene sands in wells 2/2-1 and 2/2-2, oil has also been found in Upper Jurassic
sandstones in well 2/2-1.
This study includes an organic geochemical interpretation of mainly the Jurassic section in block 2/2, and a
discussion of the Oligocene gas discoveries. Source rocks are identified in the Upper Jurassic Mandai and Farsund
Formations and in the coal-bearing Middle Jurassic Bryne Formation. Indications of migrated hydrocarbons are
evident in the Rogaland Group and older formations in all three wells. The Jurassic oil discovery correlates best
with Farsund Formation source rocks.

INTRODUCTION TECTONIC FRAMEWORK OF BLOCK 2/2


The present paper is a summary of an integrated Block 2/2 is situated on the northeastern flank of the
study by Saga Petroleum a. s. of the Jurassic sequence Central Graben and encompasses the southeastern
in Block 2/2 in the Norwegian Sector of the North part of the Hidra Fault Zone, which forms the transi-
Sea. The study resulted in a model for the structural tion between the Southern Vestland Arch (R~nnevik
and sedimentological development within Block 2/2. et al., 1975; Hamar et al., 1980) and the Central Gra-
This model, in turn, served as a base for geochemical ben (see Fig. 1).
evaluation of the block. The Hidra Fault Zone trends WNW-ESE to
Block 2/2 was a part of the Norwegian 6th round, NW-SE, and consists of several fault segments, partly
and was awarded in August 1981 to a group including arranged in an en echelon pattern. The faults are
Arco Norway Inc., Mobil Exploration, Norway, clearly influenced by halokinesis, and have to a large
Norsk Hydro A.S., Saga Petroleum a.s. and Statoil, degree been developed as growth faults separating
with Saga Petroleum as the operator. The first well rotational fault blocks. According to Skjerven et al.
(2/2-1) was drilled in 1982 in the southeastern part of (1983) a northeastward rotation of the so-called Hidra
the block, resulting in an oil discovery in an Upper Fault Blocks took place in late Jurassic times, thus
Jurassic reservoir. Subsequently, two more wells post-dating a more regional middle Jurassic uplift and
have been drilled in the block. eastward tilting of the Vestland Arch. To the south-
In addition to the present authors, several people west, the Hidra Fault Zone is bounded by the master
have been involved in the study. The structural model faults of the Central Graben. At an early stage the
is to a large extent based upon a seismic interpretation fault zone was influenced by halokinesis, when consid-
carried out in the Permian part of the sequence by erable volumes of salt flowed away from the vicinity of
Atle Edvardsen and Gyrd Sundsb~, while computer the Hidra Fault Zone in a southwesterly direction.
work was done by Knut Aanstad and Odd Alstad. The combination of halo kinesis and normal faulting
Stuart Gowland and Louis Macchi made significant associated with the Hidra Fault Zone has defined an
contributions to the tectono-sedimentological model, area not only characterized by a series of potential
and the geochemical study has benefitted from coop- hydrocarbon traps, but also by an elongate zone of
eration with Torbj~rn Throndsen. favourable reservoir sands, the 'Dla Trend' (Bailey et
Petl'Olenm Geochemistry ill EJ'p/omtioll of" the NunlJegian Shelf" © Norwegian Petroleum Society (Graham & Trotman. 1985) pp. 165-178.
166 Petroleum Geochemistry in Exploration of the Norwegian Shelf

~CONTOUR

..,.-_ QUESTIONAEILE
./ CONTOUR

o,
I
03'00'

Fig. 1. Main structural elements of the northeastern flank of the Central Graben in the vicinity of Block 2/2. Contouring in
two-way reflection time on Base Cretaceous level. Based on compilation by N. Fagerland.

al., 1981) within which the Ula Field is found in Block STRATIGRAPHY
7/12.
Roughly, three major fault segments may be iden- At present, several systems for stratigraphic nomen-
tified in the Hidra Fault Zone, for which the notations clature are in use for the Central North Sea (e.g.
A, Band C have been proposed (see Fig. 1). In detail, Deegan and Scull, 1977; Hamar et al., 1983; Hesjedal
however, these three elements again may be sub- and Hamar, 1983). However, a complete survey ofthe
divided into a series of minor fault segments. Also, a stratigraphy for the Norwegian sector south of 62 ON
series of salt pillows are found on the downthrown is in preparation by an official committee (V ollset and
(southwestern) side of the fault zone and may, Dore, 1984), and in the present paper the preliminary
together with a trend of rotated faults blocks, be said version of this proposal has been used. Vollset and
to define a discontinuous high between the major Dore (1984) discrimate between the Sleipner area,
faults of the Hidra Fault Zone and the Central Norwegian/Danish Basin and the Central Graben area
Graben. for the Jurassic part of the sequence, and a summary

BERRIASIAN
KEROGEN TYPE II

VOLGIAN OIL PRONE

.
KIMMERIDGIAN ".
z
.
>- KEROGEN TYPE II-III

OXFORDIAN
OIL/GAS PRONE

BATHONIAN
KEROGEN TYPE III

BAJOCIAN GAS PRONE

PLIENSBACHIAN

Fig. 2. General stratigraphy for the Jurassic interval in Block 2/2 with source rock units indicated.
Geochemical Interpretation of Block 2/2, Offshore Norway 167

of their scheme for the Central Graben is given in and is clearly demonstrated in the differences in the
Fig. 2. late Jurassic sedimentological development between
wells 2/2-1 and 2/2-3 (see Fig. 5).
In the following, a schematic structural and
STRUCTURAL AND SEDIMENTOLOGICAL sedimentological history for Block 2/2 is given. The
MODEL, BLOCK 2/2 names and position of the structural elements, and the
timing of activition of these features is seen in Figs. 3
In the structural evaluation of Block 2/2 all significant and 6.
faults were classified and named according to
geometry and relation to the main tectonic features in Pre-Triassic
the area (Fig. 3). By careful reconstruction on The pre-Zechstein fault pattern is dominated by
migrated, depth-converted seismic sections, the NNW-trending elements, and a considerable relief is
geometric configuration as a function of time along indicated. When trends and styles are considered
strike of the Hidra Fault zone was restored. One there is a striking difference between these faults and
example of this type of reconstruction is given in the faults mapped higher up in the sequence (see
Fig. 4. These data were compared with conventional Fig. 4). However, a pronounced hinge-line at this
decompacted, back-stripped subsidence curves from level coincides with the trend of the later Hidra Fault
wells 2/2-1 and 2/2-3. Even though the method does Zone. The fault systems of this age are associated with
not allow corrections for periods of non-deposition or the Permian sedimentary basins, and gave a surface
erosion, minimum rates of subsidence, axes of relative relief which has controlled the thickness variations of
subsidence and differences in timing of structural the evaporites.
events may be defined. The analysis, which incorpo-
rates several N-S and NE-SW trending lines, sug-
gests that the structural development of the Hidra Triassic-Middle Jurassic
Fault Zone in Block 2/2 is not synchronous. The early The fault pattern defined on Top Salt level is radically
fault activity within the Hidra Fault Zone was, furth- different from the pre-Zechstein system. The first
ermore, strongly influenced by halokinesis, and was WNW-trending fault blocks were activated in the
restricted to a few faults within the zone. This situa- Triassic, and continued to develop throughout the
tion is also reflected in the sedimentological record, Late Triassic and Jurassic. The activating agent was

STRUCTURAL ELEMENTS, BLOCK 2/2

o 1 2 3 4 SKm

Fig. 3. Some structural elements and well locations in Block 2/2. Single-lined faults are defined on 'top Ula Formation' level,
double-lined faults on Base Cretaceous level. Numbering refers to text and Fig. 7. SI-S5 indicate salt pillows.
168 Petroleum Geochemistry in Exploration of the Norwegian Shelf

? EMERGENT VESTLANO ARCH

Sandstone Member of
CENTRAL GRABEN Mandai Formation

3 ? EMERGENT VESTLA~~_~
,

Growth A,ea Oepocenlre

? EMERGENT VESTLAND ARCH


5
.::i:2/2-2
~'. _~_-:-5"blll!oral$heel S8nd;~'

r=:~=o~ 3~!~ 6~ = rI.=+. '·~~"ii.l:~~~~


Faull C 5

Faull-acll¥alild submarme
gram-liow sand bodlllS
(Oil bllaring)
6

~ ~.---.J V l ~____ CENTR~ _____

8
~ ..

[ .
\!
"
SVMIIOL CODE

".~ ;'"''''''
IROIllOSUII5H)ENCE

STABLE POINT/CENTRE Of UPI-IEAV.u !


11 OEPHI IN MErEIlS

I • '"

14 REf lASE CAET


----,

UNCOIH

I (). IUIO.l,O UPI-IEAV ... L II


INflU JUR"SSICI I
E AREA Of EROSION

! .... Flor ... TIONAL ",~rs I 1 REF

• REF
I ~ REGIONiU ROTHION

Fig. 4. Geometric reconstruction ('back-stripped basin') of a seismic N-S line crossing Fault C4 (see Fig. 3), and correspond-
ing structural/sedimentological models for the Upper Jurassic interval.
Geochemical Interpretation of Block 2/2, Offshore Norway 169

3317",

~800m

KIMMERIOGIAN

'I
km
3800m

50

100
Vertical
Scale

MIDDLE-LATE
? JURASSIC
3900m

f:22l Sand

B Mud

~ Concretionary Dolomite

Fig. 5. Schematic lithofacies correlation within the reservoir interval of Wells 2/2-1 and 2/2-3. Note lateral relation between
the Farsund Formation Source rocks and the Ula Formation sandstones in the two wells. Sedimentological correlation by Louis
Macchi.

possibly a combination of a NNE-SSW-directed ten- slowed down, and a slight south-southwesterly re-
sional stress system and halokinesis. The first activity gional rotation took place.
was restricted to a few segments of the fault zone, and The most important structural elements were the
in Block 2/2 the earliest faults to develop were C4 and salt pillows in the northern part of Block 2/2 (S3 and
its antithetic counterpart C4Al (see Figs. 3 and 6). 84, see Fig. 3) and the associated faults. These
Also, several minor faults (the P-faults of Figs. 3 and defined the structurally highest points, and may even
6) mirror the early halokinetic processes. Obviously, a have coincided with the basin margin in this time
structural low was generated between the axis of the interval. Also, some activity was occurring along
salt pillows SI and S2 and the C-fault trend, and this Fault System C, and again C4 was the most active
had already started to act as a sedimentary trap in the element. Very slight flexuring accompanied by rela-
Triassic. The Bryne Formation was deposited on a tively high rates of deposition is indicated in the Cl
swampy coastal plain during the middle Jurassic. and C2 segments, whereas faulting probably took
place along the C4 and C5 segments (see Figs. 3 and
6). Even some of the faults of the G and F systems
Callovian-Kimmeridgian
were active.
In this period the area was tectonically relatively The deposition at this time took place in a shallow
quiet. A eustatic sea level rise in early Kimmeridgian zone with a smooth relief (Fig. 4). Excepting the salt-
was followed by a slight relative emergence of the induced high in the north, the only possible positive
area towards the end of the period. Subsidence along elements were a horst-like feature immediately north
most faults of the P-system as well as the C-system of the C4-fault and the SI-area south of the Cl-fault.
>-'
--:)
o

MILL. TECTONIC e, 2A 02 0, 02 p, P3.4 P6 P6 Pl.afs,1 P11


e. e6 C6 e2A eH l r 5A 4A F5A G' G' P121p13jP141p151p16
Y S. EVENTS F" "
"' " " "
LEGEND
Active faulting

Assumed faulting
. ......I
? "t:
"" •• Flexurlng

I ? ? Colour. refer to colour


'"~
"'"w • • code InFig.2.4.
• • • • ·
'" •
·• ·· ~
z
• 6 < ~
·•••
MAASTAICH.
~
~ • • • • •
73 o'"
W , m~ • • • •
C .... MPANIAN OW
••
• z • • '""'"'
., en"'" .~~ • ·•• ·•• •••• ·· I
~~J :::> ... ~:;;1~~~ l~~' • • ~
0
• ......I
k:ENOMANIA ~
• · •• · · •• •• • ~:
"
97,5 W ~ • •
z W
• · ~
u •
...... • ·
ALBIAN '"0Z • • • ~.
~~ --
• · ·
,,~
"'" ......
," W ... APTIAN ~
,"
. BARREMIAN
z ~
""
......
H .... UTERIVIA < • • •
·• •• ·••• ···
'" U"
··• ·• a
", .. .... l"NGINIA
... •• I I • ·

". BERRIASIAN
~ ~

· "I
• · ~
".
"0
.
~
VOLG'AN
IMMEAIDG.
~ oJ;;: ~
c
z
<
• •
I
· ·· ...!
I I I II I I



... ·•• ·• ...... ...... •?


::::
.~
'" " ....I OXFOROIAN
• • t... :

......... ? ?
~ ... ...
CA.lLOVIAN "'"'
'" E ::a:. • ... I · I
BATHONIAN
'">'
,'"" ~~ BAJOCIAN " • · · ·· · . I I S'
" , ic°
,,- AALENIAN

'90
, :::>lE ~~ • ~'"'"'
TOARCIAN
..
,,. "" ... .. PliENSBACH.
..i 7fL
i':
... •
·· S·
;:!
200
<[ SINEMUAI"N
...... ·
• I
20. " • VJ.
W HETTANGIAN
~ ...
'" U" RHAETI ... ,..

· ..::;:;
"''""'
NORIAN • •
226-'"
'" en"
CARNIAN
".~ .........
2"
en'"
"d LADINIAN
23.
~ E ... ·
ANISIAN
243 ~
" SCYTHIAN " ...
240

Fig. 6. Timing of activity of different structural elements in Block 2/2. Names (CI, CIAI, etc.) refer to code in Fig. 3.
Geochemical Interpretation of Block 2/2, Qt.!'shore Norway 171

The lack of obvious structural highs during this times the configuration shifted dramatically from a
period may be surprising, considering that the sands situation of subsidence dominated by the developing
of the Ula Formation were deposited in this time Fault System C, to one where the rapid subsidence of
interval. However, the input of grain-flow deposits of the Central Graben started to control the depositional
the Ula Formation reflects the relative uplift of the pattern. Even during this period Fault System C was
northern parts of the area, possibly combined with important in acting as an axis of rotation, but the main
tectonic activity along the C4 and C5 segments. movements along this structure had ceased.
The influence of the different tectonic elements and
differences in proximity to the source areas are clearly Tertiary
demonstrated by the varying stratigraphy seen in the From the Danian through the Tertiary the deforma-
different wells. Thus, the Oxfordian-early Kimmer-
idgian interval is represented by an 80 m thick sandy tion in the 2/2 area was moderate. A southward rota-
tion continued throughout this period, but although a
sequence in well 2/2-3, while in well 2/2-1 only an 8 m
thick sandstone is present (see Fig. 5). slight flexuring above Fault System C may be traced,
these faults no longer defined an important rotational
Towards the end of Kimmeridgian times, an increas-
axis. In addition to subsidence in the south,
ing amount of fine sands were introduced into the
halokinesis affected the topography by slight doming
2/2-1 area (see Fig. 5). These are believed to have
of the strata immediately above the salt pillows.
originated as shoals along the flanks of the Vestland
Arch, and have been transported into the basin as
grain flows, possibly triggered by earthquakes associ-
ated with the emergent activity of faults C4 and C5
SOURCE ROCKS
(Fig. 4).
Two potential source rock units were deposited within
Early-middle Volgian the Block 2/2 area according to the history described
above (see Fig. 2). The middle Jurassic Bryne Forma-
During the early Volgian, a considerable relief tion was deposited during a period of slow subsidence
developed south of Fault System C. All of the struc- when the tectonic activity was restricted to a few
tural elements were now active, either as flexures or faults, and the major movements were probably con-
faults. Also, the antithetic CA-faults were activated, trolled by halokinesis. The depositional environment
defining a more-or-Iess continuous graben-like system was a swampy coastal plain. The Bryne Formation
which was filled in by very fine to coarse sands belong- contains gas-prone source rocks.
ing to the Middle 'Clean' Member (see Fig. 5). This The late Jurassic Tyne Group consists of the Far-
unit contains the best reservoir rocks in the area. The sund and Mandai Formations. These units were depo-
geometrical configuration varied considerably along sited during a period of increasing tectonic activity
strike of the graben structure. During this period the associated with Callovian to Kimmeridgian and Vol-
2/2-3 area formed a structural high above a salt pillow gian transgressive events.
(Sl in Fig. 3), and sand deposits were not preserved.
In middle Volgian times an increase in water depth MandaI Formation
is reflected in deposition of shales of the Mandai For-
mation which encompasses thin stringers of storm- The Mandai Formation consists of dark grey clay-
emplaced sands (Middle Volgian Sandstone Member of stone, and is readily recognized from the 'hot shale'
Fig. 5, see also Fig. 4). response on the gamma log. The unit is very thin in
the wells drilled in the block, and only a few reliable
Late Volgian-early Berriasian data are available.
Average organic carbon content, pyrolysate yield
During the late Volgian and early Berriasian fault (Rock-Eval Sz-peak), hydrogen index, kerogen type
activity was restricted to Fault System C. Sediments and likely products are listed in Table 1. The content
of this age are only preserved in a shallow half-graben of organic carbon is high, and the pyrolysate yield
immediately to the south of this fault system, and very confirms a high potential for generation of hydrocar-
little structural information is accordingly available bons. Corresponding high hydrogen index values indi-
from this period. To the north and south of the cate oil-source capacity.
graben-like structure non-deposition or erosion Although the unit is very thin in the wells and prob-
prevailed. ably within the Block 2/2 area in general, it is well

Early Cretaceous TABLE 1.


Throughout most of the early Cretaceous several of
the fault systems in the area were reactivated, result- Source rock ul/it Mandai Frn." FarsundFIYI. Bryne Fm./J
ing in a series of local structural highs and lows. Organic 5.72 3.67 7.5
Again, the geometrical configuration varied along carbon (wt%)
strike of Fault System C as different antithetic faults Pyroly"ate (Sz) 35.80 4.67 11.5
developed. The point of maximum subsidence was yield (mg/g rock)
associated with Fault System C, and rotation towards Hydrogen 626 144 150
this feature persisted. The area to the south of the index
graben-like structure still underwent relative uplift, Kerogen type II II-III III
whereas the areas to the north were subject to a
Predominant Oil Oil/gas Gas
period of even subsidence. hydrocarbon
product
Late Cretaceous
'One sample only.
In late Cretaceous (Campanian) to early Tertiary blnterpreten values, not measure<!.
172 Petroleum Geochemistry in Exploration of the Norwegian Shelf

-
DEPTH
+ SOURce MATURITY GUIDE

..
m SEA BED
ORGANC MATTER TYPE
LINE 55 8252 358 OIL
500
5_N
0,35
IMMATURE
0,50
1000 0,55
.... 0," EARLY
... OIL
....
1500 0,70

0,80 PEAK OIL


EARlY
0," DRY GAS
2000 WET GAS
1,00
LATE 01.
1,20

.....
1,30 COND.
COND.
2500 1,50
WET GAS DRY GAS
1,70

I,"
3000
LATE
DRY GAS DAVIlAS DRY GAS
2,10

3500

4000

4500

5000
212
SP20 SPl160

Fig. 7. Maturity cross-section N-S Block 2/2, For location see Figs. 4 and 5,

established that sufficient volumes are present in


structurally deeper areas to make it the major oil
source rock in the Central Graben. Its contribution to
the 2/2-1 oil discovery, however, appears to be of
minor importance.
A maturity cross-section is shown in Fig. 7, and a
maturity map at the Base Cretaceous seismic reflec-
tor, representing the top MandaI Formation, is pre-
sented in Fig. 8. The maturity pattern within the
block shows a variation from early mature to late
mature, implying that the MandaI Formation is an
active, although possibly small, oil source even within
the block.
Farsund Formation
The bulk of the Tyne Group is made up of the Farsund
Formation. This unit is particularly well-characterized
in well 2/2-3, where it is penetrated with a total
thickness of 353 m. The unit is relatively homogene-
ous, and consists of medium to dark grey claystones.
Average organic carbon content, pyrolysate yield
(Rock-Eval S2-peak), hydrogen index, kerogen type
and likely products are listed in Table 1. Figure 9
shows a source rock profile of the unit. The content of
organic carbon is moderate to high and ranges from
1.08 to 13.80 wt%, with an average of 3.67 wt%. The
pyrolysate yield ranges from 0.68 to 17.52 mg/g rock,
with corresponding hydrogen indices ranging from 45
to 273. These give averages of 4.67 mg/g rock and 144,
respectively, which are moderate values and suggest a
type III kerogen. It should, however, be taken into
account that the samples analysed are early to peak
Q12345Km
..... IwwI !wwI
mature, indicating that the pyrolysis values are
depressed compared with similar samples from an
Fig. 8. Maturity map, Top MandaI Formation, immature setting.
Geochemical Interpretation of Block 2/2, Offshore Norway 173

GEOLOGY ORGANIC CARBON PYROLYSATE YIELD HYDROGEN INDEX VISUAL


% wt mg/g rock KEROGEN
4 4 8 12 100 200 300

l"d'U' - I I I
FM ~

I
II

:: .~:
-~~ .::

r-p - it:, ':.:": I


t-- It:, I
,

'] \

ULA
FM I

Fig. 9. Maturity map, Top Bryne Formation.

Pyrolysis-GC analysis, on the contrary, indicates Bryne Formation


kerogen type II. A pyrogram shows n-alkane/alkene
homologues ranging beyond n-C 20 , and the abun- The Bryne Formation is developed in wells 2/2-1 and
dance of aromatic compounds is low relative to the 2/2-2, with thicknesses of 65 and 78 m, respectively.
aliphatics. The sequence consists of alternating coal, shale and
The organic matter composition is dominated by sandstone. The coal and carbonaceous shale, charac-
amorphous organic components associated with a sig- terized by densities less than 2.0 g/cm 3, make up a
nificant proportion of herbaceous and minor amounts cumulative thickness of only 14 and 5.5 m, respec-
of woody/coaly material. The amorphous components tively, in the two wells. This suggests that their
show predominantly weak brownish fluorescence, in source rock volume is restricted, and that the unit is
parts with more intense yellowish fluorescence, indi- only a secondary source for gas.
cating liquid hydrocarbon potential. The hydrocarbon
potential of the Farsund Formation can therefore pre-
sumably be rated as moderate for both oil and gas. DETECTION OF MIGRATED
The maturity maps at base Cretaceous and the top HYDROCARBONS
Bryne Formation (Figs. 7,8 and 10) together provide
information on the maturity setting of the Farsund The headspace gas composition over the pre-Chalk
Formation. The patterns indicate lateral variation section in the three wells is shown in Fig. 11 and indi-
from early mature to late mature, implying that the cates the presence of thermogenic hydrocarbon gases.
Farsund Formation is an active and relatively vol- In wells 2/2-1 and 2/2-3 the abundance of gas is
uminous source rock even within the block area. In low to moderate throughout the section. The wetness
addition, as is indicated in Fig. 5, the Farsund Forma- (C 2+/C 1+), however, is as high as 0.7, and the
tion in well 2/2-3 interacts favourably with the lower isobutane/n-butane ratio is mostly lower than 0.5.
part of the Ula Formation in well 2/2-1. This type of These characteristics can confidently be interpreted in
setting is ideal with respect to lateral drainage from terms of a thermogenic origin for the headspace gas.
the source rock directly into the reservoir rock. This, In well 2/2-2, on the other hand, the wetness
combined with effective cap rocks, can explain the (C2+/Cl+) is mostly below 0.2 and the iso-butane/n-
location ofthe 2/2-1 oil accumulation in the lowermost butane ratio is above 1.0. This points towards a pre-
part of the Ula Formation. dominantly immature origin for this headspace gas.
174 Petroleum Geochemistry in Exploration of the Norwegian Shelf

molecular composition is dominated by methane


associated with a significant proportion of ethane,
SOUflCE MATURITY GUIDE
ORGANIC M,I,TTE~ TV""

propane, iso-butane and n-butane, making up 31.1 %.


The iso-butane/nbutane ratio is 0.38, a value normally
..... 0,55
0,50
taken as indicating association with the main phase of
oil generation from the source rock.
«:: The isotope composition of the methane with a b l3C
+::::: 0,80 PUI(Oll
value of -48.6%0 supports a thermogenic origin. The
full suite of hydrocarbon gases has been analysed for
.*'::::' o,~ WETGASi
I EARLY
DRVGAS

carbon isotope composition. The relatively wide


• 1,00
1,20 LATEOIL
1,30

..... isotope separations observed in the 2/2-1 gas compo-


1,50
1,10
WET GAS ORvGAS
nents will fit the maturity diagram of James (1983) at
around an LOM (level of organic metamorphism) of
2,10 DAYGAS DRY GAS ~"
DRY GAS
8-9, suggesting that the major portion of the gas
originates from a low to moderate maturity level
within the oil window (see Fig. 12). An analogous
explanation for the origin of the gas can be inferred by
using Schoell's (1983) diagrams, as shown in Fig. 13.
The heptane and iso-heptane values of Thompson
(1983) (see Table 3) suggest a moderate thermal evolu-
tion within the oil window.

TABLE 3. Heptane ratios, well 2/2-1


Heptane-value a 22.6
Iso-heptane value" 1.28

aDefined by Thompson (1983).

C 15+ hydrocarbons
012345
....... !wew! IwwI Km
Gross composition

Fig. 10. Source rock profile of the Farsund Formation. The gross composition of the C 15+ hydrocarbons of the
oil is given in Table 4. The composition is highly paraf-
finic consisting of 79.0% saturates, 16.2% aromatics
Moreover, it suggests that the Jurassic section in and 4.8% non-hydrocarbons.
well 2/2-2 is devoid of thermogenic and migrated
hydrocarbons. Gas chromatography
A gas chromatogram of the saturated hydrocarbon is
GEOCHEMISTRY OF THE 2/2-1 OIL shown in Fig. 14. Carbon preference index and isop-
renoid hydrocarbon ratios are given in Table 4.
Some bulk properties of the gaseous hydrocarbons The gas chromatogram shows a unimodal distribu-
from well 2/2-1 are given in Table 2. The oil from this tion with a maximum below n-C 17 , a slight hump
well is light, having an API gravity of 44 0, the sulphur around n-C23 and extending out to n-C 33 • There is a
content is low (0.15%) and the trace elements nickel slight odd-over-even predominance beyond n-C 2l
and vanadium are present in very low quantities (CPI = 1.1). Based on these features the oil appears
(1 mgfl). to have a low to moderate maturity. On the other
hand, the pristane/n-C I7 ratio is only 0.4, and could
indicate a higher maturity. The pristane/phytane ratio
TABLE 2. Gas composition, well 2/2-1 of 2.2 is relatively high for a North Sea oil, and is
suggestive of a terrestrially-derived nature for the
Molecular Carbon isotope source rock organic matter.
composition composition
(%) liiSC

Methane 68.9 -48.60 TABLE 4. Well 2/2-1 oil


Ethane 16.8 -35.69
Propane 10.9 -31.29 Gross composition (%)
Iso-butane 0.9 -29.34 Saturates 79.0
N-Butane 2.4 -29.44 Aromatics 16.2
N on-hydrocarhons 4.8
(C 2 - C 4)/(C I - C 4) x 100 = 31.1.
GC-ratios
Iso-butane/N-butane = 0.38. Carbon preference index 1.1
Pristane/phytane 2.2
Pristane/n-C I7 0.4
Gaseous hydrocarbons Carbon isotope composition (1iISC)
The molecular and carbon isotope composition of the Saturates - 27.4
Aromatics - 26.1
hydrocarbon gases are also included in Table 2. The
Geochemical Interpretation of Block 2/2, Offshore Norway 175

WELL 212-1 Abundance iC4/nC4

CHALK Gp

CROMER KNOLL GP

FORTH GP

VE$TLAND GP

BRYNE FM

TRIASSIC GP

0-2

0-30000 ppm 0-100 %

WELL 212-2 Wetness IC4/nC4


ROGALAND GP

CHALK GP

CROMER KNOll GP

VESTlAND GP - BRYNE FM

TRIASSIC GP

0-30000ppm 0-100%
... 0-2

WELL 212-3 Abundance Wetness iC4/nC4

CHALK GP

CROMER KNOLL GP

III II ..
0-30000ppm 0-100% 0-2

Fig. 11. Headspace gas composition over the pre-Chalk section in the wells 2/2-1, 2/2-2 and 2/2-3.

LOM
10 12 14 16 18 Carbon isotope composition
The carbon isotope compositions for the C15+ fractions
II
CONDENSATE of the oil are included in Table 4. The b13C values of
METHANE
I HIGH TEMPERATURE
METHANE -27.4 and -26.1%0 respectively, for the saturates
and aromatics are moderate for North Sea oils.
Biomarkers
In order to determine the source rock maturity for
North Sea oils, Cornford et al. (1983) defined a set of
ratio parameters based on the high carbon number
biomarker molecules. The oils were grouped into
three categories: 'early mature', 'fully mature' and
2- 22+ 3-33+4-
'condensate' as indicative of high maturity. A com-
THERMAL Al TERArlON INDEX
parison of the 2/2-1 oil with this classification scheme
Fig. 12. Gaseous hydrocarbons from the 2/2-1 oil discovery is shown in Table 5. It is evident that the 2/2-1 oil fits
plotted in James' (1983) maturity diagram. into the early mature to fully mature category.
......
-'l
0>

813Cethane (ppt) C 2'<%)


-50 -40 -30 -20 10 20 30 50

-70 -70
1 1 '"i:l
-a. ['"
a. :;:
'"
- -60 ;§
] -60~ I c;':)
", 1 I ~
I ::c c
'";::,-
() '"
J I '"
", , () ;§
J J I J C") ~.

" I I ~
I I "0 ;;1.
I I I:;tj
~ -50~ {l
c
~.
c
;2
-40~--------\+i-{Md)~::\{ I. I. \m/ -I -40 1(· .\~~s~~-::-p.>:::-:.-:~.: ..<.:/-___________ ~ .Q.,
s.
'"<:
c
~
~
15·
-30~ ;2
en
;::,-
TT(h) ~ mixed e:;::
'"

'
Migration?
:. r·: ...
Migration?
l
- 2O r 100 90 80 70
Fig. 13. Gaseous hydrocarbons from the 2/2-1 oil discovery plotted in Schoell's (1983) diagrams.
Geochemicallntel'pretation 0.( Block 2/2, Of(shore Norway 177

In order to solve this problem a hydrous pyrolysis


test was performed according to the method of Win-
ters et al. (1983). In this experiment hydrocarbons are
artificially produced from the source rock kerogen. A
representative sample from the Farsund FOrIl1ation
was pyrolysed and the hydrocarbons produced were
found to correlate fairly well with the previously
extracted hydrocarbons, thus suggesting that the
hydrocarbons are indigenous to the formation.

DISCUSSION AND CONCLUSIONS


Fig. 14. Gas chromatogram of the saturated hydrocarbons
in a stripped oil sample from the 2/2-1 discovery. Late Jurassic shales are generally considered to be the
main source rock of the major oil discoveries in the
North Sea, although the geographical spread between
TABLE 5. Biomarker ratios, 2/2-1 oil the fields are considerable (e.g. Barnard and Cooper,
1981). Early and middle Jurassic source rocks are also
Biomarker Early FuJI!! considered to be of importance in some areas (Fuller
nllio a mature mature Condensate 2/2-1 oil 1975; Oudin, 1976). Despite general similarities in th~
history of hydrocarbon maturation and migration over
Tt 1 40 57 90 25
Tt 4 48 33 :25 8 large parts of the North Sea graben system, more
Tti) /;! 8 0 14 detailed investigations often reveal subtle differences
Ttil 61 62 66 56 in timing and development between the various parts
St 9 38 67 80 54
of the area.
aAfter Cornford et al. (1983).
In Block 2/2 the significance of local structural and
sedimentological events is clearly demonstrated.
Firstly, the distribution of source rocks and reservoir
CORRELATION rocks is controlled by the structural development
along the strike of the Hidra Fault Zone. The structur-
It has been established that the Farsund Formation is ing was in part governed by halokinesis, which in turn
the prime candidate for the possible source rock for was dependent on the configuration and sedimentary
2/2-1 oil. It is present in the area in a significant vol- thickness variations in the Zechstein basin. Local var-
ume, it is mature and has a favourable setting with iations in development of the Hidra Fault Zone caused
respect to effective drainage from source rock into the considerable facies variations within the 'Ula Trend'
r~servoir. In order to test this hypothesis, a correla-
throughout the Jurassic, when the Middle Jurassic
tIon between the 2/2-1 oil and the Farsund Formation Bryne Formation and the Late Jurassic Tyne Group
in well 2/2-3 was attempted. (which includes the source rocks of the Farsund and
Hydrocarbons extracted from the Farsund Forma- Mandai For~ations) were deposited. Simultaneously,
tion show a front-biased n-alkane distribution, slight the reservoIr sands of the Ula Formation were
od?-over-even predominance beyond n-C 2h with developed adjacent to the most active parts of the
prIstane/n-C 17 and pristane/phytane ratios of around fault zone. This has caused a geometrical juxtaposition
0.5 and 1.8, respectively. These features are also of potential source rocks of the Farsund Formation
observed in the 2/2-1 oil, see Table 6. and the reservoir rocks of the Via Formation which is
One of the more reliable methods for oil-source cor- very favourable for effective hydrocarbon drainage.
relation is the use of GC- MS 'fingerprints' of the high On the other hand, communication between the upper
carb?n ~u~ber steranes and triterpanes. The present part of the potential source sequence (the Mandai
studIes Il1dICate that there is a high degree of similar- Formation) and the reservoir rocks is not favoured.
ity in biomarkers in the oil from well 2/2-1 and rock The Upper Jurassic Mandai and Farsund Formations
extracts from the Farsund Formation of well 2/2-3. classify as marine oil-prone source rocks, whereas the
It may be argued that the extracted hydrocarbons Middle Jurassic Bryne Formation is a coal-bearing
are non-indigenous to the Farsund Formation, making gas-prone source rock. The Mandai and Bryne Forma-
the apparent correlation invalid. All of the samples tions are only developed as thin intervals within the
have very high abundances of extractable hydrocar- block area, and significant hydrocarbon contribution
bons, averaging 2043 ppm of rock, and high hydrocar- from these units has to rely upon long-distance migra-
bon/organic matter and hydrocarbon/total extract tion from outside the block area. The Farsund Forma-
ratios, averaging 0.082 and 0.69, respectively. These tio~, on the oth.er ?and, has a more favourable setting.
e~hanced value~ can be interpreted in terms of good
It IS present wlthll1 the block area in a significant vol-
011 source capacIty or, alternatively, in terms of non- ume, it is mature and has a close relationship with Ula
indigenous hydrocarbons. Formation sandstones which favours effective drain-
age into the reservoir.
The 2/2-1 oil is of low to moderate maturity and
TABLE 6. Oil-source correlation appears to originate from a source rock with a signifi-
can~ ~errestrial i.nput. Good indications are given of a
Parameter :2/:2-1 oil FarsulId Fifl. posItIve correlatIOn between the oil and the thick Far-
su~d FOrIl1ation in well 2/2-3. It should, however, be
Carbon preference index 1.1 1.1 pOll1ted out that this does not rule out the possibility of
Pristane/phytane 2.2 1.8 other source rocks, for instance a thicker development
Pristane!II-C j , 0.4 0.5
of the Mandai Formation.
178 Petroleum Geochemistry in E.rplomtioll of the Norwegian Shelf

ACKNOWLEDGEMENTS Hamar, G. P., Jacobsen, K. H., Ormaasen, D. E. and Skarpnes, O.


1980. Tectonic development of the North Sea north of the Central
Highs. In: The Sedimentation of the North Sea Re."enlOir Rocks.
The authors would like to thank Arco Norway Inc., Norwegian Petroleum Society, 1980, 23 pp.
Mobil Exploration, Norway, Norsk Hydro A. S., Hamar, G. P., Fjreran, T. and Hesjedal, A. 1983. Jurassic stratig-
Statoil and Saga Petroleum a.s. for permission to pub- raphy and tectonics of the southeastern Norwegian offshore. In:
lish this paper. Kaasschieter, J. P. H. and Reijers, T. J. A. (ells.), Petruleli III
Geology of the Southea~terl1 North Sea and the Adjacent Ol/shore
As indicated in the introduction, several people Area. Geol. Mijnbouw, 62, 103-114.
were involved in the study which formed the basis for Hesjedal, A. and Hamar, G. P. 1983. Lower Cretaceous stratigra-
the present paper, and we wish to thank them all for phy and tectonics of the south-southeastern Norwegian offshore.
enthusiasm and useful discussions during the work. In: Kaasschieter, J. P. H. and Reijers, T. J. A. (ens.), Petroleuill
Geology of the Southeastern North Sea and the Adiacent OURho/'e
Jill-Renee Sonrier M~rk and Liv Ravdal typed and Areas. Geol. Mijnbouw, 62, 135-144.
drafted several versions of the manuscript and the .James, A. T. 1983. Correlation of natural gas by use of carbon
figures, respectively. isotopic nistribution between hydrocarbon components. Alii. 118s.
Petrol. Geol. Bull., 67, 1176-1191.
Oudin, J. 1976. Geochemical study of the North Sea basin. Bull.
Centre Reck., Pau., 10,339-358.
REFERENCES R~nnevik, H. C., van den Bosch, W. and Bandlien, E. H. 1975. A
proposed nomenclature for the main structural features in the
Bailey, C. C., Price, 1. and Spencer, A. M. 1981. The Ula Oil Field, Norwegian North Sea. In: Finstad, K. G. ann Selley, R. C. (eds.),
Block 7/12, Norway. In: Proc. Norwegian Symp. on E;cploratioll, Pl'oc., Jurassic N01them North Sea Symp. Norwegian Pe-
Bergen, 1981. Norwegian Petroleum Society, NSE/18, 26 pp. troleum Society, JNNS/1R, 6 pp.
Barnard, P. C. and Cooper, B. S. 1981. Oils and source rocks of the Schoell, M. 1983. Genetic characterization of natural gases. A,n.
North Sea Area. In: Illing, L. V. and Hobson, G. D. (eds.), Pe- Ass. Petrol. Geol. Bull .. 67,2225-2238.
t1'O/eum Geology of the Continental Shelf of North- West Europe. Skjerven, J., Rijs, F. and Kalheim, J. E. 1983. Late Palaeozoic to
Institute of Petroleum, London, pp. 169-175. Early Cenozoic structural development ofthe south-southeastern
Cornford, C., Morrow, J. A., Turrington, A., Miles, J. A. and l\orwegian North Sea. In: Kaasschieter, J. P. H. and Reijers,
Brooks, J. 1983. Some geological controls on oil composition in the T ..J. A. (eds.), Petrolelllll Geology o(the Southeastcl'n NO/th Sea
U.K. North Sea. In: Brooks, J. (ed.), Petroleum Geochemistry and the Adjacent Onshul'c Areas. Geol. :W"ijnbouw, 62,35-45.
alld Exploration of Europe. Geol. Soc. Spec. Publ. No. u. Thompson, K. F. M. 1983. Classification and thermal history of
71-93. (Blackwell Scientific, Oxford.) petroleum based on light hydrocarbons. Geuchilll. COBI/IOChil/l.
Deegan, C. E. and Scull, B. J. 1977. A proposed standarrl litho- Acia, 47,303-316.
stratigraphic nomenclature for the Mesozoic of the Central and Vallset, J. and Dore, A. G. (eels.) 19R4. A revised Triassic and
Northern ~orth Sea. Rep. Inst. Geol. Sci., No. 77/:25. Norwegian Jurassic lithostratigraphic nomenclature for the )Jorwegian
Petrol. Directorate Bull., 1, 36 pp. North Sea. NPD-Bull. Nu. 8, 53 pp.
Fuller, J. G. C. M. 1975. Jurassic source rock potential and hy- Winters, J. C., Williams, J. A. and Lewan, ;VI. D. 1983. A labora-
drocarbon con'elation, North Sea. in: Finstad, K. G. and Selley, tory study of petroleum generation by hydrous pyrolysis. In:
R. C. (eds.), Proc. Jnrassic N01ihernNorthSeaSymp. Norwegian Bjol'loy, M. et al. (eels.), Advances in Organic Geochmnislry 1981.
Petroleum Society JNNSS/11, 1-18. Wiley, Chichester, pp. 524-533.
14 Organic matter quality changes
in an Upper Jurassic shale
sequence from the Viking
Graben
A. Y. Huc*, H. Irwint and M. Schoell"t
*Institut Franc;ais du Petrole, France
tStatoil, Stavanger, Norway
"tAemeler Strasse 3, D3000 Hannover, FRG

Changes in the characteristics of the organic matter in a 120 m thick relatively immature (T max = 428-436°C)
section of Upper Jurassic shales from a borehole in the Norwegian sector of the Viking Graben are presented.
Cuttings and sidewall cores have been analysed. Organic carbon concentration remains fairly high throughout the
interval studied and averages 5-6%. However, the quality of the organic matter exhibits a dramatic change with
depth, which is evident as a sharp drop in the hydrogen index (from 600 to 200) and H/C atomic ratio of the
kerogens (from 1.25 to 0.80) across the boundary between the Draupne and Heather Formations. This change in
kerogen type is associated with a decrease in bitumen content and modifications in the extractable hydrocarbons
including biomarkers. In particular, 17a(H),18a(H),21{J(H)-C 2s -C 3o bisnorhopane is absent in the 'hot' Draupne
shales, is present in the underlying 'cold' shales and increases downwards within the Heather Formation. A large
change in isotopic composition is clearly related to this organic quality pattern. The upper hot shales are depleted
in laC (OI3C = -30%,) and in deuterium (oD = -125%,); this gradually changes downwards into the lower shales to
a Ol3C of -24'h and a D-value of -100%., i.e. a comparative enrichment in carbon-13 and deuterium.

INTRODUCTION So it is now evident that more information is needed


on the time-space distribution of kerogen types and
The Kimmeridge Clay Formation in the Viking Gra- quantity. Moreover, such studies would permit a
ben of the North Sea consists generally of dark grey- deeper insight into the relationship between deposi-
brown to black carbonaceous, radioactive claystone. tional factors and organic matter composition. This
The formation varies considerably in thickness, study provides an example of vertical changes in
exceeding several hundred metres in the central part characteristic::; of the organic matter in 120 m of an
of the graben, but may be missing due to non- Upper Jurassic t3hale sequence penetrated by a
deposition or erm3ion on the outer edges of the basin borehole in the Viking Graben area.
and on the crest of some structural blocks. These
shales are often not differentiated in terms of organic
type, and are frequently considered as homogeneous
good source rock. THE SEQUENCE
As regional geochemical studies progress it
becomes obvious that such a statement can no longer The sequence examined (see Fig. 1) encompasses the
be supported. Demaison et al. (1983) have shown, on Draupne Formation (i.e. Kimmeridge Clay equiva-
the basis of over 2200 analyset3 by Rock-Eval lent) and the top of the Heather Formation. It is
pyrolysis and total organic carbon (TOC) content, that dominantly a sequence of interbedded marine shales
large variations in kerogen type can be observed later- and claystones, but changes to siltstone at the base,
ally on a basinal scale. Moreover, they pointed out when limestone and sandstone interbeds also appear.
that these changes in kerogen types can also be recog- The typical 'hot' shales occur at the top of the se-
nized along a vertical dimension in individual quence but the gamma-ray response drops dramatically
boreholes. Such an observation cannot be ignored in at 2625 m, further decreases gradually to 2655 m, and
future basin geochemical studies, particularly when then remains fairly uniform down to the base of the
prediction of regional source bed occurrence and sequence studied. The maturity of the sequence lies
organic facies variations is required by the between 0.5 and 0.55% Ro and can be termed imma-
explorationists. ture to early mature.
Petroleum Geochemistry in Exploration o(the Norwegian Shelf' © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 179-183.
180 Petroleum Geochemistry in Exploration of the Norwegian Shelf

CUT;INGS S~C ATOMIC RATIO HIC

n
....... •.
TOCI,") HYDROGEN INOEX- -TOC(%l HYDROGEN INOEX-----. +---CUITlNGS-----.-swc------.
12 to OB O£ 0.6 0.8 10 12

..
10 B 64 2 0 200 400 600 10864 20 2C() 400 600

.eA..
12~ .
".
/
.
12600

..
10

l• '" .. e •

] ·
...".
... .. "
1
§
.. ...-..•

·•• .... ~
..·
2650

.

e... .to;.

~RO.C~~MPL£S
.. KEROGENS
•• 021:------,----...,-J
o 10 20
27 00·
~_ _...J

L _ _ _ _--'
- ~_ _ ' - -_ _ _----1 }---ATOMICRATIO o/c (xl02 1---1
VITRINITE REF/ECTANCE04fi_051 TWI iKCi4l6.4J2'C
Fig. 2. Variation in the elemental composition of kerogens.
Fig. 1. Vertic.al distribution of total organic carbon (TOe)
and hydrogen mdex and a comparison between cuttings and
sidewall samples.
eral trend of the organic variation which is indicated
by both sets of samples.
The kerogen elemental composition mirrors the
EXPERIMENTAL hydrogen index trend: the atomic H/C ratio declines
~ro~ 1.2 to 0.8 with increasing depth, and probably
mdICa~es t.hat the deepest kerogen samples were
The.present study is based on analyses of cuttings and
of sIdewall cores. Organic characterization on whole deposIted III a more oxygenated environment than
rock samples has been carried out by Rock-Eval their shallower counterparts (see Fig. 2). The elemen-
pyrolysis and total organic carbon (TOC) content tal compositions indicate a transition from kerogens
measurements. comparable wit? type II toward kerogens of type III,
Kerogens have been isolated by destruction of min- as ~efined by TIsSOt et al. (1974). A comparison ofH/C
erals by acid attack then subjected to Rock-Eval ratIO values over the depth interval studied confirms
pyrolysis, elemental analysis and stable isotope eac t?e differential behaviour between cuttings and
and D) analyses. SIdewall core samples, as discussed earlier (see
Fig. 2).
The powdered shales were extracted using CH CI
then the extract was separated into saturated aro:na~ T~e isotop~c. compos.ition of sidewall core kerogens
tic and NSO fractions by thin-layer chromatdgraphy. provIdes addItIOnal eVIdence of a regular variation in
The saturated fraction was analysed by GC and the type .of organic .matter, as there is a progressive
GC-MS techniques. and SIgnIficant ennchment in both carbon-13 (b 13C
Clay mineralogy was carried out using standard from -30 to -24%0) and in deuterium (bD from -125
X-ray diffraction procedures on oriented samples. to -100%~) with increasing depth (see Fig. 3).
The ultImate causes of isotopic variation in kerogens
remain speculative (Schoell, 1984). Whatever the
RESULTS AND DISCUSSION reason for the changes, they reflect modifications in
t~e organ~c m~terial which are indicated by a tech-
TOC and Rock-Eval pyrolysis data on rock samples mque whICh IS totally independent of Rock-Eval
and kerogens are presented in Fig. 1. In the 120 m pyrolysis and kerogen elemental analyses.
secti?n s.elected, the total organic carbon is fairly high A reliable interpretation of organic facies from the
and lIes III the range 3-8%. Hydrogen index values of study of a bitumen requires that neither thermal
the rock samples and the kerogens are close to each effects nor migration phenomena have altered the
other and thus the mineral matrix effect (Espitalie et extract composition. The present example seems to
al., 198?) is limited in this set of samples. A striking fulfil such conditions. Firstly, T max values supported
feature IS the regular decrease of the hydrogen index, by elemental analyses of kerogen indicate that the
from 600 to 200, as a function of depth. It should be ?rganic ma~ter can be considered to be relatively
noted that the organic matter in this section is rela- Immature, I.e. before the oil-generation zone. This
tively immature, exhibiting Rock-Eval T values of situation minimizes the chance of thermal alteration of
426-4320 =
C. Thus, variations in the hydrogen index the initial organic molecules, their dilution by therm-
reflect changes ~n the nature of organic matter, and ally generated hydrocarbons, and also the possibility
are only very slIghtly affected by thermal alteration.
Moreover, the depth interval of only 120 m precludes ---613C(KCI~ ---6D(KCI~

any significant thermally-induced differences between .3IJ -28 -21; -\4%' -1)0 -li'O -110 -100

the top and the bottom of the section. ",'"


swe "'", swe
. When comparing the trends of decreasing hydrogen 2600 '"
mdex between cutting samples and sidewall cores (see
Fig. 1), there are significant differences. Both data '" '"
'"
sets are regularly decreasing with depth but the
decline is sharper for the sidewall core sam~les. This
'" '"
observation suggests that caving from the uppermost '" '" '" '"
part of the section might be contaminating the down- '" '"
hole cuttings samples, causing a smoothing of the '" ISOTOPE RATIOS I
IKEROGENS.
trend. This interpretation is consistent with indica- II
tions of caving based on a caliper log examination. 2700-'------------.-J

However, the caving effect does not obscure the gen- Fig. 3. Variations in the isotopic ratios of kerogens.
Organic Matter Changes in a Jurassic Shale 181

I
m/e=191

1
600
I ! ~~T!fN6S I
• CUTTINGS SWC

C~
[~
x
~ 400 o
z
w
'"
o •
~ 300
~ 0

200 • •
100 eno
o
4 5 6
------EOM !Toe (mglgJ----~

Fig. 4. Extractable organic matter (EOM) content versus


hydrogen index.

of migration effects. Furthermore, a clear relationship


is found between the hydrogen index of kerogens and
the extractable organic matter (EOM) contents of the
related sediments (see Fig. 4). Thus, there is no evi-
dence of noticeable hydrocarbon accumulation of
Fig. 6. Variations in the C2i\bisnorhopane relative content
depletion due to migration, at least in the C15+ frac-
in mass chromatograms m/e = 191, with a comparison be-
tion. tween cuttings and sidewall samples.
Assuming an in situ origin for the bitumen, it
should be noted that a regulated decrease in the
hydrocarbon content mirrors the decrease of ~ydro­ The parallel variations described above are also
carbon potential of the kerogen as a functIOn of reflected in the occurrence of the C28 nuclear-
depth in the studied section (see Fig. 5). . demethylated triterpane: 17a(H), 18a(H) ,21P(H)-28-
This quantity trend is reinforced by a quallty.trend: 30-bisnorhopane (Seifert et al., 1978). Bisnorhopane is
the n-alkane pattern exhibits a regular gradatIOn be- frequently observed in North Sea oils and sediments
tween a distribution with a marked preference for low (Mackenzie et al., 1983) and has been identified by
to medium molecular weight n-alkanes (n-C15 to GC-MS in most of the present samples (see Fig. 6).
n-C20) at the top, and a distribution with increa~ed Because it is relatively uncommon, it is thought to be
high molecular weight n-alkanes (n-C25 to n-C37) WIth useful as an oil-source correlation parameter.
a pronounced odd-even carbon p~ed~min.anc~ ~t the Depth-related variations in the relative proportions
bottom of the section. The latter dlstnbutIOn IS mher- of this C28 hopane in m/e 191 terpane 'fingerprints' is
ited from terrestrial higher plants while the fonner is quite spectacular. It is missing in the 'hot' Draupne
usually associated with marine organisms. Thus, the shales, and becomes the major peak in the deepest
changes in n-alkane 'fingerprints' may be interpreted levels of the Heather Formation where it even tends
as changes in organic input; the importance of the ter- to become the only peak in the sidewall core samples
restrial contribution increases in the lower part of the (see Fig. 6). It has been documented that the relative
Draupne and in the Heather Fonnation. This abundance of bisnorhopane decreases with increasing
interpretation is consistent with the variations in the maturity (Grantham et al., 1980; Moldowan and
chemical composition of the kerogens. The parallel Seifert, 1984). The fact that oils in the study area often
variations in kerogen properties and the extractable contain a higher relative abundance of bisnorhopane
hydrocarbon patterns is further evidence for an than the immature source rocks is difficult to explain.
absence of migration phenomena in the interval. It is even more puzzling that the Draupne shales,

__ --~.~--
HC/TOC
____ CUTTINGS
t
-t~~._.----~----~
SWC

4
CUTTINGS III
3 2
II
1 00
swc III
1 2 3 4
2589.

2589m

2600
• 0
0
,.
U
~
2601m

• 0
0 :z
0 2617m
0 ~
~
It: 2625m

2657m
2650 0 "-
0
"-
15 2633m


<I)
0 ~

••
It: 2649m
x
2700
2657m

h::;~~~~~::;::;:":;:'::::=;::::j 2673m
22.00
Fig. 5. Hydrocarbon content and composition with a 2.00 6.00 10.00 1400 laOO 2600

comparison between cuttings and sidewall samples. Fig. 7. X-Ray diffraction diagrams for sidewall samples.
182 Petroleum Geochemistry in Exploration of the Norwegian Shelf

which are usually considered to be the mOiSt favour- 3. In terms of source rock quality, these variations
able source rocks, contain no bisnorhopane at all. reflect significant differences in hydrocarbon source
The origin of this compound is controversial, and potential. The upper part of the section contains a
there is no definitive explanation available in the cur- kerogen which can be compared with type II, while
rent literature. Study of natural examples, such as the the lower part contains kerogen of type III.
one presented herein, where one can follow significant 4. The regularity of the organic change through the
variations of a biomarker content along with interval suggests that it is controlled by a gradual
documented changes in organic facies is possibly a way modification of the environment of deposition. The
to decipher its geochemical significance. Grantham et observed trend in n-alkane patterns and kerogen
al. (1980) suggest that the presence of bisnorhopane chemical composition can be explained by a declining
may reflect a reducing environment of deposition. input of terrestrially-derived organic matter with
This is contradicted by the present data, where the time. This interpretation is supported by clay mineral
most 'reducing' shales are devoid of bisnorhopane. associations and depositional history. However, it is
However, it cannot be ruled out that a particular not certain that such a trend might not be explained
depositional environment is required for the develop- by a progressive change in the redox properties of the
ment of bisnorhopane although the present data sug- depositional medium, the marine organic matter being
gest the prerequisite of an additional component, such destroyed in an oxidizing environment and the propor-
as the relationship with a specific organic facies, as tion of terrestrial organic matter therefore being pre-
described here. ferentially increased under such conditions.
Vertical clay mineralogical trends parallel the 5.It is apparent from comparison of geochemical
changes in organic characteristics. There is a down- data from cuttings and sidewall cores that there is a
ward decrease in mixed-layer clays (illite-smectite) sample quality problem. To some extent contamina-
and a distinct increase in illite and kaolinite. Traces of tion by cavings affects the lower part of the sequence
chlorite appear at 2633 m and below. The progressive when cuttings data are used. In this example it does
decline in illite-smectite and the appearance of chlo- not significantly change the general observations, but
rite can be explained by diagenesis, but the increase in it indicates that care should be taken when interpret-
illite and kaolinite is most readily explained by an ing data obtained from cuttings samples. Alterna-
increase in detrital components. Particularly, tively, sidewall cores are unambiguous in terms of
increased amounts of kaolinite can indicate proxi- depth and caving pollution, but one should be aware
mity to land. that they sample specific horizons and may not pro-
The clay mineral trend thus reflects the transgres- vide representative values for geochemical para-
sive sequence of basin development from deltaic con- meters at the formation scale.
ditions in the underlying Brent Group, through a 6. The kind of sequence which has been studied is
progressively deepening marine basin of the Heather possibly favourable for understanding the significance
Formation, leading finally to the widespread of C21lhopane in sediments, because the occurrence
restricted environment represented by the Draupne and quantity of this molecule can be matched with a
Formation. The vertical clay mineral distribution seen gradual variation in the environment of deposition.
here is expected to be expressed laterally, reflecting Thus, further studies of the sedimentological condi-
facies variations within the depositional basin. The tions during the deposition of this sequence would be
mineralogical data support the interpretation that of interest.
organic matter changes are related predominantly to
variations in the proportions of terrestrial material,
which increase downwards in the sequence. However
some of the characteristics of the organic matter may ACKNOWLEDGEMENTS
have also been influenced by the redox properties of
the environment of deposition. The Heather Forma- The authors thank Statoil and partners for their per-
tion was deposited in oxic to sub-oxic bottom waters, mission to publish these data.
whereas the Draupne Formation is supposed to have
been deposited in anoxic bottom waters. Differential
preservation in these two contrasting environments
could account for the variations seen.
It is very likely that the variations in organic matter REFERENCES
observed here are a result of both varying input of
terrestrial material and differential/preferential pre- Demaison, G., Holck, A. J. J., Jones, W. and Moore, G. T. 1983.
Predictive source bed stratigraphy; a guide to regional petroleum
servation of various organic components as a result of occurrence North Sea basin and eastern North American Conti-
varying redox conditions. nental Margin. 11th World Petroleum Congr. PD I, London,
1983.
Espitalie, J., Madec, M. and Tissot, B. 1980. Role of mineral matrix
in kerogen pyrolysis; influence on petroleum generation and mig-
ration. Am. As.s. Petrol. Geologists Bull., 64(1), 59-66.
Grantham, P. J., Posthuma,.J. and de Groot, K. 1980. Variation and
CONCLUSIONS signficance of the C27 and C28 triterpane content of a North Sea
cure and various N'orth Sea crude oils. In: Douglas, A. G. and
1. The study considers only one borehole, and no Maxwell, J. R. (eds.), Advances in Organic Geochemistry 1979.
regional generalizations can be drawn from the Pergamon Press, Oxford, pp. 29-48.
Mackenzie, A. S., Maxwell,.J. R., Coleman, M. 1. and Deegan, C.
results. K 1983. Biological marker and isotope studies of North Sea crude
2.Evidence of important vertical variations in the oils and sediments. 11th World Petroleum Congr. PD 1 London,
characteristics of the organic matter in an Upper 1983.
Jurassic shale sequence has been presented. :I1oldowan, J. M. and Seifert, W. K. 1984. Structure proof and sig-
Organic Matter Changes in a Jurassic Shale 183

nificance of stereoisomeric 28,30-bisnorhopanes in petroleum and Seifert, W. K., Moldowan, J. M., Smith, G. W. and Whitehead, E.
petroleum source rocks. Geochim. Cosmochim. Acta, 48, V. 1978. First proof of structure of a C2s-Pentacyclic triterpane in
1651-1661. petroleum. Nature, 271, 436-437.
Schoell, :11. 1984. Stahle isotopes in petroleum research. In: Tissot, B., Duranrl, E., Espitalie, J. and Combaz, A. 1974. Influ-
Advallces ill PetroleulII Geochemistry, Vol. 1. Academic Press, ence of nature and diagenesis of organic matter in formation of
215-245. petroleum . .1\.1/1. Ass. Petrol. Geologists Bull., 58(3),499-506.
15 Organic geochemistry of the
Oseberg Field (I)
B. Dahl and G. C. Speers
Norsk Hydro a.s., Research Centre, Bergen, Norway

Oseberg crude oil is typical of the highly paraffinic, high quality North Sea crudes with low sulphur, asphaltenes
and trace metal contents. A tar mat exists in part of the Oseberg Field, and is considered to result mainly from in
:;itu de-asphaltening of the crude followed by gravity segregation and concentration near the oil-water contact.
Shales from the Humber Group are the most likely source of Oseberg oil, with the Kimmeridge Clay Fonnation
being the richest source rock. Conventional geochemical parameters are unsatisfactory for maturity assessments
in well 30/6-1, so molecular parameters were employed. The latter also provide better indications of migrated
hydrocarbons and contamination.
The Humber Group is not considered mature enough in the Oseberg area to have generated the reservoir oil.
The major source area is most likely to the west and north-west of the field area.

INTRODUCTION the result of tensional rifting which took place mainly


during the Mesozoic. Fault trends are predominantly
The Oseberg Field lies mainly in Block 30/6 (formerly N-S to NNW-SSE, with some NE-SW rifting also
called the 'Silver Block') to the west of Bergen in the evident, the latter believed to be due to reactivation of
Norwegian sector of the North Sea. A portion of the older Caledonian structural trends.
field extends into the northern sector of Block 30/9 The Oseberg Field is situated in a rotated,
(see Figs. 1 and 2). eastward-dipping fault block resulting from the main
The licence for Block 30/6 was awarded in April tectonically active rifting phase (see Figs. 1 and 3).
1979 to a group comprising Statoil (operator), Elf, The reservoir is Brent Group sandstone which is trun-
Total, Norsk Hydro, Mobil and Saga. Operatorship cated on the crest of the fault block, and the uncon-
was transferred to Norsk Hydro in April 1982. In formity is overlain by Lower Cretaceous marls and
August 1982, the northern part of Block 30/9 was clays which act as a caprock for the accumulation.
awarded to a consortium of Norsk Hydro (operator),
Statoi! and Saga.
The first well was drilled in the summer of 1979, and PETROLEUM GEOCHEMISTRY
found gas at approximately 2330 m in Brent Group
sandstones. The two following wells also encountered Potential source rocks
gas in the same formation. A fourth well established
the presence of oil as well as gas, and showed that the Source rock evaluations have been made on samples
Oseherg Field had a large oil leg. The maximum from well 30/6-1 and four other wells in the area, and
height of the gas column is approximately 320 m, and are summarized in Fig. 4. The evaluations were pro-
the oil column is approximately 200 m. To date, 15 duced by a number of service organisations as well as
wells have been drilled in Block 30/6 and three wells in Norsk Hydro Laboratories in Bergen.
Block :~O/9. The results of Rock-Eval screening of cuttings
samples from the first Oseberg Well, 30/6-1, are
shown in Fig. 5. This well was drilled on the crest of
GEOLOGICAL SETTING the Alpha structure, and the Kimmeridge Clay is
absent at this location (lithostratigraphic nomencla-
The Oseberg Field lies on a structural feature called ture used according to Deegan and Scull (1977».Below
the Bergen High, west of the Horda Platform (see the Cretaceous unconformity, only a thin (approxi-
Fig. 1). The structural characteristics of the area are mately 30 m) section of silty Heather Formation is
Petroleum Geochenl1:siJ'Y ill ExplomtioJl of the Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 185-195.
186 Petroleum Geochemistry in Exploration of the Norwegian Shelf

4°E

PLAT ORM

TWT 10 Ihe
Kimmer ian Unconform ity

m
H 3000ms
4000ms

o 25 50 15 100ms
I

Fig. 1. Regional map of the Northern North Sea (59-62° N) showing location of the Bergen High (Oseberg Field).

present. Although the sampling density was rather 1983). The screening samples obtained from the
low, the results shown in Fig. 5 indicate that only the Paleocene samples in 30/6-1 are consistent with
Rogaland and Dunlin Groups of sediments have any results from the same formations in the 8tatfjord
significant hydrocarbon potential in this well. (Kirk, 1980) and Ekofisk areas (Van den Bark and
Figures 4 and 5 show that, within the Paleocene Thomas, 1980).
Rogaland Group, the Balder and 8ele Formations are Results from the Lower Jurassic Dunlin Group in
of most interest. Total organic carbon contents (TOe) 30/6-1 indicated some source potential in the Drake,
are in the range 1-1.5 wt%, and residual hydrocarbon Burton and Amundsen Formations. However, as the
potentials (8 2) are 3-3.7 kg/to Rock-Eval hydrogen Burton and Amundsen Formations contain mainly
and oxygen indices for these formations, together siltstones and clays, the results shown in Fig. 5 prob-
with some results from the Lista Formation, are plot- ably give an unrealistic indication of overall formation
ted on a modified Van Krevelen-type diagram in source potential. Assessment of the source potential of
Fig. 6. The results suggest that these formations have the Burton and Amundsen Formations could be influ-
little liquid hydrocarbon potential (probably gas/con- enced by cavings from the overlying Drake Forma-
densate), and that quality decreases with depth. tion, but this was not checked by palaeontological
Data for the Lista Formation shown in Fig. 6 are studies.
linked with TOe values of <1.0 wt% and residual hyd- The Drake Formation in 30/6-1 consists of silts and
rocarbon potentials (8 2) < 1 kg/to Kerogen from this claystones interbedded with marls, and has a moder-
formation would be classified as Type III only, but the ate hydrocarbon potential with 8 2 values ranging from
scatter of oxygen index values within the samples is 1 to 5.8 kg/t and TOe contents in the range 1-2 wt%.
not considered to be due to maturity variations, but The lower part of the Drake Formation has the high-
rather to the problems associated with determining est hydrocarbon potential, hydrogen indices in this
this parameter in whole rock samples (e.g. Katz, section indicating some liquid hydrocarbon potential
Organic Geochemistry of the Oseberg Field (I) 187

-<>11

\
\
\
\
\
\ -<>6
.10.10A \
\
\
\
\
\

~AL~~A)
\
\

"',>3 \, ,,_.1. ~~_


/'
\~

GAMMA

l
:ROSPECT

1:25000
3 4

Fig. 2. Oseberg Field: structural map.


188 Petroleum Geochemistry in Exploration of the Norwegian Shelf

w SCHEMATIC LITHOSTRATIGRAPHIC CROSS-SECTION - OSEBERG FIELD E


DEPTH
STRATIGRAPHY
- - - ~ ,----r--....--..,
MSL
----~-------------------------
r.-~~~~~~~~----~----~--~~ __~~__~~__~~__=~~~'~rr~=.~"~~~"~~~.~~~
000

,.
UTSlIIA

'000

'000

........-
...
SttfTlAMO

Fig. 3. East-West cross-section of the Oseberg Field.

(gas/condensate). The upper part of this formation is various geochemical parameters within the Kim-
entirely gas-prone. meridge Clay and Heather Formations are summar-
Most of the results obtained on later Oseberg wells ized in Fig. 5. TOe contents of the Kimmeridge Clay
are similar to those observed in 30/6-1. The ranges of Formation are in the range 5.3-10.2 wt%. The latter
the various parameters for the five wells considered in are higher than values reported for this formation
the study are summarized in Fig. 4. Some anomalous elsewhere in the North Sea region (Cornford, 1984).
results were observed in the deeper sections of one of Residual hydrocarbon potentials of the Kimmeridge
these additional wells, but these were shown by Clay formation from the Oseberg area are also high,
palaeontological studies to be due to cavings from the ranging from 26 to 71 kg/to
overlying Humber Group. A more detailed examination of the Kimmeridge
Figure 4 indicates that, apart from the Humber Clay Formation in one of the downflank wells is shown
Group, and possibly some narrow intervals in the in Fig. 7. Hydrogen indices are in the range 300-600,
Brent group, the remaining sedimentary intervals which indicates a variation in source quality from a
have poor hydrocarbon potential. The Brent Group mixed Type II/III to a Type II kerogen. It is also
consists mainly of sandstones and siltstones interbed- evident from Fig. 7 that the quality of the Kim-
ded with some shales and thin coals. The latter may meridge Clay Formation in this area varies with
well be responsible for some of the high S2 values depth. The trend is for a Type II kerogen to be pres-
observed during screening studies. ent in the upper part of the formation, decreasing to a
The Statfjord Formation is even more shaly than mixed Type II/III in the lower part of the section.
the Brent Group, and also contains some thin coals. A It should be noted that the results shown in Fig. 7
poor, gas-prone potential is shown by the generally were obtained almost entirely on cuttings samples,
low S2 values 0-2 kg/t) and hydrogen indices. and that the effect of cavings has probably resulted in
It has already been noted that the Kimmeridge Clay an 'averaging' of quality indications throughout the
Formation is absent in well 30/6-1, but eastwards and formation. This process would tend to lower the
downflank of the Oseberg structure the Upper apparent quality in the upper part of the formation,
Jurassic Humber Group thickens and up to 400 m of and increase it in the lower section. Unfortunately,
these sediments are encountered. The range of the only two sidewall cores were available for control pur-
Organic Geochemistry o/the Oseberg Field (I) 189

30/6-1 82 HYDROGEN INDEX


mg Ig OF ROCK mg HCI 9 ORG; CARBON
2 3 4 5 6 100 200 300
m I 1
UTSIRA
800 FM.

1000

1200

HOROA-
LAND
1400 GROUP

1500

1600

1800

2200

2400

3100
FM.

Fig. 4. Rock-Eval screening data from we1l30j6-1.

poses, and these results are also included on Fig. 7. with Kimmeridge Clay Formation quality variations
The sidewall core from the upper part of the formation reported by Barnard et al. (1981).
gave an 8 2 value higher than the associated cuttings, Quality decrease of the Kimmeridge Clay Forma-
whereas in the lower part of the section the 8WC tion with depth was also observed in visual kerogen
was taken in a thin limestone band and, apart from studies of the Oseberg wells. The upper part of the
giving a negligible 8 2 value, was unrepresentative of formation had a high content of amorphous material,
the lower part of the formation. A range of sidewall characteristic of a Type II kerogen, whereas the lower
cores from the Kimmeridge Clay Formation in section displayed a significant increase in the amount
another downflank well has also been examined, and of woody and herbaceous material.
this showed a similar trend to that observed in the Kerogen typing within the Kimmeridge Clay For-
previous well, but with a much more abrupt quality mation was based initially upon Rock-Eval hydrogen
decrease with depth. These observations are in line indices, as noted above. These results are supported
190 Petroleum Geochemistry in Exploration of the Norwegian Shelf

GAMMA RAY S. Hy4r~en Index


Period Toe He 20S
S_nos m'\l He/." or'll. corMn
Lithology Thi~:r;e" Ro(%)
IlPlunill mIlI-tC/~ fCC~
GrouplFM I%WT) .~=r:~) 1%)
p.-....
Rogaland Group
-;-~~ -= 225-288
*
1.00-tsQ 3.0-3.7 * 0.30-0.61 '0-20

ltJper Cretaceous
Shetland Group
26-909 0.60-0.80 < 1.0 O.3:H).n

Lower Cretaceous ~
G--143 0.25-0.50 <1.0 0.53
Cromer Knoll Group ~
Upper Jurass~
Kimmeridge 0-143 S.l-10.20 26.0-71.0 0.36-0.47 1&-24
ClayFM

Upper Jurassic 4.00-5.00


0-54' 1.4-15.0 0.39-1.02 12-19

*
HeatherFM occ<1.0

,....,
~
Middle Jurassic u.
29-162 0.56-1.80 0.3-11.4 O.43-0n I
Brent Group
~ '--
...J
lower Jurassic u
1-- ..-
.~
90-200 1.00-2.00 1.0-6.8 O.36-{)'92 '~2
Drake fM

Lower Jurassic
tUpper Trlasste .' •.:...~ 237-343 0.44-0.76 26-3,
t- - ;su; : lpatl: : . !.ord: : ,~; ,I..Fd ;Ie---1r~
, ~. ~ +---+---+---+----+---1
r-- '''' Not fully
Triassic 1.3&-0.89 0.1-{1.7 0.33-1.03
Cormorant FM ••• " • penetrated

* Results from Balder and SeIe Fms only.

Fig. 5. Osberg source rock summary. Ranges of values


were compiled from the wells analysed.

o Cuttings
/I

f
900 Fig. 7. Rock-Eval screening data from Kimmeridge Clay
Formation in a downflank well.

