2015 Book PetroleumGeoscience PDF

Knut Bjørlykke Editor
Petroleum
Geoscience
From Sedimentary Environments
to Rock Physics – Second Edition
Petroleum Geoscience
.
Knut Bjørlykke
Editor
Petroleum Geoscience
From Sedimentary Environments to
Rock Physics
Second Edition
With contributions from Per Avseth, Jan Inge Faleide,

Pål T. Gabrielsen, Roy H. Gabrielsen, Nils-Martin Hanken,
Helge Hellevang, Kaare Høeg, Jens Jahren, Ståle Emil Johansen,
Ragnar Knarud, Martin Landrø, Nazmul Haque Mondol, Jenø Nagy,
Jesper Kresten Nielsen, Jan C. Rivenæs, Hans C. Rønnevik
and Peter Sørhaug
Editor
Knut Bjørlykke
Department of Geosciences
University of Oslo
Blindern, Oslo
Norway
ISBN 978-3-642-34131-1 ISBN 978-3-642-34132-8 (eBook)

DOI 10.1007/978-3-642-34132-8
Springer Heidelberg New York Dordrecht London
Library of Congress Control Number: 2014958087
# Springer-Verlag Berlin Heidelberg 2010, 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts
in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being
entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of
this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from
Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center.
Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Cover design: deblik, Berlin

Cover illustration: Masaoki Adachi
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface
The second edition of Petroleum Geoscience is an updated and corrected version of

the first edition from 2010 and there are also four new chapters.
Petroleum geology and geophysics are not well-defined academic subjects. They
include many different aspects of the Earth sciences which are used in petroleum
exploration and production.
Nearly all types of insight can in some cases be useful in petroleum exploration
and production, but there are some disciplines that are most relevant. Since
petroleum is formed and for the most part hosted in sedimentary rocks, sedimen-
tology is critical. Palaeontology is important for dating rocks, and carbonate
reservoirs may consist mostly of fossils. Structural geology and basin analysis
are also vital for reconstructing the migration and trapping of petroleum. Geo-
chemistry and petroleum chemistry are also important. About 50% of the geos-
cientists in the petroleum industry are involved with production rather than
exploration and we have added a chapter on reservoir modelling written by experts
from Statoil Norway.
Geophysical methods are essential for logging and seismic exploration, and
recently electromagnetic methods have also been more commonly used in explora-
tion and production. This is covered in a new chapter. We have also added a chapter
on CCS (Carbon Capture and Storage). The chapter on well logging has been
rewritten and is now more comprehensive.
Most universities do not offer specific courses in petroleum geology/geophysics
and only a few have Masters or PhD programmes in this field. Oil companies
therefore recruit many geologists with little training in these subjects.
In this book, we have tried to give a basic introduction to disciplines relevant to
petroleum exploration and we have also included some aspects of petroleum produc-
tion and modelling.
Since so many different disciplines are included in this book, it is clear that it has
not been possible to make in-depth treatments of all of these. This book provides a
relatively condensed and precise presentation of the basic theories and facts in each
subject and it was therefore necessary to limit the number of field examples and
cases.
We have attempted to write a book which requires only a limited background in
geology and geophysics. Some of the chapters are therefore relatively basic, but
others are more advanced and we have then included more discussion and references
to original research papers. Each chapter is written as a separate contribution and
v
vi Preface
there may be some degree of overlap between them to avoid too much cross reference
to other chapters.
Petroleum geology and geophysics are applied disciplines and practical experi-
ence is critical. A separate chapter on petroleum exploration is written by a geologist
with 40 years of experience from the Norwegian Continental shelf (Hans Rønnevik).
The Norwegian Continental Self (NCS) has been explored for nearly 50 years and
data from wells and seismic surveys is very well documented through the Norwegian
Petroleum Directorate and is accessible through their home page (www.npd.no).
They also have a very large collection of cores.
References to the original literature had to be limited because of the wide range of
disciplines. In the past, textbooks often included very extensive lists of references
which were very useful when searching for relevant literature. With the electronic
databases available now, it is easy to search for relevant references and new text-
books.
In this textbook, we have tried to bridge the gap that often seems to exist between
geophysical and geological disciplines and there is also an emphasis on sediment
compaction, fluid flow and rock physics. The skills required for a petroleum geolo-
gist have changed greatly over the years. Traditionally the main task was to identify
reservoir rocks, structures with closure and the proximity of a mature source rock.
We are running out of “the easy to find” and “easy to produce” oil and gas, and
exploration and production technology is becoming more advanced. It is now
possible to produce oil and gas from source rocks (shales) and not only what has
migrated into a reservoir rock. This has increased the world’s petroleum reserves
very significantly in recent years. In North America, shale gas and shale oil produc-
tion has increased and has lowered the gas price also internationally.
Production of unconventional oil (tar sand, oil shale) and also tight gas reservoirs
and gas shale requires a stronger background in mineralogy, chemistry and physics.
The geophysical methods have become increasingly sophisticated and it is now
often possible to detect the presence of gas and oil prior to drilling based on seismic
data. Electromagnetic methods that were primarily used in mineral prospecting are
also used to find oil. As conventional oil is becoming more scarce, more geologists
are becoming involved with exploration and production of heavy oil, oil shales and
shale gas. Utilization of these resources may be environmentally more problematic
than conventional oil and gas and tends to cause increased CO2 emission during
production.
This requires a stronger background in the chemistry and physics of petroleum
and also in mineralogy and rock mechanics (rock physics). Physical and chemical
modelling is also very important.
Even if alternative sources of energy are being developed, the world will require
fossil fuels for several decades. It is a great challenge to limit the environmental
consequences of the production and use of fossil energy.
Until sufficient low-cost renewable energy is available, geoscientists can contrib-
ute to efficient exploration and production of oil and gas, reducing pollution and CO2
emissions as much as possible. They may also engage in developing carbon storage.
Much of the theoretical basis is the same for environmental geology and petro-
leum geology.
Preface vii
Capturing and storage of CO2 from fossil fuels in the subsurface (CCS) require
very much the same skills as production of oil and gas.
We hope that this book will be of some use also for geoscientists who work
outside the petroleum industry.
.
Acknowledgments
Adrian Read ([email protected]) has been of great assistance as Text Editor in

the preparation of both the first and the second edition of this book.
Jan Petter Holm has provided new figures for this edition.
Per Arne Bjørkum (Statoil) and Antony Spencer (formerly Statoil) have been very
helpful in reviewing large parts of the first edition of this book.
Valuable reviews of single chapters have been made by Tony Dickson (Chap. 5),
Tom Andersen (Chap. 3), and Olav Walderhaug, Statoil (Chap. 4).
Masaoki Adachi, Jon Reierstad and Jan Petter Holm have made many of the
figures in the book and figures have also been prepared by Tom Erik Maast, Øyvind
Marcussen, Olav Blaich, Brit Thyberg, Delphine Croizé Delphine and Tove Midtun.
Statoil has provided funding which has helped the preparation of this book, and
we are grateful for this support.
Fugro Geoscience Division has kindly provided good seismic data from offshore
Norway and also from other parts of the world.
Oslo, Norway Knut Bjørlykke
ix
.
Contents
1 Introduction to Petroleum Geology . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Knut Bjørlykke
2 Introduction to Sedimentology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Knut Bjørlykke
3 Sedimentary Geochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Knut Bjørlykke
4 Sandstones and Sandstone Reservoirs . . . . . . . . . . . . . . . . . . . . . . . . 119
Knut Bjørlykke and Jens Jahren
5 Carbonate Sediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Nils-Martin Hanken, Knut Bjørlykke and Jesper Kresten Nielsen
6 Mudrocks, Shales, Silica Deposits and Evaporites . . . . . . . . . . . . . . . 217
Knut Bjørlykke
7 Stratigraphy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
Jenø Nagy and Knut Bjørlykke
8 Seismic Stratigraphy, Sequence Stratigraphy and Basin Analysis . . . 255
Knut Bjørlykke
9 Heat Transport in Sedimentary Basins . . . . . . . . . . . . . . . . . . . . . . . 273
Knut Bjørlykke
10 Subsurface Water and Fluid Flow in Sedimentary Basins . . . . . . . . . 279
Knut Bjørlykke
11 Introduction to Geomechanics: Stress and Strain in Sedimentary
Basins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Knut Bjørlykke, Kaare Høeg and Nazmul Haque Mondol
12 The Structure and Hydrocarbon Traps of Sedimentary Basins . . . . . 319
Roy H. Gabrielsen
13 Compaction of Sedimentary Rocks: Shales, Sandstones
and Carbonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Knut Bjørlykke
14 Source Rocks and Petroleum Geochemistry . . . . . . . . . . . . . . . . . . . 361
Knut Bjørlykke
xi
xii Contents
15 Petroleum Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373

Knut Bjørlykke
16 Well Logging: Principles, Applications and Uncertainties . . . . . . . . . . . 385
Nazmul Haque Mondol
17 Seismic Exploration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Nazmul Haque Mondol
18 Explorational Rock Physics: The Link between Geological Processes
and Geophysical Observables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Per Avseth
19 4D Seismic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
Martin Landrø
20 Interpretation of Marine CSEM and Marine MT Data for
Hydrocarbon Prospecting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
Ståle Emil Johansen and Pål T. Gabrielsen
21 Production Geology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
Knut Bjørlykke
22 Introduction to Reservoir Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
Jan C. Rivenæs, Petter Sørhaug and Ragnar Knarud
23 Unconventional Hydrocarbons: Oil Shales, Heavy Oil, Tar Sands,
Shale Oil, Shale Gas and Gas Hydrates . . . . . . . . . . . . . . . . . . . . . . . . . 581
Knut Bjørlykke
24 Carbon Capture and Storage (CCS) . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
Helge Hellevang
25 Geology of the Norwegian Continental Shelf . . . . . . . . . . . . . . . . . . . . . 603
Jan Inge Faleide, Knut Bjørlykke and Roy H. Gabrielsen
26 Exploration Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639
Hans Chr Rønnevik
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
List of Contributors
Per Avseth, Tullow Oil, Oslo, Norway

Knut Bjørlykke, Department of Geosciences, University of Oslo, Oslo, Norway
Jan Inge Faleide, Department of Geosciences, University of Oslo, Oslo, Norway
Nils-Martin Hanken, Department of Geology, UiT - The Arctic University of
Norway, Tromsø, Norway
Helge Hellevang, Department of Geosciences, University of Oslo, Oslo, Norway
Pål T. Gabrielsen, EMGS (ElectroMagnetic GeoServices ASA OSE), Trondheim,
Norway
Roy Gabrielsen, Department of Geosciences, University of Oslo, Oslo, Norway
Kaare Høeg, Department of Geosciences, University of Oslo, Oslo, Norway
Jens Jahren, Department of Geosciences, University of Oslo, Oslo, Norway
Ståle Emil Johansen, NTNU, Trondheim, Norway
Ragnar Knarud, Statoil, Stavanger, Norway
Martin Landrø, NTNU, Trondheim, Norway
Nazmul Haque Mondol, Department of Geosciences, University of Oslo, Oslo,
Norway
Norwegian Geotechnical Institute (NGI), Oslo, Norway
Jenø Nagy, Department of Geosciences, University of Oslo, Oslo, Norway
Jesper Kresten Nielsen, North Energy ASA, New Area Exploration, Alta, Norway
Jan C. Rivenæs, Statoil, Stavanger, Norway
Hans C. Rønnevik, Lundin, Oslo, Norway
Petter Sørhaug, Statoil, Stavanger, Norway
xiii
Chapter 1
Introduction to Petroleum Geology
Knut Bjørlykke
Petroleum geology comprises those geological facies in the sedimentary basins, and from that the
disciplines which are of greatest significance for the likely presence of source rocks with a high content
finding and recovery of oil and gas. Since most of the of organic matter, reservoir rocks and cap rocks. The
obvious and “easy to find” petroleum already has been distribution and geometry of potential sandstone or
discovered it is necessary to use sophisticated methods carbonate reservoirs requires detailed sedimentologi-
in the exploration of sedimentary basins. These cal models, and sequence stratigraphy has been a use-
include advanced geophysical techniques and basin ful tool in such reconstructions.
modelling. There is also much more emphasis now The biostratigraphic correlation of strata encoun-
on enhanced recovery from the producing fields. tered in exploration wells is achieved by
Petroleum technology has made great progress and micropalaeontology (including palynology), a field
many new tools and modelling programs have been developed very largely by the oil industry. Due to the
developed, both in exploration and production. small size of the samples obtained during drilling
It is however important to understand the geologi- operations one cannot rely on macrofossils; even in
cal processes which determine the distribution of dif- core samples the chance of finding good macrofossils
ferent sedimentary rocks and their physical properties. is poor. By contrast a few grams of rock from the drill
This knowledge is fundamental to being able to suc- cuttings may contain several hundred microfossils or
cessfully apply the methods now available. palynomorphs which are small remains of plants such
It is difficult to know where to start when teaching as pollen, spores etc. These also usually provide better
petroleum geology because nearly all the different stratigraphic resolution than macrofossils.
disciplines build on each other. Reservoir rocks are mostly sandstones and
This introductory chapter will provide a short and carbonates which are sufficiently porous to hold sig-
rather simple overview of some aspects of petroleum nificant amounts of petroleum. The composition and
geology to introduce the subject and the problems. properties of other rock types such as shales and salt
Most of the other chapters will then expand on what are also important.
is presented here to provide a better background in The sedimentary environments (sedimentary facies)
relevant subjects. determine the distribution of reservoir rocks and their
Since practically all petroleum occurs in sedimen- primary composition. Sediments do, however, alter
tary rocks, sedimentary geology forms one of the main their properties with increasing overburden due to dia-
foundations of petroleum geology. Sedimentological genesis during burial.
models are used to predict the location of different Diagenetic processes determine the changes in
porosity (compaction), permeability and other physi-
cal properties such as velocity, in both sandstone and
limestone reservoirs. Chemical processes controlling
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway mineral reactions are important.
e-mail: [email protected]
K. Bjørlykke (ed.), Petroleum Geoscience: From Sedimentary Environments to Rock Physics, 1

DOI 10.1007/978-3-642-34132-8_1, # Springer-Verlag Berlin Heidelberg 2015
2 K. Bjørlykke
Organic geochemistry, which includes the study of petroleum geologist should have a broad geological
organic matter in sediments and its transformation into training, preferably also from field work.
hydrocarbons, has become another vital part of petro- Geophysical well-logging methods have developed
leum geology. equally rapidly, from simple electric and radioactive
Tectonics and structural geology provide an under- logs to highly advanced logging tools which provide
standing of the subsidence, folding and uplift respon- detailed information about the sequence penetrated by
sible for the creation and dynamic history of a basin. the well. Logs provide a continuity of information
The timing of the folding and faulting that forms about the rock properties which one can seldom obtain
structural traps is very important in relation to the from exposures or core samples. This information
migration of hydrocarbons. makes it possible to interpret not only the lithological
Seismic methods have become the main tool for composition of the rocks and the variation of porosity
mapping sedimentary facies, stratigraphy, sequence and permeability, but also the depositional environ-
stratigraphy and tectonic development. Marine ment. Image logs make it possible also to detect bed-
seismics recorded from ships have become very effi- ding and fractures inside the wells.
cient and seismic lines are shot at only a few 100 m Practical petroleum geology is not only based on
spacing or less. Because of the rapid improvement in many different geological and geophysical disciplines.
the quality of seismic data processing techniques, geo- A good background in basic chemistry, physics, math-
logical interpretation of seismic data has become an ematics and computing is also required, particularly
important and expanding field. Seismic and other geo- for different types of basin modelling.
physical data are often the only information we have, A very large percentage of the geologists and
particularly for offshore exploration where drilling is geophysicists in the world are directly or indirectly
very costly. Shooting seismic lines with a close involved with the petroleum industry which recruits a
spacing allows high resolution 3D seismic imagery large percentage of the university students in relevant
to be produced for critical parts of sedimentary basins. fields.
By repeating a 3D reservoir seismic survey during The petroleum industry also funds much research
production, one can observe how the gas/oil and oil/ which is not narrow applied research, but more general
water contacts move as the reservoir is depleted. This contributions to the understanding of the Earth’s crust
is called 4D seismic because time provides the fourth and particularly sedimentary basins. Applied research
dimension. directed towards exploration and production of oil and
Geophysical measurements may include gravime- gas has often resulted in insight into more fundamental
try and magnetometry; electromagnetic methods that geological processes.
were used mostly in ore exploration have also in the Environmental geology has become a very impor-
last 10–20 years been applied to oil exploration. Elec- tant field to deal with problems not only related to the
tromagnetic methods have been used to detect petroleum industry but generally about the distribution
sediments with low resistivity due to the presence of of pollutants in rocks and sediments.
oil instead of saline water. This method requires a few A good background in chemistry is then required.
100 m of water depth and relatively shallow accumu-
lation (<2–3 km).
Seismic surveys are more expensive on land than by 1.1 A Brief Petroleum History
ship at sea because geophones have to be placed in a
grid, often on uneven and difficult land surfaces. Dril- There are many places where oil seeps out of the
ling on land, however, costs much less than from ground. Bitumen produced from such naturally occur-
offshore rigs and a much denser well spacing can be ring crude oil has been collected and used since
used during both exploration and production. ancient times, both for lighting and medicine, and by
Indirect methods of mapping rock types employing the Greeks even for warfare. In some places, for
geophysical aids are becoming increasingly important example Germany in the 1800s, small mines were
in petroleum geology, but it is still necessary to take dug to get at the oil. Before 1859 oil was also recov-
samples and examine the rocks themselves. A ered from coal for use in kerosene lamps. It was not
until Edwin Drake’s exploits in 1859 at Oil Creek near
1 Introduction to Petroleum Geology 3
Titusville in West Pennsylvania that oil was recovered The USA maintained its position as the major
in any quantity from boreholes. He drilled a well about world producer of oil and gas well into the twentieth
25 m (70 ft) deep which produced 8–10 bbl/day, a century. Americans thus became leaders in the devel-
huge production rate compared with anything earlier. opment of oil technology, which today is strongly
A few years later there were 74 wells round Oil Creek, reflected in the industry’s terminology. The US used
and the USA’s annual production had risen to half a to be a major importer of oil despite domestic produc-
million barrels. Outside the USA the calculated total tion, though this has recently increased due to shale oil
production at that time was maximum 5,000 bbl. In and gas. The need for import has therefore been
1870 production had increased tenfold, with 5 million reduced. For energy statistics: www.eia.gov.
bbl from the USA, and 538,000 bbl from other The US consumption (18-20 million barrels/day) is
countries. In southern California oil production started a very large fraction of the total world consumption
early in 1864 (in Santa Paula), but for many years oil (94 million bbl/day).
was mined by driving shafts into the oil-bearing strata It was first in the 1930–1940 period that the indus-
because it was so heavy and biodegraded that it would try became aware of the vast oil resources of the
not flow in a well. Middle East, which now account for about 60% of
In its infancy, oil exploration consisted largely of world reserves. Since then this region has dominated
looking for oil seepage at the surface and drilling in oil production.
the vicinity, which did not require much geological In the 1950s and 1960s prospecting for oil and gas
knowledge. It was then realised that oil and gas occur was extended onto the continental shelves, opening up
where layers of sedimentary rocks form domes or new reserves. Until the development of modern seis-
anticlinal structures since petroleum is less dense mic it was not possible to effectively explore deep
than water and a low permeability (seal) layer is below the seafloor in sedimentary basins offshore. As
needed to prevent the oil and gas from rising and long as there was an abundance of oil to be found in
escaping. This led to extensive geological mapping onshore basins there was little incentive to develop
of anticlines and domes visible at the surface, particu- costly drilling rigs for offshore exploration and
larly in the USA. It was also found that oil fields had a platforms for production.
tendency to lie along structural trends defined by Since the 1970s an increasingly large share of
anticlines or faults and this “rule” was used in international prospecting has taken place offshore,
prospecting. This is also often the case with salt helped by improved seismic methods. Advanced well
domes, which became important prospecting targets. log technology in particular made it possible to gain
Oil production developed rapidly up to the end of optimal information from each well. Before the devel-
the nineteenth century, and more systematic geologi- opment of powerful computers, seismic recordings
cal principles for prospecting were gradually devel- were based on analogue methods which produced
oped. The geological information which one obtains at results very far removed from the standard of modern
the surface is often not representative of the structures seismic data. Recording of seismic data from ships is
deeper down. Structures which are not visible at the much less expensive than onshore where geophones
surface could be mapped by correlation between wells have to be laid out manually. Onshore, extensive dril-
using logs and cuttings from the drilling. One method ling may be cheaper than expensive seismic mapping.
was to measure the depth of particularly characteristic Rising oil prices and new technology have made
strata through analysis of cuttings in different wells. exploration financially attractive in areas which previ-
Improved electrical measurements (logs) from wells, ously were of little interest, including in very deep
developed during the 1920s and 1930s, made the waters. High oil prices can also pay for more enhanced
whole effort much simpler because they provided hydrocarbon recovery from reservoirs. There are now
continuous vertical sections through the rocks. The relatively few sedimentary basins in the world that
first logs were simple recordings of how well rocks have not been explored and it is getting increasingly
conduct electrical currents (resistivity), and later also difficult to find new giant fields (>500 million bbl).
gamma logs recorded the gamma-radiation emitted by There is now increasing interest in heavy oil, tar
the different sedimentary rocks. sand and oil shale.
4 K. Bjørlykke
Oil shale is a source rock exposed near the surface. The organic matter from which petroleum is
If the source rock (shale) is mature it will have a derived originated through photosynthesis, i.e. storage
characteristic smell of hydrocarbons, but it may not of solar energy (Fig. 1.1).
be mature so that hydrocarbons have not been Sunlight is continuously transformed into such
generated. If the oil shale is mature much of the oil energy on Earth but only a very small proportion of
has escaped by primary migration. Since the the solar energy is preserved as organic matter and
hydrocarbons are thoroughly disseminated in the petroleum. The oil and gas which forms in sedimen-
fine-grained sediment, oil cannot be produced in the tary basins each year is thus minute in comparison
same way as from sandstone or carbonate reservoirs. with the rate of exploitation (production) and con-
The hydrocarbons can only be obtained by breaking sumption. In practice petroleum must therefore be
and crushing the shale and heating to distill off the regarded as a non-renewable resource even though
interspersed hydrocarbons. Shales can however con- some petroleum is being formed all the time.
tain gas which can be produced when there is a net- Most of the organic materials which occur in source
work of small fractures. Gas shale is expected to be an rocks for petroleum are algae, formed by photosynthe-
important source of petroleum in the years to come, sis. The zooplankton and higher organisms that are
particularly in the US. Very large amounts of fossil also represented grazed the algae and were thus indi-
fuels are stored in organic-rich mudstones or shales rectly dependent on photosynthesis too. The energy
that have not been buried deeply enough for the which we release when burning petroleum is therefore
organic matter to be converted to petroleum. In this stored solar energy. Since petroleum is derived from
case very little hydrocarbon has excaped but these organic matter, it is important to understand how and
deposits must be mined and heated to 400–500 C in where sediments with a high content of organic matter
ovens to generate petroleum (pyrolysis). are deposited.
The Tertiary Green River Shale in Colorado, Utah The total production of organic material in the
and Wyoming represents one of the largest petroleum world’s oceans is now 51010 tonnes/year. Nutrients
reservoirs in the world. This is a lake deposit, and the for this organic production are supplied by erosion of
organic matter consisted mainly of algae. rocks on land and transported into the ocean. The
Although very large quantities of petroleum can be supply of nutrients is therefore greatest in coastal
produced from oil shale, production costs are at pres- areas, particularly where sediment-laden rivers dis-
ent too high compared to conventional oil. There are charge into the sea. Plant debris is also supplied
also serious environmental problems involved in pro- directly from the land in coastal areas.
duction from oil shale, and the process requires very Biological production is greatest in the uppermost
large quantities of water, a resource which is not 20–30 m of the ocean and most of the phytoplankton
always plentiful. growth takes place in this zone. In clear water, sunlight
The oil reserves in such deposits exceed conven- penetrates much deeper than in turbid water, but in
tional oil reserves, but the expense and environmental clear water there is usually little nutrient supply. At
issues involved with production from these types of about 100–150 m depth, sunlight is too weak for
reservoirs clearly limit their exploitation. This is par- photosynthesis even in very clear water.
ticularly true of production from oil shale. Phytoplankton provides nutrition for all other
marine life in the oceans. Zooplankton feed on phyto-
plankton and therefore proliferate only where there is
1.2 Accumulations of Organic Matter vigorous phytoplankton production. Organisms sink
after they have died, and may decay so that nutrients
It is well documented that oil accumulations are of are released and recycled at greater depths.
organic origin and formed from organic matter in Basins with restricted water circulation will pre-
sediments. Methane can be formed inorganically and serve more organic matter and produce good source
is found in the atmosphere of several other planets, but rocks which may mature to generate oil and gas
inorganic methane from the interior of the Earth is (Fig. 1.2a, b).
likely to be well dispersed and thus not form major In polar regions, cold dense water sinks to great
gas accumulations in the Earth’s crust. depths and flows along the bottom of the deep oceans
Solar energy
173 000
Reflected Transformed
shortwave radiation to heat directly Longwave radiation
52 000 81 000
40 000
Wind, waves
370
Evaporation
40 000
Photosynthesis Tidal energy
Oxidation,
40 3
0.01–0.001%
of organic
production Crystal Hot springs
preserved heat
in sedimentary rocks flux
as kerogen, coal, 32 0.3
oil and gas.
12
Numbers in (10 w)
Fig. 1.1 Transformation of solar energy to fossil fuels by is oxidised. As a result very little organic matter is buried and
photosynthesis. Only a small fraction of the solar energy is stored in sedimentary rocks and very little of this is concentrated
used for photosynthesis and most of the produced organic matter enough to become a potential source rock
towards lower latitudes. This is the thermal conveyor The production of organic matter is not limited by
belt transporting heat to higher latitudes and it keeps carbon dioxide or water, but by nutrient availability.
the deep ocean water oxidising. In areas near the Phosphorus (P) and nitrogen (N) are the most impor-
equator where the prevailing winds are from the tant nutrients, though the supply of iron can also be
east, the surface water is driven away from the west- limiting for alga production. It is this process of pho-
ern coasts of the continents. This generates a strong tosynthesis, which started 4 billion years ago, that has
upwelling of nutrient-rich water from the bottom of built up an atmosphere rich in oxygen while
the sea which sustains especially high levels of accumulating reduced carbon in sedimentary rocks as
primary organic production (Fig. 1.3). The best oil, gas and coal. Most of the carbon is nevertheless
examples of this are the coast of Chile and off West finely divided within sedimentary rocks, for example
Africa. shales and limestones, in concentrations too low to
Through photosynthesis, low energy carbon diox- generate significant oil and gas.
ide and water are transformed into high energy Energy stored by photosynthesis can be used directly
carbohydrates (e.g. glucose): by organisms for respiration. This is the opposite pro-
cess, breaking carbohydrates down into carbon dioxide
CO2 þ H2 O ! CH2 O ðorganic matterÞ þ O2 and water again, so that the organisms gain energy.
6 K. Bjørlykke
Deposition of source rocks

a
Photosynthesis
Sea level (algae)
CO2 + H2O CH2O + O2
Limestone Ox P, N nutrients
Well sorted
Red sand, potential
Organic Poorly
mat. sorted sand reservoir
rock
Grey mud
Black mud
b Sea level
Sea floor
Sediment
3–4 km Cap rock

Reservoir
rock
Cap rock
Oil
migration
120–150°C Fault
Fig. 1.2 (a) Depositional environments for potential source maturation. The carbonate trap (e.g. a reef) is a stratigraphic
and reservoir rocks. Depressions on the sea floor with little trap, while the sandstone forms a structural trap bounded by a
water circulation provide the best setting for organic matter to fault. The source rock was deposited as black mud and was
be accumulated before it is oxidised. (b) Migration of petroleum heated to become an organic-rich shale, indicated by black
from source rocks into reservoir rocks after burial and colour. Well sorted sand became sandstone reservoirs
This occurs in organisms at night when there is no biochemical source for the synthesis of lipids and
light to drive photosynthesis. Also when we burn proteins. Proteins are large, complex molecules built
hydrocarbons, e.g. while driving a car, energy is up of condensed amino acids (e.g. glycine
obtained by oxidation, again essentially reversing the ðH2 NCH2 COOHÞ).
photosynthesis equation quoted above. Oxidation Dried phytoplankton contains 45–55% carbon,
of 100 g glucose releases 375 kcal of energy. 4.5–9% nitrogen, 0.6–3.3% phosphorus and up to
Carbohydrates that are produced but not consumed 25% of both silica and carbonate.
by respiration can be stored as glucose, cellulose or Planktonic algae are the main contributors to the
starch in the cell walls. Photosynthesis is also the organic matter which gives rise to petroleum. Among
Land
Win
d
Sea level
4 Photosynthesis
Organic 3
100m Production
Organisms sink down through
400m 1 water and break down.
2 Phosphorus and other
5 nutrients are released.
Upwelling
Deposition of phosphate –rich sediment
Crystallisation of phosphate minerals
(Apatite).
Fig. 1.3 Upwelling of water rich in nutrients on a continental margin with deposition of organic-rich mud
the most important are diatoms, which have amor- Trophic level Trophic level Trophic level Trophic level
phous silica (opal A) shells. 1 2 3 4
Diatoms are most abundant in the higher latitudes 1,000 kg 100 kg 10 kg 1 kg
and are also found in brackish and fresh water. Phytoplankton Zooplankton Crustaceans Fish
Blue–green algae (cyanobacteria) which live on the
bottom in shallow areas, also contribute to the organic All animal plankton (zooplankton) live on plant
material in sediments. plankton, and in turn are eaten by higher organisms
In coastal swamps, and particularly on deltas, we in the food chain. At each step in the food chain, which
have extensive production of organic matter in the we call a trophic level, the amount of organic matter
form of plants and trees which may avoid being (the biomass) is reduced to 10%.
oxidised by sinking into mud or bog. The residues of Ninety percent of the production of organic matter
these higher land plants may form peat, which with is therefore from algae. This is why algae and to some
deeper burial may be converted into lignite and bitu- extent zooplankton account for the bulk of the organic
minous coal. But such deposits are also a potential material which can be transformed into oil. Larger
source of gas and oil. Plant matter, including wood, animals such as dinosaurs are totally irrelevant as
also floats down rivers and is deposited when it sinks sources of oil.
to the bottom, usually in a nearshore deltaic environ- The most important of the zooplankton which pro-
ment. When the trees rot they release CO2 and con- vide organic matter for petroleum are:
sume as much oxygen as the plant produced during the 1. Radiolaria – silica shells, wide distribution, partic-
whole period when it was growing. There is thus no ularly in tropical waters.
net contribution of oxygen to the atmosphere. This 2. Foraminifera – shells of calcium carbonate.
also applies to the bulk of the tropical rainforests. 3. Pteropods – pelagic gastropods (snails) with a foot
Where trees and plants sink into black mud, which has been converted into wing-shaped lobes;
preventing them from being oxidised, there is a net carbonate shells.
contribution of oxygen to the atmosphere and a This is the second lowest level within the marine
corresponding reduction of CO2 in the atmosphere. food chain. These zooplanktonic organisms are eaten
by crustaceans which themselves are eaten by fish.
8 K. Bjørlykke
The total amount of organic matter that can be 1.3 Breakdown of Organic Matter
produced in the ocean is dependent on the nutrient
supply from rivers, but river water does not only Almost all (>99%) of the organic matter which is
carry inorganic nutrients. It also contains significant produced on land and in the oceans is broken down
amounts of organic matter, in particular humic acid through direct oxidation or by means of
compounds, lignin and similar substances formed by microbiological processes. If oxygen is present,
the breakdown of plant material which are weakly organic matter will be broken down in the following
soluble in cold water. When the river water enters manner:
the sea, there is precipitation due to the increased pH
and lower surface temperature in the ocean. CH2 O þ O2 ! CO2 þ H2 O
Other plant materials, like waxes and resins, are
more chemically resistant to breakdown and are insol- Where oxygen is available, organic matter is
uble in water. Such organic particles tend to attach oxidised relatively rapidly both on land and in the
themselves to mineral grains and accompany sediment sea. As organisms die, organic material suspended in
out into the ocean. seawater sinks through the water column consuming
Most of the oil reservoirs which have been formed oxygen. If water circulation is restricted due to density
since the Palaeozoic have been uplifted and eroded, stratification of the water column, the oxygen supply
and over time vast quantities of oil have flowed will be exhausted. Instead, the bound oxygen in
(seeped) out onto the land or into the sea. In this sulphates or nitrates is used by sulphate-reducing and
sense, oil pollution is a natural process. Only a small denitrifying bacteria which decompose organic mate-
proportion of the petroleum that has been formed in rial in an anoxic environment. The first few
source rocks has actually become trapped in a reser- centimetres below the seabed are usually oxidised,
voir. One might expect this seepage to have provided a while reducing conditions prevail 5–30 cm below the
source of recycled petroleum in younger sediments, sea floor. Below this redox boundary where there is no
but petroleum breaks down extremely rapidly when free oxygen, sulphate-reducing bacteria react with
subjected to weathering, oxidising to CO2, and the organic matter as indicated below:
nutrients (P, N) that were required to form the organic
matter are released and may act like a fertilizer. 2CH2 O þ 2Hþ þ SO4 ! H2 S þ 2CO2 þ 2H2 O
On land, evaporation will remove the lighter
NH3 þ Hþ þ SO2 ! NO3 þ H2 S þ H2 O
components while bacteria will degrade the heavier
components. Fossil asphalt lakes consist of heavy
H2S is liberated, giving stagnant water and mud a
substances which neither evaporate nor can be easily
strong smell. Through denitrification we get
broken down by bacteria. In the ocean, the lighter
components will dissolve quite rapidly, while the
5CH2 O þ 4Hþ þ 4NO3 ! 2N2 þ 5CO2 þ 7H2 O
heavier asphalt fraction will sink to the bottom and
be degraded and recycled.
When the rate of accumulation of organic matter
In uplifted sedimentary basins like the Ventura
exceeds the rate of oxygen supply the redox boundary
Basin and the Los Angeles Basin in Southern
will be in the water column, separating the oxidising
California there are abundant natural oils seeps both
surface water from the reducing bottom water.
onshore and offshore.
This is typical of basins separated from the deep
On the beaches from Santa Barbara towards Los
ocean by a shallow sill, like the Black Sea and some of
Angeles there are many natural oil seeps.
the deep Norwegian fjords. Fresh or brackish surface
water floating on more saline water also helps to
maintain a stable water stratification with little vertical
mixing. Lakes may have good water stratification
because warm surface water is less dense than the layer spending longer in the bioturbation and
colder bottom water. Black mud deposited at the bot- microbiological breakdown zones, and consequently
tom of lakes may produce good source rocks. In cold less organic matter will be preserved in the sediment.
climates, however, the water in the lakes overturns in Rapid sedimentation leads to more of the deposited
the winter because the maximum water density is at organic matter being preserved but from the outset it
4 C, preventing the stable stratification required to will be highly diluted with mineral grains. Conse-
form source rocks. quently an intermediate sedimentation rate in relation
to organic production (10–100 mm/1,000 years)
results in the best source rocks.
1.4 Formation of Source Rocks As we have seen, the net accumulation of organic
matter in sediments is not so much a function of the
All marine organic material is formed near the surface total productivity, but rather of the relationship
of the ocean, in the photic zone, through photosynthe- between productivity and biogenic breakdown and
sis. For the most part this is algae. Some phytoplank- oxidation. In areas with powerful traction currents,
ton are broken down chemically and oxidised and most organic matter will be oxidised. An important
some are eaten by zooplankton. Both types of plankton source of oxygen-rich water in the deep ocean is the
are eaten by higher organisms which concentrate the cold surface water which sinks to the bottom of the
indigestible part of the organic matter into fecal pellets ocean in polar regions and flows along the ocean floor
which may be incorporated into sediments. Plankton is towards equatorial regions. This flow balances the
made up of very small organisms which sink so slowly surface flow to higher latitudes like the Gulf Stream
that they are in most cases almost entirely degraded in the Atlantic.
(oxidised) before they reach the bottom. Pellets, on the These bottom flows are of considerable magnitude
other hand, are the size of sand grains and sink more during glacial periods, when large amounts of cold
rapidly, and this organic matter is more likely to be water are sinking near ice sheet peripheries. In warm
preserved in the sediments. periods, for example during the Cretaceous, the poles
On the bottom, organic matter will be subjected to were probably ice-free and there was much less cold
breakdown by micro-organisms (bacteria). It will also surface water available to sink down and drive the
be eaten by burrowing organisms which live in the top ocean conveyor system. The deeper parts of the Atlan-
portion of the sediments. The activity of these tic experienced stagnant bottom conditions during
organisms contributes to reducing the organic content such periods.
of the sediments because most of the organic matter is Limited water circulation in semi-enclosed marine
digested when the sediment is eaten. Bioturbation also basins due to restricted outflow over a shallow thresh-
stirs up the sediments, exposing them more to the old is a common cause of stagnant water bodies
oxygen-bearing bottom water. However, if the bottom (Fig. 1.2a). The Black Sea is a good example. In
water is stagnant, the lack of oxygen and the toxicity response to an abundant freshwater supply from rivers
of H2S will exclude most life forms. The resultant lack and a relatively low evaporation rate, a low salinity
of bioturbation will thus preserve more organic matter surface layer leads to density stratification in the water
in the sediment together with perfect, undisturbed, column and a consequent reduction in circulation. In
lamination. Stagnant, or anoxic, conditions are defined basins with little precipitation and where there is net
by an oxygen content of <0.5 ml/l water. Sulphate- evaporation, the surface water will have higher salinity
reducing bacteria, however, can use a good deal of and density than the water below it, and will sink
organic matter and precipitate sulphides (e.g. FeS2). If down. This circulation brings with it oxygen from
the sediments contain insufficient soluble iron or other the surface and can give oxidising bottom conditions
metals which could precipitate sulphides, more sul- with little chance for organic matter to survive to form
phur will be incorporated in the organic matter and source rocks.
will eventually be enriched in the oil derived from Lakes or semi-enclosed marine basins often have a
such source beds. temperature- or salinity-induced density stratification
Except where the water is completely stagnant, so that oxygenated surface water does not mix with
slow sedimentation rates will result in each sediment water in the deeper part of the basin. This leads to
10 K. Bjørlykke
anoxic conditions and a high degree of preservation of represent a hazard in the form of blowouts and fire
the organic matter produced in the surface waters. This during drilling. Gas occurring at shallow depth may
aspect is therefore of considerable interest in explora- also have a deeper source generated from a gas-prone
tion for petroleum in freshwater basins, particularly in source rock (coaly sediments) or by cracking of oil,
Africa and China. The open oceans have normally had but such gas has a very different isotopic signature
oxygenated water, but during the Cretaceous most of than biogenic gas.
the Atlantic Ocean is believed to have been stagnant
during so-called “anoxic events”, and substantial
amounts of black shale were deposited in the deeper 1.6 Kerogen
parts of the ocean during these periods.
As organic material becomes buried by the accumula-
tion of overlying sediments, water is gradually
1.5 Early Diagenesis of Organic Matter expelled during compaction.
Complex organic compounds like proteins are bro-
Microbiological breakdown of organic matter in ken down into amino acids, and carbohydrates into
sediments is due to the activity of bacteria, fungi, simpler sugar compounds. These are able to recom-
protozoa, etc. and under oxidising conditions these bine to make larger compounds, for example by amino
are extremely effective. However, the porewater acids reacting with carbohydrates (melanoid reaction).
quickly becomes reducing if the oxygen is not As this type of polymerisation proceeds, the propor-
replenished. In relatively coarse-grained sediments tion of simpler soluble organic compounds diminishes
(sand), oxygen may diffuse to depths of 5–20 cm at depths of a few tens of metres down in the sediment.
below the seabed, while in clay and fine-grained car- It is these newly-formed complex organic structures
bonate mud the boundary between oxidising and which are called kerogen.
reducing water (redox boundary) may be just a few Kerogen is a collective name for organic material
millimetres below the seafloor. The pores in the that is insoluble in organic solvents, water or oxidising
sediments here are so small that water circulation acids. The portion of the organic material soluble in
and diffusion are insufficient to replace the original organic solvents is called bitumen, which is essentially
oxygen in the porewater as it gets used up by oxidation oil in a solid state.
of organic matter. Clay-rich sediments soon become a Kerogen consists of very large molecules and is a
relatively closed system, and the downward diffusion kind of polymer. When it has been exposed to suffi-
of oxygen from the seabed is very slow in fine-grained cient time and temperature these large molecules will
sediments. crack into smaller molecules, mostly petroleum. When
Aerobic breakdown is therefore much more effec- the temperature is about 100 C a long period of geo-
tive in coarse-grained sediments than in fine-grained logical time is required. In rapidly subsiding basins the
ones. In anaerobic transformation bacteria use organic exposure time is shorter and oil generation may only
matter, e.g. short carbohydrate chains. Cellulose is start at about 140–150 C. In the North Sea basin the
broken down by fungi, and finally by bacteria. The “oil window” may typically be between 130 and
end products are methane (CH4) and carbon dioxide 140 C.
(CO2). Methane, however, is the only hydrocarbon
produced in any quantity at low temperatures by bac-
teria close to the surface of the sediment. Gas occur- 1.7 Migration of Petroleum
ring at shallow depths (shallow gas) therefore consists
largely of methane (dry gas) unless there has been Petroleum migrates from low permeability source
addition from much deeper strata. Biogenic gas may rocks into high permeability reservoir rocks from
form commercial accumulations, as in Western which the petroleum can be produced (Fig. 1.2b).
Siberia and also in the shallow part of the North Sea The main driving force for petroleum migration is
basin. The presence of abundant shallow gas may buoyancy because it is less dense than water. The
forces acting against migration are the capillary forces We distinguish between primary migration, which
and the resistance to flow though rocks with low is the flow of petroleum out of the source rock and
permeabilities. If the resistance to flow (permeability) secondary migration, which is the continued flow
is very low the fluid pressure both in the water phase and from the source rock to the reservoir rock or up to
the petroleum phase will increase and cause tight shales the surface (Fig. 1.4).
to fracture, thus opening up for petroleum migration. Oil and gas may also migrate (leak) from the reser-
Migration of oil and gas will therefore nearly voir to a higher trap or to the surface. Hydrocarbons
always have an upwards component. Petroleum may have a low solubility in water and will therefore
however locally be expelled downwards from a source migrate as a separate phase. Solubility varies from as
rock into a sandstone. little as 24 ppm for methane to 1,800 ppm for benzene.
Pleistocene
Cenozoic sediments
Cretaceous mudstones and shales
m.
tF .
ren Fm
B rd
tfjo
Sta
Triassic and Permian

sediments
Reservoir sandstone with gas Migration of gas and oil from Upper Jurassic
source rocks (Kimmeridge shale) into Middle
and Lower Jurassic reservoir sandstones
Reservoir sandstone with oil (Brent and Statfjord formations).
Upper Jurassic source rock
Fig. 1.4 Schematic illustration of primary migration (expul- Sea where rifting in Upper Jurassic time produced good
sion) of petroleum from a source rock and secondary migration conditions for the formation of a source rock and also traps on
into a reservoir (trap). This example is from the northern North the uplifted fault blocks
12 K. Bjørlykke
Other compounds, such as pentane, are even less solu- H2O

ble (2–3 ppm). However, solubility increases mark- Sand grain
edly with pressure. Many hydrocarbons have
solubilities of less than 1 ppm in water.
Oil flows
It is not possible that oil should be dissolved in Oil through reservoir
water and transported in an aqueous solution, both rocks
because of the solubility and the low flow rates. It
would also be difficult to explain how the oil would
come out of solution in the reservoirs (traps). It is
therefore generally accepted that oil migrates as a
separate phase.
Gas, in particular methane, has a fairly high solu-
bility in water, especially under high pressure. If
Water flows
methane-saturated water rises to lower pressures, through reservoir
large quantities of methane can bubble out of solution. rocks
Oil is lighter than water, and oil droplets would be
able to move through the pores in reservoir rocks but
the caplliary restance is high for separate oil drops in a
Oil
water-wet rock (Fig. 1.5). In order to pass through the
narrow passage between pores (pore throat), the oil Fig. 1.5 Most sandstones are water-wet and have a thin layer of
droplets must overcome the capillary forces. When the water around the grains. A continuous oil phase will flow easily
pores are sufficiently small in a fine-grained sediment, if the permeability is relatively high and the pore throats
between the pores are relatively wide. Isolated droplets of oil
these forces will act as a barrier to further migration of will, however, be prevented from moving by capillary forces
oil. The small gas molecules, however, can diffuse
very slowly through extremely small pores and thus
escape from shales which form tight seals for oil. This
is however a very slow process when the cap rock is generated directly from gas-prone kerogen in the
tight and gas (methane) is often trapped as a separate source rock. There is also usually a large amount of
phase in reservoirs. gas dissolved in the porewater which can be released
Oil can therefore not migrate as small discrete when the pressure is reduced.
droplets, but moves as a continuous string of oil The degree of alteration of organic matter can be
where most of the pores are filled with oil rather than measured in different ways. Plant material is altered
water (highly oil-saturated). The pressure in the oil from a dull material to a material which becomes
phase at the top is then a function of the height of the more shiny with increasing temperature. This can be
oil-saturated column (string) and the density differ- quantified by measuring the amount of light reflected
ence between oil and water. from a piece of plant material (vitrinite) under the
The rate of migration is a function of the rate of microscope. This is the vitrinite reflectivity which
petroleum generation in the source rocks. The matura- increases with higher temperatures and maturity. A
tion of the source rock controlling the supply of petro- vitrinite reflectivity of 1.2 indicates that the source
leum which can migrate is a function of the rock has generated much of the oil that can be
temperature integrated over time (Fig. 1.6). generated. We will say that the source rock is in the
The temperature history is a function of the burial middle of the “oil window” (Fig. 1.7). Values
depth and the geothermal gradients. below 0.7–0.8. are found in source rocks which
Deep burial over long time will cause all oil to be have not been heated enough (immature source
decomposed (cracked) into gas. In addition gas will be rocks).
Photosynthesis
CO2+H2O Oxidation
CO2+H2O +(N,P)
CH2O+O2+(N,P)
CH2O +O2
Time
20°C
Burning of oil
and gas
Oil
3–4 km – 100°C Gas
6–8 km – 200°C Burial curve

Rate of petroleum generation from
organic matter (kerogen) follows an
Depth (temperature) arrhenius function: K = Aexp(-E/RT)
E - activation energy. R - gas const.
T - temperature
Fig. 1.6 Burial curve for source rocks determining the transformation of kerogen to oil and gas depending on time and temperature
(burial depth)
Burial
depth Vitrinite
km Temperature (°C) reflectivity
1 30° Diagenesis
Sub-bituminous coal,
Kerogen Lignite
0.5
2 60°
3 90° Bituminous
Katagenesis coal
Oil
4 120° 1.2
5 Gas 150° 2.0

Metagenesis Anthracite
Fig. 1.7 Alteration (maturation) of organic matter and genera- higher reflectivity with increasing temperature. The depth (tem-
tion of oil and gas as a function of temperature. The maturation perature) range where oil is generated is called the “oil win-
is also a function of time and this can be determined by measur- dow”. At higher temperatures source rocks will produce mostly
ing the vitrinite reflectivity. Coals become more shiny and have gas as oil will be altered into gas by cracking
Vitrinite reflectivities close to 2.0 and above indi- The Upper Jurassic Kimmeridge shale (Fig. 1.8) is
cate that the source rock has generated all the oil and the main source rock for the North Sea basin but it is
can generate only gas. not mature at its outcrop at Kimmeridge Bay in Dorset,
14 K. Bjørlykke
Oil/gas
Shale
Sandstone
Fig. 1.8 Kimmeridge Clay (Upper Jurassic) unconformably

overlain by the lower Chalk (Upper Cretaceous). This is a very Oil/gas
good source rock and equivalent shales are the main source
rocks for oil and also much of the gas in the North Sea basin
and also further north in mid Norway (Haltenbanken) and the
Barents Sea. A c.50 m.y. hiatus separates the Albian age Red
Chalk carbonate facies from the underling black shale facies.
From South Ferriby, Yorkshire, England
south England and in many parts of the North Sea

Basin. Fig. 1.9 Examples of structural traps. Simple anticlinal trap
and a fault-controlled trap
1.8 Hydrocarbon Traps (1) Structural traps that are formed by structural
deformation (folding, doming or faulting) of rocks.
Traps consist of porous reservoir rocks overlain by (2) Stratigraphic traps which are related to primary
tight (low permeability) rocks which do not allow oil features in the sedimentary sequences and do not
or gas to pass. These must form structures closed at the require structural deformation like faulting or
top such that they collect oil and gas, which is lighter folding. This may be sandstones pinching out in
than water. We can think of an oil trap as a barrel or shales due to primary changes in facies (Fig. 1.11).
bucket upside down (Fig. 1.9) which can then be filled
with petroleum which rises through the water until it is Carbonate reefs tend to form primary structures
full. The point where the petroleum can leak from this which function as stratigraphic traps.
structure is called the spill point. The closure is the There are also combinations between stratigraphic
maximum oil column that the structure can hold and structural traps (Fig. 1.12).
before leaking through the spill point (Fig. 1.10). It is important to establish when structural traps
Hydrocarbons may also leak through the cap rock. were formed in relation to the migration of the petro-
The cap rock may not be 100% effective in leum. Structures formed after the main phase of source
preventing the upward flow of hydrocarbons, but rock maturation and associated migration will not be
these will still accumulate if the rate of leakage is effective traps. In some cases traps formed late can
less than the rate of supply up to the trap. Cap rocks collect gas which normally is generated and migrates
are not totally impermeable with respect to water, but later than oil.
may be impermeable to oil and gas due to capillary Stratigraphic traps, by contrast, have been there all
resistance in the small pores. the time, and the timing of the migration is not so
Traps can be classified according to the type of important. They may however depend on slight tilting
structure that produces them. We distinguish between: of the strata involved.
Shale Oil Onlap quantitatively most important salt minerals are halite
(NaCl – density 2.16 g/cm3), gypsum (CaSO4. 2H2O –
density 2.32 g/cm3). Anhydrite (CaSO4 – density
2.96 g/cm3) is too dense to contribute to the formation
Fractured basement of diapirs.
In order for the salt to move upwards and form a
Structural trap
salt dome, a certain thickness of overburden is
required and the salt beds themselves must be at least
Gas
Oil /gas 100–200 m thick. The upward movement of salt
contact Closure
through the overlying sequence, and the resultant
deformation of the latter, is called halokinetics or salt
ir Water tectonics.
vo
s er ck The rate of salt movement is extremely slow and a
e o Oil
R
r
Oil/water Spill point dome may take several million years to form.
contact Movements of the Earth’s surface may, however,
also be recorded in recent history as is the case onshore
Shale Denmark. Salt may break right through the overlying
rocks and rise to the surface, or form intrusions in
Salt diapir younger sediments. If gypsum has been deposited,
this will be altered into anhydrite at about 1 km burial
depth, with a consequent 40% compaction and the
Oil/gas increase in density will remove the buoyancy relative
“Turtle back”
to the surrounding sediments. A comparable expan-
sion occurs when rising anhydrite comes into contact
with groundwater and reverts to gypsum.
Traps may be created (1) in the layers above the salt
Fig. 1.10 Structural traps related to salt domes and anticlinal dome, (2) in the top of the salt dome (cap rock), (3) in
folds. A basement high can also be a trap when it is covered by a the beds which are faulted and turned up against the
black shale (source rock). The basement may have some poros-
ity due to fractures or a thin sediment cover
salt structure and (4) through stratigraphic pinching
out of beds round the salt dome. Reservoirs may form
by solution and brecciation at the top of salt domes.
1.8.1 Structural Traps
Salt tectonics is of great importance in many oil-
bearing regions where there are thick salt deposits in
(a) Anticlinal Domes
the passive margin sequences of the South Atlantic
Domes formed by diapirism or other processes may
and the Gulf of Mexico. In the eastern USA we find
form closures in all directions (four-way closure). A
extensive tracts with Silurian salt, and in Texas and
simple anticline is not sufficient to trap oil. Anticlines
New Mexico we have Permian salt.
with an axial culmination are needed to provide four-
Salt layers are the ideal cap rock because of salt’s
way closure. This means that the fold axis must be
low permeability and ductile properties, which prevent
dipping in both directions (Fig. 1.9).
fracturing and leakage.
Anticlinal traps can form in association with
Salt deposits are particularly common in the
faulting. This is especially true in connection with
Permo-Triassic around the Atlantic. This is because
growth faults (roll-overs) (see below), but also with
prior to the opening of the Atlantic there were vast
thrust zones.
areas with fault-controlled basins (rifts) in the middle
of a supercontinent (America + Europe, Asia and
(b) Salt Domes Africa) with little precipitation. We find similar
Salt domes are formed because salt (specific gravity conditions today around the Red Sea and the Dead
c.1.8–2.0) is lighter than the overlying rock, and the Sea. The Permian Zechstein salt in Germany and
salt therefore “floats” up due to buoyancy. The Denmark continues below the North Sea, and
16 K. Bjørlykke
I Barrier
Clay sand island Lagoon Sand
Sea level
Ox
Red
Organic
matter
II Sea level
rise Clay
Sand
III Deposition of 2–3 km of sediments
Reservoir
rock (trap)
Migration of
hydrocarbons
>100°C
Source rocks
Water
Oil
Sand grains
Fig. 1.11 Examples of stratigraphic traps. A barrier island forms a separate accumulation of sand where the associated mudstones
(shales) may represent both source rocks and cap rocks
halotectonic movements have formed dome structures (d) Fault Traps

in the Chalk, for example in the Ekofisk area. In fault traps, the fault plane forms part of the structure
trapping the oil and hindering its further upward
(c) Growth Anticlines migration. The fault plane must therefore be sealing
These are dome-like structures formed when part of a for vertical flow of petroleum in order to function as a
basin subsides more slowly than its surroundings, barrier and a cap rock for the reservoir rocks. If the
resulting in least sedimentation on the highest part. reservoir rock is juxtaposed against a sandstone or
The sediment thickness decreases towards the dome other permeable rocks the fault must also be imperme-
centre, which also compacts less than the adjacent able for flow across the fault plane. Most frequently,
thicker sediments and thus contributes to the forma- however, the reservoir rock is faulted against a tight
tion of an anticlinal structure. Growth anticlines form shale or mudrock and the fault is then in most cases
contemporaneously with sediment accumulation, not sealing (Fig. 1.9). When there is sandstone on both
through later folding. sides of the fault plane, the permeability across it will
Growth anticlines can be formed above salt domes, amongst other things depend on how much clay has
reefs or buried basement highs, through differential been smeared along the junction fault plane from
compaction. adjacent clays during faulting. At greater depths
The name growth fault goes back to the early days of

Unconformity oil exploration without seismic data. It was noticed
that the layers had “grown” in thickness in the wells on
the downthrown side of the fault. The displacement of
Oil/gas
the beds decreases upwards along the fault plane.
Smaller, antithetic faults often develop in the opposite
direction in the beds which are turned inwards towards
Oil/gas
the main fault plane. Growth faults tend to have low
permeability and may contribute greatly to reduced
porewater circulation in sedimentary basins, and we
often find undercompacted clay, which can turn into
clay diapirs in association with growth faults.
Oil/gas
1.8.2 Stratigraphic Traps
These are traps which are partially or wholly due to

facies variation or unconformities, and not primarily
Oil/gas
the result of tectonic deformation. Porous and perme-
able sands which pinch out up-dip in less permeable
rocks, e.g. shale (Figs. 1.11 and 1.12) are good
examples. Barrier islands often form stratigraphic
traps because they may be separated from the coast
by fine-grained lagoonal facies. The main types are:
Channel sand (a) Fluvial channel sandstones may be isolated and
surrounded by impermeable clay-rich sediments,
Fig. 1.12 Combination of stratigraphic and structural traps. A
or they may be folded so that we obtain a combi-
reef forms a trap due to the primary relief and also due to later
compaction of the mud around the reef nation of stratigraphic and structural traps
(Figs. 1.9 and 1.10).
(b) Submarine channels and sandstone turbidites in
(>3–4 km) there may be diagenetic changes such as strata rich in shale. Here we will often find
quartz cementation, which can make the fault plane pinch-out of permeable layers up-dip from the
less permeable. foot of the continental slope. This will result in
The displacement along faults can be both vertical stratigraphic traps without any further folding
(normal faults) and horizontal (strike slip faults). being necessary.
Reverse faults are faults where the hangingwall is (c) Reefs often form stratigraphic traps. A reef struc-
moved upwards relative to the footwall below the ture projects up from the sea bed and often has
fault plane. These are typical of areas with high hori- shale sediments surrounding it, so that oil could
zontal stresses i.e. due to converging plate movements. migrate from the shale into the reef structure.
Growth faults are driven by gravity-sliding along (d) Traps related to unconformities. Sandstones or
curved (listric) fault planes and are typical of sedimen- other porous rocks may be overlain with an angu-
tary sequences such as deltas deposited with relatively lar unconformity by shales or other tight
rapid sedimentation. The fault plane is often (though sediments, forming a trap underneath the uncon-
not always) sealing and can stop oil and gas from formity (Fig. 1.12). Topographic highs in the base-
migrating further upward. However, oil traps are ment overlain with shales can also provide good
equally often formed in anticlines on the upper side traps in fractured basement rocks. Remember that
of the fault plane. These are rollover anticlines. oil can migrate upwards into stratigraphically
Because the faulting is active during sedimentation, lower rocks. In China there are numerous
the layers on the downthrown side will be thickest. examples of this type of trap.
18 K. Bjørlykke
Reservoir Geology is not a well-defined discipline.

It includes many aspects of geology that are of special
relevance to the production of petroleum. It is also
linked to engineering aspects of petroleum production.
Reservoir geophysics has in recent years become very
important and is now well integrated with reservoir
geology.
1.10 Porosity and Permeability
Any rock with sufficiently high porosity and perme-

ability may serve as a reservoir rock provided that
there is a source of petroleum, a structure, and a tight
Fig. 1.13 Natural oil seep at Carpenteria State Beach, cap rock.
California. Oil is flowing on land, on the beach and also offshore Sediments consist of solid grains and of fluids
on the sea floor
which for the most part are water but may be oil and
gas.
Much of the oil generated in sedimentary basins has Porosity (φ) is an expression of the percentage (or
not been trapped in reservoirs but reached the surface fraction) of fluids by volume (Vf) compared to the total
on the seafloor or on land. There it is then broken down rock volume with fluids (Vt), so that φ ¼ Vf / Vt.
by bacteria and becomes heavy oil, which is not very Porosity is often expressed as a percentage, but in
toxic. In California there are many examples of natural many calculations it is easier to express it as a fraction,
oil seeps which can be observed along roads, on the for example 0.3 instead of 30% porosity.
beach (Fig. 1.13) and also offshore. The void ratio (VR) is the ratio between pore vol-
ume (φ) and the volume of the grains (1–φ).
1.9 Other Types of Trap VR ¼ φ=ð1 φÞ
More unusual kinds of trap can be encountered. If the Void ratio is often used in engineering and it has
porewater in a sedimentary basin has sufficiently certain advantages in some mathematical expressions.
strong flow of meteoric water into the basin, the oil/ If we assume that we know the density of the
water contact may diverge markedly from a horizontal mineral grains, the porosity can be found by measur-
plane due to the hydrodynamic stresses. This has ing the density of a known volume of the sediment.
implications for calculations of oil volumes within a The density of the sediments (ρs) is the sum of the
structure, and in some instances oil can accumulate density of the grains in the solid phase, which are
without being sealed in, within a so-called hydrody- mostly minerals ρm, and the density of the fluids (ρf).
namic trap. The circulation of fresh (meteoric) water
down into oil-bearing rocks will, however, lead to ρs ¼ φρf þ ρm ð1 φÞ
biodegradation and the formation of asphalt. Asphalt
can then become a tight cap rock for the oil. Well sorted, rounded sand grains are almost spher-
At greater depths, beyond the reach of meteoric ical in shape. If we have grains of the same size, which
water, water movement is limited and any deflection are all quite well rounded and with a high degree of
of the oil/water contact is more likely to reflect pres- sphericity, we will be able to pack the grains so as to
sure differences within the reservoir. Water will also get minimum porosity. Rhombic is the densest pack-
flow then because of the pressure gradient, but unless ing, resulting in 26% porosity, but this can not be
there are low permeability barriers the pressure will obtained naturally. Cubic packing, where the grains
soon equalise. Tectonic tilting will also tilt the oil/ are packed directly one above another, results in about
water contact. 48% porosity and this does not occur in nature either.
Most well sorted sandstones have a porosity which lies flux F ¼ Q=A. F can be measured in cm3/cm2/s or in
between these two values, typically around 40–42%. m3/m2/s. This is equal to the Darcy velocity which is
Poorly sorted sand may have lower primary porosity cm/s.
and will also compact more at moderate burial depths. Well-sorted sandstones may have permeabilities
Clay-rich sediments have a much greater porosity exceeding 1 Darcy and values between 100 and
immediately after deposition, typically 60–80%. This 1,000 mD are considered to be extremely good.
means that immediately following deposition a sand Permeabilities of 10–100 mD are also considered to
bed is denser than a bed of clay or silt. However, clay be good values for reservoir rocks. Permeabilities of
and silt lose their porosity more rapidly with burial. 1–10 mD are typical of relatively dense sandstones
Porosity may be classified into different types and limestones, so-called tight reservoirs. There are
depending on its origin. also examples of rocks with even lower permeabilities
Pore space between the primary sediment grains is being exploited commercially for oil production, for
often referred to as primary porosity. Intergranular example in the Ekofisk Field where the generally low
porosity simply means porosity between the grains permeability of a chalk matrix is enhanced by
whereas intragranular porosity means porosity inside fractures which increase the overall permeability.
the sediment grains. The latter may be cavities in In the great majority of rocks, the permeability
fossils, e.g. foraminifera, gastropods, molluscs, but differs according to flow direction. In sedimentary
also partly dissolved feldspar and rock fragments. rocks the permeability is much higher parallel to the
Pore space formed by dissolution or fracturing of bedding compared with normal to the bedding. Chan-
grains is called secondary porosity. nel sandstones can also have a marked directional
Cavities formed by selective solution of sediment impact on the permeability.
grains or fossils are classified as mouldic porosity. A In well-cemented sandstones and limestones, and
typical example is when dissolution of aragonite also in certain shales, the matrix permeability is
fossils like gastropods leaves open pore spaces extremely low and the effective permeability may be
(moulds). mostly controlled by fractures if they are present.
Particularly in carbonates we may also have poros- Claystones and shales have very low permeability
ity on a large scale i.e. as caverns (karst) and in reefs. and can be almost completely tight. In the laboratory
Pore space produced by fracturing is called fracture shale permeabilities as low as 0.01 nanodarcy have
porosity. been measured. Samples from cores or outcrops can
Permeability is an expression of the ease with contain minute fissures formed in response to
which fluids flow through a rock. It will depend on unloading during retrieval to the surface and these
the size of the pore spaces in the rocks, and in particu- must be closed to replicate the in situ permeability
lar the connections between the pore spaces. Even thin prior to unloading.
cracks will contribute greatly to increasing the Most rocks are far from homogeneous. We may
permeability. measure the porosity and permeability of a hand spec-
Permeability can be measured by letting a liquid or imen or core plug, but it is not certain that these are
gas flow through a cylindrical rock sample under pres- representative of a larger volume. Fractures occur at
sure. The pressure difference P1 P2 between the two varying intervals, and range in size from large, open
ends of a horizontal cylinder is ΔP, the cylinder length joints down to microscopic cracks which can barely be
L, and the flow rate of water (or another fluid) through seen in a microscope.
the cylinder, is Q (cm3/s). A is the cross-section and μ Rocks with low porosity and permeability may
the viscosity of the fluid fracture and sufficiently increase their porosity, and
particularly permeability, to form large oil reservoirs.
k A ΔP This means that reservoirs may be good producers
Q¼
Lμ despite relatively low porosity.
Occasionally we find petroleum in fractured meta-
where k is the permeability. morphic and igneous rocks but reservoirs normally
The volume of water which flows through each consist of sedimentary rocks. Sandstones make up
surface unit in the cross-section A is thus equal to the about 50–60% of the reservoirs in the world while
20 K. Bjørlykke
carbonate reservoirs may account for almost 40%. more likely that water will break through along the
Many of the reservoirs in the Middle East are carbon- basal part and the oil will be by-passed in the finer-
ate rocks but in the rest of the world the percentage of grained upper part during production.
carbonate reservoirs is lower. Braided stream facies will tend to have higher sand/
The most important aspects of reservoir rocks shale ratios and will normally have better lateral and
include: vertical permeabilitites on a larger scale.
(1) The external geometry such as the thickness and The ratio between the intervals with high enough
extent of the reservoir rock in all directions. porosity and permeability to be produced (net or pay),
(2) The average porosity, pore size and pore and the total sequence (gross), will be mostly deter-
geometry. mined by the primary facies relationships. The net/
(3) The distribution of permeability in the reservoirs, gross ratio is often taken to be approximately equal
particularly high permeability conduits and low to the sand/shale ratio but even at moderate burial
permeability barriers to fluid flow. depths many sandstones are not reservoir rocks, due
(4) Mineralogy and wettability of the pore network. to poor sorting or carbonate cement.
The properties of sandstone and carbonate Aeolian dunes also have specific external
reservoirs are primarily linked to the depositional geometries and there are many different types. Here
environment, the textural and mineralogical composi- the net/gross will be very high. Aeolian sand is often
tion and the burial history. A good background in reworked by transgressions, accumulating as marine
general sedimentology, facies analysis and sequence sediments in drowned topographic depressions
stratigraphy is therefore important. (valleys).
Nevertheless, many of the important properties of Marine sandstones deposited as delta mouth bars,
reservoir rocks linked to changes in facies and smaller shoreface accretion and barrier islands have
faults are below the vertical resolution of exploration thicknesses controlled by the wave energy (wave
seismic (15–30 m) and it is important to establish base depth). In protected environments, particularly
relationships between facies models, diagenetic pro- inter-distributary bays, the shoreface sandstones may
cesses and reservoir properties. The properties of be very thin. Each shallow marine unit has a limited
faults are also very important factors determining oil thickness controlled by fairweather wave base. Local
flow during production. subsidence or transgressions can increase the thick-
ness of these sand deposits.
The tidal range is very important in determining the
1.11 External Geometry of Reservoir thickness and the length of tidal channel sandstones.
Rocks Tidal channels and also fluvial channels in deltas tend
to be oriented perpendicular to coastlines.
The external geometry of reservoir rocks is largely Drilling into shallow marine sandstones, it would
determined by the depositional environments, but be very important to determine whether it was a barrier
faulting and diagenesis may define the lateral or verti- island which would represent an elongated reservoir
cal extent of a reservoir. parallel to the coastline, or a tidal sandstone which
Fluvial sandstones typically represent point bar tends to be oriented perpendicular to the coastline. In
sequences in a meandering river system. The lateral some cases dipmeter logs could help to determine the
accretion of the point bar will deposit a sandstone orientation of cross-beddding and progradation direc-
layer extending to the width of the meander belt in tion of sand bars.
the valley. The thicknesses of channel sandstones are Turbidites may be laterally very extensive, but may
limited by the depth of the river. The primary thick- also be confined to narrow submarine channels. In
ness at the time of deposition is, however, reduced by either case they may form very thick sequences
10–30% or more by compaction. because there is ample accommodation space. We
The overbank muds will become tight shales which may have very thick sequences of stacked sandstone
will reduce the vertical permeability. Fluvial channels reservoir rocks in slope and deepwater facies and this
are characterised by fining-upwards sequences with may compensate for the lower porosity and permeabil-
the highest permeability near the base. This makes it ity compared to beach deposits.
Turbidites and fluvial sandstones form fining- sedimentology (depositional environments) of reser-
upwards units while marine shoreface and mouth bar voir rocks, sediment composition (provenance), dia-
sandstones are coarsening-upwards. This becomes genesis and the structural geology.
very significant during production because oil and Prior to drilling exploration wells, nearly all our
gas will be concentrated in the upper part. knowledge about a reservoir is based on geophysical
Coarsening-upwards sandstones therefore have the data. Even after data has been acquired from explora-
best properties for flow of oil and gas during tion wells, and also from production wells, prediction
production. of the reservoir properties continues to be mostly
based on geophysical methods and extrapolation
between wells.
1.12 Changes in Rock Properties During Geophysical methods including 3D and 4D seismic
Burial and Uplift (Diagenesis) now provide a much more detailed picture of the
reservoir than only 10–20 years ago. The methods
The changes in properties are due to increased burial for detecting fluid contacts, not only gas/water
and also to uplift. Both sandstone and carbonate contacts but also oil/water contacts, from seismic
reservoirs undergo diagenesis, which will cause a data have improved greatly.
reduction in porosity and permeability as a function
of increasing burial.
The reduction in porosity (compaction) may be 1.13 Carbonate Reservoirs
mechanical in response to increased effective stress
from the overburden, or chemical as a result of the Reefs stand up as positive structures and may be
dissolution and precipitation of minerals. The porosity draped by mud and form a stratigraphic trap deter-
of reservoir sandstones or carbonates may increase mined by the size of the reef structure.
with depth in certain intervals, but this is because of Carbonate reefs, and other carbonate deposits
the changes in the primary sediment composition. which can be reservoirs, form in a wide range of
Each lithology has a different porosity depth curve. environments but all require clear water without
In a uniform primary lithology the porosity and the much clay sedimentation. Reefs are deposited in high
density will be reduced as a function of burial depth energy environments along coastlines exposed to high
(temperature and stress). Overpressure causes reduced wave energy. The Bahamas carbonate platform has
effective stress resulting in less mechanical compac- well-developed reefs on the exposed eastern side but
tion. Near the surface, meteoric flow may cause disso- not on the more protected western side. Coral reefs
lution and a net increase in the porosity in carbonates also require warm water (>20 C) and do not form
(karst) and even, to a certain extent, in sandstones. where cold water is upwelling, e.g. along the coast of
In continuously subsiding basins, open faults and West Africa.
fractures will be rare because of the progressive com- Reefs build up on the seafloor and may be buried
paction processes. Shales may however fracture at beneath mud during transgressions. The reef then
high overpressures (fracture pressure) near the top of becomes a perfect stratigraphic trap, often with good
structural highs, but that may be separate from the permeability both vertically and horizontally.
fault planes. Reefs can form long continuous barriers as in
During uplift and erosion (exhumation, unloading) Australia (Great Barrier Reef). In the US much oil
the rocks will be subjected to extension, and exten- was found by following Jurassic and Cretaceous reef
sional fractures will be produced. The porosity will not trends around the Gulf of Mexico.
increase significantly but the extensional fractures will High energy beach deposits on carbonate banks
increase the permeability and thus improve the reser- may consist of well-sorted carbonate sand
voir properties. Unfortunately the cap rock may also (grainstones). Ooid sands (ooliths) are formed as
fracture during unloading, causing leakage from the beach and shoreface deposits and may have limited
reservoir. vertical thickness, reflecting the wave base. They may
To understand the properties of reservoir rock we however stack up and form thicker sequences of such
need to integrate what we know about the rocks. Ooids may also be transported from the shelf
22 K. Bjørlykke
into deeper water as turbidites and other slope oil here, but merely look briefly at some of the most
deposits, but the sorting and reservoir quality is then important principles.
reduced. When drilling commences at the surface, the diam-
Ooids are rather stable mechanically during burial, eter of the well may be 20"–30" (50–75 cm), but
but at 2–3 km porosity may be strongly reduced by decreases downwards to 3"–6" (7–15 cm) at great
cementation in the intergranular pore space. The cal- depths. Normally a roller bit is used, which crushes
cite cement may be derived by dissolution along the rock into small pieces (about 2–5 mm) called
stylolites. Stylolites have a thin layer of clay and cuttings. Core samples are only taken when drilling
other minerals that are not soluble and may present a through especially important rock strata (usually res-
barrier during oil migration and production. ervoir rocks) where large intact samples are needed for
Carbonate muds are very fine grained and do not detailed examination. A circular diamond core drill bit
have high enough primary porosity and permeability must then be used. This takes time and costs a lot more
to form reservoir rocks, but may gain porosity and per running metre, as the entire drill string has to be
permeability by fracturing and become fractured recovered to get each core section to the surface. Only
reservoirs. They can also have mouldic porosity from the most critical sections are therefore cored.
dissolved aragonite fossils. Drilling mud is pumped down through the drill
string into the well during drilling. This mud has
several functions. When one drills several hundred or
1.14 Drilling for Oil and Gas a 1,000 m down into rock, one encounters water, gas
or oil which may be under high pressure. The drilling
Drilling for oil is a costly process, especially offshore. mud acts as a counterweight to prevent the uncon-
The object of a well is to prove the presence of, or trolled gush of water or petroleum into the well and
produce, oil or gas. Sometimes wells are also drilled to up to the surface in a blow-out. The pressure exerted
inject water, chemicals or steam into the reservoir by the drilling mud must exceed the pressure of oil and
during production. Even a well which fails to find water in the surrounding formation. Heavy minerals
hydrocarbons (a dry well) is still of great value, such as barytes are frequently added to increase den-
because of the information it provides about the sity; the main components of drilling mud are mont-
rocks in the area. This information forms part of the morillonite (smectite) containing clays, with a large
basis for the geological maps and profiles which are number of different additives. The drilling mud also
used in further exploration for oil and can be sold or serves to cool the drill bit, and cuttings are brought
exchanged for data from other companies. This is the back to the surface suspended in the circulating mud.
reason why oil companies wish to keep the geological The cuttings are then washed out from the drilling mud
results of oil drilling confidential for some years after a onto a sieve (shale shaker) and the mud can be used
well has been completed. again.
The first well in a new area is called a wildcat well, The cuttings are continuously analysed on the dril-
while appraisal wells are drilled to estimate the exten- ling platform by a geologist who logs the composition
sion of an oil field. They may also become production of the cuttings, making a preliminary description of
wells. Stratigraphic wells are drilled mainly to obtain the rocks which are being drilled, and their mineral-
stratigraphical information from the basin. ogy. Samples of cuttings are usually taken every
Oil drilling used to be carried out largely on land, 10–30 ft drilled. More detailed analyses are carried
but now offshore drilling takes place not only on our out in the laboratory. The cuttings are often poorly
continental shelves, but also in deep water (1–3 km). washed and need extra cleaning to get rid of the
This type of drilling is many times more costly than drilling mud; in weakly indurated mudstones it may
drilling in shallow water or on land. This has led to be difficult to separate the drilling mud from the soft
increased efforts to gain maximum information from cuttings. Organic additions to drilling mud may cause
wells. The cost of analysing samples and logs is small problems when analysing cuttings, and sometimes oil-
in relation to the cost of drilling the well. We shall not based rather than water-based drilling mud is used,
go very deeply into the technical aspects of drilling for which confuses analyses for oil. The fossil content of
the rock fragments, largely microfossils, is used to too expensive. Horizontal drilling has revolutionised
determine the age of the strata (see Biostratigraphy). oil production.
Not all the cuttings which come up with the drilling It is now possible to follow thin oil columns
mud have necessarily come from precisely the strata laterally and to make complex wells to drain different
being drilled through at that time. There may also be compartments in the reservoir. Large oil fields with
contamination due to the caving in of overlying strata thin oil columns like the Troll field in the North Sea
into the rising drilling mud. This means we can find would have been difficult to produce without horizon-
material from younger rocks with a different composi- tal drilling. Horizontal and deviated wells may extend
tion mixed in with the formation being drilled, 8–9 km from the drilling platform, enabling produc-
together with younger fossils. This demands consider- tion from relatively small reservoir compartments
able care when making stratigraphic interpretations away from the platform.
based on microfossils identified in cuttings. The safest Earlier, geophysics was used mainly to define
way is to register the first occurrences of a species structures but the quality of the seismic data was not
when proceeding downwards from the top in the well. good enough to provide much detail. 3D seismic based
The last occurrence of a fossil may be the result of on 50 m line spacing allows the construction of a
cave-ins from younger strata. three-dimensional cube of geophysicial data, which
Since the pressure of the drilling mud is being provides much more detailed information.
monitored and adjusted to prevent oil and gas from By repeating seismic surveys during production at
penetrating into the well, significant oil and gas 2–5 year intervals, the effect of changes in the fluid
occurrences may be drilled through without being composition on the seismic data can be seen, thus
registered. This should be detected on the logs but it adding the time dimension to 3D seismics. 4D
may be advisable to carry out special tests in the most seismics has made it possible to follow the depletion
promising strata to find out if there is petroleum pres- of an oil or gas field by monitoring the GWC and often
ent, and in what quantities. also the OWC during production. By this means, parts
As drilling proceeds, the well is lined with steel of a reservoir that have not been drained by the pro-
casing to prevent rock and loose sediment falling into duction wells can be detected. If the isolated reservoir
it, but prior to casing, each section of the well has to be compartments are large it may be economic to drill
logged with different logging tools which require additional wells to drain them.
physical contact with the wall of the well. Radioactive
logs, however, can also be run after the casing has
been installed. It is useful to note when the different 1.15 Oil Reserves – How Long Will
casings are installed, because that limits the strata Conventional Oil Last?
which could have “caved in” and contaminated the
cuttings. If the well is going to produce oil, a produc- In the last 40–50 years we have had a discussion about
tion pipe is used and installed running through the how long the oil reserves will last and the famous
petroleum-bearing strata. It is then perforated by geologist M. King Hubbert predicted in 1956 that oil
shooting holes in the steel casing (in the oil column) production would peak in the United States between
so that petroleum can flow into the well. 1965 and 1970, and later Colin Campbell predicted
For an oil field to be capable of full production, that world production would peak in 2007 (see the
several wells are normally required. Peak Oil movement). There is, however, a great deal
Up to 1990–1995 most of the exploration wells and of uncertainty in the estimation of reserves since this
production wells were nearly vertical. Horizontal will depend on advances in exploration technology as
barriers due to changes in facies or faults could then well as production efficiency.
be critical barriers during production. It was difficult The price of oil and taxation policies will also
to produce oil from thin sandstones or carbonate beds determine which type of oil accumulations can be
because of vertical flow of water from below or gas exploited economically and thus reckoned as reserves.
from above. Onshore this could be compensated for by At present the world reserves are estimated to be
having a dense well spacing but offshore that would be 1:2 1012 bbl 2 1011 m3 and nearly 60% of this is