G 750

,~~
dJ
5
'"
Q
J: n Upper Kimmeridge
Clay Fm

(
600

.5'" V
x
w
0
?;
450
Z

~
W

8II:
--
0
>-
J: ,00 Mid Kimmeridge

~\.:'~ BALDER FM Clay Fm


0
(0"<:. )

"""
\0
" \ SELE FM.

" '-
-
150

- ®-~
USTA FM. \

~
--- @J -
m
~.~ -'"
- - - - - r::_~ ........
Lower Kimmeridge
Clay Fm
50 100 150

OXYGEN INDEX (mg Co,/g org. C)


•.
Fig. 6. Van Krevelen-type plot of Rock-Eval data from the Fig. 8. Pyrolysis-gas chromatography of three Kim-
Rogaland Group, well 30/6-1. meridge Clay Formation samples from a downflank well.

by the pyrolysis-gas chromatography results ('pyro- Heather Formation are present in the downflank
grams') shown in Fig. 8. The three pyrograms were wells. Geochemical data indicate that the source
taken from the upper, middle and lower sections, and potential of the Heather Formation is highly variable,
indicate a much more condensate-prone kerogen in the with the upper part being gas-prone and with low
lower part of the formation. hydrocarbon yields. Quality improves in the middle
The general observations on the Kimmeridge Clay section of the formation, which contains a more
Formation from these studies are not in agreement liquid-prone kerogen giving 82 values up to 15 kg/to
with the kerogen facies map of Barnard and Cooper The quality and hydrocarbon yield rapidly decreases
(1984) in which they indicate Type III to be dominant towards the bottom of the section, which is gas-prone.
in the Oseberg area. These indications are supported by visual kerogen
The Middle-Upper Jurassic Heather Formation studies and pyrolysis-GC results.
was penetrated in four out of the five wells used in this
stUdy. On the crest of the Oseberg structure this for-
Maturity
mation is very thin and it has not been sampled exten-
sively for geochemical studies (e.g. see Fig. 5). As In well 30(6-1, vitrinite reflectance determinations
noted previously, the Humber Group thickens were made on a number of sidewall cores from approx-
towards the east, and much greater thicknesses of the imately 1900 m to total depth. A plot of the results
Organic Geochemistry of the Oseberg Field (l) 191

for these determinations no problems exist with cav-


ings, so the wide scatter of the vitrinite data was
1800 probably due to problems in differentiating auto-
. chthonous and allochthonous components, possibly
combined with effects of bitumen 'staining'.
2000 . A more regular increase in maturation levels with
. depth was displayed by the TAl values of spores/pol-
len shown in Fig. 10. These results suggest that the
2200 . . hydrocarbon generation threshold is around 2000 m in
this well. The two rather high TAl values observed at
2392 and 2849 m appear to be related to lithology, as
2400 both occur in limestone intervals.
Although only providing a rather approximate indi-
cation of maturity, Rock-Eval pyrolysis data (T max
2600 values) suggest a generation threshold around 2400 m
in well 30/6-1. However, as these results were
obtained on cuttings, the indications may have been
2800 influenced by cavings.
Regional maturity studies (Barnard and Cooper,
1984) suggest that vitrinite reflectance values in the
3000
range 0.5-0.6% Ro would be expected for the Base
Cretaceous in the Oseberg area. In well 30/6-1, this
3100
FM. .. would suggest a generation threshold around
2200-2300m.
In view of the problems in assessing maturity varia-
TRUE VITRINITE
tions in 30/6-1 by conventional parameters, a number
.& VITRINITE FROM REWORKED MATERIAL
of biomarker molecular variations, based upon regular
Fig. 9. Vitrinite reflectance versus depth in well 30(6-1 steranes and hopane-type triterpanes were investi-
(sidewall samples). gated (Mackenzie et al., 1980; Mackenzie and Max-
well, 1981). 8terane transitions reflecting increasing
against depth is shown in Fig. 9. Even after allowing maturity are shown in Fig. 11. The transition b ~ c
for allochthonous (reworked) components, the spread corresponds to
of data points was very erratic and showed only a C2~1l'1l'20R ~ C2~1l'1l'208,
general increase with depth. A least mean squares fit
of the vitrinite data was totally unsatisfactory for any and the parameter is expressed as
reliable indications of maturity, so this parameter was 100·1l'1l'1l'208
percentage 208 = 208
not used in well 30/6-1. As sidewall cores were used Il'Il'Il' + Il'Il'Il'20R .
30/6-1 THERMAL ALTERATION The transitions (b) + (c) ~ (d) + (c) are represented
INDEX (TAIl
as:
C2~1l'1l'20R + C2~1l'1l'208 ~ C2~PP20R + C2~PP208,
And the second parameter is expressed as
Il'PP(20R + 208)·100
perc en t age Il'PP = .
Il'PP(20R + 208) + 1l'1l'1l'(20R + 208)
As a result of resolution problems in determining
these ratios, it is usual to use the C Z9 transitions rather
than those of the C27 steranes. A third maturity para-

(b) 50, 14"O,17Q ,20R


FM.
(c) 50,140,170,208

(d) 5a.14~.17B,20R

(e) 5a.14~.17f3.20S
EQUIVALENT VITRINITE
REFLECT ANCE (%)
Fig. 11. Major transitions occurring in regular sterane
Fig. 10. TAl versus depth in well 30(6-1. molecules through increasing thermal stress.
192 Petroleum Geochemistry in Exploration of the Norwegian Shelf

VIlRINITE
RElFECTACE
where it can be seen that the sterane/triterpane trans-
COl
He GENERATION 205 STERAHES «M! STERANES 229 TRtTERPANES itions correspond to different maturation levels.
('lIo) ('!Io) ('lIi)
mg HC/g carbon
Biomarker distributions were obtained on extracts
from a range of sidewall cores and cuttings samples
from well 30/6-1, and the results are shown in Fig. 13.
Abnormally high values of the various parameters
occurred at shallow depths but, after careful examina-
0.5
57-60 tion of other analytical data from these samples, it was
concluded that the shallow samples, at 910-920 m and
1610-1620 m respectively, were contaminated with
LO
migrated mature hydrocarbons. The sidewall core
from 2023 m was also found to be contaminated with
2.0 oil from the drilling mud. After allowing for these
effects, the trends shown in Fig. 13 are considered
realistic. Both the 208 sterane and 228 triterpane vari-
Fig. 12. Sterane and triterpane variations related to oil
ations suggest a generation threshold around 2500 m,
generation and vitrinite reflectance.
but the a/3/3 sterane values appeared to be too high.
The most likely explanation for this effect is the co-
meter is based upon the relative amounts of the C32 a/3 elution of a C29/3aa20R component with the C~/3/3
22R and C32 a/3 228 triterpanes, and is calculated in the steranes, giving rise to apparently higher amounts of
same way as the percentage 208 for the steranes. Vari- a/3/3 steranes in low maturity sediments (Mackenzie et
ations of the three parameters with maturity are al., 1980).
shown in Fig. 12 (Mackenzie and Maxwell, 1981) Biomarker measurements in well 30/6-1 give

30/6 -I

UTSIRA
800 FM.

1000

1200

1400

1500

1600

1800

2000

2200

2400

2600

2600

3000 ~4 '''' I I

3100

Fig. 13. Biomarker variations versus depth in well 30/6-1.


Organic Geochemistry of the Oseberg Field (I) 193

maturity indications which were later supported by when the oils are compared on an asphaltene-free
much better quality vitrinite reflectance and other basis they are quite similar in composition, although
geochemical data from adjacent wells in the Oseberg the heavy oil has a somewhat larger NSO content and
area. Biomarker variations, when combined with . a slightly decreased aromatic content.
pOSSIbly, '
other geochemical data, also give more definite indica- Further study of the occurrence of the heavy oil (or
tions of the presence of migrated hydrocarbons and 'tar mat') was undertaken by extraction of residual oil
contamination in well 30/6-1. from reservoir core samples taken in wells which
penetrated the OWC. Asphaltene contents of these
OSEBERG CRUDE OILS residual oils were determined, and the results from
one of these wells are shown in Fig. 14. In the upper
Some properties of a sample of crude oil produced part of the oil column, residual oil asphaltene contents
from the Alpha structure are given in Table 1. This oil are approximately 5 wt%, increasing very rapidly
is representative of the bulk of the crude present in near 2730 m to approximately 50 wt%. Although not
the Oseberg Field and is typical of the highly paraf- shown in Fig. 14, the bottom of the heavy oil zone in
finic low wax crudes, occurring in the Viking Graben/
J
this well is shown by a very rapid decrease in the
,East Shetlands Basin area, which are considered to amount of residual oil present at approximately
be sourced from Upper Jurassic shales (Barnard and 2743 m. These observations indicate that the tar mat
Cooper, 1984). Asphaltene, sulphur and trace metal is approximately 13 m thick at this location, and that
contents are all low , resulting in a high-quality crude traces of heavy oil are present below the OWC.
oil. .Tar mats. are uSll:ally forme? as a result ofbiodegrad-
The stable carbon isotope ratio of Oseberg oil atIo~ ~sso:Iated WIth extensIve water washing, or by
(OI3C = -28.5%0) is slightly heavier than the range of preCIpItatIOn of an asphaltic 'sludge' which concen-
-29.5 to -28.5%0 reported for oils in the East Shet- trates near the OWC as a result of gravity segrega-
lands basin area, but falls within the range of -27.5 to tion. Destabilization of the reservoir crude can be
-31.0 quoted for Viking Graben oils (Barnard and brought about by the generation of light hydrocarbons
Cooper, 1984). A pristane/phytane ratio of 1.6 is also as a result of oil maturation or, more probably of, a
consistent with values reported for North Sea crudes later stage addition of gas to the reservoir.
(Cornford et al., 1983). The Oseberg oils examined in this study show no
In th~ s~uthern part of the Oseberg Field, a heavy, evidence of biodegradation, as n-alkane and
asphaltIC 011 has been encountered near the oil-water biomarker distributions are very similar in all of the
contact (OWC). Some properties of the residue residual oil samples. As noted previously in Table 2,
::O:200°C from this oil are given in Table 2, together the asphaltene-free oils from the main reservoir crude
WIth the properties of a similar residue from the main and the tar mat are quite similar, although the higher
produ~ing interval in the same well. The most signifi- N~O content of the latter is probably due to incorpor-
cant dIfference between these oil residues is the much a.tIOn of these co~ponents in the precipitated asphal-
higher content of asphaltenes (7 compared with tIC sludge. The slIghtly lower aromatic content of the
1.4 wt%) in the sample from near the OWC. However, tar mat may also be the result of water washing pro-
cesses. It seems likely, therefore, that the heavy oil,
or tar mat, observed in the Oseberg Field is the result
TABLE 1. Properties of an Oseberg oil of in situ de-asphaltening of the crude, possibly
Specific gravity 0,8506 associated with slight water washing, rather than
Density (,API) 34,9 biodegradation.
Sulphur content (wt %) 0,29
Asphaltene content (wt %) 0,12
Wax content (,'It %) 1.9
Vanadium (ppm)
Nickel (ppm)
1
1 10 20
..
ASPHAL TENES

30 40 50 60
al3 C (%,) -28,5'
Pristane/ phytane 1,6
2690
n-Alkane CPI LO

*NB~ 22 = -29,8%"

TABLE 2. Compositional variations in two


crude oil samples from one of the Oseberg
wells where a tar mat is encountered 2710

DEPTH
(m)
Production crudes (residues >200 °C)
Asphaltenes (wt %) 1.4 2720
HPLC chromatographic fractions (wt %)
SAC 50,1 (50,8)a
AROM 20,8 (21.1)a
NSO 27,7 (28,I)a 2730

SAC/AROM ratio 2.4


Heavy oil (residue >200 °C) 7.1
Asphaltenes (wt %)
HPLC chromatographic fractions (wt %)
2740
..
SAC 45,6 (49.1)a
AROM 17.3 (18.6)a 2750
NSO 30.0 (32.3)a
SAC/AROM ratio 2.6 Fig. 1~. Asphaltene content in residual oils extracted from
cores In one of the Oseberg wells where a tar mat has
aCorrected for asphaltene content, formed.
194 Petroleum Geochemistry in Exploration of the Norwegian Shelf

BIOMARKER VARIATIONS IN OSEBERG SEDIMENTS AND CRUDE OILS

K"-'ERIDGE CLAY FM
STERANES
KIMMERIDGE CLAY FM
TRITERPANES

HEATHER FM
TRlTERPANES

OSEBERG ALPHA OL OSEBERG ALPHA OIL


TRlTERPANES

j
OSEBERG ALPHA NORTH OIL
STERANES

OSEBERG ALPHA NORTH OIL

~
17. (H).18. (H).21 ~ (H)-28.30-BISNORHOPANE

Fig. 15. Sterane and triterpane variations in two Oseberg oils (Alpha and Alpha North) and two sediment extracts (Kim-
meridge Clay and Heather Formations, respectively).

OIL-SOURCE CORRELATIONS bisnorhopane is present in the Heather Formation,


but absent from the Kimmeridge Clay Formation,
Biomarker variations have been examined in crudes whilst elsewhere in the North Sea region, this com-
from all of the oil producing wells in the Oseberg pound is observed in the Kimmeridge Clay Formation
Field. Triterpane and sterane distributions of crudes (Mackenzie et al., 1984). It can also be seen from
from the Alpha and Alpha North structures are shown Fig. 15 that large maturity differences exist between
in Fig. 15. All of the patterns are very similar, the on-structure sediment extracts and Oseberg crude
although some small maturity differences are oils, for example:
observed in the sterane distributions. Oseberg oils
contain the rather unusual triterpane C28 Osberg Alpha oil:
17a(H),18a(H),21P(H),28.30 bisnorhopane, which is 20S = 54%; app = 59%; 22S = 60%;
observed in many North Sea Oils which are generally Kimmeridge Clay Formation:
considered to be derived from the Kimmeridge Clay 20S = 24%; app = 28%; 228 = 46%.
Formation (Mackenzie et al., 1984). The ratio of this
compound relative to the C29 norhopane is 0.5, which As the Kimmeridge Clay and Heather Formations are
is again consistent with values observed in Viking not much more deeply buried in the eastern sub-basin,
Graben oils (Mackenzie et al., 1984). this region cannot be considered a major source area
Sterane and triterpane variations in extracts from for Oseberg oil, which probably originates from the
the Kimmeridge Clay and Heather Formations in the much more mature Humber Group shales located to
Oseberg area are also included in Fig. 15. The C28- the west and north-west of the Oseberg area.
Organic Geochemistry of the Oseberg Field (l) 195

CONCLUSIONS REFERENCES
(1) Source rock studies suggest that shales from the Barnard, P. C. and Cooper, B. S. 1984. Source rocks and oils of the
Central and ~orthern North Sea. In: Demaison, G. and Murris,
Humber Group are the most likely source of the oil R. J. (eds.), PetroleulII Geochemistry alld Basin Evolutioli.
found in the Oseberg structure. The richest oil-prone AAPG .l1ellloir, 35,303-314.
source rock within the Humber Group is the upper Barnard, P. C., Collins, A. G. and Cooper, B. S. 1981. Identification
section of the Kimmeridge Clay Formation. and distribution of kerogen facies in a source rock hori-
(2) Conventional parameters, e.g. vitrinite reflec- zon-examples from the ~orth Sea Basin. In: Brooks, J. (ed.),
Organic Matumtion Studies alld Fossil Fuel Explumtion.
tance and TAl, were unsatisfactory for accurately Academic Press, pp. 271-282.
assessing maturity in well 30/6-1. Molecular par- Cornford, C. 1984. Source rocks and hydrocarbons of the North
ameters based on percentage 20S steranes and 22S Sea. In: Glennie, K. W. (ed.), Introduction to the Petroleum
triterpanes are in good agreement and suggest a Geolo,qy of the North Sea. Blackwell Scientific, Oxford,
pp. 171-204.
maturity level in this well which was confirmed by Cornford, C., Morrow, J. A., Turrington, A., Miles, J. A. and
later results in adjacent wells. Brooks, J. 1983. Some geological controls on oil composition in the
(3) Biomarker studies also give more definitive indi- U.K. North Sea. In: Brooks, J. (ed.), Pe/roleulII Geochemish'Y
cations of migrated hydrocarbons and contamination and Exploration of Europe. Blackwell Scientific, Oxford,
in sediment extracts from well 30/6-1. pp. 175-194.
Deegan, C. E. and Scull, B. J. (Compilers) 1977. A standard lithos-
(4) Oseberg crude oil is typical of the highly paraf- tratigraphic nomenclature for the Central and Northern North
finic, high quality North Sea crudes with a gravity of Sea. Rep. Insl. Geul. Sci. No. 77/25; Bull. Norweqial1 Petrol.
35° API and low sulphur (0.29 wt%), asphaltene Direc/omte, No.1. HMSO.
(0.12 wt%) and trace metal contents (V = 1 ppm, Katz, B. J. 1983. Limitations of 'Rock EvaI' pyrolysis for typing
organic matter. Orqallic Geochelil., 4, 195-199.
Ni = 1 ppm). Kirk, R. H.1980. Statfjord Field, a North Sea giant. In: Halbouty
(5) A tar mat has been established in part of the M. T. (ed.), Giallt OU and Gas Fields afthe Decade 1968-1978.
Oseberg Field. This heavy oil is considered to result AAPG .tiel//Oi)', 30, 95-116.
from in situ de-asphaltening of the crude followed Mackenzie, A. S. and Maxwell, J. R. 1981. Assessment of thermal
by gravity segregation of the precipitated sludge maturation in sedimentary rocks by molecular measurements. In:
Brooks, J. (ed.), Organic JIaturatioli Stwlies and Fossil Fuel
and concentration near the oil-water contact. Some Exploratioll. Academic Press, pp. 239-254.
minor water washing of the heavy oil may also have Mackenzie, A. S., Patience, R. L., Maxwell, J. R., Vandenbrouche,
occurred. M. and Durand, B. 1980. Molecular parameters of maturation in
(6) Rocks of the Kimmeridge Clay and Heather the Toarcian shales, Paris Basin, France 1. Changes in the con-
figuration of the acyclic isoprenoid alkanes, steranes and triter-
Formations are not considered mature enough in the panes. Geochilll. Coslllochilll. Acta, 44, 1709-1721.
Oseberg area to have generated the reservoir crude Mackenzie, A. S., Maxwell, J. R., Coleman, M. L. and Deegan, C.
oil. The major source area is probably the more E. 1984. Biological marker and isotope studies of North Sea crude
mature Humber Group shales located to the west and oils and sediments. Proc. 11th World PetmleUlI? Conqress 1.98:1,
north-west of the Oseberg area. Vol. 2. Heyden, pp. 45-56.
Van den Bark, E. and Thomas, O. D. 1980. Ekofisk: first of the
giant oil fields in Western Europe. In: Halbouty, M. T. (ed.),
Giant Oil and Gas Fields of the Decade 1968-1978. AAPG
.l1emoir, 30, 195-224.
16 Block 33/6 geochemical
evaluation
B. Mamgna, * G. Zaro* and P. Pessinat
*Agip S.p.A., 20097 S. Donato Milanese, Italy
tNorsk Agip a.s., Sandnes, Norway

A geological evaluation of Block 33/6 was performed using the direct knowledge of subcrops and subsurface
conditions provided by the 33/6-1 wildcat well and by modelling the evolution of maturity in the nearby areas.
Laboratory analyses (including optical typing of kerogen, TAl and vitrinite reflectance, Rock-Eval pyrolysis and
stable isotopes) permit a definition of source rock characteristics and measurement of the present degree of
thermal maturity.
The evolution and areal extent of maturity for Humber Group rocks, generally considered as the main source
sequence in the North Sea, were studied applying the Lopatin method and the Tissot-Espitalie method.

INTRODUCTION area appears to be most attractive for petroleum


exploration because it includes many important dis-
A production licence for petroleum exploration in coveries; the Statfjord field lies just inside the Nor-
Block 33/6 covering an area of 485.5 km 2 was granted wegian border, Gullfaks and several smaller fields lie
by Royal decree on 23 December 1977 to a group con- in the Norwegian sector and many others are found in
sisting of Statoi! (50%), N orsk Agip (30%, and UK waters (Brent, Hutton, Thistle, Magnus, Murchi-
operator) and Deminex (20%). The licence expired on son, Dunlin) (see Fig. 2). All of these fields lie in the
1 January 1984 when the block was released. East Shetland Basin, which is covered in Norwegian
An exploratory well, 33/6-1, was drilled at waters by quadrants 33 and part of 34.
01°51'53"E, 61°32'14.58"N. The selected location was The tectonic evolution of the basin began in the
on the crestal area of a tilted fault block lying in a Permo-Triassic when a system of Caledonian
water depth of 306 m. The effective closure was con- NE-SW-trending faults was reactivated. Tectonic
sidered dependent upon the sealing properties of a activity continued throughout Jurassic times, charac-
NE-SW fault immediately southward of the prospect. terized by mainly eastward block faulting and syn-
Target of the well was to test well documented depositional westward rotation and subsidence of
Middle-Lower Jurassic reservoirs (Brent Group- down thrown blocks. During 'Late Kimmerian' move-
Statfjord Formation) and to reach Upper Triassic ments, the tectonic activity increased and caused the
sediments (Cormorant Formation) at total depth. The partial erosion of some previously uplifted Early to
well was entirely water-bearing and was conse- Middle Kimmerian fault blocks.
quentlyabandoned. No other wells were drilled in the During the Cretaceous and Paleocene only minor
block. tectonic activity is evident, with local reactivation of
The aim of the present study was to evaluate the Kimmerian faults and the onset of a broad general
petroleum potential of the block and, if possible, to subsidence.
discover whether geochemical and/or geological fac- The high well density in the North Sea has made
tors can explain the observation that no hydrocarbons possible exhaustive characterization of source rock
were trapped. occurrence in the area. The Kimmeridge Clay Forma-
tion seems to be the major source unit with a strongly
oil-prone organic type, while the underlying Heather
Formation has inferior characteristics, and is gener-
REGIONAL GEOLOGICAL FRAMEWORK ally regarded as mainly gas-prone (Goff, 1983).·
Several studies of the distribution of organic matter
Block 33/6 lies in the northern part of the North Sea, type have shown variations in composition which may
close to the UK-Norwegian border, which cuts the be attributed to the relative distance of the paleoshore
southwestern corner of the block (see Fig. 1). This (Cooper and Barnard, 1984).
PetroleulII Geochemistry in E:rplomtion oflhe Norwegian Sheif' © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 197-204.
198 Petroleum Geochemistry in Exploration of the Norwegian Shelf

33/2 I ;]j/6 ! 34/4- ir=l"


"
~ DEPTH ~~ TOTAL ORGANIC CARBON KEROGEN
~ 0
(m) oq.~'f> t.
COMPOSITION


N

~.
\ \
\
GROUP

, 3200

-\
61-45'N
\j1/5V jj/6 ;]4/4

, \
\
2/1/8
211//j

.5 ~ ~ \
~'
~ \ \

+' 3316- t

\
0

;-j4" 211/14 \
-<>-
",' 0'
61-30'M
211/18 -0 u 2/1//9 \j1/9 -¢;, 34/1
, " .It
-I)'

:.\.. ~
!¢- \
l' -¢" 0'
"1
.
."" .,
<);,
' ,\

.' .'.
C3AMOFlPHOUSO!.!

<i
,. 4~,
,.0

,
,
211/24
'
... I [:=J
~
IIARIt>F PHVTOPI

TERRESTRIAL HERBACE:)

~ llAR[STRI~L WOOD
AN~ TCJ'I
~ JEB~IS leAf

'R~GME'll'; ~USIN'~f
(,\JTIClE

DE3RIS
~PO~lS FGLLEr,

2/1/24

Fig. 1. Location map. Fig. 3. Well 33/6-1: TOe and kerogen composition.

Jurassic Brent Group should be excluded as a signific-


ant source rock for oil. This is due to the poor quality
of the organic matter, chiefly consisting of continental
herbaceous fragments, woody inertinitic fragments
and fusinite debris, which could, however, be a fair
ilource for gas. Such an inference is in accordance with
other data from the area reported by Goff (1983),
Barnard and Cooper (1981) and Cooper and Barnard
(1984).
A predominance of terrestrial organic matter per-
sists in Humber Group rocks, although a small amount
of marine phytoplankton and local higher concentra-
_FIELDS c:=J STUDY AREA
tions of amorphous organic matter are present.
Rock-Eval pyrolysis results are summarized in
Fig. 2. Structural sketch of Northern North Sea (redrafted Table 2. These are in agreement with the optical data,
from Skarpnes et al., 1982).
confirming a mixed quality of kerogen, largely tending
towards type III. Humber Group formations display
EXPERIMENTAL DATA FROM inferior source features (see Fig. 3) to those reported
WELL 33/6-1 by Goff for the North Sea generally, and for the East
Shetland Basin in particular.
Samples (cuttings, sidewall samples and bottom cores) The actual degree of maturity of organic matter was
from well 33/6-1, from the Lower Miocene to the determined on the basis of both thermal alteration
Upper Triassic, were analysed to measure the degree index (TAl) and mean vitrinite reflectance (see
of maturity and to evaluate source potential. Total Fig. 4). These parameters suggest early generation in
organic carbon contents of some interest were the well at a depth of 2300 m (Ro % > 0.5) and full
detected from 3536 m to 3647 m, where values rang- maturityonset(R o% > 0.55) at 2900 m. Consequently,
ing from 1% to 5% were recorded (see Fig. 3). This the Humber Group and Brent Group are considered to
section corresponds to the Middle-Upper Jurassic be mature in the study well.
Brent Group, Heather Formation and Kimmeridge Qualitative analyses on trace gas (head space and
Clay Formation. The Brent Group is characterized by carbon isotopic composition of methane) agree with
a prevailing sandy lithology. Log response suggests the evidence obtained by optical methods. These ana-
that the Rannoch Formation is present, while a very lyses were carried out on cutting samples collected at
thin, shaly Broom Formation can only be hypothe- the well site in glass jars containing a small amount of
sized. On the contrary, both the Heather and Kim- bactericide to prevent any microbial activity. Samples
meridge Clay Formations consist of mainly claystone were picked from the entire sedimentary section
with local stringers of siltstone. penetrated, and the relevant data are plotted in
An optical study of kerogen quality, as summarized Fig. 5. Results indicate several different genetic
in Table 1 and Fig. 3, demonstrates that the Middle origins for the analysed gas:
Block 33/6 Geochemical Evaluation 199

TABLE 1. Kerogen composition (all average values)

Continental Continental
Amorphous Marine herbaceous woody fragments
organic matter phytoplankton fragments and fusinite
Unit %) (%) (%) debris (%)

Kimmeridge Clay Fonnation 19 6 22 53


Heather FonnatlOn 10 10 40 40
Brent Group 10 50 40

TABLE 2. Rock-Eval pyrolysis (average values)


Unit Samples TOC (%) S2 (%,) HI Sl (%0)

Kimmeridge Clay Fonnation Cuttings 1.89 2.83 150 0.48


Heather Formation Cuttings 2.87 6.27 218 0.87
Brent Group Bottom cores 1.47 1.47 100 0.18

1. biogenic methane, distinguished by a very low


AGE
o~·

s.J:
~

Q4}.... TAl
VITRINITE REFLECTANCE
13C content, down to approximately 1500 m;
Ro '"
03 05 0.7 10 2. thermogenic gas, with higher 13C content and
(m)
significant presence of C2 +, from approximately
~~
> - 2000 m to 3900 m T.D.;
500
3. mixed gas, with intermediate isotopic and chemi-
II:
~ ~ ---
~I:::
I- ~
---
._ _ cal properties, from 1500 m to 2000 m approximately.
:i --- 1000 r--:--
-r== r-L-

,
1';'1.5
a:~--- r--
~ ~ -_- 1500 THERMAL MODELLING
~~ =-= 1.5
\
-~,ooor---
1.5.,.2 \ The establishment of a model for hydrocarbon genera-
tion requires a definition of temperature distribution
(II _-_

:~ .
~ "="
-_- 250
--
~ ~
and of its evolution through time.
The geothermal gradient in well 33/6-1 was calcu-

\
.-:. 3000
lated from log-derived bottom hole temperatures, cor-
rected according to the method described by Fertl and
-;c=:::
...i ~ _-_ Wichmann (1977) to minimize the cooling influence of
mud circulation. A value of 3.2 °C/lOO m was thus
.\ .
~~350
~ ~-:-~
~~ T2.5
determined.
T D 3900 To draw a mean temperature gradient map of
regional significance, available data from 24 wells on
Fig. 4. Well 33/6-1: maturity profile. the Tampen Spur, in the East Shetland Basin and on
the East Shetland Platform, were taken and com-
pared with other published data (Carstens and
Finstad, 1981), see Fig. 6.
~ t. Cl% CZ'% 13 C / 1Z c Gradient values range from 2.5 to 4°C/100 m; their
c( ~E 50 50 - 65 - 56 - 47 - 38
distribution appears to be strongly affected by the tec-
tonic setting of the basin. Lower gradients occur with-
>
700
in the Viking Graben and in structurally depressed
II:
areas, while maximum values are predominantly
-....
0(

associated with horsts and with the rims of structural


c:t:
w highs. A similar relation exists in block 33/6, where
.... maxima were generally recorded along the margin of
structural highs in a transitional zone towards the
northerly M0re-Tr0ndelag Basin.
(/)
A comparable regional pattern was previously
::J
0 reported by Carstens and Finstad (1981), who
w
U reviewed the geothermal gradients of the Northern
«
....
w North Sea. They suggested that subsurface fluid flow
II:
U
2800
along major fault planes could be responsible for the
;:; 3100 observed situation. This explanation would also
account for the predominance of high gradients in the
"
"~ Tampen Spur and M0re-Tr0ndelag areas.
. ~

Reconstruction of the palaeo-goethermal regime is


i obviously more difficult. Many authors have dealt
with the problem of reconstructing the past thermal
Fig. 5. Well 33/6-1: head space analysis and isotopic ratio history of the Northern North Sea, either on the basis
of methane. of experimental data, e.g. vitrinite reflectance (Yar-
200 Petroleum Geochemistry in Exploration of the Norwegian Shelf

R.
%
10

07

01h~~'--'-rrr"-~o-T-,--,--~r-,---rrr-r--rTTl-'--;"""" TTl
1 57102030 5070100 200

Fig. 7. Time-temperature index (TTl) versus vitrine


reflectance (Ro%).
and the measured vitrinite reflectance of several wells
from the region was examined to determine a general
trend for the basin, which would then allow the
interpretation of other computed maturity data. The
calibration curve is shown in Fig. 7.
Using the Tissot and Espitalie (1975) method, com-
putations with kinetic parameters for both type II
27.5 .;. 32.S"C/Km
and type III kerogen were performed. The values for
genetic potential proposed by Tissot and Espitalie
Fig. 6. Mean geothermal gradients in Northern North Sea.
were modified on the basis of actual Rock-Eval
pyrolysis results (see Table 2). The experimental
zab et al., 1983), or based on the sedimentary and observations are lower than the values proposed by
structural evolution of the basin. Tissot and Espitalie for type III kerogen.
A heating event in the past must be hypothesized if The computed transformation ratios, when com-
the stretching model of McKenzie (1978) for the for- pared with the production indices from Rock-Eval,
mation of the North Sea is accepted. In this case the suggest that the difference cannot be entirely ascribed
age of the event can be determined by the occurrence to primary migration. Hence, an experimental value
of tensional movements, as expressed in periods of of 220 mg/g of TOC' was used, bearing in mind that
normal faulting. According to Goff (1983), some dif- an underestimation could also occur. Type III kerogen
ficulties may arise in such analysis due to the presence kinetic parameters are believed to model hydrocarbon
of several phases of faulting between the Late Batho- genesis most accurately for organic matter from the
nian and Early Cretaceous. In effect, several authors Humber Group in Block 33/6. Transformation ratios
have proposed different ages for this heating event, calculated according to this model were therefore cho-
yet all accept that the heat flow has not changed in any sen to construct maps showing the evolution of matur-
systematic way during the last 100 million years. ity in the area.
Before that time, it seems unlikely that the Upper As a further step, a method which assumes a sim-
Jurassic shales in the area (i.e. Heather Formation plified scheme for kerogen thermal degradation, but
plUS Kimmeridge Clay Formation), were buried suf- employs experimentally computed kinetic parameters,
ficiently deeply to reach levels of temperature influenc- was applied. This simplified scheme involves only
ing hydrocarbon genesis. Therefore any assumption one reaction for oil genesis, instead of the set of six
about the previous heat flow should not greatly affect reactions utilized by the model of Tissot and Espitalie.
the maturity level of the Upper Jurassic shales. The results of Rock-Eval pyrolysis, considered as a
Moreover, accepting the theory that the thermal non-isothermal kinetic experiment, were used to cal-
effects induced by stretching rapidly decline after the culate a pseUdo-activation energy and the respective
event, it seems reasonable to apply a constant heat frequency factor for the hydrocarbon generating reac-
flow when calculating the maturity of the Upper tions. AGIP software was used, based on the method
Jurassic shales in the block. developed by Juntgen and co-workers (Juntgen and
An attempt was made to reconstruct the tempera- Van Heek, 1968; Hanbaba et al., 1968; Hanbaba and
ture history of the sedimentary section, taking into Juntgen, 1968). Using the values obtained in this way,
account thermal conductivity variations due to com- the hydrocarbon genesis model for the well gave
paction. However, a test performed appl~ng therm~1 underestimated results. Nevertheless, curves of the
conductivity and porosity data from the lIterature dId hydrocarbon generation rate, plotted on the basis of
not give satisfactory results for the study well. Con- the same kinetic parameters, fit in with experimental
sequently, this aspect had to be ignored, resulting in a curves quite well (see Fig. 8), thus confirming the
further simplification in the model. reliability of the method. We believe that the explan-
ation lies in the level of maturity of the Humber
Group rocks. In other words, the mature organic mat-
KINETIC MODEL ter has been depleted of mass fractions which are
characterized by low activation energies and react at
To study the maturation history and the hydrocarbon lower temperatures; hence, the computed mean acti-
generation in the study area two models were applied: vation energy, which reflects the present situation,
the Lopatin method and the Tissot-Espitalie method. turns out to be higher than the one at the beginning of
The Lopatin method (Waples, 1980) was chosen as a
first step in the study of the evolution ofmat~rity., The
relationship between the calculated maturIty mdex *Averaged value for the Humber Group formations.
Block 33/6 Geochemical Evaluation 201

E att =56.3 Kcal/mole

Ko = .25E .. ,8 sec· 1

700 800

E att == 57.2 KcaL/mole

,.A,
Ko:::: .49E+18 sec· 1

T (OK) --J 600 700 800

EXPERIMENT Al PEAK
CALCULATED PEAK

~ERDSIDNORNOOEPDSITION
Fig. 8. Well 33/6-1: thermal decomposition of Heather
Formation-comparison between experimental and calcu- F::<:d 0 - 100 m

lated curve. wm IDD-250m

maturation, and cannot be used in modelling the com- Ifff11 > 25D m

plete genetic process. _FAUlTS

Fig. 9. Block 33/6: isopach map of Upper Jurassic shales.


MATURATION AND TIMING OF
GENERATION
The evolution of maturity and its distribution over
The reconstruction of the burial history of the source the area, reconstructed on the basis of the two differ-
rocks in the study area using seismic data was a ent methods, were studied separately and compared
further step towards the modelling of maturation and (see Figs. 10-13).
hydrocarbon generation. The transformation ratio values computed for reac-
The quality of the seismic data forced two simplifi- tions involving type III kerogen were mapped. The
cations: firstly, the Kimmeridge Clay Formation and composition of this type of kerogen, in fact, parallels
the Heather Formation had to be studied together, that of vitrinite; it therefore seems reasonable to con-
and are therefore referred to as 'Upper Jurassic sider the computed transformation ratio and the TTl
shales' henceforth, as it is difficult to define the boun- values calibrated against vitrinite reflectance as prop-
dary between one formation and the other; secondly, erties of the same type of organic matter, and the two
no semi-quantitative evaluations could be carried out can easily be compared (Yiikler and Kokesh, 1984).
on the Brent Group since its base could not be recog- The results achieved applying the Lopatin method
nized. In addition, the thickness and regional distribu- show that the Upper Jurassic shales entered the
tion of the source rocks were deduced from geophysi- mature facies (R 0% equivalent >0.55) during the Late
cal data. Cretaceous-Early Paleocene. By the end of the
On this basis, areas of non-deposition or total ero- Paleocene only the base of the unit had reached matur-
sion of the Upper Jurassic shales and, in turn, sites of ity over a significant area, in a belt crossing the block
maximum accumulation were defined (see Fig. 9). from SW' to NE. This distribution was clearly con-
Their distribution is characterized by a marked irregu- trolled by the location of the main palaeo-highs
larity, which can be explained in terms of the struc- (see Fig. 10).
tural evolution of the licence block. Extensional fault- The isolated mature zones existing at the top of the
ing, tilting and a Late Kimmerian erosional phase, in Upper Jurassic shales seem to be connected to areas of
fact, promoted thickening of sediments deposited at relatively higher subsidence; the irregular tectonics
the base of fault escarpments. related to block faulting can account for their patchy
In order to describe the maturation of the Upper distribution.
Jurassic shales present in Block 33/6, several maps During the Paleocene-Oligocene interval maturity
which are relevant to the conditions at the top and the extended to the whole unit, the only exception heing
base of this unit at different times, were constructed an immature area to the southeast, which still exists,
by means of a grid of points, chosen to represent the though greatly reduced, and in central parts of the
various structural and stratigraphic features in the block the optimum maturity level (Ro% equivalent
area. -0.85) was reached (see Fig. 11). The whole central
Both R 0% equivalent values, calculated using the section is now at optimum maturity levels and even
Ro%-TTI relation for the basin (see Fig. 7), and the values close to the overmature facies have been
transformation ratio computed applying the method of recorded (see Fig. 12).
Tissot and Espitalie, were mapped. A reconstruction of the evolution of maturity based
202 Petroleum Geochemistry in Exploration of the Norwegian Shelf

CJ ~:~~I:; :~tl~::!:II.u ~ ~:~;~I~~~I~~~~~rllu


[==:J I "" ~' l U [III [J[] lallil n l II n~

Fig. 10. Block 33/6: maturity maps of Upper Jurassic shales Fig. 12. Block 33/6: maturity maps of Upper Jurassic shales
at the end of the Paleocene. at the present time.

• • !!> i" I'tIlI,l nutS


on the transformation ratio values (Tissot-Espitalie
method) leads to a pattern substantially similar to the
one shown by the vitrinite reflectance equivalent
values (Lopatin method), see Fig. 10-13.
As a next step, the two sets of results were com-
pared in order to find out just how they would corres-
pond. It was found that a transformation ratio of 0.1
roughly corresponds to the onset of full maturity
taken as 0.55 Ro% equivalent, while the optimum
maturity level threshold (0.85 R 0% equivalent) falls
between transformation values of 0.4 and 0.5.
A review of these figures compared with the rela-
tionship between vitrinite reflectance and the trans-
formation ratio for type III kerogen proposed by
~ ~~$~):; :~§\~::S:l'j(M
D l oIU 1utl 1i CIII
Yiikler and Kokesh (1984), on the basis of data from
Tissot and Espitalie (1975), allows the following con-
clusions:
(1) at the early stages of generation (low transfor-
mation ratio values) the present findings correspond
to those published by these authors; and
(2) at higher degrees of maturity (i.e. higher values
of transformation ratio), a discrepancy is evident. In
fact, much lower corresp<mding vitrinite reflectance
equivalents were obtained than would have been
obtained following the curve published by Yiikler and
Kokesh (1984).
In the opinion of the authors this behaviour can be
explained by the fact that the vitrinite reflectance val-
ues used to deduce the R o/TTI calibration are pre-
dominantly lower than 0.85; thus all the TTl values
Fig. 11. Block 33/6: maturity maps of Upper Jurassic shales corresponding to a higher vitrinite reflectance have
at the end of the Oligocene. been interpreted according to an extrapolated
Block 33/6 Geochemical Evaluation 203

t3l11 A'

~P~.~ES~EN~T~T~IM~E~ ____~FL________________~Om
se~ bottom

• lEG E K 0 r - - - - - ----1---------~ 1000

A_ Top I!rlliHJ

B _ Top Oligocene 2000


C _ fop Paleocene

0_ Top CretueOtl5

[_ Top Turonian 3ODD

G _ late llmmfll3n Uncontormity


H _ Base Upper JumSlt Shale 4000

-.-.- R~ eQ 0 55'~

- - - Ro eQ 085'1,

b:QUi MIlUR[ SOURCE ROCK


rT~.P~Ol~16~O~CE~N~E ______________________~Om

1000

holllonl~l scale

10
1
km
2000

33/1 JI/4 3000

~3/H TOP PALEOCENE

\ ¢-, ~~~==========r0m
~A'

\ r - - - - - --------/"".'\-~t 1000

\ 33/9 31 / I
2000

~-"----------- __________________..J 3000

Fig. 13. Regional cross-section through well 33/6-1.

curve. Therefore, an underestimation of the Ro% GENERATED HYDROCARBONS


equivalent is possible at high maturity levels.
Transformation ratios seem to be a more reliable It is possible to calculate the amount of hydrocarbons
means of investigating overmature situations. Zones generated (HG) along a vertical profile at any given
of wet gas and condensate generation are to be point of the basin by applying the method of Tissot and
expected where these values are higher than 0.6, as EspitaJie. This is achieved by integrating the amount
they are today in the central part of the block. The of hydrocarbons generated at each level (he) over the
base of the unit has peak values of O. 7 within a basinal thickness of the source rock using the following equa-
area to the north and, locally, in areas of thicker tion:
sedimentary accumulation at the base of fault
escarpments. The values of transformation ratio
suggest that significant ~as generation did not occur in
HG =
fZ2
ZI
he(z) dz,

the past. where ZI and Z2 are, respectively, the top and base of
204 Petroleum Geoche' tistry in Exploration of the Norwegian Shelf

levels permitting the generation of oil and, in places,


gas.
The more favourable areas of the block for hydro-
carbon generation are the central and northern
I parts, based on a partial, semi-quantitative evalua-
tion, as well as from the study of the evolution of
I maturity with time. Consequently, the relatively lean
hydrocarbon generation in the area close to the struc-
ture may be one of the reasons why well 33/6-1 is a
dry-hole.
• (it However, the authors believe that other geological
I factors, such as a lack of sealing properties in the
• \ faults defining the structure tested by 33/6-1 (see
\ Fig. 13), may have played in this case a more signifi-
• • cant role .
c:::J .ounce 1>00< c::J SOURCE AOO<
~NO\.PP'E JI.A.RAS:S I r,;.~" L E ~ IiO UPf'E.R Jl.IAo't!;SlC ·SHALE

ACKNOWLEDGEMENTS
Fig. 14. He distribution in Block 33/6.
We thank Agip and Norsk Agip for permission to pub-
the source rock. Integrating the HC values areally it lish this paper. We wish to thank our colleagues in
is possible to estimate the hydrocarbons generated in Agip S.p.A. and Norsk Agip for the co-operation and
the basin. the assistance provided. The optical and geochemical
In the present case, due to insufficient seismic data analyses were carried out in Agip Geological
(depth converted control points) covering the whole Laboratories, Milan. We are also indebted to L.
area, and a lack of areal distribution data on richness Novelli, on whose experience we have drawn heavily.
and quality of organic matter, this type of evaluation
is not possible. It was therefore necessary to restrict
evaluation to the calculation of HC values in different REFERENCES
parts of the block. The outcome is summarized graphi-
cally for kerogens type II and III in Fig. 14. The Barnard, P. C., and Cooper, B. S. 1981. Oils and source rocks of the
height of the columns is proportional to the HC value North Sea. In: IIling, L. V., and Hobson, G. D. (eds.), Petroleum
Geology of the Continental Shelf of North-West Europe. Heyden,
in each point. No distinction has been made between London, 521 pp.
oil and gas owing to the purely indicative nature of Carstens, H. and Finstad, K. G. 1981. Geothermal gradients of the
this evaluation. The HC distribution seems to be con- Northern North Sea Basin 59'-62'N. In: IIling, 1. V. and Hob-
nected with both the structural setting of the block, son, G. D. (eds.), Petroleum Geology of the Continental Shelf of
North-West Europe. Heyden, London, 521 pp.
which influenced the evolution of maturity, and the Cooper, B. S., and Barnard, P. C. 1984. Source rocks and oil of the
thickness of the source rock. A concentration of high Central and Northern North Sea. AAPG Memoir, 35,303-314.
values in the north and northwest towards the Cooper, B. S., Coleman, S. H., Barnard, P. C. and Butterworth, J.
M~re- Tr~ndelag basin is evident, while low values are S. 1975. Paleotemperatures in the Northern North Sea Basin. In:
Woodland, A. W. (ed.), Petroleum and the Continental Shelf of
found in the south-southeast in structurally higher North West Europe, Vol. 1, Geology. Applied Science Publishers,
zones. Areas of locally high H C can be seen along belts Barking, 501 pp.
bordering on the Jurassic palaeo-highs, where the Fertl, W. H., and Wichmann, P. A. 1977. How to determine static
source rocks are thicker. BHT from well log data. World Oil, 184(1), 105-106.
By calculating the HC at a point near the well and Goff, J . C. 1983. Hydrocarbon generation and migration from
Jurassic source rocks in the East Shetland Basin and Viking Gra-
assuming constant thickness and maturity of the source ben ofthe Northern North Sea. J. Geol. Soc. Lond. 140,445-474.
rock, it is possible to determine that the amount of Hanbaba, P. and Jiintgen, H. 1968. Ziir Ubertragbarkeit von
hydrocarbons generated per square kilometre are 226 Laboratoriums-Untersuchungen auf geochemische Prozesse der
and 176 thousand tons for type II and type III Gasbildung ausSteinkohle und iiber den Einfluss von Sauerstoff
auf die Gasbildung. In: Ingerson, E. (ed.), Advances in Organic
kerogen, respectively. In the same way, at points Geochemistry. Pergamon Press, 617 pp.
where the HC is highest in the block, the amount of Hanbaba, P., Jiintgen, H. and Peters, W. 1968. Nicht isotherme
hydrocarbons generated per square kilometre are Reaktionskinetic der Kohlenpyrolyse. Brennstojf-Chemie, 49(12),
1924 and 1597 thousand tons, respectively, assuming 368-376.
Jiintgen, H. and van Heek, K. H. 1968. Gas release from coal as a
in these calculations a TOC content of 2.4% and a rock function of the rate of heating. Fuel, 48, 103-117.
density of 2.3 g/cm 3• McKenzie, D. 1978. Some remarks on the development of sedimen-
tary basins: Earth Planet. Sci. Lett., 40, 25-32.
Skarpnes, 0., Briseid, E. and Milton, D. 1982. 34/10 Delta prospect
CONCLUSIONS of the Norwegian North Sea: Exploration study of an unconform-
ity trap. In: Halbouty, M. T. (ed.), The Deliberate Search for
the Subtle Trap. AAPG Memair, 32, 207-216.
The geochemical analyses carried out on samples Tissot, E., and Espitalle, J. 1975. L'evolution thermique de la
taken from the well drilled in the study area have materiel organique des sediments: applications d'une simulation
shown that the Kimmeridge Clay Formation and mathematique. Inst. FranQais Petrole Rev_, 30, 74B-777.
Waples, D. W. 1980. Time and temperature in petroleum formation:
Heather Formation have, on the average, inferior application of Lopatin's method to petroleum exploration. AAPG
source potential compared with other areas of the Bull., 64, 916-926.
Northern North Sea, because of the quantity and, Yarzab, R. F., Lerche, I. and Kendall, C. 1983. Relationship of
above all, the quality of the organic matter. vitrinite reflectance to heat flow history of North Sea. AAPG
Bull., 67(3), 574.
A study of the evolution of thermal maturity has Yiikler, M. A. and Kokesh, F. 1984. A review of models used in
demonstrated that vast areas of the Upper Jurassic petroleum resource estimation and organic geochemistry. In:
shales reached maturity as far back as the Paleocene. Brooks, J ., and Welte, D. H. (eds.), Advances in Petroleum
At present, most of this unit has reached maturation Geochemistry, Vol. I. Academic Press, London, 344 pp.
17 Maturity studies of the Jurassic
'Coal Unit' in three wells from
the Haltenbanken area
J. L. Pittion * and J. Gouadain *
*Total-CFP 218/228, Avenue du Haut Leveque, 33605 Pessac Cedex,
France
tTotal Marine Norsk, P.O. Box 4317, N.5013 Nygardstangen, Norway

Source-rock analyses including Rock-Eval, elemental composition, maceral and extract analyses were carried out
on coal samples from the Lower Jurassic 'Coal Unit' ofthe Haltenbanken area. Maturation level was determined
using vitrinite reflectance, pyrolysis temperatures (T max) and fluorescence of exinite. The source-rock potential of
the coals appears to be fair (hydrogen indices between 100 and 350 mg/g) and similar to the oil and gas source-rock
coals in the Mahakam Delta, Indonesia. The maturation stage of the 'Coal Unit' directly relates to its present
depth, i.e. it depends upon the thickness of the post-Jurassic overburden. The oil window (0.6-1.3% R~ lies
between about 3500 m and 4800 m. Oil and gas are probably generated from both Upper Jurassic shales and the
Coal Unit in this depth interval, which is most commonly found in the transition zone between the Tnmdelag
Platform and the Ml'lre Basin. The same source rocks, where more deeply buried in the basin, have probably also
generated oil at some time in the past, but this oil would have been cracked to gas unless it has migrated out of the
cracking zone. From these maturation data, there is little or no chance of finding oil in reservoir sands below
5000 m.

INTRODUCTION with shales. Organic-rich Upper Jurassic shales (equiv-


alent to the Kimmeridge Clay Formation of the
The Mesozoic sediments of the northern Norwegian North Sea) overlie these sands. Above, Cretaceous
continental shelf contain a coaly sequence about 400 m and Tertiary formations are composed mainly of clay-
thick, Rhaetic to Hettangian in age, commonly called stones which are occasionally calcareous (see Fig. 2).
the 'Coal Unit'. This formation, where coal seams are Tectonic activity started at the end of the Triassic
interbedded with sandstones and shales, is most suit- but the main 'Kimmerian' phases occurred at the end
able for a maturity study since pure coals provide of the Middle Jurassic and at the beginning of the
much more reliable data on vitrinite reflectance and Cretaceous. These tectonic events were characterized
Rock-Eval pyrolysis (Espitalie et al., 1977) than does by the formation of horsts and grabens and a general
dispersed organic matter in rocks. Moreover, in coal downfaulting of the Triassic-Jurassic formations be-
analyses, there are no secondary influences on matur- tween the Trl'lndelag platform and the V~ring and
ity indices due to changes in type of organic matter or M~re Basins (see Fig. 1). During the Cretaceous and
mineral matrix effects. Tertiary, no more faulting activity occurred, and the
differences in thickness ofthe post-Jurassic sediments
were influenced by the relative rate of subsidence.
The area studied is characterized by two main
GEOLOGICAL SETTING structural elements: the Trl'lndelag Platform and the
Halten Terrace (see Fig. 1). In the latter zone, the
In the Haltenbanken area (see Fig. 1), above marine Triassic-Jurassic formations are downfaulted towards
Permian deposits, the Mesozoic starts with a thick con- the Ml'lre Basin. Well 'A' is at the western margin of
tinental and evaporitic Triassic section (Jl'lrgensen and the Trl'lndelag Platform where the coal Unit occurs
Navrestad, 1981). An overall transgressive sequence between 2200 m and 2600 m. Well 'B' is in a faulted
characterizes the Jurassic, in which coals-Rhaetic to zone west of the Trl'lndelag Platform, where the Coal
Hettangian in age-were deposited which have a wide Unit lies between 2500 m and 2900 m. In well 'C',
regional distribution. Above this coaly sequence, which is in the western part of the Halten Terrace, the
sands of Early to Middle Jurassic age are interbedded Coal Unit is much deeper (4200-4600 m), see Fig. 3.
Petroleunt Geochemistry in Exploration of the Norwe.qian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 205-211.
206 Petroleum Geochemistry in Exploration of the Norwegian Shelf

rr----,6"r'O",,0,-'--,--.!8T'OO~'--,~IOT'O~O~'--,~12~'OO~',,----.!:!,14'O0'
~
MATURITY
TR.ENA BASIN .,~
67'00' Vitrinite reflectance
Coal beds are abundant in the Coal Unit, providing
V0RING
adequate and reliable material for vitrinite reflectance
analyses. In the three wells studied, the petrographi-
66'00'
cal composition of the coal is rather similar. The major
component is vitrinite (75-90%) and inertinite repre-
sents 5-10% of the coal. Exinite is relatively abundant
(5-20%) and is concentrated preferentially as alterna-
tions of pure vitrinite and exinite-rich beds. The exin-
65'00'
65'00' ite fraction consists mainly of sporinite. Liptode-
trinite, resinite and cutinite are also present in small
quantities.
MaRE
In well A, the vitrinite reflectance of the Coal Unit
64'00' ranges from 0.35% to 0.4%; in well B, it increases from
64'00'
BASIN
about 0.4% to about 0.55% and in well C, from 0.75%
to 1.10%.
In well C, below 4420 m, the exinite is not easily
recognizable because its reflectance approaches that
63'00' of the vitrinite, a classical feature of coals when reach-
63'00' IOOKM ing higher maturity (Ro = 1.2-1.5%). Simultaneously,
I
vitrinite becomes more fluorescent in a dark brown
10'00' 12'00' colour, making the recognition of exinite by fluoro-
escence difficult also.
Fig. 1. Structural elements of the Haltenbanken area.
Reflectance data for the three wells were plotted
against depth on the same graph (Fig. 4). Although no
reflectance data are available between 3000 m and
l000m - 4200 m, an extrapolated curve between the three sets
of results could reasonably be drawn. It is obvious
TERTIARY that reflectance is directly linked to present depth (see
Fig. 4). From the general shape of the curve, the
2.000m - reflectance gradient is linear with depth in the less
mature zone, and becomes exponential as maturity
JURASSIC
COAL UNIT increases. This increase of the reflectance gradient
------ with maturity is widely observed and is due to the
3000m- exponential increase of the kinetic reaction rate with
temperature.
From these data, 0.6% Ro is reached at around
TRIASSIC
3500 rn, and the value 1.3% near 4800 m within the
4000m-
Lower Jurassic Coal Unit. This is only an average
relation for the three wells where individual trends
may differ since in each well the burial history is not

5000m_
---PERMIAN - -
exactly the same. Thus, the 0.6% Ro and 1.3% Ru
isoreflectance lines may be a little deeper in well C
than in wells A and B because they occur in younger
Fig. 2. Generalized stratigraphy, Haltenbanken area. sediments.

TR0NDELAG
2000

\J
PLATFORM
C B A
HALTEN
E 2500

11
TERRACE

1
~
+ WELL A

J:
t-
Q.
3000 \"'1\


WELL.

WELL C
w \
C \
\
3500 \
\
\
\
\

. \ "\~
4000 \
\

4500

'--"'
20 km
NW i SE \ .20 .40 .80 .80 1.00 1.20 1.40

VITRINITE REFLECTANCE (Ro %)


Fig. 3. Haltenbanken area cross-section, with approximate
location of the studied wells (after J(Jrgensen and Navres- Fig. 4. Variation in vitrinite reflectance with increasing
tad, 1981). depth for Haltenbanken coals.
Maturity Studies of the Jurassic "Coal Unit" 207

Fluorescence of exinite reflectance, a direct relationship between T max and


increasing depth is observed (see Fig. 5).
Observations and measurements were made with a Although the T max gradient is lower than the reflec-
Zeiss u. v. illuminator FI 400 and excitation filter tance gradient, these two parameters show a rather
UG 1. In wells A and B, the different exinite compo- good correlation, as is seen in Fig. 6, where the data
nents show a strong fluorescence. Spores, which are from Haltenbanken coals are plotted in comparison
the most abundant, show variety in colour and inten- with other data from coals of various origins. All are in
sity of fluorescence. Moreover, there is often a change good agreement.
of colour beween the internal and the external part of
the spores. Surprisingly, the external part shows
brighter colour. In these wells, the wavelength cor- Other Rock-Eval pyrolysis data
responding to the maximum intensity in the spectrum In Rock-Eval pyrolysis, the Sl/0rg. C index is indica-
(,.l.ma,) varies between 560 nm and 630 nm, with a mode tive of the amount of hydrocarbons already generated
at 600 nm (Ottenjahn et al., 1975). in the rock. This parameter is a good indicator of
In well C, the fluorescence colour of the spores is maturation, since the generation of hydrocarbons' is
much darker: the intensity is too low for spectral directly dependent on the thermal evolution of the
analysis except in a sample at the top of the Coal Unit organic matter.
showing Amax around 650 nm. When the maturation When only coals are analysed, three secondary fac-
increases, spores are increasingly difficult to distingu- tors usually influencing the Sl/0rg. C results can
ish from vitrinite in their fluorescence because, as dis- largely be ignored: (1) any change in organic type; (2)
cussed previously, the vitrinite itself becomes more migration effects and (3) the mineral matrix effect. In
fluorescent. The dark brown colour of the fluorescent the case of coals, the Sl/0rg. C is only dependent on
vitrinite is of too Iowan intensity for spectral analysis. the maturity, and can be considered as a direct para-
meter to detect the beginning of oil generation. Such
Rock-Eval pyrolysis temperature (Tm.,) an approach was used in the Miocene of the Mahakam
Delta, Indonesia, where a drastic increase in the
In wells A and B, the T max values obtained from the
Rock-Eval pyrolysis of the coals do not vary much Sl/0rg. C ratio clearly indicates the beginning of the
oil window (see Figs. 7 and 8).
with depth, remaining between 420°C and 430°C. In
well C, the T max values are between 440°C and 450°C,
and increase slightly with depth. As with vitrinite
~
o
2000 r-----;----------, E -500

f
I

-E 2500

:r
f- 3000
a.. I
w
c
I
I I
BEGINNING OF OIL
Ro:06%
WINDOW

I + WELL A
3500 I -3000
I • WELL 8

I • WELL C -3500
I

..\
4000 I -4000

4500
~'-~ 20 40 80 80 100

S, /org.C (mg/g)
5000
350 400 450 500 550 Fig. 7. Variation in the Rock-Eval Sl/Org. C (mg hydro-
T max (OC) carbons/g organic carbon) with increasing depth for
Mahakam coals.
Fig. 5. Variation in Rock-Eval T max with increasing depth
for Haltenbanken coals.
350 r--------------,
5 5 0 , - - - - - - - -_ _ _---,
+ WELL A

• WELL a
400
• WELL C

500

() 450
° )(

)(
450
'E"
'"
E
f-

f- 500

400

DCOALSOFVAR'OllSOR'G,N
20 40 80 80 100
.20 .40 .60 .80 1.00 1.20 1.40 5, /org.C (mg/g)
VITRINITE REFLECTANCE (Ro %)
Fig. 8. Variation in the Rock-Eva! Sl/Org. C (mg hydro-
Fig. 6. Correlation between vitrinite reflectance and T max carbons/g organic carbon) with increasing T max for
for Haltenbanken coals. Mahakam coals.
208 Petroleum Geochemistry in Exploration of the Norwegian Shelf

2000 350

··
+

-
WELL A

¥.?
.... +++ WELL 8
+ + + + WELL C
E 2500
+ 400
.:.:
...
:I:
0-
w
3000 + WELL A ,:w.t + ++ ++

-
~
.~
• WELL 8
C
• WILL C
450
3500
.,"
E
4000
...
500

4500

5000
20 40 60 80 100 20 40 60 80 100

5,/org.C (mg/g) 5, 10rg.C (mg Ig)

Fig. 9. Variation in the Rock-Eva! Sl/0rg. C (mg hydro- Fig. 11. Variation in the Rock-Eva! Sl/0rg. C (mg hydro-
carbons/g organic carbon) with increasing depth for Ha!- carbons/g organic carbon) with increasing T max for Hal-
tenbanken coa!s. tenbanken coals.

In spite of a sampling gap between wells A, Band pond to T max = 450°C and Ro = 1%. The decrease in
C, it seems to be acceptable to extrapolate a curve SljOrg. C at higher maturities is explained by expul-
between the three wells for comparison with the sion of hydrocarbons from the coals. Nevertheless,
Mahakam example. In the three Haltenbanken wells, this decrease could also be due to part of the liquid
the SljOrg. C ratio has values between zero and about hydrocarbons being transformed into gas, which
50 mgjg. The increase of SljOrg. C is clearly depen- would then escape more easily out of the coal. In this
dent on the depth (see Fig. 9) and thus on the matura- case, the decrease of the SljOrg. C as seen in Fig. 10
tion level, at least in wells Band C. In well A, which is could be considered to be an indicator ofthe end of the
the least mature well, some high SljOrg. C values are oil window.
clearly abnormal in view of the low maturity. According to the concept of Rock-Eval pyrolysis,
Although no completely satisfactory explanation can when the maturity increases, the SljOrg. C increases
be given, the phenomenon appears to be very similar and the S2/0rg. C (hydrogen index) decreases. This is
to the abnormally high extract levels found in low true for organic types I and II, but not completely
maturity lignites (Ro < 0.3%) by Durand et al.(1975). true for type III (humic). In the study of Durand and
In any case, they cannot be considered to be the result Paratte (1983) on coals, the hydrogen indices, as a
of thermal transformation. whole, increase with the maturity up to a stage around
In the extrapolated curve from Haltenbanken, the 0.7% R 0 or T max = 435-440°C. The highest S2jOrg. C
SljOrg. C increase, which is considered to correspond values are around 325 mg/g, and hydrogen indices
to the beginning of the oil window, occurs at a reflec- decrease at still higher maturities (see Fig. 13).
tance value around 0.55-0.60% (see Fig. 10) and a The Haltenbanken coals follow this rule. In the less
T max of around 430°C (see Fig. 11). The SljOrg. C mature coals (wells A and B) showing reflectance be-
reaches its maximum value at a reflectance of around tween 0.3% and 0.5%, the hydrogen indices are scat·
1% (see Fig. 10) or a T max of 450°C (see Fig. 11). It tered, but in the normal range with values between
decreases at higher maturities. In a statistical analysis 100 and 225 mg/g (see Figs. 12-14). In the more
of Rock-Eval pyrolysis data from coals of various mature well C (reflectance between 0.8% and 1.1%),
origins, Durand and Paratte (1983) have shown high- the hydrogen indices are higher, showing values be-
est SljOrg. C values of about 70 mgjg, which corres- tween 200 and 350 mgjg (see Figs. 12-14).

..
2000
~
+ WELL ...
0
20 WELL 8 + WELL A
0 WELL C
a: 2500 • WELL B

++
w

-
U
z 3000
...«
U
.60
E
W
..J 3500
U.
W
.80
...
:I:

a: 0-
W
1.0 C
w 4000
t:
Z 1.20
a: 4500
!::
> '.40

20 40 60 80 100 200 400 600 800 1000


5, 1 org. C (mg 1 g) HYDROGEN INDEX 52/org.C (mg Ig)

Fig. 10. Variation in the Rock-Eva! Sl/0rg. C (mg hydro- Fig. 12. Variation in the Rock-Eval S2/0rg. C (mg hydro-
carbons/organic carbon) with increasing vitrinite reflec- carbons/ g organic carbon) with increasing depth for Hal-
tance for Haltenbanken coals. tenbanken coals.
Maturity Studies of the Jurassic "Coal Unit" 209

350 r----------------, H/C


2.0 r-------------.,
.........
+


WELL •

WELl C
+

WELL

WELL
A

• WI!LL C

1.5
(J
°
.,
)(

1.0
E
I-

Em COALS OF VARIOUS ORIGIN .5

• ~::T~~A~Rc~:~sa AL 1983)

HYDROGEN INDEX S2/org.C (mg Ig) o .25 O/C .50

Fig. 13. Variation in the Rock-Eval S2/0rg. C (mg hydro- Fig. 15. Evolution of elemental analyses of Haltenbanken
carbon/ g organic carbon) with increasing T max for Hal- coals (wells A, B and C) compared with coals of various
tenbanken coals (wells A, B and C) compared with coals of origins.
various origins.

and 0.45. In well C, H/C is a little less, lying between


~ 0.7 and 0.9, with OIC between 0.03 and 0.05. These
°0 .20 + WELL •
results are in agreement with the normal effects of
a: • WELL B
maturation on H/C and OIC values. As can be seen in
~~
'.
• WELL C
.40
UI
(J + Fig. 15, the Haltenbanken coals fall within the classi-
z ...t cal evolution path in a Van Krevelen diagram,
i!
(J
.60
although a few samples from well B have H/C values
UI
....J
Ii.
.80 which are a little lower than average .
UI
a:
UI
1.00 Extract analyses
!:
Z 1.20 Two extracts were performed on a sample of mixed
a:
I- coals and carbonaceous shales. One extract was taken
:; 1.40 in well B near 2600 m, i.e. in the immature zone
(Ro = 0.4%), and another in well C at 4457 m, i.e. in
200 400 600 800 1000
the mature zone near the end of the oil window
HYDROGEN INDEX S2/org.C (mg Ig)
(Ro = 1.1%). The extracts from well B show typical
Fig. 14. Variation in the Rock-Eva! S2/0rg. C (mg hydro- features of immature humic matter: high CPI value
carbons/g organic carbon) with increasing reflectance for (2.01), high pristane/C17 ratio (1.09) and abundance of
Haltenbanken coals. saturates in the C25-C30 range (see Fig. 16).

Explanations for this increase of S2/0rg. C were


proposed by Durand and Paratte (1983) and Monin et WELL B
cpr· 2.0
al. (1980) as follows. The quantities of hydrogen trans- PR/C17 = 1.1

formed into water and carbon transformed into carbon


dioxide are higher during pyrolysis than during the
natural maturity process. This phenomenon is
emphasized in the immature oxygen-rich coals. In the
more mature coals, which are poorer in oxygen, this
distortion is smaller. Thus, the hydrogen indices in the C 15 C20 C25 C:50

mature coals are closer to the 'true' petroleum poten-


tial, whereas in the immature samples they are under- cpr·1.0
estimated. PR/C17=0.6
In the Haltenbanken coals, the highest S2/0rg. C
values are around 350 mglg and correspond to a T max
of around 450°C (see Fig. 12) and a reflectance of
around 0.9% (see Fig. 14). Due to the discontinuity in
the sampling in Haltenbanken, it is not possible to
,mlu.V!IIIIIJ IIII'J "."
confirm if the peak values correspond to a maturity I
C20 C25
I
C~O

stage of around 0.7% R 0, as indicated by Durand and


Paratte (1983). Nevertheless, above 0.9% R o, the hy- Fig. 16. Saturated hydrocarbon chromatograms from Ha!-
tenbanken coa!s; extracts from well B (Ro = 0.4%) and well
drogen indices decrease clearly (see Fig. 14), indicat-
C (Ro = 1.1%).
ing the proximity of the end of the oil window.

Elemental analyses (atomic HIC and Ole ratios) The extract from well e clearly shows more mature
characteristics: CPI around 1 (1.03), lower pristanel
In wells A and B, most of the H/C values for the Coal C17 (0.62) and phytane/C18 (00.54). and predomin-
Unit are between 0.75 and 1, with OIC between 0.15 ance of light saturates (C15-C20) (see Fig. 16).
210 Petroleum Geochemistry in Exploration of the NOI'wegian Shelf

The aromatics do not show any specific differences MATURITY AND OIL WINDOW IN
between the two samples. Owing to the fact that the RELATION TO THE GEOLOGICAL
organic matter has exactly the same type in the two STRUCTURE
samples, the change of the extract character appears
to be due to maturity effects only. The structure of the Coal Unit follows the regional
pattern of the Jurassic strata, i.e. deepening towards
west. The faulting activity which generated the horst
SOURCE-ROCK POTENTIAL and graben system of this area is related to the Kim-
merian tectonic phase. It started after deposition of
Although the main source-rock in the Haltenbanken the Coal Unit and continued until the Early Cretace-
area is represented by Upper Jurassic marine shales, ous (Hollander, 1983). A period of erosion occurred at
the effective source potential of the Lower Jurassic the end of the Jurassic. Afterwards, the differences in
Coal Unit is also of interest. Until recently, coals and the rate of subsidence during the Cretaceous and Ter-
humic matter generally were considered as source- tiary influenced patterns of burial and of maturity at
rocks only for gas, the gas coming either directly from Jurassic levels.
the thermal transformation of the coal itself or from When considering the maturity process in a general
the cracking of the oil remaining adsorbed in a coal. way, two schemes are possible: (a) pre-deformational
Geochemical studies and geological considerations maturity patterns and (b) post-deformational maturity
have shown that, in some petroleum basins, the patterns. Pre-deformational maturity patterns can be
source-rocks for the oil found were coals or carbonace- summarized as follows. With a normal maturity gra-
ous shales. Such examples include the Niger Delta dient established from burial in a sedimentary sequ-
(Nigeria), Mahakam Delta (Indonesia), and the ence, the isomaturity lines are more or less parallel to
Mackenzie Delta (Canada). the stratigraphic markers. If this sequence is submit-
From the elemental analyses, the coals of the three ted to tectonic events (for example, folding), the
Haltenbanken wells show the characteristics of classi- isomaturation lines are folded with the stratigraphic
cal humic coals at somewhat different maturity stages. markers. In this case, the same stratigraphic marker
Their hydrogen contents are not especially above the shows the same maturity stage regardless of its pre-
average. In the more mature well C, the H/C ratios sent depth. A good example of pre-deformational
are a little lower than in A and B. This decrease in maturity can be found in the Tertiary of the Mahakam
hydrogen content is attributed to hydrocarbon gener- Delta (see Fig. 17).
ation. Nevertheless, it is to be recalled that the hydro- A post-deformational maturity pattern can be
gen indices from pyrolysis are higher at well C than described as follows: (a) an immature sedimentary
in A or B. These two parameters, H/C and hydrogen sequence is tectonized and eroded and (b) an overlying
index, which are dependent on the same process (i.e. sequence, when reaching sufficient thickness, will
hydrocarbon generation) vary here in an opposite induce in the previously deposited sequence a matura-
way. As seen previously, the hydrogen indices are tion gradient with isomaturation lines more-or-less
probably underestimated in the less mature coals for parallel to the stratigraphic marker of this second
analytical reasons. The elemental analysis corres- series. Then the isomaturation lines cross the stratig-
ponds better to the natural evolution path. In any raphic markers of the first sequence. In this case, the
case, the hydrogen indices in well C are high values same stratigraphic level in this sequence does not
for typical coals (200-350 mg/g), even higher than always show the same maturity stage, which depends
those of the coals from the Mahakam Delta in the same only on the present depth. This scheme can be applied
range of maturity (see Fig. 12). to the development of a basin with differential subsi-
It is well-known that the presence of exinite contri- dence.
butes to improve the petroleum potential of coal. In The regional burial history of the Coal Unit has pro-
the Haltenbanken coals, exinite represents 5-20% of duced a post-deformational maturity pattern. During
the coal. For comparison, the Mahakam coals show the Jurassic the Coal Unit was never deeply buried
similar exinite amounts. However, in the Mahakam (probably not more than 1000 m). Downfaulting dur-
example, the most abundant exinite sub-maceral is ing the major Kimmerian tectonic events at the end of
resinite (Snowdon and Powell, 1982) with minor spori- Jurassic and subsidence during Cretaceous and Terti-
nite and cutinite, while in Haltenbanken sporinite is ary induced differential burial of the Coal Unit (see
the major component. Moreover, no exsudatinite was Figs. 3 and 18) with consequent differences in matur-
observed in Haltenbanken, as it was in some coals in ity.
Mahakam. As seen previously from the Rock-Eval data, the
In all cases, the major coal component is vitrinite.
In the natural maturation process, as well as during
am
pyrolysis, vitrinite certainly contributes to hydrocar-
bon generation. The fluorescent vitrinite in the more 500
- ..... ,
mature coals is probably due to the presence of oil 1000

remaining adsorbed in the coal, and this oily micro- 1500 ',I.~.oREFLECTANCE _ - - - ....
..... - ---0.4- - - - - - MARKER
staining can be related to the higher SI/0rg. C ratio 2000
STRP...tlG~APHIC ... __ ... RS
seen in well C. 2500 ... 0.5--------
In conclusion, as noted previously, the best source- 3000 , ..... ,
\ R24
rocks of this area are probably the Upper Jurassic 3500 ',.... .... ....... ,.-'"
shales which contain an oil-prone organic matter in 4000 ......... -----0.7------
rather large quantities. Nevertheless, due to their 4500.i-_ _ _ _ _ _ _ _ _ _ _ __

relatively high hydrogen indices and exinite content, o


~
'Okm

the contribution of the Coal Unit as a potential oil Fig. 17. Example of pre-deformational maturity in the
source-rock must also be taken into account. Mahakam Delta, Indonesia.
Maturity Studies of the Jurassic "Coal Unit" 211

sw NE For oil exploration, the most interesting area is


probably the zone between the Trj'lndelag Platform
and the Mj'lre Basin, because source-rocks there have
em generated a great part of their hydrocarbons which
could have migrated into existing reservoirs. On the
1000 other hand, if liquid hydrocarbons were generated at
TERTIARY some past time in the now overmature Jurassic
000
source-rocks of the Mj'lre Basin, they could have
reached this zone by long-distance lateral migration
CRETACEOUS and may have been preserved from cracking.
In the deep Jurassic reservoirs of the Mj'lre basin,
below 5000 m, only gas would be expected. This gas
would originate either from cracking of pre-existing
oils or from thermal transformation of remaining
5000
kerogen.

Fig. 18. Regional variation in the maturity and the oil


window in relation to the geological structure, Haltenban-
ken area. ACKNOWLEDGEMENTS
beginning of the oil window in the Coal Unit itself We thank Saga and Statoil for permission to publish
corresponds to 0.6% vitrinite reflectance. The reflec- these results. Weare grateful to M. Bellet for techni-
tance values commonly accepted for the upper (shal- cal assistance and to Messrs Laporte, Oudin and
low) limit of the oil window are between 0.5% and Casanova for helpful discussions.
0.7%. Although it is more difficult to detect it directly,
the end of the oil window is probably around 1.3%.
The reflectance measurements show that in well A
the Coal Unit is immature; in well B, it is at the transi- REFERENCES
tion stage between immature and the beginning of the
oil window and, in well C, it is close to the end of the Durand, B. and Paratte, M. 1983. Oil potential of coals: a geochemi-
oil window. By extrapolation (see Fig. 4), the depth cal approach. In: Brooks, J. (ed.), Petroleum Geochemistry and
Exploration of Europe. Blackwell Scientific, Oxford,
limits ofthe oil window in the Coal Unit are, respec- pp. 255-266.
tively, 3500 m and 4800 m. These limits of maturity, Durand, B., Nicaise, G., Roucache, J., Vandenbroucke, M. and
established in the coals, can also be applied to the type Hagemann, H. W. 1975. Etude geochimique d'une serie de char-
II organic matter of the Upper Jurassic shales since bons. In: Campos, R. and Goni, J. (eds.), Advances in Organic
the maturity needed to start hydrocarbon generation Geochemistry 1975. Enadimsa, Madrid, pp. 601-632.
Espitalie, J., Laporte, J. L., Madec, M., Marquis, F., Leplat, P.,
is about the same in these two types of organic matter. Paulet, J. and Boutefeu, A. 1977. Methode rapide de caracterisa-
Thus, in well C the lower part of the Coal Unit is tion des roches meres, de leur potentiel petrolier et de leur degre
close to the end of the oil window and has probably d'evolution. Rev. Inst. Fran,ais Petml. XXXII(l), 23-42.
generated the major part of its hydrocarbons, while Hollander, N. B. 1983. Geohistory and hydrocarbon evaluation of
the Haltenbanken area. In: Spencer, A. M., et al., (eds.), Pet·
the Upper Jurassic shales have not yet reached the roleum Geology of the North European Margin. Graham and
peak of oil generation. In this well (C), hydrocarbon Trotman, London, pp. 383-388.
generation in the Coal Unit started only during the Jllrgensen, F. and Navrestad, T. 1981. The geology of the Nor-
Mid-Tertiary. Of course, in the Upper Jurassic wegian shelf between 62'N and the Lofoten Islands. In: Illing, L.
source-rocks hydrocarbon generation began even V. and Hobson, G. D. (eds.), Petroleum Geology of the Continen·
tal Shelf of North· West Europe. Heyden, pp. 407-413.
later. In wells A and B, no major hydrocarbon gener Monin, J. C., Durand, B., Vandenbroucke, M. and Hue, A. Y.,
ation has yet occurred, either in the Coal Unit or in 1980. Experimental simulation of the natural transformation of
the Upper Jurassic shales. kerogen. In: Douglas, A. G. and Maxwell, J. R. (eds.), Advances
Deeper in the basin, i.e. west of well C, the source- in Organic Geochemistry 1979. Pergamon Press, pp. 517-500.
Ottenjahn, K., Teichmuller, M. and Wolf, M. 1975. Spectralfluores-
rocks which are now over-mature have probably pre- cence measurements of sporinites in reflected light and their
viously generated hydrocarbons. If these oils were applicability for coalification studies. In: Albern, B. (ed.), Petro-
trapped very close to their source-rocks, they were .graphie de la matiere organique des Sediments, Relations avec la
probably cracked into gas. However, if the oils gener- Paleotemperature et Ie Potentiel petmlier. CNRS, Paris.
ated have migrated into reservoirs which are still Snowdon, L. R. and Powell, T. G. 1982. Immature oil and conden-
sate, modification of hydrocarbon generation model for terres-
within the oil window, they will have been preserved trial organic matter. Am. Ass. Petroleum Geologists Bull., 66(6),
from cracking. 775-788.
18 Origin of the gas-condensate
of the Midgard Field at
Haltenbanken
Arvid Elvsborg*, Terje Hagevang* and Torbjf)rn Throndsent
*Saga Petroleum A.S., Norway
tInstitutt for Energiteknikk, Norway

The Haltenbanken area has proven to be a significant hydrocarbon province. The discoveries are in Jurassic
sandstones, with gas-condensate in 6407/2-2 and 6507/11-1, oil in 6407/1-2, and oil with a gas cap in 6407/1-3.
Significant gas shows have been experienced in 6407/2-1 and 6507/11-1 in Cretaceous siltstones. Recently (Summer
1984) Shell has discovered oil in block 6407/9.
Conventional geochemical techniques including carbon isotopes were used to determine the origin of the gas-
condensate and the gas shows. A correlation is evident between the various gas-condensates, and between the
gas-condensates and the gas shows. The bulk of the gasoline fraction and heavier hydrocarbons was generated
during the main phase of hydrocarbon generation at a maturity ofaroundR o = 0.95%. The major portion of the gas
comes from a more advanced maturity level around the end of the oil window. The Lower Jurassic coal-bearing unit
is preferred as the main source for the gas-condensate and the gas shows.

INTRODUCTION SOURCE ROCKS


The Haltenbanken licence area IS located Two potential source rock intervals are recognized in
150-200 km offshore mid-Norway in a tectonically the Haltenbanken area, the Upper Jurassic Formation
complex zone that forms the transition between the H2-2 marine source rock and the Upper Triassic to
stable Trj'lndelag Platform (landward) and the deep Lower Jurassic coal-bearing Formation H1-1 (see
sedimentary Mj'lre and V j'lring Basins to the south and Fig. 4). The deep 6507/12-2 well, penetrating nearly
west (see Fig. 1). 2500 m of Late and possibly Middle Triassic deposits,
The area is a proven hydrocarbon province, with so did not prove any deep Triassic source beds (Rj'lnnevik
far a gas field (Midgard) and a gas/light oil field et al., 1983; Hollander, 1984; Wiik Jacobsen and van
(Tyrihans) (see Fig. 2). Recently (Summer 1984) Shell Veen, 1984). Consequently, it is likely that the
discovered oil in block 6407/9. gas-condensate is sourced from one or a combination
The aim of this paper is to deduce the source rock of the two recognized source-rock intervals.
and maturity of the gas-condensate in the Midgard
Field, which straddles the border of blocks 6507/
Formation H2-2
11 and 6407/2. The field has been tested by one well in
each of the two blocks. The hydrocarbons are reser- The Upper Jurassic Formation H2-2 is a black,
voired in Middle and Lower Jurassic sandstones in a organic-rich 'hot' shale equivalent to the North Sea
structural trap configuration, as shown in Fig. 3. The Kimmeridge Clay (Draupne) Formation.
reserves are estimated at 104 billion m 3 of gas and 21 Average organic carbon content, pyrolysate yield
million m3 of condensate. Hence, the condensate yield (Rock-Eval S2-peak), hydrogen index, kerogen type
is relatively low. and likely product are listed in Table 1. Pyrolysis data
The structural and stratigraphic development of the confirm a high potential for generation of hydrocar-
area has been described by Rl'Innevik and Navrestad bons and correspondingly high hydrogen indices indi-
(1977), B0en et al. (1984), Bukovics et al. (1984) and cate oil source capacity. This is supported by the
Hollander (1984). Detailf. of the reservoir sequence visual organic composition which is dominated by
can be found in Karlsson (1984) and Larese et al. amorphous, liptinitic (oil-prone) material. This
(1984). kerogen can be classified as type II.
Petroleum Geochemistry in Exploration qfthe Norwegian She(f © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 213-219.
214 Petroleum Geochemistry in E xploration of the Norwegian Shelf

4'

68'

67'

66'

65

CJ "PVTf&OI'ItII~·
C "IuJI......... ·
5 IICAMW,.,.

E ,noel"'!
b UlmJ.nIJlllIfT

li2 CI)II(:U5<lN JI~

Fig. 1. Main tectonic elements, Ml'lre-Lofoten area.

with respect to the drainage area for the Migard Field


the area and probably also the volume of peak mature
source rock appears to be too small to represent a
major hydrocarbon supply (see Fig. 5).

Formation Hl-l
The coal-bearing Formation Hl-l has been penetrated
with thicknesses of around 400 m which appear to be
relatively constant throughout the Haltenbanken
area. The sequence consists of alternating sandstone,
shale and coal lithologies deposited under paralic con-
ditions.
The frequency and abundance of coal and car-
bonaceous shale are considerable. Carbon-enriched
beds, marked by having densities less than 2.0 g/cm 3,
make up a cumulative thickness of 30-60 m. In addi-
tion, the unit contains numerous shale intervals more
or less enriched in organic carbon, grading from coalJ
carbonaceous shale to normal shales.
Average organic carbon content, pyrolysate yield
(Rock-Eval S2-peak), hydrogen index, kerogen type
and likely products are listed in Table 1. These are
interpreted average values for the unit as a whole, and
indicate a high hydrocarbon potential. The kerogen is
dominated by humic organic matter and can confi-
dently be classified as type III, with potential mainly
for gas.
A maturity map of Formation Hl-l is shown in
EiWl GAS DISCOVERY ~ OIL DISCOVERY E3 PROSPECT Fig. 6. The maturity pattern shows regional varia-
Fig. 2. Location of the Jurassic Midgard and Tyrihans
Fields. Seismic line (Fig. 3) is indicated. TABLE 1. Source rock data

The unit is relatively thin in the drilled sections, Source rock unit
with thicknesses up to around 60 m. However, suffi- FMH2- '2 FM Hl-l
cient volumes are expected in off-structure areas to
make it a significant potential source rock. Organic carbon (wt %) 10 8
A maturity map of Formation H2-2 is shown in Pyrolysate yield (mg/ g rock) 35 12
Fig. 5. The maturity pattern shows regional varia- Hydrogen index 350 150
Kerogen type II III
tions from immature to peak mature, implying that an Predominant hydrocarbon product Oil Gas
active oil source kitchen exists in the area. However,
Midgard Field Gas - Condensate Origin 215

NW Sf
10

Fig. 3. Structural cross-section over the Midgard Field (from seismic line 460), see Fig. 2.

Isoreflectance Map
Formation H2-2

RICH

RICH

Fig. 4. General stratigraphy for the Haltenbanken area


with source rock intervals indicated.

tions from immature to late mature, implying that an


active gas generation kitchen exists in the area. Gas
generation from Formation H1-1 is areally and volu- Fig. 5. Isoreflectance (Ro) map of Formation H2-2. Shaded
metrically more important in the drainage area for areas indicate oil expulsion in the drainage area for the Mid-
gard Field.
the Midgard Field than oil generation from Forma-
tion H2-2. This is clearly evident from the hydro-
carbon generation summary map shown in Fig. 7,
gas. The full suite of hydrocarbon gases has been
which takes into account both oil generation from
analysed for carbon isotopic composition. The narrow
Formation H2-2 and gas generation from Formation
H1-1.
separations recorded between the wet gas compo-
nents fit the maturity diagram of James (1983) around
a LOM (level of organic metamorphism) value of 13
(see Fig. 8), indicating that the major portion of the
GEOCHEMISTRY OF THE MIDGARD wet gas components comes from a fairly advanced
GAS-CONDENSATE maturity level in the source rock. If the gas is sourced
from the coal-bearing Formation H1-1, it should be
Gaseous hydrocarbons mentioned that James (1983) found that gases derived
The molecular composition of the gas is dominated by from coal frequently indicate a source maturity 1 to 2
methane accompanied by a significant proportion of LOM units too high. The methane value falls above
ethane, propane, iso-butane and n-butane making up the theoretical line. This discrepancy can, according to
16.2%. The iso-butanejn-butane ratio is 0.60, a value James, be interpreted in terms of a mixture between
usually taken as indicating association with the main an immature possibly biogenic gas with a fully mature
phase of oil generation from a source rock. gas, or as a result of cracking of oil. An analogous
The isotopic composition ofthe methane, with a bUC explanation for the gas origin can be inferred using
value of -41.6%n, supports a thermogenic origin for the Schoell's (1983) diagrams shown in Figs. 9 and 10.
216 Petroleum Geochemistry in Exploration of the Norwegian Shelf

2 4 18

1I'
CONDENSATE
Isoreflectance Map
Formation H1-1
METHANE I HIGH-TEMPERATURE
METHANE
o Early-Middle

o 0 ~1~~U:~XPUISlon
. Middle-Late ETHANE
.. Mature
'. Max. expulsion
PROPANE
n-BUTANE

6407/10 BP

Fig. 8. Gaseous hydrocarbons from the Midgard Field plot-


ted in a maturity diagram from James (1983).

b 13Cethane (ppt)
-50 -40 -30 -20

-70

0.
.....c. -60
640711 STATOIL '<t
J:
()
()
Fig. 6. Isoreflectance (Ro) map of Formation HI-I. Shaded ~ -50
areas indicate gas expulsion in the drainage area for the
Midgard Field.
""
-40

-30

-20

Fig. 9. Gaseous hydrocarbons from the Midgard Field plot-


ted in a diagram from Schoell (1983).

~------C2.(%)--
10 20 30 50

-70
6507110 BP
B

640711 STATOIL : ...

Fig. 7. Hydrocarbon generation summary map for the


drainage area of the Midgard Field.

Fig. 10. Gaseous hydrocarbons from the Midgard Field


plotted in a diagram from Schoell (1983).
Midgard Field Gas - Condensate Origin 217
100

'0

80

70 C15

&0

50

.0

p,
30

Czo
to

10

Fig. 11. Gas chromatogram of the saturated hydrocarbons in a stripped condensate sample from the Midgard Field.

C1S+ hydrocarbons and could indicate a lower maturity. This can, how-
ever, also reflect the predominantly terrestrially-
Gross composition derived nature of the source-rock organic matter, a
feature which is supported by the relatively high pris-
The gross composition of the C15+ hydrocarbons in the
tane/phytane ratio of 2.2.
condensate is highly paraffinic, consisting of 67.3%
A gas chromatogram of the aromatic hydrocarbons
saturates, 25.2% aromatics and 7.8% non- (AF 2 fraction) is shown in Fig. 12. The methyl-
hydrocarbons.
phenanthrene index MPI 1 and the corresponding vit-
Gas chromatography rinite reflectance value calculated according to the
method of Radke et al. (1982a, b) and Radke and
A gas chromatogram of the saturated hydrocarbons is Welte (1983) indicate that the major portion of the
shown in Fig. 11. The gas chromatogram shows a condensate was released from the source rock at a
unimodal, smooth n-alkane distribution. The distribu- maturity level of around Ro = 0.9%.
tion achieves a maximum below n-C 17 , followed by a
continuous and steep decrease of n-alkanes versus
increasing number of carbon atoms extending out to
n-C 30 • There is neither odd nor even predominance of CORRELATION
n-alkanes (CPI = 1.0). Based on these features, the
sample appears to be relatively mature. On the other A correlation was attempted between the gas-con-
hand, the pristane/n-C 17 ratio is somewhat high (0.6) densate and the source rocks based on the C15+ hydro-
carbons, using pristance/phytane ratios and carbon
isotopes. The high carbon number biomarker com-
pounds were not considered as the condensate yield is
low and the inherently very low concentrations of high
carbon number compounds are readily overprinted by
trace amounts of less mature oil or biomarkers dis-
solved by the condensate on its migration route.
They will therefore have characteristics of sediments
other than the main source rock.

Pristane/phytane ratio
Pristane/phytane ratios of the condensate and source
rock extracts are shown in Fig. 13. Extracts from
Formation H2-2 give values ranging around 0.95,
which is consistent with the marine anoxic deposi-
tional environment for this unit. The coal-bearing
Fig. 12. Gas chromatogram of the aromatic hydrocarbons Formation HI-I, on the other hand, shows values
(AF 2 fraction) in a stripped condensate sample from the averaging 4.5. Although neither of these average
Midgard Field. values is similar to the condensate, with a ratio of
218 Petroleum Geochemistry in Exploration of the Norwegian Shelf

CONDENSATE
~ FORMATION H2-2 which, with increased maturity, will approach the
gas-condensate position. It is known from laboratory
t ,,,
c::J FORMATION H1-1
maturation experiments (Lewan, 1982) that the car-
bon isotope values of the hydrocarbons produced
,
,, become heavier with increasing thermal maturity.
.t ...
The extracts from Formation H2-2 plot too far from
10

PRISTANE I PHYTANE
the condensate to correlate, even when maturity dif-
ferences are considered.
Fig. 13. Histogram plot of the pristane/phytane ratio in Carbon isotope values for kerogens are usually
source-rock extracts and the condensate. more positive than their associated bitumens. As
shown in the lower part of Fig. 14, the isotopic composi-
2.2, this can be interpreted in terms of a correlation tion of the kerogen in Formation HI-l is suggestive of
when taking into account mixing and maturity effects. corresponding bitumen values which could correlate
It can be shown that the condensate correlates with the condensate. The kerogen values for Forma-
mainly with the coal-bearing sequence, although the tion H2-2, on the other hand, are too negative to have
condensate has a pristane/phytane ratio which is associated bitumen values similar to the condensate.
lower than the average for Formation HI-I. This dis- The carbon isotopic composition of source rock
crepancy can be explained in terms of a maturity extracts, kerogens and the reservoired condensate
effect or slight contribution from Formation H2-2. It supports the other evidence that the gas condensate is
is known that the pristane/phytane ratio, at least in derived mainly from the coal-bearing Formation HI-I.
coal-bearing series, attains a maximum value around
the onset of oil generation (Ro = 0.55%) and then
diminishes with increasing maturity (Boudou, 1981; SUMMARY
Durand, 1983). This effect alone can explain the lower
values obtained for the rather mature condensate Maturity
(Ro= 0.9%) compared with the immature to early
The molecular composition, the iso-butane/n-butane
mature setting of the extracts analysed.
ratio and the isotopic composition of the hydrocarbon
gases clearly support a thermogenic origin for Mid-
Carbon isotopes
gard gas. The narrow isotopic separations between
Figure 14 shows a cross-plot of the carbon isotopic the wet gas components suggest that the gas was
composition of the C15+ saturate versus aromatic hyd- released from a source rock at a fairly advanced
rocarbons. The extracts from Formation HI-I plot in a maturity level towards the end of the oil window.
different position from extracts from Formation H2-2. The pristane/n-C 17 ratio, on the other hand, points
The Formation HI-I extracts occur in a position towards a more moderate maturity level. This is also
supported by the C15+ aromatic hydrocarbons, which
-20 . . . . . - - - - - - - , - - - - - - - - - - - - , indicate a maturity level corresponding to a vitrinite
• CONDENSATE reflectance of Ro = 0.9%.
/::. EXTRACT FROM FM H2-2 A reasonable explanation for these observations is
X EXTRACT FROM FM H1-1
that the bulk of the gasoline range and heavier hyd-
rocarbons were released from the source rock at mod-
erate maturity levels during the main phase of hyd-
rocarbon generation, whereas the major portion of the
gas comes from a fairly high maturity close to the end
of the oil window.

Source rocks
Two potential source rock intervals are positively iden-
-30~----~~~~--------~----------~
tified in the Haltenbanken area. Formation H2-2,
which is equivalent to the North Sea Kimmeridge
Clay (Draupne) Formation, is a marine sapropelic
source rock with type II kerogen and specifically high
potential for oil generation. Formation Hl-l is a coal-
bearing source rock containing type III kerogen and
-35 L -_ _ _ _ _____ _____
with high potential for gas and condensate generation.
The abundance of pristane and phytane, and in par-
~~ ~ ~

-35 -30 -25 -20


ticular the rather high pristane to phytane ratio in the
gas condensate, suggests a partially or mainly terres-
trial source rock. The evidence therefore favours
< ASSOCIATED BITUMEN
Formation Hl-1 as the main source rock. This is
further supported by the carbon isotopic composition
FM H2-2 ---/::.-/::.- - - - ---/::.-- - - - - - - -
of the C 15 + hydrocarbons which correlates with
FMH1-1---- - --)¢O(--x-------- extracts and kerogens from Formation HI-I.
I
-30
I
-25 -20
The conclusion that the coal-bearing Formation
-35
Hl-l appears to be the main source rock for the gas
condensate in the Midgard Field is compatible with
Fig. 14. Isotopic composition of kerogens and C15+ hyd- the hydrocarbon generation summary map, shown in
rocarbon fractions in source-rock extracts and the conden- Fig. 7, which delineates mature source rocks in the
sate. drainage area for the field. The relevant area is domin-
MidgaTd Field Gas - Condensate Origin 219

ated by gas-generative Formation HI-I, whereas the Karlsson, W. 1984. Sedimentology and diagenesis of Jurassic sedi-
area of oil-mature Formation H2-2 is highly ments offshore Mid-Norway. In: Spencer, A. M. et al. (eds.),
PetroleulI! Geology of the North European Margin. Norwegian
restricted. Petroleum Society/Graham and Trotman, London, pp. 389-396.
Larese, R. E., Haskell, N. L., Prezbindowski, D. R. and Beju, D.
1984. Porosity development of selected Jurassic sandstones from
the -"Torwegian and North Seas, Norway-an overview. In:
ACKNOWLEDGEMENTS Spencer, A. M. et al. (eds.), PetroleUi/! Geology of the North
Europe<tn ;Wargin. Norwegian Petroleum Society/Graham and
Trotman, London, pp. 81-96.
The authors wish to thank Saga Petroleum, Statoil, Lewan, M. D. 1982. Effects of thermal maturation on stable organic
Norsk Hydro, Shell, Arco, Agip and Deminex for carbon isotopes as determined by hydrous pyrolysis of Woodford
their permission to present this study. Shale. Proceedings of SYlllposium 'Stable Isotopes in Hydmcar-
bOll Exploratiol/', Stavanger. Norwegian Petroleum Society,
15 pp.
Radke, M. and Welte, D. H. 1983. The methylphenanthrene index
(:VIP I): a maturity parameter based on aromatic hydrocarbons. In:
REFERENCES Bjorpy, M. et al. (eds.), Advances in Organic Geochcl/listry 1981.
Wiley, London, pp. 504-512.
BOW10U, J. P. 1981. Diagenese organique de sediments deltaique, Radke, M., Welte, D. H. and Willsch, H. 1982a. Geochemical study
delta de la Mahakam, Indonesie. Thesis, University of Orleans. of a well in the Western Canada Basin: relation of the aromatic
Bukovics, C., Shaw, -"T. D., Cartier, E. G. and Ziegler, P. A. 1984. distribution pattern to maturity of organic matter. Geochilli.
Structure and development of the Mid-Norway continental mar- COSII10Ch,:III. Acta, 46, 1-10.
gin. In: Spencer, A. M. et al. (eds), Petr-oleum Geology of the Radke, M., Willsch, H., Leythaeuser, D. and Teichmiiller, M.
North European .'lIar-gin. -"Torwegian Petroleum Society/Graham 1982b. Aromatic components of coal: relation of distribution pat-
and Trotman, London, pp. 407-424. tern to rank. Geochilll. Coslllochilll. Acta, 46, 1831-1848.
Bpen, F., Eggen, S. and Vollset, J. 1984. Structure and basins of R0nnevik, H. and Navrestad, T. 1977. Geology of the Korwegian
the margin from 62 to 69 'N and their development. In: Spencer, Shelf between 62 ON and 69 ON. Geojournal, 1,33-45.
A. M. et al. (eds), Petmleum Geology of the North Enropean R0nnevik, H., Eggen, S. and Vollset, J. 1983. Exploration of the
Margin. Norwegian Petroleum Society/Graham and Trotman, Norwegian Shelf. In: Brooks, J. (ed.), PetroleulI! Geochemistry
London, pp. 253-270. and Exploration a/Europe. Geol. Soc. Spec. Publ. No. 12, 71-93.
Durand, B. 1983. Present trends in organic geochemistry in (Blackwell Scientific, Oxford.)
research on migration of hydrocarbons. In: Bjor0Y, M. et al. Schoell, M. 1988. Genetic characterization of natural gases. AAPG
(eds.), Advances in Organic Geochemistry 1981. Wiley, London, Bull., 67, 2225-2238.
pp. 117-128. Wiik Jacobsen, V. and van Veen, P. 1984. The Triassic offshore
Hollander, -"T. B. 1984. Geohistory and hydrocarbon evaluation of Norway north of 62 0-"T. In: Spencer, A. M. et al. (eels.),
the Haltenbank area. In: Spencer, A. M. et al. (eds.), PetroleulII PetmleulII Ge%qy ~(the North h,'urupeall ,WIII-gill. Norwegian
Geology of the North EU1"Opean .II;Im·gin. Norwegian Petroleum Petroleum Society/Graham and Trotman, London, pp. 317-327.
Society/Graham and Trotman, London, pp. 381-38R
James, A. T. 1983. Correlation of natural gas by use of carbon
isotopic distribution between hydrocarbon components. AAPG
Bull., 67,1176-1191.
Part IV
Surface Geochemistry
19 Surface geochemical exploration
for hydrocarbons in offshore
areas-principles, methods and
results
E. J. Schiener, * G. Stober* and E. Fabert
*Deminex GmbH, Essen, FRG
tFederal Institute for Geosciences and Natural Resources (BGR),
Hannover, FRG

Deminex has applied direct hydrocarbon detection methods in a number of surveys in the North Sea. Both regional
open-grid surveys and detailed surveys with narrow sample spacing were conducted. The advantages and
limitations of both bottom sampling and 'sniffer' are compared. The greater sensitivity of bottom sampling
outweighs the operational advantage of the 'sniffer'. As most of the areas were surveyed before drilling activities,
comparison of the early interpretations with drilling results permits a review of principles applied in
interpretation. The necessity of integrating geological/geophysical subsurface information with the geochemical
surface signature has emerged as the most important aspect. Interpretation compatible with realistic geological
models reduces source- and migration-related risk in exploration by assessing:
presence and maturity of source rocks;
presence and type of hydrocarbons in seismically defined structures and
seal integrity.
As a case history, the results of a geochemical survey in the Trrenabanken area are integrated with available
subsurface information. The marginally mature Kimmeridge Clay Formation, as the principal oil source rock in the
Helgeland Basin, is reflected in the near-surface hydrocarbon indications.

INTRODUCTION All of these methods assume essentially vertical


movement of hydrocarbons from depth, either from
The search for surface expressions of mineral deposits mature, generating source rocks or from accumula-
in the subsurface is as old as man's ability to utilze the tions. The basic assumptions have not changed during
resources. At the onset of the hydrocarbon age many the half century of application. Despite significant
of the early discoveries of oil were linked to surface advances in understanding and even quantifying some
seeps (Thompson, 1910). With increasing demand in of the subsurface processes of migration, accumulation
the 1930s, attention was turning towards means of and retention (Roberts, 1980; Schowalter, 1979;
detecting less conspicuous hydrocarbon manifesta- Leythaeuser et al., 1982) a large speCUlative element
tions at the surface (Laubmeyer, 1933; Horvitz, 1939). remains. This applies particularly to the effectiveness
From the early beginnings, the immediate goal was of various transport mechanisms, ranging from diffu-
locating and delineating accumulations, mostly in or sion to bulk migration through microfracture systems.
adjacent to areas where production had been estab- Recent advances in analytical technology as well as
lished. Although the original aim has not been aban- in fundamental and applied research (Stahl and Carey,
doned, recent emphasis has shifted towards regional 1975; Bernard, 1978; Faber and Stahl, 1983) have con-
evaluation of hydrocarbon potential and differentia- tributed towards lowering detection limits. In particu-
tion between oil or gas potential, specifically in off- lar, the measurement of stable carbon and hydrogen
shore areas. isotope ratios of gas quantities in the microlitre range
Compared with the multitude of approaches used in has increased the value of the analyses. These
onshore investigations (Philp and Crisp, 1982), in off- advances have improved the resolving power of the
shore work the direct detection and analysis of hyd- method, but have also increased the cost of data
rocarbons by exclusively organic geochemical methods acquisition.
has evolved (Faber et al., 1983; Gevirtz et al., 1983; Proprietary secrecy as well as scientifically uncler-
Faber and Stahl, 1983). documented sales talk have not contributed to improv-
Petroleuill Geochemistry in gyp/omtion of the Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 223-238.
224 Petroleum Geochemistry in Exploration of the Norwegian She((

ing the reputation of geochemical prospecting for hyd- -despite improved gear-handling equipment the pro-
rocarbons. cedure is strongly dependent on relatively calm sea
Despite increased use of state-of-the-art science, state for progress (maximum force 4-5 Beaufort
these methods remain empirical. Only by a joint effort scale), unless a large vessel equipped with a moon pool
of integrating geological/geophysical information with is used;
geochemical data will optimal benefit be obtained. -the sample recovery is strongly controlled by type
Deminex has tested direct geochemical hydrocarbon of seaf100r (poor or no recovery in areas with outcrop-
detection methods, predominantly in the North Sea. ping, consolidated sediments and sandy/gravelly
The following text is a review, with reference to some cover) and
case histories, of methods and of current ideas as -as a minor (but nevertheless possibly irritating)
developed through the co-operation between Deminex limitation, complications and delays caused by ship-
and the German Federal Institute of Geosciences and ping of sample material may occur, where analytical
Natural Resources (BGR), partially supported by the work is performed in onshore laboratories.
German Federal Ministry of Research and Technology The advantages of bottom sampling compared with
(BMFT). 'Sniffer' work are:
-sediment degassing and subsequent analysis of
gases is highly sensitive, providing information at low
DATA ACQUISITION gas concentrations;
-repeat analyses can be performed on sample splits;
The standard approach in offshore work, as applied -the standard analytical spectrum can be expanded if
also by Deminex, consists of direct detection and deemed necessary;
measurement of hydrocarbon components in near sur- -type and character of recovered sediment yield
face sediments or in the watercolumn. Two different additional information for data interpretation and
systems of investigation and data collection have assessment of data reliability;
developed: -sample position is well-defined within the accuracy
-spot sampling of bottom/ subbottom sediments by of the positioning system and
standard engineering or marine geological methods, -sampling equipment can be shipped at lower cost,
with subsequent extraction of hydrocarbons ad- can be hired or be assembled locally in logistically dif-
sorbed/ dissolved in the mineral matrix and pore ficult regions.
waters and
-continuous sampling of near-bottom waters by a Near-bottom water column 'Sniffer'
pumping system suspended from a vessel with almost
real-time shipboard extraction and analysis of dissol- Sampling is performed through a fish-mounted pump
ved hydrocarbons ('Sniffer'). which supplies a continuous f10w of water to the ship-
board analytical instruments by means of the umbilical
Each of the systems has obvious advantages and towline. Predetermined accumulated quantities of the
shortcomings, which have to be considered together water are stripped of dissolved gases by vacuum or
with the objectives of the planned survey. inert gas assisted by mechanical agitation. At regular
short intervals the gas so obtained is analysed for its
Subbottom sediments hydrocarbon components by gas chromatography. In
latest developments the accumulated methane frac-
The least complicated sampling tools are gravity cor- tion of a sufficient number of gas samples is oxidized to
ing devices, possibly with added spring-operated pis- CO 2 for later mass spectrometric determination of the
ton devices for improved penetration and core reten- carbon isotope ratios, delta 13C/I2(; (Cornford and
tion. Although it is claimed that data quality is not Dyson, 1984).
influenced by depth below the water/sediment bound- The tests performed by Deminex indicate the fol-
ary (Gerard and Feugere, 1969), most investigators lowing limitations:
recommend sampling from as deep as technically feas-
ible below the seafloor (Simoneit et al., 1979; B. D. -maximum towing depth presently does not exceed
Carey, pers. comm.). 300-400 m;
The optimum sampling depth varies, and depends -despite possible sonar-controlled automatic tow-
on grain size, composition and related physical proper- depth adjustments, the height above bottom is kept at
ties of the sediment. In clayey sediments, samples approximately 10 m to avoid snagging of tow body,
from below 1.0 m yield reproducible results, although with resulting loss of resolution (Schink et aI., 1971);
deeper penetration is generally attempted (~2. 0 m). -positions of detected anomalies do not necessarily
Increasing content of silt/sand-sized components coincide with the actual seep, owing to possible dis-
reduces penetration of the sampling tool «1.0 m) and placement by marine/tidal currents;
generates more 'noise' in the analytical results. -for low concentrations of dissolved hydrocarbons
Recovered sample material is stored and trans- the technique is not sufficiently sensitive;
ported under cryogenic conditions (in liquid nitrogen -range of analyses is limited to gas chromatography,
at between -196°C and -80°C) in order to prevent possibly with mass spectrometry.
loss of volatile hydrocarbon components. The obvious advantages listed below are not consi-
Bottom sampling is a stop-go operation with high dered to outweigh those of bottom sampling in more-
demands on vessel manoeuvrability, station-keeping detailed investigations. For reconnaissance surveys
ability, and on-deck equipment in order to secure covering large areas, the operational ease of the 'Snif-
acceptable efficiency. There is no operational limit fer' outweighs the lack of resolving power. It offers:
imposed by water-depth. In Deminex's experience the
following limitations were found (given in order of -continuous operation, allowing deployment parallel
decreasing severity): to (for example) seismic data;
Suiface Geochemical Exploration for Hydrocarbons 225

-high sampling efficiency and density (at 5 knots, Analytical procedure


approximately 400 m horizontal distance between
sample points) and For the investigations presented in this paper the
-sampling independent of type of seafloor sediments. analytical instrumentation and procedure as described
by Faber and Stahl (1983) were applied. The proce-
In test surveys where 'Sniffer' was run parallel to dure consists of the following steps:
bottom sampling only pronounced concentration
anomalies were detected by the 'Sniffer' method. -exsolution and desorption of gas from the sediment;
When results were compared, the anomalies were -quantitative determination of hydrocarbon compo-
either due to biogenic methane as shown by the 013C] nents CC C6 by gas chromatography and
values in the subbottom samples «-60%0 PDB) or -determination of stable carbon isotope ratios of
were related to 'seepage' as proven by a dense pattern methane and, where quantities permit, of ethane.
of subsurface faults (see Fig. 1). Processing bulk samples in the standard gas strip-
ping procedure frees the 'total gas', a mixture of free
and adsorbed gas. It has been established empirically
that data obtained by analysing only the adsorbed gas
are more consistent in repeat analyses (Sokolov, 1955)
and show to a lesser extent evidence of microbial modi-
fications such as oxidation. Processing for adsorbed
gas involves separation of the silt/clay fraction from
the bulk sample by sieving. The fraction <63 J.1m is
then treated according to the standard procedure.
For additional information on heavier hydrocarbons
which may be present, u.v.-fluorescence (UVF) spec-
trophotometry (fluorimetry) of aromatic compounds in
sediment extracts is performed. Two approaches are
available:
-fixed-wavelength emission intensity measure-
ment (excitation: 265 nm; emission: 320-365 nm) as
pioneered in the USSR (Kartsev et al., 1954) and
described by Horvitz (1981) and Gevirtz et al. (1983).
Aromatic compounds are recovered from the sediment
by ultrasonic extraction instead of by the time-
consuming exhaustive extraction method originally
used and
CONTOUR INTERVAL 10 pplt
-total-scanning fluorescence intensity measurement
• >60ppb
as described by Brooks et al. (1983).
=a= fAULTS OLDER THAN BASE CRETACEOUS
..... fAULTS YOUNGER THAN CRETACEOUS

RESULTS AND INTERPRETATION


The general principles applied in data interpretation
are as follows (Jones and Drozd, 1983; Faber and
Stahl, 1984):
-the chemical signature of the near-surface sediment
gas, as determined mostly by the stable carbon
isotope ratio of its methane (ethane) fraction and to
some extent by its molecular composition, is indicative
of its genetic origin (biogenic versus thermal) and
-elevated concentration values of thermal surface
gas and elevated UVF -emission intensities are the
surface expressions of migrated hydrocarbons and are
indicative of either on-going generation (presence of
mature source rocks) or of pooled hydrocarbons in the
subsurface, particularly when related to faults or
closed structures.
The geochemical data used in this paper for the
interpretation of the near-surface manifestations of
hydrocarbons consist of:
-distribution of the hydrocarbon gas components
CH 4 through C6H14 expressed in relative percentages
and absolute values (ppb, w/w), (see Figs. 2(a) and 11,
Fig. 1. (a) Surface anomaly (butane) as detected by 'Snif- and Table 1);
fer', with fault traces at base Cretaceous level. WNW-ESE -stable carbon isotope ratios of methane, and if poss-
faults penetrate into the Cretaceous sediments. (b) Corres- ible of ethane, expressed as delta 13C/12C (%0 PDB
ponding seismic section with fault pattern and well location. standard) (see Fig. 2(b» and
('0' indicates non-commercial oil.) -intensities of UVF -emission spectra at wavelengths
226 Petroleum Geochemistry in Exploration of the Norwegian Shelf

(b)

6"C.
('DO PDB STANDARD)
.2Q -JO -35 -40 -50 -50

Fig. 2_ Comparison of propane concentration distributions (in ppb) for different surveys on the Norwegian Shelf and in the
UK Inner Moray Firth (IMF). (b) Comparison of 613C1 (%0 PDB) distributions for the same areas.

of 320 nm and 365 nm, with 270 nm as excitation The usefulness of this tool for characterization of
wavelength; alternatively, total scanning fluorescence surface gas families is demonstrated in Fig. 6, where
with numerical and 3-D graphic display (see Fig. 3). the predominantly Kimmeridge-sourced near-surface
gases are clearly separated from gases of other
sources and maturities.
Data statistics
Cross-plots of single gas components (see Figs.
Comparison of data sets is essential in a method that 7(a)-(f) can provide clues to possible multiple
relies to a great extent on experience. However, as sources or to different source maturities. Well-
evident from Fig. 4 the statistical approach alone has expressed linearity from methane to pentanes is seen
only limited significance, despite claims to the con- as evidence for co-migration of the components.
trary (Jones and Drozd, 1983). Experience on the Where particularly the higher homologues deviate
Norwegian Shelf north of 62°N has shown that the from the linear trend towards higher concentrations
occurrence of more than one major source rock and the (see Figs. 7(c), (e) and (f» leakage or migration from
burial of source-rock units to different depths, with an oil reservoir rather than from a source rock can be
resulting differences in maturity, generate data that inferred (Bernard, 1980). Attempts to correlate sedi-
cannot readily be accommodated in standard concepts ment gas yields with UVF intensity distributions have
which were developed for single source rocks. been unsuccessful so far (Gevirtz et al., 1983).
Under favourable conditions (uniform surface
lithologies, uniform age of sea-floor sediments, low Inferred relationship between b13C values and
sea-bottom temperatures, greater water depths) as source maturity
found on the Norwegian Shelf, regional differences in The b13C methane values of the survey covering the
magnitude and composition are already recognizable postulated drainage area of the Midgard structure of
in statistical data sets. The differences, however, are Haltenbanken consistently indicate a source maturity
rather subtle and are limited to isotope ratios for com- of 1.8% VR equivalent (see Fig. 8). Before drilling,
position and gas yields for magnitude (see Figs. 2(a) this was unsupported by a maturity model assuming a
and (b». Reasons for the generally low yields on the Kimmeridge source. With hindsight, with an
Norwegian Shelf may be the low seismicity of the improved maturity model and seismic data combined,
Scandinavian Shield which forms the base of the the isotopic maturity fits perfectly well into the matur-
sedimentary section, combined with the widespread, ity range for a Coal Unit source (Upper Triassic to
thick and practically unfaulted Cretaceous and Terti- Lower Jurassic) for both the reservoir gas and the
ary cover sediments. In the Inner Moray Firth, where near-surface gas.
numerous faults cut the Cretaceous and where the Compared with reservoir gases, near-surface
Tertiary sediments are absent in the western parts, methane is usually isotopically heavier. One explana-
gas yields are higher by a factor of2-3 (see Fig. 2(a». tion ascribes this to mixing of gas during accumulation
in the reservoir from sources that have passed
Cross-plots for genetic characterization of gases through a range of maturities during subsidence.
Cross-plotting of data sets has been advocated for the As a second possibility, fossil biogenic methane
genetic characterization of both near-surface gases retained in the reservoir during subsidence is envis-
and reservoir gases. Of particular value are those aged as the cause of isotopically lighter signatures of
cross-plots that combine stable carbon isotope ratios reservoir gases (Faber et aI., 1984). In routine head-
with gas component ratios (Bernard, 1978; Faber et space gas analyses of wells, isotopically light methane
al., 1983; Schoell, 1983). Figure 5 shows contrasting « - 50%0) has been observed at great depths III numer-
data distributions for areas with near-surface biogenic ous wells in geographically widely spread areas. The
gas (Indonesia) and near-surface thermal gas (North signature of the near-surface gas then probably repre-
Sea). Apart from a limited area in the Inner Moray sents the most recent and most mature gas genera-
Firth, the data of which do not significantly affect the tion at depth without any biogenic input. The low
plotted mean in Fig. 6, no evidence of biogenic pressure regime of biogenic gas formation precludes
methane has been seen in near-surface gases of North migration by pressure differential and favours reten-
Sea surveys. tion in solution in the pore water (Barker, 1984).
14'.4
FJu~enc_8QIIJrrt
___ !!~l,::·!.' _ _ _ _ _ __
~
•• M<ttl M~I.
J
.t
w

1$···1ltt ? 0
- CONDENSATE SIGNATURE (46 API)
2S1 3. lSI...
l:r ~ . '.'nl,"
..se
(n)
,.
365.... 328 • . 56
F luClr"..ce1'lC"-SCNt
1 ___ ~ ... I.::!.'_ _ _ _ _ __

•• ~ ,....,~ ..
j t
.; 351 w
! r?
~.
~
g
,...... ".\ /~,),\, ') - LIGHT OIL SIGNATURE (40 0 API)
-258 - ---
[at •• ton tn'
----
365.... 328 •
.....................
.....................
1.2
'"G:l
g
1173 F luor-..ce1'lC'-SCIIN
1
5 [.
t
w £.
t>l
-@
* - OIL SIGNATURE (30 0 API) [C).
365 .... 328 • 2.67 ;:l

~
SED IMENT EXTRACT HALTENBANKEN

CONTI NUOUS UV-FLUORESCENCE SCANS:


f
!:l
"'"
- 3-0 DISPLAY (LEFT AND CENTRAL COLUMNS)
~
- TWO-DIMENSIONAL STANDARD DISPLAY (RIGH
COLUMN)

SED IMENT EXTRACT HALTENBANKEN


(NOTE SHARP GAS/CONDENSATE PEAK IN
THE SHORT WAVELENGTH FIELD WITH "OIL"
SIGNATURE EXTENDING INTO LONGER WAVE-
LENGTHS)
r., .. ,Oft f_l

Fig. 3. Comparison of u. v.-fluorimetry measurements of condensate, oils and sediment extracts.

~
"'l
228 Petroleum Geochemistry in Exploration of the Norwegian Shelf

OIL
:-41.1

,
~
,40,8
~
,,
DRY .0.
·;40,0
~
en
,,
OIL <!'>
...
CoO
w ~
·38,5:

, ., ,,
GAS
" .... ~o..
·'40,0
~\

OIL
.... ~
.I " , ·369
:t:. , "
..0 ././
DRY J~~ ~
c ..... ", "
./
36,7

en "
DRY ~
OIL , ~
!~.~

GAS ,. ~

OIL
~ ~
I
HC y lei;Jl00 3
10 '-37,-39 ,-410-43,-45
CI= • C4= 0 -10""2--"-U-'-F-::-~103 6
13 Cl 0,75 o,s5 D,95
C2= 0 C5= + 320nm 365nm
C3= • Cl1 ICn

Fig. 4. Data comparison of North Sea and offshore Argentina surveys; left column indicates the hydrocarbon types
discovered.

- - - C 2 /%1-
10 20 ~ ho 50

-70

r;:;"'
B
a. -60 Indonesia

-.!:..,
...

___
S'05 r-----------------,IBB~~ruGE..Nmlc~_=~~;;~~--~

f ...
:z:
, IORIGIN '-'
C,,)
~ I
~ -50 , To
Ul0'
)
hI.GRATION

II INDON.SIA
Argentina \

\~'" ~&i_~v
iki!!.g_~!ben Are}..

10'
i -40 ('~.: :~
, __ /
Inner Moray
Firth

"~l~'~ ~ ''''''
Tc
; "
~ TTh/ /
10' -30 Trams.flaket

/ /
10' ~~ ~\.:,'
------
I W \
\
THERMAL GASES;
I NN BLOCK 34/1- __ ---
__ ..\.4- HALTENBANK£N
/////
-20
/

l!iil \ 1 rROMS
BACKGROUND I ~, OIL - , 100 90 80 70
I ~'WINDOW'
181 I
----."",,--'-"-.",,:-----:.,1:-,-----:.•",----.,,,,----.7.,,:---."7,,:-----:1.•,
10°_ 2,,", Fig. 6. Genetic characterization of sediment gases from dif-
.!"c, (pptJ ferent Deminex surveys, after Schoel! (1983). B = biogenic
(BGRI
origin; M = mixed biogenic and thermal origin; To = associ-
Fig. 5. Bernard diagram: (P~Cl values versus ratio C1/C 2 + ated gases (oil); Tc = associated gases (condensate);
C3 of data from the Norwegian Shelf and Indonesia. TTh = non-associated gases (coa1!overmature gases).
Surface Geochemical Exploration for Hydrocarbons 229

.. " nRfn , TRRfNR~RNHN .. " RRER , TRRfNRBRNJ.EN


/,
.. .. IlR£R , TRR£NRBRNJ.EN
. 7'·
j
m / in~" ) .../

~ :1 '../ ,/ ~ j /. t /
r~"; 'I I LJ 1 . ,' ../. . I:J., ./ ../
~~:j/ I~ j /' ~-

t'i<c ~~~.._.~I ~ t</. ,,,,,~ tl(~~/ -', ,, 5


I' ,;,/ :z

,c,,!
2ll 40 50 &070eoqOIlI!! 10 30 4050 &0708090100 20 30 40 50 &0708090100 I
...
3~
METHRNE-TIELD IPPB) METHRNE-TIELD IPPB) METHRNE-YIELD (PPB)
RflER , TRRENRBRNJ.EN .. RflER , TRREHRBRI«£N .... ARER TRRENRBHEN lJI
.'
I

, III
.1 ([l
/ . '::D
/
0...
0...
/ I
o
-.JD I ~"
o
/
./
. /' :
o
w- -.J"
w-
/ ' ,./,
, ,
TI
>- /'" ./ >-
..(' -I
~ I" 7' W
I '"
/,':Y , TI
~ 1: ;>,/J'
Z
u:
r- ',Y'
=:J
([l .J-' '
//
d - '/ --....---r--T--;-~.-.-...----'
e - / f
1 J 4 5 & 7 8 9 10 10
ETHRNE-TIELD IPPB) PROPRNE-YIELD (PPB)
Fig, 7(a)-(f). Cross-plots of yield of sediment gas components, methane through butanes, with a linear relationship for
methane through propane and a second concentration trend developing for the butanes.

Mapping
CARBON ISOTOPE RATIOS un INFERRED
UTURITIES OF GAS COMPONENTS OF The traditional approach to evaluating near surface
2 IUlTEMUlIKEM WElLS AND OF
lEAR SURFACE GAS hydrocarbon data has been contouring of gridded data
-50 or their display as concentration profiles (Horvitz,
1939; Sokolov, 1955; Kartsev et al., 1954; Jones and
Drozd, 1983. Much of the early literature, but also
more recent accounts, emphasized the 'halo' concept
(Horvitz, 1980). According to this concept, hydrocar-
bon accumulations are supposed to be perfectly sealed
by, amongst other effects, pronounced secondary
cementation over the area of subsurface closure. Mig-
rating hydrocarbons would 'spill' at the margins of the
accumulation and create a halo-shaped rim of high
concentrations of near-surface hydrocarbons enclosing
an area of comparatively low concentrations. Pub-
lished evidence generally consists of surface maps
with producing wells encircled by the 'halo'. Subsur-
face structural relationships are rarely shown.
As already stated by Jones and Drozd (1983),
diapir-related structures with circular fracture
systems might well also generate halo-shaped surface
anomalies. The closest approximation to such a
relation between subsurface structure and surface
configuration was observed in a Deminex survey in
the Senja Ridge area north of 62 ON (Norway), where
shale diapirism is suspected,
0,5 -~I 1,2 ',I An alternative approach, in contrast to the halo-
VITRINITE REflECTANCE .. concept, relies on imperfect seals as a working
Fig. 8. Inferred relationship between al3 C values (methane hypothesis, In this case subsurface accumulations are
and ethane) and source rock maturity for reservoir, heads- assumed to be expressed by higher than background
pace and surface gases (Haltenbanken) after Faber et al. concentrations of near-surface hydrocarbons over the
(1984). accumulation at depth,
Petroleum Geochelni8try in Exploration of the Norwegian Shelf

The structural subsurface conditions thought to related interpretation are elevated propane and pen-
cause the various types of 'direct' anomalies are dis- tane concentrations and elevated UVF intensities;
cussed in Kartsev et al. (1954). Two statements in that methane and ethane concentrations follow the trend
publication are quite relevant for any data interpreta- but are more erratic. Although carbon isotopic ratios
tion. (alaC I ) consistently indicate thermally generated hyd-
rocarbons, the unequivocal distinction between an oil
'The shape of an anomaly, unrelated to the geologi- signature (a I3 CI < -40%0 PDB) and a gas/condensate
cal structure of the region, cannot have any great sig- signature (-40%0 < al3 CI < -37%0) for a given struc-
nificance in the interpretation of gas-survey results.' ture is not possible (see Fig. 10).
'A gas anomaly indicates the presence of a source of In several cases subsequent drilling encountered
gas but does not specify its size.' numerous good shows and impregnations high above
an expected reservoir which turned out to have high
In detailed surveys of seismically defined structures water saturations, as in the case of Fig. 9. The struc-
at Jurassic level in the North Sea, the surface data ture was drilled near a seismically indicated 'gas
exhibit direct or positive anomalies. Subsurface fault chimney', which was also expressed in the surface gas
traces, which on seismic records did not visibly affect data. In this case the intensity of the geochemical
the post-Jurassic section, commonly generate marked signals over the subsurface structure was only slightly
surface signatures, whereas over suspected accumula- stronger than outside it. Here also, non-commercial oil
tions elevated but weaker signatures were observed was encountered higher than expected in the section,
(see Fig. 9). Similarly, salt-movement-controlled with the primary objective reservoir being water-wet.
sedimentary wedging at the assumed reservoir level These observations relate immediately to the aspect
was reflected in the near-surface data (see Fig. 10). of seal efficiency, on which very little published infor-
For this case the possibility of salt-induced fracturing mation is available. Consistently high hydrocarbon
as a cause for the strong surface signature has to be yields over a seismic feature should be followed up by
considered as an alternative to up-dip micro-seepage. a close check for indications of small-scale faulting
The most consistent parameters for the structure- penetrating the assumed seal.

PROPANE •
fi"c, METHANE PENTANE -0-,0-
l ao (~pbl 18 (ppbl
320 (nO> UVF 365(nml
~ 16 16
e ®

Fig. 9. Geochemical traverse for detailed structure related surface investigations (distance between sample stations 250 m).
Surface Geochemical Exploration for Hydrocarbons 231

Fig. 10. Survey pattern for detailed investigations (distance between sample stations 250 m).

CASE HISTORY-INTEGRATION OF ity between the plotted components methane through


GEOCHEMICAL SURF ACE DATA WITH propane is maintained.
GEOLOGICAL MODEL (TRfENABANKEN Delta 13C1 values span a wide range, where the bulk
AREA) is centred on values >-40%0 (PDB) (see Fig. 2(b».
Only a few data points are < -40%n (PDB); they plot in
As an example of a regional interpretation integrated the 'oil window' section of the Bernard plot (see Fig. 12).
with available subsurface information, results of a A synoptic interpretation of the statistical data indi-
survey over the Trrenabanken area (north of 62 ON) cates the following:
are presented below. -overall low yields combined with the 'heavy' isotopic
signature (>-37%) for the majority of samples seem
Data statistics
to suggest generally unfavourable source conditions
The Trrenabanken area is characterized by low hyd- (immature type III kerogen or alternatively overma-
rocarbon yields compared with other areas (see ture type II kerogen);
Fig. 2). In the frequency distribution a bimodality -the linear relationship between gas components up
becomes increasingly better expressed from ethane to propane suggests co-genesis and co-migration of the
through butanes (see Fig. 11). The bimodality is seen C1 to C:1 components and
in the cross-plots of butanes versus the other compo- -the deviation from linearity of the butanes and the
nents (see Figs. 7(c), (e) and (f)) whilst a good linear- disproportion ally high concentration of butanes in the
232 Petroleum Geochemistry in Exploration of the Norwegian Shelf

TRAENA BANKEN io,!T!!!RAE!:!N~A..2BAN=KE~N--.-_ _ _ _ _ _ _ _----,

10 100
C2-Y1ELD PPB

TRAEHA BANKEN TRAEN, BANKEN

>-
u
~l~
o
w
a:
LL

oj]
I
1 11::>
tt.":.1 n nr
P~~
'00
10 C3-YIELD P~~O "1 C4- Yl ELD

TRAENA BANKEN TRAENA BANKEN

2
>- >-
u u
z z
~I~ w
::>
o
w
a: '"
w
LL ~l

O+---r~rM~~~+++h~~~~~
nrfmn C5-YIELD P~~
O+--rY-~~~~++~~~ro,,~
.1 I 100 .1 C6- Yl ELD P~~ 100

Fig. 11. Frequency distribution of sediment gas components (ppb, w/w). Tr:enabanken area.

few samples with 'light' isotopic ratios « -40%) sug- of 0.8, still wet compared, for example, with Halten-
gests either a more favourable source rock facies (type banken values ex = 0.86), is assumed to be related to
I/I! kerogen) and/or 'oil window' maturity of the the low yields. The greater mobility of methane and
source rock. ethane would tend to leave low concentration
'residues' of the higher homologues. In productive
To resolve the ambiguity, the data set has to be inte-
areas with strong hydrocarbon signatures it has been
grated with a geological model.
observed that the surface signatures generally appear
'drier' than the reservoir gases. This is thought to be
Mapping of geochemical data
due to the replenishment of methane from either
Analytical data or simple ratios are contoured (see mature source rocks or from hydrocarbon accumula-
Table 1). Methane, the most mobile gas component, tions. Alternatively, the signatures may be caused by
exceeds the regional background of 30 ppb as an E-W increased solubility of the C"C 4 components in the
aligned anomaly in the south of the survey area. A reservoired hydrocarbons.
second trend of elevated values runs SW-NE across The same systematic areal changes as in gas con-
the central and northern parts (see Fig. 13). centrations emerge from the distribution of ,,13C l val-
Elevated butane yields essentially mirror the ues (see Fig. 15). The areas with high gas yields and
methane trends, but have their maximum in the south elevated wetness coincide with light isotope ratios.
(see Fig. 14). Consequently, the distribution of the The background methane concentrations «30 ppb)
wetness parameter Cd LC n demonstrates a combina- towards the north are too low for a reliable determina-
tion of both the methane and the butane anomalies, tion of the stable carbon isotope ratio. However, in
where the 'wettest' signature coincides with the high samples with elevated methane yields the methane is
yield area in the south. The regional background value also isotopically heavy (>-37%) in the northern and
TABLE 1. Analytical data-Trrenabanken (b) Gas yield (ppb), Trrenabanken, Norwegian
(a) Relative gas composition, Trrenabankan, Norwegian Sea Sea

Sample No. G 1 (%) C 2 (%) C 3 (%) n-C 4 (%) iso-G 4 (%) n-G 5 (%) "lac 1 (ppt) Sample No. Cl C2 G3 C4 C5 C6
tI 66.1 5.89 2.81 12.40 2.85 2.21 -38.7 tl 59.8 10.00 7.00 50.063 8.993 0.923
t2 66.1 5.62 2.86 14.41 2.90 2.24 -45.2 t2 46.0 7.34 5.48 43.889 7.008 0.523
t3 65.7 6.54 3.15 12.60 3.17 2.60 -44.1 t3 43.5 8.11 5.73 37.821 7.729 0.462
t4 86.2 2.92 2.67 2.51 1.10 1.14 -38.8 t4 32.1 2.04 2.73 4.871 1.907 0.000
t5 54.6 4.81 3.02 24.73 4.67 2.16 -35.9 t5 18.7 3.09 2.85 36.516 3.695 0.552
t6 74.1 4.53 2.20 12.17 2.31 1.49 -37.2 t6 74.7 8.56 6.10 52.926 6.751 0.000
t7 85.9 3.07 0.97 3.30 1.21 9.94 -38.6 t7 35.0 2.35 1.09 6.672 1.962 0.000
t8 62.5 5.70 3.29 10.64 3.07 7.98 0.0 t8 41.6 7.10 6.01 33.037 24.284 5.209
t9 56.6 11.19 5.99 7.82 2.42 4.14 0.0 t9 42.9 15.91 12.49 28.141 16.658 0.000 tn
no 86.7 2.98 3.14 2.14 1.01 1.12 -37.6 tlO 66.4 4.28 6.62 8.753 4.685 0.000 .""l
tIl 75.3 6.07 3.97 3.29 1.80 1.80 -39.4 t11 36.0 5.44 5.22 8.821 4.661 0.000 '"
f:i'
<">
-35.9 t12 0>
t12 82.3 7.73 2.34 1.87 1.00 1.20 29.6 5.21 2.31 3.740 2.374 0.000
U3 76.4 6.83 2.89 3.60 0.47 2.17 -35.3 tI3 22.5 3.77 2.34 4.340 3.635 0.000 ~
tl4 78.9 6.08 3.16 2.60 1.06 1.33 -36.8 t14 24.9 3.60 2.75 4.192 2.356 2.492 0
'"
<">
t15 78.8 6.91 3.11 2.80 0.53 2.29 -35.8 t15 28.0 4.60 3.04 4.288 4.103 0.000 ;;s-
O>
tl6 76.7 8.29 4.39 2.26 1.43 1.54 -38.0 t16 27.3 5.54 4.30 4.765 2.465 0.000
-41.0
~.
tl7 83.4 6.35 2.95 1.85 0.95 0.94 t17 51.6 7.37 5.02 6.280 3.141 0.000
t18 76.6 6.68 2.73 1.41 0.95 1.67 -35.7 t18 43.8 7.16 4.29 4.892 4.780 8.717
'£."'
t19 73.5 6.44 2.76 1.64 0.97 1.65 -35.5 t19 27.5 4.52 2.84 3.538 2.772 8.530 t>:l
H
t20 73.2 7.82 3.09 1.96 1.38 1.66 -35.9 t20 41.7 8.35 4.84 6.892 4.246 10.875 "l::l
t21 75.7 7.50 3.42 1.72 1.10 1.84 -37.4 t21 75.4 14.16 9.37 10.188 8.240 11.928 C
t22 80.4 5.37 2.46 1.22 0.62 1.34 0.0 t22 52.4 6.55 4.40 4.342 3.920 9.608 ~
~

t23 85.5 5.02 2.21 1.53 0.91 1.15 -34.2 t23 44.8 4.93 3.18 4.633 2.707 0.000 '".
0
;oS
t24 72.8 6.63 2.61 1.86 0.94 1.97 -34.7 t24 26.6 4.54 2.62 3.704 3.230 7.481
t25 67.6 4.74 2.42 0.88 0.00 1.77 -31.5 t25 13.8 1.81 1.36 0.651 1.624 10.663 'c>
""l
t26 71.1 5.90 3.12 1.64 1.26 1.98 -35.2 t26 19.9 3.10 2.40 2.942 4.929 5.092 ::t::
t27 75.3 6.39 2.66 1.68 0.94 1.54 -36.3 t27 31.6 5.03 3.07 3.990 3.096 6.200
t28 69.4 6.08 2.73 1. 79 1.06 1.88 0.0 t28 26.3 4.32 2.84 3.911 3.198 10.911
""R.
C1
t29 85.3 6.48 2.40 1.37 0.61 0.00 -35.2 t29 17.6 2.50 1.36 1.477 0.000 1.049 <">
t30 83.6 4.52 1.69 0.94 0.47 0.00 -34.7 t30 40.1 4.07 2.23 2.454 0.000 9.690 ""l
0-
'"
t31 79.8 6.35 2.74 1.46 0.91 1.46 -37.8 t31 102.8 15.34 9.71 11.067 9.319 2.696 ~
Co
t32 65.4 6.15 3.30 11.61 3.25 3.76 -40.3 t32 39.5 6.97 5.48 32.541 10.206 2.983
t33 74.6 6.21 6.09 3.46 1.22 2.06 -35.0 t33 30.3 4.73 6.80 6.892 4.345 0.000
t34/1 77.6 7.03 2.88 3.51 1.58 1.52 0.0 t34/1 25.3 4.29 2.58 6.013 2.958 0.000
t34/2 84.2 7.38 3.01 1.01 0.48 0.39 -39.2 t34/2 40.7 6.69 4.00 2.610 1.064 0.000
t35/1 70.1 6.24 3.08 10.90 2.45 1.76 -38.6 t35/1 64.8 10.82 7.84 44.774 7.317 0.000
t35/2 76.6 7.05 1.90 3.21 4.18 0.00 -32.0 t35/2 30.2 5.21 2.06 10.563 0.000 0.000
t36 64.5 6.26 3.30 11.11 3.25 5.26 -42.8 t36 34.1 6.20 4.79 27.501 12.487 3.885
t37 80.6 7.78 3.00 0.51 1.49 2.05 -39.2 t37 28.5 5.15 2.91 2.561 3.650 0.000
t38 78.3 6.75 3.06 3.17 1.45 2.17 -38.7 t38 34.0 5.49 3.65 7.261 4.831 0.000
t39 70.4 6.24 2.59 1.58 0.89 1.72 -32.6 t39 47.0 7.81 4.76 5.978 5.160 14.942
t40 76.4 6.09 2.55 1.48 1.10 1.50 -35.8 t40 28.6 4.27 2.62 3.497 3.831 0.281
t41 77.9 6.89 3.20 1.60 1.04 2.11 -35.1 t41 73.6 12.21 8.32 9.045 10.065 4.058
t42 77.9 6.93 2.83 1.86 0.80 1.37 -33.8 t42 51.2 8.54 5.11 6.337 4.045 10.122
t43 76.5 5.41 2.46 1.61 1.02 3.49 -35.7 t43 24.8 3.28 2.19 3.086 5.746 4.153

~
C>O
C>O
234 Petroleum Geochemistry in Exploration 01 the Norwegian Shelf

,,.J---

"J
..J
\

,,~
,

Fig. 12. Bernard Diagram, Trrenabanken area.

I
I

W4Sr-1f-
66"15,660917
_11 /
10'
I /

/
\ 19
1\ zo
10
-1--::::_----+20
~ ~ ,IV
t/
-:J-------'-' 30
J
30

t
:,,",-J.
I
,
,

Fig. 14. Isoconcentration map: butane, Trrenabanken.


6~~---+6,.,610,"-,17 _2)

-a major fault zone, steeply downthrowing the


Mesozoic section towards the Tram Basin in the
north-west.
For an assessment of source maturity, model calcu-
lations were performed, based on a modified Lopa-
tin-Waples approach using geothermal gradients
extrapolated from the Haltenbanken area. According
to this calculation the Kimmeridge Clay as the poten-
tial oil source rock should barely reach a maturity of
0.7% vitrinite reflectance in the deepest parts of the
Helgeland Basin (see Figs. 16 and 17).
None of the five wells drilled so far in the
Trrenabanken area seems to have had any significant
Fig. 13. Isoconcentration map: methane, Trrenabanken. indications of hydrocarbons. Explanations put for-
ward (Curran, 1983; Karlsen, 1984) ascribe this to the
regional immaturity of the Kimmeridge source rock.
northwestern parts of the area, suggestive of a The results of the open grid (15/7-5 x 15 km)
deterioration of source quality. geochemical survey of the Trrenabanken area seem
A compilation of subsurface information is pre- to corroborate the geological reasoning by outlining
sented in the cross-section of Fig. 16. the marginally mature, but apparently rich, source
On the structural map of base Jurassic (see Fig. 17) rocks in the Helgeland Basin. The lack of signif-
the main features are: icant hydrocarbon migration towards the Mesozoic
-a general SW-NE structural trend; structural terrace is borne out by the drastic drop in
-the Helgeland Basin as a potential drainage area in concentration of the C2+ fraction in the sediment gas.
the south and south-east; The methane anomalies, comparatively isotopically
-a structural and depositional terrace abutting the heavy, are tentatively related to Coal Unit-sourced
hasement high of the Nordland Ridge and gas. Their alignment approximately parallel to the
Surface Geochemical E xp loration for H ydrocarbons 235

-35.2

I
I - 35

-3 .2 \ \.0
-3lE - ,:4.7J\
-_./

NOT PROSPECTIVE
(DRY GAS SIGNAT URE)

66']0' ~'!-I b609/5 -359

3, b1C
-
- -35.3
o

-3 .9

m HC-GENERATIC'I
FROM TYPE II K:- "OGEN

F':777l
tL.L:::.J F~-
H GENERATIS',
OM TYPE III KEROGEN
Fig. 15 Is
. a-ratios ,5 13C 1 (%0
L- km
==i
10
PDB) T rrenabanken,
236 Petroleum Geochemistry in Exploration of the Norwegian Shelf

c, !ppbl
50

40
BUTANE 'YIELD
30
6'3 C,
20 45

40

°f-:_3~6-.:--.::::',.....:::....---:::3S----:-39:---c-47"0------_3-9---39-----,,---~35
-33

66091111
NW SE
<0., '" ""',-'_ _ _--'!'_ _ _ _ _"'--_ _ +_-""-____---'jl_ _ _ _ ---'~ _ _ _ _ ~_ ______, ...

1=- :::::",,:::::'1
L - - - - -__________ ----l

Fig. 16. Geological section based on seismic interpretation combined with available well information, Trrenabanken area.