24 K. Bjørlykke
located in the Middle East. Europe has only got 1% A broader background in geological and engineering
and the whole of Asia about 3%. disciplines will also be required to reduce the environ-
Canada had up to recently very small reserves but mental problems with the exploitation of fossil fuels.
after the heavy oil and tar sand in Alberta was included Storage of carbon dioxide requires expertise from
it has now nearly 15% and is second only to Saudi petroleum geologists.
Arabia. The exploitation and burning of fossil fuels releases
If oil shales were to be included, however, the USA large amounts of CO2 into the atmosphere which is an
would be the country with the largest reserves. addition to what is part of the natural carbon cycle
The world consumes about 90 million bbl/day; the (Fig. 1.14). The CO2 in the atmosphere is dissolved in

US consumption ð20:7106 bbl=day makes up nearly seawater to H2CO3 and then precipitated as carbonate.
25% of that. Another part is taken up by plants, including algae,
The most important producers are Saudi Arabia and may be stored as reduced carbon.
(10.2 106 bbl/day) and Russia (9.9 106 bbl/day). The total amounts of calcite precipitated is equal to
The US is also a major producer and has increased the amounts of Ca++ released by weathering of silicate
from 7.5 million bbl/day in 2008 to 11 million barrels/ rocks (e.g. plagioclase) and transported into the ocean
day in 2014, mainly from oil shale production, and by rivers.
now provides for >50% of the country’s consumption.
China has become a major importer of oil with a
consumption of 9.0 106 bbl/day, while their produc- 1.16 The Future of Petroleum
tion is 4 106 bbl/day. Geoscience
It is clear that consumption of oil in Asia will rise
and it will be very difficult to meet this demand. Petroleum geoscience is geology and geophysics
Norway’s oil production is about 1.9 106 bbl but applied to petroleum exploration and production. In
the domestic consumption is comparatively small this book we will try to show the wide range of
(0.22 106 bbl/day), so Norway is a major exporter. disciplines that are relevant and useful for this
Gas production has increased, supplying more than purpose.
20% of Europe‘s natural gas consumption. Many of the disciplines in the geosciences are
It will probably be difficult to meet the demand for highly specialised and there is often too little commu-
conventional oil in the next decades. There are, however, nication between the different fields. Most researchers
very large reserves of fossil fuels in terms of gas, heavy naturally focus on a very small area because of the
oil, tar sand and also coal. All these types of fossil fuel can requirements with respect to methods and analytical
be used for heating and transport. Particularly in North techniques, and the demands of following the litera-
America there is much oil shale and also gas shale. Gas in ture. Applied petroleum geoscience requires a broad
fine-grained siltstones and shales is expected to be a overview of substantial parts of geology and geophys-
major source or energy in the years to come. ics and provides good training in the integration of
In recent years coal methane and shale gas have different types of data and models. These skills are
become important sources of such energy. Gas in solid also applicable in many types of environmental
form (gas hydrates) may also represent a future source research and when solving practical environment
of hydrocarbons. There are, however, many environ- problems.
mental problems connected to the production and The petroleum industry employs a large percentage
utilisation of these resources and this represents a of the world’s geologists and geophysicists and funds
great challenge, including for geoscientists. much of the research in this field.
It will probably take a long time before fossil fuels Most of the obvious petroleum-bearing structures
can be replaced by other sources of energy. As the have already been found in the explored sedimentary
demand increases, oil exploration and production will basins and there are now rather few areas that have yet
become more and more sophisticated technologically to be explored seismically and by drilling. The large,
and also geologically. easy-to-find structures did not usually require very
advanced methods and geological skills.
Geochemistry of carbonate minerals
CO2 Dissolution of
carbonate rocks
CO
Photosynthesis
CO2 Oxidation of
organic matter
Transport of Delta
CO2 exchange
organic matter Addition of organic matter
by rivers + nutrients to the ocean
Organic Photosynthesis
plant matter Breakdown
in delta CO2
of biogenic
sediment carbonate
Oxidation of
CO2 organic matter
Deposition
of carbonate
CO2 Ocean floor
Oxidation and
biological processing Deposition of
of organic material organic material
C
Kerogen in sediments
CO2
Carbon in
sedimentary rocks
Fig. 1.14 Illustration of the carbon cycle. Carbon from organic is limited by the supply of Ca++ and Mg++ from weathering of
matter and carbonate rocks are the major sinks for carbon (CO2). silicate rocks brought in by rivers
The rate of precipitation of carbonate in the ocean by organisms
More and more sophisticated methods are therefore

used in modern exploration. We are on a global basis 1.17 Conventional and Unconventional
not finding enough new oil fields oil to replace the Oil and Gas
produced oil. Global reserves have however not
changed very much because of higher estimates of Reservoir rocks have traditionally been defined as
recovery from existing fields and because unconven- rocks with sufficient permeability and porosity for
tional oil like tar sand is now included in the reserves. petroleum to be produced economically by drilling
There is a major challenge for geoscientists to wells. They have most commonly been sandstones
develop ever better exploration methods and to opti- and limestones but fractured basement rocks may
mise production. also serve as reservoirs. Sandstone reservoirs normally
Even if the global production of conventional oil have porosities >10–12% and permeability >0.01
may be reduced there will be significant production for Darcy. Gas may be produced from tighter rocks. Petro-
many decades. This will mostly come from the tail leum generated in organic-rich source rocks must have
production of giant fields and from small reservoirs, migrated to a reservoir rock to be produced as conven-
but it is rather labour intensive. This is also the case tional oil and gas.
with unconventional oil (tar sand and oil shale) and Exploration is then limited to structural traps or
also tight gas reserves and shale gas. pinch-outs trapping the hydrocarbons.
Until enough alternative sources of energy are In many mature sedimentary basins such traps have
developed it is necessary to extract fossils fuels from already been explored and produced.
these resources. This should be done with as little In the last 5–10 years it has however been possible
environmental damage as possible and this requires a to produce oil and gas directly from shales which are
new generation of highly skilled geoscientists. source rocks even if their porosity and permeability
26 K. Bjørlykke
WELL FOR Cap rock

Conventional Tar
UNCONVENTIONAL OIL
reservoir (Shale) sand
Well AND GAS
(trap)
Heavy
oil
Shale gas/
Shale oil
Fig. 1.15 Simplified illustration of the difference between produced by vertical wells. After uplift and erosion to about
conventional and unconventional oil and gas. Conventional oil 1.5–3 km depth, near-horizontal wells can follow shales and
and gas is limited to structural closure or other traps and can be produce from large parts of the basin
are rather low. This has increased gas and oil produc- heated to 400–500 C to generate oil. This requires
tion dramatically, particularly in the US, and is often large amounts of water and produces very great
referred to as unconventional oil and gas. Unconven- volumes of waste which is difficult to store.
tional oil and gas however can be produced from Heavy oil and tar sand is found at rather shallow
shales covering much of the basins, independent of a depth where oil has been biodegraded and become
structural trap (Fig. 1.15). This has been possible using highly viscous (Fig. 1.16). Bacteria capable of break-
long horizontal wells and methods to fracture shales ing down oil exist at the surface and in sediments
by injecting water at high pressure. Many of the shales which are buried to depths with temperatures less
producing oil and gas in the US have a significant than about 80 C. When oil migrates into sandstones
content of silt and also carbonate cement, which and limestones at these depths bacteria can start eating
influences their fracture properties. the oil; the light components will be consumed first,
Shales representing source rock buried to 3-4 km leaving the heavy component.
(>120–140 C) will contain some oil which has not In reservoir rocks which have been buried more
been expelled. After uplift to 1–2.5 km depth they deeply (T >80 C) bacteria have not survived and
have been unloaded and fracture more easily because these rocks are pasteurised. Oil can then be preserved
of differential stresses. Sand or other granular material for a long time at relatively shallow depth (a few
is used to prevent the fractures from closing. hundred metres) without being biodegraded (Fig.
Organic-rich shales that have been buried to 1.18).
120–150 C have generated oil which has migrated Such oil may be produced by injection of steam to
upwards to a reservoir rock or to the surface. Some heat the oil to make is less viscous so that is can flow to
of the oil generated is however not expelled and can be the wells. Tar sand may also be excavated and the oil
produced as unconventional oil. Shales that have been is then separated from the sand by using hot water.
buried to greater depth (>150–180 C) contain mostly Steam and hot water require energy which in most
shale gas. (Fig. 1.16). Some of the injected water will cases is produced by burning oil or gas which results
be produced with the oil and gas and this water may be in increased CO2 emissions.
highly saline and contain toxic components, causing Organic matter rich in plant material (humic kero-
environmental concerns that it may pollute the gen Type III) undergoes thermal alteration to brown
groundwater. coal (lignite) and coal (Fig. 1.17). This type of kerogen
Shales that have not been buried deeply enough for (vitrinite) changes from dull to more shiny material
the organic matter (kerogen) to mature into oil or gas with increasing temperature. Under the microscope
are called oil shales. These must be excavated and the percentage of reflected light is a measure of the
Organic matter - mostly algae

Surface (0)/25°C Geological time
Tar sand
Shallow gas
Heavy oil
Shale gas
Shale oil
Oil-
3 km/100°C
shale
Oil Oil window

4-5 km/150°C
Condensate
Gas
6-8 km/200°C
Depth /Temperature
Fig. 1.16 Alteration of organic material, mostly algae, with increasing depth and temperature. Shallow gas is formed at shallow
depth (low temperature) by bacteria. In oil shale the temperature has been insufficient to generate petroleum from the kerogen
Plants
Surface (0)/25°C Geological time
Turf, bog
Coal bed methane

Brown
coal
Lignite
3 km/100°C
Coal
4-5 km/150°C
Anthracite
6-8 km/200°C
Depth/Temperature
Fig. 1.17 Alteration of plants and woody material with increasing burial, temperature and time
thermal alteration in a sedimentary basin (vitrinite migration. This involves modelling based on the kinet-
reflectivity). ics and temperature history.
Before drilling, the position of a trap and a reservoir
rock has to be determined.
1.18 Summary The porosity and permeability of the reservoir rock
is critical. If the porosity is 30% it may at the best
In the exploration for oil and gas we need to predict the contain about 250 l of oil and 50 l of water for each m3
occurrence and distribution of source rocks which is a of rock, and assuming 50% recovery, 125 l of oil can
function of the sedimentary environment and climate be produced. If the porosity is only 15% it may contain
at the time of deposition. The burial history has to be about 100 l of oil and the recovery may then be lower
reconstructed to predict the timing of maturation and so that less than 50 l will be produced.
28 K. Bjørlykke
Effect of heang (pasteurisaon) of reservoir

rock prior to oil migraon and filling
Seafloor
or land
surface
Sand reservoir Biodegraded oil/tar sand

with some B Reservoir
quartz cement
loose sand
A 80° C
Normal oil - little

biodegradation.
Reservoir rock Reservoir A may aer
Migration
of oil from pasteurised at upli and erosion contain
deeper temperatures >80° C oil which is not very much
source biodegraded even at
rocks relavely shallow depth .
Reservoir B may be an
exposed tar sand.
Fig. 1.18 Pasteurisation effects as a function of maximum burial temperature on migrating oil. Oil migrating into reservoirs that
have not been buried deeply enough to be pasteurised will contain bacteria and is likely to be biodegraded relatively fast
The quality of reservoir rocks depends on the depo- refined over many years. This book will include an
sitional environments and the primary mineralogical introduction to geophysical methods and also 4D
and textural composition, and on the diagentic pro- seismic methods. Interpretation of geophysical data
cesses that change the reservoir properties during requires that the physical properties of sedimentary
burial. rocks are known or can be predicted. We have there-
For sandstone reservoirs the distribution and geom- fore put some emphasis on rock mechanics and rock
etry of sand is critical because much of this is below physics.
the resolution of seismic methods, and even after dril- Unconventional oil such as heavy oil, tar sand,
ling the information from each well has to be shale oil and shale gas represents new challenges and
extrapolated in 3 dimensions. requires also different training and background from
For this reason I have included a chapter on sedi- that of the conventional petroleum geologist. The
mentary structures and sedimentary facies and also on organic chemistry of petroleum and the detailed min-
carbonates. There are, however, many textbooks eralogy of the rocks become increasingly important.
which will give a more detailed presentation of clastic Estimates of global petroleum reserves vary greatly
sedimentology. To be able to predict the reservoir with time and are dependent on the price of oil and gas
properties we must understand the principles of sand- and other energy sources and also on exploration and
stone diagenesis and this is discussed in Chapter 4. production technology. There are also very large
This also involves chemical reactions driven by ther- reserves of oil shale e.g. the Eocene Green River
modynamics and kinetics. Reservoir quality can to a shale (in the southern US) which must be heated
certain extent be modelled as a part of basin before oil and gas can be extracted.
modelling, if the primary mineralogical and textural It is however clear that it is becoming more difficult
composition is known. to find new reserves to compensate for production and
Exploration and production of conventional oil has that there are relatively few remaining unexplored
become well established and technologies have been sedimentary basins. Production of oil and gas from
unconventional sources represents a serious environ- Further Reading

mental challenge to limit local pollution and also the
release of CO2.
Much updated information can be found on the Allen, P.A. and Allen, J.R. 2005. Basin Analysis. Principles and
Applications. Blackwell, Oxford, 549 pp.
Internet: Beaumont, E.A. and Foster, N.H. 1999. Exploring for Oil and
International Energy Agency(IEA) www.iea.org Gas Traps. Treatise of Petroleum Geology. AAPG Special
US geological survey (http://www.usgs.gov/) Publication 40, 347 pp.
Peak oil (http://peakoil.com/) Biju-Duval, B. 1999. Sedimentary Geology. Sedimentary
Basins, Depositional Environments, Petroleum Formation.
Norwegian Petroleum Directorate (http://www.npd. Institut Francais du Pétrole Publication, Rueil-Malmaison
no/) Editions Technip, 642 pp.
Geological Society of London (http://www.geolsoc. Glennie, K.W. 1998. Geology of the North Sea: Basic Concepts
org.uk) and Recent Advances. Blackwell, Oxford, 636 pp.
Groshong, R.H. 1999. 3D Structural Geology. A Practical Guide
IFP (http://www.ifp.fr/) to Surface and Subsurface Interpretation. Springer, New
Units for petroleum reserves: York, 324 pp.
1 Sm3 (One standard cubic metre) ¼ 6.293 bbl Hunt, J.M. 1996. Petroleum Geochemistry and Geology. Free-
1 Sm3¼ 35.3 standard cubic feet. man and Co., New York, 741 pp.
Miall, A.D. 1996. The Geology of Fluvial Processes, Basin
1,000 Sm3 of gas ¼ 1 Sm3 o.e Analyses, and Petroleum Geology. Springer, New York,
1 Sm3 of oil ¼ 1 Sm3 o.e 582 pp.
1 tonne of NGL ¼ 1.3 Sm3 o.e Salvador, A. 2005. Energy: A Historical Perspective and 21st
o.e – oil equivalents are used to sum up the energy of Century Forecast. AAPG Studies in Geology 54, 207 pp.
oil, gas and NGL (Natural Gas Liquids).
Chapter 2
Introduction to Sedimentology
Sediment Transport and Sedimentary Environments
Knut Bjørlykke
Sedimentology is the study of sedimentary rocks and sediments. We also have a genetic nomenclature,
their formation. The subject covers processes which which names rock types according to the particular
produce sediments, such as weathering and erosion, way in which we think they have formed. Examples of
transport and deposition by water or air, and also the these are fluvial sediments (which are deposited by
changes which take place in sediments after their rivers) and aeolian (air-borne) sediments. The descrip-
deposition (diagenesis). Diagenesis includes mechani- tive nomenclature is used as a basis for an interpreta-
cal sediment compaction and mineral reactions as a tion of how the rock was formed. When we are
function of temperature during burial. Changes in sed- reasonably confident about their origin we may use
imentary rocks at temperatures of over 200–250 C are the genetic nomenclature.
called metamorphic processes and are not dealt with Sedimentology covers studies of both recent (mod-
here. In this chapter we shall discuss primarily trans- ern) sediments and older sedimentary rocks. By study-
port and deposition of clastic sediments and sedimen- ing how sediments form today we can understand the
tary environments. These processes determine the conditions under which various sedimentological pro-
distribution and geometry of reservoir rocks in a sedi- cesses take place. From such observations we may be
mentary basin and also the changes in rock properties able to recognise older sediments which have been
during burial. Accumulation of organic-rich sediments formed in the same way. This is called using the
which may become source rocks is also an integral principle of uniformitarianism, which has been of
part of sedimentological models. Since both source great importance in all geological disciplines since
rocks and reservoir rocks normally are sedimentary its proposal by James Hutton (1726–1797).
rocks, sedimentology is a very important subject for Conditions on the surface of the Earth have
petroleum exploration and production. Sedimentolog- fluctuated widely throughout geological history and
ical research is to large extent funded by the petroleum the principle of uniformitarianism cannot be applied
industry. without reservation. One important aspect of sedimen-
Like all natural sciences, sedimentology has an tological research is attempting to reconstruct changes
important descriptive component. In order to be able in environments on the Earth’s surface throughout
to describe sedimentary rocks, or to understand such geological time. This applies particularly to climate,
descriptions, it is necessary to familiarise oneself with vegetation and the composition of the atmosphere and
quite an extensive nomenclature. There are specialised the oceans.
names for types of sedimentary structures, grain-size Palaeontology is important to sedimentology, not
distributions and mineralogical composition of only for dating beds, but also because organisms are an
important component of many sedimentary rocks (par-
ticularly limestones), and organic processes contribute
to weathering processes and the precipitation of
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway dissolved ions in seawater. Many organisms make
e-mail: [email protected] very specific demands of their environment, and

32 K. Bjørlykke
fossils are consequently a great help in reconstructing electron microscope) or chemical analyses. Pure
the environment in which the sediments were depos- descriptions of sedimentary rocks are useful because
ited. Palaeoecology is the study of ecological they increase the data base on which we can build our
conditions as we are able to reconstruct them on the interpretations.
basis of remains or traces of plants and animals in Systematic studies of recent environments of sedi-
rocks. Traces of animals in sediments have proved to mentation are important to find connections between
be very useful environmental indicators. the environment and the sediments which accumulate.
Studies of recent and older sedimentary rocks pro- The environment governs the sedimentological pro-
vide a fruitful two-way exchange of information in cesses which determine what sort of sediments are
sedimentology. From studies of recent environments formed and deposited. The connection we are trying
we can learn about the conditions that particular pro- to understand in modern environments is thus: envi-
cesses require. In older rocks, however, we can study ronment ! process ! sediment.
sedimentary sections which encompass many millions Today a large number of modern environments of
of years of sedimentation, offering us a completely sedimentation have been studied in great detail. These
different record of the way sedimentological processes include aeolian and fluvial environments, deltas, beach
can vary as a function of geologic time. As a result, zones, tidal flats and carbonate banks. Deep sea
studies of older rocks also contribute to our under- environments have naturally not been so easy to
standing of the recent environment and offer non- study, but modern sampling and remote sensing equip-
uniformitarian explanations. ment and underwater TV cameras have made it easier
When we study rocks, we should attempt to give to gather observations from this environment, too. In
objective descriptions of the composition, structure recent years systematic drilling through sedimentary
etc. of the rocks, and on the basis of these try to layers on the ocean floor (Deep Sea Drilling Program –
interpret how they were formed. However, it is impos- DSDP, and Ocean Drilling Program – ODP) has
sible to give a completely exhaustive, objective provided an entirely new picture of the geology of
description of a rock. Nevertheless it is often most the ocean depths. Specially constructed diving ships
fruitful to have a theory or hypothesis against which (e.g. ALVIN) make it possible for geologists to
to test our observations. Data collection can then be observe the ocean floor directly at depths of up to
focused on observations and measurements which can about 3,500 m and take samples of surface sediments.
support or disprove the hypothesis. In addition geophysical, particularly seismic, surveys
We know from experience that we have a tendency provide one of the most important bases for under-
to observe what we are looking for, or what we antici- standing the stratigraphy and geometry of sedimentary
pate finding. basins.
Early descriptions of sedimentary sequences con- In studying older rocks we base our approach on
tain few observations about sedimentary structures certain features which we can observe or measure, and
which we would consider fairly conspicuous and attempt to interpret the processes that produced them.
important today. We observe sedimentary structures Particular variations in grain size and sedimentary
because we have learned to recognise them and under- structures in profiles can be interpreted as having
stand their genetic significance. been formed through particular processes, e.g. aeolian,
Many sedimentologists use a standard checklist for tidal or deltaic processes. The recognition of such
what they should observe in the field, so that their sedimentary processes helps us to reconstruct the
descriptions are as comparable as possible. Neverthe- environments. The sequence of interpretation in stud-
less, it is important that field observations do not ies of older sedimentary rocks is thus: description of
become too much of a routine. Facies analysis should sedimentary rock ! processes ! environment.
be a creative process and the various depositional Applied geology has always been important, even
models should be kept in mind when making the for purely scientific research. The interests of eco-
observations. It is also desirable to quantify field nomic recovery of raw materials from sedimentary
observations as far as possible, for example by surveys rocks create a demand for sedimentologists and sedi-
in the field and a range of laboratory analyses. These mentological research. Exploration for and recovery
might be texture analyses (e.g. grain-size distribution), of raw materials also provide important scientific
microscope analyses (perhaps using a scanning information. Sedimentary rocks contain raw materials
2 Introduction to Sedimentology 33
of considerably greater value than those we find in The depositional environment determines to a large
metamorphic and eruptive rocks. The most important extent the shape and extension of potential reservoir
are oil, gas and coal deposits, but a very large amount rocks such as sandstones and limestones. Sequence
of the world’s ore deposits is also found in sedimen- stratigraphy of the sequences filling a sedimentary
tary rocks and many types of ore have been formed basin often provides the key to predict the distribution
through sedimentary processes. Limestone, clay, sand of source rocks, migration pathways and traps. The
and gravel are also important raw materials which internal properties of reservoir rocks such as porosity,
require sedimentological expertise. density and velocity depend however very much on
The petroleum industry employs a very high per- the mineralogical and textural composition at the time
centage of the world’s professional geologists. This of deposition.
industry has a particular need for research, and also Sandstones may vary greatly with respect to grain
has the financial capacity to invest in it. Because oil size, feldspar and clay content and compact very dif-
and gas are found largely in sedimentary rocks, explo- ferently. The source of clastic minerals (provenance)
ration for and recovery of hydrocarbons is based to a is an important part of basin analyses which will help
large extent on sedimentology. Much of what we now to predict rock properties at greater burial depth.
know about the world’s sedimentary basins and their Mudstones and shales should also be analysed with
regional geology is derived from seismic profiles respect to grain size (silt content) and clay mineralogy
which have been shot in connection with oil explora- because different types of clay and mud are important
tion and drilling for oil and gas. The oil industry has parts of sedimentary sequences, particularly as source
also helped to stimulate pure sedimentological rocks and cap rocks.
research, and significant contributions to research in Limestone rocks are to a very large extent com-
this area are published by the research laboratories of posed of material precipitated biologically and the
the oil companies. Research based on economic types of marine organisms forming reefs and other
interests is also useful from a purely scientific point carbonate reservoir rocks reflect to a large extent the
of view, because it often focuses on particular biological evolution through time.
questions which may be quite fundamental. Also in the clastic sediments like sandstones and
Petroleum geology requires close teamwork shales biologically precipitated carbonate and also
between reservoir engineers and geologists, to estab- silica are important components controlling rock
lish in great detail the geometry and distribution of properties. Quartz-rich sandstones may have very little
porosity and permeability in reservoir rocks. We also porosity because of carbonate cement and calcareous
need very much to know more about the physical and mudstones and shales may be very tight and brittle.
chemical properties of reservoir rocks, for reasons This also influences the seismic data recorded in sedi-
which are discussed at the end of the book. Geophys- mentary basins.
ical methods provide most of the information used in Sedimentology and facies analyses represent an
petroleum exploration and production and many petro- important basis for the different petroleum related
leum geologists rarely examine real rocks in cores and disciplines discussed in the following chapters in this
cuttings. It is however important to know something book.
about the textural and mineralogical composition of
the sedimentary sequences. The geophysical data
rarely provide unique solutions when inverting seis- 2.1 Description of Sedimentary Rocks
mic and log data to rock properties. Facies models
provide ideas about how different lithologies may be 2.1.1 Sediment Composition. Mineralogy
distributed in a sedimentary basin. With few wells far and Other Components
between it is difficult to make a realistic representation
in 3D. The diagenetic processes which change rock Clastic sediments consist of grains and fine particles
properties after deposition are equally important. Sed- which may be composed of single minerals, rock
imentary facies and the primary sediment composition fragments and also some amorphous compounds
represent however the starting material for burial dia- such as organic matter, biogenic silica and volcanic
genetic reactions determining the rock properties. glass. The sediment composition is a function of the
34 K. Bjørlykke
rocks being eroded or weathered (provenance), sedi- downwards. A sample is put in the uppermost sieve
ment transport, and depositional environments. The and the bank of sieves is shaken (Fig. 2.1a). Grains
grains-size distribution is very much a function of which are larger than the mesh size will remain, while
the mineralogy in the source rock, and shales and smaller grains will fall through and perhaps remain
volcanic rocks will produce clay-rich sediments. lying on the next sieve. By weighing the fraction of the
Rapid erosion of granite and gneisses produces sample which remains on each sieve, we can construct
feldspar-rich sand (arkoses) while weathering results a grain-size distribution curve. The lower practical
in sand corresponding to the quartz crystals in the limit for sieve analyses is 0.04–0.03 mm; finer
granite and clay composed of kaolinite from particles exhibit much more cohesion, which makes
weathering of feldspar. Volcanic rocks produce smec- it difficult for them to become separated and pass
tite-rich clays which are very fine grained and have through the finer sieves.
cohesive properties very different from kaolinitic Fine silt and clay fractions can be analysed in a
clays. The grain-size distribution and the sand/shale number of ways. Most classic methods are based on
ratio as well as the composition of clay play an impor- measurements of settling velocity in liquids, and are
tant role in the depositional processes. based on Stokes’ Law:
Descriptions of outcrops and cores should not be
limited to studies of lithology and sedimentary v ¼ cgR2 Δρ=μ
structures but should also include sampling and
analyses of the textural and mineral composition by Here c is a constant (2/9) and μ is the viscosity of the
microscope (optical microscope and SEM) and by X- water. R is the radius (cm) of the grain and Δρ is the
ray diffraction (XRD). This will provide a basis for the density difference between the grain and the fluid (water).
interpretation of diagenetic reactions occurring during When the settling velocity of grains (falling
progressive burial in a sedimentary basin and also the through water, for example) is constant, the resistance
distribution of pore space between the minerals. to the movement (friction), which acts upwards, must
be equal to the force of gravity, which acts downwards
(Fig. 2.2).
2.1.2 Textures
6πRvμðfrictionÞ ¼ 4=3πgR3 ΔρðgravityÞ
The textures of clastic sediments include external
characteristics of sediment grains, such as size, shape v ¼ c gR2 Δρ=μ
and orientation. These properties can be described
relatively objectively and say a great deal about the Log v ¼ 2 log R þ cða constantÞ
origin and conditions of sediment transport and
deposition. The settling velocity is sensitive to temperature
By grain size we normally mean grain diameter, but variations, which affect the viscosity of the water (μ).
the two are only strictly synonymous in the case of We can measure the settling velocities of sediment
completely spherical particles. Most grains are not grains indirectly by measuring the density of the water
spherical, however, and it is difficult to identify a with suspended sediment sample with a hydrometer,
representative diameter, particularly in the case of which registers the fluid density. We disperse the
elongated or flat grains. For this reason we have sample in a cylinder with a mixer so that at a start
adopted the concept “nominal” diameter (dn), defined time To we have an even distribution of all grain sizes,
as the diameter of a spherical body which has the same and therefore of density, throughout the cylinder. The
volume as the grain. In practice we are seldom in a individual sediment grains then sink to the bottom at a
position to measure the volume of individual grains, rate which is a function of their size. The change in the
and we therefore use indirect methods to measure the fluid density as progressively fewer grains remain in
distribution of grain size within a sample. suspension is therefore a function of the grain-size
Sand and gravel can most simply be analysed by distribution. This applies for small particles, where
means of mechanical sieving. A bank of sieves the flow of the liquid around the grain is laminar and
consists of sieves with mesh sizes which decrease the concentration of grains is low.
a Sediment sample
X1
Weight
X2
X3 Sieves with
different
X4 mesh sizes Sediments Time
depositing Cumulative
X5 on the balance grain-size
distribution
X6 curve
Amount of sediment on Sedimentation balance