SW-NE fault trends on the southeastern side of the interpretation of geochemical surface data is essential.
Nordland Ridge is characteristic for fault-controlled Interpretation performed in close co-operation with
migration. The few sample stations with very low gas geophysicists and geologists will contribute towards
yields north-west of the Nordland Ridge, despite the assessing:
marked faulting, would indicate a poor hydrocarbon
-presence and maturity of source rocks;
potential for the sedimentary section towards the
-possible migration fairways from kitchen to struc-
Trren Basin. Even when overmature, source rocks at
tural/ stratigraphic play;
depth could be expected to have generated strong
-presence and type of hydrocarbons in seismically
positive yield anomalies of isotopically heavy methane
defined structures and
in the near surface sediments. However, experience in -seal integrity.
the Haltenbanken area has shown that interpretations
of survey results cannot be extrapolated beyond the Obviously these positive aspects cannot be universally
area with actual data control. For the Trren Basin too realized. They are well-balanced by negative factors
little structural and stratigraphic information is avail- that render data acquisition and/or interpretation dif-
able to permit definite conclusions concerning its ficult or impossible:
prospectivity.
-the conventional sampling technique of gravity
coring of subbottom sediments is rendering ineffective
by an indurated or gravelly/sandy sea-floor;
CONCLUSIONS
-if abundant organic material (recent and/ or
reworked) is incorporated in the near surface sedi-
Applying geochemical surface investigations offshore
in the manner described above provides an apprecia- ments, e.g. in tropical near shore environments and
-elevated water temperatures favour microbial activ-
tion both of its use as an additional tool in hydrocarbon
exploration and of its present limitations. The possibil- ity with resulting modification of gas composition and
isotopic ratios of methane.
ity of stable carbon isotope measurements on very
small gas quantities has significantly reduced Despite the use of state-of-the-art analytical methods
ambiguities as experienced in earlier interpretations and established scientific principles, resulting data are
of compositional data only. Refinements in UVF by frequently 'noisy' with unacceptably wide scatter.
analysing a fluorescence spectrum of aromatic compo- Further improvements in analytical techniques such
nents provide added support for interpretation. A as measurement of D/H ratios, routine measurement
close relationship between subsurface structure and of (l13e of ethane and propane, and chromatography of
surface geochemical manifestations has been liquid extracts are required in order to improve con-
observed. sistency and reliability of results. Similarly, sample
Empirically established relationships between acquisition would have to be modified to overcome the
measured isotope ratios of methane/ethane of reser- indicated shortcomings, to improve efficiency and to
voir gases and maturity and type of respective source widen the application of the method.
rocks are, with certain limitations, valid for near- Properly conducted and interpreted geochemical
surface gas. surveys have the potential for reducing the source-
The need for a thorough integration of all available and migration-related risk elements in both regional
subsurface information in order to achieve a realistic and play-related exploration. Even at costs which
Surface Geochemica
t E xptoration for Hy 237
drocarbons

~ BASE JURASSI(
L- LJ wIThin top 01\ ...lOdow

ttrilllli\ ~~~nlR~0!l~[ oil window


A- B ~~:~:~ ~~~!IO~H81_02_1.191

F; g. 17. s,m,""'') in"'''''''''''''


of b" " J", "" "
wi,b m.m ,'m,"u
Trrenabanken area. m f" ,'U "" ", d """' ," " hy
"""",""n kitchon,
approach those of ap pr ec ia te th e
seism pe rmission of th e gr
cation as an additio ic data acquisition, th ei r appli- Saga_Statoil-Nor oup
nal exploration to
ol is considered sk Hydro-Norsk consisting of
N or ge -N or sk Ag e Shell-Arco
an advantage. ip
stable carbon isoto -Deminex Norge to publish
pe da ta of gases fro
discovery. m the Midgard
Gas analyses were
ACKNOWLEDGEM Manfred Schmitt performed by GCA-Laboratory,
ENTS trophotometry by, Le hr te , FRG and UVF spec
D -
The permission by
Deminex managem tu te for Geoscienc r 1. Dumke in the Fe de ra l Insti-
this paper is gr es an d Re sources (BGR), Ha
atefully acknowled ent to publish FRG. nnover,
ged. We also
238 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Horvitz, L. 1980. K ear-surface evidence of hyrlrocarbon environ-


REFERENCES ment from depth. In: Roberts, W. H., III and Cordell, R. J.
(erl~.), Problems of Petroleum Migration. AAPG Stndies ill
Barker, C. 1984. Development of abnormal and subnormal pres- Geology No. 10, pp. 241-269.
sures in reservoirs containing bacterially-generated gas (ab- Horvitz, L. 1981. Hydrocarbon geochemical prospecting after forty
stract) In: AAPG Research Conference 'Geochemistry of Natural years. In: Proc. 2nd Southern Jlethodist University Dallas
Ga~es', San Antonio, 12-15 February 1984 (unpublished). Symp. 'Unconventional Methods in Exploration for Petro leu III
Bernard, B. B. 1978. Light hydrocarbons in marine sediments. and Natural Gas II', pp. 83-95.
Thesis, Texas A & M University College Station. Jones, V. T. and Drozd, R. J. 1983. Predictions of oil and gas poten-
Bernard, B. B. 1980. Geochemical prospecting for petroleum and tial by near-surface geochemistry. AAPG Bull., 67, 932-952.
natural gases. US Department of Energy Symp. St Petersburg, Karlson, W. 1984. Trrenabanken's prospectivitet. Oral presentation
Florida, 15-17 December 1980 (unpublished). at Norwegian Petroleum Society (Oslo Chapter) Meeting, 15
Brooks, J. M., Kennicutt, M. C.,II, Barnard, L. A., Denoux, G. J. March 1984 (unpublished).
and Carey, B. D. Jr 1983. Applications of total scanning fluores- Kartsev, A. A., Tabasaranskii, Z. A., Subbota, M. 1. and Mogilevskii,
cence to exploration geochemistry. Contrib. No. OTC 4624, 15th G. A. 1954. Geochemical Methods qf Prospecting and Explora-
Annual Offshore Technology Conf., Houston, pp. 393-400. tion for Petroleum and Natural Gas. English translation by
Cornford, C. and Dyson, A. V. R. 1984. Seeps Point the Way. The Witherspoon, P. A. and Romey, W. D. 1959, University of
Role of Surface Prospecting in Oil and Gas Exploration. Hyd- California Press, 349 pp.
rochem Surveys International Ltd, sales brochure, 15 pp. Laubmeyer, G. 1933. A new geophysical prospecting method,
Curran, P. M. 1983. Modelling of hydrocarbon generation, especially for deposits of hydrocarbons. Petroleum, 29, 1-4.
Trrenabanken, Norway. Oral presentation at Geological Society Leythaeuser, D., Schaefer, R. G. and Yiikler, A. 1982. Role of
of London Petroleum Group Meeting 20 September 1983 (unpub- diffusion in primary migration of hydrocarbons. AAPG Bull., 66,
lished). 408-429.
Faber, E. and Stahl, W. 1983. Analytical procedure and results of Philp, R. P., and Crisp, P. T. 1982. Surface geochemical methods
an isotope geochemical surface survey in an area of the British used for oil and gas prospecting-a review. J. Geochem. Explora-
North Sea. In: Brooks, J. (ed.), Petroleum Geochemistry and tion, 17, 1-34.
Exploration qf Eumpe. Geol. Soc. Spec. Publ. No. 1Z (Blackwell Roberts, W. H. 1980. Design and function of oil and gas traps. In:
Scientific, Oxford.), pp. 51-63. Roberts, W. H. and Cordell, R. J. (eds.), Problellis of Petroleulli
Faber, E. and Stahl, W. 1984. Geochemical surface exploration for Migration. AAPG Studies in Geology, No. 10, pp. 217-240.
hydrocarbons in the North Sea. AAPG Bull., 68, 363-386. Schink, D. R., Guinasso, N. L., Sigalove, J. J. and Cima, B. E.
Faber, E., Rehder, S. and Stahl, W. 1983. Surface sediment gas 1971. Hydrocarbons under the sea-a new survey techno signe.
data from petroliferous regions of the North Sea and offshore OTC Contribution No. 1339, Srd Annual Offshore Technology
California. Erdal und Kohle-Erdgas-Petrochemie, 36, 357-360. Conference, Houston.
Faber, E., Whiticar, M. J., Stahl, W. and Schiener, E. J. 1984. Schoell, M. 1983. Genetic characterization of natural gases. AAPG
Biogenic methane in gases from oil reservoirs? (abstract). In: Bull., 67,2225-2238.
AAPG Research Cor(ference 'Geochemi.~try of Natural Gases', Schowalter, T. T. 1979. Mechanics of secondary hydrocarbon migra-
San Antonio, 12-15 February 1984 (unpublished). tion and entrapment. AAPG Bull., 63, 723-760.
Gerard, R. E. and Feugere, G. 1969. Results of an experimental Simoneit, B. R. T., Mazurek, M. A., Brenner, S., Crisp, P. T. and
offshore geochemical prospecting study. In: Schenck, P. A. and Kaplan, 1. R. 1979. Organic geochemistry of recent sediments
Havenaar, 1. (eds.), Advances in Organic Geochemistry from Guaymas Basin, Gulf of California. Deep Sea Res., 26A,
Pergamon Press, Oxford, pp. 355-371. 879-891.
Gevirtz, J. L., Carey, B. D. Jr and Blanco, S. R. 1983. Surface Sokolov, W. A. 1955. Die Gasvermessung. VEB Verlag Technik,
geochemical exploration in the North Sea. Methodology with an Berlin, 228 pp.
example. In: Brooks, J. (ed.), Petroleum Geochemistry and Stahl, W. and Carey, B. D., Jr 1975. Source-rock identification by
Exploration of Europe. Geul. Soc. Spec. Publ. No. 12 (Blackwell isotope analysis of natural gases from fields in the Val V ~rrle and
Scientific, Oxford.), pp. 35-50. Delaware basins. West Texas. ehell/. Geol., 16,257-267.
Horvitz, L. 1939. On geochemical prospecting. Geuphysics, 3, Thompson, A. B. 1910. Petroleum Milling and Oil Field Develop-
108-109. ment. Crosby, Lockwood and Son, London, 362 pp.
20 Geochemical exploration on the
Norwegian Continental Shelf by
analysis of shallow cores
R. H. Emmel*, M. Bjor¢yt and G. van Grasst
*Texas Eastern Corporation, Houston, Texas, USA
tContinental Shelf Institute (IKU), Trondheim, Norway

A surface geochemical exploration survey has been carried out on 250 (near)-surface sediments which were
collected in Quadrants 34 and 35 on the Norwegian Continental Shelf. This paper reports the results of the occluded
and adsorbed gas analyses and the relationship between the gas data and several sediment parameters.
Variations in occluded gas concentrations and composition correspond to variations in sediment characteristics.
This, coupled with the presence of significant amounts of unsaturated compounds, indicates that the occluded gas
mainly has an in situ origin. Adsorbed gas data are not strongly related to sediment characteristics and neither
does this gas fraction contain significant amounts of unsaturated compounds. This indicates the absence of a major
contribution of in situ produced hydrocarbons to the adsorbed gas.
A preliminary interpretation of areal distributions of adsorbed gas concentrations shows some areas with high
concentrations which can be related to different hydrocarbon sources in the deeper sediments. al3C values of
methane to a certain extent correspond with the adsorbed gas concentrations. Finally, the combination of a seismic
line with geochemical variations at the surface shows the relationship between changes in the deeper sediments
and hydrocarbon data at the surface.

INTRODUCTION and 35 on the Norwegian Continental Shelf. A total of


250 sediment samples was collected along a regular
Surface geochemical exploration for oil and gas is grid. These samples have been subjected to extensive
based on the concept that no reservoir cap rock is analysis including adsorbed and occluded (interstitial)
impermeable, and that hydrocarbons can escape from gas content, carbon isotope ratio of methane, sedi-
the reservoir and migrate towards the surface by a ment parameters such as grain-size class distribution,
combination of diffusion and effusion mechanisms. total organic carbon and carbonate content. Although
Hydrocarbons reaching the surface can be measured a wider range of analytical techniques has been emp-
directly-both in the sediment itself and in the overly- loyed (see Experimental section) the discussion in this
ing air or water-or through geochemical changes paper will focus on the occluded and adsorbed gas data
they induce (Gevirtz et al., 1983; Jones and Drozd, and their correlation with the occurrence of different
1983; Philp and Crisp, 1982). types of sediments in order to elucidate the extent to
Most reported studies on organic geochemical which variations in hydrocarbon gas data are influ-
exploration concern the analysis of gaseous hydrocar- enced by variations in sediment characteristics. Addi-
bons in recent sediments because these compounds tionally, some areal distributions of hydrocarbon con-
migrate easier than the larger, liquid hydrocarbons centrations which are thought to originate from
(Sokolov et al., 1970). In marine sediments two gas migrated compounds are briefly discussed and the
fractions are generally distinguished: (1) interstitial correlation between seismic data and geochemical var-
gas which is dissolved in the pore water and can be iations observed at the surface is shown.
released by mechanical disintegration of the sediment
(Kvenvolden et al., 1981) and (2) adsorbed gas which
is bound on mineral surfaces and can be released by
acid-treatment (Horvitz, 1939, 1980). If the interstitial
EXPERIMENTAL
gas is not removed first then the gas liberated upon
Sampling
acidification of the sediment is often referred to as
total gas (Faber and Stahl, 1983, 1984). The samples were collected along a square grid with a
This paper reports the results of a surface geochem- spacing of 4.5 km (see Fig. 1). Sample stations are
ical exploration survey carried out in Quadrants 34 identified by a four-digit code with the fIrst two digits
Petroleum Geochemistry in Exploration of the Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 239-246.
240 Petroleum Geochemistry in Exploration of the Norwegian Shelf

()2 00 00 02 50 00 out on adsorbed gas methane using a similar proce-


dure to that of the GCA laboratory in Lehrte (West
Germany).

GC conditions
C j -C 5 + analysis
This analysis was performed on Carlo Erba Fractovap
2150 and 2350 gas chromatographs equipped with
2 m x ~ inch stainless steel columns filled with
!!! ...., i Porapack Q on Chromo sorb using nitrogen as carrier
"'
'f" gas. The oven temperature was 150 DC. After elution
N

;;;
! '--., of n-butane the column was back-flushed and C5 + was
'''-, recorded. A standard gas containing methane, ethane,
r,,! II! I "11.,"
propane, n-butane, n-pentane and n-hexane was used
, ,
..
_ ~/_._._ ... _ •. _o-._o-._+-._o--._ ...... _a-._

+ for quantification.

.
II II II., II

, , CCC IO analysis
-~6~
,
;
g , g CCC IO analysis was carried out on a HP 5880A gas
;
o
o
o
o chromatograph equipped with a 50 m x 0.2 mm (Ld.)
;;;~34~/l~O__~~~__~~~~~____~~__ -+;;; fused silica column coated with OV-1Ol. Helium was
02 00 00 02 50 00
••••• GEOGRAPHICAL (EII5Tl •••••
O~ .0 00
used as carrier gas at 1 mlfmin. The inlet split ratio
was 1: 60. The temperature program was 35 DC
Fig. 1. Sample grid and bathymetry of survey area. Sample (5 min)-8 DC/min-200 DC (2 min), which gave good
stations are indicated by small squares. Concession blocks resolution from Cg upwards. Quantification was car-
are outlined by broken lines and numbered. ried out using the same standard gas as for C j -C 5 +
analysis.

increasing from west to east (01 to 21) and the last two
digits increasing from south to north (01 to 22). The RESULTS AND DISCUSSION
sampling device used was a gravity corer equipped
with a 3 m barre!. Core liners were made of chemically The sampling grid (see Fig. 1) is located in Quadrants
inert polypropylene. 34 and 35 on the Norwegian Continental Shelf. The
On deck, the bottom part of the core liner was cut water depth varies between 350 m and 412 m for the
off and taken into an onboard transportable labora- major part of the area, but in the south-west the slope
tory. After removal of the liner, the outside of the core, of the Norwegian Trench is reached and water depths
which had been in contact with the liner, was scraped decrease to less than 220 m. The decreasing water
off and four samples were taken for analysis. One depth coincides with the occurrence of a thick carpet
sample for organic geochemical analyses (700-1000 g)
was put in a can. Distilled water was added to leave a
headspace volume of 100-150 m!. The can was sealed
and stored in a freezer at -80 DC until analysis. Two
samples (300-400 g each) were taken for isotopic
analysis. Samples were put in plastic bags, which
were then placed inside cloth bags. Two cloth bags
containing duplicate samples were tied together,
attached to a string and stored in liquid nitrogen. One
sample (200 g) was taken for grain-size analysis and
stored in a plastic bag.

Gas analyses
The cans were thawed out overnight. A septum was
attached to the can, a sample of the headspace gas was
taken and analysed for C), C2, Ca, i-C 4, n-C 4 and C 5+
(see below). If any C5 + was detected, a second sample
was taken and analysed for CCC lO compounds (see
below).
The can was opened, headspace volume, water vol-
ume and sample weight were measured, and subsam-
pIes for occluded and adsorbed gas analysis, water and
carbonate determination, and fluorescence spectros-
copy were taken. The remaining sample was stored in
a freezer at -80 DC. For occluded gas analysis about ••••• GEOORAPH ICAL (EAST) •••••

30 g of sample was milled in a ball mill together with Fig.2. Distribution of total organic carbon contents (wt%) in
water. The evolved gas was analysed as described for survey area. Values higher than the mean value (x = 0.6%)
headspace gas. All samples were analysed for their are indicated by contour lines. Values higher than the mean
adsorbed gas content using the method described by value plus one standard deviation (x + (J = 0.8%) are
Faber and Stahl (1983). Isotopic analyses were carried indicated by a shaded area.
Geochemical Exploration by Shallow Core Analysis 241

D34D

8
8

05 <to
••••• GECJ3R.KPHICAl {EAST) ••••• • ••.. GECGRAPHICAL (EASTl ••.••

Fig. 3. Distribution of carbonate contents (wt%) in survey Fig. 5. Distribution of sand contents (wt%) in survey
area. For explanation of contouring see Fig. 2 (x = 13%, area. For explanation of contouring see Fig. 2 (x = 9%,
x + (J = 17%). x + (J = 19%).

« 15%) and low in total organic material


(TOC < 0.6%). The northern and eastern parts of the
sample area contain post-glacial marine clays which
are finer (sand <10%), carbonate-richer (> 15%) and
have higher TOC contents (TOC > 0.6%) than the
glaciomarine clays.
The clay fractions from 50 stations have been sub-
jected to XRD analysis (see Table 1). All of the sam-
ples examined have a rather similar, although com-
plex, mineralogical composition. The frequency of
each mineral phase may, however, differ to some
extent. These differences are mainly a result of varia-
tion in grain sizes and early diagenesis. Any signific-
ant trend in mineralogical variations as a result of
geographical positioning is not registered, except with
respect to calcite, which varies in the way described
above for carbonate.
The results from the gas analyses will now be dis-
cussed and correlations between gas data and the
occurrence of the two sediment types will be checked
for. For reasons of clarity a few words are necessary
on the terms occluded and adsorbed gas that are emp-
••• •. GEOORAPH ICAL [EASTl •••••
loyed here. Occluded gas is released by mechanical
Fig. 4. Distribution of clay contents (wt%) in survey area.
disintegration of the sample (see Experimental sec-
For explanation of contouring see Fig. 2 (x = 54%, tion) and should be very similar to what other authors
x + 1l(J = 69%). indicate as interstitial gas. Adsorbed gas is released
by acid-treatment of the total sediment (termed total
gas by Faber and Stahl (1983».
of cover sands which limits the extension of the sampl- Areal distributions of occluded. gas results give dif-
ing grid towards the southwest. ferent results for methane compared with the C2-C 4
Figures 2-5 show the areal distributions of some hydrocarbons. Methane concentrations (see Fig. 6)
sediment characteristics: total organic carbon (TOC), are highest in the western part of the sample grid
carbonate, clay and sand content (all expressed as whereas the CC C4 hydrocarbons (see Fig. 7; the
wt%). In these and later figures the areas with values diagrams for propane and the butanes are very similar
between the mean and the mean plus one standard to that of ethane, which is shown) have the highest
deviation are contoured, while areas with higher val- concentrations in the eastern and northern part of the
ues are shaded. grid. A more detailed inspection ofthe gas chromatog-
The data in Figs. 2-5 reveal the existence of two rams reveals the presence of significant amounts of
sediment types in the sampling area. The southwest- unsaturated compounds. Due to insufficient GC-
ern part consists of glacio-marine clays, which are resolution, the alkenes are not detected as separate
relatively coarse (sand> 10%), low in carbonate peaks but in many samples peak shapes indicate their
242 Petroleum Geochemistry in Krp/oration of the Norwegian She!f"

TABLE 1. Mineral composition of the samples «63 11m fraction) as determined


by XRD. Values are expressed as percentages after normalization
Sample
ident. M:r.l Mica/ill Hbl Chi Kao Qtz K-jld Plag Cal Dol Sid Py
0110 2.7 27.6 2.0 7.7 6.4 30.2 5.5 6.2 7.0 2.8 0.0 1.9
0113 5.0 19.6 1.5 9.2 3.7 27.1 7.2 14.3 7.1 2.4 0.4 2.4
0214 1.7 23.3 0.0 9.3 4.7 36.4 2.9 8.7 7.3 2.6 0.5 2.6
0217 0.7 16.0 2.4 6.3 2.1 48.0 7.4 7.1 5.7 2.4 0.9 1.0
0219 3.3 9.3 0.0 3.0 6.0 41.9 7.9 15.0 6.2 4.4 0.0 2.9
0310 2.9 26.1 Ll 8.6 2.9 30.3 6.9 n.8 6.0 2.2 0.0 1.2
0315 2.9 28.4 1.8 7.2 7.2 29.7 4.4 7.0 8.5 2.3 0.0 0.5
0317 3.4 13.6 0.0 7.2 1.8 36.1 8.6 13.1 7.9 3.5 0.7 4.0
0407 2.1 19.9 0.0 7.6 3.8 38.0 4.7 11.3 10.2 1.4 0.2 0.9
0408 1.9 2Ll 0.0 12.9 3.7 29.9 7.7 12.7 7.4 2.2 0.0 0.5
0412 1.8 24.6 0.0 7.2 4.8 31.2 8.8 lO.O 7.8 2.7 0.0 Ll
0413 3.1 30.1 2.9 11.5 5.7 25.2 1.8 8.3 8.0 2.0 0.0 1.3
0418 2.9 17.7 1.3 6.4 3.2 41.0 7.5 11.5 5.2 3.3 0.0 0.0
0506 3.7 30.3 3.4 10.4 3.6 25.9 1.6 11.2 7.2 1.6 0.0 0.8
0516 2.1 23.6 4.8 6.8 6.8 25.6 8.2 11.3 5.9 2.5 0.0 2.4
0611 6.1 18.0 0.0 7.8 5.2 37.1 3.9 9.2 8.3 3.0 0.0 1.4
0707 2.9 22.8 0.0 7.6 5.0 37.3 3.9 10.1 6.5 2.8 0.0 1.2
0715 2.8 22.2 0.0 8.2 8.2 31.5 2.3 12.5 8.7 2.4 0.0 1.0
0717 4.1 19.1 1.4 4.2 2.8 34.1 5.5 13.2 11.2 2.8 0.4 1.4
0719 1.3 18.2 2.3 8.7 1.7 31.3 8.1 11.7 10.6 3.7 Ll 1.4
0804 3.3 13.8 2.3 4.7 1.9 34.3 20.6 9.6 5.2 2.6 0.0 1.6
0809 2.3 15.1 1.6 6.4 3.2 45.4 8.7 8.8 4.4 3.0 0.2 0.9
0814 3.5 19.2 0.0 6.8 4.5 37.6 5.5 11.3 7.4 3.1 0.0 Ll
0903 2.3 14.4 2.6 4.8 2.4 44.9 ILl 8.0 5.5 2.8 0.3 0.9
0911 3.6 19.5 5.3 8.6 2.1 31.3 8.5 ILl 5.0 3.1 0.3 1.6
0913 2.8 18.8 0.0 4.9 4.9 40.3 3.7 13.1 6.0 2.7 0.2 2.3
0918 3.0 22.6 0.0 7.0 3.5 33.9 7.5 9.8 7.7 2.8 0.2 2.0
0921 3.2 lO.8 1.6 6.7 1.5 45.2 5.9 9.5 9.5 3.7 0.4 1.9
1001 3.8 16.8 0.0 5.2 2.6 37.2 8.9 13.8 5.5 2.9 0.5 2.8
1004 2.2 12.9 4.8 6.1 2.0 46.1 6.6 9.4 5.7 4.3 0.0 0.0
1008 2.3 16.2 2.0 5.9 1.5 37.6 13.5 11.0 5.2 2.6 0.1 2.1
1016 5.2 21.9 Ll 6.4 2.6 34.6 7.0 8.5 5.9 4.7 0.5 1.6
1022 3.6 22.7 0.0 9.5 1.9 27.3 8.1 11.9 9.0 2.5 0.3 3.1
1105 5.8 8.7 3.1 3.7 3.1 43.9 4.6 15.5 6.9 2.5 0.2 1.9
1119 7.2 18.6 1.9 7.3 3.7 34.4 1.8 10.9 10.6 2.2 0.4 Ll
1212 9.2 21.8 0.0 7.0 2.3 27.2 5.4 11.0 11.9 2.2 0.4 1.6
1304 2.6 10.7 2.5 4.6 0.9 44.4 10.0 12.0 6.6 3.6 0.0 2.0
1317 2.5 13.7 1.0 5.8 1.4 40.0 7.7 12.9 10.3 2.2 0.0 2.5
1407 10.6 20.4 1.5 6.7 2.7 27.0 4.1 10.9 12.0 2.3 0.1 1.7
1414 5.9 11.3 2.0 6.5 2.2 3Ll 6.4 15.1 14.0 3.0 0.1 2.3
1501 4.0 13.5 0.0 6.0 1.5 36.9 8.1 13.2 8.9 2.9 0.6 4.5
1504 2.6 13.8 1.4 5.4 1.4 44.4 0.5 12.6 11.7 2.9 0.4 3.0
1610 6.9 14.9 2.2 5.3 2.1 31.4 7.5 ILl 13.9 2.5 0.4 1.9
1701 6.4 14.5 0.6 7.8 1.6 31.7 7.6 12.5 12.4 2.6 0.3 2.2
1706 6.5 13.9 Ll 6.0 2.0 33.7 6.2 14.4 11.4 2.6 0.2 2.0
1808 7.4 22.5 0.0 7.9 3.2 24.2 5.5 9.9 14.5 2.4 0.3 2.3
1905 5.1 18.3 2.0 6.7 2.7 35.0 4.1 10.2 11.9 2.8 0.2 1.0
2002 7.8 15.1 2.0 5.8 1.9 28.7 (i.9 10.3 17.3 2.5 0.2 1.6
2103 8.8 19.4 1.2 7.7 3.8 24.5 4.0 13.1 12.8 2.5 0.0 2.2
2107 3.0 8.3 2.5 4.2 2.1 45.5 9.8 14.9 4.3 3.0 0.4 1.9

Mx.1 = mixed layer clay minerals, Mica/ill = mica and illite, RbI = hornblende, ChI = chlorite, Kao =
kaolinite, Qtz = quartz, K-fld = K-feldspar, Plag = plagioclase. Cal = calcite, Dol = dolomite, Sid =
siderite, Py = pyrite.

concentration to be equal to or larger than the coelut- Thus, the interstitial gas in sediments with low clay
ing alkanes. The area with high C~-C4 concentrations and TOC contents contains more methane and less
(see Fig. 7) is thus also rich in unsaturated com- C2-C 4 hydrocarbons (western area), than is found in
pounds. sediments with higher clay and TOC content (eastern
The presence of C2-C 5 alkenes in surface sediments to northern band). Furthermore, the CZ-C 4 hydrocar-
has been reported by several authors (Horvitz, 1972; bons contain significant amounts of unsaturated hyd-
Hunt and Whelan, 1979; Whelan and Hunt, 1983). rocarbons. The higher amounts of these are associated
These compounds are thought to be the result of in with the larger TOC and clay values. This is inter-
situ bacterial activity. In fact, ratios of alkanes to preted as indicating a higher biological activity. The
alkenes are used to distinguish biogenic from pet- difference in methane concentrations between the two
rogenic gas (Cline and Holmes, 1977; Kvenvolden et gases cannot be explained by the effects of selective
al., 1981; Sandstrom et al., 1983). According to these degassing as described by Faber and Stahl (1984).
criteria our occluded gas has a significant contribution In summary, the variations of the occluded gas data
from microbial sources. A comparison of the sediment in this survey primarily reflect variations in the type
parameters (see Figs. 2-5) and the occluded gas data of surface sediment. Additionally, they contain a sig-
(see Figs. 6 and 7) shows that the two areas with dif- nificant contribution from microbial sources. For
ferent occluded gas composition roughly coincide with these reasons these data have not been used in further
the occurrence of the two sediment types. investigations.
Geochemical Exploration by Shallow Core Analysis 243

03 iO

..'.:-: ...•...:............. .

:@
i . L1
j a @!l D

"·-··'~"""''''I5:···-·-t·- ... ·-o-·-o.·-r·-D-·-II-·-


<f ,

.t
;
!
8
8

00 00
• • • • •GEOGRAPH I CAL (EAST) ••••• ••• , .GEOGRAPH ICAL I EASTl •••••

Fig. 6. Distribution of occluded gas methane concentrations Fig. 8. Distribution of adsorbed gas methane concentrations
(Ill gas/kg dry sediment). For explanation of contouring see (Ill gas/kg dry sediment). For explanation of contouring see
Fig. 2 (x = 141 Ill/kg, x + (1 = 215 Ill/kg). Fig. 2 (x = 979 Ill/kg, x + (1 = 1469 Ill/kg).

03 ~O 02 00 00 02 50 00 03 "to 00

_._._-_._. __, ..... _._ ...

¢,
.
,
,
! ,

0
0
!':
2l 2l
;;; ~~--------------~~~~~4-~~~~-+;;;
02 00 00 02 50 00 03 ~O 00 00 00
••.•• GEOGRAPH I CAL (EAST) ••••• • ; ••• GEOGRAPH ICAl (EIIST) •••••

Fig. 7. Distribution of occluded gas ethane concentrations Fig. 9. Distribution of adsorbed gas ethane concentrations
gas/kg dry sediment). For explanation of contouring see
(Ill (I.d gas/kg dry sediment). For explanation of contouring sep
Fig. 2 (x = 32 Ill/kg, x + (1 = 54 Ill/kg). Fig. 2 (x = 73 Ill/kg, ;i; + (1 = 108 Ill/kg).

The adsorbed gas compositions, yields and <5 13C of testing of the adsorbed gas alkanes (C 1-C 5) shows
adsorbed gas methane have been compared with the them to be correlatable and to represent a homogene-
data published by Faber and Stahl (1984). No evidence ous series. There is no strong correlation with the
of sediment degassing fractionation is detected. sediment parameters, such as exists for the occluded
Unsaturated compounds are rarely found, and those gas data. On the contrary, there is a moderate inverse
present were at a concentration approaching the sen- correlation between adsorbed gas concentrations and
sitivity of the instrument. The concentrations of the TOC, clay and carbonate contents. This indicates that
saturated species are much larger (5x) than was the adsorbed gas data are not primarily reflecting var-
measured for the occluded gases. Additionally, a iations in sediment characteristics but that they are
number of higher alkanes and aromatics are detected. influenced by other processes, among which is the
Figures 8-10 display the areal distributions of the migration of hydrocarbons from deeper strata.
concentrations of adsorbed gas methane, ethane, and Although it is not the intent of this paper to present
pentanes and higher compounds (C 5+). Statistical an in-depth interpretation of surface geochemical
244 Petroleum Geochemistry in Exploration of the Norwegian She(f

03 '0 00
Block 34/8 (see Figs. 8-10). Contrary to the earlier
case, all hydrocarbons (CCC lO) show high concentra-
tions in this area.
Carbon isotopic ratios (b I3C) of adsorbed gas methane
are presented in Fig. 11. Two value ranges are indi-
cated following the suggestions by M. Schoell (pers.
comm.) and related to the work by Faber and Stahl
(1984). The areas with vertical shading indicate bl3C
values between -38.2 and -36.2%0 (relative to PDB)
representing methane from a very mature source .

.(l
!
These areas coincide nicely with the observed occurr-
ence of high concentrations of adsorbed gas C I -C 3
1
17 b,

u.
alkanes (see Figs. 8 and 9) confIrming the suggested
gas-source origin. bl3C values ranging from -46 to
-40%0 are shaded horizontally. These values are indica-
tive of a less mature source for the methane (oil to
condensate). One large continuous area with these
values occurs in Blocks 35/8 and 35/11 (see Fig. 11). In
8 this area some condensate discoveries have been
8 g made. No distinct trend in isotopic values matches the
.. +-----~------~~ '""-.......-.........,'-----+ .. area with high adsorbed gas concentrations in Blocks
00 00
34/4, and 34/5 and 34/8 (see Figs. 8-10).
In order to illustrate that the surface geochemical
Fig. 10. Distribution of adsorbed gas C5 and higher data reflect processes occurring in the deeper sedi-
hydrocarbons (~l gas/kg dry sediment). For explanation of ments, Fig. 12 combines a seismic lines with the varia-
contouring see Fig. 2 (x = 9 ~l/kg, x + (1 = 17 .ul/kg). tions of some geochemical parameters at the surface.
The two parallel lines added to the seismic diagram
r:epresent a proposed oil window (150-300 OF),
02 50 00 03<000 which results from a statistically generated geother-
mal gradient (1.62 °F/100 ft), based on data obtained
from bottom hole well temperatures. The dark line
appearing as a 'wave' represents the Kimmerian
unconformity (with its associated Kimmeridge Shale).
This illustration shows that when the Kimmeridge lies
within the oil window, low molecular weight gas (Cl>
C2 and C3) concentrations remain low and higher

;i~-':::t~";r-- :
molecular weight species have a concentration which
is large (Block 34/7). When the source bed material
drops below the oil window (as in Block 34/9) the con-
tVoa til It i It
centration of low molecular weight species increases
and that of the higher order species decreases. This
trend continues through Blocks 34/9, 34/12, 35/10 and
'. II 'II ., II II II • •

~~:-:t-:1 ; '11,.0
35/11. The isotopic values also support this trend;
however, their numerical values are not in agreement
with other reported data (Faber and Stahl, 1984).
''', .
g 'II...,.! 8
8 8

00 00 CONCLUSIONS
•• "••• GEOGRAPH I CAL (EAST) •••••

One of the main objects of this study is the investiga-


Fig. 11. Distribution of (113C values (%n relative to PDB) tion of the relationship between sediment characteris-
of adsorbed gas methane. Values between -38.2 and tics and the amount and composition of the occluded
-36.2%0 are shaded vertically, values between -46 and
-40%0 are shaded horizontally.
and the adsorbed gas phases. The results of sediment
gas analyses can be affected by various sediment
parameters (grain-size distribution, TOe, etc.), and
this process interferes with the interpretation of sur-
data in an exploration context, several trends in face geochemical data for exploration purposes.
the adsorbed gas data are interesting enough to be dis- The area under investigation contains two sediment
cussed in some detail. types: glaciomarine clays-which are relatively
An area with higher concentrations of C I-C 3 coarse, low in carbonate and organic material-and
alkanes stretches from Block 35/10 to the eastern part marine clays-which are more fine-grained and have
of Block 34/9 (see Figs. 8 and 9). This area is less higher carbonate and TOe contents. The variations in
distinct in the diagrams of the higher hydrocarbons the occluded gas data (amount and composition) cor-
(see Fig. 10) which can be indicative of a gas source. respond to a large extent with the occurrence of the
This area is located directly north of the Troll Field. two types of sediments. Furthermore, the occluded
Another area with high adsorbed gas concentrations gas contains unsaturated hydrocarbons suggesting a
is located in Blocks 34/4 and 34/5, and continues into significant microbial contribution to this gas fraction.
Geochemical Exploration by Shallow Core Analysis 245

02 00 00 02 50 00 03 ~O 00 (a)

i5
~+-------~------~~~~~~----+-------+~
02 00 00 02 50 00 03 .0 00
..... GEOGRAPH I CAL (EAST I •••••

PENTANE (PLUS) METHANE FRACTION ((((C5) + ( C5))/ClIl000


35.84 -- --- - - - - ------ - -r-- - -- - - - -- - - - r - - -- - -------1--- --,-- ---- - -------- --r --- -- ----- -- - --- -- (b)
_ _ _ _ __ _ ___ : :
_ _____ L ________________ 1- _____________ -1 ______ -1._ __ :: i
_ ___________ .1 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

: i : I :
I '

~ ,
: : PROPANE METHANE FRACTION HeJ/e1) 1000) 1 !
40.32 - -------- ------ --T- --- --- ----- --,-- ------ ----- -"1 -- - -- 1---- ---- ------ ---- ..... -- --- ~ ------ - ----..,
I : : :
----~-----+--------
i
---------r--------
!
I
,
I
'
I I
,
I
25.51 - ----------- I ---- I -------i------ -------1

Y'-------------!-------------!----------:~--__t,-------------'-------------_i:
: l li
18.11

t------------- --,----- - -------T--- -- r- ------- ---- ---i-- -- --- --- --------:


: : ETHANE METHANE FRACTION ((C1IC1) 1000) ,
96.61 ,------------- - -- - -

-----------------~---------- ----I -- ----- ---4------~---- ------~-----------------~


69.95 --- :
---""1-------- -----1-----------
: : : --~-----------------j:
-----r------------: 1

56.62 '"""'""-----------=;----------~,-----------i----+---------------r-'-------------i:
I :

-37.29 1 -- - -
I
---1-- -- - -
CAR~ON ISOTOPIC SHIFT ((OEl13CI12C) PO B)
- - - - - ---~ - - - - -- --- ---r ----,- - - --- - - ----- - -
i !
T---- ------- - --- -\
I 1 1 I I
----- ------ -----~- ---------------- -- ----~-----~-------- -----1--------- ---------1
-39_81- ---------, : ---------- __ :I ____________ J1_____ ~---------------l---------- 1 ------J I

: I I i
-41.07f..,------------...;.....-----------i----------7--------;.,--------------+----------t~-"

,
BLOCK 14/7 BLOCK 34/8 BLOCK 34/9 BLOCK: BLOCK 35/10 BLOCK 35/11
: 34/12

Fig. 12. (a) Location of three seismic lines that have been compared with geochemical data. (b) Combination of seismic line A with
geochemical surface data. Line 1 in bottom diagram indicates the Kimmerian unconformity (base Cretaceous), lines 2 and 3
indicate, respectively, the top and bottom of the oil window. For further explanation see text.

As the occluded gas is to a large extent derived in situ, tribution of Ol3C of adsorbed gas methane corresponds
it has not been used for further interpretations. in some areas to the gas concentration data, but gives
The adsorbed gas fraction does not show a strong different results in other areas.
relationship with sediment characteristics, neither A combination of geochemical surface data with a
does it contain significant amounts of unsaturated seismic line shows that changes at the surface can be
compounds. Areal distributions of gas concentrations related to variations in the deeper sediments, thus
reveal the presence of several areas with high confirming that at least part of the hydrocarbons at
amounts of adsorbed gas. Depending on the hydrocar- the surface have migrated from below.
bon range which shows high concentrations, such In conclusion it can be stated that surface geochemi-
areas can be related to oil or gas sources. Areal dis- cal exploration is a useful tool that can yield informa-
246 PetToleum Geochemistry in Exploration of the Norwegian She!f

tion on the presence of hydrocarbons in the deep sub- Gevirtz, J. L., Carey, B. D. and Blanco, S. R. 1983. Surface
surface. However, due to several aspects which are geochemical exploration in the North Sea. In: Brooks, J. (ed.),
Petroleum Geochemistry and E.rplorat£on. of Europe. Geo/.
not yet fully understood, it should not be used by Soc. Spec. Pub. No. 12, pp.35-50. (Blackwell Scientific,
itself, but always in combination with other tech- Oxford.)
niques. Some particular subjects that merit further Horvitz, L. 1939. On geochemical prospecting. Geophysics, 4,
attention as indicated by this study are: (1) the influ- 210-225.
ence of sediment characteristics on hydrocarbon data Horvitz, L. 1972. Vegetation and geochemical prospecting for pet-
roleum. Bull. Am. Asse. Petroleum Geologists, 56, 925-940.
and (2) the observation that different analytical tech- Horvitz, L. 1980. Near-surface evidence of hydrocarbon movement
niques yield results which are sometimes similar, but from depth. In: Roberts, W. H., III and Cordell, R. J. (eds.),
are different in other cases. Research concerning Problelils of Petroleum Migration Am. Ass. PetroleulII
these problems is currently in progress in our respec- Geologists Studies in Geology 10,241-269.
Hunt, J. M. and Whelan, J. K. 1979. Volatile organic compounds in
tive laboratories. Quaternary sediments. Org. Geochem., I, 217-224.
Jones, V. T. and Drozd, R. J. 1983. Predictions of oil or gas poten-
tial by near-surface geochemistry. Hull. Alii. Ass. PelroleulII
ACKNOWLEDGEMENTS Geologists, 67, 932-952.
Kvenvolclen, D. A., Vogel, T. M. and Gardner, J. V. 1981.
The managements of Texas Eastern Corporation and Geochemical prospecting for hydrocarbons in the outer continen-
tal shelf, South Bering Sea, Alaska. J. Geochell/. ExploT., 14,
Texas Eastern Norwegian Inc. are gratefully acknow- 209-219.
ledged for their permission to publish these data. Philp, R. P. and Crisp, P. T. 1982. Surface geochemical methods
used for oil and gas prospecting-a review. J. Geochelil.
EJ:plm·., 17, 1-34.
REFERENCES Sandstrom, M. W., Meridith, D. and Kaplan, I. R. 1983. Hyd-
rocarbon geochemistry in surface sediments of Alaskan outer con-
Cline, J. D. and Holmes, M. L. 1977. Submarine seepage of natural tinental shelf: Part 2. Distribution of hydrocarbon gases. Ru II.
gas in Norton Sound, Alaska. Science, 198,209-219. Am. Ass. PetTOleUin Geologists, 67, 2047-2052.
Faber, E. and Stahl, W. 1983. Analytical procedure and results of Sokolov, V. A., Geodenkyan, A. A., Grigoryev, G. G., Krems, A.
an isotope surface survey in an area of the British North Sea. In: Y., Stroganov, V. A., Zorkin, L. M., Aeiddson, M. I. and Vain-
Brooks, J. (ed.), PetTOleUin Geochemistry and Exploration of saum, S. J. 1970. The new methods of gas surveys, gas investiga-
Europe. Geol. Soc. Spec. Publ. No. 1;2, pp. 51-63. (Blackwell tions of wells and some practical results. Can. Inst. Min. M etall.,
Scientific, Oxford.) 11, 538-543.
Faber, E. and Stahl, W. 1984. Geocehmical surface exploration on Whelan, J. K. and Hunt, J. M. 1983. Volatile C1-C s organic com-
hydrocarbons in the North Sea. Bull. Am. Ass. Petroleulil pounds in sediments from the Peru upwelling region. Org.
Geologists, 68, 363-384. Geochelll., 5, 13-28.
21 Regional geochemical analysis
of the southern portion of
the Norwegian Sector of
the North Sea
Joel L. Gevirtz*, B. D. Carey, Jr* and Stephen R. Blanco t
*Tenneco Oil E&P, Houston, Texas, USA
tTenneco Oil Co. Norsk A/S, Sandnes, Norway

To test the usefulness of surface geochemical methods as regional evaluation tools, a study was undertaken by
Tenneco Oil Company, comprising, in part, the southern quadrants of the Norwegian North Sea.
A suite of 345 samples collected from the bottoms of cores taken at block corners and centres over most
quadrants south of 60 oN was analysed for adsorbed light hydrocarbons (C l , C2+ and Cs), indicators of aromatic
compounds (fluorescence intensities at 320 nm and 365 nm wavelengths), and methane carbon isotopes. These
data, when mapped, indicate patterns coinciding with major Jurassic structural trends. Application of a variety of
bivariate and multivariate statistical techniques to these data to attempt to enhance the patterns further
demonstrates relationships between geochemical signatures seen at the surface and underlying subsurface
geology.
Geochemical parameters mapped at the smface represent a weighted average of generation and migration
occurring at depth. The signature seen at the surface most closely represents the dominant process occurring at
depth. At the sampling density used here, the data provides an idea of the type of hydrocarbons most likely to be
found in a !,riven basin. Because the sampling grid is widely spaced, the results of this study cannot be interpreted
in terms of hydrocarbon accumulation in specific structures, but instead should be viewed as a surface
representation of source rock and subsurface fluid phenomena occurring throughout the stratigraphic section.
Maps developed here, then, represent a summation of hydrocarbon generation throughout the stratigraphic
section, but weighted by dominant hydrocarbon-producing strata.
A regional study, such as the one described here, can provide important exploration data concerning the regional
hydrocarbon potential in a basin. A study of this kind should ideally precede detailed seismic shooting in a region so
that seismic acquisition can be concentrated in areas with the highest geochemical potential.

INTRODUCTION leakage points along geological discontinuities. There-


fore, proper interpretation of these data require their
Trace amounts of hvdrocarbons collected from near integration with geological and geophysical know-
surface marine sedi~ents and soils are believed to be ledge. In the past, surveys were designed to examine
indicators of hydrocarbons reaching the surface from potential reservoirs and have been limited in scope,
depth, and have been used by the oil industry as evi- often covering only features of immediate interest
dence of reservoired petroleum. However, until with a closely spaced sample grid. Because virtually
recently their origin has remained controversial. all samples contain some hydrocarbons and, based on
Recently, methane carbon isotopic ratios measured in existing evidence, at least some of these compounds
sea-floor sediments show values that strongly suggest originate by catagenic processes at depth, these data
a catagenic origin for these compounds. The processes can also be used within a regional context for basin
by which these compounds get to the surface and are evaluation. This study, then, intends to demonstrate
retained there in fine-grained sediments are poorly how near surface geochemical data can be used for
understood; nevertheless, observations presented by evaluation of regional hydrocarbon potential.
earlier workers (Horvitz, 1972, 1980; Rosaire, 1940; Earlier workers (Horvitz, 1939, 1980; Rosaire,
Scull and Myhre, 1976) suggest that anomalously high 1940) differentiate between 'anomalous' and 'back-
concentrations of hydrocarbons recovered from sedi- ground' yield values and, of course, are more
ments or soils below a level where pollution or other interested in the high yield or anomalous samples. In
man-made phenomena can be ignored, represent mic- this work, all values, high or low, are believed to carry
roseeps. Often these microseeps are indicative of pet- information related to subsurface conditions. Subtle
roleum accumulation, but they can also indicate high differences in source-rock facies and maturity are
Petroleum Geuchemi,tr-!J i'li E.,·p/omlion ofthr Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 247-261.
248 Petroleum Geochemistry in Exploration of the Norwegian Shelf

km
i
50

~---=-~--:~.--~:----:-h---~156'N
2'E 3'E 4'E 5'E 6'E 7'E

Fig. 1. Base map of Norwegian North Sea south of 59 oN, showing Jurassic structural elements and location of surface
geochemical samples.

reflected by different gas compositions and fluores- population. Loosely bound gases produce less reliable
cence intensities. In a regional context, no differentia- results. For this study, then, adsorbed gases will be
tion can be made between anomalous and background used exclusively.
concentrations; but instead, only the magnitudes and The study area covers southern Norwegian waters
their relationship to each other have meaning. south of approximately 59 ON (see Fig. 1) and is part
There are several different gas popUlations in soils of a larger study carried out in the Norwegian and UK
and sediments. These different populations often have sectors of the North Sea by Tenneco Oil Company in
different origins, and therefore convey different 1979. Within the South Norway portion of the study
interpretive meanings. Collection procedures area, 204 locations were sampled, generally at block
designed for each population have their own inac- corners and centres. Held acreage constraints prohi-
curacies, and often different populations cannot be bited regular sampling of Quadrants 1 and 2. Because
sampled in the same operation. Of importance to this these areas have an abundance of installations such as
study are two populations occurring in marine sedi- platforms and pipelines that have disturbed the sea
ments. One population of gases occurs within sedi- floor, it is doubtful that they would yield representa-
ment interstices, loosely bound, probably by Van der tive samples in any case.
Waals' forces, to grain surfaces. This population,
termed 'loosely bound' hydrocarbons by Horvitz
(1969) represents both gases that are now reaching
the surface and those being produced there. Because REGIONAL SUMMARY
the mechanical process by which samples are collected
from the sea-floor disturb the sediments and their Because it is tectonically and stratigraphically
interstitial water, gas is lost and the population is diverse, the Norwegian continental shelf is considered
never reliably censused. Also, because marine sam- to be a good test area for surface geochemistry. Var-
ples are collected from depths below the sedi- ied tectonic elements-including deep grabens, shal-
ment-water interface at which substantial anerobic low basins, structural highs and stable plat-
and/or methanogenic activity occurs, these gases are forms-yield a variety of source rock-reservoir rela-
often enriched in biogenic methane, and such mixtures tionships. Released well control is fairly evenly
are more difficult to relate to processes occurring at spread throughout the area, allowing calibration of
depth. surface geochemical data by correlation with available
The second popUlation, termed adsorbed gases by downhole geochemical information. Within the study
Horvitz (1972), are those gases sorbed within clay area, three major tectonic/stratigraphic features can
mineral interlayers (Barrer and MacLeod, 1954, 1955; be delineated: the Central Graben, the Vestland
Barrer and Perry, 1961; Eltantawny and Arnold, Arch/Fiskebank Platform, and the Egersund Basin
1972). This population is unaffected by mechanical col- (see Figs. 1 and 2). This provides an essentially
lection methods and storage, as evidenced by the fact tripartite division of the southern Norwegian North
that considerable chemical processing, designed to Sea into the barren Vestland Arch/Fiskebank Plat-
digest the clay lattice, is necessary to recover them. form separating two oil generating basins. The Cen-
Experience has shown that for marine sediments, tral Graben is characterized by major oil and gas
reproducible results are obtained for the adsorbed accumulations, thick sequences of mature and over-
Regional Geochemical Analysis in the North Sea 249

REFLECTOR' :
<D"'" IIIOCINE iii INTRA JUIIAS
PROFILE I%> T. I'ALEOCENE ill T. TRIASSIC
LOCATION ~ T. CHALK It T. ZECHSTEII
MAP ®1. CHALK iIll. ZECHSTEII
$ •. VALANGINIAN

.r
'"

100 'm
50'

" " "


A A' •
3'3D'E 4'E 4'30'E
1.0

2.0

4.0

MITAMORPHIC
1.0 FLEKKEFJOtID HIGH IASEMEIT I.'

Tw.-wIY
I,ml ti..
Ii ..... 1 _ _-=====::J___
a""'IK_ L--'~==- __ II ••
~I

B B'
0.0 1'"'_ _ _ _ _ _ _ _......
3....
E _ _ _ _ _ _ _ _ _......
3'3..._·E_ _ _ _ _ _ _ _ _.....
" ..

2.0

CEITUL GRABEl
HIORA fAULT
BLOCKS VESTLAIO ARCH
.
After Skjerven, Rijs 8. Kalheim, 1983.

Fig. 2. Regional cross-sections across Jurassic structural features of the Southern Norwegian North Sea, (after Skjerven et
al., 1983). Striped interval depicts salt.

mature oil-prone source rocks of the Kimmeridge Clay BjJrglum Member of the Bream Formation become
Formation, and several reservoir horizons occur, vary- mature enough to have generated oil. To date, disco-
ing from Triassic through Paleocene. In contrast, the vered accumulations have been small (Bream, Brisling,
Kimmeridge Clay is thin or absent on the Vestland 18/10), and most structures do not appear to be filled
Arch; Jurassic reservoirs are generally absent. As the to structural spill-point.
Vestland Arch grades north-eastward to the Fis- The gross isopach of the radioactive portion of the
kebank Platform, both Kimmeridge Clay and Middle Kimmeridge Clay Formation and Borglum Member of
Jurassic Haldager Sands thicken, but the Kimmeridge Bream Formation shows the two basins separated by
Clay is still generallv immature. Further north-east, the intervening high (see Fig. 3). This is closely paral-
in the Egersund BasIn, the Kimmeridge Clay and the leled by the vitrinite reflectance trend indicating (see
250 Petroleum Geochemistry in Exploration of the N01wegian Shelf

RADIOACTIVE SHALE
ISOPACH MAP
Bj1Irglum Member
of Bream Formation
and Kimmeridge Clay
~~\~------~------~-------+~~~-157'N
Formation

C.I.:50ft.

km
i ,
o 50

2'E 3'E 4'E 5'E

Fig. 3. Gross isopach map of radioactive portion of Kimmeridge Clay Formation. Note differences in thickness between
Egersund Basin and Central Graben.

~ VITRINITE

\9.\
\ ~ 'P REFLECTANCE
at Top of B9rglum Member
~ of Bream Formation

::::::'"f.'f=:'~~-++--__\-__-1---~____+::;:"J't4!l~"~Y.i----j/
r--'-t;~::' [it"'" 57.Nand Kimmeridge Clay
./ Formation

C.1. .1%

km

:---~,=-----:-!-:----_L ____L ___-,J.56'N


2'E 3'E 4'E 5'E 6'E 7'E

Fig. 4. Vitrinite reflectance at top Borglum Member (of Bream Formation) and Kimmeridge Clay Formation. Note that the
Egersund Basin is less mature than the Central Graben.

Fig. 4) that, although both Central Graben and the tom using a modified gravity coring device that recov-
Egersund Basin are at an oil generating stage of ers up to 2 m of section. Core lengths varied from
maturity, Jurassic source rocks of the Central Graben 0.5 m in hard sand bottoms to a full core barrel in soft
are more mature than those in the Egersund Basin. clay. Two samples were collected from each core; one
was packed in an inert plastic bag and shipped to Hor-
vitz Research Laboratories in Houston, Texas, for
METHODS analysis of adsorbed gases and fluorescence inten-
sities. A second portion of each core, also packed in
Samples were collected from the lower portion of shal- inert plastic bags, was frozen in liquid nitrogen
low, 3 inch diameter cores obtained from the sea bot- (at - 196 °C) for later analysis by another laboratory
Regional Geochemical Analysis in the North Sea 251

for whole gas molecular composition and methane 30 30


isotopic ratios. Of the 204 regional cores, 172 provided
sufficient sample material for liquid nitrogen preser-
..
t;20
C1
20
zw
vation. The details of this work have been reported :>
elsewhere (Faber and Stahl, 1984). "~10
"-
In the laboratory, sediment samples for adsorbed
gas analysis, stored at room temperature, were agi-
tated in a blender to release any remaining interstitial 500 1000
ppo
1500 100 200
ppb
300

gas and screened with a 62 /-lm sieve to remove coarse


material. Resultant fines were dried by evaporation, 30
were weighed and then digested in hot phosphoric acid
under partial vacuum to desorb light hydrocarbons
..
>-
~20
from the sediment (Horvitz, 1969, 1972). Evolved gas w
:>
was purified and analysed by hydrogen flame "If
w
10
chromatography to ascertain its relative composition
(Horvitz, 1969, 1972). Results were expressed in parts
per billion dry weight of sediment for methane (C l ), 25 50 75 100
ethane and heavier hydrocarbons (C 2 + Cs + iso ppb

- C4 + n - C4 ), referred to hereinafter as C2 +, and


pentanes (n+iso+neo+cyclo). To estimate the occurr- 30 -320 30

ence of heavier hydrocarbons, a portion of fine mater- ..


>-
. 365
ial was extracted using n-hexane, and was subjected ~20 t;20
z
to fluorimetric analysis using a 265 nm excitation "'
:> w
:>

wavelength and recording emission intensities at ""'~ 10 "w 10


~

320 nm and 365 nm.


Dudenbostel and Priestly (1954), Lloyd (1971a,b,c), 25 50
~25 50
Wakeham (1977) and Law (1981) indicate that u.v. PEAK HEIGHT PEAK HEIGHT

emissions at 320 nm and 365 nm, respectively, rep-


resent two-ringed and three-ringed aromatic com- Fig. 5. Histograms showing distribution of yield values for
pounds including the triaromatic steroid hydrocar- adsorbed gases and intensities oftwo measured fluorescence
bons. However, these authors point out that recogni- wavelengths. Note that gas yield distributions are bimodally
tion of individual compounds by fluorimetric methods distributed.
can only be accomplished in simple mixtures. Because
mixtures of aromatic compounds derived from pet-
roleum are complex, fluorescence measurements indicate a single gas yield popUlation. Histograms of
obtained from surface sediments are reported as rela- fluorescence peak heights are unimodal and approxi-
tive intensities. For this study, the 320 nm peak is mate a log-normal distribution (see Fig. 5).
taken as an index of lighter, two-ringed aromatic com- Based upon analysis of the histograms, adsorbed
pounds and their derivatives occurring in gasoline gas yields can be separated into a popUlation charac-
fractions, and the 365 nm peak is taken as an index of terized by low gas yields from samples that contain
three-ringed aromatic compounds and their deriva- high sand percentages and those containing low sand
tives found in higher boiling fractions of natural oils. percentages and showing high gas yields (see Fig. 6).
Samples preserved in liquid nitrogen were analysed To determine the 'best' sand percentage at which to
for whole gas composition using gas chromatographic divide the samples into two distinct groups, a mul-
methods (Faber and Stahl, 1984). Methane thus reco- tivariate analysis of variance (MANOVA), using the
vered was also subjected to isotopic analysis using a three gas yields as discriminating variables, was per-
mass-spectrograph. Gas yield data were expressed as formed. Wilks' Lambda is a statistic computed as a
a percentage of the gas fraction. Isotopic analysis ratio of determinants of the total variance and the
yielded data expressed as the ratio ,j3C/ ,j2C in parts within-groups variance matrices, and is obtained as a
per thousand PDB*. However, fractionation of gases result of the MAN OVA (Cooley and Lohnes, 1971). As
recovered in some high sand samples resulted in a loss this ratio becomes smaller, the hypothesis that the
of methane, and therefore measurements for some adsorbed gas yields are sufficient discriminators of the
samples were unreliable. two groups is rejected with increased confidence
(Cooley and Lohnes, 1971). As the value of this statis-
tic increases, the discrimination hypothesis is rejected
RESULTS with less and less confidence, indicating that the vari-
ables are not sufficient discriminators. Therefore, if
Data obtained from the various analytical techniques the MANOVA is performed using groups separated at
are presented as histograms and as a selection of several different sand percentages, the 'best' division
bivariate cross-plots. Gas yield histograms are is obtained where Wilks' Lambda has the lowest
bimodal, representing a mixture of two distinct popu- value. The variation of Wilks' Lamdba with sand per-
lations (see Fig. 5). These contrast strongly with centage is shown in Fig. 7, and reveals that the 'best'
those presented by Horvitz (1980) for the Gulf of Mex- division ranges between 80% and 85% sand.
ico, and those examined by Tenneco in other areas of Goodness-of-fit tests applied to each sediment-
the world where bottom lithology is uniform, which dependent population are significant at the 95% confi-
dence level, strongly indicating that they are normally
distributed. Because the popUlations overlap, it is dif-
* al3 C --
eSc/ 12C) sample - (13C/ 12C) standard
1000
ficult to assign to their proper popUlation those sam-
t13C/12C) standard x %,. ples containing adsorbed gas yields of 0-500, 125-150
252 Petroleum Geochemistry in Exploration of the Norwegian Shelf

> 80% SAND <80% SAND each of the distributions. To place all yield values on a
single basis, one population (usually the low sand) is
.. 30
>-
----- METHANE .. 30
>-
- METHANE- selected as a standard, and values comprising the
~
(J
z other are transformed so that these yields fall into the
~20+-~------------~ 20+----------;c;----
o
~
distribution of the standard group. The following equa-
w iil0:
a:
... 10 ... 10+------1
tion has been derived and used to correct high sand
values to approximate those values in the low sand
population:
250 SOO 750 1000 1250 0 250 500 750 1000 1250
ppb ppb
Y .8sYm
Y corr = Y.8m '
.. 30 C2+- .. 30 -----C2+
~
>-
(J
zw 20
where Y m = mean yield in mud, 8", = standard
ffi 20 -'- deviation in mud, Y s = mean yield in sand, 8 8 =
"a:
ow
r "o
W
0: standard deviation in sand and Y s = yield for given
... 10 -I ... 10
sand sample.
,!l I~
100 200 300 400 500 a 100 200 300 400 500 600 Application of this method to an area of mixed sedi-
ppb ppb
ment type is fairly robust. However, there are several
uncertainties involved. First, because distributions
.. 30 -- PENTANE- .. 30 -PENTANE overlap, yields falling at the high tail of the high sand
>- >-
(J
z
(J
z distribution and at the low tail of the low sand dis-
~ 20 ~ 20
tribution are, at best, uncertain. Second, if the two
0
w iil0:
a:
... 10 l ---- ... 10
component distributions are of widely differing statis-
tical character, the conversion will be inaccurate for
~ the entire range. Because light alkanes are most prob-
25 50
ppb
75 100 125 25 50
ppb
75 100 125
ably adsorbed differentially by different clay mineral
species, and because the adsorbtivity of clays-espe-
Fig. 6. Results of dividing sample set into two cially those of the smectite group-depends on the
sediment-dependent groups at 80% sand. Note that these condition of the lattice at the time of adsorption
histograms approximate a normal distribution. (Eltantawny and Arnold, 1972), a more realistic
method of obtaining a basis on which to map adsorbed
gas yields should be based on quantitative aspects of
.80,j----,-------,---- I
I
-- - T sample mineralogy, rather than on the crude grain
size analyses as done here. However, until the adsorp-
tion mechanism becomes better understood, the nor-
~--
70
malization method derived and used in this paper must
8 suffIce, although some caution should be exercised
60 'u;-
Z
"' when interpreting resultant maps.
To compare adsorbed gas data obtained from South

Lrr
Norway with those measured elsewhere, a cross-plot

,I I ! I was prepared of methane versus C2+ uncorrected yields


for several areas of contrasting source-rock types,

"tJ=_!±j~j_LL
maturity levels and production potential. These
include an area off southern California, six Halten-
banken (Tr~ndelag II) blocks in the Norwegian Sea,
ECMA Trinidad, and several areas in the Gulf of Mex-
% ~ ~

% SAND
00 ~ 70 ~
ico (see Fig. 8). Southern California is underlain by

Fig. 7. Variation of Wilks' Lambda ().) with division of


sample set at various sand percentages. The lowest values '00

occur at 80-85% sand, indicating that the 'best' division


occurs there. • SOUTH NORWAY
u SANTA. YNEZ
(> HAlTENBANI(

.. ,
c

.,....-
0
<\' ECMA TRINIOAO
<l SOUTHTIMBAlIER
o GREEN CANYON
and 25-30 ppb for methane, C2 +, and pentane, ~

respectively, and sand percentages of 75-85%. .a 240


~
In areas of contrasting bottom lithology, raw +
0 200
adsorbed gas yield maps partially resemble the sand
percentage map. These maps partially reflect the rela-
tionship between surface sediments and gas yields,
thus obscuring any relationship to subsurface
phenomena. In order to examine any relationship
between adsorbed gas yields and subsurface
phenomena, values must fIrst be corrected for these
sediment effects. Several different statistical and O~O~~C-~"~O~~3~20~.~00~48~O~~~Z640~7~20~8~00~8~80~%~O~1O'~O-m"20
METHANE ppb
graphical techniques have been tried. The technique
described below is empirical, and is the most robust. Fig. 8. Cross-plot of uncorrected methane yield versus
It is essentially a transformation relying solely on the uncorrected C2+ yield, comparing South Norway sample
parameters (mean and standard deviation) describing area to other hydrocarbon provinces of the world.
Regional Geochemical Analysis in the North Sea 253

TABLE 1. Area comparison, CJC 2 + ratios. 1~~-------------------------------,

Area Sauree rock facies

South Norway 2.60 Sapropelic-humic mixture


Santa Ynez 2.07 Sapropelic
Haltenbanken 3.28 Humic-sapropelic
ECMA Trinidad a 6.66 ? 750
South TimbaIier 3.63 Humic?
Green Canyon 2.77 Sapropelic-humic

aRatio approximate because gas yields are extremely low.

Oligocene and Miocene source rocks composed


entirely of type II (marine sapropelic) kerogen. 500 o
• SOUTH NORWAY
Maturity varies slightly with structural depth and pos-
l::\D 0
o SANTA YNEZ
<> HALTENBANK
ition. The six Haltenbanken blocks sampled are under- ~ ECMA TRINJOAD
SOUTH TJMBALIER
lain by mature humic-sapropelic source rocks. No
LI
o GREEN CANYON

mature source rocks were penetrated to drilled depth


in ECMA Trinidad, but mature and overcooked
sources may .exist at deeper horizons. Gas produced
here is of biogenic origin (~lac of CH 4 = -60%0). The 250

source sections underlying the sampled areas of the


Gulf of Mexico have not yet been recovered by dril-
ling, but regional geological considerations suggest o
o
0

that kerogens should be of mixed humic-sapropelic


composition.
The methane versus C2+ cross-plot comparing
100 200
these diverse terrains shows several distinct trends 320 nm Intensity
expressing differences in both kerogen type and
maturity level. Each trend can be approximated by a Fig. 10. Cross-plot of 320 nm intensity versus 365 nm
straight line through the origin, whose slope reflects intensity, comparing the South Norway study area to other
hydrocarbon provinces of the world.
the gross ratio between the gases. Average ratios for
each area were given in Table 1 and range between
2.1:1 and 6.7:1. cates that this interpretation is consistent and that the
Available evidence concerning kerogen facies and statistically derived value of 2.6 seen in the plot rep-
thermal maturation for these areas strongly suggests resents the dominant kerogen type.
that average CdC 2+ ratios reflect the dominant The range of gas yields exhibited in the methane
kerogen facies. The plot demonstrates that for more versus C2 + plot results in part from differences in
sapropelic source rocks, the CJC 2+ ratio decreases. surface sediment type. To demonstrate this, the
For the South Norway area, the methane versus methane versus C2+ plot was constructed using dif-
C2 + cross-plot (Fig. 9) shows a trend indicating a ferent symbols for each of the two gas populations (see
CdC 2+ ratio of 2.6. This value suggests that most of Fig. 6). The high sand samples are restricted to the
the adsorbed gases recovered here originate from low portions of the trend while the low sand samples
admixed sapropelic and humic kerogen with the sap- are located in the high portions of the trend. However,
ropelic material predominating. Comparison of the differences in gas yields are not directly related to
surface data from which this diagram was constructed sand percentage, but rather to mineralogical composi-
with source rock data obtained from well samples indi- tion of the fine fraction itself, as suggested earlier.
In contrast to gas yields, fluorescence intensities
400
show no relationship to bottom sediments. Fluores-
cence intensity histograms are strongly unimodal and
• >80%SAND are lognormally distributed. To compare the South
" <BOSSANO
Norway region with others worldwide, a plot was
prepared showing the variation of intensities for the
, ,~

two wavelengths for the same areas as those used for


,,~o
~240
"
gas yields (Fig. 10). This plot shows that a relation-
&
+ ship exists between the intensities, and also indicates
0'200
that intensities differ by an order of magnitude with
source facies. Southern California samples show
.... .' .
160
'.
, , )
fluorescence intensities up to 10 times greater than
"
those found in other areas. Evidently, on a regional
.. ..r:,:.'
• '.
.
'!~••

80 basis, fluorescence intensity is also related to source


a~·· composition and thermal maturity. The plot suggests
f'l'" .. :.
that sapropelic source rocks are represented by sub-
stantially higher intensities of both 320 nm and
365 nm wavelengths.
Fig. 9. Cross-plot of uncorrected methane versus To compare fluorescence intensities measured in
uncorrected C2+ for the South Norway study area. The South Norway, a cross-plot was prepared for the two
linear trend shown in this plot reflects a C,/C 2+ ratio of2.6. wavelengths (Fig. 11). This plot, in contrast with the
Note differentiation of samples by sediment type. gas yield plot presented above (Fig. 9), shows consid-
254 Petroleum Geochemistry in Exploration of the Norwegian She{f

TO
• >80%SAND DRY GAS GAS ASSOCIATED GAS DRY GAS
o <80%SAND
BIOGENIC WITH OIL COHO THERMOGENIC
60 ,\
\
\

30
50
....
~
z
.0 ••• ~20
• :.~ 0
.. filII:
u.

10

~~
20

10
-60 -50
I -401-38;361 -30
/)13C/12C

ooL----,~0--~2~0----3~0----.~0----5~0----T60---,70 Fig. 13. Histogram showing the distribution of methane


320 nm Intensity
carbon isotopic ratios obtained from frozen portion of bottom
Fig. 11. Cross-plot of 320 nm intensity versus 365 nm samples from South Norway. Note that the bulk of the val-
intensity for tile South Norway samples. Note lack of ues fall within the liquid range.
differentiation by sediment type.
versus 320 nm and corrected methane versus 365 nm
are -13.26% and -39.50%, respectively. The me-
erable scatter. Analysis of duplicate samples in other thane versus 320 nm correlation is not significant
areas suggests that some of the scatter is due to while the methane versus 365 nm is significant at the
laboratory and sampling errors. Because, as will be 95% confidence level. Plots of C2 + and pentane
shown below, fluorescence measurements are valu- against the two wavelengths show similar relation-
able interpretive data, better laboratory technology is ships, and have been therefore omitted for brevity.
required. The remainder of the variation is related to Carbon isotopes measured for methane obtained
maturity and source composition. To demonstrate the from frozen samples collected on the regional grid are
independence of fluorescence intensities and bottom displayed in histogram form in Fig.13. These values
lithology, these data were plotted using different range between -42%0 and -20%0, thus presenting con-
symbols for each of the two sand populations (Fig. 11). vincing evidence that the gases collected from these
High and low sand samples are intermixed in this plot sediments are of catagenic origin. The bulk of the
and, in contrast to the gas plots, are not separable on isotopic ratios fall in the oil region as defined by Faber
the basis of fluorescence intensity. and Stahl (1984).
Cross-plots were prepared to show the relationship A cross-plot prepared for sand percentage versus
between gas yields corrected for sand content and carbon isotopic ratio (Fig.14) shows no obvious rela-
fluorescence intensities. Two of these plots are pre- tionship. However, carbon isotopic ratios that are
sented here (Figs. 12(a) and (b». Although they unreasonably heavy (-26%0 to -10%0) are those from
appear similar, correlation analysis indicates that high sand samples, suggesting that oxidation and frac-
while methane shows no relationship to 320 nm inten- tionation are more likely in sandy samples.
sity, a weak inverse relationship with 365 nm inten- To relate whole gas measurements obtained from
sity is suggested. Correlations for corrected methane frozen samples to adsorbed gases obtained from

r :-13.26 r--39.5O

o
00 o

....
00

:-, 00
00

..
...' . -:'.. .
:
... ... ,....-
o

....-.:... ~ .
• • etA : : .
.... .,.... ..,..,.... .~.:
o 0

.
.. :.-
\:o1r~.::r \~o
\
-:.: If· • •• a ••:.:~tl:::-::. 00

.. ..; ..- .
0

00
: .
: . ..-.':. ·":-·-i •.. : ..
.... :. -. . ,
400 500 600 ..,. ... 600
C1 ppb ICorrectedl C1 ppb ICorrectedl

Fig. 12.(a) Cross-plot of corrected adsorbed methane yields versus 320 nm intensities for South Norway sample area.
Correlation coefficient (-13.26) is not significant at the 95% confidence limit. (b) Cross-plot of corrected adsorbed methane
yields versus 365 nm intensities for South Norway sample area. Correlation coefficient (-39.5) is significant at the 95%
confidence interval, demonstrating an inverse relationship.
Regional Geochemical Analysis in the North Sea 255

-10 1360

-12 1280
00
-14 1200
-16 1120
-16 1040
o 80 0
-20 .60 o 0 08 00

-22 00 0

..:l:
880
-24 800
-26 o
8 00
w 720
u Z
"w
(') -28 00 640
J:
~@ -30 a 0 ctbce I-
560

g.8J
c 08 ~ 6'
:Ii
-32 480

-34 0 °er o cPo~~. 400


0
0 320

06'",0% 0 0 ~~ 240 o
o
o o~ 0 0 ° 0 g~oo 160
8 0°
80
o 0
-44
_~~~~~~_~~~~44~~ a 4
-46
DELTA 13C