Shaker each sieve X1 − X6
Sieve analysis
b Wentworth Scale
mm j = – log2d
Boulder
256 –8
Cobbles
64 –6
Pebbles
4 –2
Granules –1
2
1 V. Coarse sand
Coarse sand +1
0.5
Medium sand
Sand 0.25
Fine sand
0.125
V. Fine sand +4
0.0625
Silt
0.004 1 )
( 256 +8
Clay
Fig. 2.1 (a) Sketch showing the principles involved in sieve velocity, i.e. weight increase as a function of time. This is
analysis and use of a sedimentation balance. Sieve analysis is therefore a cumulative grain-size distribution. (b) Grain-size
usually used for grain sizes down to 0.03–0.02 mm, but with classification of clastic sediments. The grain size (d) is often
wet-sieving even finer sediment grains can be sieved. The sedi- described in terms of φ values (φ ¼ –log2d)
mentation balance gives us a direct expression of settling
When the grains are larger than about 0.1–0.5 mm, The rate of density reduction in the suspension is a
the settling velocity increases, and turbulence function of the grains’ size. By taking successive
develops around the grains. The frictional resistance readings of the density we may plot a curve which
therefore increases, and in the case of larger grains expresses density reduction as a function of time.
(>1 mm) the settling velocity increases approximately Since density reduction is a function of settling veloc-
in proportion to the square root of the radius. It is not ity, this curve can be calibrated to give a grain-size
practicable to measure each grain, but the settling distribution curve.
velocity can be measured indirectly. A hydrometer When we analyse fine-grained sediments with a
floating in a suspension of sediments and water large clay fraction, or separate out clay fractions, it
measures the density of the suspension through time. may be useful to use a centrifuge. We then increase the
36 K. Bjørlykke
2.2 Grain-Size Distribution in Solid

Rocks
Resistance due to Lightly-cemented sandstones can be disintegrated by

friction. Laminar flow; means of ultrasound in the laboratory and then analysed
6πRvμ (Stokes law) as loose sediment in the normal manner. Carbonate-
cemented rocks may be disintegrated using acids. How-
R ever, we must bear in mind that new clay minerals may
have formed through post-depositional alteration (dia-
genetic processes), and that some of the original
Force of gravity:
minerals may have been broken down mechanically or
4/3πR3Δρ dissolved chemically. Consequently it is not certain that
we are dealing with the original grain distribution.
Diagenesis must be taken into account.
Well-cemented rocks must be analysed in thin sec-
tion by means of a petrographic microscope. It is
difficult to analyse the finer fractions (fine silt and
Fig. 2.2 The velocity of a falling grain in water is controlled by clay) in this manner and we must always remember
the gravity forces directed downwards and the resistance to the the “section effect”, i.e. that in most cases we will not
flow around the grain which is directed upwards (6πRvμ). The be seeing the greatest diameter of the grains. With
force of gravity is a function of the volume of the grain
(4/3πgR3) and the density difference (△ρ) between the grain
spherical grains, the relation between the real diame-
and the fluid (ρg – ρf) ter, dr, and the observed diameter, do, can be expressed
statistically: dr ¼ 4do =π.
acceleration term g in Stokes’ Law. There are also

“sedimentation balances”. Sediments suspended in a 2.3 Presentation of Grain-Size
cylinder fall through the water column and accumulate Distribution Data
on a balance pan at the bottom of the cylinder. This
balance records and writes out the increase in weight, Grain-size distribution is one of the many types of
which is the precipitation from suspension, as a func- natural data which must be presented on a logarithmic
tion of time. This gives a direct cumulative curve. scale for convenience. Wentworth’s scale is based on
Other methods are based on the refraction or dis- logarithms to the base 2, and this is now the one most
persion of a laser beam passed through suspensions widely found in geological literature (Fig. 2.1b).
producing a characteristic “scatter” which is calibrated For the sake of convenience, these data are com-
against samples of known grain size. These machines monly plotted against a linear scale. The phi (φ) scale,
use very small samples and have a high degree of where φ ¼ log2 d, allows convenient interpolation
repeatability. Equipment has also been developed of graphic data. The reason this negative logarithm is
which uses X-rays instead of light to produce the used is that normally most of the sediment grain
characteristic scatter patterns. diameters (d) are less than 1 mm, so these will have
It is important to note that no method measures the a positive phi value (Fig. 2.3). It is convenient to plot
nominal diameter. In methods which measure settling grain-size distribution data as a function of phi values,
velocity, grain shape is a significant factor. A large, especially on cumulative curves. In normal
thin mica flake has a settling velocity which descriptions of grain size, however, it is more helpful
corresponds to that of a considerably smaller spherical to state grain size in mm, so the reader does not have to
grain. The diameter of a spherical grain with the same calculate back from phi values.
volume and settling velocity is called the effective The simplest, and visually most informative, way
diameter (de). With the scatter method, flaky grains of presenting grain-size distribution data is by means
are assigned a different, probably greater, diameter of histograms (Fig. 2.3). These show the percentage,
than that indicated by the settling velocity method. by weight, of the grains falling within each chosen
a %
30
20
10
0
2 6 20 60 0.2 0.6 2 6 20
µ µ µ µ mm mm mm mm mm
b %
100
90
80
75
70
60
50
40
30
25
20
10
0
0.002 0.006 0.02 0.06 0.2 0.6 2 6 20 100
O25 Md M O75
(Sk) a
– +
(So)
c %
99
90
80
75
70
60
50
40
30
25
20
10
5
1
–3 –2 –1 0 1 2 3 4 5 6 (phi)
8 4 2 1 1/2 1/4 1/8 1/16 1/32 1/64 mm
Granules
25 75
Pebbles
Sand Silt
Fig. 2.3 Presentation of the same grain-size distribution data as logarithmic normal distribution, like a Gaussian curve, will
(a) histogram, (b) cumulative curve, and (c) cumulative curve plot as a straight line
on probability paper. When plotted on probability paper, a
38 K. Bjørlykke
0.01 2.4 Grain-Size Distribution Parameters

0.1
Fine-grained Phi ðφÞ ¼ log2 d (after Folk and Ward 1957) where
1 suspended
material
d is the grain diameter in millimetres (as previously
5
defined). The percentage of grains larger than a certain
Cumulative percentage
10
20 grain size (φ) is called the percentile. φ30 means that
30% of the grain population by weight is larger than
40 Well-sorted
60 sand the grain size. For φ ¼ 4 the grain size is 0.0625 mm so
70 that 30% of the sample is sand or larger grains.
80
90
95 Coarse
99.0 material 2.5 Significance of Grain-Size
(creep)
Fossils etc. Parameters
99.9 Phi ( )
99.99 The mean diameter is an arithmetically calculated
–8 –6 –4 –2 0 2 4 6 8 10 12
average grain size. The median diameter is defined
Sorting
by the grain size where 50% by weight of the sample
Fig. 2.4 Grain-size distribution curve presented as a function grains are smaller, and 50% are larger. Only in the
of grain size in φ values against a logarithmic cumulative case of completely symmetrical distribution curves
percentage will the mean diameter (M) and the median diameter
(Md) coincide. The mean will otherwise shift further
subdivision of the size range. It is then easy to see how than the median in the direction of the “tail” of the
well sorted the sediments are, and whether the distri- distribution. If the sample has a wide spread (tail)
bution of grain sizes is symmetrical, or perhaps bi- or towards the fine grain sizes (larger phi values) and a
polymodal, i.e. with two, or more, maxima. relatively sharp delimitation at the large grain-size
A cumulative distribution curve shows what per- end, we say that the sample has positive skewness.
centage of a sample is larger or smaller than a particu- This will be typical of fluvial sediments. There will be
lar grain size. The steeper the curve, the better the a fairly definite upper limit to the grain sizes that rivers
sorting. Note that engineers use the inverse term can transport as bedload, while there will be no sorting
“grading”, whereby well graded ¼ poorly sorted. of the fine fractions which are transported in suspen-
If we use probability paper, distributions which are sion. Major variations in flow velocity, for instance
lognormal (following a logarithmic distribution) will during floods, will give poorer sorting.
plot as straight lines and the slopes of these will reveal Aeolian (wind) deposits are very well sorted
the degree of sorting. Even if the whole distribution is (Fig. 2.5). They also have positive skewness because
not lognormal, it often appears that the curve can be there is an upper size limit to the grains which can be
regarded as a composite of 2–3 lognormal grain-size transported. Although the finest particles may be
populations. These populations generally overlap, so removed selectively, there will still be a “tail” of fine
that some sections of the curve represent a combina- material. The fine material in dunes may also be
tion of parts of two populations, each of which may be protected by a cover of larger particles (lag) against
lognormal. Each population may represent a different further erosion and transport. Beach sand deposits, on
mode of grain transport, for example saltation, rolling the other hand, are clearly negatively skewed, i.e. the
(bedload) or suspension (Fig. 2.4). distribution curve shows a definite lower limit, while
It is important that we collect representative there is often a “tail” of larger particles, i.e. granules
samples for grain-size distribution analysis, i.e. each and pebbles. The hydrodynamic conditions on a beach
sample only has material from one bed. This ensures are such that each wave brings some sediment in
that it represents deposition by a single sedimentary suspension. Whereas sand grains, particularly medium
process. If we take a sample at the boundary between to coarse sand, will rapidly settle from suspension and
two beds, we will often get false bimodal distributions be deposited on the beach again, fine sand, silt and
which can easily be mistaken for naturally produced clay will remain in suspension longer. This finer mate-
bimodal sediments, leading to interpretation errors. rial will be transported further out and at a depth of
+
Turbidites
River sand
Aeolian
Skewness 0
Beach
sand
– Sorting
0 1.0 (Standard deviation)
Better sorting Poorer sorting
Fig. 2.5 Sorting and skewness are grain-size distribution environments of deposition. Turbidites and fluvial sediments
parameters which are suitable for distinguishing between have a tail of small sediment particles while beach sand may
sediments deposited through various processes and in different have coarser grains such as pebbles
some metres (1–50 m), depending on how strong the sizes. Chemical weathering of acid rocks (granites),
waves are (and consequently the depth of the wave for example, will lead to the formation of sediments
base), we will have poorer sorted deposits because the consisting of quartz grains corresponding in size to the
fine fractions will be deposited and mixed with quartz crystals in the granite, and clay (kaolinite,
coarser-grained sediments which are carried out dur- smectite, illite) formed through weathering of the
ing storms. feldspars. Weathering of basic rocks (basalts and
Sediments deposited from suspension have poor gabbros) will produce almost exclusively clay
sorting and positive skewness. This is very typical of minerals, and practically no sand grains.
turbidites. There are however some turbidites that are The grain-size distribution in a depositional area
composed of relatively well sorted sand. They may also depends on the transport mechanisms carrying
represent beach or shelf sand which has been reworked sediment to the area. The grain-size distribution we
and deposited on a submarine slope. Clay suspensions observe in a sediment therefore reflects both the
which are deposited on land as high-density hydrodynamic conditions and the grain-size popula-
suspensions (mud flows), have negative skewness tion of sediments available from the source area. Fur-
because they often contain large clasts. thermore, although a particular grain-size distribution
Kurtosis is an expression for the spread of the may be characteristic for a type of deposit, it does not
extreme ends of a grain-size distribution curve in point unambiguously to a particular environment of
relation to the central part. This distribution parameter deposition because similar hydrodynamic conditions
is used somewhat less than the others, but relatively can exist in different environments. Statistical com-
small amounts of silt and clay can have a significant parison of high resolution (minimum ½ Phi sampling
effect on the properties of coarse-grained sand. interval) grain-size distributions, applying pairs of
Pebbles and stones in otherwise fine-grained sediment Folk and Ward or Krumbein parameters, can provide
may also be important. some diagnostic criteria for hydrodynamic
One of the most important points to bear in mind interpretations.
when interpreting grain-size distributions is the avail- Sediments may change their grain-size distribution
ability of grain sizes supplied to the area where the by diagenetic processes at quite shallow burial as well
process is taking place. Strong currents or wave energy as at greater depth. The formation of the clay minerals
will only be able to deposit coarse-grained sediments kaolinite and smectite from feldspar and rock
if there is coarse-grained sediment present in the area. fragments (volcanics) in a sediment certainly results
A source area where sediments are generated by ero- in very different grain-size distributions from those at
sion and weathering will often supply specific grain the time of deposition.
40 K. Bjørlykke
2.6 Grain Shape (or aerodynamic) principles involved. When a liquid

or gas flows in a channel or pipe it exerts a force (shear
We distinguish between three parameters: stress) against the walls or bottom. This force is
1. Roundness is a property of surface shape – whether counteracted by friction from the walls.
it is smooth or angular. A visual scale is most Pure water without suspended sediment is a New-
commonly used. tonian fluid which obeys Newton’s law:
2. Sphericity is an expression for how much a particle
deviates from a spherical form, and is defined as the τ ¼ μ dv=dh
ratio between the diameter of a circumscribed cir-
cle round the grain and the diameter of a sphere of A Newtonian fluid has no shear strength, so it will
the same volume (the nominal diameter). be deformed even by an infinitely small shear stress
3. We also use various expressions for grain shape ðdv=dhÞ.
such as (a) discoid or bladed for grains which are τ ¼ shear stress, which is an expression of force per
flat, (b) Prolate or roller for grains with one dimen- unit area (N/m2). μ is the dynamic viscosity expressed
sion considerably greater than the two others, in poise (g/cm/s or 0.1 N s/m2), dv=dh is the change in
(c) equant for grains with three relatively equal velocity (dv) or velocity gradient (deformation velocity)
dimensions and (d) oblate for grains with one as a function of distance from the boundary (dh). The
large, one medium and one small dimension. viscosity of pure water decreases with increasing tem-
4. Surface textures are concerned with the nature of the perature. Suspended material may also affect viscosity,
surface itself, whether it is rough, smooth, pitted, but the concentration of suspended material must be
scratched etc. Some textures are diagnostic of spe- quite high (15–25%) before the viscosity increases sig-
cific modes of transport, and superimposed texture nificantly. If the water contains a large percentage of
features may reveal the transport history of a grain. swelling clay minerals (smectite), however, the viscos-
The surface texture of grains can best be studied ity will increase at lower concentrations. The kinematic
under the scanning electron microscope. Aeolian viscosity v is the dynamic viscosity (μ) divided by
sand grains may develop fine pitting on their density ρ, i.e. v ¼ μ=ρ and units are cm2/s.
surfaces due to the collisions of grains during trans- We distinguish between laminar flow, where each
port, clearly visible under a binocular microscope. point in the liquid moves along a straight line parallel
Large grains become rounded far more rapidly than to the bed, and turbulent flow, where each point
smaller ones because the impact energy released in follows an irregular path so that eddies form
collisions with other grains declines in proportion to (Fig. 2.6). Reynold’s number (Re) is a dimensionless
the cube root of the radius. Blocks may be rounded number which describes flow in channels and pipes.
after only a few hundred metres or several kilometres It is defined as:
of transport. Grains less than 0.1 mm in diameter
undergo little rounding even when carried very long Re ¼ vhρ=μ
distances in water, for example by tidal currents.
The grain size and sorting of sand grains at the time Here v is the mean velocity, h is the depth of a
of deposition play an important role determining the channel or the diameter of a pipe in which fluid is
rate of compaction with increasing overburden stress flowing, ρ is the fluid density and μ its viscosity. If
(See chapters 4 and 6). The grain-size distribution may Reynold’s number exceeds a certain value, about
also change due to grain crushing during compaction 2,000, the flow changes from laminar to turbulent.
and chemical alterations e.g. dissolution of felspar and The density of water is 1 g/cm3 and the viscosity is
precipitation of clay minerals like kaolinite. one centipoise (0.01 Ps ¼ 0.01 g/cm/s). We see that
the boundary between laminar and turbulent flow
corresponds to 20 cm/s.
2.7 Sediment Transport This means that for the flow of water to be laminar
the product of velocity (cm/s) and depth (cm) must not
Sedimentary grains can be transported by water or by exceed 20. If the velocity is 1 cm/s, there will be
air. In order to understand the transportation processes turbulence if the depth (h) is greater than 20 cm. In
we must know a little about the hydrodynamic practice, then, flow in rivers and channels is always
dv/ h dv/
dh dh
h
Laminar flow Turbulent flow
Reynold’s number (R) = vhρ < c. 2000 Reynold’s number (R) = vhρ > c. 2000
µ µ
Air
Shear stresses against air

Max. water velocity about 0.6 − 0.7D
D
Mean velocity about 0.4D
Shear stresses against bed
Bed
Fig. 2.6 Diagram showing principles of turbulent and laminar flow and the shear stress against the underlying bed
turbulent. For turbulent flow the expression for shear Flow in a channel
stresses which applies to laminar flow, is no longer
adequate. The shear stress in turbulent flow will then F2 L Force of gravity = F1 = ρg.L.A.sin
F1
increase as a function of velocity because of the eddies Frictional forces = τ.P.L
A
which produce an eddy viscosity (η) (Fig. 2.6). The P
total shear stresses will then be: τ ¼ ðη þ μÞ=dv=dh. Flow in a pipe
H1 – H2
H1 sin = L
L
Darcy-Weisback equation:
2.8 Flow in Rivers and Channels H2 2
D sin = f V.
D 2g
For all types of water flow the forces acting on the water
must be in equilibrium. In most cases it is the force of Fig. 2.7 Flow of water in channels is controlled by the ratio
gravity which balances bed frictional forces. In order to between the gravitational forces and the shear stress against the
bottom of the channel
understand geological processes in connection with the
erosion, transport and deposition of sediments, it is
bed in a section along the channel. If the water flow
important for us to be aware of the relationships
has a steady velocity, the force of gravity F1 will just
which govern the flow of water in channels.
equal the frictional force F2 (Fig. 2.7). Consequently:
If the channel has a cross-section A and we look at a
stretch L of the channel, the force of gravity will be: τ L P ¼ ρg L A sin α
F1 ¼ ρ g L A sin α; or
where ρ is the density of water, g is the force of gravity A
(constant) and α is the angle of slope of the channel. τ ¼ρg sin α
P
The resistance to flow consists of frictional forces
against the bed and against the air. If we disregard A/P is the cross-section of the channel divided by
friction against the air, the frictional forces are: the wet perimeter, and we call this the hydraulic
radius, R. For flow in a pipe:
F2 ¼ τ L P
R ¼ D=4
where τ ¼ shear stress (force per unit area) and L P is
the area of the bed on which the forces are acting. P is The shear stresses (τ) R ¼ D/4 increase in propor-
the wet perimeter and L is the length of a line along the tion to the square of the velocity (τ ¼ c v2 ).
42 K. Bjørlykke
Dissolved ions bedding develop. This is called the lower flow regime.
(Na+, Cl− etc.) When the velocity, v, becomes high in relation to the
Dissolved load
depth of water, h, rapid or shooting flow develops
Turbulence
where the waves come into phase with the boundary
Saltation Suspended load irregularities (Fig. 2.9); this represents the upper flow
Bottom transport regime.
Bed load
The transition between lower and upper flow
regimes corresponds to a Froude number of 0.6–0.8.
Fig. 2.8 Different forms of transport in water. Sediment grains

may be carried in suspension if the vertical component of the
turbulence is equal to the falling velocity of the grains. Larger 2.9 Sediment Transport Along the Bed
grains are carried along the bottom due to the shear stress Due to Water Flow
This relation between shear stress and flow velocity What actually gives flowing water the capacity to
can also be used for flow in channels where we have carry sediment, and how are sediment particles
bedload transport (Fig. 2.8). Solving the two equations transported?
above with respect to the velocity (v) we obtain: We have seen that flowing water exerts shear forces
against the stream bed. Frictional forces are converted
into turbulence in the overlying water, and have the
v ¼ CðR sin αÞ1=2 effect of transporting particles along the stream bed.
Under moderate flow conditions the largest particles
will be transported along, or just above, the bed as
This is the Chezy equation and C is the Chezy bedload (Fig. 2.8). This takes place partly through
number. rolling or slow creep, partly through saltation, i.e. the
The value of C depends on the roughness of the bed grains jump along the bed.
and on the shape of the channel, particularly its sinuosity. Saltation can be partly explained through
Often used in engineering for calculating the veloc- Bernoulli’s equation:
ity of water in channels, is Manning’s formula:
v2
Pþghþ ¼ CðconstantÞ
2
v ¼ R2=3 ðsin αÞ1=2 =η
Here P ¼ pressure, h ¼ height above the stream
where n is the coefficient of roughness of the bed: n ¼ bed, and v ¼ velocity. We see that water which flows
0.01 corresponds to a smooth metal plate and n ¼ 0.06 over a sediment grain on the bed will have a greater
to a shifting bed of gravel. It is of great practical velocity than water which flows under the grain.
importance to be able to calculate water velocity and Bernoulli’s equation predicts that the pressure above
thereby the erosion potential of artificial channels. the grain must be less than the pressure adjacent to the
The Froude number is a parameter which is often grain (P), and when this difference becomes suffi-
used to describe water flow: ciently great it will be possible to lift the grain from
the stream bed. This “airplane wing effect” does not
F ¼ v=ðg hÞ1=2 work once the grain is in the water above the stream
bed, and the grain will then drop to the bed again.
where v is the average velocity, h is depth of water and The condition for sediment grains being
g the force of gravity. The Froude number is the ratio transported in suspension is that their settling velocity
between the kinetic energy of the water masses (which must be less than the upward vertical turbulence com-
is proportional to the square of the velocity) and the ponent. This means that the grain must be transported
force of gravity, which is proportional to the depth, h. upwards through the water at least as fast as it falls
For low Froude numbers the water flows out of phase downwards. The magnitude of the vertical turbulence
with the bedforms, and current ripples or cross- upwards will be a function of the horizontal velocity
a Bedforms as a function of flow velocity and grain size
Greater than 1
Average velocity cm/s Anti-dunes

ing)
b
edd
Froude number
tion (B
amina
llel L
r Para
Uppe
Dunes
Fr = V
Less than 1
g.L Lower
Current ripples (Cross-bedding) parallel
lamination
No sedimentary
transport
0.1 0.2 0.3 0.4 0.6 0.8 1.0 1.5 2.0 mm

Grain size
d
Formation of flaser structures
c
Flood High tide Ebb tide

Current ripples or Deposition of New current
wave ripples formed suspended material ripples formed
Flat current ripples

Erosion of current ripples at low tide
Formation of flat (eroded) current ripples
Fig. 2.9 (a) Sedimentary structures as a function of flow veloc- clay pellets accumulating between the ripple crests. (c) Formation
ity, grain size and water depth. The Froude number (F) is an of current ripples and truncated ripples in a tidal environment. (d)
expression of the velocity as a function of depth. (b) Ripples with Dunes formed on a coastline with high wave energy
(about 1:8). Under normal flow conditions (< about 1 200
m/s) only clay and silt will be transported in suspen- 100 øm

lstr
sion. Under high flow energy conditions, e.g. during >0 Erosion Hju
.2 m
floods, sand and gravel may also be transported – at m
least partly – in suspension. 10
Erosion and transport are a function of shear stress For
Flow velocity cm/s
m ula
against the stream bed. This in turn is a function not V~
D 1/2
only of water velocity but also of depth. There is a
connection between the flow velocity and the size of 1.0
the sediment grains which can be transported but the Transport
water depth is also important. Hjulstrøm’s diagram Sedimentation
Sto
(Fig. 2.10) applies to channels about 1 m deep. Other

0.1
ke’s
factors which complicate these relationships are the

V~
viscosity (and hence temperature) of the water and the

D
2
density and shape of the sediment grains.

With small grain sizes (silt and clay) the flow 0.01
velocity required to sustain transport is far less than 6 2 1 6 2 0.1 6 2 0.01 6 2 0,001
Grain size mm
the velocity needed to erode a particular grain. This is
because cohesion between sediment particles, particu- Fig. 2.10 Hjulstrøm’s curve showing relations between grain
larly clay, is such that once they are deposited, it is size, flow velocity, erosion and sedimentation. Fine-grained
difficult to erode them again. sediments stay in suspension and are transported at low
velocities, but require higher velocities to erode than silt due
to cohesion
44 K. Bjørlykke
Turbidity Debris flow, Grain flow Fluidised flow

currents mud flow
ρ
ρ ρ ρ
α α
Δρ α•
α
ρ ρ ρ ρ ρ
Δ τ
τ
Fig. 2.11 Sketch showing different types of transport on slopes
Note that fine sand (about 0.1 mm) is the easiest the suspension. We call this phenomenom gravity flow
sediment to erode. On the other hand, finer-grained (Fig. 2.11) and it includes turbidity currents and debris
particles remain in suspension for a long time at low or mass flows. The force of gravity causes movement
velocities. Flocculation of small clay particles to form of sediment/water mixtures because they have a higher
larger ones in seawater increases the settling velocity density than their surroundings, i.e. they are not in
of clays. Also “pelletisation”, through clay being eaten equilibrium with the ambient clear water mass. Grav-
by organisms, is important for the formation of many ity flow is thus distinguished from traction currents by
fine-grained clay sediments. the fact that it can take place in otherwise still water.
However, there are transitions between these two fun-
damentally different processes and we often have
2.10 Different Types of Sediment combined effects.
Transport
We have shown that water or air flowing over a surface 2.11 Turbidity Currents
exerts shear forces on the substratum so that sediment
can be transported by what we refer to as traction Sediment in suspension will be carried down subma-
currents. When we have a relatively low concentration rine slopes because the suspension is heavier than the
of sediment in water (or air) there is little increase in surrounding clear water, forming a turbidity current.
the density and viscosity of the fluid phase. The flow It may be started by river water containing
will then still have approximately the same suspended material entering a sedimentary basin. In
characteristics as it had without the sediment. marine basins the difference in density between river
Another type of sediment transport is due primarily (fresh) water and salt seawater is so great that even if
to the density difference between a water mass carry- river water carries sediment in suspension, it will in
ing suspended sediments and the clear water outside most cases not be denser than seawater.
In consequence there will not be a positive density becomes large, we also get dispersive energy because
contrast, which is the prerequisite for the formation of collisions between the grains (see Sect. 2.13).
of turbidity currents; instead the flow may become an We therefore find that both the concentration of
overflow plume. In lakes, on the other hand, river sediment in suspension, and maximum grain size in
water is often heavier, due both to suspended material suspension, decrease upwards from the bottom. If we
and to its being colder than the lake surface water. It disregard internal friction, we obtain
will then be able to follow the bed downslope, and
become a turbidite. g Δρ V sin α ¼ ðτ1 þ τ2 Þ A
Submarine slides may also quickly evolve into tur-
bidity currents. River sediment entering a marine basin where A is the area of contact with the bottom and the
will mix with the seawater so that clays flocculate and overlying water. The ratio between the volume (V) and
are deposited on the delta slopes. If the slope angle the contact area A is approximately the thickness of the
becomes too steep, we get slides, which may result in flow H. The shear stresses are proportional to the
turbidity currents. When fine-grained sediments are square of the velocity ðτ ¼ cv2 Þ.
deposited on slopes from suspension they have a The velocity (v) of a turbidity current is then:
very high water content. Compaction, sometimes
caused by earthquakes, will cause upward flow of v ¼ c ðg Δρ H: sin αÞ1=2
porewater and may result in liquefaction. This causes
the sediments to begin to flow even on gentle slopes Here the coefficient c includes the coefficient of
because friction is reduced, and they may then turn resistance for friction against the seafloor and against
into turbidity currents. the overlying water. This corresponds to Chezy’s
The forces driving a turbidity current are: number for fluvial flow, so in many ways we can
regard a turbidity current as an underwater river.
F1 ¼ g Δρ V sin α We see from the above equation that thick turbidity
flows will have a higher velocity than thin ones and
where g is the gravity constant, Δρ is the difference that thick flows can flow on gentler slopes than thinner
between the density of the current and that of the ones. This is because the shear stress against the bot-
surrounding water, V is the volume of water along a tom and the overlying water is nearly independent of
certain length of the channel, with the cross-section the thickness of the flow. The flow velocity also
(A) of a turbidity current with length L, and α the angle increases with increasing density of the sediment/
of the slope. Acting against the movement are fric- water mixture in the flow, but high density flows will
tional forces (F2) which, as long as the current is not have higher internal friction and require higher
accelerating, must be equal to the gravitational forces. velocities to keep the material in suspension.
These are shear forces against the bed, τ1, and against A turbidity current can be divided into head, neck,
the overlying water, τ2, plus internal friction and tur- body and tail. The sediment particles in the head area
bulence within the current which keep the sediments in move somewhat faster than the front of the current
suspension. In order for the sediment grains to remain itself. This leads to sediment being swept upwards and
in suspension, the turbulence must be sufficiently then backwards towards the neck, where it mixes with
strong to have an upward component which water from the overlying water mass. From there it is
corresponds at least to the settling velocity of the carried backwards to the body and tail, where we find a
coarsest grains. Turbulence is greatest near the bottom finer-grained, thinner suspension. When the turbidity
of the current, where change in velocity as a function current loses velocity, the largest particles in the head
of height above the bottom (velocity gradient) is will settle out of suspension first because of reduced
greatest. The largest grains in suspension will thus be turbulence. Gradually smaller and smaller grains will
concentrated near the bottom of the current. Near the settle and we get deposition of a bed which is fairly
bed, in addition to turbulence, we also have shear massive, without internal structure, but which
stresses which will transport the grains in virtually becomes finer upwards. In most cases, apart from in
“pseudo-laminar” flow in a thin layer over the bottom. proximal turbidites, we also find deposition of some
If the concentration of large sand grains along the bed
46 K. Bjørlykke
a b
Bouma divisions Interpretation
E Fines in turbidity current, followed

(D) ? by pelagic sediments 3
C Lower
Traction in Flow regime
B Upper
2
Rapid deposition, ? Quick bed

A
1
Fig. 2.12 (a) Bouma sequence in turbidites. (b) Turbidites in a Late Precambrian sequence at Lillehammer, Norway. The sequence
is younging to the right
fine material in this layer, so there is poor sorting. An it may be difficult to distinguish between distal
example of this is Unit A of the Bouma Sequence turbidites and alternations between silt and clay
(Fig. 2.12). formed by traction currents at great depths. Proximal
As settlement from suspension slows down, the turbidites which are well sorted may also resemble
water will have time to sort the grains further, and coarser sediments deposited by powerful traction
we find lamination and bedding structures. The B Unit currents in submarine channels.
exhibits parallel lamination which may be due to flow At the base of turbidity sequences, particularly at
just above the upper flow regime boundary. The C Unit the base of the A Unit, we often find well-developed
exhibits current ripples and convoluted laminae and erosion structures, particularly flute casts and groove
represents further velocity reduction, with deposition casts. Flute casts are formed by the turbulence of
in the lower flow regime. The D Unit has parallel turbidity currents when they pass over a substratum
lamination and was probably deposited from the tail, which consists of finer-grained sediments. The
which consists of very fine-grained sediment. The E structures, which point upcurrent, are produced by
Unit consists mainly of pelagic material, fine-grained vortices in the turbulent flow eroding into the sediment
clay and fossils that accumulated on the seabed during surface. Groove casts are formed by larger grains
the long periods (often thousands of years) between being dragged along the bottom. Even though flute
turbidity flows (Fig. 2.12). The E Unit is therefore not casts and groove casts are typical of turbidites, they
necessarily a part of the turbidite sequence. cannot be used as proof that we are dealing with
The Bouma sequence, first described by Arnold turbidites because similar structures can also be
Bouma in 1962, is an ideal sequence in the sense that formed by various types of traction currents where
in most cases we do not find all the units developed. In there is turbulence and transport along the bottom,
some sequences, particularly those thought to have e.g. in fluvial environments. Sequences resembling
been deposited close to the base of submarine slopes Bouma sequences may also be produced by processes
where the gradient is still fairly steep, we will find only other than turbidity flows, for example rapidly
the coarsest parts of a turbidity current deposited, Unit accelerating fluvial flows. To assist our interpretation
A or a sequence of A + B. We call these proximal we should therefore look at the entire sequence and
turbidites.The finest-grained fractions of a turbidity also try to obtain palaeo environmental information
current tend to be deposited beyond the foot of the from fossils or oriented grains. On submarine slopes
slope or out on the ocean abyssal plain. In these areas there may also be very swift traction currents, particu-
we often only find alternations between C-D-E, or just larly in submarine valleys due to the focusing of the
D-E. These are called distal turbidites. In many cases tidal forces, and relatively strong currents may go up
the canyon as well. These traction currents are capable Debris flows and mud flows normally have thixo-
of transporting coarse sand and at times even coarser tropic properties: the shear forces must reach a critical
material. threshold before deformation (shear) takes place, and
the material loses much of its shear strength following
deformation. In clay containing smectite (montmoril-
2.12 High Density Mass Flows – “Debris lonite) this property is particularly well developed.
Flows” and “Mud Flows” Under shear stress, water will be released. The
house-of-cards packing of clay minerals is destroyed,
Debris flows occur both on land and under water, and the clay particles tending to develop parallel align-
represent a type of mass transport where the sediment/ ment, with the release of water which will reduce the
water ratio is very much greater than in turbidity shear strength causing shear weakening. Some of the
currents, resulting in high viscosity and high internal water in the bottom layer in which deformation is
friction during flow. This also means that the density taking place will be lost to the other sediments, how-
of the mass is from 1.5 to 2.0 g/cm3, while most ever, and friction will mount again. If large clasts enter
turbidity currents have a density of 1.1–1.2 g/cm3 or the basal shear zone, friction will also increase and the
less. The high density of a debris flow means that all flow may stop.
clasts have increased buoyancy because of the dense The stability of mud on slopes and the flow
matrix. The high viscosity of the matrix also means properties of mud flows depend on the clay mineral
that large stones do not sink rapidly towards the bot- composition of the mud and on the geochemistry of
tom of the flow. In flows with high density and viscos- the porewater and the ions adsorbed on the clay
ity, blocks may remain near the surface of the flow minerals. The presence of potassium and sodium
until it solidifies, through loss of water or reduced tend to stabilise the mud.
gradients, and becomes quite rigid as a result of Debris flows are often described in continental
increased density and cohesion. The shear strength of deposits, but are also found on submarine slopes. In
the matrix is often referred to as matrix strength. water the density difference between sediment and
Debris flows may be rich in stones and other coarse surroundings is far less than on land, so that the
material because of the matrix strength. Mud flows angle of the slope must be greater for flows with the
typically have a more clay-rich matrix. But there is same internal friction. Submarine mud flows, on the
no clear distinction between the two. other hand, will not dry up and they can easily take up
Because of the high matrix strength, little sorting more water as they move.
takes place in debris flows (Fig. 2.11). Large blocks Debris flows are particularly common in desert
are often concentrated at the front or on the sides of the deposits. This is because of the often powerful
flows, and there may be less coarse material near the rainstorms which mobilise sediments that in a wetter
base of the flow because of the shear movements. climate would have been transported by fluvial pro-
On land, the density difference (Δρ) between the cesses. In addition, there is little vegetation in deserts
flow and its surrounding fluid (air) is much higher than to bind the sediments, so they are more easily set in
under water. Flows with a particularly low water con- motion. Also of great significance is the fact that the
tent and high shear strength will only flow slowly clay mineral smectite is formed particularly through
down a slope, and we get transitions into what we weathering in desert environments. Clay containing
call solifluction (creep). Sediment flows with a higher smectite will expand when it begins to rain, preventing
water content can move faster, however. In debris the water from filtering rapidly through the soil profile.
flows with their high internal friction most of the Instead, water will be bound to the sediments and the
shear forces will be released along the bottom of the viscosity may be reduced enough for mud flows with
flow, so that the overlying mass moves more or less as thixotropic properties to form. In continental
a coherent mass with little internal deformation. This environments with freshwater the content of
helps to reduce the total frictional resistance to move- stabilising salt (K+, Na+) is low.
ment. Shear strength and viscosity tend to decrease The term “quick clay” is used for extremely thixo-
with increasing rate of shear, and this means that when tropic clays. Undisturbed clays have a relatively high
a flow first gets going it will tend to accelerate. shear strength, but after shaking or some other type of
48 K. Bjørlykke
deformation they can flow like liquids, with a very low grain flow. Grain flow may also occur on beaches and
internal friction. In Scandinavia, Holocene marine in shallow marine environments.
clays which have been uplifted by glacio-isostatic
rebound have been slowly weathered by percolation
of rainwater so that sodium has been leached out, 2.14 Liquefied Flow
making them more prone to landslides and to form
mud flows. Liquefaction is the name given to a process whereby
Systematic surveys using modern coring and sediments lose most of their internal friction, and
remote sensing techniques, such as underwater consequently act almost like fluids. This is the case
cameras, side scan sonar and 3D seismic time slices, when the pore pressure is equal to the weight of the
have shown that large-scale debris flows are rather overburden. When sediments are deposited, they have
common on continental slopes. a high water content and the sediment grains are
On the eastern slope of the Norwegian Sea a huge packed in an unstable manner. As the overburden
slide (the Storegga slide) occurred about 8,000 years increases, the stress on the grain contacts increases,
ago, involving about 3,500 km3 of sediment. The slope and the framework of the sediment grains may col-
scar stretches for nearly 300 km and parts of the flow lapse suddenly. Earthquakes produce tremors which
extended up to 800 km across the deep ocean floor. may cause this structure to collapse, but it can also
The transport mechanism was chiefly debris flow, take place purely as a result of stress (loading). When
where the sediments were riding as a plug on a the packed framework of grains which was formed
wedge of water that reduced the friction against the during deposition is destroyed, the grains can pack
bottom. more closely together. For this to be able to happen,
however, water must flow out of the bed as the poros-
ity decreases. This leads to an upward flow of
2.13 Grain Flow porewater and fine sediment particles, which may be
as great as or greater than the settling velocity of the
Grain flow is flow of relatively well-sorted sediment grains. This process is called liquefaction. The force of
grains which remain in a sort of suspension above the gravity, acting on the contact between the grains, is
substratum due to collisions between the grains. We therefore neutralised, and friction between the grains
see this if we make a little landslide in a dry sandpit or tends towards zero, resulting in liquefaction. If we
pour sugar out of a bag. Grain flow can develop only measure the pressure in the porewater, we find that it
when the initial flow is near the angle of repose (about increases during settlement (compaction) when the
34 ). Bagnold (1956) described how collisions unstable grain framework is destroyed. At one stage
between sediment grains led to a dispersive stress. the pore pressure will be approximately as great as the
However, this stress is only significant near the base weight of the overlying sediments. We can then use
of a flow, where we have rapid variation in flow Coulomb’s Law:
velocity as a function of height above the base
(dv/dh). Here grains with very different velocities τ ¼ C þ ðσ v PÞ tan φ;
will strike one another, and transfer velocity
components to one another. Higher up in the flow the where τ ¼ shear strength, C ¼ cohesion, σ v is the
dispersive stress due to collisions between grains will weight of the overlying sediment, P ¼ pore pressure
be considerably less as the grains have far more simi- and φ is the angle of friction (about 34 ). (σ v P) is
lar velocities, despite turbulence. The dispersive pres- the effective stress. When the pore pressure, P,
sure developed near the base cannot support a thick approaches the weight of the overlying sediments
layer of overlying sediment and therefore grain flows (σ v), the friction component (σ v P) tan φ,
have an upper thickness limit of about 5 cm. approaches zero. Fine-grained sediments like clay
Sand grains which avalanche down the lee side of have considerable cohesion (C), and this will often
sand dunes form small grain flows and are probably prevent clay sediments from sliding even if there is
one of the few significant examples of natural pure little friction. However, once a deformation plane
forms, there will often be movement mainly along it
Grain sizes prevented by drilling wells which release the excess

susceptible pore pressure so that the effective stress and the fric-
Permeability
Co to fluidisation
Cohesion
he tion increases.
y
si
lit
on
bi
ea
rm
Pe
2.15 Sedimentary Structures, Facies and
Sedimentary Environments
Clay Silt Fine Medium Coarse Congl.
sand sand sand
It is difficult to observe or take measurements of rocks
Fig. 2.13 Relation between permeability and cohesion in entirely objectively and consistently. Most types of
sediments. In coarse-grained sediments water escapes quickly, measurements and observations have a considerable
preventing build-up of overpressure, and in fine-grained clayey degree of inherent uncertainty, and the validity and
sediments the cohesion prevents mobilisation. Fine-grained usefulness of results often depend on the experience
sand and silt is therefore most mobile and likely to be liquefied
and also injected as sand dykes (injectites) and skill of those carrying out the field work. It has
turned out to be very difficult to observe structures
due to cohesion in the rest of the clay. At relatively which one does not recognise and understand the
shallow depth clays may have a house of cards struc- significance of.
ture (Fig. 6.2) which will be deformed when subjected A good description of a stratigraphic profile
to shear stress. Clay minerals will then be more paral- depends on good theoretical knowledge of sedimen-
lel oriented resulting in a shear softening. The excess tary processes, and of experience from studies of sim-
porewater from the denser packing could produce an ilar rocks.
overpressure reducing the effective stress so that the It is easy to forget to record or measure some of the
frictional forces are reduced. Smectitic clays are much properties of a rock. In order to obtain a more compre-
more fine grained and have higher cohesion than hensive description and avoid forgetting anything, it
e.g. kaolinitic clays. If we have coarse-grained may be a good idea to have a well-established routine
sediments, i.e. coarse sand and gravel, compaction or even a checklist. Photographs from outcrops or
will lead to excess water flowing out so rapidly that cores may help when writing final reports. If our
the overpressure will drop very quickly, assuming the investigation has a definite and limited objective, we
high permeability has allowed it to build up properly measure only the properties we think will be relevant.
in the first place (Fig. 2.13). It is therefore silt and fine A list of features which can be observed (measured or
sand, the fractions most susceptible to liquefaction, registered) in sedimentary rocks:
that are likely to generate high-velocity subsea flows. 1. Textures – grain size, sorting, grain shape etc.
Liquefaction can, as already mentioned, be triggered 2. Grain orientation – fabric.
by tremors, e.g. earthquakes, and stress. Stresses on 3. Sedimentary structures and their orientation.
sediments (soils) due to buildings, fills, etc. can lead to 4. Fossils.
collapse of the grain frameworks and cause liquefac- A. Preservation or impressions, casts, or the
tion. Lowering of the groundwater table on a slope, for fossils themselves, and their mode of
example down towards the coast, has a similar effect occurrence.
because of reduced buoyancy in part of the sediment B. Trace fossils.
column. Extremely low tides or a combination of a 5. Colour.
strong ebb and a land wind can trigger a slide in 6. Resistance to weathering and erosion.
otherwise stable coastal sediments. This is because 7. Composition (a) Mineral (b) Chemical.
the effective stress in the sediments increases when 8. Thickness and geometry of beds.
the sea level is low. 9. Variations in texture and composition within a
The stability of slopes can be estimated by calcu- bed, e.g. increase in grain size upwards or down-
lating the gravitational forces acting on a particular wards in the bed (grading or inverse grading).
volume of sediment in relation to the frictional forces. 10. Type of contact between beds (e.g. erosional con-
During construction work, slides may sometimes be tact, conformable contact, gradational contact).
50 K. Bjørlykke
11. Association or any tendency to statistical period- called inverse grading. Normal grading may be due
icity in the features of the strata in a profile – bed to deposition from suspension, when the largest
types, structures. particles tend to fall to the bottom first, as with
These observations form the basis for defining turbidites, or to flow velocities dropping off during
facies, which are a synthesis of all the data listed deposition in a river.
above which can be used to group certain types of Inverse grading may be due to increasing flow
rocks. They may be genetic facies, i.e. strata which velocity but if the increase in velocity is too high, the
one assumes have formed in the same manner. All result will be erosion. The supply of coarse material
strata which contain criteria which indicate that they during transport and deposition may also produce
were deposited in shallow water can be described (in inverse grading. In high density sediment currents
reality interpreted) as shallow water facies. In the (debris flows) we may get inverse grading; smaller
same way we have fluvial facies, deep water facies, particles sink more easily to the bottom between the
evaporite facies, and so on. The facies concept can large particles. Massive beds, or at any rate reasonably
also be used to distinguish between different rock massive beds without visible lamination or bedding,
compositions (lithologies), e.g. carbonate facies, sand- may be formed during very rapid deposition of
stone facies. sediments from suspension. X-ray photography of
apparently massive sediments nevertheless usually
reveals the presence of weak lamination even in
2.16 Sedimentary Structures sand. Massive sand beds occur due to rapid fall out
from suspension. In mudstones the primary lamination
By sedimentary structures we mean structures in sedi- may be destroyed by intense bioturbation.
mentary rocks which have formed during or just after On the surface of laminated sediment we may find
deposition. We distinguish between primary structures erosion structures.
which are formed at the time of deposition of the Water running over a plane surface, e.g. beach
sediments, and secondary structures which are formed sand, will produce small-scale, branched (dendritic)
after deposition. erosion marks called rill marks. The flow of water
may be due to runoff from big waves or from ground-
water seepage at low tide. These structures are good
2.17 Layering and Lamination indicators of inter- or supratidal environments, but
they are seldom preserved because they are usually
Most sedimentary rocks exhibit some lamination or destroyed when the water level rises again. Raindrop
bedding, but we also have massive (unlaminated) imprints are often preserved on bedding surfaces, and
rocks. Lamination records variations in the sediment are good indicators of subaerial exposure.
composition as successive depositional layers are
draped over the contours of the sedimentary surface.
The variations may reflect different grain sizes, 2.18 Bedforms
sorting, mineral composition or organic matter.
Laminae are less than 1 cm thick. Units greater than Bedforms are morphological features resulting from
1 cm are called beds. A bed will contain sediments the interaction between particular types of flow and the
which have been deposited by the same sedimentary sediment grains on the bottom. A specific type of
processes. Some sedimentary processes, for example bedform will only form within a limited flow velocity
deposition of a turbidite bed, may be fairly rapid. range and is also dependent on the availability of grain
Migration of a sand dune to give cross-bedding sizes which can be moved by that flow.
takes somewhat longer, while deposition of a clay Current ripples form in fine-grained sand when the
bed from suspended material may take a very long velocity exceeds the lower limit for sediment
time. movement.
Graded beds have a grain size which tends to Ripples and dunes have a stoss side upstream where
decrease upwards within the bed. The opposite is erosion takes place and a lee side on which deposition
Crest
Straight dune
Sediment
transport
Fore set
Erosion,
Deposition, beds
stoss side
lee side
Dune
Fig. 2.14 Principle illustration of dune and cross-bedding
takes place (Fig. 2.14). They therefore migrate as a

result of the combined effects of erosion and deposi- Fig. 2.15 Aeolian dunes several metres high in the Navajo
tion. Current ripples form in fine-grained sand when Sandstone (Jurassic)
the velocity exceeds the lower limit for sediment
movement. Sections through ripples show inclined
foreset laminae, a structure called “small-scale cross-
lamination”. Ripples may also form in coarser sand
but there is an upper limit of 0.6–0.7 mm for the grain
diameter. Ripples are less than 3–5 cm high, and may
have a wavelength of up to 40 cm. The ripple index is
an expression of the ratio of the wavelength divided by
the wave height, and varies between 10 and 40. Waves
may generate oscillatory flow that produces symmet-
rical ripples, which have foreset laminae pointing in
both directions.
Dunes are similar to ripples in shape and structure,
and form in coarse-, medium- and fine-grained sand,
but require significantly higher flow velocities for their
formation. Dunes range in height from 5 cm up to
several metres, and wavelengths may exceed 10 m
(Fig. 2.15).
Cross-bedding (large-scale cross-stratification) is
seen in cross-sections through dunes (Fig. 2.16).
Each lamination is called a foreset bed and represents
the lee-side surface of a migrating dune. If the dunes
have straight crests the foreset beds on the lee side will
Fig. 2.16 Cross-bedded sandstone which represents the filling
form a straight transverse plane (tabular cross-bed- of a fluvial channel above a coal bed. The grain size fines
ding). Curved dunes have rounded foreset beds, and upwards somewhat. It is largely trough cross-bedding
represented here. From the Tertiary sequence of Spitzbergen.
(Photo A. Dalland)
52 K. Bjørlykke
in trough cross-bedding the laminae have a rounded Tongue-shaped (linguoid) ripples have a very high
surface which is concave in the downstream direction. sinuosity and asymmetry and are usually formed in
The foreset lamination may form a relatively sharp shallower water or under higher velocities than
angle with the underlying bed, or may have a more straight-crested types. Wave ripples in particular
tangential contact. The latter is typical of trough- may split laterally into two ripples. This is called
shaped sets. bifurcation. In intertidal zones ripples formed at
Aeolian dunes may be many metres high, and their high tide may be eroded at low tide, and the crests
cross-bedding will then be correspondingly large become flattened. When the tidal flat is submerged at
(Fig. 2.15). high tide it may also be below the normal wave base,
Whatever the size of the bedform, ripple movement resulting at slack water in deposition of clay which
provides the basic mode of migration. Seabed survey tends to collect in the ripple troughs. Current ripples
profiling records clearly show smaller ripples climbing with thin lenses of clay between them constitute
the stoss sides of larger ones (so-called “megaripples”, flaser bedding. Wind may generate waves moving
i.e. small dunes), which in turn are climbing the stoss in different directions, particularly at very low
sides of sandwaves. water, so that we find two or more sets of ripples at
Cross-sections through dunes show “large-scale an angle to each other (interference pattern). In most
cross-stratification”, which is often referred to as cases each bed with current ripples represents a sort
“cross-bedding”. This can be observed in real time of equilibrium with deposits reflecting current
on seabed video recordings made during periods of patterns. Isolated sand lenses in clay are called len-
strong tidal current flow, where gradual forward ticular bedding.
movement of “megaripples” is seen, caused by sand Normally current ripples form completely horizon-
cascading down the lee side after reaching the crest. tal beds. In some cases, however, we find examples of
“Plane beds” (upper stage) may form when the current ripples appearing to climb downstream in rela-
shear stress against the bed exceeds the values which tion to the horizontal plane. They form several sets of
produce dunes. In cross-section we only see planar cross-laminated beds delimited by erosion boundaries,
lamination, which is an internal structure, but on the but with small internal erosion planes. These are called
bedding surface we may see very small ridges, which climbing ripples and are due to sedimentation taking
define a lineation called primary current lineation, place so rapidly that, in contrast to normal ripples,
parallel to flow. there is no equilibrium between erosion and sedimen-
At even higher velocities in relatively shallow tation. Climbing ripples are therefore typical of
water, standing waves may produce antidunes when environments with rapidly declining flow velocity
the Froude number exceeds 0.8. The antidunes which and consequently a high rate of sedimentation.
are produced when standing waves are in phase with Sandwaves are large-scale transverse bedforms,
the bedforms develop resulting in low-angle cross- generally 2–15 m high, with a wavelength of
lamination which dips up-current. 150–500 m. They may be formed at flow velocities
Both ripples and dunes are formed through sand of 65–125 cm/s. Cross-stratification may be symmet-
being transported along the bottom and deposited in rical or asymmetrical on both sides of the sandwave,
sloping strata on the lee side of the structure. In con- depending on the relative strength of the opposing
sequence they always have dipping laminations currents.
(foreset beds) which may lie at an angle (angle of At velocities of less than 1 m/s sand ribbons may be
repose) of up to 35 to the surface of the bed, though deposited – longitudinal bedforms developed parallel
such high angles are rather rare. Current ripples in plan to the currents. Sand ribbons are typical of subtidal
view may be straight, or form curved patterns (sinuous environments (20–200 m), and they may be up to
crests). Ripples with a symmetrical cross-section 20 km long, 200 m wide and less than a metre thick.
(symmetrical ripples) are formed by waves as a rule. Relatively low-energy environments (<50 cm/s)
Asymmetrical ripples are formed by a predominantly are characterised by sand patches and mud. The sand
unidirectional current and their steeper side faces forming the patches probably only moves during
downstream. Ripples with a high sinuosity are also storms. Lateral structures are typically current ripples
asymmetrical in most cases. and small dunes.
0.5 m
Storm surge
n
sio
Fairweather wave base Ero Mean sea level
Hummocky deposition
Storm wave base
Turbidite deposition
Graded rythmite deposition
(simple fallout)
Fig. 2.17 Hummocky stratification (After Harms et al. 1975, Walker 1982)
Sandwaves often have current ripples (“smaller 2.19 Erosion Structures on the
dunes”) on their surface. This may also be the case Underside of Sand Beds (Sole
with dunes. Bedforms are a function of both flow Structures)
velocity and grain size. Current ripples are formed
only in silt and fine- to medium-grained sand, while When sand is deposited rapidly over finer sediments (silt
dunes require medium to coarse sand. Upper flow and clay) there will also in very many cases be an initial
regime plane beds which have an internal structure erosional phase, so that erosion hollows are formed in the
of planar lamination are formed when the Froude substrate. These hollows fill with sand, and we get a sort
number is about 0.6–0.8. Plane beds typically develop of cast. We see these erosion structures on the underside
in beach sand. A well-developed lineation on the bed- (sole) of sandstone and conglomeritic strata. We seldom
ding surface parallel to the direction of sediment trans- observe these surfaces unless the bed is steeply inclined
port is typical of plane beds (Fig. 2.9). Antidunes are or inverted. Wherever there are overhanging rocks, it is
formed by higher velocities in the upper flow regime; important to study the underside of the bed.
the absolute flow velocities required are lower when Flute casts are formed by static vortices in the
the water is shallow. water eroding into the underlying sediments. Their
Below the ordinary, fair weather wave base we find sharp end points upstream and they broaden in the
different types of sedimentary structures from those downstream direction. Flute casts (Fig. 2.18) are
formed through constant wave action. During storms good indicators of current direction, which is
in nearshore areas traction currents may develop and measured along an axis of symmetry through the
carry fine material (fine sand) from the beach out to structure. They are typical on the underside of turbi-
greater depths. This material is deposited as hummocks dite beds, but also occur in fluvial sandstones as a
consisting of parallel laminae which form dome- result of fluid turbulence.
shaped structures (Fig. 2.17). Because deposition of Similar structures are formed around objects which
beds of this type takes place during short periods project up from the bottom, such as stones and large
during storms, they are often reworked by bioturbation fossils.
near the top. Hummocky stratification is common in Gutter marks are longitudinal grooves up to 20 cm
sections through some shallow marine deposits and deep and rather narrow (less than 20–30 cm) with a
wave-dominated deltas. spacing of 1 m or more. They are the result of
54 K. Bjørlykke
Convolute lamination
Convolute lamination
structures within a bed
formed during deposition.
Erosion of these structures
before the next bed is deposited.
Slumping
Folding of a number of beds
simultaneously due to gravity-
induced sliding following
further deposition.
Dish structures
Clay- and silt-enriched, rounded
50 cm dish-shaped laminae formed by
deposition from upward-flowing
Fig. 2.18 Flute casts on the lower surface (sole) of a coarse- water immediately following
deposition.
grained sandstone bed in the Ring Formation, Rena, South
Norway. Note the small flute casts on the large flute cast struc-
ture. Flute casts are casts of the erosion structures formed in the Fig. 2.19 Sedimentary structures due to liquefaction and soft
finer-grained underlying bed by vortices sediment deformation during the deposition process. See text
channelised flow, and flute casts or groove casts may 1. Shear stress due to water or sediment movement,
be found along their margin. e.g. convolute lamination.
Ridges and furrows may also be formed through 2. Expulsion of porewater (liquefaction, dewatering),
erosion of the substratum. Some are U-shaped or V- e.g. dish structures, clastic dykes.
shaped channels. Erosion structures formed by objects 3. Heavier beds above lighter beds (inverse density),
which are transported along the bottom are called tool e.g. load casts and ball-and-pillow structures.
marks. The objects may be fossil fragments or larger 4. Gravitational deformation. Gravitation-induced
sediment particles. sliding, folding and faulting on a slope, e.g.
Groove casts are long, narrow erosion furrows due slumping.
to something being dragged along the bottom. 5. Shrinkage, e.g. due to dessication, or permafrost
Chevron marks are erosion furrows with V-shaped (ice wedges).
structures in the clay sediments on either side. The V- It is important to distinguish between these types of
structure closes downstream, and is due to a cast deformation structures because they have completely
forming in cohesive clay. different implications for interpreting depositional
Prod marks show where an object has dug down environments. They are not mutually exclusive, how-
into the clay and then been plucked out again by the ever, and may be found in close association in the
current. As a result the steep side of prod marks is the same sequence.
downstream side. Convolute lamination forms in fine sand or silt, and
Bounce marks are rows of more symmetrical marks is due to laminae, e.g. current ripples, being deformed
due to objects being swept or bounced along the almost as they develop. Folded and inverted
bottom. (overturned) lamination is typical, and the structure
is deformed in the downstream direction due to the
stress induced by the water movement and the insta-
2.20 Deformation Structures bility (almost a state of liquefaction) of the sediments
(Fig. 2.19).
Sediments are often unstable immediately after depo- Dish structures (Fig. 2.19) are thin, clay- and silt-
sition, and later movements will deform the primary enriched dish-shaped laminae in sandy sediments.
structures. Their structure is due to the porewater which flows
Deformation may be caused by four main factors: upwards immediately after deposition, transporting
clay and silt which become trapped in these thin
laminae. Sand dykes (or clastic dykes) are intrusions of the bedding is proportional to the sine of the angle of
sand upwards into cracks in a finer-grained sediment, dip, and acts as a compaction force which can slide
due to porewater overpressure. The overpressure and fold the beds. In the upper part of the slide,
reduces the friction between the grains and injects tensional deformation is prominent, producing
water with sand into vertical fractures produced by faulting, while compression occurs near the base of
high overpressures or follows bedding planes as sills. the slide and produces folding. The result is called
Overpressurised porewater with sand may rise right to slumping (Fig. 2.21). Slumping occurs most readily
the surface and form sand volcanoes. where we have rapid sedimentation and therefore rel-
Well-sorted sand has at the time of deposition an atively thick beds with a high water content and low
initial porosity of 40–45%, whereas recently deposited shear strength. Deformation takes place when the
mud contains 50–80% water. When sedimentation is shear stress exceeds the shear strength. The shear
rapid, the mud has little time to lose its excess water, stress increases with the thickness of the unconsoli-
and a sand bed deposited on top may then sink down dated sediments, but there is not usually a
into the underlying silt and clay and form load corresponding increase in shear strength with increas-
structures. On the lower surface of a sand layer we ing thickness. Slumping may resemble convolute lam-
often see pillow-shaped depressions surrounded by ination, but normally affects more than one bed.
clay which has oozed up around them. If this process Gravitational deformation also leads to faults on a
continues, it will form isolated sand pockets in the greater or lesser scale. Sliding of large volumes of
underlying clay: ball-and-pillow structures. Primary sediment down slopes produces slope scars. Growth
structures such as flute casts often sink in the underly- faults and other types of “listric” faults are a result of
ing mud and are deformed by loading (Fig. 2.20). large-scale gravitational deformation in the upper part
This mechanism also operates on a larger scale, for of an area which is under tension.
example channel sand will sink down into surrounding Dessication cracks or mud cracks are examples of
clay (see “deltas”). Poorly compacted clay and silt will contraction or shrinkage of sedimentary beds due to
be lighter than surrounding sediments, and flow dehydration. Cracks frequently form regular polygons,
upwards to form clay diapirs. often hexagons or orthogonal sets. Dessication cracks
Gravitational deformation occurs in sediments form only in clay and silt, and the cracks often become
which are deposited on slopes. These forces can be filled with sand, resulting in good contrast. The forma-
resolved into a vector normal to the bedding, and a tion of dessication cracks requires that the beds be
shear stress parallel to it. The vector which acts along exposed to the air so that the sediments can dry out.
Fig. 2.21 Sediment beds which have been folded immediately

Fig. 2.20 Load cast structures at the base of a sandstone bed in after deposition (through slumping) due to sliding on submarine
the Late Precambrian Ring Formation at Rena, Southern slopes. Note that the overlying beds are undeformed showing
Norway. The picture covers an area of 3 4 m. (Bjørlykke that this is not tectonic folding. From the Ridge Basin (Miocene-
et al. 1976) Pliocene), California. Scale, John Crowell
56 K. Bjørlykke
In certain cases shrinkage structures may also form The source of the silica is usually amorphous biogenic
underwater through dehydration of clay minerals silica, frequently sponge spicules.
(smectite) as a result of variation in the salinity of
the porewater (syneresis). Shrinkage structures are
less regular than desiccation cracks and are not usually 2.22 Trace Fossils
interconnected. In the smectite-rich Eocene sediments
from the North Sea basin and the Norwegian Sea, Trace fossils are structures in sedimentary rocks which
seismic data show large (several hundred metres have been left by organisms that lived on and/or
wide) polygons which have been interpreted as shrink- burrowed in the sediment. Such organisms are
age cracks. extremely sensitive to changes in the composition of
When the porewater in sediments freezes to ice and the nutrient content, the sedimentation rate and bottom
remelts, we also find expansion and contraction which currents, and are therefore useful indications of the
results in polygonal surface marks and associated ver- environment (Fig. 2.22a,b). Trace fossils are therefore
tical ice wedges. good indicators of the depositional environments.
They can be classified taxonomically, i.e. according
to the animal which left the traces, but it may not be
possible to determine which animal was responsible.
2.21 Concretions The same species may form several different types of
trace depending on the sediment composition and on
Concretions are round, flat or elongated structures its mode of life. Furthermore, different animals may
which consist of cement which has been chemically leave traces which are so similar that they are classi-
precipitated in the pores of the sediment. The most fied as one trace fossil. For these reasons a descriptive,
common types of concretion are carbonate (calcite and morphological classification of trace fossils is used.
siderite) and silica (chert). Sulphides, particularly Trace fossils can also be classified according to
pyrite, also form concretions. where they occur in relation to the bed:
A characteristic feature of concretions is that any 1. On top of beds (e.g. a thin sand or carbonate layer
laminations in the sediment pass through the Epichnia).
concretions. This shows that the concretion has been 2. Within the bed (Endichnia).
formed through passive filling of its pores. As the 3. On the lower surface of the bed (Hypichnia).
overburden increases, the sediments around the con- 4. Outside the bed (Exichnia).
cretion will be subject to compaction, while the con- The impression made by an animal may create a
cretion cannot be compressed because the pores are mould or a cast in the overlying bed. Organisms which
full of cement. A concretion therefore has a cement burrow into sediments often secrete a cement which
content which corresponds to its porosity at the time of ensures that the walls of their burrows do not collapse.
formation. Carbonate concretions in clay may have a These secretions also contribute to preservation.
carbonate content reflecting 50–70% porosity if the When worms eat sediment, for example, it passes
matrix does not contain carbonate. Concretions in through their digestive organs and their burrow refills
calcareous rocks, i.e. marls, contain clastic or biogenic with a sediment which has a somewhat different com-
carbonate in addition to carbonate cement, and there- position from the surrounding sediments. This is par-
fore have more carbonate than the matrix. In these ticularly noticeable in burrows at the interface
cases the carbonate content cannot be taken as an between two strata with different compositions. The
indication of the porosity at the time of formation. amount of bioturbation reflects nutritive conditions
Concretions often contain fossils, showing no sign of and the sedimentation rate. With very rapid sedimen-
compaction while the same fossils are severely tation there will be less time for organisms to burrow
deformed and sometimes also dissolved outside the through the sediments. Where we have very slow
concretion. In carbonate sediments, particularly sedimentation or an hiatus, the sediments will often
chalk, there are silica concretions (chert). These are be thoroughly churned up by bioturbation, and thereby
formed through precipitation of finely divided amor- homogenised. Bioturbation is most widespread in
phous silica to form a type of chert called flint. marine environments, but can also be found to a lesser
55 56 57 58 59 60 61 62
(f) Rocky coast association
re
river y sho
Lake Lake
shore Rock Deep sea fan
y
Sand 0m
Wave base
200 m
Supratidal Intertidal Subtidal
2000 – 4000 m
Continental area Continental shelf Cont. slope and apron Ocean basin
Non-marine zone Tidal zone Neritic zone Bathyal and abyssal zones
(a) Scoyenia association (b) Skolithos and Glossifungites (c) Cruziana association (d) Zoophycus association (e) Nereites association
Diplichnites
Rusophycus
Scoyenia Skolithos
Cruziana
Cylindrichnus Helminthoida
Zoophycos
Planolites
Phycodes
Diplocraterion Oldhamia
Cylindricum Nereites
sensu stricto
Dikaka Urohelminthoida
Isopdichnus
Cosmorhaphe
Beaconites Rhizocorallium
(Rusophycus
type)
Spirophyton Spirophycus
Spirorhaphe
(Cruziana
type)
0 1m
Thallassinoides
Merostomichnites
Palaeodictyon Dictyodora Palaeomeandron
Arenicolites Teichichnus Scalarituba