o ..
.ro~w~~~~
SAND PERCENT
~~~~ron~~~~_
Fig. 15. Cross-plot of methane carbon isotopic ratio
obtained from frozen portion of South Norway samples
Fig. 14. Cross-plot of percentage sand versus methane versus corrected adsorbed methane yield. Abnormally
carbon isotopic ratios obtained from frozen portions of South heavy isotopic ratios are associated with low adsorbed gas
Norway samples. Samples showing abnormally heavy yields. The remaining samples form an almost vertical trend
isotopic ratios in high sand percentage samples are probably with larger methane yields having lighter isotopic
fractionated. composition.

unfrozen samples, a plot of bl3C versus adsorbed by the bl3C versus methane plot, and can be inter-
methane yield (Fig. 15) was prepared. This plot shows preted similarly. Although it is not possible at this
that isotopic ratios become heavier as methane yields time to establish definitely the nature of the relation-
decrease, suggesting that low methane yields repre- ship between adsorbed gas yields and isotopic ratios,
sent more-mature or coaly source rocks, or that many these ratios have proved valuable as independent
of these heavier values are fractionated. This diagram methods of assessment in the North Sea and other
demonstrates that the variance in isotopic ratios is areas.
greatest at lower methane yields, lending support to Cross-plots prepared for methane carbon isotopes
the notion that low yield samples are fractionated. and the two fluorescence intensities show no obvious
Cross-plots of bl3C versus C2+ and pentane (omitted relationship (see Figs. 16(a) and (b». However, corre-
for brevity) show similar relationships to those shown lation analysis reveals a weak negative correlation,

60
80

72
72

64 64

56 56

46 48

o
00

o 00
o 0000
o o:c
,
000 00
o 00 ,
o 0
~ 0
o 0 0 0 0
o 0'
o 0
o 0 =
o 00
00 00000
000 CII 00 00
o 0
00 00000
o 0 00 0 00 00 0 0
24 OCXiOOO 0 0 o

.0.
!XI 0 0 (DO 0
o 00 o 000
00 0 00 0

16
GOOD
o ao
GOO
Oal

o
o
0
0
0
0

16
o :~~l~oo
~Oo~O(;Qoo
0

00 00 atlOOOOO
00. OCXll><ll:XXmO 0 0
o 0 o O~OCII~ 00 00
00 00
Oal ClIIIDlO 0 (ll
~-=
(l])OlIJI:OCXXXlOOQ
O_~
=~.
0
ctl°oQ~a:l: ilia CD o
00=00000
o ~ 0
o
__~__~__~__~____
0~~4---'~56----'4~8---'~40-----3r2---'~24----"-6---_~8--~ O~--~--~--~~

-1S4 -56 -48 -40 -32 -24 -11 -8


b13c (METHANE) b '3 c (METHANE(

Fig. 16.(a) Cross-plot of methane carbon isotopic ratios obtained from frozen portions of South Norway samples versus 320 nm
intensities. (b) Cross-plot of methane carbon isotopic ratios obtained from frozen portions of South Norway samples versus
365 nm intensities.
256 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Sand %

o >eo"
1O·eo"
o <10"

2'E 3'E .t'E S'E 6'E

Fig. 17. Map of South Norway sample area showing distribution of sand.

indicating that light isotopic ratios occur with higher which major source section has been removed by ero-
fluorescence intensities. sion, show significantly lower adsorbed gas yields.
Corrected adsorbed gas yields, fluorescence inten- These areas include the Vestland Arch/Fiskebank
sities, and methane isotopic ratios obtained from re- Platform.
gional samples were mapped over the southern Nor- In contrast, fluorescence intensity maps (Figs. 20
wegian portion of the North Sea. These maps were and 21) show high intensities over the platforms and
compared with a regional tectonic and structural map relatively low intensities over areas highlighted as
(Fig. 1) and with several published cross-sections generating basins by adsorbed gases. The Beatrice Oil
(Fig. 2) to relate surface geochemistry to geology. Field study (Gevirtz et al., 1983) has demonstrated
Source rock evaluations, where available, were also that fluorescence intensities indicate the types and
incorporated into the interpretation. A regional sand amounts of hydrocarbon liquids carried by upward-
percentage map has been included for comparison migrating water reaching the surface, and provide
(Fig. 17). evidence of the subsurface hydrodynamic regime. In
Maps for each variable highlight the major tectonic South Norway, areas of high fluorescence intensities
elements over which sufficient samples were col- occur over the Vestland Arch/Fiskebank Platform
lected. The Vestland Arch, the Central Graben, and (see Fig. 1). The Arch, almost devoid of Jurassic
the Egersund Basin are clearly shown by either high rocks, is covered by a Tertiary section containing con-
or low values of several variables. The remarkable siderable blanket sand deposits. It is believed that
coincidence between patterns formed by geochemical oil-rich water originating in the source basins to the
variables and the configuration of various tectonic east and west rises at the edge of the Vestland Arch
elements is believed to represent an expression of the and is carried by Lower Tertiary sands across it.
structural discontinuities defining these tectonic ele- Because these Tertiary sediments are only partially
ments, such as faults and/or fracture zones of large consolidated and a fairly uniform leakage rate can be
lateral extent. These discontinuities provide avenues postulated, the Vestland Arch shows uniformly high
for leakage to the surface of both gases and water of fluorescence intensities.
deep origin. Structural discontinuities defining highs Comparison of the 320 nm intensity map (Fig. 20)
tend to focus upward water movement (Roberts, with the 365 nm intensity map (Fig. 21) confirms sub-
1980), thus concentrating water-borne hydrocarbons tle differences between distributions of two- and
over them, while gases tend to migrate in a more ver- three-ringed compounds, as indicated by cross-plots
tical manner; therefore, higher yields occur closer to and correlation analysis. While both fluorescence
the source. wavelengths show high intensities over structural
Corrected adsorbed gas maps (Figs. 18 and 19) for highs, higher 365 nm intensities occur in a narrow
South Norway show higher gas yields over major hyd- band over the eastern edge of the Fiskebank Plat-
rocarbon generating provinces such as the Central form. Experimentally derived maturation curves for
Graben and the Egersund Basin, thus confirming the various kerogen types (Tissot and Welte, 1978; Hunt,
idea that these light hydrocarbons migrate in a more 1979) suggest that for most oil-generating kerogens,
vertical direction. Areas that have been shown to be three-ringed aromatic compounds predominate in
immature on the basis of well cuttings analysis, or in less-mature source rocks while two-ringed aromatic
Regional Geochemical Analysis in the North Sea 257

Corrected
Adsorbed C l Yield

C.l . : 250ppb

57'N

0 >150 ....
D soo . 750 poll
0 <500ppb

,
0
km
,
50

56'N
2'e 3'E .'E S'E o'E 7'E

Fig. 18. Map showing methane yield distribution over sample area. High sand samples have all been corrected by method
discussed in the text.

compounds are dominant at higher thermal maturity the source rocks in which these fluorescing compounds
levels. originate. Therefore, because the Central Graben area
Very mature kerogens, those in the gas-generating is richer in 320 nm than the Egersund Basin is, the
range, produce little or no two- or three-ringed Central Graben is probably generating wet gas. This
aromatic compounds. Therefore, assuming that gas reaches the surface more or less vertically, and
kerogen types are grossly similar where they occur carries with it two-ringed aromatic compounds that
throughout the study area, these fluorescence inten- show up in surface samples as higher 320 nm inten-
sity distributions should reflect relative maturity of sities. The Egersund Basin, on the other hand, does

Corrected
Adsorbed C2+ Yield

C. I. ' 50ppb

o >250ppb

o 150· 250ppb

o <1S0ppb

:
o
km
,
50

'-----~---L---.,.L,---_,.L---_:l SO'N
2'e 3'E "E S'E o'E 7'E

Fig. 19. Map showing C2+ yield distribution over sample area. High sand samples have all been corrected by method
discussed in text.
258 Petroleum Geochemistry in Exploration of the Norwegian Shelf

320 nm Intensity

C. I. : 10 unll.

D >30 unit.

D 20·30 unlla

o <20 unit •

....
50

:------::---~_=__--~::__--~::__--_:!S6'N
2'£ J'E 4'E 5'E 6'E 7'£

Fig. 20. Map showing distribution of intensities of 320 nm wavelength fluorescence. Note the higher intensities over the
Vestland Arch and a portion of the Central Graben, see Figs. 1 and 2.

365 nm Intensity

C.I.: 10 unll.

D >40uollo
[23 30 · 40 unlta

o < 30unllo

:
co .'" 50
,

2'E 3'E A'E 5'£ 6'E

Fig. 21. Map showing distribution of intensities of 365 nm wavelength fluorescence. Note the higher intensities over the
Vestland Arch, see Figs. 1 and 2.

not show high 320 nm fluorescence intensities. This The Vestland ArchJFiskebank Platform shows high
evidence suggests that source rocks in the basin are intensities of both wavelengths. These materials, the
less thermally mature-perhaps at the beginning of two-ringed compounds originating mainly in the Cen-
the oil-generating stage-and are generating three- tral Graben and the three-ringed compounds from the
ringed aromatic compounds that are being carried by Egersund Basin, are carried by water moving through
water to the western basin edge. Here they come shallow (Tertiary?) aquifers away from their respec-
to the surface and are measured as high 365 nm tive generating sites, and reach the surface over this
intensities. shallow basin-platform complex. It is important to
Regional Geochemical Analysis in the North Sea 259

C.I.: 5"-

0>40%.
035.40"-

0<35%0

Om
i i
o 50

- - - - ' - - - - - - ' - -- - _ __ _ _L-_ _ ~56·N


2'E 3'~ 4'E S'E 6' E 7'E

Fig. 22. Map showing the C2 +IC! x 100 ratio. All adsorbed gases in high sand samples were corrected using method described
in text. Note location of wetter (shaded) areas.

Whole Gas Methane


Carbon Isotope Ratio

D >-36
0 < -38

, km
i
0 tiD

ZOE 3'E j'E S'E /I'E

Fig. 23. Map showing distribution of methane carbon isotopic ratios obtained from frozen portions of South Norway samples.
Note general conformation of heavy ratios with Fiskebank Platform/Vestland Arch.

recognize that high fluorescence intensities-particu- text, a map was prepared for South Norway and com-
larly for the higher wavelengths-do not indicate a pared with regional tectonic and isopach maps. The
generating basin, but instead mark the basin edge. map shows that the South Norway basins sampled are
The Beatrice Oil Field study (Gevirtz et aI., 1983) generally in the oil to gas-condensate generating
suggests that the expanded ratio C2 +/C 1 X 100 range. They are more mature over the deep graben,
(Fig. 22) expresses subtle variations from the rela- with maturity decreasing as structural highs are
tionship indicated in the cross-plot (Fig. 9). To approached. The Vestland Arch/Fiskebank Platform,
examine the variation of this ratio in a regional con- where Jurassic rocks are thin or absent, shows grea-
260 Petroleum Geochemistry in Exploration of the Norwegian She(f

tel' maturity than expected, due to additions of compounds are carried with water and reach the sur-
methane possibly originating in deeply buried Per- face through discontinuities such as faults and frac-
mian or Carboniferous sources responsible for large tures of broad geographic extent.
gas deposits to the south, outside the sample area. 5. Differences in distributions of two- and three-
However, low yields over this platform complex argue ringed aromatic compounds are apparently related to
that it is not a prospective oil area. differences in their solubilities in both water and gas.
North Sea basins contain several source rocks. 6. Isotopic ratios measured in South Norway
However, the Upper Jurassic/Lower Cretaceous strongly suggest that the methane sampled is of
Kimmeridge Clay Formation and similar units such as catagenic origin. No ratios indicative of biological
the Borglum Member of the Bream Formation activity were observed in the study area.
are considered to be the predominant source of the 7. A complex relationship exists between adsorbed
hydrocarbons in the Central Graben and Egersund gas yields and sediment mineralogy. This relationship
Basin. Where they have reached thermal maturity, must be fully understood in order to relate adsorbed
the Middle Jurassic coal sequence of the Haldager gas yields more precisely to subsurface geology and
Formation and thin organic-rich units, such as the hydrodynamic regime.
Kupferschiefer in the Zechstein, add to the overall 8. Ratios of adsorbed gases gathered over a region
hydrocarbon generation. Because all sources contribute reflect source rock facies and maturity levels in the
to gases and fluorescence intensities measured at the underlying basins.
surface, the picture presented in this ratio map is 9. A calibration survey ofthis type can be invaluable
unclear. in showing which geochemical parameters should be
Studies in other areas of the world indicate that mapped in wildcat basins without well control in order
isotopic ratios for methane ranging between -40%0 to rate the prospectivity of the basin.
and -50%0 are indicative of a thermal regime in the oil
generation range. Ratios lighter than -50%0 are indi-
cative of methane resulting from biogenic activity tak- ACKNOWLEDGEMENTS
ing place at temperatures below the pasteurization
point (87 DC) (Faber and Stahl, 1984). Isotopic ratios The authors would like to express thanks to the many
heavier than -35%0 are indicative of extreme thermal people at Tenneco Oil Company, Exploration and Pro-
maturation in coaly material, and are usually found duction, who helped us prepare data and maps, and for
where catagenic gas is produced. the numerous discussions on the geology and geophys-
Because isotopic fractionation is more likely in small ics of this part of the North Sea. We would like to
and/or sandy samples than in large clay ones, values in thank Cindy Olinger for preparation of the final graph-
sandy samples heavier than -36%0 were grouped to ics, and Rhona Jenkins who typed the manuscript.
avoid confusion. The isotopic ratio map (Fig. 23) pre- The senior author would like to further acknowledge
pared from regional data also highlights the major P. R. Glasgow who helped with the final editing.
structural features in this portion of the North Sea Finally, we would like to express thanks to the man-
(see Figs. 1 and 2). The Vestland Arch/Fiskebank agement of Tenneco Oil Company, Exploration and
Platform is characterized by values heavier than Production, for permission to publish this work.
-36%0. Most of the scatter occurs in this area. These
values are indicative of extreme thermal maturity.
However, spore colouration and vitrinite reflectance REFERENCES
data obtained from well samples indicate that this area
is immature at the Jurassic stratigraphic level, sug- Barrer, R. M. and MacLeod, D. M. 1954. Intercalation and sorption
gesting a much deeper source for the methane from by montmorillonite. Trans. Faraday Soc. 50, 980-989.
Barrer, R. M. and MacLeod, D. M. 1955. Activation of montmoril-
which isotopic ratios were gathered. These immature lonite by ion exchange and sorption complexes of tetra-alkyl
Jurassic source rocks probably do not generate suffi- ammonium montmorillonites. Trans. Faraday Soc. 51,
cient methane to substantially alter the isotopic signa- 1290-1300.
ture arising from much deeper source materials, Barrer, R. M. and Perry, G. S. 1961. Sorption of mixtures and
selectivity in alkylammonium montmorillonites. Part I, Mono-
perhaps at Permian levels. Methane yields across the methyl ammonium bentonite. Chell!. Soc. Lond. J., 842-849.
Vestland Arch are low, also suggesting that this mat- Cooley, W. W. and Lohnes, P. R. 1971. Multivariate data analysis.
erial is coming from sources producing low gas yields. Wiley, New York, 364 pp.
Dudenbostel, B. F., Jr and Priestley, W., Jr. 1954. Ultraviolet
spectra of hydrocarbons, in Brooks, B. T., Boord, C. E., Kurtz,
S. S. and Schmerling, L. (eds.), The Chemistry of PetroleulII
SUMMARY AND CONCLUSIONS Hydrocarbons. Reinhold, New York, pp. 335-350.
Eltantawny, l. M. and AmaIn, P. W. 1972. Adsorption of n-alkanes
1. Surface geochemistry is an extremely valuable by Wyoming montmorillonite. Nature, 237, 123-125.
tool for regional evaluation when integrated with Faber, E. and Stahl, W. 1984. Geochemical surface exploration for
more conventional geological, geophysical and hydrocarbons in North Sea. Bull. Am. Ass. Petroleum Geol-
ogists, 68(3), 363-386.
borehole geochemical data. Gevirtz, J. 1., Carey, B. D., Jr and Blanco, S. R. 1983. Surface
2. Several geochemical methods are far superior to geochemical exploration in the North Sea. In: Brooks, J. (ed.),
one, and contribute to the understanding of hydrocar- PetroleulII Geochemistry and Exploration of Europe. Geol. Soc.
bon generation and migration throughout a series of Lond. Spec. Publ. 12, pp.35-49. ll(Blackwell Scientific, Oxford.)
Horvitz, L. 1939. On geochemical prospecting for petroleum.
basins in a given province. Geophysics, 5(10), 487-493.
3. A preliminary broad-brush approach such as the Horvitz, 1. 1969. Hydrocarbon geochemical prospecting after thirty
one described here can aid the explorationist in defin- years. In: Unconventiunal JIethods in Exploration for Pet-
ing areas that merit further interest and those that roleum and Natural Gas, Southern Methodist University,
show little hydrocarbon potential. Dallas, Texas, pp. 205-218.
Horvitz, 1. 1972. Vegetation and geochemical prospecting for pet-
4. This study shows that gases migrate to the sur- roleum. Bull. Am. Ass. PetroleulII Geolugists, 56(5), 925-940.
face more or less vertically, while heavier aromatic Horvitz, L. 1980. Near surface evidence of hydrocarbon movement
Regional Geochemical Analysis in the North Sea 261

from depth. In: Roberts, W. H. III and Cordell, R. J. (eds.), In: Roberts, W. H., III and Cordell, R. J. (eds.), Problems of
Problems of Petroleum Migration. AAPG Studies in Geology, Petroleum Migration. AAPG Studies in Geology, No. 10,
No. 10, pp. 241-269. pp. 217-240.
Hunt, J. M. 1979. Petroleum Geochemistry and Geology. W. H. Rosaire, E. E. 1940. Geochemical prospecting for petroleum. Bull.
Freeman, San Francisco, 617 pp. Am. Ass. Petroleum Geologists, 5, 1418~1426.
Law, R. J. 1981. Hydrocarbon concentrations in water and sedi- Skjerven, J., Rijs, F. and Kalheim, J. E. 1983. Late Paleozoic to
ments from U. K. Marine waters, determined by fluorescence early Cenozoic structural development of the south-southeastern
spectroscopy. Marine Pollut. Bull. 12(5), 153-157. Norwegian North Sea. Geol. Mijnbouw, 62, 35-45.
Lloyd, J. B. F. 1971a. The nature and evidential value of the Scull, B. J. and Myhre, L. A. 1976. A combination of geochemical
luminescence of automobile engine oil and related materials. I. and geophysical prospecting for petroleum, offshore Norway. In:
Synchronous excitation of fluorescence emission. J. Forensic Sci. Empresu Nacional Adaro de Investigaciones Mineras,
Soc. VII, 83-94. pp. 561-569.
Lloyd, J. B. F. 1971b. The nature and evidential value of the Tissot, B. P. and Welte, D. H. 1978. Petroleum Formation and
luminescence of automobile engine oil and related materials. II. Occurrence: A New Approach to Oil and Gas Exploration.
Aggregate luminescence. J. Forensic Sci. Soc. 11, 153-170. Springer-Verlag, New York, 538 pp.
Lloyd, J. B. F. 1971c. The nature and evidential value of the Wakeham, S. G. 1977. Synchronous fluorescence spectroscopy
luminescence of automobile engine oil and related materials. III. and its application to indigenous and petroleum-derived hyd-
Separated luminescence. J. Forensic Sci. Soc. 11, 235-253. rocarbons in lacstrine sediments. Envir. Sci. Technol. 11(3),
Roberts, W. H., III 1980. Design and function of oil and gas traps. 272-276.
22 Recently formed methane-
derived carbonates from
the North Sea floor
Martin Hovland, * Mike Talbot, t Snorre Olaussen * and Lars Aasberg*
*Statoil, P.O. Box 300, N-4001 Stavanger, Norway
tGeologisk Institutt avd. A, University of Bergen, N-5000 Bergen,
Norway

Numerous slabs and indurated crusts of limestones and carbonate cemented sandstone were found on the floor of
pockmarks in the North Sea. Sea-floor pockmarks are local depressions found in unconsolidated fine-grained
sediments. Their origin has been ascribed to percolation of gas or liquids through the sea floor. A group of five
pockmarks believed to be associated with shallow gas, located in Norwegian Block 25/7, were studied by the use of
a remotely operated vehicle (ROV). Petrographic studies of the cements in one of these slabs (the '25j7-slab') show a
variety of aragonite and Mg-calcite crystals. Aragonite crystals are often perfectly preserved and are comparable
with aragonite cements found in modern reef structures.
The alaC-value of the aragonite cement in the 25/7-slab is -58.9%, (PDB) which indicates a methanogenic origin. It
is concluded that the 25/7 -slab is a response to an active diagenetic zone on the floor of the pockmark, whereby a
glauconitic and fossiliferous sand is transformed to a carbonate cemented sandstone and sandy limestone. Scanning
electron microscopy has revealed globular and crystal structures in the slab's contact zone with the sea floor,
which suggest that the diagenetic process was active when the slab was retrieved from the sea floor.
Sea-bed sediment samples were also obtained from the vicinity of the pockmarks. The yields of methane through
hexane contained in the upper soft sediments together with the JloCI-value (-39%, to -45%, PDB) suggest a
thermogenic origin for the shallow gas.
It is concluded that both the sea-floor pockmarks and the carbonate crusts are probably caused by the seepage of
hydrocarbon gases through the sea-floor.

INTRODUCTION POCKMARK CARBONATES


During the visual survey of the area, the floors of the
pockmarks were found to be littered with blocks and
protruding slabs of carbonate cemented sandstone and
During reconnaissance pipeline route surveys for the sandy limestone (see Fig. 3) (Hovland, 1984a). In
Statpipe project in 1981 an area with five isolated some areas the pockmark floors seemed to be encrusted
pockmarks was found south of the Heimdal Field in with an irregular layer of the material. The crust is
Block 25/7 (see Fig. 1). The area is of interest due to dusted with finely grained sediments and has numer-
the associated occurrence of shallow gas in soft, ous holes and cavities. A slab about 30 x 20 x 10 cm
layered sediments (Hovland, 1984b). The five pock- and weighing about 10 kg in air (the '25/7-slab') was
marks vary from 20 m to 40 m in diameter and are recovered from the centre of one of the pockmarks
2-4 m deep. They are situated above a possible gas (see Fig. 4). It was located over an area where the
'cloud' seen on shallow seismic records between 38 m subsurface gas anomaly apparently reaches the sea-
and 0 m beneath the sea floor (see Fig. 2). At the floor surface. The slab protruded out over a cavity, or
pockmark location in Block 25/7 the water depth is a minor crater, which seems to have been formed by
127 m, and the sea floor, consisting of fine sand, is some erosional process.
generally flat. Another characteristic of these pock-
marks is their highly reflective floors which appear as
dark patches on side-scan sonar records. MINERALOGICAL AND GEOCHEMICAL
In order to study the pockmarks in more detail, a RESULTS
visual geological and geochemical survey employing
an ROV (Remotely Operated Vehicle) was carried out Laboratory analysis of the 25/7-slab has provided a
in 1983. number of interesting results. A section of the slab is
Petroleun! Geochemistry in Exploration Qf'the Norwegian Shelf' © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 263-266.
264 Petroleum Geochemistry in Exploration oftke Norwegian Shelf

Fig. 1. Location map. Occurrence of pockmarks is indicated by stippled areas.

o
..

t:::::::!::::::
N
_
o 200m

c:J ~~= 6~U:ER SEOIiIlENTS 0 POCKMAilK 1_ 5m OEEP

Fig. 2. Shallow seismic section across pockmarks and


~ OCCURRENCE OF CA.~80NAfe 0 MOUND I-1m HIGH)

SMAL.L POCItMARICS tlDmWI DE


shallow gas clouds at the 25/7-location. G = shallow gas. "fJ~

Fig. 4. Local map showing pockmarks, shallow gas and


occurrence of calcium carbonate.

Fig. 3. Photograph of sea-floor inside pockmark where the Fig. 5. Section of the 25/7 slab. The assumed active
25/7-slab was retrieved. The slab is seen at the left, lower diagenetic zone is seen at the lower part of the slab.
centre.
Recently Formed Methane-derived Carbonates 265

Fig. 6. SEM photograph showing (a) migration duct for solutions or gas in the active zone showing globular calcite or
aragonite and (b) crystallization process in the assumed active zone (contact zone with sea-floor) of the 25/7-slab.

shown in Fig. 5. The main surface of the slab is very brown stained, inclusion-rich zone which is sometimes
uneven and jagged, is full of cavities and has lobate also marked by possible micro-borings.
protrusions (some of which suggest exposure to a Aragonite from the 25/7-slab was analysed for car-
reducing environment). A portion of the underside of bon and oxygen stable isotope ratios. The 8l3 C-values
the slab was, however, completely different. This were - 58.9%" and -59.0%0 PDB and the alSO-values
zone, which was probably the contact zone embedded 2.52%0 and 2.65%0 PDB (Hovland, 1984a). The carbon
in the sea-floor sandy sediments, displays several atoms thus found are obviously not taken from a nor-
uneven lithified sandstone ribs. Scanning electron mal marine environment. The a l3C-values are, how-
microscope (SEM) studies of this zone show that it ever, ofthe same order of magnitude as those found in
was probably an active zone of accretion and aragonite carbonates of assumed methane-derived origin from
cementation (see Fig. 6). other locations in the North Sea (J0rgensen, 1976) and
The slab generally consists of irregularly shaped on the continental margin off the northeastern USA
clasts of grey, very fine sandstone in a very fine to fine (Hathaway and Degens, 1970; Allen et al., 1969).
shelly sandstone matrix. Quartz is the dominant clas- Additional geochemical information from the sur-
tic mineral. Both the grey clasts and matrix contain a rounding sea-floor sediments was gathered by drop-
sprinkling of rounded glauconite grains. Feldspar, core sampling. Six dropcores were retrieved and 200 g
mica and amphibole are also present. While the clasts sediment subsamples preserved from the lowest por-
only contain poorly preserved foraminifera, the basal tion of the core (1-1.5 m sub sea-floor). The subsam-
sandstone (the presumed active zone) has a rather pies were packed in plastic bags and stored in liquid
diverse assemblage of foraminifera, molluscan and nitrogen (-196 DC) until geochemical analyses were
echinoderm shell fragments. carried out onshore. Samples were analysed for
Clasts and matrix also differ in style of cementation. adsorbed and occluded concentrations of light hyd-
The clasts are cemented by a granular Mg-calcite mic- rocarbon gases. Geochemical results are shown in
rospar with crystals typically 5-30 !lm in diameter, Table 1.
whereas the surrounding sandstone has a cement of The highest concentration of light hydrocarbons is
acicular aragonite. Of particular interest is the pres- found in core S16/2 which was retrieved from the soft
ence, in large pores under sandstone clasts (but not in sediments within the pockmark where the 25/7-slab
the active zone) and within shell fragments, of well- was recovered . This is also within the location where
developed botryoidal aragonite crusts and spherulites. the shallow gas apparently comes very close to the sea
These are apparently closely comparable with some of floor. The light hydrocarbon concentrations and a l3 C
the aragonite cements found in modern reef structures isotope values from the other five samples are of the
(James and Ginsburg, 1979; James, 1983). Some bot- same magnitUde as those described by Faber and
ryoids contain up to three generations of aragonite Stahl (1984) as being background values within this
needles, each generation being separated by a part of the North Sea. The gases found in the sedi-

TABLE 1. Geochemical results from the pockmarked area in Block 25/7


(ppb)
Methane Ethane Propane Butane Pentane Hexane
Ethene Propene
Sample Cl C2 C:l C4 C5 C6 %C 2 %C3 61SC I ClLCn

812/ 2 43 6 3 3 9 0 0.25 0.00 -40.0 0.858


816/ 2 128 21 10 8 7 0 0.09 0.09 - 43.8 0.874
816/1 50 7 4 3 10 1 0.34 0.09 - 40.4 0.855
817 58 8 5 4 11 0 0.45 0.19 -38.2 0.858
818 42 6 3 3 6 0 0.19 0.00 -39.8 0.868
819 49 7 4 3 10 0 0.23 0.14 -39.3 0.860
S21/2 92 15 9 7 9 0 0.12 0.00 -40.2 0.859
822 41 6 3 2 5 0 0.18 0.13 -40.6 0.870
Stone -58.9
266 Petroleum Geochemistry in Exploration of the Norwegian Shelf

ments have all the characteristics of being of a ther- being pumped (by cyclic loading?) through the floor of
mogenic origin. Since the isotope value of the carbon the pockmarks. The cementation process is thus most
within the 25/7-slab is lighter (-58.9%0 PDB) than the likely a result of biogenic methane oxidation taking
value of the methane in the sediments (approximately place in the unconsolidated sediments immediately
-40%0 PDB) it is assumed that either a geochemical below the benthic boundary.
isotope fractionation has occurred during the
diagenesis of the carbonate. or that the original
thermogenic methane in the sediments was biologi- REFERENCES
cally modified before participation in the diagenetic Allen, R. C., Gavish, E., Friedman, G. M. and Sanders, J. E. 1969.
process. Aragonite-cemented sandstone from outer continental shelf off
Delaware Bay: Submarine lithification mechanism yields product
resembling beachrock. J. Sediment. Petrology, 39, 136-149.
Faber, E. and Stahl, W. 1984. Geochemical surface exploration for
CONCLUSIONS hydrocarbons in North Sea. AAPG Bull., 68,363-386.
Hathaway, J. C. and Degens, E. T. 1969. Methane derived marine
From the study in Block 25/7 of the local occurrence of carbonates of Pleistocene age. Science, 165, 690-692.
five pockmarks associated with a restricted 'cloud' of Hovland, M. 1984a. Seabed features caused by shallow gas in the
North Sea. AAPG Research Conference, San Antonio, Abstr.
shallow gas the following conclusions may be drawn:
No. 19.
The five pockmarks in Block 25/7 have a causal rela- Hov land, M. 1900. Gas-induced erosion features in the North Sea.
Earth Surface Processes and Landforms, 9, 209-229.
tionship with the presence of the sub sea bed shallow James, N. P. 1983 Reef environments. In: Scholle, P. A., Bebout,
gas. The pockmarks are thus probably formed by a D. G. and Moore, C. H. (eds.), Carbonate Depositional Envi-
gas-induced erosion process (Hovland, 1984b). ronrnents. AAPG Memoir, 33, 346-462.
The occurrence of the lithified pockmark floor crusts James, N. P. and Ginsburg, R. N. 1979. The Seaward margin of
and carbonate slabs has a causal relationship with Belize barrier and atoll reefs. Int. Ass. Sedimentologists, Spec.
Publ., 3, 191 pp.
the pockmark forming process. This means that ther- Jllrgensen, N. O. 1976. Recent high magnesian calcite/aragonite
moge!Jic gas or a low density liquid solution containing cementation of beach and submarine sediments from Denmark. J.
light thermogenic hydrocarbons is emanating from or Sediment. Petrology, 46, 940-951.
Part V
Methods
23 Integrated kerogen typing in the
recognition and quantitative
assessment of petroleum source
rocks
S. R. Larter
Union Oil Science and Technology Division, 376 South Valencia Avenue,
Brea, California 92621, USA

Many of the kerogen typing procedures used in petroleum source-rock evaluation today are derived from basic
principles which were developed before this decade. Quantitative high-resolution pyrolysis-gas chromatography is
a recent technique which allows more detailed evaluation of kerogen in source rocks. Such detailed molecular
kerogen typing data may now be integrated with the older established chemical and petrographic methods. An
example is given of the application of integrated petrographic, bulk chemical and high-resolution chemical methods
to the quantitative evaluation of 'actual source beds' (source beds that have expelled oil and gas) using data from
the Viking Graben province of the North Sea. The limitations on the various kerogen typing methods are discussed
and restrictions on the value of conventional source rock analysis are considered in the light of our limited
understanding of the primary migration process.

INTRODUCTION analyses aimed at determining the gross origin and


organic facies of older sediments, and their catagene-
The characterization of kerogen, as the precursor for tic hydrocarbon potential, have been an area of exten-
petroleum, is now a major part of many contemporary sive study by geochemists in the previous decade (Tis-
exploration programs. Whilst the most common sot et al., 1974; Durand, 1980; Tissot, 1984; Tissot and
routine chemical methods of kerogen analysis used in Welte, 1978; Demaison and Murris, 1984).
petroleum geochemistry date from the middle seven- The concept of organic facies, which integrates the
ties (Durand, 1980; Tissot and Welte, 1978, and refer- environment of deposition and type of organic matter
ences therein), many of the basic concepts can be in a sediment and its chemical composition, is now
traced back to the coal science literature of previous coupled with the 'generative basin' approach, to pro-
decades. In particular, the petrographic maceral con- vide the basis for play and prospect evaluation in
cept (Stopes, 1935) and the investigation of chemical many exploration companies (Demaison, 1984). Whilst
variations between individual macerals and kerogens this approach is tried and tested (Demaison and Mur-
by CHO atomic ratio plots were well-established by ris, 1984, and references therein) there seems to be a
the 1960s (Huntjen and Van Krevelen, 1957; Van general movement towards application of more
Krevelen, 1961). Many current kerogen-typing proce- quantitative and absolute approaches to the evalua-
dures are the result of merging older concepts with tion of petroleum source beds. These more quantita-
the application of newer analytical technology. tive methods are aimed at a better understanding of
Kerogen-typing studies may have a variety of the complexities of the petroleum migration process
objectives of which those related to petroleum (Durand, 1983; Welte and Yukler, 1981; Ungerer et
geochemistry are listed in Fig. 1 and summarized al., 1984). In order to discuss the role of integrated
below. Evaluation of the depositional environment of a kerogen typing in the evaluation of source rocks,
sediment using reflected and transmitted light mic- objectives and definitions need to be clarified. In this
roscopy (Gutjahr, 1983) may have important implica- context the dual objectives of kerogen typing are
tions either for source-rock potential or in terms of primarily considered to be discrimination of the source
reservoir development and quality. organic matter for a given facies, and its significance
Kerogen analysis and typing related to organic in terms of the actual source potential of the sediment
diagenetic studies of the formation of chemical species (see Fig. 1).
in modern sediments (see Fig. 1), whilst of some The two major factors controlling the petroleum
theoretical importance, has not been a major area of potential of any sediment at any rank level are (1) the
interest for exploration geochemists. In contrast, quantitative fraction of the sediment that contains
Petroleum Geochemistry in Exploratioll or the Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 269-286.
270 Petroleum Geochemistry in Exploration orthe Norwegian Shelf

OBJECTIVES PETROGRAPHIC TYPE KEROGEN RANK LEVEL


1. DEPOSITIONAL FACIES DETERMINATION - (MACERAL COMPOSITION, (VITRINITE REFLECTANCE,
"PALYNOFACIES" SOURCE OF ORGANIC MAnER) TAl, TIl-THERMAL HISTORy)
APPLICATION - DETERMINATION OF SOURCE/RESERVOIR
SYSTEM GEOMETRY AND PALAEOGEOGRAPHY

2. DIAGENETIC INTERPRETATIONS
~ BULK CHEMICAL TYPE
/
(GROSS DISTRIBUTION OF H,C,O; GROSS HYDROCARBON
APPLICATION - UNDERSTANDING ORIGIN AND EARLY
YiElD; H/C-O/C RATIOS; HYDROGEN INDEX

!
FATE OF SEDIMENTARY ORGANIC MAnER INCLUDING
KEROGEN. UNDER A VARIETY OF DEPOSITIONAL
ENVIRONMENTS

3. ORGANIC FACIES DETERMINATION


MOlECULAR TYPE
APPLICATION - DETERMINATION OF THE NATURE OF THE
ORGANIC MAnER IN A SEDIMENT. COUPLED WITH (DISTRIBUTION AND QUANTITY OF POTENTIAL
MATURITY AND RICHNESS DATA. USE OF THIS HYDROCARBON SKElETON TYPES REMAINING IN THE KEROGEN)
INFORMATION IN THE GENERATIVE BASIN APPROACH TO
REGIONAL AND PROSPECT SCALE SOURCE POTENTIAL
EVALUATION AFTER LARTER AND SENFTlE (1984)

SOME WORKING DEFINITIONS


Fig. 2. Elements of source-rock kerogen type.
SOURCE ROCK POTENTIAL: -

THE FRACTION OF A SEDIMENT. AT ANY MATURITY LEVEL. ition (classical chemical type), the detailed molecular
THAT CONTAINS RELEASABLE (BY CATAGENESIS OR OTHER composition, its rank level (maturity) and the deposi-
MECHANISMSI. EXPELLABLE PART STRUCTURES RELATED TO
THOSE CHEMICAL SPECIES FOUND IN PETROLEUM tional environment of the host sediment (see Fig. 2).
SOURCE ROCK POTENTIAL f (ORGANIC RICHNESS. BITUMEN
Applications of kerogen typing methods to explora-
tion geochemistry fall into two categories: (1) screen-
0

AND KEROGEN COMPOSITIONI


ing analyses aimed at identifying the gross organic
MOLECULAR KEROGEN TYPE: -
type in a large number of sediment samples and (2)
A CHEMICAL DESCRIPTION OF A KEROGEN BASED ON THE characterization in detail of relatively few samples of
QUANTITATIVE RELATIONSHIP BETWEEN THE DIFFERENT
STRUCTURAL ELEMENTS COMPOSING THE MATERIAL prime interest in the principal source intervals. The
need to screen samples before time-consuming
ACTUAL SOURCE ROCK detailed analysis is made obvious when one considers
ANY SEDIMENT THAT HAS PRODUCED AND EXPELLEO that 5000 samples are available when 10 wells are
HYDROCARBONS AND RELATED SPECIES analysed at 10 m sampling intervals assuming 5 km
Fig. 1. Kerogen typing-objectives and definitions. average total depth. In this context, petrographic and
bulk chemical typing procedures (chiefly Rock-Eva!)
are the most suitable for screening kerogen studies.
catagenetically releasable structures related to those Elemental analysis, NMR and pyrolysis-GC methods
chemical species found in petroleum and (2) the expul- are suitable for more detailed evaluations.
sion efficiency of the individual bitumen structural
types from the host sediment. The second factor,
which is a complex function of organic content, organic PETROGRAPHIC TYPE
type, host mineralogy, sediment rank level and hyd-
rodynamic potential, is the ultimate parameter gov- The value of optical kerogen typing in source rock
erning effective source potential. analysis has been well documented (Staplin, 1969;
Crude oils (and gases) are often described in precise Robert, 1971; Powell et al., 1982; Gutjahr, 1983).
compositional terms. In contrast, the inability of many Whilst there are broad general relationships between
of the historically established kerogen-typing techni- petrographic kerogen typing and chemical kerogen
ques (such as elemental analysis) to define the detailed composition parameters (Powell et al., 1982), the wide
molecular nature of kerogens accurately has held back range of chemical properties displayed within the
current typing schemes to general descriptions of bulk same maceral group (Bailey, 1981; Horsfield, 1984;
kerogen composition. Whilst these schemes, based Larter, 1984) renders classification by solely petrog-
mainly on the work of Tissot and co-workers (Tissot et raphic means very hazardous, and often actually mis-
al., 1974; Durand, 1980) have provided a solid founda- leading. However, petrography (maceral description)
tion for new approaches to typing kerogens, their currently provides the best way of identifying precur-
usage alone as the major base for chemical typing is no sor organic matter, and provides the only practical
longer consistent with the analytical resolution now way of attempting the typing of very high rank mater-
obtainable. Ideally, the characterization of kerogens ial. At high maturities the chemistry of the kerogen
should attain the same level of molecular detail as is assemblages changes significantly and the values
possible with crude oils or source-rock bitumens (Tis- obtained from both bulk (Tissot et al., 1974) and
sot and Welte, 1978). detailed chemical typing (Horsfield, 1984; Larter,
It is the application of quantitative kerogen-typing 1984) may become misleading. With high rank sections
studies to the evaluation of the source rock potential in explorationally immature basins (where drilling of
and recognition of actual source rocks (see Fig. 1) that source rocks over the entire thermal evolutionary
is the purpose of this paper. track has not occurred), often the only way to attempt
reconstruction of the maturation pathway through
time of a possible source rock is via petrographic typ-
INTEGRATED KEROGEN TYPING ing, although this must be used with caution.
Whilst petrographic data cannot a priori evaluate
A complete, integrated description of a kerogen organic matter chemically, once calibrated for a par-
should include information relating to its visual com- ticular basin, petrographic data provides a rapid and
position (petrographic type), its bulk chemical compos- inexpensive approach to determining the source of the
Assessment Using Quantitative Kerogen Typing 271

~ 1~ ,- ________,

.... -
l( ~ flOOUIi
~.,,.-. ~
Tl PE LOG
........
-..".... ...

II)
cr
....ww
::;;
o
'"

L ~ ~
0 .0<1 0 .1110 1.20 l to ~.AO 300
~ O"GA NIC Colli ION
" ' OO, N""'" .. . ,;.. ""'" '20001100<>20000
.... .,..
=
=
=
INERTINITE CHEMICAL KEROG EN TYPE , - 600 ' C PYROGRAM
<= VITRINITE
EXI NITE
PYROGRAM PEAK LEGE ND 1600' C/ l SECOND PYROLYSIS )
1.
AMOR PHINITE 2.
Tolu ene
nCB al kene + alkane

MAJOR EOM RICH ZONE CORR ESPO NDS TO ZONE WHEN HIGH TOC AND!' ~r~~ ~k:~benz. nes,
LARGE PERC ENTAGES OF FLUOR ESC ENT AMORPHOUS KEROG EN OCCUR 5: nC 20 aikene + alkane

Fig. 3. Optical and chemical typing data for section near peak oil generation, Kimmeridge Clay Formation, Central Graben.

organic matter in a sediment, and helps in making their dominantly vitrinite-rich nature (Yarzab et al.,
preliminary predictions of the possible source poten- 1980). It is important to consider that many different
tial. Frequently the chemical properties of a source petrographic compositions may result in a similar bulk
bed are closely related to the amount and distribution chemical character.
of exinite and fluorescent (under u. v. illumination) The Brent Group coals in the Viking Graben of the
amorphous kerogen types in a sediment. In this paper, Norwegian North Sea fall into type 3 although, as
the sum of the fluorescent amorphous kerogen species discussed below, the chemical variance in coals from
(presumed exinite/liptinite-derived) and the mor- different provinces is very great, and the Brent coals
phologically defined exinite macerals are summed into themselves have relatively high hydrogen indices for
a parameter termed '%AMEX' (Senftle, 1983). This 'coals' and may not fall on the conventional Tissot et al.
parameter is frequently found to be the most critical in (1974) type 3 evolution curve. Brent coals may,
relating petrographic and chemical source rock data. indeed, have some liquid hydrocarbon generating
Figure 3 shows data from a well in the Central Graben potential (M. Bjorl'lY, pers. comm.; Ungerer et al.,
of the Norwegian North Sea, illustrating an example 1984; Durand and Paratte, 1983).
of the correspondence between the chemical richness
and petrographic parameters. This type of approach is
of greatest value in the initial stages of basin explora- MOLECULAR TYPE
tion (Cooper and Barnard, 1984). It must be emphas-
ized that the effective use of petrographic data is The detailed molecular configuration of a kerogen
strongly dependent on a cross-calibration of petrog- cannot be obtained from examination of its bulk chem-
raphic and chemical data (see Fig. 3), and on the com- ical properties, due to the isomeric complexity poss-
bined use of reflected light on whole rock samples and ible with high molecular weight macromolecules. The
reflected and transmitted light on kerogen concen- mechanically complex (poly-maceralic) nature of most
trates, both in white and u. v. illumination. It is the kerogen assemblages further renders any 'average-
use of this multi-mode illumination (Senftle, 1983) that structure' determination for a kerogen a somewhat
is enhancing the value of petrography in an explora- academic exercise. However, the total kerogen reser-
tion context. Simple white transmitted-light petrog- voir of precursors of, for example, a petroleum species
raphy does not have the resolution consistent with the such as normal alkanes, is a primary factor in deter-
chemical and petrographic variations evident in mining the molecular kerogen type. This is so even if
source-rock kerogen components. the alkane precursors (such as kerogen-bound alkyl
groups) are anisotropically distributed between dif-
ferent macerals. Thus, in a molecular chemical type it
BULK CHEMICAL TYPE is the total quantity of a kerogen structural component
in the kerogen mixture as a whole that is significant,
As commonly defined (Tissot et aI., 1974), the three rather than whether the parent alkyl group is bound
main kerogen types are broadly analogous to the prin- to a sporinite, amorphinite or a vitrinite.
cipal coal macerals alginite, exinite and vitrinite, the
atomic ratio evolution paths for both kerogens and
macerals being similar (Huntjen and Van Krevelen, DETERMINATION OF MOLECULAR
1957; Van Krevelen, 1961). A fourth kerogen, type 4 KEROGEN TYPE
defined by Harwood (1977), is analogous to the inerti-
nite coal maceral group with very low values of atomic In defining a basis for molecular kerogen typing any
HIC ratio at any rank level. potential method should be based on elements
Most kerogens are poly-maceralic assemblages. (species) in a kerogen that will ultimately end up, after
Thus coal kerogens are frequently mixtures of compo- catagenesis, as a significant component of crude oils
nents ranging from types 1 to 4, but for many and gases (see Fig. 1). Secondly, the method must be
Palaeozoic-Mesozoic coals bulk analyses would indi- based on kerogen molecular elements that can be
cate a type 3 composition, this being consistent with satisfactorily and reproducibly analysed. The general
272 Petroleum Geochemistry in Exploration of the Norwegian Shelf

~~ x
~
(i!~ ~
%u Z
~
uz
~o 8z
::lU
d (i!%'" ~
<0
(!)W
\:!
(!)
z5 z '"
:5
~~ ~ '"
~~ U'"~
~

--PROCESS YIELD-

Fig. 4. Kerogen type and process yields.

A B
REPRESENTATIVE KEROGEN PYROGRAMS RELATIVE DISTRIBUTION OF
TYPE IV KEROGENS n-OeTENE. m + p XYLENE
CAR80NIFEROUS - SCOTLAND - WES11'IELD FUSAIN AND PHENOL IN 600°C
KEROGEN PYROLYSATES

N OCTENE
100%
POLYETHYLENE

AREA OF PETROLEUM
ASPHALTENES

0= n-oclene
M = (m + p) xylene
P = phenol

TYPE III

100%
PHENOL

PYIIOGRAMS OBTAINED ON A MODIFIED CDS 120 PYROPROBE.


1 SECOND PYROLYSIS (N2) AT 600"C. PYROLYSATE CHROMATO-
GRAPHED FROM 4()'280'CAT 4'!MINUTEONASP2100FUSEDSILICA
WCOl COLUMN.

Fig. 5. Discrimination of different organic matter types using pyrolysis gas chromatography.
Assessment Using Quantitative Kerogen Typing 273

abundance and ease of resolving normal and other Figure 5 shows how the basic kerogen types can be
aliphatic hydrocarbons in flash kerogen pyrolysates distinguished in terms of both pyrogram fingerprint
(Larter, 1984; Horsfield, 1984) suggests that these and the relative distribution of aliphatic, aromatic and
components are a good candidate for such a method. phenolic (n-octene, xylene, phenol) species in the
Miknis et al. (1982) have recently discussed the pyrograms. A complete treatment of this procedure is
importance of the abundance of the aliphatic content of given in Larter (1984).
a kerogen (carbon-bound hydrogen content or the The kerogen fields shown are based on the classic
more easily determined and somewhat equivalent scheme of Tissot et ai. (1974). Some overlap of type 3
aliphatic carbon content) to high temperature thermal and 4 kerogens is evident, but a fairly clear discrimi-
conversion to gases and liquids. In addition, extensive nation between the hydrogen-rich types 1/2 and more
source-rock kerogen data, studies of various pet- hydrogen-deficient types 3/4 is evident. Intermediate
roleum basins and the similarities between high kerogen compositions plot accordingly. Samples cover
temperature-induced laboratory experiments and the vitrinite reflectance range 0.3-1.3% Ro with most
natural catagenic processes (Tissot and Welte, 1978, samples having values of <1.0% Ro. A major strength
and references therein) all suggest that the aliphatic of pyrolysis methods is that while quantitative yields
carbon content of a kerogen is the critical factor in of kerogens are maturity-related (Espitalie et al.,
determining catagenic hydrocarbon yields in the 1977) the qualitative pyrogram fingerprints obtained
Earth's crust. While total hydrocarbon yield (i.e. pro- are relatively rank independent over much of the oil
cess yield in Fig. 4) is proportional to the aliphatic window (Van Graas et al., 1980).
content of the kerogen, the product distribution (in The main distinguishing features of type 3/4 versus
gross terms the gas-oil ratio) is dictated by the dis- type 1/2 kerogens are the relative abundances in their
tribution of the various structural elements in the pyrograms of aromatic and phenolic species versus
kerogen, i.e. the molecular type (see Fig. 4). aliphatic species, respectively (see Fig. 5). Most
The aliphatic fraction of kerogen has been deter- importantly, enormous variation in detailed composi-
mined by solid-state carbon NMR methods (Zilm et tion can occur, even for kerogens of very similar ele-
al., 1981; Miknis et al., 1982) and these methods are mental or Rock-Eval analysis. This is the major
becoming more routinely used in the evaluation of reason for calibrating bulk methods with high-
source kerogens for exploration purposes. However, resolution methods.
the most commonly used method for estimating the While discrimination within or between the various
proportions of discrete structural elements in kerogen types is possible using either the pyrogram
kerogens would seem, at present, to be pyrolysis fingerprint or relative distributions of pyrolysate
methods, in particular pyrolysis-gas chromatography components, it is not easy to relate, directly and quan-
(Larter and Douglas, 1982; Horsfield, 1984; Larter, titatively, this information to source-rock potential.
1984). This is because, as with other contemporary chemical
Analytical kerogen typing methods have greater typing methods, no absolute measurement of the
chemical resolution at the molecular level than routine quantitative ability of a kerogen to generate specific
methods such as elemental analysis, bulk flow petroleum hydrocarbons is readily available.
pyrolysis (e.g. Rock-Eval) or optical spectroscopic The major success of bulk flow methods such as
methods. Further, since high-resolution methods Rock-Eval has been that they allow some estimate of
directly monitor specific chemical species in a kerogen the amount of hydrocarbons generated by a kerogen
pyrolysate, such data are more pertinent to the type. Orr (1983) suggested that hydrocarbon potential
determination of kerogen aliphatic fractions, and thus may be indirectly related to the difference in bulk flow
kerogen behaviour under natural processes. Tradi- pyrolysis yield for kerogens of different rank but con-
tional bulk chemical methods cannot determine the stant chemical type. However, while combined
distribution of carbon types in a system, except in a chromatographic/bulk flow pyrolysis systems (Martin,
most general way. Rock-Eval, in its most basic form, 1977; Gormely and Mukhopadahay, 1983; Horsfield,
can determine the kerogen aliphatic content, but not 1984; Dembicki et ai., 1983) have enabled improved
its distribution. characterization, the ultimate aim of a completely
The typing of kerogens by determination of the quantified on-line analysis of kerogen pyrolysates on a
relative quantities of specific fragments in kerogen routine basis has not yet been achieved.
pyrolysates is an old idea, this concept having first Ultimately, as discussed earlier, classification of
been applied in coal studies (Romovacek and Kubat, kerogens should be performed on the same molecular
1968; Larter and Douglas, 1982). The more basis as that used for crude oils and rock extracts.
hydrogen-rich exinite coal maceral kerogens generate Ideally kerogen analyses should enable the generation
fewer aromatic pyrolysates than the more hydrogen- of data analogous to that used by Tissot and Welte
deficient vitrinite macerals. More recently this work (1978) in their proposed oil classification, namely the
has been extended to kerogens from a broader range distribution of hydrocarbon and non-hydrocarbon
of sedimentary facies (Van de Meent et al., 1980; Lar- types within a kerogen pyrolysate. Such data allows a
tel' and Douglas, 1980; McHugh et ai., 1978; Giraud, direct assessment of the quantity and type of carbon
1970; Horsfield, 1984; Larter, 1984). skeletons potentially available in a kerogen, and
Analytical pyrolysis techniques measure specific allows the petroleum geochemist and explorationist to
carbon: hydrogen: oxygen assemblages. They pro- attempt more-sophisticated mass balance approaches
vide a reliable kerogen typing method in that the to quantifying oil-expulsion phenomena.
aliphatic carbon-bound hydrogen content of a kerogen, Some idea of the need for such a quantitative basis
the factor primarily responsible for high hydrocarbon for detailed molecular kerogen typing is illustrated in
yield during catagenesis, is more directly assessed. Fig. 6, which shows pyrograms of four kerogens con-
Using pyrogram fingerprint data, it has been possible taining predominantly non-marine organic matter.
to distinguish the major variations between kerogens These are (A) a kerogen derived from the alga BotJ'yo-
readily, even those of the same bulk chemical type. coccus, (B) a kerogen derived from higher plant cutic-
274 Petroleum Geochemistry in Exploration of the Norwegian Shelf

A) SAMPLE 9 TYPE I,ALGINITE - PERMIAN, FRANCE

B) SAMPLE 32 TYPE II,CUTINITE RICH KEROGEN- TERTIARY ALASKA

LEGEND
S " INlERNAL STANDARD
9 " nCQ ALKENEI ALKANE
p" PRIST·1·ENE
22 " nC22 ALKENE! ALKANE

C) SAMPLE 20 TYPE III.VlTRINITE RICH KEROGEN' TERTIARY INDONESIA

D) SAMPLE 2 TYPE III,COAL VlTRAIN-CARBONIFEROUS. PENNSYLVANIA

'JJ~J~Jl.)~.
I \1 ~'
LEGEND
S = INTERNAl.; '.-.: ·'..h~D
9 = nC9 ALKENE/ALKANE
P = PRIST·1·ENE
22 " nC22 ALKENEI ALKANE

Fig. 6. Pyrograms of kerogens containing primarily non-marine organic matter.


Assessment Using Quantitative Kerogen Typing 275

A. COMPOSITE S1 + S2 CHROMATOGRAM OF IMMATURE KIMMERIDGE


CLAY FM. SHALE (PYROLYSIS FOR 3 MINUTES AT 250°C FOLLOWED BY--
PROGRAMMED TEMPERATURE PYROLYSIS FROM 250-600°C AT I
25°CjMINUTE IN HELLIUM.

~
__ _ ~ nC11 alkane, alkene

52
phytane
I
\11
I I
I
nC8
I
I ;I
'II I
EFFECTIVE
RANK LEVEL

)'=: I I I' I II I I II I ~I D.3Uo

""" I I, I '!I I,IIII~~.'I,I~I:'I!I" i i i


_uJ1U ~~,!~.J N~II! 11~11~~~!ill :!~lJ!' il,Jk,lll; ~,!J.l;l i "
:1 1/

, .' !.

Tmax
-~----~=+--C-O-M-PO-SI-TE-S-1-+-S--2-C-H-R-O-M-A-T-O-G-RA-M-O--:F=-A--:R--:T=IF:-:IC=IA=-L:-:-LY:-:-::-:M=A=TU-=R:::=ED -r---
SHALE (HYDROUS PYROLYSIS - 3 DAYS AT 325°C)
i nC8 I
I
I I !~. , .
EFFECTIVE
RANK LEVEL
~O.8% Ro.

.,, ,

~ III, I' I ' I 'I II ,

/---~~ ~~DmONS. CO,"~N !~~~~I,iII1i~l,i,j',.111 J: J': II :


, I I

25m OV1D< FUSED SILICA WCOT II IiIIW


PROGRAMMED FROM O°C-280°C AT 4°C/MINUTE.

16 18 2121 36 4121 42 I
• . GAS CHROMATOGRAM OF THE SATURATED HYDROCARBONS IN
N 8121.
n·alkane
carbon # ~ THE ARTIFICAL OIL PRODUCED AND EXPELLED FROM THE SHALE

,
pristane I

....
L
; l 60. phytane
"m
L
,
a. ~ 40.

121. 4. B. 12. 16. 2121. 24. 28. 32. 36. <10. 44. 48. 52. 56. 6121. 54. 68. 72.

Fig. 7. Selective generation of isoprenoid species in artificial maturation experiments.

ular material and deposited in a marine setting, (C) bon basis, the amount of specific components in a
a kerogen typed petrographically as dominantly vitri- pyrolysate (Larter and Senftle, 1984). In the following
nite, deposited in a marine setting and (D) a kerogen discussions emphasis is placed on the quantification
consisting entirely of vitrinite picked from a coal of normal hydrocarbons in kerogen pyrolysates, since
seam. All the kerogens are at rank levels below peak these components are a major fraction of many crude
oil generation (Le. <0.8%Rd). Samples C and D have oils, particularly those of higher plant origin (Tissot
very similar bulk chemical analyses. and Welte, 1978).
The large peak in each pyrogram represents an By directly determining the total reservoir of
internal standard (methyl styrene) generated in the specific pyrolysate elements, direct assessment of
pyrolysis process from a polymer (poly- actual petroleum potential for any kerogen becomes
methylstyrene) quantitatively mixed with the kerogen possible (see Fig. 1) if this type of information can be
before pyrolysis (Larter and Senftle, 1984). The abso- integrated into quantitative migration studies of the
lute quantity (relative to starting kerogen weight) of type discussed by Mackenzie et al. (1983) and Leyth-
the internal standard in each pyrogram is constant, aeuser et al. (1983). Further, this method allows the
approximately 0.25% by weight of the kerogen (see systematic discrimination of kerogens previously
Appendix 1 for details). Using such a technique it is grouped together, on the basis of elemental analysis or
now possible to quantify, on a kerogen or organic car- hydrogen/oxygen index, as of similar type.
276 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Thus, for example, the two type 3 kerogen pyrog-


rams shown in Fig. 6 (C,D), while both consisting pet-
40~-----------------------------------.
t AUSTRALIAN TORBANln IPERMIA.)

rographically of 'vitrinite'-dominated materials, are ERROR BARS REPRESENT


2uVARIATIOII FROM MEAN
clearly different in their pyrolytic chemical details: 36
e.g. the sample from the deltaic shale contains a signif-
icant amount of potential normal hydrocarbons. DATA FROM 800 0 C120 SECOND PYROLYSES

Ultimately, a complete chemical characterization of ~ 32


the kerogen pyrolysate is desirable such as that
'"<I:c
z
obtainable by quantitative pyrolysis-mass spec- ;:
trometry. With gas-chromatographic methods, accu- ...'"c 28 o
8
TYPE I
TYPEII
ALGlNITES fROM AUSTRALIA
MIOCENE CAlifORNIA. TRIASSIC N. SEA JURASSIC PARIS BASIN,
rate quantification is limited to species easily iden- ~ JURASSIC N. SEA, TERTIARY ALASKA. CRETACEOUS, ALASKA. CARBONIFEROUS UK

tified and resolvable chromatographically. These ~ 24 o


::e TYPEIII JURASSIC N. SEA, MIOCENE INDONESIA, CRETACEOUS CANADA
methods allow the direct quantification of specific
'"
X TYPE IV CARBONIFEROUS UK

+
c
structural elements in kerogens and their quantitative z
I AUSTRALIAN TASMANITE (PERMIAN)
evolution with maturity. Figure 7 shows '"
z 20
programmed-temperature pyrograms of a Kim- .,
c
meridge Clay Formation sample, both before and 5'"

+f--+
after a simulated maturation experiment (hydrous ~ 16
c
pyrolysis, 3 days at 325°C in helium, cf. Lewan, >
:c
.....
1983). The kerogen in the initial sample is rich in <I:

isoprenoid structures capable of producing prist-1- ~ 12

~(~+::
c
z
ene. Pyrograms run in the same manner as those of on
....
Fig. 6, with a standard incorporated, indicate a con- U
I

centration of about 0.025 wt% of prist-1-ene precur-


sors in the initial material. In contrast, after hydrous
S
IN
8
+ ' . . _.. -/ SPORINITES

pyrolysis (see Fig. 7) the remaining material is rela-


tively devoid of isoprenoid species, while the expelled
oil is rich in pristane and phytane. For the Kim-
meridge Clay Formation in the North Sea, and adja-
1 2 J 4 5
cent basins, the absolute isoprenoid concentrations in 1:C6-C8 AROMATICS NORMALIZED TO STANDARD = 1
kerogens vary from around 0.02 wt% down to zero 1 STANDARD UNIT APPROXIMATELV EUUALS O,Z5% BV WEIGHT. OF KEROGEN

near the base of the oil window. The ability to quantify QUANTITATIVE DATA OBTAINED BV AUTOMATIC PEAK RECOGNITION, PEAK AREA DETERMINATION

the generation of specific structural species from Fig. 8. Normalized abundance (by weight of kerogen) of
kerogens and to monitor their appearance in associ- aliphatic and aromatic hydrocarbons in low maturity
ated bitumens, coupled with mass balance approaches kerogen pyrolysates «0.8% VR).
to stUdying oil migration phenomena (Mackenzie et
al., 1983) enables the geochemist to now study the
complete source-rock kerogen/bitumen system. pyrolysates of type 2/3 kerogens have also been made
Although the relative compositions of kerogen previously by Tissot and Vandenbrouke (1982).
pyrolysates have been widely used to indicate kerogen Figure 9 shows pyrograms of kerogens which are
typical of distinct molecular kerogen types at low rank
type (Giraud, 1970; Larter and Douglas, 1982, and
references therein) it is important to remember that level. Petrographic types are also included for these
the major change in pyrolysis behaviour when going components and rank levels are shown in Table 1. The
basis for identifying the various end-member types is
from, say, kerogen type 1 to type 3, is the increase in
involatile residue during this transition. Thus, shown in Fig. 10, which is a plot of paraffin precursor
although type 3 kerogens (particularly coal vitrain vit- content, i.e. the paraffin index (based on measure-
ments of peak intensity for normal hydrocarbons and
rinites) have the highest relative aromatic content in
their pyrolysates, it is the more hydrogen-rich the internal standard in 800 °C/20 s pyrograms) ver-
sus hydrogen index for several kerogens. The split
kerogens that produce, in absolute terms, the most
hydrocarbons of any type. In other words, the more into specific kerogen type is based on examination of
many pyrograms (Larter and Senftle, 1984).
hydrogen-rich a kerogen is, the more hydrocarbons of
The recognized kerogen type end-members are as
any type, including gases and aromatics, are produced
follows.
under pyrolytic (or indeed catagenic) processes. How-
ever, the ratio of aromatic to aliphatic hydrocarbons in Type lp (paraffinic) kerogens are characterized by
the thermal product always increases in the progres- having high hydrogen indices and highly paraffinic
sion from type 1 to type 3. Figure 8 shows the rela- pyrolysates. They are typified by the Botryococcus
tionship between the absolute yield of aromatic -derived boghead coals (Larter and Douglas, 1980).
(expressed as total benzene, toluene and Cs aromatics) Type lpn (paraffinic/naphthenic) kerogens, while
and paraffinic (total Cr C25 n-alkanes + n - alkenes) hydrogen-rich, have the hydrogen distributed be-
species for a variety of kerogens. Interestingly, the tween paraffinic and naphthenic structures. Typical
absolute paraffin content of a kerogen (on a examples include kerogens derived from Tasmanitid
pyrolysis-GC basis) seems relatively constant for algae and many mixed type 1 kerogens (Philp et al.,
many immature type 2/3 kerogens. It is the larger 1982). The term paraffinic/naphthenic is used as mass
branched chain + naphthenic contributions of marine spectral data indicate that kerogens of this type have
systems that provide the higher hydrogen indices of abundant naphthenic contributions to their pyroly-
type 2 amorphous kerogen, such as are found in the sates (Philp et al., 1982; Larter, unpubl. data).
Kimmeridge Clay Formation of the Central and Vik- Type In (wholly naphthenic) kerogens are unusual.
ing Grabens. Similar observations on the relative con- The example shown here is a somewhat artificial
stancy of the normal hydrocarbon contents of kerogen illustration as the sample is a coal resinite.
Assessment Using Quantitative Kerogen Typing 277
~ SAMPLE 4 TVPt: I P ALGINITE E) SAMPLE 12 TYPE II/III P VITRINITE, EXINITE

KUPFERSCHIEFER-NORTH SEA
9 S
9 S

F) SAMPLE 24 TYPE II/III pp SPORINITE


SAMPLE 8 TYPE I pn, ALGINITE

22

LECHND LECHND
S INTERNAL STANDARD
S ~ INTERNAL STANDARD
9 nC9 ALKENE/ALKANE
(800'C/20 SECOND PYROLYSES) 9 ~ nC9 ALKENE/ALKANE
P PRIST·HNE
p ~ PRIST·1-ENE
22 nC22 ALKENE/ALKANE
22 ~ nC22 ALKENE/ALKANE

G) SAMPLE 22 TYPE IIIpp VITRINITE


SAMPLE 28 TVPt: I n, RESINITE S I

i·lL,
11_t----hPHENOLS
NAPHTHENIC SPECIES I I P

" I; !
III
I: 22
l~ llJJ.LUU,I,;lJ,.~
H) SAMPLE 2 TYPE III ap, VITRINITE

S
'I 1 'I' r -r '~d'" 71'"" -- '>"

I' I I ! PHENOLS

.~J.~u~~-
>AMPLE 5 TYPE II pn, AMORPHINITE

KIMMERIDGE CLAY FM, ONSHORE UK.


5

I) SAMPLE 23 TYPE IV, FUSINITE

SUBSTITUTED NAPHTHALENES
I
LEGEND
S = INTl:RNAL STANDARD
LEGEND 9 ~ nC9 ALKENE/ALKANE
S = INTERNAl STANDARD p = PRIST·1-ENE
9 :: nC9 ALKENE! AlKANE 22 ~ nC22 ALKENE/ALKANE
P :: PRIST-1-ENE
22 = nC22 ALKENE/AlKANE

Fig. 9. Principal kerogen types as defined by pyrolysis-G.C. (a) Type I, (b) types I and II, (c) type II/I11 and (d) types III
and IV.
278 Petroleum Geochemistry in E~>ploration of the Norwegian She((

TABLE 1. General sample information (refer to Fig. 10)


Bulk Rank
Sample chemical Principal (>JO%) level
No. Age /location type maceral contributors (VR)
1 Tertiary-Alaska 2/3 Amorphinite, vitrinite 0.6
2 Carboniferous-Pennsylvania 3 Vitrinite 0.6
3 Cretaceous-West Africa 2 Amorphinite 0.7
4 Permian-Australia 1 Alginite
5 Jurassic-North Sea 2 Amorphinite 0.3
6 Jurassic-North Sea 2 Amorphinite, vitrinite 0.8
7 Tertiary-Canada 2/3 Amorphinite 0.4
8 Permian-Australia 1 Alginite 0.4
9 Permian-France 1 Alginite 0.4
10 Jurassic-France 2 Amorphinite 0.3
11 Jurassic-N orth Sea 3 Vitrinite 0.8
12 Permian-North Sea 2/3 Vitrinite, exinite 0.6
13 Tertiary -California 2 Amorphinite 0.3
14 Jurassic-North Sea 2 Amorphinite 1.1
15 Jurassic-Alaska 2 Amorphinite 1.3
16 Cretaceous-Canada 3 Vitrinite 0.5
17 Cretaceous- North Sea 3 Vitrinite 0.7
18 Cretaceous-North Sea 3 Vitrinite 0.6
19 Ordovician-USSR 2 Alginite
20 Tertiary - Indonesia 3 Vitrinite, amorphinite 0.6
21 Cretaceous-Canada 3 Vitrinite 0.6
22 Tertiary - Indonesia 3 Vitrinite, amorphinite 0.3
23 Carboniferous-UK 4 Fusinite
24 Carboniferous-UK 2/3 Sporinite 0.9
25 Permian-North Sea 2/3 Vitrinite, exinite 0.6
26 Cretaceous-Alaska 2 Amorphinite 0.5
27 Jurassic-France 2 Amorphinite
28 Tertiary - Indonesia 1 Resinite 0.4
29 Carboniferous-UK 2/3 Sporinite 0.6
30 ,Jurassic-North Sea 3 Vitrinite 0.5
31 Tertiary-Alaska 2 Cutinite, vitrinite 0.5
32 Tertiary-Alaska 2 Cutinite, vitrinite 0.5

Type 2pn (paraffinic/naphthenic) kerogens are the (type 3pp) or vitrinites from the Carboniferous coals of
most common type 2 materials, being representative Northern Europe and the USA (Type 3ap). These
of kerogens found in rich marine source beds such as types are end members of a continuum and do not
the grabenward part of the Kimmeridge Clay Forma- represent the entire field of kerogen occurrence.
tion in the North Sea, Toarcian of the Paris Basin, Type 4 kerogens are generally aromatic in pyroly-
Miocene of Southern California basins and the Lower sate terms with occasional remnant phenolic contribu-
Cretaceous of the Alaskan North Slope region. Such tions, although inertinites in some Australian coals
kerogens are characterized by high hydrogen indices, may have significant aliphatic contents (Evans et al.,
but relatively low paraffin contents. The pyrolysates 1984).
are dominated instead by abundant naphthenic/ Quantitative information of the type described
naphtheno-aromatic components (Tis sot and above can be used to directly assess the total quantity
Vandenbrouke, 1982). of specific species released from a kerogen during any
Type 2p (paraffinic kerogens) are typified by maturation pathway history. This information is
reduced hydrogen indices but relatively high paraffin relevant to the assessment of petroleum migration
contents. They generally have a substantial exinite concepts. However, the analytical method presented
component in the kerogen. here does not completely define the total paraffinic
Type 2/3pp (paraffinic/phenolic) kerogens, in addi- carbon content of a kerogen, in that no estimate of the
tion to exinite contributions, may have phenolic con- relative amount of long-chain alkylated naphthenic or
tributions to their pyrolysates resulting from land- aromatic species is made. This is significant, since
plant structural polymers which may be lignin or long-chain alkyl-aromatic hydrocarbons are abundant
sporopollenin/suberin-derived. components of some kerogen pyrolysates (Larter et
Type 3 materials can be most easily split into two al., 1978; Solli et ai., 1980; Philp et ai., 1982) and crude
broad categories represented by a relatively paraffin- petroleums (Solli et al., 1980). Additionally, no allow-
rich vitrinite type (type 3pp paraffinic/phenolic) and ance is made for the fact that substantial n-alkadiene
more aromatic vitrinitic kerogens from coal measures yields are evident on pyrolysis of type 1p kerogens
(type 3ap-aromatic/phenolic). Type 3pp kerogens are (Larter and Douglas, 1978). Nevertheless, the method
typified by non-marine organic matter deposited in a does allow kerogens to be directly compared on a more
marine setting and correspond, in general terms, to systematic basis than with elemental analysis or
vitrinite 2 of Gutjahr (1983), while type 3ap kerogens Rock-Eval alone.
correspond to classic coal vitrains (vitrinite 10fGutjahr Also shown in Fig. 10 is the schematic evolution
(1983». path for initially hydrogen-rich kerogens from the
Typical examples of type 3 kerogens are vitrinite- grabenward North Sea Kimmeridge Clay Formation
dominated kerogens from marine or deltaic shales, (from immaturity through to the base of the conven-
e.g. Miocene of Indonesia, and many of the carbonace- tional liquid window, 0.5-1.3%R u). Movement along
ous Middle Jurassic shales of the North Sea Basin the evolution track on the paraffin index-hydrogen
Assessment Using Quantitative Kerogen Typing 279

100 assessment of the actual quantity of paraffins (and


ultimately, using similar approaches, other kerogen
ii:
,(
pyrolysate products) potentially available in a source
w
C system. Clearly the method needs to be extended to
~
z the other hydrocarbon species in kerogen pyrolysates,
ii:
u. but the abundance of normal alkanes in many crude
..:
a:
..: petroleums suggests that even in the form described
"-
I here the method is applicable in routine source-rock
...cw analysis.
;: 10 Modern exploration geochemistry has advanced to
z
c., the point where quantitative assessment of the local
a:
..: oil potential in a generative basin is now a feasible
()
0
a:
objective (Goff, 1983; Demaison and Murris, 1984).
C
> Thus the recognition of 'actual source rocks' (see
.....:
1:
Fig. 1) is becoming a reality .
:;;
a:
0
z
g
()

I
()
0> RECOGNITION OF ACTUAL SOURCE
C
W
ROCKS
N
:::;
..:
:;;
a:
0
Conventional source rock analytical methods have
Z
()
been very successful in recognizing areas of high pet-
....0 roleum potential. However, they are mainly aimed at
determining the current richness of sediments, while
ultimately it is the realized potential of the sediment
~~ ~ ~ ~ ~ 1~
that is of prime interest to the explorationist. In this
CALCULATED HYDROGEN INDEX context an actual source rock is defined as 'a sediment
KEROGEN TYPE - TOC NORMALIZED PARAFFIN YIELD
VERSUS HYDROGEN INDEX that has produced and released mobile hydrocarbons
and related materials' (see Fig. 1).
PARAFFIN INDEX-FROM PYROlYSIS-GC.1 UNIT; 0.25'1. BY WT.
Many previous studies (Brooks and Thusu, 1981;
HYDROGEN INDEX-FROM ROCKEVAl.UNITS; mg' HC/gm TOC. Barnard et al., 1981; Goff, 1983; Cornford et al., 1983;
Cooper and Barnard, 1984) have demonstrated that
REPRESENTATIVE KEROGEN TYPES
the Kimmeridge Clay Formation is the principal
source unit in the Central and Viking Grabens of the
Norwegian North Sea. Using data from these pro-
TYPE Ip - TORBANITE (BOTRYOCOCCUS) KEROGENS
vinces the conceptual problem of assessing the extent
RICH IN ALGINITE. of 'actual source rocks' is addressed schemat-
TYPE Ipn - TASMANITE (ALGINITE) RICH KEROGENS. ically within the context of available kerogen typing
TYPE In - RESINITE DERIVED MATERIALS. procedures.
TYPE II pn - AMORPHINITE RICH KEROGENS FROM In the simplest case the amount of organic material
SOURCE ROCKS CONTAINING MARINE
ORGANIC MAnER. expelled from a source sediment (the expelled organic
TYPE II/III P - EXINITE RICH KEROGENS matter-'XOM') is simply represented by the differ-
TYPE III pp - VITRINITE RICH KEROGENS FROM ence in genetic potential of the sediment at the pres-
MARINE/DELTAIC SHALES.
TYPE III ap - VITRINITE RICH KEROGENS FROM
ent time from its initial genetic potential. The genetic
COAL SWAMPS. potential of a sediment is defined as the sum of free
TYPE IV - INERTINITE RICH KEROGE~S. and potential hydrocarbons (and associated non-
hydrocarbons) in the sediment at any condition of bur-
Fig. 10. Paraffin index-hydrogen index cross-plot. ial. Thus:
XOM = prematurity organic content - present
organic content,
index cross-plot allows estimates of the total yield of
hydrocarbons (a hydrogen index) and the change in or, using Rock-Eval nomenclature,
paraffin content with maturity for the generated mat-
erials (a paraffin index). XOM = (S,i + S2i) depth i - (SIX + Szr;) depth X,
In defining source-rock kerogen type in terms of where SIi is the corrected initial free 'hydrocarbon
both bulk chemical composition (hydrogen/carbon content', S 2i is the initial S 2 yield of the sediment,
ratio, hydrogen index), detailed chemical composition S IX is the corrected free 'hydrocarbon content' at
(paraffin index, i.e. summed normal hydrocarbons in depth X and S ~ is the current S 2 yield of the sediment.
kerogen pyrolysates) and the source of the organic Thus, the basic concept is extremely simple; the
matter (petrographic type) the requisite information is expelled organic fraction is the difference between the
available to define the organic facies of the system. present and original organic fractions. In the normal
Coupling this information with maturity data enables oil window the full genetic potential is never realized,
an estimation of the petroleum potential. and thus the genetic potential of a sediment is not a
The quantitative approach to defining the composi- reliable indication of the actual yield. In addition two
tion of a kerogen described here is believed to be practical problems arise. Firstly, the original organic
directly relevant to defining better the petroleum content is unknown and, secondly, several analytical
potential of source kerogens. The model involves an limitations affect chemical data obtained from conven-
280 Petroleum Geochemistry in Exploration of the N01Wegian Shelf

S2/TOC VS TOC
Ii , A. KIMMERIDGE CLAY FM. - HOT SHALE 5.0t-----..----,--,---"'--,---\
Iii i I, i, 2350 M - HYDRDGEN INDEX 610 4.5
I,' - PARAFFIN INDEX 7.18 (KERDGEN BASIS)
III,I ,11·1
i I·
I Ii
'",I. 'i. ,I', li"II'11
1
11~1~,.,I~~\'N;.J.~I~'Ji.I,,,:.wu.J...Li.lL'-',~ '" ..'
S2/TOC 3.0
HYDROGEN
INOEX 2.5

m~:~~~oi :~:;rE 2.0


Iji) I "
DATA PRODUCED FROM
SYNTHETIC ROCKS CONTAINING
B KIMMERIDGE CLAY FM "HDT SHALE 1.5 o TYPE II KEROGEN SIMILAR TO
,i' , 2400 M - HYDRDGEN INDEX 640 THAT IN THE KIMMERIDGE CLAY FM

I ' ," I - PARAFFIN INDEX 7 25 1.0 0° (KEROGEN TYPE IIpn) MIXED


WITH PURE MINERAL MATRICES
1I'Ij I' II' il I I ' 0.5
IIII1 'II ,111'1 'I I I I , ,
0.0 0
~'i'~I)/II~;~\'\IU.I,~"1"';'"",,)"\..J_L.LU",,-u~LLc. _. -'
TOTAL ORGANIC CARBON % "ROCK"

Fig. 12. Matrix effects-normalized pyrolysate yield

I~III'
(hydrogen index).
C HEATHER FM
\1 ,
1

I,'I,I' 'I; ,
2690 M - HYDRDGEN INDEX 640
type 2pn amorphous-rich kerogen similar to that in the
1

I
" " I'"11111 1 'IJ ' , Kimmeridge Clay Formation. Of importance to the
111~11'1IJlll".'.,II~~~\Jl~ .. ,..u..,.:1....l....IL.j~--.J ~" .... present study is the observation that for very
organic-rich shales, such as the Kimmeridge Clay
Formation, the measured hydrogen index for the sed-
iment approaches within 10% of the value for an iso-
lated kerogen. Other matrix effects influencing yields
in source-rock pyrolysates are reviewed and summar-
ized in Fig. 13, and more fully by Larter (1984). Thus,
in many cases for the Kimmeridge Clay Formation,
hydrogen index matrix phenomena can be ignored and
the 8 2 Rock-Eval parameter used as a measure of
Fig. 11. Pyrograms of Upper Jurassic kerogens from the potential hydrocarbons in the kerogen. For example,
Viking Graben, North Sea. in the Viking Graben the Kimmeridge Clay Formation
has a mean TOe of 4.6%, with local values as high as
10.8%.
tional rotary cuttings. These problems, and possible Rock-Eval hydrogen index, in the absence of mat-
solutions to them, are discussed below. rix effects, is a powerful typing tool, as many samples
Two types of data relevant to this type of study are
available in most oil company files. The first includes
500
detailed kerogen analyses and chromatographic data
from the solvent-soluble organic fractions of sedi-
'" KEROGEN· NO MATRIX
ments. The second group includes less detailed, but ©- ~
more numerous, quantitative assessments of the vol- CARBONATE . O~NIC INTERACTIONS - INCREASED C02 YIELD.

atile organic and kerogen contents (e.g. Rock-Eval


data), and is generally more abundantly available.
Routine, repeated, cross-calibration of low- CLAY ABSORBTIVE AND CATAlYTIC
CRACKING/COKING MATRIX EFFECTS - DECREASED S2 YIELD.
resolution kerogen typing methods using higher-
\

o
resolution techniques (e.g. pyrolysis-GC) is manda-
tory as kerogen facies variations within lithostrati-
graphic units are the rule rather than the exception.
Figure 11 shows pyrograms of kerogens from
Humber Group sediments from a well in the Viking KEROGEN/PVROLY5ATE
OXIDATION MATRIX EFFECTS
Graben, North Sea. While the Kimmeridge Clay For-
mation is of relatively uniform composition, Heather
formation sediments are clearly quite variable. This .. OXIDATION OF KEROGEN AND
type of information is critical to the accurate interpre- 52 TO C02

tation of 'average data' for any stratigraphic unit in a


basinal context. Mapping of whole formation geochem- INORGANIC MATRIX

ical data in cases where extreme organic facies varia- a ILLITE 100%
100 • 90% ILLITE + 10% PYRITE
tions occur is potentially misleading. Once a calibra- 6 90," IlliTE, 10'" HAEMAIITE
tion has been made, the quantitative use of screening .. 90\' IlliTE + 10% LIMONITE
.. CALCITE 100%
data becomes useful in certain cases.
Several problems relate to the use of Rock-Eval
data in a quantitative manner, the principal one being
the 'matrix effect' as discussed by Espitalie et al. 50 100 150 200 250 300

(1980); Horsfield and Douglas (1980) and Larter S3/Toe OXYGEN INDEX mg. e02i gm TOe

(1984). Modified offer lorter. 1984.

Figure 12 shows the effect of mineral matrix and Fig. 13. Representative matrix effects, organically lean
organic content on the measured yield of hydrocar- sediments. Rock-Eval data for a synthetic rock series
bons during pyrolysis of synthetic rocks mixed with a containing 1.2-1.3% organic carbon ofa type IIpm kerogen.
Assessment Using Quantitative Kerogen Typing 281

A/RAW DATA B/WEIGHTED AVERAGE DATA - BY WELL


1500

1750

2000
I/)

I/)
ffi 2250
a: I-
LU 2500
LU
I- :2
LU
:2 I 2750

I t::
2 3000

....
LU ;:)
u.. 3250
0.
:2 o
<t :t 3500
I/) I-
0. 3750
u..
o LU
o 4000
:t 2
b:LU <t
LU l!250
o :2 4500

11750

100 200 300 400 500 600 100 100 200 300 400 500 600 700

HYDROGEN INDEX (-MGS HC/gm TOC) WEIGHTED MEAN HYDROGEN INDEX OF FM


(-MGS HC/gm TOC)

Fig. 14. Hydrogen index data, Kimmeridge Clay Formation, Viking Graben.

can be analysed on a regional basis and a significant


data set obtained cheaply. Figure 14(A) shows hyd-
rogen index data from the Kimmeridge Clay Forma-
tion in several wells in the Viking Graben which cover
Tf---COLDSHALE'~
a range of maturities down to the base of the conven- V'J HOTSHALE
tional oil window. A range of kerogen bulk composi-

l
a:
w
tions is evident, although the picture is substantially 0-
w
:;;
clarified if the data are reduced to weighted mean val- o
o
ues. This procedure places greater weight on the rela- COLDSHALE 2
DIAGRAM SHOWING
VERTICAL VARIATION IN
tively more uniform richer samples in the unit, as fol- KEROGEN TYPE FOR A
HYDROGEN RICH
lows: - - - - - - - - -_ _ _ _ ORGANIC FACIES
KIMMERIDGE CLAY

itl
SECTION

S2i)/n '---------'----~
10 90 150 210 270 330 390 450

(tl TOCi)/n
weighted mean hydrogen index = HYDROGEN INDEX

Fig. 15. Hydrogen index data for a well near peak


hydrocarbon generation (0.8% Ro, 3500 m depth), Kim-
The weighted mean hydrogen index for a formation, in meridge Clay Formation, Viking Graben.
cases where the cuttings are well-picked and sampled
at closely spaced intervals (10 m or less), best repres-
ents the hypothetical 'typical average formation sam- occurring during natural catagenesis (Tissot and Van-
ple'. Figure 14(B) shows data from several Viking denbrouke, 1982). However, differences in kerogen
Graben wells and illustrates the clarification obtained. behavior under Rock-Eval pyrolysis between samples
Such plots indicate that the organic facies for the buried to different depths but of constant organic
Kimmeridge Clay Formation varies, with a humic facies, still reflect the net catagenetic effect experi-
platform facies locally in the Horda Platform/Utsira enced by the more-mature kerogen. Thus, for
High areas and with an initially hydrogen-rich facies kerogens of constant organic facies, the difference in
in the Viking Graben and on its flanks. Further sub- the hydrogen index is proportional to the net amount
division of the Kimmeridge Clay Formation is, of of soluble species produced over the maturation path-
course, possible; Figs. 15 and 16 also show vertical way. This is so even though the analytical method
facies variations in the kerogen quality in the Kim- itself is in no way simulating natural processes. This
meridge Clay Formation. These trends often, but not line of argument, of course, assumes that catagenesis
always, correlate with radioactivity (y-ray) logs. involves extensively cleavage reactions rather than
Studies of the basin-wide variations in the properties condensation processes. While both occur, the evi-
of the 'hot shale' and 'cold shale' units allow more accu- dence is clear that normal catagenesis is dominated by
rate assessment of the real source potential of the unit fission rather than condensation processes (Tissot and
in any area (M. Bjor~y, pers. comm.). Figure 16 shows Welte, 1978).
depth-related trends in the 'hot shale' and lower 'cold Many approaches for estimating the oil volume
shale' units of the Kimmeridge Clay Formation. expelled from a source bed assume a constant kerogen
During high temperature pyrolysis of kerogens, the type, and also generally an initial (pre-burial) value for
operative processes are quite different from those the organic carbon content for the prospective source
282 Petroleum Geochemistry in Exploration of the Norwegian Shelf

1500
100
1150
90 o

2000 80
\ "" •• v.:~
\"'.... ' · v"
70
rJl
2250
~ f: •••
II: %
~~. , 4 C1tS <
60
W 2500 '\
~
W
2750
AMEX 50
FACIES 1 ! "
::r! 40
.~
I 3000
'x
3D I 0 0 0 0 0 000

t:
Z 3250 20 AMEX-ESTIMATED ~ - :0':;
:>
II. 3500 , 10 AREAL % AMORPHOUS KEROGE~ (AMORPHIN~
+AREAL % EXINITE KEROGEN
0 x' O~--~--~~~~~~---'----r0---r
J: 3750 1250 1750 2250 2750 3250 3750 4250 4750
~
11-
W
llGOO
HOT SHALE SECTION GENERALLY HAS A HIGHER DEPTH 1M)
C HYDROGEN INDEX IN ANY WELL THEN CORRESPONDING
4250
Z COLD SHALE SECTION. 0.0· 69.7 () 348.5·487.9
c( o 69.7·209.1 (OJ 487.9·627.3
W 4500
::r! • 209.1 ·348.5 () 627.3 - 697.0
4750

5000
Fig. 17. Hydrogen index as a function of 'AMEX' and depth,
2 2 3 3 4 ') 5 6 6 ? Kimmeridge Clay Formation, Viking Graben.
5 usc 5 0 o S 0 S 0
0 0 o 0 0 0 [] o J 0 0 0

WEIGHTED MEAN HYDROGEN INDEX OF FM index, maturity (represented as depth) and the
(-MGS HC/gm TOC)
organic facies for the graben ward facies of the Kim-
UPPER COLD SHALE + LOWER COLD SHALE x HOT SHALE * meridge Clay may be obtained. With such an
Fig. 16. Hydrogen index as a function of stratigraphy, approach, a relationship of the form
Kimmeridge Clay Formation, Viking Graben. hydrogen index =JCpetrographic type)(depth)
can be obtained. Figure 18 shows the relationship
rock. In the case of the Kimmeridge Clay Formation between hydrogen indices calculated from petrog-
in the Viking Graben neither assumption is valid. raphic data and those measured from Kimmeridge
The difference in hydrogen index between two Clay Formation samples in six wells.
kerogens of identical organic facies but different burial Unfortunately, this method is only found to be reli-
histories can be used to estimate the carbon conver- able for small, areally restricted data sets. On a reg-
sion during catagenesis and thus the ancient carbon ional scale an empirical approach was used to derive a
content (Orr, 1983). The definition of kerogen in this maturation curve for any depth in the range
context refers to all non-volatile organic matter in a 2500-4500 m. This range was chosen since in the
sediment, and thus includes kerogen + non- North Sea it includes the most data from the Upper
volatile-non-distillable material (ideally all non- Jurassic, and thus the evolution curves are most reli-
petroleum components) some asphaltenes and other able over this interval. The 2500 m depth is also the
non-migratable, non-hydrocarbons. This redefinition point at which significant hydrocarbon generation
is not an arbitrary one, as oils are largely hydrocarbon commences, corresponding to a true (vitrinite 1) mat-
in nature, and distillable in most Viking Graben cases. uration level of around 0.5%Ro. At shallower levels,
Thermal modelling experiments indicate that the only major changes in kerogen composition involve
weakly migratable NSO and asphaltene fractions are heteroatom loss rather than hydrocarbon production,
considered to be an equivalently potent source of hyd- thus 2500 m is taken as the zero datum for calculation
rocarbons as the conventionally defined kerogen in the of generated hydrocarbon quantities. An initial hyd-
Kimmeridge Clay Formation. rogen index can be defined for the kerogen in each
The variation in kerogen chemistry within a given sample at 2500 m.
formation is a function of both maturity and organic HII represents the hydrogen index for each sample
facies. A plot of data from many wells indicates a con- computed at a datum of 2500 m using predictive rela-
sistent relationship between vitrinite reflectance and tionships such as those discussed above.
depth of burial for the Kimmeridge Clay Formation of
the Viking Graben (see also Goff, 1983). Similar con- 700 r-----,-----,----.,..----,-----,-----,..-----,
clusions were reached by Cornford et al. (1983) who
indicated that locally, depth of burial (or more cor- 600

rectly, formation temperature) is a practical maturity


indicator for the Kimmeridge Clay Formation in the 500

UK sector. c
5"' 400
The variation in source organic facies in a
65:
somewhat variable sediment such as the Kimmeridge ~ 300

Clay and over a range of maturity levels is best incIi-


(J

cated by careful petrographic kerogen typing. This is 200

because petrographic typing in transmitted light is


less affected at low maturity levels «0.8% Ro) by
mass changes due to conversion of kerogen to liquid
o
and gaseous hydrocarbons. The data in Fig. 17, how- 100 200 JOO
HI
400 500 600 700

MEASURED
ever, indicate a trend of reducing 'AMEX' (areal %
amorphous kerogen + exinite kerogen) with depth at Fig. 18. Predicted versus measured hydrogen indices,
higher maturity levels. Using a computerized regres- based on petrographic data from six wells (Kimmeridge
sion procedure, the relationship between hydrogen Clay Formation, Viking Graben).
Assessment Using Quantitative Kemgen Typing 283

For the kerogen-only system, the fractional loss of organic material retained in the source bed. This is
carbon through production of hydrocarbons from bur- critical since, if this relationship is evaluated wrongly,
ial beyond 2500 m may be conservatively estimated by the value of the final migration efficiency will be in
the method of Orr (1983), as follows: error. Using quantitative oil models, flash pyrolysis
and hydrous pyrolysis data, the relationship between
conservative carbon loss (f) = (hydrogen
observed S] and actual hydrocarbon contents can be
index 1 - hydrogen index2) x 0.85/1000, estimated as a function of maturity.
where the hydrogen indices are for kerogens of the Once the relationship between measured hydrocar-
same facies at different maturities and the Rock-Eval bon content (S] or C]5+ fraction) and the probable
evolved hydrocarbons consist of approximately 85% in-sediment value is obtained, the expelled organic
carbon. The 1000 represents a conversion factor from matter in the sediment system approximates to
hydrogen index to absolute fractional units.
Analytical data provide current kerogen-based TOC XOM =
. (ANCAR x
( S]1 + 100
HII)) - (S2 + S1F x S]),
contents and corrected hydrogen indices. Using mod-
elled or observed kerogen composition data, an esti- where S1F is a factor relating observed S1 to actual
mate of HII (hydrogen index at 2500 m) can be sediment C1-C" hydrocarbon content. S]F is a func-
obtained, and thus the prematuration kerogen-based tion of maturity, and reaches high values at the base of
carbon content of the sediment can be approximately the oil window. In a condensate-producing section, gas
estimated. (C 1-CJ may represent over 70% of the total migrating
Returning to the original mass balance equation, material, and S1F would be correspondingly high.
the expelled fraction of the sediment is estimated as General procedures of the type outlined above have
follows: substantial errors associated with their assumptions
but they nevertheless provide the explorationist with
XOM = (S] + S2 at 2500 m) - (S] + S2 at current useful estimates of mobile oil and gas in a basin. In
depth). general, the approach is more reliable when dealing
Substituting, with a source section of maturity up to around 0.8% R 0
('black oil' generation and expulsion in the Viking
XOM = (S] at 2500 m + ANCAR x HII/100) Graben). At greater maturities, reliable estimates of
- (S2 + S] at current depth). the retained gas content of the sediments become
more difficult to make.
The hydrocarbon content at 2500 m can be estimated Such simple models coupled with regionally exten-
from plots of SJTOC for the immature Kimmeridge
sive geochemical data bases, source-rock structure