Anthichnium
Neonereites
0 5 cm Phyllodocites
Ophiomorpha
Chirotherium
Lophoctenium
Energy Variable: waves, Higher energy, Lower energy, Lower energy, Lower energy, infrequent No wave mixing,
(waves, river, air currents constant wave frequent wave frequent wave mixing wave mixing, some density some density flows,
wind) mixing mixing flows ocean current flows
Eh Oxidising, except Oxidising Oxidising Oxidising O2 reduced, high organic Little O2, e.g. from density
at lake bottoms content flows, even anoxic
Salinity Low, rarely high Mostly normal Normal Normal Normal Normal
Temperature Extreme variety Daily changes Daily changes Seasonal changes < 10°C, no changes Down to 2°C, no changes
Light Daily changes Daily changes Daily changes Daily changes in upper part None None
Substrate Aqueous or Sand > mud, less Mud = sand, Mud = sand, stable, Mud, stable except in Pelagic mud, most stable
sediment subaerial; sand, firm, reworked firmer, reworked rarer reworking and ripples failure and density flow but for density and ocean
silt, mud current flows
Diversity Lower than in Low Low High Low Low, but higher than in
marine realm (d), higher through time
Abundance High in places Dense Dense High Low Low, but seems higher
due to slow deposition
Dominant Arthropods (insects) Arthropods, molluscs, echinoderms, Arthropods, molluscs, Arthropods, polychaete worms, hemichordates
organisms, molluscs, corals, ‘worms’ echinoderms, corals (e.g. acorn worms), echinoderms
traces vertebrates, Unbranched Vertical and Horizontal crawling – Complex horizontal grazing
Casts, crawling, grazing
small shallow vertical burrows inclined grazing traces > inclined and shallow feeding traces,
traces, spreiten planar
burrows, trails, ear-shaped burrows spreiten inclined in sheets,
and on surface
tracks, burrows ribbons, spirals – bioturbation
herbivores, carnivores Mostly suspension feeders Mostly sediment feeders Sediment churners, feeders Sediment grazers, farmers
Fig. 2.22 Common trace fossils in different sedimentary facies (from Collinson and Thompson 1982)
58 K. Bjørlykke
extent in freshwater sediments. In poorly oxygenated sequence of the Oslo area we find Nereites facies in
environments there is, however, little bioturbation. the shales, but the water depth was probably not more
In high energy environments, e.g. above the fair- than 100–200 m, perhaps even less. In the deep oceans
weather wave base, we only have vertical trace fossils, the Nereites facies may correspond to a depth of sev-
e.g. Skolithos or Diplocraterion. The high current eral thousand metres.
velocity prevents these organisms from crawling Trace fossils are very useful facies indicators and
around on the bottom; they have to burrow down into should be noted whenever sedimentary sections are
the sediment and live by filtering nutrients out of the examined for facies interpretations. Certain trace
water. To remain at the same level beneath the sedi- fossils can be linked with animals that have fairly
ment surface they must move upwards or downwards specific environmental requirements.
in their holes, depending on whether erosion or sedi-
mentation is proceeding in the area. Structures or
fillings which reflect such adjustments are called 2.23 Facies and Sedimentary
“spreiten”. Environments
In modern marine environments we can study a
number of organisms which create bioturbation The word “facies” is used in a number of geological
structures. The most common are worms like disciplines. A term such as “metamorphic facies” is
Arenicola which make U-shaped traces in fine-grained thoroughly entrenched. Sedimentary facies have also
sand and silt. Such traces (arenicolites) are also to be long been identified in sedimentology to distinguish
found in older rocks. Burrowing bivalves create vari- between sedimentary rocks which differ in appearance
ous types of trace as well. Arthropods like crabs and and have formed in different ways. The term facies can
prawns make burrowing structures in beach sand be used both descriptively and genetically. We use
(ophiomorpha). These vertical trace fossils are terms like “sandy facies”, “shaly facies”, “carbonate
grouped together in an ichnofacies called the Skolithos facies” when we are describing properties of the rock
facies. Thallassinoids are more horizontally aligned that can be observed or analysed objectively. We use
networks of arthropod burrows. terms such as “shallow water facies”, “deep water
Below the wave base and in other protected facies”, “turbidite facies”, “deltaic facies”, “intertidal
environments, for example the intertidal zone, we facies”, “aeolian facies”, “reef facies” etc., depending
find traces in the horizontal plane from organisms on which environment we believe the rocks represent.
which live off blue-green algae and other organic In these examples, the word “facies” represents an
material on the surface of the sediment. In this neritic interpretation, and is therefore not very suitable for
zone (Fig. 2.22a,b) we find a number of different types describing sedimentary rocks objectively. For this rea-
of traces from organisms which eat their way through son it is important that we define the objective criteria
sediments, and which form different patterns. This is (observations) on which we are basing our
called the Cruziana facies. Rusophycus and Cruziana interpretations.
are typical of the neritic zone and represent horizontal What we can do, then, is first describe and take
traces of arthropods which feed on the sediment sur- measurements on a series of beds in the field, and on
face. Rhizocorallium and Teichichnus are other trace the basis of certain criteria divide the series into facies.
fossils that occur below the Cruziana facies. In deeper The criteria will generally be texture, sedimentary
water where wave and current energy is even lower, structures, mineral composition, and, if present, also
we find Zoophycus and Nereites facies. It is important fossils. Interpreting a facies in terms of depositional
to remember that these environments are primarily a environment is often very difficult, especially since
function of wave and current energy, and cannot sim- few criteria are unambiguously diagnostic of one par-
ply be correlated with absolute depth. In shallow ticular environment. In some cases it may be useful to
enclosed seas with a shallow wave base, e.g. the mod- use statistical methods for distinguishing between dif-
ern Baltic, we find an effective wave base of only 5–10 ferent facies and for describing facies sequences.
m in many areas, while elsewhere we may have stron- What we observe and measure is a selection of the
ger currents along parts of the deeper trenches. In the properties of the rock. When we describe a sedimen-
epicontinental Cambro-Silurian marine sedimentary tary rock, we observe the results of processes which
Fig. 2.23 Schematic representation of sedimentary facies on a passive margin
have acted in the environment in which the rock was facies may represent a facies association typical of
deposited. The sedimentary structures we observe tell delta progradation.
us something of the hydrodynamic conditions during Other facies associations may represent
deposition. Organic structures and fossil content help transgressions or regressions in a shallow marine
us reconstruct the ecological conditions at the time of environment.
deposition in the basin. Evidence of chemical pro- If we have a continuous sequence without breaks
cesses such as weathering, diagenesis and precipita- (unconformities) the vertical succession represents the
tion of authigenic minerals also supply important lateral succession of environments. This, in essence, is
information about the depositional environment, as Walther’s Law, named after the German geologist
well as its early post-depositional history. Figure Johannes Walther.
2.23 is a diagrammatic representation of the major
depositional environments on a continent and the var-
ious transition stages to deep water. 2.24 Alluvial Fans
Following an interpretation of the facies of each
bed or sequence of beds, we may try to group these Alluvial fans are accumulations of sediment which has
into what we call facies associations. These been transported by fluvial processes or different
associations of facies can be related to large-scale forms of mass transport (e.g. mud flows and debris
processes in particular environments. An association flows) and deposited in valleys or on slopes fairly near
of facies coarsening upward from marine shales into the erosion area. The fans form in areas that have
delta front sand or into coal beds of the delta plain considerable relief and are usually associated with
60 K. Bjørlykke
faults which keep uplifting the erosion area in relation through which finer-grained sediments can pass. In
to the area of deposition. This tectonic subsidence of this way the sediment is sieved, and sieve deposits
the alluvial fans is a necessary condition for their are formed.
being preserved in the geological series. Alluvial fan Downslope on the fan, channels usually split into a
deposits in older rocks may therefore supply important number of smaller channels. This reduces the hydrau-
information about tectonic movements during deposi- lic radius of the channels and in consequence their
tion. Alluvial fans which have formed in areas with a velocity, and hence capacity for carrying sediment, is
high relief but little tectonic activity, will merely rep- lowered. Sediment will therefore become finer-
resent a stage in the transport of the sediments. For grained downslope, even if there is no reduction in
example, in glaciated areas we may find alluvial fans the gradient.
on slopes steepened by glacial erosion. They will be The water table will be deepest at the top of the fan,
eroded again if they are not rapidly covered by a and shallowest at the foot. Alluvial fans are good
transgression. groundwater reservoirs, easy to tap because of their
Erosion on the uplifted block will form V-shaped porosity and permeability. The circulation of ground-
valleys (or canyons), and these will drain into the water through an alluvial fan leads to strong oxidation
valley. The apex of the alluvial fan is usually near of at least the upper part of the sediments, giving them
the main fault plane and sediment transport from a red colour due to iron oxides. In most cases any
here will tend to follow the steepest slope downwards organic material will be completely oxidised.
so that the sediments will be spread out in a fan In arid climates there will be a great deal of evapo-
(Figs. 2.24 and 2.25). If there is only a short distance ration from the groundwater which emerges from the
between adjacent valleys, and consequently between fan, and the ions in solution will be precipitated as
fan apices, the fans will coalesce. If major drainage carbonate (caliche) and iron oxide. Water flows
systems develop, larger fans will form further apart beyond the arid alluvial fans in ephemeral rivers,
from one another. In areas with a relatively humid which only exist after heavy rain, and may then collect
climate, fluvial processes will account for sediment in playa lakes which dry up each year, forming evapo-
transportation even high up on the fan. In arid climates rite deposits (Fig. 2.25).
the water table under the fan will be deep down, and Humid fans will be dominated on the lower part by
when it rains the water will rapidly filter down into the fluvial channels. These may drain a major portion of
upper part of the fan. The slope of the fan may then the fan, and thus have a large hydraulic radius and a
increase due to deposition on the upper part. As a greater capacity for transporting sediment, even if the
result these sediments may be transported with a slope of the fan is not very great. In the lower part of a
high sediment/water ratio as debris flows or mud humid fan cross-bedding will be fairly pervasive,
flows. The upper part of the fan may consist of large while the upper part will tend more to consist of
blocks or cobbles which form an open network system massive conglomeratic beds. The foot (distal end) of
the fan will merge into the other sediments which
cover the valley floor. These may be lacustrine or
fluvial deposits.
A characteristic feature of transport and sedimenta-
tion across and around arid zone alluvial fans is that
sedimentation takes place during short periods in con-
nection with the rains. However, the water rapidly
disappears into the ground, increasing the surface sed-
iment/water ratio, and eventually dries up to leave
poorly sorted conglomeratic beds. It is important to
remember that the sedimentary structures and sorting
we observe are only representative of the final deposi-
tion phase.
The grain size of the sediments on the fan is a
Fig. 2.24 Alluvial fan. Death Valley, California function of weathering and erosion in the source area
Apex
Alluvial
plain
Playa
lake Basement
block
Subsidence
Fig. 2.25 An alluvial fan developed along a basement fault. The degree of progradation of the fan into the basin depends on the
relief along the fault, but is also dependent on the rainfall and the catchment area on the upthrown block
and of transport capacity outwards across the fan. the water table is shallow (less than 1 m), there is little
With fluvial transport, the velocity v and thereby trans- direct evaporation, but some water will be drawn up by
port capacity is both a function of the slope and the capillary forces and evaporate.
depth (hydraulic radius) of the channels (see Chezy Plants are important in connection with evaporation
equation, p. 42). of groundwater as tree roots may penetrate more than
The velocity is therefore much higher during 10 m below the surface. Trees can be a major drain on
flooding. groundwater reserves because they use water which
could otherwise have seeped into wells, rivers or lakes.
Dense forests may use water corresponding to
2.24.1 The Water Budget 200–500 mm of precipitation. In areas with limited
water reserves it may consequently be necessary to
A part of the total rainfall will sink into the ground and limit vegetation, particularly of varieties of trees and
supply the groundwater. Another part will evaporate bushes with a relatively high transpiration rate and no
on the surface or flow on the surface into rivers useful function. However, vegetation is important for
(Fig. 2.26). In hot, desert areas where the air is very stabilising the topsoil and thereby preventing erosion.
dry, evaporation may correspond to 2–3,000 mm/year, Much of the water evporated from vegetation may
i.e. more than the normal amount of precipitation in return as rain locally. Vegetation may also have an
most places. Plants also contribute to water loss effect on the albedo, i.e. the amount of sunlight which
through transpiration, and in areas with vegetation, is reflected from the Earth’s surface, and this in turn
evaporation and transpiration may reach up to may affect the climate.
2,000 mm/ year. Without vegetation we find a signifi- Runoff is the fraction of precipitation which reaches
cant amount of evaporation only when there is free the rivers. The percentage of precipitation which filters
water or damp ground right up to the surface. Even if down to and recharges the groundwater is of particular
62 K. Bjørlykke
X1 precipitation X3 evaporation
X4 runoff
Evaporation
Water table
X2 infiltration Lake
ov Island
M
em Sea level
en
t of
groundwater
Fresh groundwater
Marine porewater
Fresh groundwater
X1=X2+X3+X4
Fig. 2.26 Diagram showing the circulation of groundwater on land and under marine basins
importance. Rivers also have groundwater added to freshwater under islands. Freshwater may also flow
them if their surface is below the water table. If their from the continents, following permeable beds
surface is above the water table they will lose water. In beneath the continental shelf. Freshwater of high qual-
consequence, rivers cannot be considered in isolation ity (little pollution) is scarce in many regions, and in
from groundwater reserves. recent years in particular a large number of geologists
Infiltration of surface water into the ground have been involved in mapping groundwater reserves.
depends on the permeability of the soil cover and on
the vegetation. Clayey sediments allow far slower
infiltration of surface water than sand or gravel. 2.25 Desert
Where there are soil types with swelling clay (smectite
group), which are very common in desert areas, the Deserts are areas with little or no vegetation and we
first rains will cause these minerals to swell, further can distinguish between different types of desert:
reducing their permeability. Consequently, infiltration 1. Deserts due to low precipitation and high evapora-
is reduced, and the runoff which creates floods tion in hot climates.
increases. If the earth is dry above groundwater 2. Deserts due to low precipitation in cold climates.
level, water has to overcome the capillary forces 3. Deserts due to soil erosion.
which act against water percolating down through The first two types of desert are largely governed by
dry, fine-grained sediments. We may find a layer meteorological factors. Areas which lie on the lee-side
with air in the pores between the groundwater table of major mountains will be dry. At about 30 N and
and the water filtering down from above (trapped air). 30 S high pressure areas predominate, resulting in low
Alluvial fans are porous and have a relatively good precipitation. Since the prevailing wind at these
potential as water reservoirs, and we often find springs latitudes will come from the east, we find most of the
at the foot of fans. In dry areas the depth to the water deserts on the western side of mountain chains. This is
table can be most simply charted through seismic true of Asia Minor, the Sahara, California and Nevada
recordings. in the northern hemisphere, and Australia, South
Because freshwater is lighter than saltwater, it will Africa and Chile in the southern hemisphere. Inland
flow over saline porewater and form pockets of areas surrounded by mountains tend to have a desert
climate. During the glacial periods there was less 4. Lack of fossils or organic material.
evaporation and therefore less rainfall. North Africa The lack of thin silt or clay beds between the cross-
and the Sahara were then very much drier than the bedded layers, and the general association with
present day. overlying and underlying sediments, are also impor-
Iceland has good examples of desert regions which tant criteria. One of the best-known examples of an
are due to a cold climate, and has problems with soil aeolian sandstone is the Navajo Sandstone (Jurassic)
erosion due to lack of vegetation in many areas. in the Rocky Mountains. In Northern Europe there are
Shortly after the withdrawal of the ice across aeolian sandstones in the Lower Permian
Scandinavia there was a desert with considerable aeo- (Rotliegendes) and the equivalent Yellow Sand in
lian deposits, which formed before the vegetation northeastern England (Fig. 2.27).
cover developed. Loess deposits are fine-grained In pre-Devonian times all continental environments
sediments which form largely through aeolian erosion were deserts in the sense that they lacked vegetation.
and transport from deserts and also from glacial Also in these rocks, however, we can distinguish
sediments exposed after glacial retreat. Large areas between dry and wet climates, largely from the nature
of China, for example west of Beijing, are covered of the fluvial or aeolian deposits.
by up to several hundred metres of Quaternary loess. Sand dunes are always moving and may end up in
The lack of vegetation naturally means that aeolian the sea, still partly preserving an aeolian sorting and
transport and deposition are important in deserts, but grain-size distribution.
only a relatively small part of the desert areas are
covered by wind-blown sand.
Large areas consist of bare rocks and mountains 2.26 Lacustrine Deposits
with little sediment. In other regions there is only wind
erosion (deflation), which leaves the ground covered How Are Lakes Formed?
(armoured) by a layer of stones which protect it from In principle we have three types of lake:
further erosion. Even in desert regions like the Sahara 1. Lakes of glacial origin:
many areas are dominated by a fluvial drainage pat- (a) lakes formed through glacial erosion,
tern. Although several years may pass between rains in (b) lakes formed through damming by moraines or
this area, a heavy rainstorm may transport so much by the glaciers themselves.
sediment that the fluvial drainage pattern survives for 2. Lakes of tectonic origin:
many years. (a) lakes formed in areas of rapid tectonic subsi-
Large expanses of wind-blown sand are called ergs. dence (rifting) or more uniform subsidence,
They may be formed by the coalescence of different (b) lakes formed as a result of damming by horsts
aeolian bedforms. Barchans are crescent-shaped which have been elevated through faulting, or
isolated aeolian dunes with a convex erosion side damming by lava etc.
and a concave lee side (Fig. 2.27). Barchans are 3. Lakes formed by sedimentary processes, e.g.
found mainly on the edges of the erg area, where oxbow lakes in fluvial environments, and delta top
there is not enough sand to form a continuous thick lakes.
cover. Transverse dunes are common within ergs. Seif Lakes of glacial origin will have a relatively short
dunes are long, straight sand dunes which may occur lifetime by geological standards. In 10,000–100,000
within an erg but are also found in areas with incom- years most of the lakes in Scandinavia and North
plete sand cover. In the central zones very large dunes America will have filled up with sediment, if we do
form which may be over 100 m high and 1 km long. not have another glaciation.
Ancient aeolian deposits are recognised by: Lakes of tectonic origin, however, will continue to
1. Their large-scale cross-bedding, up to 20–30 m and subside. If the rate of subsidence keeps pace with the
often with wind ripple marks on top of slanting rate of sedimentation, a lake will continue to exist.
cross-bedded surfaces (foreset beds). Extensive carbonate beds (freshwater carbonates)
2. High degree of oxidation which gives a red colour. and diatom deposits are also common in lacustrine
3. Good sorting – largely medium- to fine-grained basins. Some of the largest lakes formed by rifting are
sand. found in East Africa, where a number of lakes occur in
64 K. Bjørlykke
a c
Windward side
b erosion
Cross-bedding (foreset)
Leeward side
deposition
Erosion
(Deflation)
Fig. 2.27 (a) Aeolian dune from the desert in Death Valley, “Yellow Sand” of northeast England (Old Quarrington Quarry,
California. (b) Schematic representation of an aeolian dune Durham). This sand is equivalent to the “Rothliegendes” sand-
(barchan). (c) Aeolian cross-bedding 4–5 m high in the Permian stone of the North Sea
the actual rift valley system. Lake Victoria, which lies and less dense water near the surface. This may
between two rift valley systems, was formed by tectonic restrict oxidation of organic material and promote
movements about 100,000 years ago. the development of organic-rich sediments (source
In humid climates all lakes will have outlets in the rocks) in the deeper parts of the lakes. In warm
form of rivers or via the groundwater, but in arid climates the temperatures of the lake waters are
climates lakes develop into evaporite basins without always above 4 C which makes the water stratifi-
outlets. Normal, non-saline lakes differ from ocean cation rather stable. In colder climates the water
basins in several ways: column will be inverted when the temperatures falls
1. Low salinity leads to slower flocculation of clay to 4 C which is the highest density. This results in
particles than in marine basins, producing more an oxygen supply to the lake bottom.
distinct lamination. 4. River water will generally have approximately the
2. Low wave and tidal energy and weaker currents same density as lake water, and will therefore mix
mean less erosion and resedimentation. Seasonal well and deposit sediments rapidly. Cold (glacial)
variation in the influx of sediments may produce river water or river water with a lot of suspended
annual cycles in lacustrine sediments. material will be heavier than lake water, however,
3. Limited water circulation makes it easier for the and will form turbidity currents along the bottom.
water to develop layering based on temperature 5. Lake sediments have a distinctly different fauna
(thermocline) and therefore also density from marine basins. The geochemistry of lake
(pycnocline) between dense layers at the bottom sediments and the composition of carbonates and
evaporites are also different from those of marine Rifting associated with the opening of the Atlantic
sediments. Ocean in Mesozoic times resulted in a large number of
By comparison with marine deltas, lacustrine deltas lacustrine basins, and sediments deposited in such
are among the most constructive of all (e.g. the Gilbert basins now underlie the continental shelves. In dry
Delta) because erosion in the basin is so limited. regions these turned into evaporite basins.
Nevertheless, wind conditions, the size of the lake
and the composition of the sediments will be crucial
factors. In large lakes we might have relatively high 2.27 River Deposits
wave energy which could erode sediments from the
river mouth and deposit them laterally on beaches. Transport of sediments in rivers depends on the gradi-
Wind will also cause water to well up from the ent and cross-section of the river, which determine the
bottom, increasing the circulation of oxygen along flow rate, and on the composition and concentration of
the bottom. sediment. In rivers which flow only during and directly
In dry regions we may get evaporitic lakes and after the rains, but are otherwise dry (ephemeral
lakes which dry up after each flood (ephemeral lakes). streams), equilibrium between flow and sediment
Lake deposits are characterised by: load is not achieved, so that even fine-grained sedi-
1. Lack of marine fossils, but diatoms and algae may ment is deposited in the channel when it dries up,
be important components. leaving a mud-cracked surface. Intensive oxidation
2. Very fine laminations in clay and silt sediments. of sediments, including silt and clay beds, is typical
3. Bioturbation structures may be found in lacustrine of such fluvial deposits. There is little organic produc-
sediments, but they are less common than in marine tion in areas with a dry climate, and clay and silt
sediments. particles will therefore contain little organic material
4. Chemical analyses of freshwater sediments will which could act as a reducing agent. A low water table
reveal low concentrations of the trace elements during dry periods also contributes to oxidation of
which are enriched in seawater (e.g. Cl, Br and organic matter.
B). Minerals formed in freshwater lakes may have Most rivers have marked seasonal and annual flow
characteristic isotopic compositions. Evaporites in variations. The flow velocity increases appreciably
lakes consist mainly of carbonate minerals; with flow volume because the friction per unit volume
sulphates and chlorides are normally less common. of water is inversely proportional to the water depth.
5. The content of sulphur-bearing authigenic minerals When water flow is greater than the capacity of the
must be relatively low since there is little sulphate channel, the water flows out over the banks. The water
in freshwater. However, organic material in lakes outside the channel will normally be very shallow and
contains a fair amount of sulphur which can be have a low flow velocity and fine-grained sediments,
precipitated as sulphides by sulphate-reducing bac- which have been transported in suspension, are then
teria. Sulphur may also be derived from volcanic deposited as what we call overbank sediments.
sources. Overbank sediment may build up into elevated banks
6. Lack of tidal structures and similar marine called levees. The resulting soil along modern rivers is
structures is also an important indicator. often very rich and ideal for cultivation. In post-
Amongst the most important ancient lacustrine Devonian sedimentary rocks we often find traces of
deposits are the Karoo deposits of South and East plants, commonly roots, in levees. Levee deposits
Africa, from the Carboniferous to the Jurassic periods. have parallel lamination and in some cases also cur-
These sediments contain important coal deposits. In rent ripples, particularly climbing ripples, which are
eastern China there are large areas of Cretaceous and typical of rapid sedimentation. The primary sedimen-
Tertiary lacustrine sediments, which are also impor- tary structures, however, will often be partly or wholly
tant petroleum-bearing sediments. destroyed by traces left by plant roots.
The Green River Formation (Eocene) of Colorado, Major floods may cover the areas beyond the levees
Wyoming and Utah, is a lacustrine sediment of great as well, and clay and silt will be deposited on these
extent which represents one of the most important flood plains. Many rivers which carry a large amount
petroleum source rocks in the world. of suspended sediment gradually build up their beds
66 K. Bjørlykke
through deposition in their channels and on the levees, therefore also typical of areas where the velocity
so that the surface of the water in the channels may be declines somewhat, e.g. when a river widens onto a
considerably higher than the surrounding plain. Dur- plain after passing through a narrow valley. The veloc-
ing floods the water may then break through the levee ity and flow will in most cases vary considerably with
and flow down from the channel to the plain, forming time. When the river is low it will flow round bars of
temporary lakes several metres deep. The lower sand and gravel, while at high water and with strong
Mississippi River is considerably higher than the currents sand bars will migrate as they are eroded on
surrounding area, including New Orleans. The major the upstream side and accumulate deposits on the
Chinese rivers have also built themselves up above the downstream side. Gravel bars, on the other hand,
surrounding countryside, so floods can cause very have rather complex patterns of erosion and
great damage. deposition.
However, the stability of the levees will determine Meandering rivers have a wavelength which is a
how much a channel can build itself up, and clay-rich function of the breadth of the river. The wavelength is
sediments form stronger levees than sand and silt approximately 11 times the breadth or 5 times the
because of their cohesiveness. When the levees give radius of the river. There is also a relatively regular
way, sand will flow out of the channel and deposit ratio (about 7:1) between the breadth and depth of the
sandy sediments in fans or crevasse splays on the flood river.
plain. They are characterised by thin sand beds, often Meandering rivers move in loops, with the greatest
fining upwards, with an erosional base in the proximal velocity at the outer bank so that erosion is
part (nearest the channel). However, we may also find concentrated there. The velocity along the inner bank
small sequences which coarsen upwards as a result of is much lower, so sediment is deposited there. The
progradation of this fan, because grain size decreases flow velocity and shear forces against the bottom
towards the edge of the fan. also decrease upwards towards the top of the bank on
the inner side of the bend, and sediments are deposited
in a point bar which reflects the hydrodynamic
2.27.1 Channel Shapes conditions. At the lowest point there is a lag conglom-
erate followed by large-scale cross-bedding due to the
We distinguish between relatively straight fluvial flow being in the upper part of the lower flow regime.
channels and curved channels. Curved channels are Then come current ripples, and finally fine-grained
shaped rather like a sine curve. To describe the degree sand, silt and clay sediments corresponding to the
of curvature we speak of high and low sinuosity. lower part of the lower flow regime at the top of the
Sinuosity is defined as the ratio between the length profile (overbank sediments). This produces a fining-
along the channel and the length of a straight line up sequence from sand to silt and clay (Fig. 2.28).
through the meander belt, i.e. the length of the river
valley. Meandering streams have a single channel and
by definition a sinuosity greater than 1.5. Lithology Process
Mud Overbank deposition
Braided streams have a branched course, but in during flooding
most cases the river channels are fairly straight. Fine-grained Flow in the lower
These rivers are branched because the river channel sand and silt part of the lower
with ripple flow regime,
is not very stable, and because sediment is deposited in cross-lamination formation of
current ripples
the middle of the channel, forming small islands or
bars. Braided streams typically develop in coarse Medium to coarse Flow in the upper
sand with large-scale part of the lower
sediments containing little clay or silt which can cross-lamination flow regime,
often trough cross- formation of dunes
form stable levées. A higher stream gradient gives bedding and troughs
greater energy and greater erosion of the sides of the
Plane beds Higher flow regime
channel. An abundant supply of sediment leads to
Basal lag congl. Erosion.
sediment being deposited more rapidly than it can be
eroded and transported onwards, resulting in the for- Fig. 2.28 Fining-upwards point bar sequence deposited by a
mation of deposits in the channel. Braided streams are meandering river
The secondary flow, in a vertical section at right Rapid subsidence and low sand/mud ratios will
angles to the downstream flow direction, moves from increase the stability of the channels and we may
the outer bank where erosion takes place along the find an anastomising channel distribution where there
bottom, and up the inner (point bar) bank where depo- is little or no lateral acretion of the channel (Fig. 2.30).
sition takes place. This is a result of a difference in When the whole river channel shifts course (avul-
hydrostatic pressure because the surface of the water sion) the abandoned channel will fill up with mud and
slopes inwards towards the inner bank due to centrifu- form a clay plug.
gal forces. If we combine this movement with the main
flow of water down-river, we find a corkscrew or
helical movement (Fig. 2.29). At each bend in the 2.28 Summary of Fluvial Sedimentation
river the helical flow reverses direction. The point
bar becomes asymmetrical, with coarser material on Fluvial processes are fairly simple in principle. We
the upstream side, so that a perfect fining-upward know the physical laws which govern the flow of water
profile is not developed. The upper fine-grained part in channels and transport of sediment in water. The
and the overbank deposit will also be lacking. great variation in composition, structure and geometry
Variations in the depth of water in the channel will demonstrated by fluvial deposits is due to all the
also lead to departures from the ideal fining-upward variables which influence transport and sedimentation.
sequence. As we have seen, the most important factors are:
During floods the water may flow over the point bar 1. Climate, particularly precipitation and seasonal dis-
and form a little channel, or chute. This may be wid- tribution of precipitation
ened by further erosion to become the main channel. (a) in catchment areas
This process is called chute cut-off. The sinuosity may (b) along river valleys – vegetation stabilises river
also become so high that erosion cuts a channel banks.
through a narrow neck (neck cut-off), making a 2. The drainage area
straight course to a lower bend in the river. The (a) size
whole meander will then be abandoned by the river, (b) type of rock being supplied
and a small, curved oxbow lake is left, which will fill (c) topography – tectonic uplift.
with clay, silt and organic matter. 3. Subsidence of the alluvial plain.
In order for thick fluvial series to be deposited and
preserved, the fluvial plain must be located in a tec-
tonically subsiding area.
Catchments with a lot of shale and other fine-
Erosion grained sedimentary rocks will produce sediments
Point bar
with a high clay and silt content. The products from
weathering of eruptive rocks will contain a consider-
able amount of clay, mainly kaolinite, illite and smec-
Cla
Fi
tite, which increases the cohesiveness of the sediments
y
ne
sa
an
nd Sand and stabilises the fluvial channels. The climate along

d
sil
the river plain may be very different from that in the

t
Erosion drainage area but some vegetation can be supported by

Point bar the groundwater close to the river also in dry climates.
In humid climates the vegetation will help to stabilise
Deposition the river channel and reduce the velocity of the flood-
water outside the channel.
The flow of rivers on the river plain will also
depend on the groundwater table. The river will con-
tribute water to the groundwater if it is lower than the
Fig. 2.29 Diagram showing how the flow in a vertical section surface of the river, and groundwater will flow into the
reverses direction in each bend (helical flow) river if the reverse is the case.
68 K. Bjørlykke
1 Braided
2 Meandering
3 Anastomosing
Fig. 2.30 Different types of fluvial channels. The type of fluvial channel determines the distribution and geometry of sand bodies
and depends on the sediment composition and slope variations in the water flow
2.29 Delta Sedimentation rocks will give sandier sediments because of their
quartz and feldspar contents. Deltas prograde out-
Large deltas require drainage areas with sufficient wards into a sedimentary basin and form a surface
precipitation to produce high runoff. They are formed called a delta top (delta plain) near the lake or ocean
along passive plate margins where extensive drainage level. The waves break against the delta front,
systems can unite into large rivers. The greater part of beyond which is the delta slope.
the drainage from North and South America and The formation of a delta can be depicted as a battle
Africa flows into the Atlantic Ocean. Only minor between the fluvial development of the delta and its
rivers enter the Pacific Ocean from the American erosion by marine forces. We therefore distinguish
continent, and relatively little of the drainage from first between river-dominated deltas, which prograde
Africa enters the Indian Ocean. The major rivers fol- far out into the basin and consist largely of fluvial
low old drainage systems, the main features of which sediments, and deltas which are eroded more rapidly
have existed since the Mesozoic, and which often by marine forces (tide and waves) and consist largely
seem to have been governed by Mesozoic rifting of marine sediments (Fig. 2.31).
along the Atlantic continental margin. River-dominated deltas are often referred to as
As mentioned in connection with fluvial constructive deltas, since they tend to prograde more
sediments, the erosion area will determine the com- rapidly into the basin without much marine reworking.
position and grain size of the sediments which are Tide- and wave-dominated deltas are often called
transported by rivers and deposited on the delta. The destructive deltas.
Mississippi, for example, drains extensive tracts of We may therefore distinguish between three main
Palaeozoic and younger sediments which contain a types of deltas: (1) river-dominated, (2) tide-
large percentage of shale, and the river therefore dominated and (3) wave-dominated. Most deltas fall
transports a high percentage of clay and silt in sus- somewhere between these three extremes. The
pension. Areas of metamorphic and acid eruptive Mississippi, however, comes near the river-dominated
High-destructive
Wave
Strandplain
High-constructive
Lobate
High-destructive
Tide
High-constructive
Elongate
Major sand facies Fine-grained facies
Fluvial sands Delta plain
Delta front Prodelta and shelf
Channel mouth bars Tidal flats
Distal bar Splays (on map view)
Sheet sands and shoreface
Tidal sand bars Vertical scale greatly exaggerated
Fig. 2.31 Classification of delta types based on the relative strength of waves and tidal energy in relation to the rate of sediment
input. After Fisher et al. 1974
end in a marine environment because the sediments The difference in density between the river water
are fine-grained and rather cohesive. The delta front is and the water in the marine basin plays a major role. In
also protected from big ocean waves by a shallow most cases river water will be lighter than saltwater,
shelf. even if it contains a good deal of suspended material.
High rates of sediment supply also favour fluvially- In deltas, river water will therefore flow far out over
dominated deltas. The Rhine is fairly wave-dominated the salt water (hypopycnal flow) before salt- and fresh-
as well, because the sediments are rather coarse- water mingle. If the river water and the water in the
grained and the sedimentation rate is lower. The basin have the same density (homopycnal flow) there
Ganges and Mekong deltas are typical tidal deltas. will be more rapid mixing of the water masses in axial
70 K. Bjørlykke
flow, and the sediments will settle out of suspension these sediments are stabilised by vegetation. In the
more rapidly. Hyperpycnal flow, where the river water Mississippi delta, sedimentation is very rapid so that
is denser than the water in the basin, takes place only the channel and the levees build up above their sur-
in lakes as a rule and leads to the flow of river water roundings. This means that the average gradient of the
along the bottom of the delta slope, with erosion of the channel decreases. Sooner or later the channel will
delta front and formation of turbidites on the basin have to find a new, shorter route (through avulsion)
floor. to the ocean, and which therefore has a somewhat
In studies of modern deltas we must also take into greater slope. At any given time most of the sediments
account the fact that most deltas are rather out of are deposited in one delta lobe prograding into the
balance as regards progradation in relation to sea ocean until the slope becomes to low. A major avul-
level because the Holocene transgression after the sion will then start the formation of a new delta lobe.
last glaciation raised sea level by more than 100 m. We can see that the Mississippi has constantly
In the Niger delta we find beyond the present delta shifted course, and has consequently been a focus of
front deposits which are 12–25,000 years old, which sedimentary activity in historic as well as modern
corresponds to the last advance of the last glaciation. times (Fig. 2.32). The present course has extended
the modern delta far out and should have been aban-
doned for a shorter course towards the southwest, to
2.30 River-Dominated Deltas the Atchafalaya basin. However the flow in this direc-
(Mississippi Type) tion has been artificially limited because of its impor-
tance for transport to and from towns like New Orleans
The Mississippi drains a huge area of the North Amer- which lie on the present channel.
ican continent (about 3.2 106 km2) with an average When a delta lobe is abandoned, it slowly subsides
precipitation of 685 mm/year. It carries vast quantities because of compaction and tectonic subsidence, while
of sediment (about 5 108 tonnes/year) with a high sedimentation takes place elsewhere. The abandoned
clay and silt content, and the gradient of the lower part lobe may sink below sea level before the fluvial supply
of the river is extremely low (about 5 cm/km). returns to this part of the delta and another delta lobe is
In the 200–300 years during which bathymetric deposited in the same area, allowing intervening depo-
measurements have been taken in the area, it has been sition of thin marine beds. Thin layers of carbonate or
possible to record considerable progradation. The high shale represent periods of local transgression (aban-
clay and silt content gives the sediments great cohesion, donment facies) which are time equivalent with
making the fluvial channels relatively stable. Clay and regressions in the prograding delta lobes. Between
silt sediments which are deposited on the delta plain the delta lobes in the interdistributary bay facies
have a high porosity and water content (60–70%), but wave energy is very low and there may be little or no
they loose much of their porosity by compaction at beach (sand) deposits between the marine mud and
shallow depth. Sand will be deposited chiefly in the mud deposited above sea level.
channels, and in mouth bars were the channels enter In vertical profile we observe alternations of fluvial
the sea. Well-sorted sandy sediments, which have only channel sediments, levee deposits, crevasse splays and
about 40–45% porosity immediately after deposition, possibly marine sediments (Figs. 2.33 and 2.34).
will then sink into the underlying clay because of their Even small variations in sea level have a strong
higher density. The fluvial channels therefore sink into influence on delta sedimentation. In a section where
the mud and this contributes to their stabilisation, with a fluvial facies gives way to a marine bed, it is often
the consequence that channels change course (avulsion) difficult to know whether this represents a transgres-
less frequently. Because of this subsidence while sedi- sion due to a rise in sea level or whether this part of the
mentation is continuing, the channel sand may be delta was abandoned by the fluvial system and is
thicker than the depth of the channel. The long strings subsiding. Only if we can correlate a transgressive
of sand which may then be preserved in the mud-rich bed over the whole area can we assume that it is due
environment are called bar-finger sands. to changes in sea level. Transgressive carbonate or
During floods large quantities of silt and clay are thin sandstone beds are the most useful for regional
deposited in overbank areas between channels, and correlation.
Modern Mississippi River subdeltas

a b A Dry Cypress Bayou complex
Prairie Terrace B Grand Liard complex
Baton
Rouge C West Bay complex
M D Cubits Gap complex
N iss E Babtiste Collette complex
.R F Garden Island Bay complex
. 2
New
B E
30° Orleans 1874
4 A D 29°15'
1862
3
1838 utre
s Lo
6 Pas
1 N 1891
5
F
So
C
uth
Garden
Pa
29° 7 West Bay Island
s
Bay
ss
Gulf of Mexico
as
East Bay
tP
Port 29°00'
1 Sale Cypremort 4 St. Bernard 7 Balize
es
Lads
2 Cocodrie 5 Lafourche
thw
0 50 km Burrwood 0
3 Teche 6 Plaquemine 15 km
u
So
92° 91° 90° After Kolb & Van Lopik, 1958 89° 89°30' 89°15' Carto Sect.,Csi,Lsu 89°00'
Fig. 2.32 Sedimentation in the modern Mississippi delta. When the river water breaks through the levees, crevasse channels and
splays are formed, which help to fill the areas between the channels (Coleman and Prior 1980). (a) Delta lobes which show how the
sedimentation has changed during the past 7,000 years. Each lobe of the delta appears to be active for 1,000–1,500 years (Coleman
and Prior 1980). (b) Sedimentation in the last few 100 years
Levée Swamp
Delta plain
silty sand and clay
Pure sand
Clay Delta front
diapir Sand and clay and silt,
silt little fauna
Prodelta clay
rich in fauna
Fig. 2.33 Fluvial channels in a clay-rich delta environment. The channel sand is denser than the clay and will sink into the
underlying mud. Clay diapirs may form as a result
2.31 Delta Front Sedimentation mouth. The wedge reduces the cross-section of the
river, and the velocity increases somewhat in conse-
At the channel mouth the fluvial water flow rapidly quence (Fig. 2.35). Spring tides will force the river
loses its energy, and all the sand which has been water up the channel and may cause it to overflow its
transported along the bottom (the bed load) is depos- banks.
ited in the form of a channel mouth bar (Fig. 2.34). The channel mouth bar itself is a very characteristic
Most of the suspended material is also deposited deposit which grades upwards from delta mud to well-
fairly rapidly, at a relatively short distance from the sorted sand on top. Because of brackish water and the
mouth. While the river water, which is lightest, flows high rate of sedimentation, there are few organisms
out of the channel, salt and brackish water flow back which live right at the mouth of the channel, but
in along the bottom, and may encroach a long way up bioturbation may occur in the surrounding sediments.
the channel when the rate of fluvial flow is low. The delta front is the area where fluvial and marine
Marine fossils can therefore be transported quite a forces meet, and we can virtually quantify the marine
way into the fluvial environment. During floods the influence on the delta. Tidal forces are a function of
saltwater wedge is forced back over the bank at the the tidal range, and from wave measurements we can
72 K. Bjørlykke
Fig. 2.34 Vertical section through modern Mississippi delta sediments. From Coleman and Prior (1980)
a Water flow out through a channel

Standing internal waves
strong turbulence Quick speed
Low rate of flow
reduction Ocean
Fresh water
Salt water
Mouth bar
At low flow of water a wedge of salt water may reach far up

in the fluvial channel
b Flood Strong turbulence
River water
Salt water
Erosion
At flood the wedge of salt water is forced back over
the mouth bank of the channel
Fig. 2.35 Flow of fresh and salt water at the mouth of a river channel (after Wright and Coleman 1974)
also estimate the wave power. While wave power near has a shallow profile beyond the delta front, most of
the beach is estimated to be 0.034 107 erg/s on the wave power dissipates before it reaches the delta
average for the Mississippi delta, it is 10 107 erg/s for front. Deltas which develop into deep water right by
the Nile, and 20.6 107 erg/s for the Magdalene River the continental slope are exposed to the greatest
(Wright 1978). In rivers like the Mississippi, which wave power, because waves are not damped before
they reach the delta front. The River Magdalene in diminished, and in consequence so is the shear
Venezuela is a very typical example. The thickness strength of the sediments. If the pore pressure attains
and the lateral extent of the delta-front sand is a the pressure exerted by the overlying sediments, the
measure of wave power. Powerful storms may, how- effective intergranular stresses will be equal to zero.
ever, erode vast quantities of sediment, far more There is then no friction between the grains, and
than evenly distributed wave power would have the sediments can flow like a liquid (liquefaction).
achieved. The resulting instability may cause diapirs of mud to
Wave-dominated delta front facies have a typical be squeezed up into the sand bed (Fig. 2.33). The
coarsening-upward sequence: from pro-delta clay to Mississippi delta is characterised by rapid sediment-
increasingly sandy sediments and finally well-sorted ation, and we find diapirs of overpressurised clays.
cross-bedded sediments with a low angle of dip, i.e. a Clay diapirs rise like salt diapirs because they are
beach facies profile. We often find traces of plant roots less dense than the more compact clay, silt or sand,
at the top of such sequences, and this shows that the which have lower water contents.
sand bank has had vegetation right down to the shore- The stability of sediments also depends on the
line. It was probably protected by shoreline barriers chemical composition of the porewater. In freshwater,
which absorb most of the wave energy. Vegetation can clay mineral particles with negatively charged
offer protection against both fluvial and tidal erosion, surfaces will repel one another. In saltwater these
and mangrove swamps such as those in the Niger delta surface charges will be neutralised by cations (Na+,
are particularly effective. There will always be some K+ etc.) so that clay minerals flocculate.
erosion on a delta front. Whether the delta progrades Deltas which prograde out into deep water will
or is broken down by marine forces depends on the develop a slope which may vary greatly. The force of
supply of sediment. The delta front will be fed with gravity acting on the sediments in the delta front and
sediment – particularly sand which migrates from the on the slope produces shear stresses in the sediments.
channel mouth bar – along the beach in the wave zone. When these stresses exceed the shear strength of the
We often call this “strike feeding” because sediment sediments, the sediments will be deformed by some
transport is parallel with the strike of the shoreline, i.e. sort of gravity-governed process. This may take place
along a horizontal line. This is in contrast to the through sliding and slumping which in turn may gen-
transport in channels, which is parallel with the dip erate turbidity currents, or steep fault planes may
of the deposit. develop, i.e. growth faults (Fig. 2.36). The name was
If there is little erosion of the channel mouth bar, introduced during early oil exploration and refers to
the channels will extend far into the sea, and little the fact that beds would thicken (grow) on the
sediment, least of all sand, will be supplied to the downfaulted side. The growth fault plane gradually
rest of the delta front. deflects and flattens out with depth.
2.32 Stability in a Delta Direction of progradation
Antithetic Rollover Main

Sediments which are deposited in a delta possess very fault anticline growth fault
high porosity, and clay- and silt-sized grains form a
A A
very unstable structure after deposition. Clay and silt
B
have a low permeability, however, and will expel
B B
water only very slowly. If sedimentation is rapid the C
sediment load will increase faster than the water can C C
flow out, and overpressure will develop in the D
D
porewater. This means that more of the overburden is D
E Possible development
carried by the porewater, reducing the effective
of clay (shale diapir)
stresses between the sediment grains and as a result
also the compaction. The friction between grains, Fig. 2.36 Main features of growth faults. The rollover
which is a function of the effective stresses, is greatly anticlines may be traps for oil and gas
74 K. Bjørlykke
Fig. 2.37 Section through a prograding delta system showing potential traps for oil and gas (Brown and Fisher 1977)
The growth fault may form near the delta front and material will be transported up-river. The difference
lead to this part of the delta subsiding rapidly, resulting between ebb and flood decreases gradually up the
in very thick deposits with delta front facies stacked on river, and the periods of rising water (when the tide
top of one another (Fig. 2.37). comes in) will be shorter, right down to 2–3 h. This
must be compensated by higher flow velocities. Con-
sequently it is often only the flow up the river channel
2.33 Tide-Dominated Deltas during high tide which gives rise to shear forces
against the bottom strong enough to transport a
A large tidal range affects delta sedimentation very bedload. Sand deposited in such rivers has structures
strongly and in a number of ways. The lower part of a (bedforms) which indicate transport up the river. It is
major river has a very low gradient, so when the sea very important to remember this when studying older
level rises and falls by 6–8 m it causes great rocks.
fluctuations in the gradient, with the tidal range often Rivers in areas with high tidal ranges broaden
affecting river flow as much as 50–100 km upstream. markedly as they approach the sea because of the
During flood tide the river surface gradient is reversed ever greater volumes of water to be transported out
and a great deal of sand (bedload) and biological and in. At the outermost point a broad estuary forms,
with clay and sand banks cut by channels which trans- unambiguous proof that one is dealing with a delta,
port tidal water. The Thames and the Rhine are however, is if the beach ridges are found to be cut by
examples of modern rivers with well-developed tidal fluvial channels. In addition to the Rhone and the
estuaries. Niger deltas, the Nile delta is also a good example of
The mixing of fresh and marine water also causes a wave-dominated delta Fig. 2.38. Here sediment sup-
flocculation of mud and deposition in the more ply to the delta has been greatly reduced through the
protected parts of the estuary between sand bars. Estu- building of the Aswan Dam, which traps sediments,
arine deposits are therefore characterised by a mixture and the balance between sediment supply and wave
of sandstone and mudstone, often with marine fossils erosion has been disturbed so that the delta is now
which may be transported several kilometres upstream retreating. This is a modern example of how
by the salt wedges along the bottom. Tidal flats are progradation or destruction (transgression) of a delta
often found in the inner parts of the estuary. Tidal depends on a very sensitive balance between sediment
channels are oriented at a high angle to the coastline, supply and erosion.
but in the outer part waves may produce elongated
bars which are oriented parallel to the coastline. The
geometry and orientation of the sandbars are very 2.35 Coastal Sedimentation Outside
different from those associated with wave-dominated Deltas
deltas and coastlines, where the sand bars are parallel
to the coast. Most sediment transport to the sea takes place via
Tide-dominated deltas are characterised by the fact rivers which drain into deltas. In fluvially-dominated
that the erosion due to tidal currents is very strong deltas, the greater part of the sediment is deposited
compared to erosion by waves. The fluvial channels there.
have cross-sections which are too small to allow trans- On destructive deltas, a great deal of sediment is
port of water out and in, and the result is the develop- eroded and transported out into the sea or along the
ment of a broad belt of tidal channels, which farthest coast. Coastal areas between deltas are very largely
out are separated only by narrow ridges of sand and supplied with sediment which has been eroded from
clay. Very often separate ebb and flow channels deltas and transported along the coast by waves and
develop, and in each channel one transport direction coastal currents. Since the surface of the coast slopes
will dominate. Signs of transportation in two opposite outwards, the direction parallel with the shoreline can
directions have often been used as a criterion for tidal be defined as the strike of the deposit, and the direction
deposits. This will apply to the whole area, but we normal to the coast is called the dip.
cannot always expect to find cross-bedding with oppo- Sediment supply parallel with the coast is often
site current directions in the same channel deposits. called “strike feeding”. If the supply of sediment
along the coast is greater than the erosion rate, the
coast will build out into the sea and we will get
2.34 Wave-Dominated Deltas regression through coastal progradation. If sediment
supply is less than the rate of erosion, the coast will
Wave-dominated deltas have a vertical sequence very retreat, and we will get transgression.
like that of a beach. Little of the fluvial part of the delta Because there is a finite amount of sediment being
is preserved, and instead of distinct deposits at each transported along any stretch of the coast, engineering
channel mouth (distributary mouth bars), the waves constructions which are sometimes built to accumulate
distribute the sediments in a continuous beach ridge sediment and prevent erosion at a particular spot
along the delta. The result is a coarsening-upward (groynes) will as a rule lead to increased erosion
profile which is not always easy to distinguish from a along other parts of the coast. This has great practical
beach deposit outside a delta. In a wave-dominated significance in areas with high coastal erosion. Beach
delta, however, beach ridges form right in front of the sedimentation must therefore be seen, not in isolation,
delta top deposits, which often contain coal beds from but in the context of the overall sediment budget along
vegetation, whereas beach profiles formed by barrier the coast.
islands have a lagoon behind them. The first
76 K. Bjørlykke
Mediterranean Sea
us
rull
e bu
Lak
L
Ma ake
nza Port Said
la
ku
e ld
Alexandria Lak
of Ni
Mary le
ke
La
)
ch
an
Br
(Ros
ett
a
(Daniett
a bra
nch)
Nile
Distributary channel facies
Coastal barrier - strand plain
Coastal basin
Coastal lowlands (salt marsh-alkali flat) 0 10 20 km

Scale
Delta plain
Shelf and prodelta Cairo
Fig. 2.38 The Nile delta system is an example of a wave-dominated, destructive delta system. (Fisher et al. 1974)
The Mississippi delta is the greatest source of sedi- uplifted tectonically, and where the coast erodes
ment to the Gulf of Mexico, although there are a metamorphic or eruptive rocks.
number of minor rivers in Texas. Here sediment trans- 2. Sand or Gravel Beach. This is the most common
port takes place along the coast from east to west. form of beach zone where we have active sedimen-
Because the active part of the Mississippi delta lies tation or erosion of older sediments (makes up 33%
as far east as it does today, less sediment arrives at the of the coastline of North America).
Texas coast. The fact that this coastline is subsiding at 3. Barrier Island. This is a beach which is separated
the same time leads to transgression of the coastline. from the main coastline by a lagoon. Common in
This example shows that if the supply of sediment due North America (22%), but less common in other
to strike feeding cannot match subsidence, the result areas.
will be a transgression. 4. Muddy Coastlines. Beach zones which consist basi-
cally of clay are formed where we have sediments
with a very low sand content, and where there is
2.36 The Shore Zone little wave activity to wash out or enrich any sand
the sediments might contain. We find this in parts
The shore zone is where land and sea meet, and we can of clay-rich deltas and estuaries and where there is
distinguish between different types: abundant vegetation along the beach zone which
1. Rocky Beach. The beach is covered with pebbles protects the clay sediments against erosion (e.g.
and blocks or solid rock representing ancient, resis- mangrove forest).
tant bedrock. Much of the coast of Norway is of this 5. Cheniers are isolated sandy beach ridges on coastal
type, but only 2% of the coastline of North Amer- mud flats. They are abundant along the coastline
ica. Rocky beaches form in areas which have been around major deltas like the Mississippi and the
Amazon. They require low tidal ranges, moderate
wave energy and abundant mud, and a limited In the French Riviera (e.g. Nice) the beaches are
amount of sand. often full of rounded pebbles with less sand. This is
The composition of beach sediments varies greatly, because they are near mountains supplying coarse
depending on the materials available and the grade of clasts and also close to a shelf edge. Sand is then easily
mechanical and chemical breakdown of the minerals. eroded from the beaches during storms while the
Quartz sand is the most widespread because quartz is coarse clasts remain.
the most stable of the minerals we find in most beach
sediments. But on volcanic islands, which consist
largely of basalt, there is no quartz, and we get sand 2.37 Prograding Beach and Barrier
derived from basalt or volcanic glass. Carbonate sand Sequences
is formed locally from the broken skeletons of
carbonate-secreting organisms. This does not only The progradation of a beach will produce a character-
happen in tropical regions, e.g. the Bahamas, but also istic coarsening-up sequence (Fig. 2.39) which will
along coasts with cold climates. In many parts of obey Walther’s Law of facies succession. The vertical
Norway, for example, the beach sand consists very sequence will represent the environments from the
largely of carbonate sand from molluscs, barnacles, shelf to the shoreline. The transition between shelf
bryozoans and calcareous red algae. mud and fine-grained sandstones with ripples may
Beach profiles are formed by wave power acting on represent the wave base. Isolated sand layers in mud
the coast and depend on a number of different factors: may have been deposited near the storm wave base,
1. The composition of the available sediments. whereas the transition to continuous fine-grained sand
2. Wave power may represent the fairweather wave base. The thick-
a. Average wave height ness of the sequence from the fairweather wave base to
b. Size and frequency of storms the foreshore is an expression of the wave energy at
c. Angle between the commonest orientation of the the coastline when the sequence was deposited.
waves and the beach. Bioturbation occurs in the lower part of this
3. Tidal range. sequence. Below the wave base it usually takes the
4. Vertical profile off the beach. form of horizontal feeding traces, and in the lower and
5. Supply of sediment from land or along the coast. middle shorefaces, where there is relatively high wave
The morphology of the beach zone and nearshore energy, as vertical traces (Skolithos facies). As erosion
areas is the result of interaction of various features and reworking intensify, the preservation potential of
related to wave activity. As waves approach land, bioturbation is reduced and it becomes less frequent.
they will be affected by friction against the bottom. The formation of sand bars and erosion surfaces on the
The depth at which this starts depends on wave height upper shoreface results in cross-bedding, usually
and length. When the depth becomes less than about trough cross-bedding, representing flow in the upper
half the wave length, friction against the bottom will part of the lower flow regime.
be great, the circular wave motion (orbit) will be In the breaker zone there is an upper flow regime,
distorted and oscillatory sediment transport will affect producing a planar facies which in vertical section will
the seabed, producing ripple marks. The depth at appear as very low-angle cross-bedding. On the beach
which this occurs is called the wave base. Dunes are we often have a beach bar which is flooded only during
formed of sediments which are deposited when the storms, and a depression behind it called a runnel,
waves break, and are at the same time the reason for which helps to drain the backshore area. Because the
the waves breaking precisely there. There is thus an exposed beach is a rich source of sand, the wind will
interaction between the wave regime and the bottom tend to blow sand from the beach and redeposit it as
geometry in beach sediments. In addition to breaking aeolian dunes, usually where it is trapped by vegeta-
on the foreshore itself, we often find that waves break tion. Aeolian sediments therefore often cap ancient
at two or three places offshore, and at each of these beach profiles, but the aeolian dunes may also be
places we find a sand bar. eroded and not be preserved in the geological record.
78 K. Bjørlykke
Aeolian sand dunes

Plant roots, foreshore bedding,
small dip
0m
Upper shoreface, cross-bedding,
some vertical bioturbation
2-4 m
Middle shoreface,
a good deal of bioturbation
4-6 m
Grain Lower shoreface,
fairly intensive bioturbation
6-10 m
size
Transition zone, intensive bioturbation,

primary lamination often destroyed
15-20 m
Offshore mud, shelf mud (clay and silt)
In some cases also sand,
often residual sand,
sediments completely bioturbated
200 m
Coarse sand
Medium sand
Fine sand
V. fine sand
Silt
Clay
Depth
Be unn
ac el
R
hb
ar
Sea level, high tide Sea level, ebb tide

2-4m
4-6 m
6-10 m
Ae
Ba
Up
Fo
Lo
Mi
oli
ck
pe
res
Tra
we
an
sh
dd
rs
Of elf m
ho
ore
le
Sh
n
rs
du
ho
fsh
sit
re
sh
ho
ne
ref
ion
ore ud
ore
ref
s
ac
zo
ac
fac
e
ne
e
e
Fig. 2.39 Diagrammatic section through a sand deposit with some typical sedimentary structures. Just under normal fair weather
wave base, 5–20 m, “hummocky” stratification occurs
2.38 Barrier Islands Georgia and North Carolina. There are several
indications that barrier islands are due to transgressive
Barrier islands are beach deposits which are separated conditions such as those of Holocene times. This is
from the mainland by a lagoon. They form long, thin certainly the case along the eastern coast of the south-
sand ridges which are often only a few hundred metres ern North Sea. When the ocean rises in relation to the
to a couple of kilometres broad, and which rise up to land, beach deposits can continue to grow through
5–10 m above sea level. gradual deposition of sand in the beach zone, but the
A vertical section through the part of a barrier areas behind them sink below sea level and form a
island facing the sea resembles an ordinary beach protected lagoon with clay sedimentation. More
deposit (Fig. 2.40). We find a coarsening-upwards localised transgressions may be caused by compaction
sequence from marine clay to beach sand, often with and subsidence of sediments along a coastline where
vegetation on top. On the lee side, facing the lagoon, the sediment supply is insufficient to keep pace with
there is little wave power and clay, mud and often the subsidence.
oyster reefs are deposited in the lagoon. Barrier islands One prerequisite for forming barrier islands is an
are very well developed along long stretches of the adequate supply of sand, so that the island can grow
coast of North America, particularly off Texas, and keep pace with the transgression. This sand cannot
Transgression
6
Lagoon
1 6 Ocean
2 3 4 5 7
5 Sea level
4
7
Relative sea-level rise

Legend
7 - Nearshore marine
6 - Barrier Stra
3 tigra
Holocene
5 - Back-barrier marsh unit phic

s
4 - Sublagoonal barrier 2
3 - Lagoon
1
2 - Fringing marsh Estimated time lines
1 - Pre-holocene
Regression
Lagoon 1 Ocean
5 Sea level
1
2
5
6
2
Legend 3
sea-level
Relative
Change
1 - Weathered,
laminated, Stratigraphic
fine to very units 3
fine barrier
sands
Holocene
4
2 - Laminated very fine
barrier sands
3 - Bored and bedded shoreface 4
barrier, very fine sands and silt
4 - Nearshore marine, interbedded sands
and silt-clay with shells
5 - Lagoonal, mixed sands and silt-clay
6 - Pre-holocene 6
Estimated time lines
Fig. 2.40 Transgressive and regressive barrier systems. (from Kraf and John 1979)
be transported across the lagoon, and must be added landward of the inlets small sandy deltas may develop
along the length of the islands parallel to the coast in response to ebb and flood currents respectively
(strike-feeding from deltas or eroding coastlines). (Fig. 2.41). Flow at ebb tide will normally be stronger
During storms or hurricanes the sea level may rise than that at flood tide. This is due to the profile of the
due to wind stress, and waves may break over and lagoons. At high water in the lagoon the volume of
through the barrier island. An erosion channel is then water which must flow out to compensate for a specific
formed through the island, and a washover fan delta lowering of sea level is greater than the volume needed
develops at the rear, out into the lagoon. A delta of this to raise the water level in the lagoon correspondingly
type can form in a matter of hours during a hurricane. at low tide when the area of the lagoon is smaller. The
Barrier islands may extend for tens of kilometres, but tidal inlets are therefore capable of transporting more
they will not form a continuous belt along the coast. sediment out during the ebb, and structures indicating
Water has to circulate between the lagoons and the this flow direction may predominate (Fig. 2.41). Inlets
ocean through gaps between barrier islands. The gaps are characterised by an erosional base, and lateral
are called tidal inlets, and the distance between them migration of inlets produces a characteristic fining-
will be a function of the tidal range. Both seaward and upwards sequence. The strong currents in tidal inlets
80 K. Bjørlykke
Lagoon
lta
l de
t ida
od
Flo
a
elt
a ld
b tid
Eb
Ocean
Fig. 2.41 Tidal channels with tidal deltas forming between accretion. Note that the ebb-tidal delta outside the barrier is
barrier islands. The barrier islands and the channels will migrate much more exposed to waves than the flood-tidal delta in the
laterally and deposit channel facies sediments by lateral lagoon
often generate sand waves which tend to migrate in the sediments or cohesive clays, erosion due to lateral
ebb direction, but they may also be modified by flood migration of tidal channels results in intraformational
currents. breccias.
Ebb-tidal deltas consist of a channel dominated by Barrier island deposits thus consist of a long, thin
ebb currents with smaller flood-tide channels on the body of sand. The thickness of the sand layer will
sides. At the ocean end of the channel sediment is correspond to the depth of the wave base plus a few
deposited in a sand ridge which is similar to a channel metres which correspond to the height it builds up to
mouth bar in an ordinary delta. This sand ridge, which above sea level.
is called a terminal lobe, is subject to wave erosion, In areas with a larger tidal range, this lateral migra-
and smaller swash bars may form, which reach above tion will be rather pronounced, and fining-upwards
sea level. In areas with strong wave power, ebb-tidal sequences will also be common.
deltas will be less obvious because of erosion and If the barrier islands are drowned by a transgres-
further transport along the barrier ridges. Ebb-tidal sion, a carpet of clay and silt will be deposited over
deltas will be characterised by greater water depths these sandstone deposits. This represents the ideal
than flood tidal deltas. stratigraphic trap for oil and gas. Compaction or tec-
Flood tidal deltas form inside the lagoon and are tonic tilting will cause the sandstone deposits to
well protected from wave erosion. Here the water interfinger with mud from the lagoon deposits, which
flows into flow channels which branch inwards in a are a good source rock. Oil will be able to collect in the
flood-tidal delta, where the sediments are deposited on top of the barrier ridge sand or in flood-tidal delta
a tidal flat. Ebb currents move back along the edges of deposits (or washover fans) which represent pinch-
the outer side of this delta and may form small spill- outs in the muddy lagoon sediments.
over lobes when ebb-currents penetrate over the edge
of the flood-tidal delta. Flood-tidal deltas are
associated with shallower channels than ebb-tidal
deltas and are not eroded very much by waves. 2.39 Tidal Sedimentation
Tidal channels fill with sand which forms an
upward-fining sequence overlain by tidal flat Tidal range is an important factor in coastal sedimen-
sediments (Fig. 2.42). In areas with carbonate tation. We distinguish between:
a Regressive (prograding) Transgressive b

Barrier model Barrier model
10 10 Sandstone, siltstone,
Backshore Argillaceous, coal lenses Marsh-tidal
Dune –Dune
8 8 Shale, brown, silty, flat
Washover Fissile up section
Backshore Sandstone, F.G., Planar & Subtidal
(Back barrier)
6 Foreshore 6 15 Cross laminated, burrowed lagoon
Metres
Metres
St.Mary river fm.

Marsh
Lagoon
Sandstone, silty,
Upper Washover Argillaceous, plant debris Washover
Channel
4 4 Tidal flat Oyster coquina bed
Shoreface
Flood tidal Shale, medium gray, Subtidal

Middle delta Friable, oysters lagoon
2 2 Sandstone, siltstone,
Subaqueous Coal lenses
lagoon Marsh-tidal flat
Lower Erosional surface
0 0
Sandstone, F. to M.G.,
Barrier-inlet 10 bipolar small to medium
Legend scale trough crossbeds Shallow
(Channel-delta)
model
Metres
abundant, some planar channel
Tidal inlet
Dune Flasers crossbeds & rare horizontal
10 Bioturbation, trace fossils
Spit beach plane beds, abundant
Shells, shell debris ophiomorpha burrows
Blood reserve fm.

8 Welded ridge Roots, organic debris
(Spit platform) Sandstone Deep
channel
Silty shale
6 Shallow channel 5
Metres
Coal lenses Erosional surface

Erosional surface
4 with lag deposit Sandstone, F.G., parallel Upper shoreface-
Plane beds laminations, plane beds foreshore
Barrier
beach
Deep channel Planar crossbedding ophiomorpha & arenicolites
2 burrows Middle
Trough crossbedding sandstone, V.F. to F.G., silty,
shoreface
Ripple laminae parallel laminations,
0 microcross laminations,
0 bioturbated, flasers
Fig. 2.42 (a) Interpretation of a vertical section in a tidal environment (From Walker 1979). (b) Section through a barrier beach cut
by a tidal inlet channel into the lagoonal facies
1. Microtidal environment (tidal range <2 m). friction against the bottom. When the water depth and
2. Mesotidal environment (tidal range 2–4 m). the velocity decline, the height of the tidal wave will
3. Macrotidal environment (tidal range >4 m). increase, so that the total energy flux is maintained.
The average tidal range in the open sea is only The tidal range, which is thus the height of the tidal
about 50 cm. Along the coasts, however, we often wave, therefore increases inwards across the shelf.
get increased interference by tidal currents. This is Where there are embayments along the coast, tidal
particularly true around large islands where tidal waves become focused, so that the tidal range
waves converge on the lee side and can build up, and increases towards the middle of the bay. This is true
also in bays along the coasts. In long, narrow bays we of the east coast of the USA, off Georgia, and of the
may get a high degree of resonance. This occurs if the German Bight in Europe.
bay has a length and depth which cause tidal waves If the shelf becomes even wider, as in an epiconti-
which are on the rebound to reinforce the next incom- nental sea, tidal power will gradually be exhausted due
ing tidal wave. The highest tidal range which has been to the high frictional resistance. This is the case on the
measured is 16.3 m in the Bay of Fundy in Canada. In present Siberian continental shelf where the tidal
some areas around the British Isles the tidal range may range is very low (<10–20 cm). Wind stress may set
be up to about 12 m, and there are also large ranges in up rather large storm tides. When the water is very
the German Bight and adjacent parts of the North Sea. shallow, as on the Bahamas Bank, friction damping
Tidal waves also move anticlockwise in N hemisphere takes place over much shorter distances.
around centres with zero tidal range (amphidromic The great epicontinental seas, in Ordovician and
points), with the tidal range increasing radially with Cretaceous times for example, were also characterised
the distance from the centre. This is typical of the tidal by low tidal ranges.
pattern in the North Sea. About 1/3 of the world’s coasts have tidal ranges
The width of the continental shelf plays a major greater than 4 m (macrotides), 1/3 have mesotides
role in determining tidal ranges. When tidal waves (2–4 m) and 1/3 microtides (less than 2 m). Small
enter shallow water, their velocity is reduced due to inland seas, like the Mediterranean, the Black Sea
82 K. Bjørlykke
and the Baltic Sea, are too small to keep pace with the such as in certain fluvial environments. On tidal flats,
attraction of the moon and the sun, and have small clay and silt will settle out at high tide to be deposited
tidal ranges. This is also true of lakes. between ripples formed by the ebb and flood currents.
1. Tidal Channels. The infills may resemble fluvial Here the fine sediment may consist partly of clay
channels or submarine channels in that they form pellets (faecal pellets from marine organisms) which
fining-upwards sequences. Channels formed in settle out faster than clay-sized particles. Flaser bed-
estuaries in fact are often connected to fluvial channel ding belongs to a type of structure which we get with
systems. Tidal channels which are not part of a river mixtures of sand and clay. Lenticular bedding
delta, however, tend to be filled with sandy sediments represents isolated laminae or lenses of sand in mud.
from the surrounding tidal flat, because they have no The inner parts of a tidal shelf often have
supply from land. Channels will often erode their embayments consisting of very muddy sediments, usu-
banks and cause them to collapse, and this may result ally with abundant bioturbation. Mollusc shells in the
in the formation of intraformational conglomerates if mud (Fig. 2.43) are often eroded and deposited as shell
the sediments are slightly lithified. lag.
Lateral migration of tidal channels may produce 3. Tidal Bundles. The best identifying feature for
typical epsilon cross-bedding which is the result of tidal environments is regular lamination consisting of
lateral accretion of point bars in the tidal channel. fine sand and mud, making up tidal couplets that each
Tidal channels often contain a bed of marine fossils represent a tidal cycle. Both modern and ancient tidal
at the base, and marine trace fossils. Channels on tidal sediments show a regular variation in the thickness of
flats which are not associated with deltas (i.e. those in such couplets, reflecting the energy levels of spring
estuaries) will not receive much clastic material from and neap cycles. Regular laminations reflecting tidal
land. Conglomerates and breccias in these tidal cycles are often called tidal bundles (Fig. 2.44).
channels will therefore typically be of the intrafor-
mational type, derived by local reworking of tidal
flat sediment. Because of their early lithification, car- 2.40 Shallow Marine Shelves
bonate beds in particular can be reworked to form
intra-formational conglomerates and breccias. The shelf extends from the nearshore environment to
Tidal cycle duration is about 12 h and 25 min, with the shelf edge, where there is a rather abrupt increase in
currents switching direction every 6 h, and we some- slope, usually at a depth of 200–500 m. The width of the
times find good examples of cross-bedding with oppo- shelf varies considerably, and may exceed 1,000 km.
site current directions. This is not always the case in Continental shelves are generally very flat areas which
tidal environments, however. Some tidal channels are may be cut by deeper channels transporting sediments
dominated by ebb flow and others by flood currents. across the shelf from nearshore or deltaic environments.
This is because the ebb and flood often find different The study of sedimentation on modern continental
dominant pathways. Bipolar cross-bedding is there- shelves is complicated by the fact that sea level was
fore not an essential feature of tidal channels. Storms more than 100 m lower only 10,000 years ago. This
with opposing wind directions may also produce some means that most shelf areas have not yet reached an
form of bipolar cross-bedding. equilibrium with respect to the modern environment.
In a regressive sequence tidal channels will be Sandy shelf deposits are much more difficult to core
overlain by lagoonal sediments (Fig. 2.42). than muddy sediments and the commonly used gravity
2. Flaser Bedding. Consists of clay laminae in a corer has to be replaced by vibro-core equipment.
matrix of sandstone with ripple cross-lamination. The Most shelf areas are below the wave base for nor-
clay occurs mainly as infill in ripple troughs, but may mal waves (fairweather wave base) and sedimentation
also drape over the ripple crests as well. Flaser bed- is governed largely by tidal currents and storms.
ding forms as a result of alternating periods of currents When the wind is landward, waves will usually
or wave activity, and slack water. The clay settles out approach the beach obliquely, resulting in wave
in the slack water periods in a tidal environment, refraction effects, particularly on relatively steep
though this type of bedding may also form in other beaches with high wave energy. This will produce
environments where there is rhythmic sedimentation, rip currents, where the water piling up against the
a b
Fig. 2.43 (a) Tidal channel on a tidal flat in the muddy facies of the inner part of a tidal flat (near Wilhelmshaven, Germany).
Erosion by tidal channels in this mud produces a lag of mollusc shells. (b) Mollusc in living position
effect of wind stress, tides and the low barometric

pressure associated with storms. The increased poten-
tial (elevation) of the coastal water due to these storm
surges will result in strong bottom currents which are
capable of transporting sediment further out onto the
shelf. Storm surges may transport fine sand and mud in
suspension, but not as true turbidity currents. In this
case the increased potential of the coastal water is the
driving force and not the density difference between
the current and the surrounding water, as with
turbidites. Hummocky cross-bedding is a characteris-
tic sedimentary structure produced by the deposition
of coarse particles from suspension during storms, and
rounded, undulating sand surfaces tend to form. Finer
particles will be transported further, into areas with
lower energy.
In many modern shelf areas tidal currents are
important transport mechanisms, in combination with
rare strong storm currents. Sediments deposited in this
environment are characterised by rather abrupt
transitions from well-sorted sand to mud and from
bioturbated to non-bioturbated strata.
Fig. 2.44 Tidal bundles. These are bundles of laminae which
The main characteristic of tidal shelves is the
reflect the tidal cycles between spring tides. From the Late
Precambrian Wonoka Formation, Patsy Spring, Flinders mobility of the clastic sediments, on scales ranging
Ranges, Australia from the diurnal tidal cycle to annual (storm aug-
mented) cycles and gradual long-term movement of
beach has its return flow seawards. Rip currents are the largest bedforms. In shelf areas with relatively
often rich in suspended material and transport material strong tidal currents (>150 cm/s) we may get furrrows
from the beach towards the shelf. Another component and gravel waves. At velocities of less than 1 m/s sand
of wave energy is transmitted parallel with the beach ribbons may be deposited – longitudinal bedforms
as longshore currents, contributing to the longshore developed parallel to the currents. Sandwaves are
drift of sediment transport. large-scale transverse bedforms, generally 2–15 m
During onshore storms the sea level near the coast high, with a wavelength of 150–500 m. Sandwaves
may be raised by several metres due to the combined
84 K. Bjørlykke
require current velocities exceeding 60–70 cm/s. We also find low oxygen content in the water column
Cross-stratification may be symmetrical or asymmet- on continental slopes, allowing much of the organic
rical on both sides of the sandwave, depending on the matter produced to be retained in the sediments. There
relative strength of the opposing currents. will also be high productivity on shallower slopes in
Relatively low-energy environments (<50 cm/s) front of deltas because of the large nutrient supply
are characterised by sand patches and mud. The sand from river water, but the organic matter may be greatly
forming the patches probably only moves during diluted by rapid clastic sedimentation. On the conti-
storms. nental shelf the supply of nutrients is small and the
Shelf sediments characteristically accumulate at prevalence of stronger currents and turbulence means
relatively low overall sedimentation rates (1–10 mm/ that most of the organic production there will be
1,000 years). oxidised.
Out in the open ocean basin organic production is
relatively low, due to a limited supply of nutrients.
2.41 Continental Slopes Much of the planktonic organic matter is oxidised near
the seafloor by deep currents of cold, oxygenated
Continental slopes are the areas between the edge of water from the polar areas.
the mostly very flat continental shelf which commonly Sediments deposited on the continental slope are
lies at a depth of 200–500 m and the continental rise, therefore more promising as source rocks for oil than
where the ocean deep begins at a depth of 2–4,000 m. shelf and deep-water facies.
The continental slope gradient is typically 2–6 , and it
is 20–100 km broad. The gradient is a function of a
number of different factors, but the stability of the 2.43 Sediment Transport on Submarine
shelf edge constitutes a major control factor. The Slopes
steepest slopes are therefore to be found off carbonate
banks with well-cemented coral reefs and carbonate Gravitational processes are naturally important on
beds which have high shear strength. In areas with submarine slopes.
rapid sedimentation, loose sediments have little shear Gravity forces can be represented by forces
strength and submarine slides, slumping and formation (vectors) normal to, and parallel to, the slope. The
of turbidites occur, maintaining relatively gentle component parallel to the slope consists of shear
slopes (1–2 ). Where sedimentation is slower the forces which may overcome the shear strength of the
sediments have more time to consolidate, and will be sediment, causing slumping. Sliding of large volumes
more stable. The steepest submarine slopes in clastic of sediments downslope produces extensional faulting
sediments (greater than 10 ) are therefore found in in the upper part of the slope and compression in the
submarine canyons, where erosion cuts into older, lower part. Gravitational instability on the slopes may
well-consolidated sedimentary strata. also develop into debris flows and turbidity currents.
Along passive continental margins the continental Collapse of the sediment grain framework may cause
slope is associated with the transition from continental sudden compaction and liquefaction of the slope
crust to oceanic crust. In areas with a large supply of sediments.
sediment, the shelf may have prograded beyond this Traction currents may, however, also play a part on
boundary. the submarine slopes, particularly in canyons but also
near the toe of the slope, where we may have
contourites – deposited by currents flowing parallel
2.42 Organic Sedimentation on the to the slope contours.
Slope of the Continental Shelf
The continetal slopes are enriched in organic matter

compared to the shelf and the deep ocean.
This is because the slope is where we have the
greatest upwelling of nutrients from the deep.
2.44 Submarine Canyons turbidity currents, but traction currents, which trans-
port and deposit better-sorted material. Turbidity
These are valley-shaped depressions which extend currents have been observed in submarine canyons as
from the top to the bottom of the slopes, down to well, but definite (observed) examples were only low-
2,000–4,000 m. In some cases they may start in shal- velocity, low-density turbidity currents with a maxi-
low water near the beach, in others close to the edge of mum velocity of 70–100 cm/s. In submarine canyons
the shelf. The height from the bottom of the canyon to we thus have both traction currents, which are con-
the top of the slope on each side may be up to 2,000 m. trolled partly by tidal forces, and turbidity currents.
We are dealing with enormous topographical features, Downward-moving currents driven by tidal forces
which would have been very impressive indeed if they may, if they contain much suspended material, turn
had been on land, towering structures on the scale of into turbidity currents. The downward-moving
the Grand Canyon. currents have both a component of traction and
Shepard et al. (1979) systematically gathered data gravitation.
on currents and sediment transport in submarine The relief of submarine canyons is due partly to
canyons and found that tidal currents are of great erosion down into the underlying sediments, and
importance also at great depths in submarine canyons. partly to lack of deposition in the canyon while the
Current meters have shown that currents flow both up adjacent beds were being deposited. During low sea
and down the submarine canyons, and that they switch level stands rivers may prograde closer to the shelf
every 6 h like tidal currents in shallow water. Current edge and hence supply more sediment to the subma-
velocity is often only 10–20 cm/s, but in many rine canyons, thus feeding submarine fans. At sea level
canyons velocities of up to 40 cm/s occur sometimes, highstands, currents in the submarine canyons and the
powerful enough to transport fine to medium-grained shelf may erode shelf and slope sediments and deposit
sand. The flow velocity tends to be greatest in the pure sand onlapping an erosional unconformity at the
upper part of the canyon and diminish downvalley. toe of the canyon.
Most sediment transport takes place during these epi- Most of the canyon itself is an area of sediment
sodic and unusually high flow velocities which may be transport and erosion. Deposition takes place where
linked with storms which create wind-induced shear there is a change of slope near the basin floor. Here the
forces which sweep the water up against the coast channel defined by the canyons splits up into several
(storm tides) and may cause currents to develop channels which build depositional lobes called
along the bottom and down the submarine canyons. suprafan lobes (Fig. 2.45).
However, high flow velocities have also been As the lobes build up, the gradient of the slope is
measured without it being possible to associate them reduced and a new channel will form in a part of the
with storms or wind stress. The most powerful flow fan where there is a steeper slope (Fig. 2.46). This
velocity is most commonly directed downvalley, but produces lobe-shifting similar to that observed in
upvalley-directed streams have been observed with fluvially dominated deltas. Each lobe will tend to
velocities of up to 90 cm/s. build a fining-upwards sequence, with conglomerate
Detailed measurements in the submarine canyons and coarse sand near the base. On the sides of the
off the coast of California reveal that the highest flow channels fine-grained material in suspension is depos-
rates are oriented up the canyon in a way which seems ited as thin-bedded turbidites. The levee builds up on
to indicate that they are generated by internal waves both sides of the channel, and resembles a river levee.
from the ocean basin, and not by gravitational forces. On the submarine Amazon delta slope there are well-
Currents may then develop when the waves “break” developed meandering channels. The distal fan is also
against the coast or the continental shelf. dominated by fine-grained sediments deposited as
Powerful currents in submarine canyons are capa- thin, graded fine sand, silt and clay. Progradation of
ble of transporting sand, sometimes in rare instances submarine fans may also produce an upwards-
even coarser material. These are frequently not coarsening sequence (Fig. 2.47).
86 K. Bjørlykke
Shelf edge Slope trough