Clay Formation. This is a valid approach, as few low
and isopach maps, can be used to assess readily charge
molecular weight species have been generated at availability to a given trap. Comparison of observed
2500 m and thus the S] value is a true estimate of the
reservoir fills and calculated potentially migrated oil
distillable soluble fraction of the sediment at this low
volumes permits the assessment of prospects in a sys-
rank level. tematic manner.
In contrast, once hydrocarbons are being generated
Such models further allow the geologist, for exam-
from kerogen and soluble non-hydrocarbon fractions,
ple, to evaluate whether the rotated fault block traps
substantial amounts of gasoline range and gaseous
in the Viking Graben were filled from source beds
molecules are produced. However, during return of
largely within the block, the oil migrating up the dip
cuttings to the surface and subsequent sample wash-
slopes of the blocks and then laterally into the Middle
ing, handling and storage, substantial amounts of
Jurassic reservoirs, as proposed by Goff (1983) for the
these components are lost (Espitalie, 1984). This
East Shetland Basin. Alternatively, sourcing may
results in a very low estimate of the total retained
occur from Jurassic source sections in deeper down-
hydrocarbon content for the sediment, especially at
faulted blocks, with migration along bounding fault
depth, and thus in an overestimate of the expelled
planes. Such questions are relevant to the prospec-
fraction. This is a critical observation. Additionally, tiveness of untested fault block traps adjacent to the
the mode of operation of the Rock-Eval instrument
and the presence in a sample of chromatographically deeper Viking Graben/M0re Basin areas of the Nor-
wegian North Sea.
active inorganic species such as clays, results in very
high-boiling components being selectively retained by
Oil migration effects-restrictions on the use of
the sediment in the pyrolysis oven (Tarafa et al., 1984;
geochemical data
Horsfield, 1984). Solli et al. (1984) and personal obser-
vations suggest that a typical S] fraction from a stored Using a mass balance model conceptually similar to
cuttings sample represents material in the boiling that described above, it is estimated that up to 25% of
range of about CjQ-C 25 normal-alkanes, corresponding the organic carbon may have been expelled as hydro-
approximately to a kerosene through gasoil distillate carbons by peak oil generation' (0.8% Ro) in the
cut for an oil. Kimmeridge Clay Formation of the Viking Graben.
The hydrocarbon-rich nature of Viking Graben oils, High calculated expulsion efficiencies, high resis-
and also of the S I fraction of sediments, suggests that tivities on E-Iogs, overpressuring, and the heterogen-
the 'petrogen' (kerogen + other non-migratable mat- ous distribution of organic matter in Kimmeridge
erial) concept is applicable to quantitatively evaluat- shales all strongly suggest that continuous oil phase
ing the retained organic matter (S2 precursor) in a migration is taking place throughout much of the
sediment. generative zone beyond peak oil generation (Goff,
The accurate estimation of the concentration of free 1983).
gaseous and gasoline range components in a source
rock is critical in assessing the relative contribution of
the analyzable S] (or EOM) fraction to the total *Perhaps more aptly termed peak oil retention!
284 Petrolenrn Geochemistry in E.1::ploration of the NOiwegian Shelf

At shallow depths, the organic-rich sediment, con- CONCLUSIONS


taining up to 15% or more by volume of organic mat-
ter, supports part of the sedimentary load on its lami- Petrographic, bulk chemical and detailed molecular
nated organic layers. As burial continues, part of the kerogen typing can provide the explorationist with
kerogen is converted to fluids, chiefly oil below 0.8% much rele.vant info~ation on the nature and type of
R 0, resulting in overpressuring and lateral migration the orgamc matter m source rocks. Some preliminary
along kerogen sediment inferfaces (cf. Meissner, 1978; quantitative estimate of actual oil yields from rich
Du Rouchet, 1978; Durand, 1983). source beds is obtainable using these data. Con-
With deeper burial (>3500 m) significant gas gen- straints on the final prospectiveness of a source unit
eration occurs along with liquid hydrocarbons, the cannot be established using kerogen type and matur-
increases in volume during this process (Ungerer et ity data alone, as poorly understood migration
al., 1983) possibly assisting in maintaining fluid over- phenomena add significant uncertainty to the effective
pressure and migration processes. Lateral oil and gas interpretation of source rock data.
migration will occur until either the pressure field is
exhausted or, more probably, until relatively perme-
able conduits such as faults are encountered. Once in
the fracture system oil migration will follow local and ACKNOWLEDGEMENTS
regional fluid potential fields (Ungerer et al., 1984).
As the sediments become more mature, the I am grateful to Union Norge A/S and Union Oil
kerogens will inevitably become more hydrogen- Science & Technology Division management for en-
deficient and less plastic (Van Krevelen, 1961). couragement and permission to publish these ideas.
Further, as the organic contents decrease there will Many scientists have been involved in discussions of
be reduced opportunity to sequence oil directly from many of the concepts described here, and I acknow-
the sediments, as grain-grain overburden load sup- ledge the contributions made by Drs R. E. Sweeney,
port becomes more important. J. T. Senftle, B. W. Bromley, G. H. Smith, J. A.
These processes, and others such as phase change Curiale, J. C. Ellice-Flint and N. A. Briedis. Malvin
Bjor~y (IKU), on a sabbatical research program at
and aqueous solubility behaviour relative to tempera-
Brea, contributed greatly to ideas concerning Viking
ture and pressure (Goff, 1983; Ungerer et al., 1984)
Graben geochemistry. Manuscript review was aided
may lead to changes in the effective migration
by comments from John Fox and John Ellice-Flint.
mechanism and variations in the composition of the
expelled fluids at depth. The nature of the expelled
hydrocarbons cannot be predicted on the basis of
maturity or kerogen type alone. Gas only may be APPENDIX 1: PYROLYSIS METHODS
expelled from more deeply buried 'oil-source' sedi-
ments even within the 'liquid windows'. Further, at Kerogens were isolated and purified using standard
the onset of oil expulsion, any compositional fractiona- methods (Durand, 1980). Pyrolysis analysis was car-
tion between organic matter expelled (XOM) and ried out using a CDS Pyroprobe system modified for
retained in the source bed (EOM) will be at a max- use with fused silica high-resolution capillary-GC
imum (Kontorovich, 1984). Early generated oils will columns (Larter, 1984). Unless indicated otherwise
be the hardest to correlate to source beds using cur- on the figures, pyrolysis conditions were as follows:
rent approaches. pyrolysis of up to 500 f1g of extracted kerogens at
Durand (1983) concisely describes the limits which 800 DC for 20 s in flowing helium, followed by gas
migration mechanisms may place on the composition chromatography on an SP2100-coated, 50 m,
of fluids expelled from source beds. Due to the princi- Durabond fused silica capillary column programmed
ple of relative permeability, the effects of water and from 40°C to 300°C at 4°C/min. The quantitative
oil saturations in a shale pore fluid on oil expUlsion pyrolysis-GC experiments were carried out on
efficiency are dramatic. Durand suggests, for exam- samples pre-mixed with weighed amounts of
ple, that in the case of hydrocarbon generation in poly(alpha)methyl styrene. The polymer was added
rapidly buried organically-lean, water-wet sedi- quantitatively in methylene chloride solution to the
ments, the expelled fluid may be dominantly gas, weighed extracted dried kerogen and thoroughly
despite ample higher hydrocarbon generation in the mixed with the material. The solvent was then evapo-
source bed. The low relative permeabilities of the rated before pyrolysis. Programmed temperature
source section to oil (due to high water saturations in pyrolysis was performed on a CDS 820GS geological
the source sediment) favour preferential gas migra- sample analysis system.
tion in aqueous solution.
Durand (1983) clearly demonstrates that, despite
the availability of accurate and sophisticated methods REFERENCES
of characterizing the type and maturity of sediments,
migration phenomena, understood so poorly at pres- Bailey, J. J. L. 1981. Hydrocarbon potential of organic matter. In:
ent, may be the prime factor in deciding the final com- Brooks, J. (ed.), Organic Malumlioll Stlldie" and Fo""il Fnel
[i),l'p/oratioll. Academic Press, New York, p. 283.
position of expelled and reservoired fluids. Thus, Barnard, P. C. and Cooper, B. S. 1981. Oils and source rocks ofthe
despite the ability to effectively type kerogens, and Korth Sea area. In: Illing L. V. and Hobson, G. D. (eds.), Pet-
with some understanding of hydrocarbon generative mlculIl Geology of the COlltif/elltal She(t" of Norlh West Europe.
mechanisms under some conditions, the composition of Institute of Petroleum, London, pp. 169-17.1.
the fluid expelled from a source bed often cannot be Barnard, P. C., Collins, A. G. and Cooper, B. S. 1981. Identification
and distribution of kerogen facies in a source rock hori-
predicted accurately using these data alone. zon-examples from the North Sea Basin. In: Brooks, J. (e(L),
Orgallic Matumtioll Studies alld Fossil Fuel E.rp/omtioll.
Academic Press, London, pp. 271-282.
Brooks, J. and Thusu, B. L. 1981. Oil-source identification and
Assessment Using Quantitative Kerogen Typing 285

characterization of the Jurassic sediments in the Northern North Larter, S. R. and Douglas, A. G. 1982. Pyrolysis methods in organic
Sea. Chelil. Geol., 20,283-294. geochemistry-an overview. J. Analyt. Appl. Pyrolysis, 4,
Cooper, B. S. and Barnard, P. C. 1984. Source rocks and oils of the 1-19.
Central and Northern North Sea. In: Demaison, G. and Murris, R. Larter, S. R. and Senftle, J. T. 1984. Practical high resolution
J. (eds.), Petroleum Geochemistry and Basin Evaluation. AAPG kerogen typing for petroleum source rock analysis. AAPG Bull.
Memoir, 35,303-314. (submitted).
Cornford, C., Morrow, J. A., Turrington, A., Miles, J. A. and Larter, S. R., Horsfield, B. and Douglas, A. G. 1977. Pyrolysis as a
Brooks, J. 1983. Some geological controls on oil composition in the possible means of determining the petroleum generating-
UK North Sea. In: Brooks, .J. (ed.), Petroleum Geochemisb'Y and potential of sedimentary organic matter. In: Jones, C. E. R. and
Exploration of Europe. Geol. Soc. Spec. Pap. No. N., Cramers, C. A. (eds.), Analytical Pyrolysis. Elsevier. Am-
pp. 255-265. (Blackwell, London.) sterdam, pp. 189-202.
Demaison, G. 1984. The generative basin concept. In: Demaison, G. Larter, S. R., Solli, H. and Douglas, A. G. 1978. Analysis of
and Murris, R. J. (eds.), Petroleum Geochemistry and Basin kerogens by pyrolysis-gas chromatography-mass spectrometry
Evaluation. AAPG Memoir 35, 1-15. using selective ion detection. J. Chromatogr., 167, 421-431.
Demaison, G. and Murris, R. J. (eds.) 1984. Petroleum geochemis- Lewan, M. D. 1983. Effects of thermal maturation on stable organic
try and basin evaluation. AAPG lvlemoiT 35, 426 pp. carbon isotopes as determined by hydrous pyrolysis of Woodford
Dembicki, H., Horsfield, B. and Ho, T. T. Y. 1983. Source rock shale. Geochim. Cosmochim. Acta. 47, 1471-14S,Q.
evaluation by pyrolysis-gas chromatography. AAPG Bull., 67(7) Leythaeuser, D., Mackenzie, A. S., Schaefer, R. G., Altebaumer, F.
1094-1103. J. and Bjorpy, M. 1983. Recognition of migration and its effects
Durand, B. (ed.) 1980. Kerogen-insoluble Organic :lIatter From \\~thin two core-holes in shale/sandstone sequences from Sval-
Sedimentary Rocks. Editions Technip, Paris, 519 pp. bard, Norway. In: Bjorpy, :vI. et al. (eds.), Advnnces in Organic
Durand, B. 1983. Present trends in organic geochemistry in Geochemistry 1981. Wiley, London, pp. 136-146.
research on migration of hydrocarbons. In: Bjorpy. :VI. el al. Mackenzie, A. S., Leythaeuser, D., Schaefer, R. G. and Bjor0Y,:vI.
(eds.), Advances ill Orgallic GeochelllistJ~y 1981. Wiley, London, 1983. ExpUlsion of petroleum hydrocarbons from shale wmce
pp. 117-128. rocks. Nature, 301, 506-509.
Durand, B. and Paratte, :VI. 1983. Oil potential of coals, a geochemi- Martin, S. J. 1977. Thermally evolved hydrocarbons from the bitu-
cal approach. In: Brooks, J. (ed.), PetroleulII Geochellristry alld men and kerogen constituents of whole rock. In: Campos, R. and
E.cploration of Europe. Geol. Soc. Spec. Pap. No. 11, Goni, J. (eds.), Advances in Orga.nic Geochemistry 1975.
pp. 255-265. Blackwell, London. Enadimsa, Madrid, pp. 677-692.
Du Rouchet, J. 1978. Elements d'une theorie geomecanique de la Maters, W. L., Van de :vIeent, D., Schuyl, P. J. W., De Leeuw, .J.
migration de l'huile en phase constituee. B1l11. Cent. Rech. W. and Schenk. P. A. 1977. Curie point pyrolysis in organic
Ex·ploT.-Pmd. Elf-Aquitaine, 2, 337-373. geochemistry. In: Jones, C. E. R. and Cramers, C. A. (eds.),
Espitalie, J. 1984. Geochemical logging. In: Voorhees, K. (ed.), Analylical Pyrolysis. Elsevier, Amsterdam. pp. 20:3-214.
Analytical pryolysis-Methods and Applications. Butterworth, McHugh, D. J., Saxby. J. D. and Tardiff, J. W. 1978. Pyrolysis-gas
London, p. 275. chromatography of carbonaceous material from Australian sedi-
Espitalie, J., Laporte, J. C., Madec, M., :Vlarquis, F., Le Plat, P., ments. Cheill. Geal., 21, 1.
Paulet, J. and Boutefeu, A. 1977. Methode rapide de character- McIver, R. D. 1967. Composition of kerogen. Proc. Sellenlh. Wurld
isation des roches meres, de leur potential petrolier et de leur Petroleulli COllgr., 2, 25-36.
degre d'evolution. ReI!. lnst. Fmw;ais Petrole, 32, 23-42. Meissner, F. F. 1978. Petroleum geology of the Bakken Formation
Espitalie, J., Madec, M. and Tissot, B. 1980. Role of mineral matter Williston Basin, :-,Torth Dakota and Montana. :Honleuw Grol. Soc.
in kerogen pyrolysis. AAPG, Bull, 64, 59-66. Proc. 197!? Willistoll Basin "!flllp. pp. 207-227.
Evans, E. .J., Batts, B. D. and Smith, J. W. 1984. Determination of Miknis, F. P., Netzel. D. A., Ward-Smith, J., Alisament. M.
the hydrocarbon prospectivity of sediments by hydrogenation. and }laciel, G. E. 19~2. C-l:l NMl{ measurements of the
APEA Journal, 24(1), 222-229. genetic potential of oil shales. Geocili III. Coslilicililll. Acl",
Giraud, A. 1970. Application of pyrolysis and pyrolysis-gas 46, 977.
chromatography to the geochemical characterization of kerogen Orr, W. L. 1983. Comments on pyrolytic hydrocarbon yields in
in sedimentary rocks. AAPG Ball., 54,439-455. source rock evaluation. In: Bjor0Y, M. et al. (eds.), Advallces in
Goff, J. C. 1983. Hydrocarbon generation and migration from Orgallic Geocheillist;'!J Iii!!/, Wiley, London, pp. 77.S-788.
Jurassic source rocks in the East Shetland Basin and Viking Gra- Philp, R. P. and Russel, N. J. 1980. Pyrolysis-gas chromatogra-
ben of the North Sea. ./. Gml. Suc. Loue/., 140,445-474. phy-mass spectrometry of batch autoclave products derived from
Gormely. J. R. and Mukhopadahay, P. K. 19i13. Hydrocarbon coal macerals. In: Douglas, A. G. and Maxwell, .J. R. (eds.),
potential of kerogen types by pyrolysis-gas chromatography. In: Advances ill Organic Geochcmislt7! 1979. Pergamon Press,
Bjorpy M. et al. (eds.), Advances in Organic Geochemistry 1981. Oxford, p. 653.
Wiley, London, pp. 597-606. Philp, R. P., Gilbert, T. D. and Russell, N. J. 1982. Characterisa-
Gutjahr, C. C. M. 1983. Introduction to incident light microscopy of tion by pyrolysis-gas chromatography-mass spectrometry of the
oil and gas source rocks. Geol. Mijubollw, 62,417-425. insoluble organic residues derived from hydrogenation of Tasl/la-
Harwood, R. J. 1977. Oil and gas generation by laboratory pyrolysis IIdes sp. oil shale. Fuel, 61, 22l.
of kerogen. AAPG Bull, 61,2082-2102. Powell, T. G., Creaney, S. and Snowdon, L. R. 1982. Limitations of
Horsfield, B. 1984. Pyrolysis studies and petroleum exploration. In: the use of organic petrographic techniques for identification of
Advances in Petroleum Geucheltlisll~Y, Vol. 1. Academic Press, petroleum source rocks. AAPG Buli. 66(4), 430-435.
pp. 247-292. Robert, P. 1971. Etude petrographique des matieres organique:;
Horsfield, B. and Douglas, A. G. 1980. The influence of minerals on insolubles par la mesure de leur pouvoir reflecteur. Contribution
the pyrolysis of kerogens. Geochim. Cosmochim. Acta, 44, a i'exploration petroliere et II la connaissare des basins sedimen-
2087-2095. taires. Rev. Ilisi. Fmll,ais 1111 Pamle, 26(2), 106.
Huntjen, H. J. and Van Krevelen, K. 1957. Chemical structure and Romovacek, J. and Kubat, J. 1968. Charactel~zatioll of coal sub-
properties of coal-composition of the individual macerals (vitri- stances hy pyrolysis-gas chromatography. Alla/yt. Cftelll., 40,
nites, fusinites, micrinites, exinites). Fuel, 33, 79. 1119.
Kontorovich, A. E. 1984. Geochemical methods for the quantitative Senftle, J. T. 1983. Optical analysis of kerogen-its role in an inte-
evaluation of the petroleum potential of sedimentary basins. In: grated approach for kerogen typing-. Paper and Abstract Pre-
Demaison, G. and Murris, R. J. (eds.), Petroleum Geochemistry sented at the North Alllel'ieall Coal Pelroqmphical }leetillg,
and Basin Evaluation, ilAPG Memoir, 35, 79-109. Merrilville, Indiana.
Larter, S. R. 1984. Application of analytical pyrolysis techniques to Sigleo, A. C., Hoering, T. C. and Helz, G. R. 1982. Composition of
kerogen characlerization and fossil fuel exploration/exploitation. estuarine colloidal material: organic components. Geochilll. Cos-
In: Voorhees, K. (ed.), Analytical Pyrolysis-Melhods and lIIochilil. Acla. 46, Hil9-1626.
Applications. Butterworth, London, pp. 212-275. Solli, H., Larter, S. R. and Douglas, A. G. 1980. Analvsis of
Larter, S. R. and Douglas, A. G. 1978. Low molecular weight kerogens by pyrolysis-gas -chromatography-mass' spec-
aromatic hydrocarbons in coal maceral pyrolysates as indicators trometry-Part I, Alkyhenzenes. In: Douglas, A. G. and Max-
of diagenesis and organic matter type. In: Krumbein, W. E. (ed.), well, J. R. (eds.), Advallces ill Orgallic Geachelllisll·y1979. Per-
Environmental Biogeochemistry and Geoll1icrobiolo.QY, Vol. 1. gamon Press, Oxford, p. 59l.
Ann Arbor, Michigan, pp. 373-386. Solli, H., Bjor0Y, lVI. Leplat, P. and Hall, K. 1984. Analysis of
Larter, S. R. and Douglas, A. G. 1980. A pyrolysis-gas chromatog- organic matter in small ruck samples using combined thermal
raphic method for kerogen typing. In: Douglas, A. G. and Max- extraction and pyrolysis-gas chrumatography . .1. lillnlyt. Appl..
well, J. R. (eds.), Advances ill Ol~ganic Geochemistry 1979. Per- Pyrolysis (in press).
gamon Press, Oxford, 579-584. Staplin, F. L. 1969. Sedimentary organic matter. organic
286 Petroleum Geochemistry in El'ploratioll of the Norwegian Shelf

metamorphism, and oil and gas occurrence. Call. Petroleuill geochemical models in oil exploration, principles and practical
Geoloqists Bull., 17,47-66. examples. In: Demaison, G. ancl Murris, R. J. (eels.), Pell'OleulII
Stopes, M. C. 1935. On the petrology of banded bituminous coal. Geochelllist1'yand Basin Evaluation. AAPG .llellloir, 35, 5:3-77.
Fuel, 14,4-13. Van ele )!eent, D., Brown, S. C., Philp, R. P. and Simuneit, B.
Tarafa, M. E., Hunt, J. M. ane! Whelan, J. K. 1984. AAPG Bull., (in 1980. Pyrolysis high resolution gas chromatography and
press). pyrolysis-gas chromatography-mass spectrometry of kerogens
Tissot, B. P. l!:J84. Recent advances in petroleum geochemistry and kerogen processes. Geochilll. Cos IllUchi III. Acta, 44,
applied to hydrocarbon exploration. AAPG Bull., 68(5), 999-1014.
545-563. Van Krevelen, D. W. 1961. Coal. Elsevier, Amsterdam.
Tissot, B. P. and Vandenbrouke, ]\II. 1982. Geochemistry and Van Graas, A., de Leeuw, J. W. and Schenk, P. A. 1980. Analysis of
pyrolysis of oil shales. Preprillt" Aili. Cheill. Soc., Div., Pet- coals of different ranks by Curie-point pyrolysis-gas chromatog-
m/eulI/ Chelll., 28(1),92-99. raphy-mass spectrometry and Curie-point pyrolysis-mass spec-
Tissot, B. P. and Welte, D. H. 1978. Petruielwi FUl'/lw/'ioll alld trometry. In: Douglas, A. G. and Maxwell, J. R., (eds.),
OceluTellce, A New Approach to Oil alld Go" e,rploratioll. Advallces ill (h.llllllic Geuehelliistl'i/, 1.979. Pergamon Press,
Springer-Verlag, Berlin, 583 pp. Oxford, p. 485.
Tissot, R. P., Durand, B., Espitalie, J. and Combaz, A. 1974, Influ- Welte, D. H. and Yukler, :vI. A. 1981. Petroleum origin and accumu-
ence of the nature and diagenesis of organic matter in the forma- lation in basin evoluation-a quantitative model. AAI'G Bull., 65,
tion of petroleum. l1APG Rill!., 58(:i), 499-506. 1387-1396.
Ungerer, P., Behar, K and Discamps, D. 1983. Tentative calcula- Yarzab, R. F., Given, P. H., Spackman, W. and Davis, A. 1980.
tion of the volume expansion of organic matter during hydrocar- Depenclence of coalliquefaclion behavior on coal characteristics 4.
bon genesis from geochemistry data-implications for primary Cluster analysis for characteristics of 104 coals. Fu.el, 59, 81.
migration. In: Bjor0y, :vI. et 01. (eds.), AI/l'al/ces ill Orgallic Zilm, K. W., Pugmire, R. J., Larter, S. R., Allan, J. ancl Grant, D.
Geocilellli.,tI'Y 1981. Wiley, London, pp. 129-135. M. 1981. Carbon-13 cross polarised/magic angle spinning NMR
Ungerer, P., Bessis, F., Chenet, P. Y., Durand, E., Nogaret, K, spectroscopy of coal macerals. Fuel, 60, 717.
Chiarelli, A., Oudin, J. L. and Perrin, J. F. 1984. Geological and
24 Hydrous pyrolysis of
asphaltenes
N. Telnms*, G. C. Speers*, A. Steen* and A. G. Douglast
*Norsk Hydro Research Center, Box 4313, 5013 Nygaardstangen,
Bergen, Norway
tOrganic Geochemistry Unit, Drummond Bldg, University of Newcastle
upon Tyne, Newcastle upon Tyne, NE1 7RU, UK

Asphaltenes are defined as small, soluble parts of the kerogen, and may be used as kerogen models.
Hydrous pyrolysis in small stainless steel bombs at elevated temperatures (280-370 DC) releases hydrocarbons
that are very similar to the oil/bitumen from which the asphaltenes were precipitated. As asphaltenes are not
affected by biodegradation, this is a possible method for generating an oil unaffected by biodegradation.
Hydrous pyrolysis of asphaltenes can also facilitate the study of mineral matrix effects in hydrocarbon
generation. Asphaltenes can be dissolved and slurried with mineral matrices in order to obtain an intimate
organic/inorganic contact.
The effect of hydrous pyrolysis on the sterane distribution is studied in detail using computerized GC-MS. The
sterane distribution from the pyrolysate of the asphaltenes derived from a biodegraded oil have an unusually high
contribution of the biologically derived R-epimer of aaa ethylcholestane.

INTRODUCTION especially in the upper Green River Formation. The


sample of Gilsonite used in this study was from the
Asphaltenes are defined as that part of petroleum pre- 50 cm wide Nigger Bay Vein (Rainbow Group),
cipitating upon addition of excess non-polar solvents located in Uinta County.
such as n-pentane or n-heptane. Asphaltenes are thus Asphalts and bitumens have been classified by
a highly complex solubility class, without a well- Abraham (1945) mainly on the basis of solubility in
defined molecular structure. carbon disulphide and their fusibility. According to
The composition of asphaltenes may be studied by this scheme, Gilsonite belongs to the soluble, but dif-
various spectroscopic techniques or by chemical/ ficult to fuse, 'Asphaltite' class.
thermal degradation of either the total materiar or Crude oil samples were all from the same area of the
fractions separated by liquid chromatography. Earlier North Sea, and are considered to be sourced from the
work on the pyrolysis of asphaltenes has been mainly Upper Jurassic Kimmeridge Clay Formation.
concerned with flash-pyrolysis, and the use of the The Gilsonite was dissolved in dichloromethane and
resulting pyrograms in oil-oil and oil-source rock cor- asphaltenes were precipitated by addition of a 40-fold
relations (Tissot, 1984; Huc et al. 1984). excess of cold n-pentane, and left at room temperature
In these studies, asphaltenes are regarded as small overnight. Crude oil asphaltenes were precipitated
soluble fragments of kerogen with a composiiton in- with a 40-fold excess of cold n-hexane, and left at room
herited from the parent kerogen (Tissot, 1984). Pre- temperature overnight. The crude asphaltenes were
liminary investigations were concerned with the extracted in a soxhlet for 24 h with petroleum ether
hydrous pyrolysis of asphaltenes, with special (40-60 DC). Gilsonite asphaltenes were reprecipitated
emphasis on biomarker compositions and their pos- from a clean solvent, and further soxhlet extracted.
sible application in correlation studies. Asphaltene The Gilsonite asphaltenes were subjected to hydrous
samples were derived either from Gilsonite or a range pyrolysis in small stainless steel bombs, one-third filled
of North Sea crude oils. with water and in a nitrogen atmosphere, for 72 h at
temperatures varying from 280°C to 370 DC.
A solution of Gilsonite asphaltenes was slurried
EXPERIMENTAL with calcium carbonate or sodium montmorillonite in a
1 : 3 ratio and, after evaporation of the solvent, the
Gilsonite is a solid, black, naturally occurring 'bitu- impregnated solids were subjected to hydrous'
men' found extensively in the Uinta Basin, USA, pyrolysis as above.
PetTOleum Geochemistry in Exploration of'the Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 287-292.
288 PetTOleum Geochemistry in Exploration of the Norwegian She!f

C-30, but the pristane/n-C-17 and the phytane/n-C-18


ratios are still > 1. Increasing the pyrolysis tempera-
ture to 350 cc gives rise to a bimodal n-alkane dis-
tribution with maxima at C-17 and C-25. The
phytane/n-C-18 ratio approaches unity at this temper-
ature. These results are indicative of a thermal crack-
ing of the higher molecular weight n-alkanes observed
in the pyrolysate at 330 cc. One of the interesting
aspects of using asphaltenes as model compounds for
kerogen is the possibility of studying the effects of
mineral matrices on the generation of hydrocarbons.
Asphaltenes in solution can be slurried with mineral
matrices in order to achieve an intimate contact bet-
ween the organic and inorganic phases followed by
evaporation of the solvent. This will enable studies to
be carried out on the catalytic effects of single miner-
als or representative mixtures of minerals on the con-
Fig. 1. Separation scheme for the precipitation of version of asphaltenes to petroleum-like hydrocarbons
asphaltenes and fractionation of pyrolysates. under laboratory simulation experiments such as
hydrous pyrolysis (Ivanov and Shcherban, 1983).
Hydrous pyrolysis in the presence of calcium carbo-
The crude oil asphaltenes were pyrolysed at 310 °c nate or sodium montmorillonite as a mineral matrix
and 330 cc. After pyrolysis, the stainless steel bombs gives pyrolysates with a lower abundance of n-alkanes
were extracted with dichloromethane, and the hexane relative to the pyrolysate at 330 cc without the min-
soluble fractions were separated on a MPLC system eral matrix. For both the carbonate and the clay min-
(see Fig. 1). Gas chromatography was carried out on eral situations, there is a bimodal distribution of the
an OV-1 fused silica column, programmed from 50 cc n-alkanes in the saturate fractions of the pyrolysates,
to 320°C with a ramp of 4 cC/min. GC-MS work was with maxima at C-17 and C-25. The odd-even pre-
conducted on a VG7070E mass spectrometer in the dominance in then -alkanes found in the pyrolysates at
multiple ion detection mode, with an electron impact 280 cc and 300°C is not apparent at 330 cc.
energy of 70 eV. Steranes are known to be very sensitive to thermal
stress (Seifert and Moldowan, 1978) and these com-
pounds were used in order to study the maturity vari-
RESULTS AND DISCUSSION ations of the pyrolysates in more detail. Figure 3
gives the m/z 218 fragmentograms from the saturate
Gilsonite is a highly biodegraded and waterwashed fractions of some of the pyrolysis experiments, as well
natural asphalt. A gas chromatogram of the saturate as the original bitumen. Table 1 is a summary of the
fraction (see Fig. 2) shows almost complete depletion GC-MS results for the C-29 sterane ethylcholestane.
of the II-alkanes, and high concentrations oftetracyclic The steranes in the natural bitumen have a distribu-
steranes and pentacyclic triterpanes. In addition to tion which appears unusually mature for this low rank
these compounds, fJ-carotane(s) are also abundant deposit (Grantham, 1974). With increasing pyrolysis
(marked with an asterisk in the chromatogram). One temperatures the steranes start epimerizing from
of the reasons the asphaltenes from this bitumen were 5a(H), 141X(H), 171X(H) to 51X(H),14fJ(H),17fJ(H). The
chosen for these experiments is that co-precipitation IXfJP/ aaa + afJfJ ratio for the ethylcholestane changes
of n-alkanes with the asphaltenes could be virtually from 0.23 at a pyrolysis temperature at 280 cC to 0.47
neglected. at 370 cc, reflecting the relative increase in the ther-
At a pyrolysis temperature of 280°C, the saturate modynamically more stable afJfJ isomer. It is interest-
fraction is dominated by C-14 to C-20 isoprenoids, ing that even at 370°C this reaction has not reached
with pristane as the most abundant compound. its equilibrium value. With increasing maturation, the
Steranes and triterpanes are also present in abun- biologically inherited C-29 20R isomer (peak 8) is trans-
dance. A relatively uniform distribution of n-alkanes
up to C-35, with a slight odd-even predominance, is
also observed.
At 300°C the relative amount of II-alkanes in the TABLE 1. A summary of the C-29
pyrolysate has increased, with a maximum around sterane GC-MS data for the Gilsonite
C-25. Pristane is still the most abundant component, bitumen, and the mlphaltene pYl'olysates
and an odd-even predominance in the alkanes is still Sall/pie /lame Ratio 1 Ratio'!
evident. The high abundance of triterpanes/steranes
in the pyrolysates at 280°C and 300 cc might indicate G-BIT 0.21 0.25
a major contribution from co-precipitated material in G-280 0.23 0.33
G-300 0.28 0.39
these pyrolysates. Heating the asphaltenes in the 0.42
G-BI0 0.29
injection port of a gas chromatograph at 320 cc, how- G-330 0.;)1 0.36
ever, gave no detectable release of any co-precipitated G-340 0.42 0.51
material (if present). When pyrolysing asphaltenes, G-350 0.46 0.56
coals and kerogens, it is generally very hard to distin- G-:no 0.47 0.51
G-:i30C 0.32 0.41
guish between release of occluded material and true G-330Yl 0.39 0.45
pyrolytic products.
At 330°C the pyrolysate is dominated by an Ratio 1; C-29 af3f3! xaa + af3f3 ethyl cholestanes.
homologous series of II-alkanes with a maximum at Ratio 2; C-29 20S!20R + 20S aaa ethyl cholestanes.
Hydrous Pyrolysis of Asphaltenes 289

G BITUMEN
G 330
Na-montmorll1onlte

*
I
I

I
I
II,;, ,IJi III i'"II!I!
• " ,I, .. ,
.' -' .~

G 330
Ca-Carbonate

G 300
G 350

G 330

!
, '.
I I
i ~
I I, !I./ I I' !
',1,,111,., ",,"
I '11 ",' ,1.,.r.:1,'..,.",J II. ' · '

1,1","

Fig. 2. Gas chromatograms of the saturate fractions from Gilsonite bitumen and asphaltene pyrolysates at various
temperatures.

formed into the more stable C-29 20S isomer (peak 5). be negligible compared with the corresponding experi-
This leads to an increase in the 20S/20R + 20S ratio ment without the presence of a mineral matrix.
from 0.33 when pyrolysis is carried out at 280°C, to Asphaltenes derived from crude oils have a more
0.50 at 350 °C and 0.51 at 370°C. The equilibrium for uniform composition than asphaltenes from source
this reaction is reported to be 0.5 (MacKenzie and rocks. In the North Sea region, where most of the oils
Maxwell, 1981) which is reached at maximum are sourced from the Upper Jurassic Kimmeridge
pyrolysis temperature. The high value at 350°C may Clay Formation, it was considered of interest to inves-
be due to some coelution of an unknown component. In tigate whether hydrous pyrolysis of the corresponding
Fig. 4, the afJfJ/aaa + afJfJ ratio is plotted against asphaltenes would give significant differences in the
pyrolysis temperature. The effect of a mineral matrix composition of the pyrolysates.
is quite pronounced for the sodium montmorillonite, The two crudes chosen for this part of the study
and shows that this mineral has a catalytic effect on the have the same origin, but one is partially biodegraded
epimerization of steranes. The superacid effect of clay (Oil B) and the other (Oil A) is not.
minerals on geochemical transformations is well Gas chromatograms of the saturate fraction of these
documented (Sieskind et al., 1979). The effect of cal- oils and the pyrolysates are given in Fig. 5. Oil A
cium carbonate on the sterane epimerization seems to shows a smooth unimodal n-alkane distribution with-
290 Petroleum Geochemistry in Exploration of the Norwegian Shelf

M/Z 218 BITUMEN IZ 218 330

M/Z 218 350'


M/Z 218 300

M/Z 218 330


Montmorillonite

7
6

Sa(H),14a(H) ,17a(H) Steranes

Sa (H) ,146 (H), 176 (H) Steranes

Diasteranes

Fig. 3. Mass fragmentograms of rn/z 218 from Gilsonite bitumen and asphaltene pyrolysates at various temperatures.
Compound identification is given in Fig. 8.

out any odd-even predominance, typical for a mature In the pyrolysate from Asphaltene B, there is an
crude oil. Oil B, on the other hand, shows typical signs n-alkane maximum at C-29 with a slight odd-even
of biodegradation with a depletion of the n-alkanes in predominance. The pristanejn-C-I7 ratio is much
the gasoline range. This gives rise to a pristanejn-CI7 smaller than in the corresponding oil, and very similar
ratio > 1, and an apparent bimodal n-alkane distribu- to that of Oil A. The pyrolysate shows no evidence of
tion. The gas chromatograms of the saturate fractions the biodegradation affecting the oil from which the
of the two pyrolysates are distinctly different. In the asphaltenes were precipitated, but there is a marked
pyrolysate from Asphaltene A at 330°C there is the contribution to the pyrolysate from a higher molecu-
same smooth unimodal n-alkane distribution as in the lar weight fraction, maximizing around C-29.
original oil. The pristanejn-C-I7 ratio is also identical The mjz 218 mass fragmentograms of the various
to that of the corresponding oil. samples are given in Fig. 6, and the results for the
C-29 ethyl cholestane ratios are summarized in
:: 29 aBB, aaa i a96 FTrlYL.-CHDl_C:ST :.~,,::S
Table 2.
The sterane distributions of the two original oils are

)
very similar, lending support to the theory of a com-
mon origin for these crudes. The 208j20R + 208

TABLE 2. A summary of the C-29


Ii
x
::0- carto'la~~
·,a-~Gnl ..,o~; ••
sterane GC-M8 data for the two crude
oils, and the asphaltene pyrolysates
Sample name Ratio 1 Ratio 2

Oil A 0.61 0.63


Asph. A 310 0.55 0.46
Asph. A 300 0.54 0.46
OilB 0.64 0.59
Asph. B 330 0.40 0.22

Fig. 4. The C-29 IIPP/ IIIIII + IIPP ratio for Gilsonite asphal- Ratio 1: C-29 a{J{J/aarz + rz{J{J ethyl cholestanes.
tene pyrolysates plotted against pyrolysis temperature. Ratio 2: C-29 208/208 + 20R rzrzrz ethyl cholestanes.
Hydrous Pyrolysis of Asphaltenes 291
OIL B

I I I ,

ASPH. A 330
ASPH. B 330

Fig. 5. Gas chromatograms of the saturate fractions of the crude oils and the asphaltene pyrolysates from the pyrolysis at
330 cC.

ratios (ratio 2) are 0.63 and 0.59, respectively, for Oils asphaltene pyrolysates, particularly in the early part
A and B, which is higher than the expected equilib- of the fragmentograms in the diasterane region. In
rium value for this transformation. This may be due to both pyrolysates, ratio 2 has reached the expected
co-elution of the 208 isomer with an unknown compo- equilibrium value of 0.5, but they are less mature than
nent. the original oil as measured by the rxPP/rxrxrx + rxpp
The fragmentograms of the pyrolysates from ratios. These results suggest that steranes released
Asphaltene A at 310 ~C and 330 DC are very similar, from asphaltenes during hydrous pyrolysis of asphal-
and the maturity-as measured by sterane epimeriza- tenes have retained more of their biologically inher-
tion-is identical. There are some apparent differ- ited stereochemistry.
ences in the sterane distributions in the oil and the The m/z 218 fragmentogram from the pyrolysis of

, ,,
2 ,
2
M/Z 218 OIL A
M/Z 218 OIL B

, • 7
2

M/Z 218 ASPH. A 310

" " M/Z 218 AsPH. B 330'

M/Z 218 AS PH. A 330

Sa (H) ,14a (H) ,17a (H) Steranes

Sa (H) ,146 (H) ,176 (H) Steranes

Diasteranes

Fig. 6. Mass fragmentograms of m!z 218 from the crude oils and the asphaltene pyrolysates. Compound identification is
given in Fig. 8.
292 Petroleum Geochemistry in Exploration of the Norwegian Shelf

graded oil B, there is a contribution from a suite of


higher molecular weight alkenesjalkanes. Apart from
ASPH.8
this observation, the two pyrograms are strikingly
similar.

CONCLUSIONS
These investigations suggest that hydrous pyrolysis of
asphaltenes separated from either natural bitumens or
crude oils can be used to improve understanding of a
number of geochemical processes, such as the follow-
ing.
(i) Asphaltenes from biodegraded crudes or bitu-
mens can be used to provide indications of the likely
ASPH. A
composition of the non-degraded material. The com-
position of the hydrous pyrolysis products may also
reveal material derived from micro-organisms.
(ii) If asphaltenes can be regarded as soluble frac-
tions of kerogens, then they can be used as a substrate
for stUdying the catalytic effects of various mineral
matrices. The present studies showed very large dif-
ferences in the effect of clays and carbonates on hy-
drous pyrolysis products.
(iii) The relative ease of separating asphaltenes in
Fig. 7. Pyrograms from the anhydrous pyrolysis-gas comparison with kerogens provides a simpler means of
chromatography of the two crude oil asphaltenes. Pyrolysis simulating the process of hydrocarbon generation
temperature 600°C. from kerogen-like material.
(iv) Biomarker compositions of hydrous pyrolysis
products may provide another means of assessing
crude oil maturity, as well as improving oil correlation
R=CH 3 : 1(20S) , 4 (20R)
studies.
R=C 2H S :S(20S),8(20R)
A significant problem in the use of asphaltenes in
hydrous pyrolysis studies is the difficulty of distingu-
SaCH),14a(H) ,17a(H)-Steranes ishing occluded material from 'true' pyrolytic pro-
ducts, and further work is necessary on this problem.
R=CH 3 :3(20S),2(20R)

R=C 2H S :7 (205) ,6 (20R) REFERENCES


Abraham, H. 1945. Asphalts and Allied Substances. Van Nostrand,
5a(H),148CH),178(H)-Steranes New York.
Grantham, P. 1974. Ph.D. Thesis, University of Newcastle.
Fig. 8. Compound identification. Huc, A. Y., Behar, F. and Roussle, J. -C. 1984. Geochemical vari-
ety of asphaltenes from crude oils. In: Symp. on Characterization
of Heavy Crude Oils and Petroleum Residues, Lyon, 25-27 June
Asphaltene B at 330°C is distinctly different from 1984. Technip, pp. 99-108.
that of the original oil, in that the pyrolysate appears Ivanov, V. V. and Shcherban, O. V. 1983. Mineral matrix influence
on the dynamics and products of organic matter catagenetic
much more immature. There is a possibility that the transformation. Org. Geochem., 4(3/4), 185-194.
low maturity-as measured by sterane epimerization MacKenzie, A. S. and Maxwell, J. R. 1981. Assessment of thermal
and the presence of higher molecular weight n-alkanes maturation in sedimentary rocks by molecular measurements. In:
with an odd-even predominance in the pyrolysate of Brooks, J. (ed.), Organic Maturation Studies and Fossil Fuel
Exploration. Academic Press, London, pp. 239-254.
the asphaltenes from the biodegraded oil-may reflect Seifert, W. K. and Moldowan, J. M. 1978. Application of steranes,
a contribution from the microorganisms responsible terpanes and monoaromatics to the maturity migration and
for the biodegradation. This will certainly help to source of crude oils. Geochim. Cosmochim. Acta, 43, 77-95.
explain the high relative abundance of the C-29 20R Sieskind, 0., Joly, G. and Albrecht, P. 1979. Simulation of the
isomer in this pyrolysate. This suggestion is sup- geochemical transformations of sterols: superacid effect of clay
minerals. Geochim. Cosmochim. Acta, 43, 1675-1680.
ported by anhydrous pyrolysis-GC data on Asphal- Tissot, B. 1984. Geochimie des resines et asphaltenes. In: Symp. on
tenes A and B which are shown in Fig. 7. In the Characterization of Heavy Cmde Oils and Petroleum Residues,
pyrogram of the asphaltenes derived from the biode- Lyon, 25-27 June 1984. Technip, pp. 3-18.
25 Aromatic hydrocarbon
v'ariations in North Sea Wells
P. B. Hall, L. Schou and M. Bjoroy
Continental Shelf Institute (IKU), Trondheim, Norway

An investigation was made into the use of aromatic hydrocarbons in routine geochemical analysis of North Sea
wells.
Generally, in routine geochemical analysis only saturated hydrocarbons are analysed by gas chromatorgraphy.
However, aromatic hydrocarbons can also supply useful information on maturity, and also on environment of
deposition and organic input. Aromatic maturity ratios such as methyl phenanthrene index have already been
employed in other areas. Certain aromatic hydrocarbons in some oils have been related to specific organic inputs;
for example, cadalene from land plant sources. Observations are made on some of the variations in aromatic
hydrocarbon distributions and the possible causes of those variations.

INTRODUCTION ancient sediments are, in part, derived from non-


aromatic biogenic precursors (Laflamme and Hites,
Aromatic hydrocarbons have been identified in all 1978, Wakeham et al., 1980). Some of the processes
recent and ancient sediments (Maxwell et al., 1971; involved in the formation of the polycyclic aromatic
Anders et al., 1973; Giger and Blumer, 1974; Blumer hydrocarbons in recent sediments during early
and Youngblood, 1975; White and Lee, 1980; Radke diagenesis have been elucidated (Blumer and Young-
and Welte, 1981; Radke et al., 1982a,b). However, blood, 1975; Youngblood and Blumer, 1975; Greiner et
unlike saturated hydrocarbons, aromatic hydrocarbon al., 1976; Laflamme and Hites, 1978, 1979; Wakeham
distribtuions were until recently only treated in gen- et al., 1979, 1980).
eral terms, and related to gross changes rather than The suggested precursor compounds for aromatic
individual compound variations. Early investigations hydrocarbons include unsaturated fatty acids (Breger,
of aromatic hydrocarbons in source rocks have concen- 1960; Erdman, 1961), carotenoids (Day and Erdman,
trated mainly on the types of aromatic hydrocarbons 1963; Ikan et aI., 1975; Gallegos, 1981), polyhydroxy-
in coals (e.g. Hofmann and Damm, 1922; Sakabe and quinones (Blumer, 1965) and terpenoids (Mair, 1964;
Sassa, 1952; Danyuskerskaya, 1959; Streibl and Her- Albrecht ct al., 1976; Greiner ct aI., 1976).
out, 1969; Birkofer and Pauly, 1969). More recently, Polycyclic terpenoids and steroids produced
general trends in aromatic hydrocarbon distributions biogenetically do not have methyl groups in positions
with maturity for both oil- and gas-prone source rocks which would correspond to the C-4 or C-5 positions of
have been assessed (Tissot et al., 1971, 1974; Albrecht phenanthrene (Mail', 1964). Until recently the 4- and
et al., 1976). Specific aromatic ratios, particularly of 5-methyl phenanthrenes had not been identified in
mono-aromatized steranes, in both oils and oil-prone petroleum. Recent studies using high resolution spec-
source rocks have also been shown to be maturity- trofluorimetry (Ewald et ai., 1983) suggest that
sensitive (Seifert and Moldowan, 1978; Mackenzie et 4-methyl phenanthrene is present in minor amounts in
ai., 1981). Artificial maturation experiments on recent some petroleums. However, the proposition that the
marine kerogens by Ishiwatari and Fukushima (1979) phenanthrenes in petroleum are derived to some
showed that the relative distribution of alkylphenanth- extent from polycyclic terpenoids and steroids is able
renes changes with increased maturity. Radke et al. to account for the general low concentrations of 4- and
(1981, 1982a,b) have shown that several aromatic hyd- 5-alkyl derivatives. The relatively small content of
rocarbon ratios can be used as maturity parameters. compounds with an anthracene structure is an addi-
Biosynthesis of polycyclic aromatic hydrocarbons tional support for the claim that the terpenoids and
has been shown to be unlikely (Hase and Hites, 1976). related steroids (containing the phenanthrene struc-
The aromatic hydrocarbons found in recent and ture) are major precursors (Mail', 1964; Greiner et al.,
Petl'O/elllll Geocilcllli.,tl:q ill E.rp/omtioll anhe Norwegian Shelf © Norwegian Petroleum Society (Graham & Trotman. 1985) pp. 293-301.
294 Petroleum Geochemistry in EJ"plomtion of the Norwegian Shelf

1976). One specific source for alkylated phenanthrenes and Lee, 1980). Sulphur aromatics, such as ben-
may come from the terpenoids derived from con- zothiophene and dibenzothiophenes found in coals, can
stituents of higher plants, such as the resins of conifer be produced by the reaction of elemental sulphur or
species. Diterpenoid acids are major constituents of pyrite with hydrocarbons (Deroo and Hodgson, 1978).
higher plant resins (Thomas, 1969), and several recent The abundance of sulphur aromatic compounds in
studies have indicated that aromatic hydrocarbons in- marine type II kerogens, compared with type I and
cluding retene and pimanthrene have been derived by III kerogens, is probably related to environmental
diagenetic aromatisation of such acids (Skrigan, 1964; conditions during deposition. Type II kerogens gener-
Stonecipher and Turner, 1970; Laflamme and Hites, ally form in anoxic marine waters with abundant
1978; Wakeham et al., 1980). Radke et al. (1982a) have available sulphur in the form of sulphates, and in
suggested that a substantial source for alkyl phenanth- which sulphur bacteria play an important role.
renes is through direct methylation of phenanthrene The main objective of this investigation has been to
during catagenesis. assess the usefulness of aromatic hydrocarbon ratios
Other tricyclic aromatics, such as alkylated as maturity parameters in North Sea wells. These
fluorenes, have been identified in petroleum (Mair, ratios are known from other studies to give useful
1964). Sources for such compounds would come from information about maturity (Radke and Welte, 1981;
alkaloids or steroids (Mair, 1964) or at very high Radke et al., 1982a,b).
temperatures from phenanthrenes (Yu and Hites, This paper deals firstly with the general changes in
1981). aromatic hydrocarbon profiles and secondly with vari-
Another possible source of di- and tricyclic aromatic ations in individual aromatic hydrocarbon ratios
compounds may come through cyclization and dehyd- (MNR, MPI 1, etc.).
rogenation of unsaturated compounds such as certain Nine wells were chosen for the study and basic
fatty acids and carotenoids (Erdman, 1961; Ikan et al., organic geochemical data were obtained for all the
1975; Cane, 1976; Gallegos, 1981). Alkylated naph- wells to evaluate the source rock potential, maturity
thalenes are common constituents of coals (Birkofer and possible migration effects.
and Pauly, 1969; Radke et al., 1982b), petroleum
(Carruthers and Douglas, 1955; Yew and Mair, 1966) RESULTS AND DISCUSSION
and shales (Murphy, 1969; Anders et al., 1973; Radke
et al., 1982a). It has been postulated that naph- The samples used in this study for evaluation of
thalenes, e.g. eudalene, are dehydrogenation products aromatic hydrocarbons were mainly of Jurassic age
of sesquiterpenoids (Murphy, 1969; Bendoraitis, from nine different wells in the North Sea, together
1974). Alternatively, alkylated naphthalenes could be with some outcrop samples to cover the whole matur-
the products of diagenetic aromatization of ity range and spread in kerogen types.
carotenoids, as mentioned above. Mair and Douglas When dealing with cutting samples from wells it is
(1965) have shown that the reaction of elemental sul- extremely important to evaluate problems arising
phur with terpenoids and steroids produces alkylated from caved material, turbodrilling, mud additives and
naphthalenes, sulphur aromatics plus other aromatic migrated hydrocarbons. This will not be discussed in
hydrocarbons. this paper, but reference will be made to these prob-
Polycyclic aromatic hydrocarbons with more than lems where they arise.
three rings which have been identified in recent sedi-
ments and coals include fluoranthene, pyrene, benzan- Variations in aromatics with kerogen
thracene and perylene (Aizenshtat, 1973; Orr and type/depositional environment
Grady, 1967; Wakeham et al., 1980). Non-alkylated
There are three main groups of peaks seen in the
polycyclic aromatic hydrocarbons such as fluoran- aromatic hydrocarbon gas chromatograms of most
thene, pyrene and chrysene may require longer samples (aromatic fraction obtained by extract frac-
exposure to higher temperatures for formation, and tionation by MPLC). These are the following.
involve aromatization and dealkylation of partially
hydrogenated aromatics (White and Lee, 1980). Group 1. Naphthalene and alkyl naphthalenes.
Variations in aromatics in sediments have been Group 2. Phenanthrene and alkyl phenanthrenes.
mostly related to variations in source and maturity Group 3. Aromatised steranes and triterpanes plus
rather than in depositional environment. In type III other polycyclic compounds.
kerogens (as defined using Rock-Eval hydrogen indi- Figure 1 is a typical example showing the three
ces) early aromatization during deposition or organic main groups, and tentative identifications for some of
matter in relatively acid, mildly reducing of oxidizing the main peaks in groups 1 and 2. Another common
environments leads to the early formation and domi- group of compounds are the organic sulphur com-
nance of pure aromatic compounds. Thus Vanden-
pounds such as dibenzothiophene and methyldiben-
broucke et al. (1976) have observed that in the Douala zothiophenes.
shales with type III kerogens the aromatics consist If only the groups consisting of fully aromatized
dominantly of pure aromatic molecules. These are components (i.e. 1 and 2) are <,onsidered, there are
mainly naphthalene and phenanthrene types with several useful maturity parameters which can be
short alkyl chains, which show little evolution with used. However, before discussion of variations within
depth regarding different cyclic structures or carbon the groups there are many variations in the relative
numbers. However, in type II kerogens, as found in amounts of the three different groups which can be
the Toarcian shales of the Paris Basin, there is consid- roughly related to a combination of maturity and type
erable evolution in pure aromatic compounds with of kerogen.
depth due to rearrangement and breaking of alkyl
chains in polycyclic naphthenic aromatics which are (a) Group 1 compounds (naphthalene and alkylated
abundant in the immature samples. naphthalenes) are absent or in low concentration in
The abundance of sulphur aromatics certainly type IV kerogens and in highly mature (>2% R 0)
appears to be related to environmental factors (White kerogens of all types (see Fig. 2). The level of matur-
Aromatic Hydrocarbon Variations 295

DBT
FPD

MDBT
~

MN
Il
2
DMN
Il
1,6
TMN
FID ,--,

1,3 + 1,7

\ 1,4+2,3 MP
2.6 I I"j'1
+ 2.7 /.5 c
1,2 {J

~ ________~______~I LI____________~I LI___________________


2
Fig. 1. identification of main aromatic components, Vitrinite reflectance Ro = 0,6%.

MP

DMP
rt

P ; Phenanthrene
MP ; Methylphenanthrene
DMP; Dimethylphenanthrene

'"

Fig. 2. Overmature type III kerogen,


296 Petroleum Geochemistry in Exploration of the Norwegian Shelf

""
6.7%TOC Contaminated saturated fraction ..
Type II kerogen
Immature
...... .. "
..
x-hrtiallyArOlllillisedcompoundTet.al,"

MNR 1

.'
o Brent/Sleipner
Coaly claystones

3
Pris/Phyt
Fig. 3. Aromatic hydrocarbons, immature Kimmeridge
Clay Formation shale.
MNR = Methylnaphthalene ratio = 2/1 MN

ity at which group 1 compounds become absent in gas


chromatograms probably depends on the richness and Fig. 5. Plot of MNR versus pristane/phytane.
type of the kerogen. Samples with poor type III
kerogens probably lose all group 1 compounds at
a lower maturity level than type I and II kerogen To indicate the degree to which aromatic hydrocar-
samples. bon ratios vary with kerogen type and/or depositional
(b) Group 3 compounds, which include aromatized environment, plots have been made for certain aroma-
steranes and triterpanes are absent or in low concent- tic hydrocarbon maturity ratios versus the pristane/
ration relative to the group 2 compounds in the case of phytane ratio calculated from the saturated hydrocar-
type IV kerogens and highly mature kerogens (see bon fractions of the same sample. It is already recog-
Fig. 2). Aromatic steranes and triterpanes are par- nized that pristane/phytane ratios are affected by
ticularly prominent in immature samples (e.g. Fig. 3, depositional environment as well as maturity (Brooks
less than 0.5% R 0> which have type I or II kerogen. et al., 1969; Powell and McKirdy, 1973; Lijmbach,
(c) Group 1 and 2 compounds are absent or very 1975; Didyk et al., 1978). If aromatic ratio~ ~re n?t
minor components in very immature sediments, i.e. affected by type differences, then any VarIatIOns m
less than 0.4% R 0 (see Fig. 3). Common components pristane/phytane ratio will not reflect in the aromatic
in such sediments are hydro-aromatic (partially ratios and a straight-line plot should result. These
aromatized) compounds. Groups 1 and 2 are mainly plots are shown for one well, chosen because the
formed during thermal degradation of the kerogen in Jurassic interval analysed appears to be least affected
the sediments (see Fig. 4). by contamination and migrated hydrocarbons. Also,
(d) Group 1 compounds are relatively more abun- there are only minor changes in maturity over the
dant (compared with compounds of groups 2 or 3) in interval under consideration from moderately to early
samples containing mature type I and II kerogens mature (i.e. 0.5-0.6% R 0).
than in samples with type III kerogens, including Figures 5-7 show the variation in methyl naph-
coals. However, there are some samples which do not thalene ratio (MNR), ethyl naphthalene ratio (ENR)
follow this rule. and dimethyl naphthalene ratio (DMNR)-see
(e) Oils, and particularly condensates, are much Table 1 for actual ratios used. These show a noticeable
richer in group 1 compounds than group 2 or 3 com- increase in values with increasing pristane/phytane,
pounds. The presence of migrated condensate hyd-
rocarbons in immature shales can give misleading gas
chromatograms (although this is probably clearer on
examination of saturated hydrocarbon traces).

Contaminated
..
ill ..
saturated fraction
'-" "
MN
.
" ENR 1

o Brent/Sleipner
6.3%TOC
Type II/iliofiyinaliv Coaly claystones
typellke.ogen?
OM'

'" E,rlym.lu,e.

OOT MP
,"
2 J
Pris/Phyt
ENR = Ethylnaphthalene ratio =2/1 EN
C2

Fig. 4. Aromatic hydrocarbons, early mature Kimmeridge


00
Clay Formation shale. Fig. 6. Plot of ENR versus pristane/phytane.
Aromatic Hydrocarbon Variations 297

3611 m/ ~ \ Migrated hydrocarbons at base Heather Fm. (1)

. .... ..
{ I
\~/ 3569m

.. MPII 1

DMNR 1
o Brent/Sleipner
Coaly claystones

o Brent/Sleipner
Coaly claystones
0.5 . ._• • .a ..
3
3 Pris/Phyt
Pris/Phyt MPI1 ~ 1.5 x (3 MP + 2 MP)
~ P+9MP+1MP

00
Cl Cl DMNR 2,6 + 2,7 Dimethylnaphthalene
1,5 Dimethylnaphthalena ~ MP' Methyl phenanthrene

Fig. 7. Plot of DMNR versus pristane/phytane. Fig. 8. Plot of MPI 1 versus pristane/phytane.

particularly when a comparison is made between the


non-marine coaly shale and marine shales. This sug-
gests that there are slight differences in these ratios
between the mainly non-marine Brent coaly clays-
tones and the marine Kimmeridge claystones used in
this stUdy.
Figures 8 and 9 show the variation in MPI 1 and 2/1 MP 1 Migrated hydrocarbons at the
2/1 MP versus pristane/phytane. In both wells there :'.';- base olthe Heather Fm. (?)
is very little change in these parameters, regardless of
organic matter type/depositional environment (par-
ticularly in the MPI 1 values). The only problem with
MPI 1 is that at low maturities there appears to be
0.5 '.
• \. :

....,. ..
,'
. •
0 Brent/Sleipner
Coaly claystone

interference from abundant sulphur aromatics present


in extracts from marine shales with type II kerogens. 2 3

'" ".""." '"~.....


A plot of the variatirm in the sulphur aromatic ratio Pris/Phyt
4/1 MDBT versus pristane/phytane was also made
(Fig. 10). This indicates a fairly good correlation be- ~" ~

TABLE 1. List of maturity ratios, indicating those Fig. 9. Plot of 2/1 MP versus pristane/phytane.
used in this paper

2-methylnaphthalene
*MNR
I-methylnaphthalene
biphenyl
*BPHR
1,6 dimethyl naphthalene
2-ethyl naphthalene
*ENR
I-ethyl naphthalene
2,6 + 2,7-dimethyl naphthalene
.,. ..
*DMNR
1,5-dimethyl naphthalene 4/1 MDBT
4MDBT
DBT
4-methyl dibenzothiophene
dibenzothiophene
II •
o Brent/Sleipner
*4MDBT 4-Methyl dibenzothiophene Coaly claystones
IMDBT I-methyl dibenzothiophene
3/2MDBT
IMDBT
3 + 2-methyl dibenzothiophene
I-methyl dibenzothiophene
..
2-methyl phenanthrene 3
*2/1 MP
I-methyl phenanthrene Pris/Phyt
1.5 x (3MP + 2MP)
*MPll
P + 9MP + IMP where P = phenanthrene and ~ 4/1 MOBT • Methyl dibenzothiophene
MP = methylphenanthrene

*Ratios used. Fig. 10. Plot of 4/1 MDBT versus pristane/phytane.


298 Petroleum Geochemistry in Exploration of the Norwegian Shelf

DMN
m
Diesel contamination
3000
,.1 0

~.
•• •
... 0
MN ::
te
",
• .1; ... I> tJ.
Sequence more mature at
.S • ~., • • Oil show? shallower depths
TMN .c '.~~ 0
r:-1 MP
i • ~ cP

... ..
m c
0
Coal
4000 \

... .
DMP
\
"- ........... _ _ 6.2

-- • __ ~.3

Turbodrill affected or @T .'


5000 ..L.::hi",9h:,ch:.:ea;:;tin:!.g:.:rat::.e.:.,?r--,----.----,,--L-
lI.DI 41.... ".11
RETENHDI'I TIME ("nlllES)
DMNR
Fig. 11. Post-mature Kimmeridge Clay Formation. Fig. 13. Variation in dimethylnaphthalene ratio with depth.

tween the two ratios, and suggests that the 4/ 1 MDBT (BPH) and methylbiphenyl (MBPH) is also seen at this
ratio is affected by variations in organic matter and/or high maturity level.
depositional environment. However, for Kimmeridge
A list of the various maturity ratios that were applied
Clay Formation samples this and other sulphur
is given in Table 1. The various parameters will be
aromatic ratios show a marked change after peak oil
generation, which is a useful rough guide to maturity. discussed separately, and plots of some of the paramet-
Individual compounds are known to be source- ers versus depth have been made for the data in all the
wells in the study.
specific. Pyrene and fluoranthene and other polycyclic
aromatics are most common in coals and non-marine Methyl naphthalenes. Two ratios were calculated
type III kerogens and in sediments of very high from the distribution of the methyl naphthalenes, the
maturity. Other polycyclic aromatic compounds such methyl naphthalene ratio (MNR) and dimethylnaph-
as retene were found in immature to moderately mature thalene ratio (DMNR) (see Table 1 for actual com-
sediments and indicate a major land plant input. pounds used in determination of ratios). They are both
fairly simple to calculate, but are susceptible to inac-
Aromatic hydrocarbon variations and maturity curacies due to work-up procedures, and are affected
by hydrocarbon contaminants and migrated hydrocar-
The main variations in aromatic hydrocarbon distribu-
bons. They are therefore only rough maturity para-
tions with increasing maturity can be summarized as
meters. The MNR changes from less than 1 in imma-
follows.
ture to early mature samples «0.7% R 0) to greater
1. Immature samples are often dominated by hy- than 1, increasing to 2 from 0.7 to 1.3% Ro. The
droaromatic compounds (partially saturated aromatic DMNR changes from values between 1 and 1.5 in
compounds, see Fig. 3). immature to early mature samples to 2+ in mature
2. The characteristic aromatic hydrocarbon gas samples. A depth plot of the DMNR shown in Fig. 13
chromatogram is seen in moderately mature to oil shows a large difference between coal and claystones
window mature samples (e.g. see Fig. 4). Particularly over a narrow depth interval within one well. This
prominent are group 1 and group 2 compounds. probably indicates that the type of kerogen affects
3. Disappearance of group 1 compounds begins at these maturity ratios.
maturities above 1.3% Ro and particularly above Biphenyl. The ratio of biphenyl (BPR) and one of
2% Ro (see Figs. 11 and 12). Prominence of biphenyl the dimethyl naphthalenes (1,6 DMN) was seen to

"0

"
lOOO
f @ Coal
® Coal

~I···
~

~
" Sequence more mature at shallower depths

9
.c
MP i
c
.."
~ Coal
@
4000
\

TMN
r;--,
-Methyl

5000 ...L-_ _ _ _~..,__-----.,___

11.11 tt.1I 1iI.1I


RETENTION TInE (MUfJTESl
Bph
Fig. 12. Dry gas zone, Kimmeridge Clay Formation. Fig. 14. Variation in biphenyl ratio with depth.
Aromatic Hydrocarbon Variations 299

(a) The methyl phenanthrene ratios are easy to


measure often without requiring further group sep-
aration, although sulphur aromatics make measure-
..
Diesel affected
3000 'I
ment difficult in moderately to early mature samples .
:.~:
.~ I
These ratios suffer less from evaporation losses than
.~:t
~ ~c
' do the naphthalene ratios, but are affected by conta-
minants such as diesel as are the other ratios.
, " *t.,
•• Sequence more mature
~ at shallower depths
~

~c Sulphur aromatics. The sulphur aromatic ratios are


0
.... \ U 0
'\ Coal
easy to measure (if the chromatograph is equipped
!- 4000
.:\
\ -
Go with a sulphur-sensitive detector, e.g. flame photome-
'" '.".' . . ter), but they are affected by type of kerogen and
'-.~
' .. environment. They show a general change in marine
type II kerogens from values less than 2 in
moderately-early mature samples to values greater
than 10 past peak oil generation, i.e. >0.8% Ro. The
5000...L------,--------, ratios which we used are the 4/1 methyldiben-
zothiophene and 2 + 3/1 methyldibenzothiophene.
2/1 MP
It is apparent that many of the ratios discussed
change radically in value only after peak oil genera-
(b) tion. This is, perhaps, a feature similar to that seen on
pyrolysis of kerogens and analysis of biomarker com-
0?J-Diesel pounds produced (Seifert and Moldowan, 1978). This
3000
~ Sulphur aromatic - rich has shown that whflst 'free' (i.e. not kerogen-bound)
hydrocarbons are able to isomerise with increasing

~
'1" " .,"0" .0••••• " •• "o~, ....
maturity, the kerogen-bound compounds are not.
Similarly, for aromatic compounds isomerization may
occur in the soluble bitumen and when released from
~
.:
.c
'.t,. \ ~ Coal • Approximate the kerogen. When most of the hydrocarbons have
been generated, then isomerization processes domi-
~ 4000 '\' Peak oil
'" ....: (!) 0 generation interval nate.
..:./a",.. .. ", There appear to be variations in the trend for ratios
such as the MPI 1 with vitrinite reflectance due to
6~' ~.
~'.(iii! '-.
• ~ differences in heating rate. Figure 16 shows the
~,
(il, trends for a Western Canada well (Radke et al.,
1982a), from the Bramsche Massif (Radke and
5000 ~---r----r---..-----'
05 1.5 Willsch, 1981), from the North Sea wells used in this
MPII study (excluding those samples considered to be con-
taminated and/or turbodrill affected) and artificially
Fig. 15. Variation in methylphenanthrene ratio with depth. matured Kimmeridge type II samples also from this
(a) 2/1 Methylphenanthrene; (b) MPI 1.

change with maturity. However, the difference in the MPll


ratio for coal and claystone samples from the same
depth interval suggests that type of organic matter
plays an important role in this context. Values in
claystones increase from <0.1 in immature to moder-
ately mature samples to 0.3 at about 1% R 0, and 0.5
rapidly increase to values greater than 1 above 1% R 0
(see Fig. 14).
o
o
--- --- --- --
Methyl phenanthrenes. Two ratios were deter- o
o
o
- ............ ,
,,
mined from the distribution of phenanthrenes, the
0

o ,
2/1 methylphenanthrene, and the MPI 1. The gen- \
\
eral trend observed in most samples is for a decrease 1.0 0' \
in phenanthrene and the 9- and I-methyl phenant- \
\
hrenes with increasing maturity relative to the 3- and \

2-methyl isomers (at least to about 1.3% Ro) (see


Figs. 15(a) and (b».
If the important process in the formation of methyl - - - Bramsche Massif - Radke and Willsch (1981)
1.5 - - - . North Sea Well data (this work)
phenanthrenes is by methylation of phenanthrene, - - Radke and co·workers (1981 - J)
then in this instance the I-methyl phenanthrene is o Variation for type II kerogens (Kimmeridge) artificially matured
preferred. Radke et al. (1982b) indicate that intense
generation of phenanthrene occurs before methyl
phenanthrene generation and during peak oil genera-
tion. Therefore migration of phenanthrene before
methyl phenanthrene generation could account for 0.5
%Ro
1.0 1.5

phenanthrene depletion relative to the methyl


phenanthrenes. Fig. 16. Variation in MPI 1 index with vitrinite reflectance.
300 Petroleum Geochemistry in Exploration of the Norwegian She(f

study. This suggests that in areas of high heating Organic geochemical parameters of palaeoenvironmental condi-
tions of sedimentation. NatuTe, 272, 216.
rates the vitrinite reflectance increases before MPI 1 Erdman, J. G. 1961. Some chemical aspects of petroleum genesis as
values change, unlike in more normal thermal related to the problem of source bed recognition. Geochilll. C0811/-
regimes. It is therefore possible that rapid heating ochilll. Acta, 22, 16-36.
due to turbo-drilling may also have the same effect as Ewald, M., Lamotte, M., Garrigues, P., Rima, J., Veyres, A. and
natural high heating rates, e.g. due to intrusives. Laponyade, R. 1983. Determination of isomers of monomethyl
phenanthrene extracted from petroleum by capillary gas
chromatography coupled to mass spectrometry and by high resol-
ution spectrofluorimetry in alkane crystals at a temperature of
CONCLUSIONS 15 K. In: Bjorfly, M. (ed.), Advances ill Orqallic Geochemistty
1981. Wiley, pp. 705-709.
Both source input and heating rate affect aromatic Gallegos, E. J. 1981. Alkyl benzenes derived from Carotenes in
coals by GC/MS. J. Ckroll/lltog. Sci., 19, 177-182.
hydrocarbon distributions. The naphthalene ratios Giger, W. and Blumer, lVI. 1974. Polycyclic aromatic hydrocarbons
MNR and ENR are useful, although changes from the in the environment. Isolation and characterization bv chromato-
true values can occur due to work-up procedures, pre- graphy, visible, ultra-violet and mass spectromet~y. Annlyt.
sence of migrated light hydrocarbons and contamina- Chelll., 46,1663-1671.
Greiner, A. Ch .. Spyckerell, C. and Albrecht, P. 1976. Aromatic
tion. The most useful ratios employed appear to be hydrocarbons from geological sources 1. New naturally occurring
DMNR and the methyl phenanthrene ratios, particu- phenanthrene and chrysene derivatives. Tetrahedrull, a2,
larly the MPI 1 index. At higher maturities the 257-260.
biphenyl ratio (BPHR) appears to be useful, although Hase, A. and Hites, R. A. 1976. On the origin of polycyclic aromatic
there are not enough data to confirm the trend. hydrocarbons in recent sediments: biosynthesis by anaerobic bac-
teria. Geochilll. COSlllOchilll. Acta, 40, 1141-1143.
Hofmann, F. and Damm, P. 1922. The pyridine extraction of Upper
Silesian Coals. Brenll~toff-Chelll., 3, 81-91.
ACKNOWLEDGEMENTS Ikan, R., Aizenshtat, Z., Baedecker, M. J. and Kaplan, I. R. 1975.
Thermal alteration experiments on organic matter in recent
We thank the following oil companies for supplying marine sediment-I. Pigments. Geochilll. Cosmochim. Acta, 39,
the samples and for their permission to publish this 173-185.
Ishiwatari, R. and Fukushima, K. 1979. Generation of unsaturated
paper: Mobil Exploration Norway Inc., Norsk Hydro and aromatic hydrocarbons by thermal alteration of young
a.s., Statoil, Unionoil Norge A/S, Elf Aquitaine kerogen. Geochim. Coslllochilil. Acta, 43, 1343-1349.
Norge AjS, Phillips Petroleum Co. Norway, FINA Laflamme, R. E. and Hites, R. A. 1978. The global distribution of
Exploration Norway U/A and Texaco Exploration polycyclic aromatic hydrocarbons in recent sediments. Geuch; 11/.
C08ll/Ochilil. Acta, 42, 289-303.
Norway A/S. Laflamme, R. and Hites, R. A. 1979. Tetra- and pentacyclic, natur-
ally occurring, aromatic hydrocarbon in recent sediments.
Geochilll. Coslllochilll .•4.eta, 43, 1687-1691.
REFERENCES Lijmbach, G. W. M. 1975. On the origin of petroleum. Pmc. Ninth
Wul'ld Petroleu III COllq., Spec. 5. p. ;,57.
Aizenshtat, Z. 1973. Perylene and its geochemical significance. Mackenzie, A. S., Maxwell, J. R., Coleman, M. L. and Deegan, C.
Geochilli. Cosllloch·illl. Acta, 37, 559-567. E., 1983. Biological marker and isotope studies of North Sea
Albrecht, P., Vandenbroucke, M. and Mandengue, M. 1976. crude oils and sediments. E/C1!ellth World PetTO/eUlIl COUqT.
Geochemical studies on the organic matter from the Douala Basin Panel Discussion 1. London, 1~ pp.
(Cameroon)-I. Evolution of extractable organic matter and the Mackenzie, A. S., Hoffman, C. F. and Maxwell, J. R. 1981. Molecu-
formation of petroleum. Geochim. CoslI/ochim. Acta, 40, lar parameters of maturation in the Toarcian shales, Paris Basin,
791-799. France-III. Changes in the aromatic steroid hydrocarbons.
Anders, D. E., Doolittle, F. G. and Robison, E. W. 1973. Analysis Geochim. C08l1/Ochilll. Acta, 45,1345-1355.
of some aromatic hydrocarbons in a benzene-soluble bitumen from Mail', B. J., 1964. Terpenoicls, fatty acids and alcohols as source
the Green Rive~ Shale. Geochim. Coslllochim. Acta, 37, materials for petroleum hydrocarbons. Geochilil. Cosllluchilll.
1213-1228. Acta, 28, 1303-1321.
Bendoraitis, J. G. 1974. Hydrocarbon of biogenic origin in Mail', B. J. and Douglas, A. G. 1965. Sulfm': role in genesis of
petroleum-aromatic triterpenes and bicyclic sesquiterpenes. In: petroleum. Sciellce, 147, 499-50l.
Tissot, B. and Bienner, F. (eds.), Advances 'i" orq(! II ic Maxwell, J. R., Pillinger, G. T. and Eglinton, G. 1971. Organic
Geochemistl'y, 1978. Editions Technip, Paris, pp. 209-224. geochemistry. Chell}. Soc. LOlld. Q. Rev., 25(4), 571-G28.
Birkofer, L. and Pauly, W. 1969. Gas chromatographische und mas- Murphy, :VI. T. J. 1969. Analytical methods in organic geochemis-
senspektroskopische Untersuchung von Kohleextrakten. try. In: Eglinton, G. and Murphy, M. T. J. (eds.), Ol:qanic
Brenl/stoff-Chelll., 50,376-382. Geochemistl7f. Springer-Verlag.
Blumer, M. 1965. Organic pigments: their long-term fate. Science, Orr, W. L. and Grady, J. R. 1967. Perylene in basin sediments of
149, 722-726. southern California. Geochilll. CowlOckilli. Acta, 31, 1201-1209.
Blumer, M. and Youngblood, W. W. 1975. Polycyclic aromatic hy- Powell, T. G. and McKil'dy, M. D. 1973. A relationship between
drocarbons in soils and recent sediments. Science, 188, 53-55. ratio of pristane to phytane, crude oil composition and geological
Breger, I. A. 1960. Diagenesis of metabolites and a discussion ofthe environment in Australia. Natnl'e Phy~. Sci., 243, 37.
origin of petroleum hydrocarbons. Geuchilli. COSI/wchilii. Acta, Ranke, ~1. and Welte. D. H. 1981. The methylphenanthrene index
19, 297-308. (MP!): a maturity parameter based on aromatic hydrocarbons. In:
Brooks, .J. D., Goukl, K. and Smith,.J. W. 1969. Isoprenoid hydro- BjorflY, M. et a/. (eds.), Adt'(Ulers ill OrfitUlic (;f'ocilf'llIi.,tr!lJ!18.!.
carbons in coal and petroleum. Natute, 222, 257. Wiley, Chichester, pp. 504-512.
Cane, H. F. 1976. The origin and formation of oil shale. In: Yen, T. Radke, Nl. and Willsch, H. 1981. Maturity parameters based on
F. and Chilingarian, G. V. (eds.), Oil Shale. Developlilelits ill aromatic hydrocarbons from the Bramsche Massif area, NW-
PetmleulIl Science, 5, Chapter 3. pp. 27-60. (Elsevier.) Germany. KFA Report .Vo . .50.2681. (unpublished).
Carruthers, W. and Douglas, A. G. 1955. The constituents of high Radke, ;VI., Welte, D. H. and Willsch, H. 1982a. Geochemical study
boiling petroleum distillates. Part II. Trimethylnaphthalenes in a on a well in the Western Canada Basin: relation of the aromatic
Trinidad oil. J. Chem. Soc., 1847-1850. distribution pattern to maturity of organic matter. Geochilll.
Danyusherskaya, A ..J. 1959. The chromatography of hydrocarbons CO'lIIochilll. Acta, 46. 1-lO.
found in the bitumen of the Sangarsk coal. Nauchll Isslei/. Illst. Radke, M., Willsch, H., Leythaeuser, D. and Teichmiiller, :\1.
Geol. Arktiki, .11ini. Geol., OkhrallY, NedI'. SSSR, 98,120-129. 1982b. Aromatic components of coal: relation of distribution pat-
Day, W. C. and Erdman, J. G. 1963. Ionene, a thermal decomposi- tern to rank. Geuchilll. CuslllOcililli. Acta, 46, 1831-1848.
tion product of /l-Carotene. Science, 141, 808-809. Sakabe, T. and Sassa, R. 1952. An aromatic hydrocarbon isolated
Deroo, J. and Hodgson, G. W. 1978. Geochemical origin of organic from coal bitumen. Bul!. ('hem. Soc. Jpll, 25,353-355.
sulfur compounds; thiophene derivatives from ethylbenzene amI Seifert, W. K. and Moldowan, J. M. 1978. Application of Steranes,
sulfur. Cheili. Geol., 22, 71-78. Terpanes amI Monoaromatics to the maturation and source of
Didyk, B. M., Simoneit, B. R., Brassel, S. C. and Eglinton, G. 1978. crude oil. Geochilll. Cosllluchilil. Acta, 42,77-95.
Aromatic Hydrocarbon Variations 301

Skrigan, A. J. 1964. Preparation and utilization of fichtelite and Wakeham, S. G., Schaffner, C. and Giger, W. 1979. Perylene in
retene. Trudy. Vses. Nauchl/. Tech'll. Soveshch. Gorki 1963, sediments from the Nambian Shelf. Geochim. Cosmochim. Acta,
108-115; Chem. AbRtr., 62, 10664. 43, 1141-1144.
Stonecipher, W. D. and Turner, R. W. 1970. Rosin and rosin deri- Wakeham, S. G., Schaffner, C. and Giger, W. 1980. Polycyclic
vates. Encycl. Polymer Sci. Techrwl., 12, 139-161. aromatic hydrocarbons in recent lake sediments. II. Compounds
Streibl, M. and Herout, V. 1969. Terpenoids-especially oxygenated derived from biogenic precursors during early diagenesis.
mono-, sesqui-, di-, and triterpenes. In: Eglinton, G. and Mur- Geochim. Coslllochim. Acta, 44, 415-430.
phy, M. T. J. (eds.), Organic Geochemistry. Springer-Verlag, White, C. M. and Lee, M. L. 1980. Identification and geochemical
pp. 401-424. significance of some aromatic components of coal. Geochim. Cos-
Thomas, B. R. 1969. Kauri resins-modern and fossil. In: Eglinton, II/ochilll. Acta, 44, 1825-1832.
G. and Murphy, M. T. J. (eds.), Springer-Verlag, pp. 599-618. Yew, F. H. and Mair, B. J. 1966. Isolation and identification of C13
Tissot, B., Califet-Debyser, Y., Deroo, G. and Oudin, J. L. 1971. to C l7 alkylnaphthalenes, alkylbiphenyls, and alkyldibenzofurans
Origin and evolution of hydrocarbons in early Toarcian shales, from the 275 to 305 °C dinuclear aromatic fraction ofpetroieum.
0

Paris Basin, France. Bull. Alit. Ass. Petroleum Geologists, 55, Anal1ft. Chelll., 38, 231-237.
2177-2193. Youngblood, W. W. and Blumer, M. 1975. Polycyclic aromatic hy-
Tissot, B., Durand, B., /<;spitalie, J. and Combaz, A. 1974. Influ- drocarbons in the environment: homologous series in soils and
ence of the nature and diagenesis of organic matter in the forma- recent marine sediments. Geochilli. Cosmochill1. Acto, 39,
tion of petroleum. Bull. Alii. A.,". Petrolelull Geologists, 58, 1303-1314.
499-506. Yu, Ming-Li and Hites, R. A. 1981. Identification of organic com-
Vandenbroucke, M., Albrecht, P. and Durand, B. 1976. Geochemi- pounds on Diesel engine soot. A.llaiyt. Chell1. 53, 95-954.
cal studies on the organic matter from the Douala Basin
(Cameroon) comparison with the Early Toarcian shales, Paris,
Basin, France. Geochilli. Coslllochilil. Acta, 40, 1241-1249.
26 Quantification of steranes and
triterpanes by gas
chromatographic-mass
spectrometric analysis
B. Dahl, * G. C. Speers, * A. Steen*, N. Telnms* and J. E. Johansent
*Norsk Hydro Research Centre, Bergen, Norway
tChiron Laboratories, Trondheim, Norway

Quantitative determination of en aaa 20R cholestane has been investigated using a deuterated internal standard
and combined gas chromatography-mass spectrometry. Low resolution Voltage Selected Ion Recording (VSIR) of
a major fragment ion and parent ion have been compared. The new technique of Selected Metastable Ion
Monitoring (SMIM) has also been investigated and compared with VSIR results.
The dellterated cholestane 3D"C 27 ~~~ 20R was also used as an internal standard in comparisons with a
synthesized triterpane, 30,29,22 Trisnor 17~(H)-Hopane (T",). Improvements should be made in quantitative
GC-MS before applying statistical techniques for interpretation.

INTRODUCTION a particularly accurate method. A much better proce-


dure for estimating absolute concentrations of various
The use of steranes and triterpanes in petroleum biomarkers has recently been outlined by Rullkotter
geochemistry is now well established (e.g. Mackenzie, et al. (1984) using a commercially available alkyl-
1984; Philp, 1982). By determining the composition, perhydroanthracene as an internal standard.
isomerization, rearrangement, aromatization and Use of an internal standard for quantification will
degradation of these compounds, it is possible to compensate for variations in sample preparation tech-
obtain information on their environment of deposition, niques, e.g. losses during separation procedures, as
maturation and migration history, biodegradation and well as for instrument response. Results are clearly
stage of thermal decomposition. dependent upon the similarity in behaviour between
Early methods for the separation and analysis of the internal standard and the compound, or com-
these compounds were complex and time-consuming pounds, being determined. By analogy with the tech-
(Speers and Whitehead, 1969) and tended to rely upon nique of isotope dilution analysis, the best internal
either gas chromatography (Pym and Whitehead, standard would be a stable isotope-labelled version of
1975) or mass spectrometry for determination of the the compound being determined, followed by the use
various components. Improvements in equipment and of SIR for analysis.
techniques resulted in simplification of the analytical A small number of deuterium-labelled biomarkers
method and the widespread acceptance of the use of have recently been produced by Chiron Labora-
combined gas chromatography-mass spectrometry tories in Trondheim, including cholestane,
for the characterisation of these compounds (Kimble et Cz7 5a(H),14a(H),17a(H) 20R (C 27 aaa 20R) with two
ai, 1974; Wardroper et aI., 1977). deuterium atoms in the 3-position (see Fig. 1). The
As part of a programme studying the systematic mass spectra of C27 aaa 20R and 3D 2-C 27 aaa 20R are
variations of biomarkers in a number of sedimentary shown in Fig. 2.
sequences, it was necessary to examine existing As an initial step in establishing an analytical
methods for the quantitative determination of method, it was necessary to construct a calibration
steranes and triterpanes. Relative concentrations are curve for the determination of cholestane, and to
usually determined by integrating peak areas of mass examine any factors influencing the results. An alkane
fragmentograms (SIR traces), but absolute concentra- fraction from the natural bitumen Wurtzilite was then
tions are only infrequently determined. Some indica- used as an 'unknown' sample of cholestane as it con-
tion of concentrations can be obtained using a combi- tained a relatively high concentration of this com-
nation of Mass Spectrometric Group Type Analysis pound.
(MSGT A) and mass fragmentograms (Seifert and Quantification was achieved by adding a fixed
Moldowan, 1978), but MSGTA is not considered to be amount of the internal standard to the sample and
Pe/miellill Geochemistry ill Exploration of'the Nonvegian She/( © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 303-307.
304 Petroleum Geochemistry in Exploration of the Nmwegian Shelf

(SMIM). The latter method is based upon metastable


transitions occurring in the first field-free region of
the instrument. To be sure of comparable results, all
analyses were at the same sampling rate.

RESULTS AND DISCUSSION


(27 ScC(H), 14::dH), 17::dHl -
As noted previously, a calibration curve for
20 R - CHOLESTANE
C227 cxcxcx 20R was first prepared using varying concen-
= C27 UUQ 20R
trations of cholestane and a fixed amount of the deuter-
ated cholestane internal standard (107.4 pg) and
measuring the ratios of the appropriate ion intensities.
Detector linearity was checked over the range
10-550 pg. Quantification and repeatability checks
were carried out using a sequence of six samples of the
Fig. 1. Molecular structure of C27 crcrcr 20R and the deuter- SAC fraction from the Wurtzilite. The contribution of
ated internal standard, 3-Dz-C 27 crcrcr 20R. the isotopic signal to the internal standard was estab-
lished by analysing cholestane standards in the
absence of the deuterated internal standard. The
lee l 27 respective contributions were found to be 16.9% of the
m/z 217 fragment ion and 0.9% of the m/z 372 parent
50
I (27 MCl 20 R
ion.
Figure 3 shows that the calibration curves for VSIR
determinations of the m/z 217 fragment ion and the
m/z 372 parent ion displayed good linearity over the
concentration range investigated. In Fig. 3, the lines
are fitted to computer-determined algorithms and the
derived functions are displayed in the figure. Using
50 the m/z 217 fragment ion, the cholestane concentra-
359 tion was found to be 97.3 pg/JlI, and six consecutive
runs gave a standard deviation of 2.7%. The corres-
ponding values for the m/z 372 parent ion were,
respectively, 104.7 pg/ JlI and 2.0% standard deviation.
Fig. 2. Mass spectra of C27 crcrcr 20R and 3-Dz-C 27 crcrcr 20R. Deviation of the two lines in Fig. 3 could indicate
that fragmentation depends upon concentration. This
may be due to the concentration range affecting
measuring the ratios of the corrected ion intensities source vacuum and ion beam focussing. Higher con-
which are proportional to the amounts of the compo- centrations might also result in ion-molecule interac-
nents present. Both Voltage Selected Ion Recording tions similar to chemical ionisation effects. However,
(VSIR) and the new technique of Selected Metastable in the low and limited concentration ranges, the devia-
Ion Monitoring (SMIM), see Warburton and Zum- tion is more likely the result of experimental error
berge (1983), were used in these studies.

Cal1bration curve for Cholestane


EXPERIMENTAL 217 and 372 m/z I D2-Chol.

High purity samples of C27 cxcxcx 20R and I·OT

3Dz-C 27 cxcxcx 20R were provided by Chiron


Laboratories, Trondheim.
'.0*
The sample used for quantification studies was an
alkane concentrate obtained from a sample of Wurtzi-
lite from the Uinta Basin. Following exhaustive
~ 04,0-;-
extraction with dichloromethane and pentane de- o
r;;
I

asphaltening, the sample was separated into broad rr


N
3.!5-

compound types using a conventional HPLC Group o


"'- 31.0-;-
Type separation.
5 2.!5~
rl

The saturated alkane (SAC) fraction was analysed


using a combination gas chromatography (HP 5790)
double focussing magnetic mass spectrometer (VG .
:
L
2.0--

*
372 m/z Cho 1.
70E) coupled to a data handling system (VG 11/250). - - 0.01063*X +0.0192
The solvent used was iso-octane, and gas chromato- o 217 m/z Chol.
- - 0.00975*X -0.0073
graphic separation was carried out on a 25 m OV-101
bonded stationary phase capillary column. Inlet to the
mass spectrometer was via a 70 eV 6kV Electron 0.0 ---f- -t 1 -- - - ----+- -- - I
o tOO 200 JOO <GO .,.

Impact Source.
pg Cholestane inj.
Quantitative estimations were made using both
low-resolution Voltage Selected Ion Recording Fig. 3. Calibration curve for steranes. VSIR fragment ion,
(VSIR) and Selected Metastable Ion Monitoring mjz 217, and parent ion, mjz 372.
Quantification of Steranes and Triterpanes by GC -MS 305

Metastable ion from Cholestane


372-217 m/z transition .,,,
STERME TR ..... SITIOII-IO~S
ClJ+ C:2g-.217",Z

!!.!!~

!I. 0"';'"

T"""SlTIOII_IOII
1j()()·217/1i1

3.5t

3.0+
I
2.0+ I'll,. I' '.
" 1 5. . 1111S11

TAANSITlOll-IOll
"2·211 IVZ

M Data po ints
_ _ O. 00983>X -0.0391

0.0
o
-+---~~t~~--+-----+--~
soo 200 JOG ...00
f
500
--f
600
I.
Fig. 6. SMIM transitionograms of C27 , C29 and C27 + C29 •
pg Cho lestane inj.

Fig. 4. Calibration curve for steranes. SMIM metastable steranes present in the Wurtzilite SAC fraction. It
transition ion m/z 372-217. would obviously be unwise to assume that the calibra-
tion curve for C27 iXiXiX 20R could be used to obtain
resulting from subtraction of the very large isotopic quantitative values for C27 iXiXiX 208, C29 iXiXiX 20S and
contribution to the m/z 219 internal standard peak. C29 iXiXiX 20R, but determination of the response values
Figure 4 shows a calibration curve derived from for these various compounds showed them to have a
SMIM measurements of the transition m/z 372-217. better than 6% standard deviation.
The determined concentration was 103.3 pg/JlI and the Figure 6 shows the 8MIM traces for C27 and C29
excellent linearity resulted in a standard deviation of steranes, together with the sum of these 'transi-
only 1.0%. The metastable fragmentation should be tionograms'. It is interesting to note that the
very sensitive to instrumental variations and ion responses relative to C27 iXiXiX 20R are not the same for
interactions, but the observed 1% standard deviation V8IR and SMIM. The probable explanation is that
indicates exceptionally good repeatability. In SMIM two different types of reactions are being observed.
measurements, it should also be noted that resolution V8IR is detecting the ionization of molecules by elec-
is defined by the instrument conditions used for this tron impact, whereas SMIM is detecting the collapse
type of work, including open slit mode, and not by of a metastable ion. In the results shown in Fig. 6, the
deflection of a homogeneous ion beam through narrow C27 iXiXiX 208 response has increased relative to V8IR,
slits. The method is, therefore, more open to false but for C29 iXiXiX 20R and C29 iXiXiX 208 the situation is
signals and, hence, the observed repeatability is quite reversed (cf. Fig. 5). The difference between VSIR
outstanding. The sensitivity for SMIM determination and SMIM results is also illustrated in Fig. 7.
of steranes is, however, only in the range of about 1% This quantification study also included examination
of the m/z 217 base peak from low-resolution VSIR. of the response of the triterpane 30,29,22-Trisnor
Figure 5 shows a VSIR m/z 217 trace for the 17a(H)-Hopane (T,,) , also synthesized by Chiron

I YS1R.2Vl1Iz TRAce I
SHUN( T~ANSITIOil-IOllS

'27 + (28 + '29 .2UI'II!

I VS1R, 2111l1l n~c~ I

C:2,,,."2DR

-----~-~~---I~----~

Czg .M20S

Fig. 7. Relative response of SMIM transitionogram,


Fig. 5. VSIR fragmentogram, m/z 217. C27 + C28 + C29 , and VSIR fragmentogram, m/z 217.
306 Petroleum Geochemistry in Exploration of the Norwegian Shelf

"'-'lUilI
Calibration curve for TM
• tn' I11IN), 21'(81, 22_ -iIOf'4II( 191 and 370 m/z ID2-Chol .

I I
.. ,
lO,19,22-hI5_-J1.IN)-~''''''[ ¥SIR, 1~1 "-'1 IU~(
."
17.IHI.21_IHI-I!orut: • 370 m/z TM
_ _ o. 00668*X -0.0288
o 191 mlz TM

..
_ _ 0 0121S*X -0.0037

I".
III to.l8allll'
I. I •• .... . ~ !I.O
.c
~ 4.5
ru
0 4.0

I ';'SIR, 217K1l nACl

P9 TM in]

Fig. 8. VSIR fragmentograms of steranes, mlz 217, and Fig. 10. Calibration curve for triterpane (Tm). VISR frag-
triterpanes, mlz 191. ment ion, mlz 191, and parent ion, mlz 370.

Laboratories. Concentrations were varied over the CONCLUSIONS


range 13-590 pg injected. In the absence of a deuter-
ated internal standard, 3Dz-C 27 aaa 20R was used. The selectivity and sensitivity of GC-MS has resulted
Figure 8 shows the m/z 217 and m/z 191 VSIR traces in its widespread acceptance as the main tool for the
for the Wurtzilite saturate alkane fraction. The area analysis of complex organic mixtures. In organic
outlined in Fig. 8 containing Tm is shown enlarged in geochemistry it is the single, most useful technique for
Fig. 9. a wide range of analyses, especially in the biomarker
The use of an unrelated internal standard is strictly field. However, use of VSIR of a major fragment ion,
incorrect. However, this situation resulted in the e.g. m/z 217 in sterane analyses, often results in a
calibration curves illustrated in Fig. 10. The variation very complex distribution. Improved selectivity can
in line slopes for fragment and parent ions can readily be achieved by detecting parent ions, and the reduced
be explained by the different fragmentation patterns intensity of the latter can be enhanced using chemical
for CZ7 aaa 20R and Tin' Divergence of the curves is ionization techniques. SMIM has been shown to be a
more likely the result of an unsuitable internal stan- simple and repeatable method for resolution of com-
dard, but a noteworthy observation from the results plex sterane mixtures. Improved specificity in this
shown in Fig. 10 was a repeatability with less than 5% type of analysis can be obtained using the more-
standard deviation. sophisticated high resolution VSIR.
In any quantitative method it is important to use a
suitable internal standard to compensate for instru-
mental and other variations, e.g. separation/recovery
efficiencies. The MS-technique allows the use of stable
isotope labelling, and a deuterated version of the com-
pound being determined is one of the best types of
standard. The low mass number of deuterium ensures
strong similarity in behaviour of the labelled com-
pound. However, as outlined in this paper, sufficient
IVSIR,}91M1ZTRACEI
mass difference needs to be incorporated into the
molecule to minimize interference.
The SMIM technique will allow better resolution of
complex mixtures, e.g. steranes, and will improve the
application of biomarkers in organic geochemistry.
In all MS techniques, the separation and estimation
of compounds in complex mixtures requires better
quantitative methods, which are a necessary prelimi-
nary to the application of pattern recognition and mul-
tivariable statistical analysis to interpretational prob-
lems.

REFERENCES
Fig. 9. VSIR fragmentogram of 30,29,22-Trisnor-17a(H)- Kimble, B. J., Maxwell, J. R., Philp, R. P. and Eglinton, G. 1974.
Hopane (T m), mlz 191. Identification of steranes and triterpanes in geolipid extracts by
Qual1t~tlcatioll of Steranes and Triterpane.~ by GC -MS 307

high-resolution gas chromatography and mass spectrometry. Seifert, W. K. and Molclowan, J. M. 1978. Application of steranes,
Chelli. Geol., 14, 174-198. terpanes and monoaromatics to the maturation, migration and
Mackenzie, A. S. 1984. Applications of biological markers in pet- source of crude oils. Geochilll. CoslIlochilll. Acta, 42, 77-95.
roleum geochemistry. In: Brooks, J. and Welte, D. H. (eds.) Speers, G. C. and Whitehead, E. V. 1969. Crude petroleum. In:
Advances in Peiroleuill Geochemistry, Vo!. I. Academic Press, Eglinton, G. and Murphy, M. T. J. (eds.), Organic Geochellli.~try.
pp. 115-215. Springer-Verlag, pp. 638-675.
Philp, R. P. 1982. Application of computerised gas chromatog- Warburton, G. A. and Zumberge, J. E. 1983. Determination of
raphy/mass spectrometry to fossil fuel research. Spectra, 8(2/3), petroleum sterane distributions by mass spectrometry with selec-
6-31 (Finnigan MAT Pub!.). tive metastable ion monitoring. Analyl. Chem., 55, 123-126.
Pym, J. G. and Whitehead, E. V. 1975. Petroleum triterpane fing- Wardroper, A. M. K., Brooks, P. W., Humberston, M. J. and Max-
erprinting of crude oils. Analy. Cheill., 47, 1617-1620. well, J. R. 1977. Analyses of steranes and triterpanes in geolipid
Rullk6tter, J., Mackenzie, A. S., Welte, D. H., Leythauser, D. and extracts by automatic classification of mass spectra. Geochilll.
Radke, M. 1984. Quantitative gas chromatography-mass spec- Cos III ochi III. Acta, 41,499-510.
trometry analysis of geological samples. In: Advances in 01'ganic
Geochemish'y 1988 (in press).
27 Characterization of sedimentary
organic matter using nuclear
magnetic resonance and
pyrolysis techniques
Hauk SolliaJ Liv Schou, * Jostein Krane, Tore Skjetnet and Paul Leplat+
*Continental Shelf Institute (IKU), P.O. Box 1883, 7001 Trondheim,
Norway
tDepartment of Chemistry, AVH, University of Trondheim, Norway
+Petrofina, Exploration Department, Rue de la Loi 33, B-1040 Brussels,
Belgium

The aromatic hydrocarbon fractions from a series of sediment extracts have been isolated and subjected to proton
nuclear magnetic resonance spectroscopy eH
NMR). The NMR results were correlated with maturity data such as
Rock-Eval Tmax values, vitrinite reflectance and specific biomarker ratios. Evidence is presented to indicate that
the percentage of aromatic protons in the aromatic fraction (PAP) is proportional to the level of maturity of the
sediment. Some kerogens have been subjected to l3C NMR (CPjMAS) and flash pyrolysis-gas chromatography
(Py-GC) analysis to provide complementary structural information about the kerogens. The l3C NMR and
pyrolysis results are compared with maturity and other organic geochemical data.

INTRODUCTION distribution in the kerogen macromolecular matrix


(Miknis, 1982; Solli et al., 1984a). However, at present
Optical techniques are well-established, and are used pyrolysis techniques are the most commonly used
to characterize the type and evolutionary stage of methods for characterization of kerogens and their
organic matter in sediments. Vitrinite reflectance is applications in organic geochemistry and hydrocarbon
one of the most widely used parameters for measuring exploration have recently been reviewed (Larter and
source rock maturity (Dow, 1977; Alpern, 1980; Pow- Douglas, 1982; Larter, 1984; Horsfield, 1984).
ell et al., 1982). The maturation stage is also obtained The authors have recently reported on the analysis
from pyrolysis methods, for example the Rock-Eval of kerogens in a homogeneous sedimentary column
T ma , parameter, the temperature corresponding to the (Miocene, Angola) by using pyrolysis-gas chromato-
peak of kerogen pyrolysis (Espitalie et al., 1977). graphy (Py-GC), pyrolysis-mass spectrometry
Recently chemical parameters based on alterations (Py-MS) and CP/MAS l3C NMR. The type of organic
of biological markers (isomerization and aromatization material which has been depotlited is expected to be
of steranes/triterpanes) have been proposed for etlti- uniform and the extractable aromatic hydrocarbons
mation of thermal maturation in the subsurface (Mac- from this sedimentary column appear to be suitable
kenzie et al., 1980, 1981). Radke et al. (1982) have for the tltudy of the maturity-dependent changes in
introduced the use of specific aromatic ratios as the NMR properties of aromatic compounds. A
another chemical maturity parameter. Alexander GC-MS study of Middle Triassic source rocks from
et al. (1980) have used proton nuclear magnetic reso- Svalbard has been reported recently (Schou et al.,
nance CH NMR) to monitor the maturity changes of 1984) and the IH NMR analysis of the aromatic frac-
the extractable aromatic compounds and evaluated tions of these samples is also included in this study.
their potential as maturation indicators. In connection with ongoing Py-GC studietl of
NMR has also been utled in coal and kerogen studies kerogens and asphaltenes we have used l:lC NMR as a
by utling Cross Polarization/Magic Angle Spinning complementary technique in order to obtain additional
(CP/MAS) l3C NMR operating on the solid phase and structural information about these macromolecules
providing structural information about the carbon and a few examples are included in this paper.
The purpose of the present paper is to report some
preliminary results obtained by using NMR techni-
"Present address: Department of Geology, University of Geology, ques in combination with analytical pyrolysis and gas
University of Oslo, P.O. Box 1047, Blindel'll, Oslo 3, Norway. chromatography methods to study maturity changes
Petroleum Geochemishy in Exploration of the NomJegian She({© Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 309-317.
310 Petroleum Geochemistry in Exploration of the Norwegian Shelf

and to characterize sedimentary organic matter. ISIOm

However, the authors wish to emphasize that the PAP=15.0 TMS

results presented here were obtained on a few sets of


Aliphatic Proton,
samples only, and that refinement and evaluation of
the methods is currently being investigated.

EXPERIMENTAL
Sample information and details of the experimental
conditions have been described elsewhere (Solli et al.,
1984a, b; Schou et al., 1984) and are also given in the
text and figure legends.
Chem;cal SllIh {ppm)

RESULTS AND DISCUSSION


801m Art,fjc'ally MalUted
Representative gas chromatograms of the aromatic 350oCil11oh,s Allphat,cProto", TMS

fractions from a homogeneous sedimentary column PAP = 24.4

(Miocene Angola; Fu 2-series) are shown in Fig. 1.


The chromatogram of the shallow immature sample
(801 m; Ro = 0.35%) shows a high abundance of hydro-
aromatic compounds (e.g. tetrahydronaphthalenes
and aromatized steranes and triterpanes) together
with alkylated naphthalenes. The abundance of
phenanthrenes is low. This is in contrast to the GC
trace of the deeper-lying early mature sample ArollliltlC Protons

(2483 m; R 0 = 0.60%) which is dominated by alkylated


naphthalenes and phenanthrenes. This difference

ChemicalShilt(ppml

Fig. 2. 400 MHz IH-NMR spectra of typical aromatic


CS·N
fractions. Sample identification and percentage of aromatic
801m
Ro"O.3S%
protons, PAP, are indicated.

reflects the aromatisation of cycloalkyl aromatics


(hydroaromatics) and the breaking oflong alkyl chains
which take place with increasing maturation. The
chromatograms of samples from intermediate depths
(not shown here) show a gradual change between the
two extremes which are shown.
IH NMR has been used to monitor the changing
proton chemical environments caused by increased
maturation (Alexander et al., 1980). A proton NMR
spectrum of a typical aromatic fraction shows well-
separated signals from aromatic and aliphatic protons
(see Fig. 2). The NMR spectrum of the aromatic frac-
tion of an artificially matured sample (801 m heated in
24BJm
Ro =0.6% nitrogen for 18 h) is also shown in order to demons-
trate significant differences. The integral due to aroma-
tic protons (6.7-9.0 ppm) has increased relative to
that of non-aromatic protons. Changes are also
observed in the aliphatic region compared with the
CloP
unheated sample; signals from methyl and methylene
~~·I' protons adjacent (rl- and p-positions) to an aromatic
nucleus (approximately 2-3 ppm) have increased rela-
tive to other aliphatic protons, reflecting a relative
decrease in chain length of n-alkyl substituents with
increasing maturation.
The integral due to aromatic protons is expressed as
a percentage of the total integral from the spectrum,
and the percentage of aromatic protons in the aroma-
Fig. 1. Gas chromatograms of aromatic fractions (Miocene
Angola; Fu 2 series). Sample depths and maturity are tic fraction (PAP) is shown in Tables 1 and 2 for the Fu
indicated. Conditions: Varian 3700; Carrier gas H2 with inlet 2 and Svalbard samples, respectively. The PAP-
pressure 12 psi; 50 m x 0.25 mm I.D. fused silica capillary values for sediment extracts were plotted against T max
column coated with Se-54; split ratio 1: 10. Peak and vitrinite reflectance values (reflectance data were
identification: N = Naphthalene; P = phenanthrene; CII C2, not obtained for the Svalbard samples) and their rela-
etc., are alkyl substituents of that carbon number. tionships are shown in Figs. 3 and 4. The regular
NMR-Pyrolysis Characterization of Sediment 311

TABLE 1. PAP: percentage aromatic protons PAP versus Reflectance for the FU2 profile (·)with 2 artificially
matured samples ( .. ) and fora manne shale ($, R. Alexander et
of total amount of protons, Fu 2 samples at, Nature 1978)

FU1
Depth (m) Ro (%) T max tC) PAP (%) 350 DCl18h
/
801 0.35 418 11.2
1075 0.37 424 13.2
1401 0.37 426 14.1
15.0 PAP
1810 0.38 432
10 3000[;/40 h Manne shale
2483 0.60 436 20.6

Artificial maturation

801 m/300 "C/40 h 19.7


801 m/350 "C/18 h 24.4
15

TABLE 2. Percentage aromatic protons, Svalbard


samples
10

Lomtion
(number qf'samples) T max> mean (S.D.) PAP, mean (S.D.)
0.1 0.4 0.6 0.' Ro

1(5) 429 (1. 7) 12.8 (1.5)


II (6) 436 (3.3)
Fig. 4. Plot of PAP-values against vitrinite reflectance
13.2 (0.7)
III (9) 439 (2.7) 19.8 (2.6) (Ro). Sample code is indicated. Marine shale data are taken
IV (12) 450 (3.0) 17.6 (5.4) from Alexander et al. (1978).
V (4) >463 (-) 21.3 (4.4)

were plotted versus their PAP-values. These data are


presented in Fig. 5, and show reasonably good linear
increase in PAP ,vith increasing maturity is approxi- relationships in the range below the individual equilib-
mately linear and is in accordance with the results rium values of these parameters.
obtained by Alexander et al. (1980) and with the GC In the present work CPjMAS 13C NMR and Py-GC
results described above. The difference in slope of the have also been used, in order to study some oil shales
PAP versus T max and vitrinite reflectance is probably and a kerogen from the North Sea. The present and
caused by a combination of difference in organic other authors have shown that the Py-GC charac-
source material and reaction kinetics (heating rates teristics of kerogens change with increasing matura-
and catalvtic effects) in the sediments. Standard PAP tion, i. e. the abundance of aromatics and isoprenoids is
scales p;obably cannot be developed; however, a reduced and the aliphatic n-alkenejn-alkane homology
wider range of organic matter types (kerogens) in dif- becomes more dominant with increasing sample depth
ferent geological situations needs to be investigated (Solli et at., 1984a; van Graas et al., 1981). By using
before a definite conclusion can be drawn regarding CPjMAS 13C NMR it has also been observed that the
their affect on the slope ofthe curves. The same consid- relative carbon distribution in kerogens becomes more
erations are also relevant when biomarkers are used aromatic 'with increasing maturity (Solli et al., 1984b).
as a maturity parameter (Tissot, 1984). Figure 6 shows the pyrograms of kerogen concen-
The use of NMR to evaluate percentage of aromatic trates isolated from three samples of Kimmeridge
protons can be regarded as a bulk chemical aromatic Clay. Basic organic geochemical data and peak iden-
maturity parameter. In order to compare the P AP- tification are noted on the figure and in the figure
values with other chemical maturity parameters, legends. Two of the kerogens (K7 and K5, Dorset,
some GC-MS biomarker ratios for the Fu 2 samples South England) are genetically similar (input of the
same type of organic source material) with different
maturity (R 0 = 0.35% and 0.45%, respectively).
PAP versus Tmax for the FU2 profile (.) and the mean values oftype
II kerogen samples from Svalbard.(*) The pyrogram of K7 shows an n-alkenejn-alkane
homology ranging from C7 to about C27 • The abun-
PAP FU1
10

'5

./
'0
/

"0 440 460 Tmax_


,<1S
HOPANES

Fig. 3. Plot of PAP-values against T max' Sample code is


indicated. Fig. 5. Plot of PAP-values against various biomarker ratios.
312 Petroleum Geochemistry in Exploration of the Norwegian She(f

Py- GC
Kerogen
Kimmeridge Clay Formation
K7 Ringtead Bay. Dorset
S. England
Ro = 0.35%
HI = 719
Tmax = 422 0
Ow

Py- GC
Kerogen
Kimmeridge Clay Formation
K5 Kimmeridge Dorset
s. England
Ro ±, 0.45%
HI = 565
Tmax = 424 0

Py - GC
Kerogen
Kimmeridge Clay Formation
K2
Marton Yorkshire
N. England
Ro =0.35%
HI = 528
Tmax = 426 0

r'J
I

Fig. 6. Flash (600°C) pyrograms of kerogen concentrates. Sample data are indicated. Peak identification: 1 = toluene;
2= em+ p)-xylene; 3 = o-xylene; 4 = C3 -alkylbenzenes and -thiophenes and phenol; 5 = C4-alkylbenzenes and alkylphenols;
6 = C4- and C5-alkylbenzenes and naphthalene; 7 = 2-methylnaphthalene; 8 = Cwisoprenoid alkene; Pr = prist-l-ene; C 7, Cs,
etc. are n-alkene/n-alkane of that carbon number. Peaks marked with an asterisk are isoprenoId hydrocarbons.
NMR-Pyrolysis Characterization of Sediment 313

The Kimmeridge Clay kerogen K2 (Y orkshire


K7 North England) has the same maturity level as K7.
Ro = 0.35%
HI = 719
However, the pyrogram of K2 shows a higher relative
Tma. = 422 0 abundance of the n-alkenejn-alkane homology com-
pared with the pyrogram of K7, reflecting a difference
in the organic source material. This is also consistent
with the difference in Rock-Eval hydrogen index (HI)
and T max parameters.
The three Kimmeridge Clay kerogen concentrates
were subjected to solid state CPjMAS lac NMR
analysis and their spectra are shown in Fig. 7. The
200 100 ppm spectra are dominated by a resonance band in the
aliphatic region (0-50 ppm) with a maximum at
30 ppm and with only a minor band in the aromatic
region (100-180 ppm). However, some differences are
observed in the spectra; the spectrum of K7 shows a
small shoulder in the 50-100 ppm region which
K5 includes aliphatic carbons bound to oxygen and nit-
Ro = 0.45%
rogen (not carbonyl and carboxylic acid derivatives).
HI = 565
Tmox=424° This shoulder is not observed in the spectrum of the
more mature sample K5, and might be explained in
terms of loss of functional groups with increasing
maturity.
The most common type of information provided by
l3C CPjMAS studies of coals and kerogens is their
apparent aromaticities if:), the fraction of organic
carbon atoms that are part of aromatic rings (Miknis et
al., 1981). The f: values were calculated from the
integrated spectra (see Table 3) and show a slight
200 100 ppm
increase in aromaticity going from K7 to the more
mature K5. This is in accordance with the generally
accepted concept that kerogens become more aromatic
with increasing maturation (Tissot and Welte, 1978),
i.e. the relative carbon distribution in kerogens
becomes more aromatic. It is also observed that the
K2
Ro = 0.35%
NMR spectrum of K2 (same maturity as K7) shows a
HI = 528 higher aromaticity than that of K7. This is in agree-
Tma. = 426 0 ment with the Py-GC and Rock-Eval data suggesting
a difference in the organic source material for the two
kerogens.

TABLE 3. Aromaticities (j~) from


CPj MAS l3C NMR spectra
zoo 100 ppm Sample f~

Fig. 7. 50 MHz CP/MAS l3C NMR spectra of kerogen K7 0.08


concentrates. Sample data are indicated. K5 0.13
K2 0.17
GRS 0.05
HS22 0.51
HS22 (350°C) 0.61
dance of aromatic and isoprenoid compounds is rela-
tively high, and the Py-GC trace shows an unresolved
naphthenic envelope in the CJ:rC27 range. These fea-
tures are characteristic of the pyrograms of immature
marine type II kerogens. The pyrogram of K5 is By visual inspection of the spectra of K7 and K2 it
dominated by an aliphatic homology ranging from C7 appears that the spectrum of K7 possesses a broader
to C27• The pyrograms of K7 and K5 demonstrate the triangulated aliphatic resonance band than that of K2
maturity-dependent Py-GC characteristics which which shows a narrower aliphatic band with peak max-
were observed for the kerogens in a homogenous imum at 30 ppm. It might be somewhat speCUlative to
sedimentary column, i.e. the relative abundance of draw too broad conclusions based on these small spec-
aromatics and prist-1-ene is reduced and the tral differences. However, the sharp resonance band
n-alkenejn-alkane homology becomes more dominant at 30 ppm can be assigned to methylene groups (CH 2)n
with increasing maturity. It is also observed that the in long-chain normal alkanes and n-alkyl substituents
abundance of the naphthenic envelope is reduced in on aromatic rings, while the broader aliphatic band at
the more mature sample. lower and higher ppm values (relative to 30 ppm) indi-
314 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Py GC

c, K7
Kerogefi
Kimmeridge Cia V FormatIOn
Rlngstead BaV, Dorset
S.ENGlANO

Py- GC
Kerogen
GRS Green RIVer Shale
Eo&ene
USA

c"
p,

Fig. 8. Pyrograms of kerogen concentrates. Sample data are indicated. For peak identification see Fig. 6.

cates methyl branching and condensed alicylic (naph- the view that branched and cyclic hydrocarbons (iso-
thenic) structural moieties in the kerogen matrix prenoids) are preferentially generated during the early
(Hagaman et al., 1984). This observation is consistent stages of diagenesis/catagenesis (Tissot and Welte,
with the Py-GC results; the pyrogram of K2 shows a 1978).
higher relative abundance of the n-alkene/n-alkane Alginites are known to produce mainly normal hydro-
homology compared with that of K7, suggesting a carbons upon pyrolysis (Larter, 1984; Douglas et
higher content of long-chain alkanes relative to naph- aI., 1983). In this study a Green River Shale kerogen
thenic structures in the macromolecular matrix of K2 (GRS) was used as a representative for algal type I
compared with K7. The same difference is observed in kerogens. For the purpose of comparison, the pyro-
the spectra of K7 and K5; K5 shows a narrower alipha- gram of GRS is shown together with that of K7 in
tic resonance band than K7. This is in agreement with Fig. 8. The Py-GC trace of GRS is totally dominated
NMR-Pyrolysis Characterization of Sediment 315

Kl with a high abundance of aromatic hydrocarbons and


phenols (peak identification is noted in the figure
legend). The pyrogram is characteristic for terres-
trial type III kerogens deposited in a marine environ-
ment. The main distinguishing feature compared with
pyrograms of type I and II kerogens is the high abun-
dance of phenols in the pyrograms of type III
kerogens (Larter, 1984). The Py-GC trace of HS 22
also shows a high abundance of n-alkenes/n-alkanes in
the C I8 to C25 region which indicates an input of lipid-
rich plant wax material.
The pyrogram of the artificially matured sample
(350 DC/18 h) shows the same trend as for naturally
matured kerogens; a reduced relative abundance of
aromatics and prist-l-ene compared with the unheated
sample, and also a reduced abundance of phenols. The
abundance of n-alkenes/n-alkanes in the high molecu-
lar end of the pyrogram is reduced drastically com-
pared to the unheated sample.
The CP/MAS 13C spectra of HS 22 and the heated
GRS residue (see Fig. 11) show a drastically higher aroma-
HI·835
Tmax=4500 ticity than the oil shale spectra which have been dis-
cussed previously. The spectrum of HS 22 shows a
sharp resonance band at 30 ppm which is due to
methylene groups (see above) and a significant shoul-
der in the 50-100 ppm range which might be assigned
to C-O bonds in phenols and ether linkages. The spec-
trum of the heated sample shows an increased relative
aromatic carbon distribution which is caused by the
loss of the sharp component at 30 ppm, suggesting
that alkane chains have been released during the simu-
lated maturation process. This is also in agreement
with the Py-GC results which suggests a reduced
relative chain length of n-alkyl groups attached to the
heated kerogen matrix.
200 ppm
Reports have shown that the aromatic carbon con-
Fig. 9. 50 MHz CP/MAS 13C spectra of kerogen tent of oil shales is relatively unchanged following
concentrates. Sample data are indicated. retorting (Hagaman et al., 1984) supporting the
hypothesis that it is the aliphatic carbon which is
largely responsible for, and the critical factor in
determining, the quality and amount of oil which is
by the n-alkene/n-alkane homology ranging from C7 to produced during retorting and catagenesis in the sedi-
C28' with only a minor unresolved naphthenic hump. ments.
This is in contrast to the pyrogram of K7, which is Using HS 22 as an example, the Py-GC and NMR
described above. Using the same arguments as above, data show that the increased aromaticity of the heated
the pyrolysis results are confirmed by the C I3 NMR sample is caused by the loss of aliphatic structural
data (see Fig. 9). The spectrum of GRS shows a much moieties and demonstrates the high potential of this
narrower resonance band with maximum at 30 ppm combined kerogen analysis approach.
compared with that of K7 (Fig. 9). The GRS spectrum
is very similar to that of the alginite maceral (Zilm et
al., 1981). It is also relevant to note the significantly
higher Rock-Eval T max value of GRS compared with CONCLUSIONS
that of K7 (450 DC and 422 DC, respectively), which
reflects the higher activation energy required to break In the present study it has been demonstrated that
the aliphatic C-C bonds in the GRS matrix than that non-destructive NMR techniques provide information
required to break the labile mixed bond types in K7 on sedimentary organic matter which is complemen-
(Tis sot and Welte, 1978). tary to that obtained by the more commonly used gas
In connection with a model study concerning the chromatography and pyrolysis-gas chromatography
influence of thermal stress on the kerogen structure, a methods.
large number of different kerogen types were heated Proton NMR of the aromatic fraction isolated from
to different temperatures for specified time intervals sediment extracts gives information about the relative
(Solli et al., 1984a). As a final example of the com- amounts of aromatic and aliphatic protons, and it is
plementary information which is provided by Py-GC shown that the percentage of aromatic protons in the
and CP/MAS studies, a type III kerogen from the aromatic fraction (PAP) is proportional to the level of
North Sea (HS 22) was selected, and its artificial mat- maturity of the sediment.
uration residue (HS 22 heated at 350 DC for 18 h). The Solid-state CP/MAS 13C NMR represents a good
pyrogram of HS 22 (Fig. 10) shows an aliphatic complementary tool to pyrolysis-gas chromatography
n-alkene/n-alkane homology ranging from C 7 to C29 in kerogen studies, as it provides information on the
316 Petroleum Geochemistry in Exploration of the Norwegian Shelf

Py - GC
Jurassic
HS 22 N. Sea
Ro =0.45%
Tmax = 439 0

8
7

Cg e11

HS 22 Heated 350 0 CJ18 h

Fig. 10. Pyrograms of kerogen concentrates. Sample data are indicated. Peak identification: 1 = toluene; 2 = (m + p)-
xylenes; 3 = o-xylene; 4 = C:ralkylbenzenes; 5 = phenol; 6 = Cralkylbenzenes; 7 = a-cresol; 8 = (m + p)-cresol;
9 = dimethylphenols and naphthalene; 10 = C13 + methylnaphthalenes; 11 = C14 + dimethylnaphthalenes; Pr = prist-l-ene.
NMR-Pyrolysis Characterization of Sediment 317

HS22
HI" 158 REFERENCES
Alexander, R., Gray, M. D., Kagi, R. I. and Woodhouse, G. W.
1980. Proton magnetic resonance spectroscopy as a technique for
measuring the maturity of petroleum. Chem. Geol., 30, 1-14.
Alpern, B. 1980. Petrographie du kerogene. In: Durand, B. (ed.),
Kerogen. Editions Technip, Paris, pp. 339-371.
Douglas, A. G., Hall, P. B. and Solli, II. 1983. Comparative organic
geochemistry of some European oil shales. In: Miknis, F. P. and
Mackay, J. F. (eds.), ACS Symposium Series, No. 230.
Geochemistry and Chemistry of Oil Shales, pp. 59-84.
Dow, W. G. 1977. Kerogen studies and geological interpretations.
J. Geochem. Exploration, 7, 79-99.
Espitalie, J., Laporte, J. C., Madec, M., Marquis, F., Leplat, P.,
Paulet, J. and Boutefey, A. 1977. Methode rapide de character-
isation des roches meres de leur potential petrolier et de leur
200 100
degre d'evolution. Rev. Inst. Fran~ais Ntrole, 32, 23-42.
s
ppm
Hagaman, E. W., Schell F. M. and Cronauer, D. C. 1984. Oil shale
analysis by CP/MAS I C NMR spectroscopy. Fuel, 63, 915-919.
Horsfield, B. 1984. Pyrolysis studies and petroleum exploration. In:
Advances in Petroleum Geochemistry, Vol. I. Academic Press,
HS22 pp. 247-292.
Heated 350oC/18 h Larter, S. R. 1984. Application of analytical pyrolysis techniques to
kerogen characterization and fossil fuel exploration/exploitation.
In: Voorhees, K. (ed.), Analytical Pyroly.qis-Methods and
Applications. Butterworth, London, pp. 212-275.
Larter, S. R. and Douglas, A. G. 1982. Pyrolysis methods in organic
geochemistry. An overview. J. Analyt. Appl. Pyrolysis, 4, 1-19.
Mackenzie, A. S., Patience, R. L., Maxwell,J. R., Vandenbroucke,
M. and Durand, B. 1980. Molecular parameters of maturation in
the Toarcian shales, Paris Basin. France, I. Changes in the con-
figuration of acyclic isoprenoid alkanes, steranes, and triter-
panes. Geochim. Cosmochim. Acta, 44, 1709-1721.
Mackenzie, A. S., Hoffman, C. F. and Maxwell, J. R. 1981. Molecu-
lar parameters of maturation in the Toarcian shales, Paris Basin,
France. III. Changes in the aromatic steroid hydrocarbons.
Geochim. Cosmochim. Acta, 45, 1345-1355.
Miknis, F. P. 1982. NMR studies of solid fossil fuels. Magnetic
Resonance Rev., 7, 87-121.
Miknis, F. P., Sullivan, M., Bartuska, V. J. and Maciel, G. E. 1981.
200 100 ppm Cross-polarization magic-angle spinning 1oC NMR spectra of
coals of varying rank. Org. Geochem., 3, 19-28.
Fig. 11. 50 MHz CP/MAS 13NMR spectra of kerogen Powell, T. G. S., Creany, and Snowdon, L. R. 1982. Limitations of
use of organic petrographic techniques for identification of pet-
concentrates. Sample data are indicated. roleum source rocks. AAPG Bull. 66, 430-435.
Radke, M., Welte, D. H. and Willsch, H. 1982. Geochemical study
on a well in the Western Canada basin: relation of the aromatic
distribution pattern to maturity of organic matter. Geochim.
Cosmochirn. Acta, 46, 1-10.
distribution of the total amount of aliphatic and Schou, L., Mllrk, A. and BjorllY, M. 1984. Correlation of source
aromatic carbon atoms in the kerogen whereas Py-GC rocks and migrated hydrocarbons by GC-MS in the Middle
only analyses the volatile pyrolysis products. Triassic of Svalbard. In: Advances in Organic Geochemistry 1983
Both NMR techniques are still at the development (in press).
Solli, H., Graas, G. van, Leplat, P. and Krane, J. 1984a. Analysis of
stage and further refinement and evaluation of the kerogens of Miocene shales in an homogenous sedimentary col-
methods with a large number of samples is still umn. A study of maturation using flash pyrolysis techniques and
required. carbon-13 CP/MAS NMR. In: Advances in Organic Geochemistry
1983 (in press).
Solli, H., BjorllY, M. Leplat, P. and Hall, K. 1984b. Analysis of
organic matter in small rock samples using combined thermal
extraction and pyrolysis-gas chromatography. J. Analyt. Appl.
Pyrolysis (in press).
ACKNOWLEDGEMENTS Tissot, B. P. 1984. Recent advances in petroleum geochemistry
applied to hydrocarbon exploration. Bull. Am. Ass. Petroleum
The authors would like to thank the management of Geologists, 68, 545-561.
Tissot, B. P. and Welte, D. H. 1978. Petroleum Furmation and
Petrofina for permission to publish this paper, and Occurrence. Springer-Verlag, Berlin.
R. L. Skogstad for recording the CP/MAS 13C NMR van Graas, G., de Leeuw, J. W., Schenck, P. A. and Haverkamp, J.
spectra at the National NMR Laboratory, University 1981. Kerogen of Toarcian shales ofthe Paris Basin. A study of its
of Oslo. The spectra were recorded on NMR spec- maturation flash pyrolysis techniques. Geochim. Cosmochim..
trometers obtained through a grant from the Nor- Acta, 45, 2465-2474.
Zilm, K. W., Pugmire, R. J., Larter, S. R., Allan, J. and Grant, D.
wegian Research Council for Science and Humanities M. 1981. Carbon-13 CP/MAS spectroscopy of coal macerals. Fuel,
(NAVF). GO,717-722.
28 Study of the diagenesis,
catagenesis and metagenesis of
coals and dispersed organic
matter with a high-temperature
modified Rock-Eval
Paul Leplat and Jacques Paulet
Petrofina Exploration, Brussels, Belgium

A series of model coal samples ranging in rank from high volatile bituminous to anthracite was studied for the
relationship between vitrinite reflectance and Rock-Eval pyrolysis. A small modification in the electronic part of
the temperature programmer allows it to reach a temperature of750 °C. The study of the hydrocarbons and carbon
dioxide released, as well as the increase of Tmax of the S2 peak in borehole samples, shows a relationship with burial
depth. This modified Rock-Eval enables a rapid estimation of diagenesis and catagenesis, as well as metagenesis,
of dispersed organic matter in small quantities of whole rock samples. The oil and gas potential on whole rock
samples has been evaluated by a combination of thermal extraction and pyrolysis-gas chromatography.

INTRODUCTION -to reach a temperature of 750°C the electronics of


temperature programmer (PTF -7) were modified and
Rock-Eval studies on various types of coals and -to avoid damage of a special joint for closing the
mature sediment samples have provided the basis for oven and to obtain an adequate sealing of the mechani-
a rapid method for characterization of the various cal system, an improvement in the water-cooling sys-
types of organic matter and estimation of their degree tem was necessary.
of maturation. However, in the case of highly mature
(b) Preliminary tests were performed as follows:
organic matter, it is necessary to heat the sample at
higher temperatures for total and quantitative crack- -checking the temperature resistance of the oven and
ing of the kerogen by using a modified Rock-Eval its heating element (pyrotenax);
principle. -ensuring good response of the FID detector;
-checking optimal FID response (no baseline drift)
and
EXPERIMENTAL -calibrating the oven temperature programming
(25 °Clmin, see Fig. 2).
Sample description (c) the CO~ trap was replaced by an infrared anal-
With a modified Rock-Eval (Fig. 1), the pyrolysis of yser for CO 2 in series with a second infrared analyser
various coals with different maturities for CO 2 (Leplat et al., 1983).
(Ra = 0.5-4.2%) and maceral compositions was per- (d) Operating procedure was as follows.
formed at high temperatures. The same procedure A known quantity of crushed sample (20 mg is used
was also applied to 'dispersed organic matter' in rock for coals, and 100 mg for rock samples) was weighed
samples taken from a petroleum exploration well. into the crucible.
Starting at 250°C \vith a heating rate of
Instrumentation 25 °Clmin, the final temperature of 750°C is reached
after approximately 20 min. Including oven-cooling
Modified high temperature Rock-Eval
time, the complete analysis takes half an hour (see
(a) The pyrolysis device is the standard 'Rock-EvaI' Fig. 3).
working in normalized conditions which has been mod- The different analytical steps are completely auto-
ified as follows: matic and the principal ones are:
Petroleum Geochemistry in EJ'plomtian aftile NOlwegian Shelf' © Norwegian Petroleum Society (Graham & Trotman, 1985) pp. 319-326.
320 Petroleum Geochemistry in Exploration of the Norwegian Shelf

FLOWSHEET OF HIGH TEMPERATURE


MODIFIED ROCK-EVAL

/
,/

/
/

-,----- --,------,------,----
" ,n
-, ,
,
.
Fig_ 1. Operational flowsheet of the apparatus.
Fig. 3. Analytical cycle.

peak PI = free hydrocarbons, which are vaporized column (20 m x 0.25 mm LD.) coated with OV-l;
at moderate temperatures (250 "C); sample split ratio 1: 30; oven temperature prog-
peak P2 = hydrocarbon resulting from the cracking ramme from 40°C to 280°C at 4°C/min; helium car-
of kerogen (or coal); rier gas inlet pressure 8 psi.
peak P3 = Oxygenated compounds evolved as CO 2
which is desorbed at the end of the
cycle; in some cases, the CO 2 trap was
replaced by an IR analyser not shown in RESULTS AND DISCUSSION
Fig. 1;
T max maximum of peak P2. A series of coals (Table 1) with known vitrinite reflec-
tance (R 0) and volatile matter content (VM) have been
Pyrolysis-gas chromatography analysed with the modified high-temperature
The instrumentation has been described in detail (Solli Rock-Eval (see Fig. 4).
et aI., 1984). The system combines a rapid simultane- The data reported in Table 2 show an increasing
ous analysis of CS-C 15 free hydrocarbons by thermal T max with respect to coal rank up to Ra = 4.45%, and
extraction and temperature-programmed furnace its significance in terms of hydrocarbon generation.
pyrolysis-gas chromatography using focused As described by Alpern (1970), Robert (1974),
cryogenic trapping on capillary columns. The ther- Teichmiiller (1971) and Teichmiiller and Durand
mally extracted whole-rock samples or coals are (1983):
pyrolysed.
Conditions: temperature-programmed furnace
-the first coalification jump (R a 0.6°) corresponds =
to the onset of catagenetic generation of hydrocarbons
pyrolysis (320-550 at °C/min) in helium (30 ml/min) fol- in source rocks.
lowed by focused cyrogenic trapping (liquid nitrogen)
and gas chromatography on a fused-silica capillary
-the second coalification jump (Ra 1.2%) is the =
beginning of oil cracking and its transformation into
condensate and gas.

f
HYDROCARBON By plotting T max values (in °C) of coal as well as of
rock samples against vitrinite reflectance a good cor-
t----~.
ANALYSIS
Purge Sample pyrolysis relation throughout the evolutionary path is shown in
, (programmed heating)
Fig. 5.
" I
The geochemical method is simple and is applied
~
> directly on sediments without pretreatment as in the
standard Rock-Eval method (Espitalie et al., 1977).
In a short time this analytical procedure and the use of
~' Fig. 5 provides a preliminary estimate of the hydro-
~<::2_...
j-::::-= carbon potential of possible source rocks.
As an example, a sedimentary series of rock sam-
ples with depths ranging from 296 m to 2810 m were
TIME(mn) analysed with the high-temperature modified
10 20 '0 Rock-Eva!' The samples, which contain total organic
Fig. 2. Oven temperature calibration. carbon contents of less than 1%, have already gener-
Rock-Eval Study of Coals and Organic Matter 321

%
R€~U:.CTIIiITY
TABLE 2. Stages of evolution and hydrocarbon
generation for reference coals

Vitrinite Volatile
reflectance matter Tmax Evolution Hydrocarbon
(R o(%) (% d.af) (OC) stages generation

Early Biogenic

--r--
0.31 54.7 425
diagenesis gas
(0.55)- - - - -- - - - -(435)- - - --------
0.77 41.0 443
r'O,
1,38

0.75 39.0 441


0.92 35.1 444 Oil
1.04 29.6 447

JV\~ (1.10)- - - - - - - - - - - (455) - - - Catagenesis - - - - - - - - -

I
1.15 30.2 464
1.38 25.3 472 Wet

lo
1.70 20.3 492 gas
.'\~m" 1.83 12.9 529

--f--
. l!\~
I,
, \ 2.00 15.1
(2.00) - - - - - - - - - (530) - -
509
--------
''J\ 2.25 9.3 588
2.45 10.0 559
2.65 7.9 582 Dry
(3.00)- - - - - - - - - - - - - - Metagenesis -----gas
4.45 3.6 653 t
2,65

depth shows a complete chromatographic picture with


the n-alkenejn-alkane doublet homology dominating in
the range from Cs to C\9' The prominence ofn-alkanes
in the whole rock sample pyrogram can be explained in
terms of its maturity.
Fig. 4. Characterization of the degree of evolution in a By visual inspection of the pyrograms of mature to
series of coal samples. over-mature samples (see Figs. 8-10), it appears that
dominance of the aliphatic doublet homology
decreases with depth and maturation. The normalized
ated some hydrocarbons, and the T max values are in data of the n-C ll and n-C 12 n-alkane relative concent-
good agreement with the vitrinite reflectance data rations are given in Table 4. Table 5 shows the rela-
(see Table 3). tive concentrations of aromatic compounds (benzene,
Figure 6 shows the Rock-Eval traces of the rock toluene, xylenes). As can be seen in these tables, the
samples with increasing depth and the vitrinite reflec- n-C ll and n-C 12 n-alkane concentrations in the thermal
tance. It seems important to have a closer look at the extracts (Sl detailed peak) increase as they decrease
morphology of the peaks because, as shown for exam- in the pyrolysates (S2 detailed peak). The sum ofn-C ll
ple in Fig. 7, the 'hump' is due to the fact that this and n-C 12 n-alkane concentrations produced in the two
sample is a mixture of 95% of a mature (Ro = 0.60%) peaks (Sl + S2) are similar; they increase a little with
sedimentary rock (T max = 438 °C), with 5% of a high maturation (see Table 4, columns 2 and 6).
volatile bituminous coal (R 0 = 1.15%, T max = 464 °C). This trend is in accordance with results obtained on
The Rock-Eval data (see Table 3) for three rock samples from Miocene Shales (Solli et al., 1985), and
samples (out of the series of seven) are complemented suggests an apparent absence of migration of the
with temperature-programmed pyrolysis-gas n-alkanes. The concentration of the n-alkanes in the
chromatographic traces (Figs. 8, 9 and 10); as can be S2 peak decreases as it increases in the Sl peak; thus
seen, the mature sample (Ro = 0.87%) taken at 296 m the sum ofn-C ll + n-C 12 in the Sl + S2peaks remains
almost the same with increasing depth.
The relative concentrations of the aromatics
TABLE 1. Rank data for a set of reference coals decrease from mature to over-mature samples (see

Vitrinite
Organic carbon reflectance, Volatile matter TABLE 3. Stages of evolution and hydrocarbon
Coal sample (%) Ro(%) (% d.af) generation for a series of sediment samples
R1 57.8 0.31 54.7 Vitrinite
R2 81.9 0.77 41.0 Depth Tmax reflectance Evolution Hydrocarbon
R3 67.1 0.75 39.0 (m) (OC) (%) stages generation

-- :~----:=-- ----~:~- ---- -t- -----~~ ~~~)


R4 79.2 0.92 35.1
R5 76.6 1.15 30.2
R6 75.6 1.04 29.6
R7 58.0 1.38 25.3
R8 88.4 1.70 20.3 786 469 1.29 Catagenesis
R9 89.1 2.00 15.1 960 479 1.37 I Wet gas
RIO 86.7 1.83 12.9 1190 504 1.73 +
Rll
R12
90.5
84.7
2.45
2.25
10.0 - 1~~ - --~~ - -- - -- ;.~8---- --f- ---- --------
9.3
R13 91.8 2.65 7.9 - - - - - - - - - - - - - - - - - - - Metagenesis - - - - - Dry gas
R14 69.8 4.45 3.6 2810 630 3.32 t
~

~
[
~
;:!
~
o
'";>-
~.
J
~.

t"l
~
:s-
~.
o
~

~
;;:.
""
~
~

~
C/)
;>-

illQ!Q ~
• COALS

O-DISPEltSEO" O"GAHIC ....TTEIt IN SEDIMENTS

Fig. 5. Relation of physical and optical parameters to maximum temperature (T maX> in modified Rock-Eval procedure.
Rock-Eval Study oj Coals and Organic Matter 323

%
REFLECTIVITY

Fig. 8. Pyrogram (temperature-programmed) of thermally


extracted whole rock sample taken at 296 m depth. Condi-
tions: temperature-programmed furnace pyrolysis
(320-550 °C at 30 °Cjmin) in helium (30 mljrnin) followed by
focused cryogenic trapping (liquid nitrogen) and gas
chromatography on fused silica capillary column
(20 m x 0.25 mm LD.) coated with OV-1; sample split ratio
1,73
1 : 30; oven temperature programme from 40°C to 280 °C
at 4 °Cjmin; helium carrier gas inlet pressure, 8 psi. Peak
identification: B = benzene, T = toluene, X = xylenes, °7,
°8, etc. are n-alkanes, X, X, etc. are n-alkenes,
MN = methyl naphthalene, DMN = dimethyl naphthalene,
P = phenanthrene, MP = methyl phenanthrene, DMP-
= dimethyl phenanthrene, DBT = dibenzothiophene.

Fig. 6. Characterization of the degree of evolution in a


series of sediment samples.

Fig. 9. Pyrogram (temperature programmed) of thermally


extracted whole rock sample taken at a depth of 960 m.
Conditions and peak identification as in Fig. 8.
"hump"

Fig. 7. Mixture of rock sample with coal.

Table 5, column 13). The drastic change from 71 to 38


is the result of a modification of structure in the
kerogen during metagenesis over a vitrinite reflec-
tance range from Ro = 0.87 to Ro = 3.37 (Table 5,
column 2). The sample taken at 2810 m depth is prob-
ably polluted with gas-oil; indeed, the n-alkane range
from n-C 14 to n-C 20 is typical of gas-oil.
In order to compare the effects of the evolution from
'mature' to 'over-mature', coal samples of similar
maturities to the rock samples were analysed
(Table 6). Fig. 10. Pyrogram (temperature-programmed) of thermally
The characteristic 'oil' pyrogram fingerprint of coal extracted whole rock sample taken at a depth of 2810 m.
B (see Fig. 11) shows a continuous n-alkene/n-alkane Conditions and peak identification as in Fig. 8.
324 Petroleum Geochemistry in Exploration of the Norwegian Shelf

TABLE 4. n-C l1 and n-C 12 n-alkane relative concentrations


n-C l1-n-C 12 n-C l1--n-C 12 Sumo!
Vitrinite concentration concentration n-C l1-n-C 12
Depth reflectance Tmax in Sl in S2 in Sl + S2
(m) Ro (%) (oC) detailed peak detailed peak peaks

296 0.87 450 11 9 20


960 1.37 479 16 6 22
2810 3.32 630 25 1 26

TABLE 5. Aromatic compound relative concentrations

'"
c
:;:; '"
.:;l
'"
c
:;:;
.~ .~ E .~ .~ E .~ .~ E
'"
.:;l .§ ''"'"c" "
.:;l .:;l'" ~
<.l
.§ '"
.:;l "'"
''"" '"
'->
~ ]~ ~"'~ Ol.b~ '"
c ~
'"
~'III" '" ]~ E~ .~
'" ,;
<.l '" C '"

'"
<.l '"
'"
<.l~
"'~
~ ,; "'~ ~ ,; ~~ ]g
il'" '~" '"'" '"~
,;
'"'->g~'"
III ,;
III
,;
~~ '-' ~ ~~ '"
~~ III '"
"'~III

,,'"~
~
't>
III '"
"'~ g~ ~~ C'~
~~
c~ c~
'" ~'" "'~'"
! "
~'"
.2 Cf:l-
-.,
'"~~'" .ct/J '-'.!:l "'.!:l c(,):-;: c(,)~
'" 'It/J ,;

."
'-'
'~~
" '" '" '"
~~
~ ~ '&+ ""~'",; '&+ "- '"
",'~ ",'~

"'.l:i '&+ ,;
",'" c· '"
-:S :~~ ~ ",,'",;
~~ ~tl
",,'" ,; ,;
!O
,,'-- ~~ !O~ ~~ !O~ ~~ ~~ !O~ ~
~ s "' .... ~~ ~~ J3 .~
c:)
.~ 0
~O:: 0.., CCit/J "''''
CCit/J J3 .~ ~~ ~~ J3 .~ ;:l "
t/J8

296 0.87 450 2 25 27 8 16 24 4 16 20 71


960 1.37 479 3 20 23 8 9 17 9 11 20 60
2810 3.32 630 7 4 11 12 2 14 11 2 13 38

TABLE 6. Vitrinite reflectance, high-temperature modified Rock-


Eval and continuous IR analysis data
Continuous
Ir analysis
Rock-Eval CO~CO

TOe Ro T max
Sample (%) (%) Sl S2 S3 HI 01 ("C) 01 CO 2 01 CO

Coal B 73.50 0.87 6.20 138 2.65 189 4 443 7 2


Coal C 66.72 1.37 4.60 46 1.70 69 3 480 2 1

COALA

COAL B

Fig. 11. Pyrogram (temperature-programmed) of thermally


extracted coal sample B. Conditions and peak identification Fig. 12. CO 2 and CO pyrograms for coal samples A, Band
as in Fig. 8. C.
Rock-Eval Study of Coals and Organic Matter 325

TABLE 7. Maceral compositions for three coal samples and corresponding


high-temperature modified Rock-Eval and continuous IR analysis data
Continuous
IR analysis
Maceral composition Vitrinite Rock-Eval CO 2-CO
reflectance
Sample Exinite Inertinite Vitrinite Ro (%) T maX (OC) HI 01 CO 2 01 CO

Coal A 17 36 47 0.54 429 150 30 7


Coal B 10 64 26 0.87 443 189 7 2
Coal C 87 13 1.37 480 69 2 1

, T
(Fig. 12(b» to one of the generalised CO pyrograms in
Daly and Peters (1982).
Significant also in the pyrogram fingerprint of the
pyrolysate of coal B is the presence of aromatic com-
pounds such as naphthalenes, phenanthrenes and,
MN
n particularly, methyl and dimethyl naphthalenes.
, These aromatic compounds are not visible in the
~ r;r MP pyrogram of the rock sample taken at a depth of
'? " o~
296 m, where only the width at the base of the n-C J3
and n-C J4 peaks gives some indication of aromatic
compounds (see Fig. 8).
"
It must be kept in mind that vitrinite reflectance
(R 0) only gives data relating to the vitrinite maceral;
in contrast the apparent scattering of T max is thought
to be due to variations in maceral composition (see
Fig. 13. Pyrogram (temperature-programmed) of thermally Table 7). This is also indicated by the CO 2 and CO
extracted coal sample C. Conditions and peak identification pyrogram fingerprints (see Fig. 12) and the pyrogram
as in Fig. 8. of coal C (see Fig. 13), with a vitrinite reflectance of
Ro = 1.40, which has few alkanes; most of the peaks
are representative of aromatic compounds. This con-
trasts with the coal B fingerprint (Fig. 11) and the rock
sample taken at a depth of 960 m, which are quite
paraffinic. Fig. 14 shows the CO 2 and CO pyrograms
of the sediment samples taken, respectively, at depths
" \"' of 296, 960 and 2810 m (refer Table 3), showing the
effects of increasing maturity with depth (most evi-
--~ 296 m DEPTH
dent in the CO 2 trace).

CONCLUSIONS

(1) The high-temperature modified Rock-Eval


~
960m DEPTH device represents an important screening method
which is complementary to optical studies on concen-
trated as well as disseminated organic matter.
(2) The high-temperature modified Rock-Eval
/\C02 device is one way of estimating the evolutionary stage

~
of highly mature samples, compared with the time-
2810m DEPTH
consuming vitrinite reflectance analysis.
(3) The fingerprints of CO 2 and CO can be used to
Fig. 14. CO 2 and CO pyrograms for three sediment samples. distinguish quickly the general maturity levels of
organic matter.
(4) The combined thermal extraction and
distribution between n-C 7 and about n-C 30 • This coal B
pyrolysis-chromatography data are available within a
indeed has some 'oil potential' for three reasons.
period of one hour of the high-temperature Rock-Eval
(1) The n-alkane distribution has a typical 'oil-like data.
fingerprint'. The odd carbon number preference in the (5) Aromatics, if present in the pyrogram, may sup-
n-C23 range is considered to be due to a higher land ply useful information about specific input of organic
plant input. material and specific environment. This conclusion is
(2) The reflectance (Ro = 0.87) is situated within necessarily tentative at this stage and studies in this
the range of the 'oil window'. direction are now in progress in the authors'
(3) The similarity of the present CO pYrogram laboratories.
326 Petruleum Geochemi8try in Exploration of the Norwegian Shelf

M. (ed.), AdmYlces in Organic Geochemistry 1981. Wiley,


ACKNOWLEDGEMENTS pp. 613-619.
Robert, P. 1974. Analyses microscopiques des charbons ct des
We thank Mr G. Fortems, manager of Petrofina bitumes disperses dans les roches et mesure de leur pouvoir
Exploration and Mr P. Rousseau for the permission to reflecteur. Application it l'etude de la paleogeothermie des bas-
publish this paper. sins sedimentaires et de la genese des hydrocarbures. In: Tissot,
B. and Bienner, F. (eds.), Advances in Organic Geochemistry
1973. Technip, pp. 549-569.
Solli, H., BjorfJY, M., Leplat, P. and Hall, K. 1984. Analysis of
REFERENCES organic matter in small rock samples using combined thermal
extraction and pyrolysis-gas chromatography. J. Analyt. Appl.
Alpern, B. 1970. Classification petrographique des constituants PYl'Olysis. (submitted).
organiques fossiles des roches sedimentaires. Rev. Ilist. Fmn~ais Solli, H., van Grass, G., Leplat, P. and Krane, J. 1985. Analysis of
Petrole, 25(11), 1233-1266. kerogens of Miocene Shales in an homogeneous Sedimentary Col-
Daly. A. R. and Peters, K. E. 1982. Continuous detection ofpyrolv- umn. A study of maturation using fla"h pyrolysis techniques and
tic carbon monoxide: a rapid method for determining sedimenta;y carbon-13 CP/MAS NMR. In: Advances in Organic Geochemistry
organic facies. AAPG Bull., 66(12), 2672-2681. 198J. Pergamon Press (in pre"s).
Espitalie, J., Laporte, J. C., Madec, M., Marquis, F., Leplat, P., Teichmiiller, M. 1971. Application des methodes petrographiques
Paulet, .J. and Boutefeu, A. 1977. Methode rapide de caracterisa- des charbons it la prospection de l'huile et du gaz nature!. E I'd61
tion des roches meres, de leur potentiel petrolier et de leur degre Khole, 24, 69-76.
d'evolution. Rev. Inst. Fran~ais PC/mie, 32(1), 23-42. Teichmiiller, M. and Durand, B. 1983. Fluorescence microscopical
Leplat, P., Paulet, J. and Melotte, M. 1983. Continuous infrared rank studies of liptinites and vitrinites in peat and coals, and
analysis of temperature programmed pyrolysis products of comparison with results of the Rock-Eval pyrolysis. Int. J. Coal
laboratory-oxidized kerogens and borehole samples. In: BjorfJy, Geol., 2, 197-230.
29 Analysis of source rocks,
reservoir rocks and cap rocks by
combined thermal extraction
and pyrolysis-gas
chromatography
Malvin Bjor¢y*a, Hauk Solli*\ Keith Hallt and Paul Leplatt
*Continental Shelf Institute (IKU), P.O. Box 1883, 7001 Trondheim,
Norway
t(GC)2 Ltd, P.O. Box 30, Altrincham, Cheshire, UK
tPetrofina, Exploration Department, Rue de la Loi 33, B-1040 Brussels,
Belgium

A combined thermal extraction/pyrolysis-gas chromatography system has been developed for the multiple
evaluation of potential source rocks, variation of hydrocarbons within a reservoir, residual oil and the effectiveness
of a cap rock. The determination of maturity and organic richness is based on previously published methods. The
evaluation of the type of hydrocarbons a source rock is capable of producing is based on the relative abundance and
distribution of aliphatic and aromatic compounds in the pyrolysis-gas chromatography trace of the kerogen and the
gas chromatogram of the thermally extracted free hydrocarbons in the sample.
Hydrocarbon variations within a reservoir are detected by the gas chromatographic analysis of the thermal
extract. If care is taken during sampling, a complete gas chromatogram (C 6-C 3zl can be obtained by this method.
The sealing effectiveness of a cap rock and residual oil in a reservoir is evaluated by a combination of thermal
extraction and pyrolysis-gas chromatography. The distance that hydrocarbons from a reservoir penetrate into a
cap rock can be detected by comparing the thermal extract from the top of the reservoir with those from different
intervals within the cap rock, together with the use of pyrolysis-GC to evaluate whether the thermally extracted
material was produced in situ.

INTRODUCTION These techniques will normally be used on different


subsamples, and a direct comparison of the pore
The petroleum generation process, i. e. production of gases, extract and kerogen is therefore difficult.
bitumen from kerogen, can be studied by various The C 1-C 30 fraction comprises those hydrocarbons
methods. The type of kerogen is normally defined which are most mobile during the migration process,
either by microscopic or by pyrolytic methods. Bitu- and it is therefore of prime interest to have techniques
men is usually studied either by gas chromatographic that will analyse the whole range of hydrocarbons pre-
analysis of the pore gas or by solvent extraction fol- sent in a rock sample at the same time as the kerogen
lowed by liquid and gas chromatographic techniques, composition is determined. The gaseous compounds
or a combination of both. Due to evaporation losses, (CI-C~ can be analysed by headspace analysis (Hunt,
the extraction and fractionation technique restricts 1975). Heating and grinding, either under reduced
the analyses to compounds containing 15 carbon atoms pressure (Gedenk, 1966) or with the aid of an auxiliary
or more, which are normally classified as the 'C I5 + gas (Durand and Espitalie, 1972; Philippi, 1975; Le
fraction'. The analysis of pore gas normally gives Tran, 1975) has given satisfactory results in studying
information on CC C7 compounds and there is, there- the thermal mobilization of hydrocarbons from the
fore, a lack of information of the C S-C 14 compounds. rock matrix at elevated temperatures. All these tech-
niques have large sample requirements, and they will
only analyse hydrocarbons up to C l5 due to use of
"p,.esent add,.ess: Geolab Nor., Hornebergveien 5. 7000 Trondheim,
lower temperatures. Several micro-scale techniques of
Norway. combined thermal mobilization and gas chromatog-
bpresenl address: Department of Geology, University of Oslo, P.O. raphic analysis have been developed (Jonathan et al.,
Box 1047, Blindern, Oslo 3, Norway. 1975; Schaefer et al., 1978; Saint-Paul et al., 1980) as
Petruleum Geuchemistry in Explomtion of the Nom'egian Shelf © Norwegian Petroleum Society (Graham & Trotman, 19S5) pp. 327-337.
328 Petroleum Geochemistry in Exploration of the Norwegian She!f

well as analysis of the C1-C 15 hydrocarbon range in EXPERIMENTAL


the pore water by a combined thermal mobilization/
gas capillary GC system (Schaefer et aZ., 1983). Samples
The application of pyrolysis methods in organic
geochemistry and petroleum exploration has recently A number of claystone samples were used in these
been reviewed by Larter and Douglas (1980), Larter studies, spanning a range of immature to lower oil
(1984) and Horsfield (1984). Pyrolysis methods for the window maturity, as determined by vitrinite reflec-
investigation and characterization of organic matter in tance (Vassoevich et al., 1974) and T max on Rock-Eval
sediments can be broadly split into two categories: (a) pyrolysis (Espitalie et al., 1977). Various types of
bulk-flow methods in which only quantity or rate of kerogen were determined by visual kerogen examina-
evolution of the pyrolysates is monitored with little or tion (Staplin, 1969; Burgess, 1974) and Rock-Eval
no detailed separation, as used in the well-known pyrolysis. The samples were also extracted by a rapid
rotary system using dichloromethane as solvent,
IFP-Fina Rock-Eval process (Espitalie et al., 1977)
and (b) analytical pyrolysis techniques in which chromatographically separated into saturated, aroma-
tic and NSO fractions by using medium pressure liquid
kerogen or whole rock pyrolysates are characterized
in detail. Two methods that have been developed and chromatography (MPLC) (Radke et al., 1977), and the
employed in a number of laboratories are saturated fraction was analysed by capillary gas
pyrolysis-gas chromatography (Py-GC), alone, or in chromatography. The results from these routine
organic geochemical analyses are shown in Table 1.
combination with mass spectrometry (Py-GC-MS),
and pyrolysis-mass spectrometry (Py-MS) (Larter
Instrumentation
and Douglas, 1980). The success of the commercial
Rock-Eval instrument as a preliminary method for The combined thermal extraction and temperature-
determining petroleum source rock potential attests programmed furnace pyrolysis system, shown dia-
to the widespread acceptance of pyrolysis methods by grammatically in Fig. 1, comprises a sample probe,
the petroleum industry (Orr, 1981). The Rock-Eval furnace tube ",rith controls and a micro-switching val-
instrument uses the bulk-flow approach monitoring ve/ splitting system coupled to two fused-silica capillary
the amounts of thermally extracted free hydrocarbons columns. The programmed pyrolysis assembly is
(bitumen) and kerogen pyrolysates in rock samples mounted into a Varian 3700 gas chromatograph fitted
without prior detailed separation. Using the terminol- with three flame ionization detectors.
ogy associated with the Rock-Eval method, the ther- The sample probe (see Fig. 1) was constructed from
mally extracted hydrocarbons at 250-300 DC are rep- a 3.2 x 1.8 mm I.D. stainless steel tube with boat-
resented by a peak, S]) and the high-temperature shaped section formed at one end by cutting the tube
(300-550 DC) kerogen pyrolysate by the S2 peak. over a 1 em section to expose the bore of the tube. The
Martin (1977) used a thermal extraction procedure bore at both ends of the section was brazed to form a
coupled with Py-GC to evaluate the source rock sample cup. An alignment bar was brazed in a horizon-
potential of sediments. Dembicki et al. (1983) and tal plane to the sample cup section to allow correct
Whelan et aZ. (1980) used a similar technique to study alignment of the sample cup in the furnace tube. Nit-
source rocks and marine pollution. rogen is passed over the sample (30 ml/min) through a
Py-GC analysis of coal macerals and a large number gas inlet in the furnace tube.
of kerogens from different sedimentary facies have The furnace tube was made from 6.4 x 4 mm bore
been shown to produce characteristic fingerprint HPLC-grade stainless steel. The furnace assembly
pyrograms (Larter and Douglas, 1980; Douglas et al., was fabricated from an aluminium block heated by a
1983). 150 W cartridge heater. The furnace tube assembly
Solli et al. (1984) have recently reported the was mounted into a stainless steel enclosure to ensure
development and application of a pyrolysis-GC sys- a good draught-free environment. The temperature
tem which provides simultaneous analyses of the control of the furnace was achieved via a laboratory-
released hydrocarbons during thermal extraction (to constructed temperature programmer which enabled
C;j) and the pyrolysis products of the kerogen in small variable ramp rates between 5 and 150 DC/min to be
size rock samples by using cryogenic trapping on two selected. The temperature range normally used was
capillary columns (see 'Experimental' below). This 320°C for thermal extraction and from 320 DC to
paper describes the application of the system on 550 °C at 30 DC/min for temperature-programmed
selected samples to estimate (a) source rock quality, pyrolysis.
(b) generation of hydrocarbons from a source rock, Dried, finely-ground whole rock (20-30 mg) was
(c) remaining hydrocarbons in a source rock, (d) hyd- placed in the sample cup, inserted into the furnace
rocarbon variations within a reservoir and (e) the tube and held at 320 DC for about 3 min during the
effectiveness of a cap rock. All of the analyses for this thermal extraction step (S1). The furnace temperature
paper have been undertaken on whole rock samples was then increased to 550°C at 30°C/min during the
from the North Sea area and UK coast. The effect of pyrolysis step (8:0. The effluent from the furnace was
minerals in whole rock pyrolysis has been discussed directed to the required column system via a six-port
elsewhere (Larter, 1984; Solli et al., 1984) where it is microvalve (Val co) which was operated at 300 DC
suggested that, due to possible mineral-orgaLic mat- using the injector temperature control of the gas
ter interactions, the Py-GC character of kerogens in chromatograph. The two split exit lines (sample! split
whole rocks is somewhat more difficult to determine ratio = 1 : 30) were heated by a heater tape during the
than with kerogen concentrates. On the other hand, thermal desorption and pyrolysis stages and were led
whole rock pyrolysis might, in some cases, give a bet- to flame ionization detector C. The output from flame
ter indication of the type and distribution of hydrocar- ionization detector C recorded the bulk, unresolved,
bons likely to be produced under natural catagenic thermally extracted material and kerogen pyrolysis
conditions, but more work needs to be done in this products on a pen-recorder a8 two broad peaks, 8 1 and
area by studying well-documented geological models. S~ (see Fig. 1). The thermally extracted and kerogen
Analysis of Source, Reservoir and Cap Rocks 329

TABLE 1. Basic organic geochemical data on analytical samples


Kerogen Sl S2 HI 01 EOM EHC TOC (%)
Origin of sample type Vito Refi. T max (OC)
~~g) ~~g)
Sample No.
(mg He) (mg Hr)
gCorg gCorg

Kimmeridge Clay Fm. K-5 II 0.42 424 0.46 202.21 565 70 21.9 3.1 lS.l
(Dorset)
Kimmeridge Clay Fm. M-386 II 0.51 429 1.56 17.42 396 12 S1.1 44.3 4.4
(North Sea)
Lower Jurassic M-5141 III 0.37 426 0.19 33.74 284 11 51.3 11.3 I1.S
(North Sea)
Lower Jurassic M-5247 III 0.50 434 0.17 2.51 95 12 49.8 2.8 2.6
(North Sea)

Bulk-Flow
trace
,
JJl
S2 Thermal extract
Chromatogram

Kerogen Pyrogram
llllllil.lll.lL
111

k2
IIII1 t , . S1

FlO C FlO B FlO A

, PRESSURE REGULATOR
2 CARRIER T PIECE
3 SAMPLE PROBE
4 FURNACE
5 MICRO SWITCHING VALVE
6 SPLITTER UNIONS
7 HEATED TRANSFER LINE
8 LlOUID NITROGEN COLD TRAP
COLUMNS
9 GAS INLET
12
3 S?~------------------------~----~
Fig. 1. Schematic diagram of the thermal extraction and pyrolysis-gas chromatography system.

pyrolysis products were cold-trapped in matched at 4 DC/min; helium carrier gas inlet pressure 8 psi.
bonded OV-l fused silica capillary columns using a Peak identification: X = (m + p)-xylenes; Pr = pris-
thin-wall PTFE tubing Ginch O.D.) which was tane; Ph = phytane; numbers are n-alkanes and iso-
flushed with liquid nitrogen during the desorption and prenoids of that carbon number.
pyrolysis stages. GC analysis of the trapped material
was achieved by turning off the stream of liquid nit- Saturated fraction
rogen and starting the GC oven programme. GC con- Conditions: 20 m x 0.25 mm. J.D. fused-silica capillary
ditions are noted in the figure legends. The output column coated with OV-l; sample split ratio 1: 20;
from flame ionization detectors A and Band C was oven temperature-programmed from 80 DC to 280 DC
monitored on two recorders. The data from the three at 6 DC/min; hydrogen carrier gas inlet pressure 5 psi.
detectors was also acquired in raw data format into a For peak identification see above.
V.G. Laboratory Systems Minichrom Chromato-
Pyrolysis-gas chromatography
graphy data system for further manipulation.
Conditions: temperature-programmed furnace
pyrolysis (320-550 DC at 30°C/min) in helium
Gas chromatographic conditions and peak (30 mljmin) followed by focused cryogenic trapping
identifications (see Figs. 2-9) (liquid nitrogen) and gas chromatography on a fused-
Thermal extraction silica capillary column (20 m x 0.25 mm I.D.) coated
with OV-l; sample split ratio 1: 30; oven
Conditions: 20 m x 0.25 mm J.D. fused-silica capillary temperature-programmed from 40 DC to 280 DC at
column coated with OV-l; sample split ratio 1 : 30; 4 DC/min; helium carrier gas inlet pressure 8 psi. Peak
oven temperature-programmed from 40°C to 280 DC identification: T = toluene; X = (m + p )-xylenes;
330 Petroleum Geochemistry in Exploration of the Norwegian Shelf

C3-, C4- and C;;B are alkylbenzenes of that carbon are far more prominent in the thermal extracts than in
number; P = phenol; CIP = alkylphenols; N = naph- the solvent extracts for this sample (see Fig. 2(B».
thalene; CIN = methylnaphthalenes; numbers are Analysis of solvent extracts of Kimmeridge Clay sam-
n-alkenes/n-alkanes of that carbon number with the ples have shown large variations in the relative abun-
n-alkenes eluting before the corresponding n-alkanes; dances of n-alkanes and isoprenoids in samples col-
Pr = pristene. lected within short intervals (Williams and Douglas,
1980) showing these differences to be clearly source
dependent. Some of the variations found in the rela-
RESULTS AND DISCUSSIONS tive abundance of isoprenoids relative to n-alkanes for
the two extraction techniques could therefore be due
Source rocks
to inhomogeneities in the samples. However, this can-
Capillary gas chromatograms of the thermal extract, not explain fully the large differences observed, and it
C15 + fraction and pyrolysate of immature samples is therefore believed that solvent extraction will yield
containing kerogen of types II and III are shown in relatively less isoprenoids than thermal extraction.
Figs. 2 and 3 respectively. The gas chromatograms of The pyrogram (see Fig. 2(C» shows an alkene/
the thermal extracts and the saturated fraction of the alkane homology ranging from Cg to about C30 on top of
solvent extracts generally show some overlap in the an unresolved hump, together with C3-, C4- and C5-
C lf,C 25 range. The GC of the thermal extract of the alkylbenzenes and isoprenoids which elute between
immature kerogen type II sample (Kimmeridge Clay) the alkene/alkane doublets. This fingerprint is quite
is shown in Fig. 2(A). The sample produced relatively characteristic of immature type II kerogen (see,
small amounts of C I-C 8 hydrocarbons and very small Douglas et al., 1983). The Py-GC trace also shows a
amounts of n-alkanes in the C15+ range. Isoprenoids high abundance of peaks eluting between the
are very prominent in this sample with iso-C l6 as the n-alkene/n-alkane doublets. These are unsaturated/
most prominent peak. Similar results were also found saturated isoprenoids (Burnham et al., 1982) similar to
by Winters et al. (1983) in their hydrous pyrolysis the pristene described by Larter (1984) and Solli et al.
experiments on Kimmeridge Clay. The isoprenoids (1984). This is in good agreement with the thermal

x
I

nC 13 A

nC 20 ne 20
t ;~""W/o!JII',IIIIVJAWAJlJ...4.LLLLLtil ~--

Pr! Ph B
Ph IB
ie,s
nC 15
Steranes + Trlterpanes
nC20
i i Pr
ie t6 + ne 25
nC28

c Ie

20
25
Fig. 2. Gas chromatograms of thermal extract (A), satu-
rated hydrocarbons of solvent extract (B) and pyrolysate (C)
of immature kerogen type II (Kimmeridge Clay Formation, Fig. 3. As Fig. 2 of immature kerogen type III (Lower
Dorset). Jurassic, North Sea).
Analysis of Source, Reservoir and Cap Rocks 331

extract shown in Fig. 2(A) and hydrous pyrolysis of ible cause might be that the thermal extraction
the same type of material (Winters et al., 1983) which releases entrapped isoprenoids which are not
indicates an abundance of isoprenoids during the first extracted by solvents (Solli et al., 1984).
stages of maturation. It is concluded that type II The pyrogram (see Fig. 4(C» shows an alkene/
kerogen therefore generates unsaturated/saturated alkane homology ranging from C8 to Cao and the alipha-
isoprenoids during pyrolysis. The solvent extract (see tic homology is more pronounced compared with the
Fig. 2(B» shows a large abundance of biomarkers in less mature Kimmeridge type II kerogen. This differ-
the sterane/triterpane range. These are not detected ence is in agreement with previous observations of
in the thermal extract gas chromatogram, due to poss- Py-GC characteristics with increasing maturity (van
ible mineral retention effects (Larter, 1984). Graas et al., 1981; Solli et al., 1984), i.e. the abundance
The immature sample of type III kerogen is of of isoprenoids and aromatics is reduced with increas-
Early Jurassic age. The sample has a relatively high ing maturity resulting in a more pronounced aliphatic
hydrogen index (see Table 1) as determined by doublet homology. This is also consistent with the
Rock-Eval pyrolysis, but visual kerogen examination early generation of isoprenoids and aromatics during
did not show any indication of amorphous material. maturation (Burnham et al., 1982).
The gas chromatogram of the thermal extract of this A sample of approximately the same maturity but
sample is shown in Fig. 3(A). This shows the sample with kerogen type III is shown in Fig. 5. The gas
to contain a large abundance of light C I-C 4 hydrocar- chromatogram of the thermal extract shows o/m-
bons and m-/p-xylenes. The abundance of CIO and Cll xylenes to be the largest peak, with low concentra-
n-alkanes is high with a sharp decline for CI2 and tions of isoprenoids and n-alkanes from C12 upwards.
higher. Due to the large abundance of m-/p-xylenes This is similar to the less mature sample of type III
the higher hydrocarbons can hardly be detected in the kerogen (see Fig. 3) and suggests that there is hardly
thermal extract, but these can easily be seen in the any production of hydrocarbons from this type of
saturated fraction of the solvent extract (see kerogen at such a low maturity (Ro = 0.50%). This is
Fig. 2(B» which contains abundant high molecular also shown in the extractable hydrocarbons; the gas
weight n-alkanes with a high CPI value which is typi- chromatogram of the saturated fraction of the
cal of immature kerogen type III (Albrecht et al., extractable hydrocarbons is shown in Fig. 6(B). The
1976). GC trace shows a higher proportion of high molecular
The pyrogram (see Fig. 3(C» shows an alkene/
alkane homology ranging from Cs to about C30 ; how-
ever the intensity of the doublets is drastically reduced
after C 15• This is in contrast to the type II kerogen
pyrogram (see Fig. 2(C». The high abundance of
phenol, alkylbenzenes and naphthalenes is charac-
teristic of type III kerogens derived from terrestrial
material (Larter and Douglas, 1980; Douglas et al.,
1983). It must be emphasized that the pyrograms
which are discussed above were obtained from whole
rock samples. However, the results are also in general
in good agreement with Py-GC analysis of kerogen
concentrates (Larter and Douglas, 1980; Douglas
et al., 1983).
Now that the characteristics of immature samples of
kerogen types II and III have been established, the
effects of increasing maturity must also be considered.
A sample of more-mature kerogen type II from the ne,s PI' B
Kimmeridge Clay Formation in the North Sea has
Ph
been analysed and the results are shown in Fig. 4(A). ) nC 20
The increase in the maturity is easily detected in the
thermally extracted product, Fig. 4(A). The amount
of product has increased markedly, i.e. the kerogen in
this sample has produced a significant amount of
hydrocarbons even at a relatively low maturity
(Ro = 0.5%) (see Table 1). The distribution of the hyd-
rocarbons in the thermal extract is more oil-like than
was observed at a maturity of Ro = 0.4%, but distinct T
features of low maturity, e.g. high pristane/n-C 17 c
ratio, are still observed.
The gas chromatogram of the saturated hydrocar- a
9
bon fraction of the solvent extract (Fig. 4(B» shows a
distribution of the CH,C 30 range similar to the thermal
extract. The main difference is the larger abundance
of biomarkers (steranes/triterpanes) in the solvent
extract, which are released during thermal extraction,
and a higher ratio of n-alkanes to isoprenoids in the
solvent extract. Typical examples are pristane/n-
C l7 and phytane/n-C I8 • This phenomenon IS regis-
tered for all the samples examined, and there is pres- Fig. 4. As Fig. 2, of moderately mature kerogen type II
ently no adequate explanation for it. However, a poss- (Kimmeridge Clay Formation, North Sea).
332 Petroleum Geochemistry in Exploration of the Norwegian Shelf

ave~a~e con~en.t and will not give any indication of


A
VarIatIOns withm the reservoir. Analyses of extracts
~rom cor~ samples will, however, give more-detailed

r
mformatIOn on the compositional changes throughout
the ~eservoir. A solvent extraction of a core sample
provIdes good chromatographic results for the C ,+
fraction, ~ut the lighter components will be lost duri~g
evaporatIOn of the solvent. It will be shown how a
combination of thermal extraction, solvent extraction
and pyrolysis with gas chromatographic analysis can
~omplement each.ot~er in reservoir rock analysis. To
Illustrate the VarIatIOns often found within reservoir
fluids, samples from a number of different reservoir
B
zones were selected.

~I The first suite of samples represents a Middle


Jurassic reservoir from a well in the North Sea. This
reservoir covers a depth range of 90 m and consists of
high porosity, high permeability s~ndstones with
some thin layers of coal and claystone. The reservoir
produced cond~nsate during testing which, upon gas
chromatographIc analysis was found to contain
nC 30 n-alkanes up to C40 • The sample, which was almost
colourless, went solid when kept in a refrigerator. A
total of 20 samples of core chips from the reservoir
zone were analysed by the three methods described
above. All of the samples showed an almost complete
absence of hydrocarbons lower than n-C I4 , as seen in
the gas chromatograms of the thermal extracts (Figs.
6(A), (B), (E) and (F». The reservoir is in a high
pressure zone and the cores were reported to 'bleed'
hydrocarbons while being handled on the rig. Only
small differences were found in the hydrocarbons
d~te~ted .by thermal extraction, mainly in the n-alkane
distrIbutlOn from the C 15-C25 region, as illustrated in
Figs. 6(A), (B), (E) and (F). Some of the samples show
a cpr (Cl'rC2J greater than 1.0, while others are less
than 1.0. This feature was found repeatedly in anal-
yses of s~~~les from this reservoir which have good
reproducIbIllty. The samples show a decrease in the
cpr ~th increasing depth which probably represents
a fractIOnation effect within the reservoir. Similar but
Fig. 5. As Fig. 2, of moderately mature kerogen type III not so evident variations in n-alkane distributions are
(Lower Jurassic, North Sea). also noted in the gas chromatograms of the saturated
h~drocarbon fraction of the solvent extracts (see
FIgS. 6\C) , (D?, (G) and (H». The samples from this
reserVOIr sectlOn were also pyrolysed, but produced
weight n-alkanes with a high CPI value which is typi- no pyr~lysate, thus showing that all of the organic
cal of kerogen type III (Albrecht et al., 1976). The matter m these samples is mobilized at the tempera-
main fractions of hydrocarbons in this chromatogram ture used for thermal extraction.
are of a higher molecular weight than those detected
A second suite of reservoir samples was chosen
in the thermal extract, which is dominated by C1-C 15 from another Middle Jurassic section in the North
hydrocarbons which are lost during solvent extrac-
Sea. Core chips covering an interval of 40 m were
tion. This sample also shows a high pristane/phytane
analysed as described above. The gas chromatograms
ratio, far higher than expected by comparison with the
of the thermal extracts show large variations in the
slightly less mature sample (see Fig. 3), and may hydrocarbon distributions above C IO for the different
indicate some type differences between the two
samples analysed. Below ClO only traces of hydrocar-
samples. bons were detected, probably due to losses during
The pyrogram (Fig. 5(C» shows an alkene/alkane
s~mpling and transport. Some of the samples, espe-
homology ranging from Cs to C30 ' together with a Cially at t~e top of the reservoir, showed C lO-C 16 to
high abundance of phenols and methyl naphthalenes. be the maJor components with a significant decline in
This is characteristic of type III kerogens. The high C17 and higher n-alkanes. This would indicate a
abundance of high molecular weight hydrocarbons condensate-type field or a sample from near a gas/oil
might indicate cuticular material- and/or mineral- contact (see Fig. 7(A». A number of the analysed
retained hydrocarbons which are released during the samples had gas chromatographic distributions for the
pyrolysis step (Solli et al., 1984). thermal extract as shown in Fig. 7(B), with large
unresolved envelopes in the range n-C24~n-C30 and
Reservoir rocks with a large abundance of high molecular weight
When fluid samples are collected from wide intervals n-alkanes. The light end of the gas chromatogram is
within a reservoir, these most often represent an similar to that in Fig. 7(A), suggesting that the free
Analysis of Source, Reservoir and Cap Rocks 333

A B
Ph
nC,51 nC 20
I Pr
nC 15 ' il~h
iC,6 nC 20

I nC 25 iC,6

IIII~~ '?"
,

~t
nC lO
~1~C30 ~.
nC,O
I ,I
tI
~~-! -'--

C D
Pr
Pr
Ph nC 20 nC,,5 IPhrnC 20

nC'5
• I ' Ii ;nC 25
,C 25 iC,6
III
iC'l ,-nC 30 1\
I .-nC30

JillIII(nc 35nC 40 II1I 111(C 35


III
nC 40
~ I --1..llI

E Pr Ph F
Pr
Ph nC'5 i i (C 20
/nC 20 !
I
nC,
,

iC,6 iC,6
/nC 25

~
nC 25
\
1.1 X nC
I '0 1 Irl ll '?- L ~_
X nC
I. i
lO 1 llJiL~,
G H
P Ph ,/nC 20
Pr Ph
..--nC 20 nC '5
nC 25
nC'51 I
!
nC 25 II
I
iC,s
nC 30

~'I ,nC 30
I
I

,II
Ill -4140
nC 35
C35
IIIr
~
11(1 nC 40
+ ~
Fig. 6. Gas chromatograms of thermal extracts «A), (B), (E) and (F)) and saturated hydrocarbon fraction of solvent extracts
(G) and (H)) of four different intervals in a Middle Jurassic sandstone reservoir, North Sea,
((e), (D),

hydrocarbons detected from thermal extraction of middle part of the reservoir shows a similar large
sample B may be a mixture of those present in sample unresolved envelope as seen in the thermal extract of
A and another fraction with components of much the sample (see Figs. 7(B) and (H)). The samples from
higher molecular weight. The third type of gas the lower part of the reservoir have a pattern for the
chromatogram for the thermal extracts is shown in pyrolysate similar to those from the upper part of the
Fig. 7(C). This shows the normal distribution for a core (see Figs. 7(G) and (I)). Solvent extraction with
mature oil with a slight even/odd n-alkane predomi- liquid chromatographic separation showed that the
nance. samples from the upper and lower part of the reser-
The gas chromatograms of the saturated hydrocar- voir contained 9.1% of non-hydrocarbons, whilst the
bon fraction of the solvent extracts are almost identi- samples from the middle part of the reservoir con-
cal for the various samples from the reservoir (see tained 35-40% non-hydrocarbons. Thermal evapora-
Figs. 7(D), (E) and (F)). The gas chromatograms of tion with subsequent pyrolysis will release this heavy
the pyrolysates show significant differences. The material and generate new hydrocarbons. This
samples from the top of the reservoir show some very residue probably represents asphalthenes or heavy oil
light compounds together with very small amounts of which normally were not produced, but may subse-
the alkane/ alkene doublets and an unresolved en- quently clog the pores of the reservoir rock.
velope of high molecular weight n-alkanes. These are, The above technique was also tried on samples close
however, small and can only be detected at the highest to an oil/water contact, which were taken at 0.5 m
sensitivity. The pyrolysate of the sample from the intervals from the same reservoir as those shown in
~

A B c

nC'5 C
nC,o A- n ~
Phi nC 25 ~
~
• nC 30 If
I~lliJj ;i
c;':)

o ;c 13- iC 1. iC 15 E iC13 .....


~;C FI ""c
/ iC '~"'
;i
1;;-

00,,111111 J
C ;;-
nc,o!iCl~ r-}II·~'~C
11 ~ III,
~ I nC 25
~~.
j
1, ,I r~I~; ;':C O oC. ~
nC lS ~
I a5-
;:!
G H ~
~
iT
~
8 j
I Xg 1'>-
;:!
Cr.;
, I • q
i • l l II .qU
~t}3 2~____ ~ I"
~ !' ~
~
.~
Fig. 7. Gas chromatograms of thermal extracts «A)-(C», saturated hydrocarbon fraction of solvent extracts «D)-(F» and pyrolysates
«G)-(I)) from a Middle Jurassic sandstone reservoir, North Sea_ (A),(D),(G): Top of reservoir; (B),(E),(H): middle of reservoir,
(C),(F),(I): lower part of reservoir_
Analysis of Source, Reservoir and Cap Rocks 335

A D

B E

x c F

Fig. 8. Gas chromatograms of thermal extracts «A)-(C» and saturated hydrocarbon fraction of solvent extracts «D)-(F»
from a Middle Jurassic sandstone reservoir, North Sea. (A),(D): 65.5 m below top of reservoir; oil zone; (B),(E): 66.0 m below
top of reservoir; oil/water contact; (C),(F): 66.5 m below top of reservoir; water zone.

Fig. 6. The gas chromatograms of the thermal extract have been dissolved in the water phase. The saturated
and the saturated hydrocarbon fraction of the solvent components will only be a minor fraction in the ther-
extract of a sample 65.5 m below the top of the reser- mal extract.
voir are shown in Figs. 8(A) and (B). These have simi-
lar distributions to those found higher up in the reser-
Cap Rock
voir (see Fig. 6). The gas chromatograms of the sam-
ples from 66.0 m are shown in Figs. 8(B) and (E), The effectiveness of the caprock for a reservoir is an
while those from 66.5 m are shown in Figs. 8(C) and important factor in exploration. This can be investi-
(F). gated in terms of the amount and type of hydrocar-
The samples from 66.0 m (Figs. 8(B) and (E» have a bons found within a caprock and how this varies with
distribution of heavier hydrocarbons similar to those distance from the top of the underlying reservoir. An
found higher in the reservoir. This is particularly seen example is given in Figs. 9(A) and (D), where
in the saturated hydrocarbon fraction of the solvent Fig. 9(A) represents the thermal extract from a cap-
extract (see Fig. 8(E». The light hydrocarbons are rock sample 5 m above the top of the reservoir, and
significantly different, as seen in the gas chromato- Fig. 9(D) represents a sample 10 m above the top.
gram of the thermal extract, Fig. 8(B). There is a sig- The gas chromatogram of the thermal extract of a
nificant loss of C ta-C 22 n-alkanes while there is some sample from the top of the reservoir is shown in
increase in lighter hydrocarbons compared with Fig. 6(A). These examples clearly show how the hydro-
higher in the reservoir. The latter could be due to carbons from the reservoirs have penetrated some-
slightly different permeability in the reservoir rock, what into the cap rock, but they cannot be detected
while the former effect could be due to the sample 10 m above the reservoir. The kerogens in the two
being just below the oil/water contact (assuming that cap rock samples are almost identical. The thermal
the oil/water contact as reported from logs is slightly extract shown in Fig. 9(D) represents in situ hydro-
inaccurate). The sample from 65.5 m below the top of carbons related to the kerogen, while those shown
the reservoir has a completely different character in Fig. 9(A) represent a mixture of in situ hydrocar-
from that found higher in the reservoir. Figures 8(E) bons and those migrated in from the reservoir rock.
and (F) clearly show that this sample is below the Gas chromatograms of the saturated hydrocarbon
oil/water contact. The main reason for the large dif- fraction of the solvent extract of the two cap rock sam-
ference between the thermal extract (see Fig. 8(C» ples are shown in Figs. 9(B), and (C). These show the
and the saturated fraction of the solvent extract (see same trend as seen in the thermal extracts.
Fig. 8(F», is that the thermal extract contains mainly The chromatograms of the pyrolysates indicate
aromatic hydrocarbons and NSO compounds which some differences, particularly in the high molecular
336 Petroleum Geochemistry in Exploration of the Norwegian Shelf

A I IX o
nC,.

X nC,.
nCro

B E

,.
iC nC"

lO'
nC"
nC"
: I

.IJ"J,c,I'1~,:
nCro
I
nC"

J]llTI~l~36
C T, T F
I'T

,i

lex C,1,N
17
10 C2 N
n

Fig. 9. Gas chromatograms of thermal extracts «A) and (D», saturated hydrocarbon fraction of solvent extracts «B) and (E»
and pyrolysates «C) and (F» of a caprock. (A)-(C): 5 m above top of reservoir; (D)-(F): 10 m above top of reservoir.

weight region where the sample containing migrated molecular weight aromatic compounds, are produced
material (see Fig. 9(C» has a large abundance of high at an earlier stage of maturity than the n-alkanes. The
molecular weight alkanes. These have probably been n-alkanes can only be detected by overloading the sys-
retained in the sample during thermal extraction, and tem. Kerogen type III has a large relative abundance
are now found in the pyrolysate. The results from the of low molecular weight aromatics than found in other
pyrolysates are in good agreement with what was kerogen types. Variations in the thermal extract, i.e.
found in the thermal extracts and solvent extracts, if the hydrocarbons already produced from the kerogen,
the migrated hydrocarbons found in the sample closest and in the pyrolysate, i.e. the hydrocarbons which the
to the reservoir are excluded. kerogen is still capable of generating at increased
maturity, are clearly illustrated. With increasing
maturity the percentage of higher molecular weight
CONCLUSIONS compounds, especially n-alkanes, increases for all
types of kerogen. At higher maturities, an increase in
The combination of thermal evaporation and pyrolysis light and medium molecular weight relative to the
gas chromatography is a rapid method for screening high molecular weight hydrocarbons is observed. The
both source rocks and reservoir rocks. This paper has pyrolysates exhibit similar changes. The amount of
shown how, with as little as 20-30 mg of rock, the high molecular weight hydrocarbons in the pyroly-
amount and type of hydrocarbons present in the rock, sates decreases with increasing maturity, clearly
the type of hydrocarbons which can be generated from showing that these are U'ansferred from the kerogen.
the rock and the maturity of the rock can be evalu- In analyses of core samples from a reservoir, varia-
ated. The analyses will also show if the hydrocarbons tions in the reservoir fluid composition are easily
present in the rock are produced in situ or if they have detected, as well as residual oils and the oil/water con-
migrated into the rock. The system is also a good tool tact. With the amount of material needed for the
for detection of contamination. analyses being so small, cuttings, sidewall cores and
The first part of the present study confirmed that core chips are all suitable for analysis which takes only
branched and cyclic paraffins together with low 1-1~ h per sample, although care must be taken in
Analysis of Source, Reservoir and Cap Rocks 337

sampling, especially of coarse material, since light Larter, S. R. and Douglas, A. G. 1980. Typing of kerogens by
hydrocarbons are easily lost. pyrolysis capillary-gc. In: Douglas, A. G. and Maxwell, J. R.
(eds.), Advances ill Organic Geochemistry 1979. Pergamon
The thermal extraction/pyrolysis-gas chromato- Press, Oxford, pp. 579-583.
graphy system may be used on its own for the types of Le Tran, K. 1975. Analyse et etude des hydrocarbures gayeux
problems described in this paper, but it is advisable occlus dans les sediments. Exemple d'application it l'exploration
that such analyses are backed up by conventional sol- petroliere. Bull. Celltre Rech. Pau, SlYP A (1975), 223 pp.
Martin, S. ,J. 1977. Thermally evolved hydrocarbons from the bitu-
vent extraction and gas chromatographic analyses, at men and kerogen constituents of whole rock. In: Campos, R. and
least for selected samples. This will give a complete Goni, J. (eds.), Advances in Organic Geochelllistry 1.975.
coverage of all the hydrocarbons in the sample. Enadimsa, :Vladrid, 677 pp.
01'1', W. L. 1988. Comments on pyrolytic hydrocarbon yields in
source-rock evaluations. In: Bjol'j;jy, M. (ed.). Advances in
REFERENCES Organic Geucheillist;·y 1981. Wiley, p. 775.
Phillipi, G. T. 1975. The deep subsurface temperature controlled
Albrecht, P., Vandenbroucke, M. and Mandengue, M. 1976. origin of the gaseous and gasoline-range hydrocarbons of pe-
Geochemical studies on the organic matter from the Douala basin troleum. Geochim. Cosmochim. Acta, 39, 1353-1373.
(Cameroon); 1. Evolution of the extractable organic matter and Radke, M., Sittard, H. G. and Welte, D. H. 1971:1. Removal of solu-
the formation of petroleum. Geochilll. COSl/IO-chilll. Acta, 40, ble organic matter from rock samples with a flow through extrac-
791-799. tion cell. Allal!!t. Cheill., 50, 663-665.
Burgess, J. D. 1974. Microscopic examination of kerogen (dispersed Saint-Paul. C., Monin, J. C. and Durand, B. 1980. Methode de
organic matter) in petroleum exploration. Geol. Soc. Alii. Spec. caracterisation rapide des hydrocarbures de C 1 it C:15 conte nus
Paper 1.5.1, pp. 19-30. dans les roches sedimentaires et dans les huiles. Rev. Inst.
Burnham, A. K., Clarkson, J. B., Singleton, M. F., Wong. C. M. Fmll(;ais Petrole 35(6), 1065-1078.
and Crawford, R. W. 19i12. Biological markers from Green River Schaefer, R. G. and Poock, H. 1983. Thermal mobilization of hyd-
kerogen decomposition. Geoch;lIl. CO,sIlIOChilll. Acta, 46, rocarbons from small-size rock samples. Application in petroleum
1243-1251. geochemistry. Chl'Olliatographica, 16,257-260.
Dembicki, H., .Jr, Horsfield, B. and Ho, T. T. Y. 1983. Source rock Schaefer, R. G., Weiner, B. and Leythauser, D. 1978. Determina-
evaluation by pyrolysis-gas chromatography. Bull. Alii. A,%. tion of sub nanograms per gram quantities of light hydrocarbons
Petrole/lill Geolagistii, 67, 1094-1103. (C~-C9) in rock samples by hydrogen stripping in the flow system
Dougla;;, A. G., Hall, P. B. and Solli, H. 1988. Comparative organic of a capillary gas chromatograph. Analyt. Cheill., 50, 1R48-1854.
geochemistry of some European oil shales. In: Miknis, F. P. and Solli, H., Bjor0Y, M., Leplat, P. and Hall, K. 1984. Analysis of
Mackay, J. F. (eds.), ACS SylliposiulI/ Serie" No. 2.10, organic matter in small rock samples using combined thermal
Geochemistr!! and Chrlnistry o(Oil Shales, pp. 59-84. extraction and pyrolysis-gas chromatography. J. Allal!!t. Appl.
Durand, B. and Espitalie, J. 1972. Formation et evaluation des Pymlysis (in press).
hydrocarbones de C 1 it C 15 et des gas permanents dans les ang'iles Staplin, F. L. 1969. Sedimentary organic matter, organic
clu toarcian du bassin de Paris. In: \'on Gaertner, H. and Weiner, metamorphism and oil and gas OCCUl'l'ence. Bull. Can. Petrol.
H. (eds.), A.dvances in Orqanic Geochemistry 1971. Pergamon Geol., 17(1),47-66.
Press, Oxford, pp. 455-468. Van Graas, G., de Leeuw, J. W. and Schenck, P. A. 19110. Analysis
Espitalie, J., Laporte. J. C., Madec, M., Marquis, F., Leplat, P., of coals of different rank by Curie-point pyrolysis-mass spec-
Paulet, J. aIHI Boutefeu, A. 1977. Methode rapide de character- trometry. In: Douglas, A. G. and Maxwell, J. R. (eds.), Advances
isation des roches meres, et de leur potential petrolier et de leur in Organic Geochemish'Y 1!!79. Pergamon Press, Oxford,
degre d'evaluation. Rev. inst. Fmnr;ais Petmle, 32, 23-42. pp. 485-494.
Gedenk, R. 1966. Auf geochemischen Beurteilung und Genese von Vassoevich, N. B., Akl'amkhodaghaev, A. M. and Geodekyan, A.
Kohlenwasseransammelungen. Errlol. Kohle, Erdgas Petl'O- A. 1974. Principal zone of oil formation. In: Tissot, B. and Bien-
chell!., 19, 720-727. ner, F. (eels.), Advances in Organic Geochemistry 197';. Technip,
Horsfield, B. 1984. Pyrolysis studies and petroleum exploration. In: Paris, pp. 809-814.
lidvauees ill Petmlell.lII Gmchmnistry, Vol. I. Academic Press, Whelan, J. K., Hunt, J. M. and Hue, A. Y. 191:10. Applications of
pp. 247-292. thermal distillation-pyrolysis to petroleum source rock studies
Hunt, J. M. 1975. Origin of gasoline range alkanes in the deep sea. and marine pollution . .!. Analyt. Appl. Pyrolys;", 2, 79-96.
Natnre, 254. 411-413. Williams, P. F. W. and Douglas, A. G. 1980. A preliminary organic
.Jonathan, D., I'Hote, G. and clu Rouchet, J. 1975. Analyse geochemical investigation of the Kimmeridgian oil shales. In:
geochimique des hydrocarbures h\gers par thermoevaporation. Douglas, A. G. and Maxwell, J. R. (eds.), Advances in Organic
Ret·. Illst. Fmll9ais Petrole, 30, 65-88. Geochelilistry 1979. Pergamon Press, Oxford, pp. 531-546.
Larter, S. R. 1984. Application of analytical pyrolysis techniques to Winters, J. C., Williams, J. A. and Lewan, M. D. 1983. A laborat-
kerogen characterization and fossil fuel exploration/exploitation. ory study of petroleum generation by hyrlrous pyrolysis. In:
In: Voorhees, K. (ed.), AWllytiml PyrolYiiis-Metlwds and Bjor0y, :VI., et 01. (eds.), Adl'ollres ill 01',qllllic Geochemistry
Applicatiolls. Butterworths, London, pp. 212-275. 1981. Wileu, Chichester, pp. 524-533.

You might also like