Slope trough levee
s
Sh
nd
lf Filled trough
margin sa
e
A A'
Distal slope
e sands
tiv ad
ac
Su are
Leveed In fan lo
pr a
n
io on)
a
slope trough s
n iti
fa
A A'
pe os Proximal slope wedge
n
us p
(S de
Migrating Active
B slope troughs B' Migrating slope
Fan plain troughs
fan Active fan
Distal slope B B'

C sands C' Intermediate slope wedge
Bed load transport

No scale
Extreme vertical C C'
exaggeration Distal slope wedge
Slope fan model
Fig. 2.45 Model of submarine fans deposited at the foot of levée deposits resembling those along fluvial channels. The
submarine slopes. Submarine fans have channels which some- pattern of shifting depositional lobes resembles that of deltas.
times meander. On the sides of the channels we find fine-grained (Brown and Fisher 1977)
Feeder channel
Debris D-B Slumps
flows CGLS.
Slope into Basin

Thin bedded turbidites
Conglomerates: Upper on levee
Inverse-to-normally Fan
graded Pebbly SSTS.
Terrace
Graded-bed
Massive SSTS.
Graded-
stratified
Braided
Mid fan
Suprafan lobes
Smooth
Incised
Proximal channel
Classical turbidites
New suprafan
Lower fan lobe
Basin
plain
Thin bedded No relative scale implied
Fig. 2.46 Submarine fan model showing progradation and shifting of lobes similar to delta lobe shifting. One of the main
differences between submarine fan and delta facies is the absence of wave reworking. (Walker 1984)
Facies Sequence Interpretation are capable of transporting silt and fine sand. Clay
should not produce ripples, but the clay is often com-
SL. posed of pellets of clay aggregates, often also with
some organic matter. These clay particles behave like
D.F. F-U Inner fan
channel fill silt but they have lower densities and may be
CGL
transported by traction currents forming small-scale
CGL
ripple cross-lamination.
There may also be significant contributions of aeo-
Channelled lian dust from deserts and of volcanic ash. In
M.S. portion of
F-U suprafan Palaeozoic times, before there were many land plants,
lobes aeolian dust was more important, but even today much
P.S.
Suprafan lobes on mid fan

(CGL) of the sedimentation in the South Atlantic and Pacific
M.S. F-U
P.S. Channelled Ocean is aeolian.
C-U
C.T.
Smooth
2.46 Sedimentation Along Island Arcs
M.S.
and Submarine Trenches
C-U
C.T.
Smooth Submarine trenches form along converging plate
portion of
suprafan boundaries where oceanic lithospheres are
lobes disappearing into a subduction zone. Along these con-
C.T. C-U
verging plate boundaries sediment basins with very
special deposition environments are formed. There
are three main sources of sediment:
C.T. C-U Outer fan 1. From the continent.
Fining- or 2. From island arcs, which may consist of continental
coarsening- crust, oceanic crust and/or volcanic rocks.
upward
3. Pelagic sediment, including biogenic sediment and
Fig. 2.47 Vertical sequence through a submarine fan (Walker wind-blown volcanic ash.
1984) These have quite different compositions. Sediment
which is added from the continent is deposited in
2.45 Sedimentation Along Continental deltas and in turn fills the basin behind the arc (back-
Margins and in Epeiric Seas arc basin). Sediments which are formed on the island
arcs are rich in volcanic material, and this will charac-
As discussed above most of the sediment derived from terise deposits in small basins on island arcs and fore-
land is deposited in river deltas and distributed rela- arc basins. Back-arc basins may also receive volcanic
tively close to the shore line by longshore transport. sediments from the island arcs. During the initial sub-
There is, however, also significant sedimentation fur- duction phase fore-arc basins will tend to be
ther away from the deltas and the coastlines. Turbidites characterised by turbidites deposited in relatively
and debris flows can in some cases transport sediments deep water, but sediment filling may convert them
hundreds of kilometres offshore. Debris flows represent into a shallow water environment which may include
a very efficient type of sediment transport which can carbonate sediments. A structural high separates the
carry sediment far beyond the continental slopes fore-arc basin from the actual slope down to the sub-
because there is little internal shear deformation, only marine trench.
at the base and the top of the flow. Beyond the deep sea trenches a significant amount
In fine-grained shales we may find evidence of of pelagic sedimentation takes place on a relatively flat
ripples indicating traction currents. Storms and spring ocean crust. Some of the sediments get scraped off
tides may produce rather high velocities (20–40 cm/s) the subducting volcanic crust and stacked up in what
on the seafloor far away from land and these currents are called “accretionary prisms” (Fig. 2.48). Some
88
W Aves ridge Grenada trough St. Vincent Tobago trough Barbados E

(extinct arc) (back arc basin) (active arc) (fore arc basin) (fore arc ridge) Undeformed sediment
on ocean floor
Volcanics Crustal rocks Lavas & pyroclasts Undeformed sediment Small fore-arc basin Trench-slope break (mainly turbidites)
10
20
Decollement
Magma Eastward verging
30 Anomalously thick chamber Old basement Metamorphosed Oceanic layer
thrusts and folds
ocean crust surface and deformed in deformed sediment Oceanic layer 3
40 Lower crust of island arc oceanic Igneous sediments
Moho
Moho layer 3 & basic & ultra basic
Depth Km
50 intrusives
60 Ascending magma
and volatiles
70
80
100 200 300 400 500 600 700
Km east of long 64° W
Fig. 2.48 Section through an area with converging plate boundaries in the Caribbean Sea. Note that sediments deposited on oceanic crust are scraped off the descending plate and
imbricated in the accretionary prisms (see Fig. 2.49). From Leggett (1982)
K. Bjørlykke
N
2
4
SL
1 8
2 9 12
4 6
3 5
17
MTL 6
9 11
BTL 10 15
13
16
Acc. M
14
OP
Fig. 2.49 Sedimentation in a submarine trench (accretionary relative movements (thrusting) of the rock units piled up in the
prism) near Japan. Sedimentation is very much influenced by the prism (from Taira et al. 1982)
sediment may also be carried down with the very steep locally, and conglomerates and fan deposits
subducting plates. The supply of sediment to the may become unstable and slide. Lithified carbonates
deep-sea trenches themselves is often very limited, and sandstones will break up and form large blocks in
which is why they do not fill up with sediment. They a more clay-rich matrix. Volcanic rocks may also be
represent the greatest depths in the ocean (up to included in this package of broken-up sediments and
10 km) and this can be explained isostatically by the be incorporated into coarse conglomerates with large
fact that the oceanic plate which is undergoing sub- blocks called olistostromes. The blocks, which lie in a
duction is cold, and therefore heavy. The downward matrix of clay sediments, may be from a few metres up
movement acts against the direction of heat flow, to several hundred metres across. The result is called a
resulting in low geothermal gradients and therefore tectonic melange.
dense crust. The sediments in an accretionary prism are
The sediments may be pelagic oozes or distal subjected to strong tectonic deformation prior to
turbidites. Since the oceanic plate is moving towards deeper burial, and soft sediment deformation is a
the island arc, the sediments on the oceanic crust have very characteristic feature of such deposits. If we
been deposited further away from the sediment source, look at the total package of imbricated wedges, there
and in consequence we do not normally have very is a younging in the opposite direction, towards the
thick sedimentary sequences in the subducting plate. subduction zone. This is a feature that can be used to
The accretionary prism consists of a series of sliding recognise this depositional environment.
faults which are steepest near the surface and have a
lower gradient downwards. They are often draped with
a blanket of pelagic sediments (Fig. 2.49). 2.47 Summary
Listric faults of this type are similar to those we find
in plate boundaries with tension (rifting) but the rela- The study of sedimentary processes and facies
tive movements are in the opposite direction (reverse relationships is important for the prediction of the
faults). Sediments which are still not very consolidated distribution of different facies and rock properties.
tend to deform along the imbricated faults and develop We are interested in the geometry and distribution
various kinds of drag folds. Continued movements of of sedimentary facies and also the internal properties
the imbricated fault planes cause this slope to become of the sediments as they change during burial.
90 K. Bjørlykke
We therefore need to reconstruct sedimentary Davis, R.A and Dalrymple, R.W. 2012. Principles of Tidal
basins with respect to water depth, wave and tidal Sedimentology. Springer Verlag, 621 pp.
De Blasio, F.V., Engvik, L., Harbitz, C.B. and Elverhøi, A.
energy, and climate at different gelogical times. 2004. Hydroplaning and submarine debris flows. Journal of
The grain-size distribution and the mineralogy are Geophysical Research 109, C01002, doi:10.1029/
to a large extent controlled by the source area being 2002JC001714.
eroded and by transport processes. Fisher, W. L., Brown, L. F., Scott, A. J., & McGowen, J. H.
(1974). Delta systems in the exploration for oil and gas. A
The primary sediment compostion with respect to research colloquium. Austin, TX: Geology Building, Uni-
mineralogy and texture is very important with respect versity of Texas campus. 78 pp.
to the diagenetic process during burial. This will James, N.P. and Dalrymple, R.W. 2010. Facies Models 4.
deterime the rock properties recorded seismically, GEOtext 6. Canadian Sedimentology. Geological Associa-
tion of Canada, 586 pp.
and also reservoir quality. Sedimentology provides Kraft, J.C. and John, C.J. 1979. Sedimentary patterns and geo-
an important basis for basin analyses and for logic history of Holocene marine transgression. Geological
predicting rock properties, such as porosity, ahead of Society of America Bulletin 63, 2145–2163.
drilling. It is also very close to sequence stratigraphy. Leggett, J.K. 1982. Trench-Forearc Geology: Sedimentation and
Tectonics on Modern and Ancient Active Plate Margins.
See chapter 7. Geological Society Special Publication 10, 576 pp.
Much of the interest in applied sedimentology was Miall, A.D. 1984. Principles of Sedimentary Basin Analysis.
traditionally concentrated on predicting the distri- Springer, New York, NY, 490 pp.
bution of reservoir rocks, mainly sandstones and Mutti, E., Bernoulli, D. and Ricci Lucchi, F. 2009. Turbidites
and turbidity currents from alpine ‘flysch’ to the exploration
limestones, and their properties. Now with the of continental margins. Sedimentology 56, 267–318.
increased interest in unconventional oil and gas, shales Nichols, G. 2013. Sedimentology and Stratigraphy. Wiley-
have become very important, and the mineralogy is Blackwell, Chichester, 432 pp.
very critical. Nichols, G., Williams, E. and Paolola, C. 2008. Sedimentary
Processes, Environments and Basins. Wiley-Blackwell,
The changes from clay to mudstones and shales can Chichester, 648 pp.
not be understood without detailed clay mineralogical Perry, C. and Taylor, K. 2009. Environmental Sedimentology.
analyses and geochemistry. The primary mineralogy John Wiley, 460 pp.
and textural relationships determine the physical Reading, H.G. 1996. Sedimentary Environments. Processes,
Facies and Stratigraphy (3rd. ed.). Blackwell, Oxford, 688 pp.
properties during burial in sedimentary basins. This Shepard, F. P., Marshall, N. F., McLoughlin, P. A., & Sullivan,
is important for the interpretation of seismic data. G. G. (1979). Currents in submarine canyons and other sea
valleys. AAPG Studies in Geology, 8. 173 pp.
Taira, A., Okao, A.H., Whitaker, J.H.McD. and Smith, A.J.
1982. The Shimanto Belt of Japan. Cretaceous-lower Mio-
Further Reading cene active margin sedimentation. In: Legget, J.K. (ed.),
Trench-Forearc Geology: Sedimentation and tectonics on
modern and ancient plate margins. Geological Society Spe-
Allen, P.A. and Allen, J.R. 2009. Basin Analysis; Principles and cial Publication 10, pp. 5–26.
Applications. John Wiley, 560 pp. Talling, P.J., Lawrence, A., Amy, A. and Wynn, R.B. 2007.
Bjørlykke, K., Elvsborg, A., & Høy, T. (1976). Late Precam- New insight into the evolution of large-volume turbidity
brian sedimentation in the Central Sparagmite Basin of currents: Comparison of turbidite shape and previous
South Norway. Norsk Geologisk Tidsskrift, 56, 233–290. modelling results. Sedimentology 54, 737–769.
Brown, L.F. and Fisher, W.L. 1977. Seismic – Stratigraphic Tinterri, R., Drago, M., Consonni, A., Davoli, G. and Mutti, E.
interpretation of depositional systems. Examples from the 2003. Modelling subaqueous bipartite sediment gravity
Brazilian rift and pull-apart basins. In: Payton, C.E. (ed.), flows on the basis of outcrop constraints: First results.
Seismic Stratigraphy – Application to Hydrocarbon Explo- Marine and Petroleum Geology 20, 911–933.
ration. AAPG Memoir 26. Tulsa, OK, pp. 213–248. Tucker, M.E. 2009a. Sedimentary Petrology: An Introduction to
Coleman, J.M. and Prior, D.B. 1980. Deltaic Sand Bodies. the Origin of Sedimentary Rocks. Blackwell, Oxford, 272 pp.
Continuing Education Notes Series 15. AAPG, Tulsa, OK, Tucker, M.E. 2009b. Sedimentary Rocks in the Field. A Practi-
171 pp. cal Guide. John Wiley 288 p.
Collinson, J.D. and Levin, J. 1983. Modern and Ancient Fluvial Walker, R. G. (1984). Facies Models. Geoscience Canada.
Systems. International Association of Sedimentologists, Reprint Series, 211 pp.
Special Publication 6. Blackwell, Oxford, 575 pp. Wright, L.D. and Coleman, J.M. 1974. Mississippi River mouth
Collinson, J.D. and Thompson, D.B. 1982. Sedimentary processes: Effluent dynamics and morphologic develop-
Structures. Allen and Urwin, New York, NY, 194 pp. ment. Journal of Geology 82, 751–778.
Chapter 3
Sedimentary Geochemistry
How Sediments are Produced
Knut Bjørlykke
The composition and physical properties of sedimen- minerals and water at relatively low temperatures. At
tary rocks are to a large extent controlled by chemical temperatures above 200–250 C these processes are
processes during weathering, transport and also during referred to as metamorphism which is principally sim-
burial (diagenesis). We can not avoid studying chemi- ilar in that unstable minerals dissolve and minerals
cal processes if we want to understand the physical which are thermodynamically more stable at certain
properties of sedimentary rocks. Sediment transport temperatures and pressures precipitate.
and distribution of sedimentary facies is strongly At low temperatures, however, unstable minerals
influenced by the sediment composition such as the and also amorphous phases may be preserved for a
content of sand/clay ratio and the clay mineralogy. long time and there may be many metastable phases.
The primary composition is the starting point for the Many of the reactions associated with the dissolu-
diagenetic processes during burial. tion and precipitation of minerals proceed so slowly
We will now consider some simple chemical and that only after an extremely long period can they
mineralogical concepts that are relevant to sedimento- achieve a degree of equilibrium.
logical processes. Reactions will always be controlled by thermody-
Clastic sediments are derived from source rocks namics and will be driven towards more stable phases.
that have been disintegrated by erosion and The kinetic reaction rate is controlled by temperature
weathering. The source rock may be igneous, meta- and distance from equilibrium. The rate is zero at
morphic or sedimentary. The compositions of clastic equilibrium.
sediments are therefore the product of the rock types Silicate reactions are very slow at low temperature
within the drainage basin (provenance), of climate and and this makes it very difficult to study them in the
relief. The dissolved portion flows out into the sea or laboratory.
lakes, where it is precipitated as biological or chemical Biological processes often accompany the purely
sediments. Weathering and abrasion of the grains chemical processes, adding to the complexity. Bacte-
continues during transport and sediments may be ria have been found to play an important role in both
deposited and eroded several times before they are the weathering and precipitation of minerals. Their
finally stored in a sedimentary basin. chief contribution is to increase reaction rates, partic-
After deposition sediments are also being subjected ularly during weathering.
to mineral dissolution and precipitation of new In this chapter we shall examine the processes
minerals as a part of the diagenetic processes. For the between water and sediments from a simple
most part we are concerned with reactions between physical-chemical viewpoint. A detailed treatment of
sediment geochemistry is however beyond the scope
of this book.
Water (H2O) consists of one oxygen atom linked to
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway two hydrogen atoms, with the H-O-H bonds forming
e-mail: [email protected] an angle of 105 (Fig. 3.1). The distance between the

92 K. Bjørlykke
H H Ionic potential (I.P.) may be defined as the ratio

between the charge (valency) Z and the ionic radius R:
O
H
Z
H IP ¼
R
O Na + O 105° The ionic potential is an expression of the charge on

the surface of an ion, i.e. its capacity for adsorbing
H ions. Small ions carrying a large charge have a high
H
ionic potential while large ions with a small charge
O have a low ionic potential (see Fig. 3.2). The ionic
potential is strictly a function of the surface charge and
H therefore a function of R squared.
H
Ions with low ionic potential are unable to break the
Fig. 3.1 The strong dipole of water molecules causes them to bonds in the water molecule and therefore remain in
be attracted to cations which thereby become hydrated. Small
solution as hydrated cations (e.g. Naþ ; Kþ ). This
cations will be most strongly hydrated and less likely to be
adsorbed on a clay mineral with a negative charge means that the ion is surrounded by water molecules
with their negative dipole towards the cation
(Fig. 3.1).
O and the H atoms is 0.96 Å, and between the hydro- This is because the O–H bond is stronger than the
gen atoms 1.51 Å. Water molecules therefore have a bond which the cation forms with oxygen (M–O bond-
strong dipole with a negative charge on the opposite ing, M ¼ metal); this is particularly true of alkali metal
side from the hydrogen atoms (Fig. 3.1). This is why ions (Group I) and most alkaline earth elements
water has a relatively high boiling point and high (Group II, I.P. <3). Metals with an ionic potential
viscosity, and why it is a good solvent for polar only slightly lower than that required to form M–O
substances. Another consequence of this molecular bonds, namely Mg2+, Fe2+, Mn2+, Li+ and Na+, will be
structure is that water has a high surface tension, the most strongly hydrated. The hydration strongly
important for enabling particles and organisms to be affects the chemical properties of the ion and its capac-
transported on its surface. The capillary forces which ity to be adsorbed or enter into the crystal structure of a
cause water to be drawn up through fine-grained soils mineral. Since the ions are surrounded by water
are also a result of this high surface tension. molecules, we can use the expression “hydrated
A number of concepts are particularly useful for radius” to describe the space occupied by the ion and
describing and explaining geochemical processes: its water molecules within a crystal structure
1. Ionic potential (Fig. 3.3).
2. Redox potential Eh If the M–O bond is approximately equal in strength
3. pH to the O–H bond (I.P. 3–12), the metal ion replaces
4. Hydration of ions in water one of the hydrogen atoms to form very low solubility
5. Distribution coefficients compounds of the type M(OH)n (see Fig. 3.2).
6. Isotopes Examples of these so-called hydroxides that we com-
monly encounter in sedimentary rocks as a result of
weathering are Fe(OH)3, Al(OH)3 and Mn(OH)4.
3.1 Ionic Potential These hydroxides have very low solubility.
Ions with high ionic potential (>12) form an M–O
Ionic potential is a term introduced by V.M. bond that is stronger than the H–O bond, giving solu-
Goldschmidt to explain the distribution of elements ble anion complexes such as SO4 , CO3 , PO43
in sediments and aqueous systems. It must not be and releasing both of the H+ ions into solution.
confused with ionisation potential. Recent authors This approach can be used to explain the behaviour
have proposed the term “hydropotential” for the con- for elements on both sides of the Periodic Table (elec-
cept, to avoid confusion. tropositive and electronegative) which form ionic
3 Sedimentary Geochemistry 93
Fig. 3.2 Ionic radius and 2.0

charge (valence) for some Soluble cations e.g. Na+
geochemically important Cs
elements. Ions with low ionic 1.5 Rb
potential are soluble as cations Ba Hydrolysates e.g.
K
(e.g. Naþ ; Kþ ) while ions
Ionic radius
Fe(OH)3 .Very low
Sr
with intermediate ionic Th solubility
1.0 Na Ca R.E U
potentials will bond with OH
Mn Sc Zr
groups and have very low Fe
Li Fe Ti Nb.Ta
solubility, forming Mg Ga Mn V
hydrolysates (e.g. Al(OH)3), 0.5 Al
Si Soluble anion
Fe (OH)3). High ionic Be P S
complexes e.g.
potentials make soluble cation B
C N SO4--
complexes like CO3 and
SO4 . The ratio between 0 1 2 3 4 5 6 7
these parameters - the ionic Charge (valency)
potential - can be used to
explain their behaviour in
nature
Fig. 3.3 Ionic radius (in Naked radius (r)

Ångstrom units) of hydrated Hydrated radius (R)
and non-hydrated (“naked”)
ions of alkali metals and
alkaline-earth metals. The
smaller ions have higher ionic
Li +, r = 0.6 Å Na +, r = 0.95 Å K +, r = 1.33 Å Rb +, r = 1.48 Å Cs +, r = 1.69 Å
potentials and form stronger
R = 3.8 Å R = 3.6 Å R = 3.3 Å R = 3.2 Å R = 3.2 Å
bonds with water molecules so
that they become hydrated.
This hydration effect is
reduced with increasing
temperature
Mg ++, r = 0.65 Å Ca ++, r = 1.0 Å Sr ++, r = 1.13 Å Ba ++, r = 1.43 Å

R = 4.2 Å R = 4.0 Å R = 4.0 Å R = 3.0 Å
bonds. The elements in the middle, however, have a potential and are much more soluble that Fe3+ and
greater tendency to form covalent bonds in which the Mn4+.
strength of the M-O bond is not merely a function of The most soluble ions remain in the seawater until
the valency and radius, and the picture becomes far they are precipitated as salt when seawater is
more complex. The concept of ionic potential is nev- concentrated during evaporation. In addition to the
ertheless still useful; we see that during weathering, chlorides (i.e. NaCl, KCl), these are mainly salts of
elements with low ionic potential remain in solution cations with low ionic potential, and of anions with
along with the anionic complexes of metals and non- high ionic potential, e.g. CaSO4.2H2O, Na2CO3 and
metals with high ionic potential. This is reflected in the carbonates such as CaCO3 (calcite), CaMg(CO3)2
composition of seawater. The hydrolysates, on the (dolomite) and MgCO3 (magnesite).
other hand, become enriched on land as insoluble The principle of ionic hydration and the size of the
residues or through weathering (Al3+, Fe3+, Mn4+, ionic radius are capable of explaining a whole range of
Ti4+, etc.). Note also that Fe++ and Mn++ which geochemical phenomena. Among the Group I
occur in reducing environments have lower ionic elements of the Periodic Table, we know that Li+ and
94 K. Bjørlykke
Na+ are enriched in seawater. This because the strong Even if Mg is preferred in the carbonate structure and
hydration prevents absorption on clay minerals which also in the clay minerals, very little magnesium is
usually have a negative surface charge. K+, Rb+ and usually available in the porewater in the deeper parts
Cs+, on the other hand, have larger ionic radii and of sedimentary basins except in the presence of
consequently are less strongly hydrated. This leaves evaporites with Mg salts.
them with a more effective positive surface charge
which facilitates their adsorption onto clay minerals,
etc. 3.2 Redox Potentials (Eh)
This is demonstrated in nature during weathering
and transport. While similar amounts of potassium and Oxidation potential (E) is an expression of the tendency
sodium are dissolved during weathering of basement of an element to be oxidised, i.e. to give up electrons so
rocks, the potassium concentration in the sea is much it is left with a more positive charge. This potential can
lower (K/Na ratio of only 1:30). This is because K+ is be measured by recording the potential difference (pos-
more effectively removed by adsorption because it is itive or negative) which arises when an element
less protected by hydration. The same is true to an functions as one electrode in a galvanic cell. The other
even greater extent for Rb+ and Cs+, which are electrode is a standard one, normally hydrogen. The
adsorbed even more readily. These ions therefore oxidation potential of the reaction H2 ¼ 2Hþ þ 2e
have a relatively short residence time in seawater, (electrons) is defined as E0 ¼ 0:0 V at 1 atm and H+
between being delivered by rivers and then removed concentration of 1 mol/l at 20 C. Different conventions
by accumulating sediment. The ratio between the have been used to assign plus and minus values. In
average concentration of an ion in river water and geochemical literature, metals with a higher reducing
the concentration in average sea water is an expression potential than hydrogen are assigned negative values,
of the residence time. e.g. Na ¼ Naþ þ e ¼ 2:71V, while strongly
The large potassium ions can only fit in to certain oxidising elements are given a positive sign, e.g.
crystal structures. In mica and illite each potassium ion 2F ¼ F2 þ 2e ¼ 2:87V. A list of redox potentials
is surrounded by twelve oxygen or OH groups. shows which elements will act as oxidising agents,
With regard to Group 2 elements, Mg++ for example and which will be reducing agents. Reactions which
will be more strongly hydrated than Ca++ because it is result in a negative oxidation potential (E) will proceed
a smaller ion. As a result, Mg++ has a greater tendency spontaneously, while those which have positive voltage
to remain in solution in seawater and has a residence will be dependent on the addition of energy from an
time which is 13 times that of calcium. However, outside source. We can predict whether a redox reaction
despite the fact that the Mg/Ca ratio in seawater is 5, will occur by using Nernst’s Law (see chemistry
it is calcium carbonate which is the first to form textbooks).
through chemical and biological precipitation. Dolo-
mite or magnesite do not precipitate directly from 3.3 pH
seawater and this is in part due to the strong hydration
of Mg++. Normally, if we had naked (unhydrated) ions, The ionisation product for water is
MgCO3 and FeCO3 would be more stable than CaCO3 þ 14 +
½H ½OH ¼ 10 . The concentration of H in
because Mg++ and Fe++ have greater ionic potentials
neutral water will be 10 7. pH is defined as the nega-
and stronger bonding to the CO23 ion. However with tive logarithm of the hydrogen ion concentration activ-
increasing temperature the hydration declines because ity, and is therefore 7 for neutral water (at 25 C).
the bonds with the dipole of the water molecules However, the ionisation constant varies with tempera-
become weaker. Mg++ is then more likely to be ture, e.g. at 125 C the ionisation constant for water is
incorporated into the carbonate mineral structures. ½Hþ ½OH ¼ 10 12 . In other words, neutral water
Therefore during diagenetic processes at 80–100 C, then has a pH of 6. It is important to remember this
magnesium carbonates precipitate more readily even when considering the pH of hot springs or in deep
if the Mgþþ =Caþþ and Feþþ =Caþþ ratios are low. wells, for example oil wells.
In nature the pH of surface water mostly lies corresponds to Eh ¼ 1:22 0:059 pH, beyond which
between 4 and 9. Rainwater is frequently slightly free oxygen would be released from the water. If we
acid due to dissolved CO2, which gives an acid also set pH limits at 4 and 9 in natural environments,
reaction: we can divide the latter into four main categories:
1. Oxidising and acidic
H2 O þ CO2 ¼ H2 CO3 ðcarbonic acidÞ 2. Oxidising and basic
H2 CO3 ¼ Hþ þ HCO3 ¼ 2Hþ þ CO23 3. Reducing and acidic
4. Reducing and basic
Variations of pH and Eh are the major factors
Humic acids may give the water in lakes and rivers
involved in chemical precipitation mechanisms in sed-
a low pH. Sulphur pollution from burning oil and coal
imentary environments where there is not strong evap-
gives SO2, which is oxidised in water to sulphuric
oration (evaporite environments).
acid:
The stability of some important compounds in
porewater may be presented in an Eh - pH diagram
2SO2 þ O2 þ 2H2 O ¼ 2H2 SO4
(Fig. 3.4).
The solubility of many elements is highest in the
In areas with calcareous rocks or soils this sulphuric
reduced state and they are precipitated by oxidation.
acid is immediately neutralised and the water becomes
This is particularly characteristic of iron and manga-
basic, as is the case across much of Europe. By con-
nese, whereas others such as uranium and vanadium
trast, in areas with acidic granitic rocks as in the south
are least soluble in the reduced state.
of Norway and large areas of Sweden, the rock does
not have sufficient buffer capacity to counteract acid
rain or acidic water produced by vegetation (due to
humic acids). Organic material also contains a certain 3.3.1 Distribution Coefficients
amount of sulphur, and drainage of bogs, or drought,
can produce an acidic reaction. This is because H2S When a mineral crystallises out of solution, the com-
from organic material is oxidised to sulphate when the position of the mineral will be a function of the com-
water table is lowered, allowing oxygen to penetrate position of the solution and the temperature and
deeper in these organic deposits. pressure. Trace elements which are incorporated in
The water near the surface of large lakes and the sea the mineral structure are particularly sensitive to
can have a high pH because CO2 is consumed due to variations of these factors. With constant temperature
high organic production (photosynthesis). If the and pressure, the concentration of a trace element
organic material decomposes (oxidises) on its way to within a mineral which is being precipitated, is pro-
the bottom, CO2 is released again, causing the pH to portional to its concentration in the solution. The ratio
decrease with depth since the solubility of the CO2 between the concentration of an element in the mineral
increases with the increasing pressure. and its concentration in the solution (water) is called
CO2 is also less soluble in the warm surface water the distribution coefficient (Kd).
than in the colder water at greater depth. A number of elements substitute for Ca++ in the
Seawater is a buffered solution, with a typical pH calcite lattice: Mn++, Fe++ and Zn++ have distribution
close to 8, though this varies somewhat with tempera- coefficients ðKd >1Þ. This means that they will be
ture, pressure and the degree of biological activity. captured, so that the mineral becomes enriched in
Eh and pH are important parameters for describing these elements relative to the solution.
natural geochemical environments, and the diagram
obtained by combining these two parameters is partic- Mnþþ =Caþþ ðmineralÞ ¼ k Mnþþ =Caþþ ðsolutionÞ
ularly useful.
The lower limit for Eh in natural environments is
defined by the line Eh ¼ 0:059 pH, because other- k ¼ KdðMnÞ=KdðCaÞ
wise we would have free oxygen, and the upper limit
96 K. Bjørlykke
Fig. 3.4 Eh-pH diagram showing the upper and lower limits for Eh and pH in natural environments. Note the stability fields of
hematite (Fe2O3), siderite (FeCO3) and magnetite (Fe3O4). (Constructed by Helge Hellevang)
k is here the ratio between the distribution 3.4 Isotopes

coefficients of Ca++ and Mn++) which is about 17,
that is to say the manganese concentration in the A number of elements occur in nature as different
calcite is 17 times greater than in the solution. isotopes: the atomic number (protons) is constant but
At low temperatures (25 C) Mg++, Sr++, Ba++ and there are different numbers of neutrons. They there-
+
Na have distribution coefficients <1. This means that fore have the same chemical properties although their
the mineral phase will contain proportionately less of masses are slightly different. Isotopes which are radio-
these elements than the aqueous phase. For Sr++, k is active (unstable) break down at a specific rate charac-
about 0.1 (0.05–0.14) in calcite, such that the Sr con- teristic for the isotope species (the decay constant). By
tent in calcite is relatively low. The Sr content in analysing the reaction products formed in the minerals
aragonite is considerably higher because the Sr++ ion, they can be dated. The 87 Rb 87Sr and the 40 K 40Ar
which is larger than the Ca++ ion, is more easily methods have been used in determining the age of
accommodated within the lattice. By analysing trace rocks. This has been replaced by argon-argon datings
elements in minerals like calcite we can infer some- 40Ar/39Ar. The ratios between lead isotopes can also
thing about the environment when the minerals be employed because of the 235 U 207 Pb, 238 U 206Pb
precipitated. Limestones with a high content of stron- and 238 Th 208 Pb reactions. Now age dating of zircons
tium may have had much primary aragonite which was is extensively used. Such age dating can be used to trace
replaced by calcite. Strontium Carbonate (strontianite) the source (provenance) of sandstones and other sedi-
may also form from the excess Sr. Calcite containing mentary rocks.
significant amounts of iron must have precipitated Dating sedimentary rocks is a complicated proce-
under reducing conditions because only Fe++ would dure and the results are often difficult to interpret.
be admitted into the calcite structure. Separating the newly formed (authigenic) mineral to
be dated, can be particularly challenging. Volcanic ash The concentration of stable isotopes in minerals or
beds may provide the best dates for a sedimentary water is given as a ratio compared with certain
sequence. standards. The standards can then be analysed on a
It is important to be able to identify the source of mass spectrometer together with the samples. The δ
clastic minerals in sedimentary rocks so that the areas value thus represents the deviation from the isotopic
of erosion and sedimentation can be reconstructed. ratio in the standard.
Minerals like zircon can be characterised and dated, This standard may be the average composition of
based on their isotopic composition (Nd isotopes and seawater, called SMOW (Standard Mean Ocean
U-Pb). Water). Another commonly used standard is PDB
(Pee Dee Belemnite), which is the composition of
calcite in a Cretaceous belemnite. The calcite
(CaCO3) was precipitated in the sea and its composi-
3.4.1 Light stable isotopes
tion was in equilibrium with the seawater at normal
temperatures (15–20 C). There is more 18O in calcite
The fact that isotopes have different masses causes
than in the water (positive fractionation), but with
fractionation to take place through both chemical and
higher temperatures the less effective fractionation of
biological processes. The simplest example is water,
H2O, which contains two oxygen isotopes and two oxygen lowers the δ18 O values. The relationship
hydrogen isotopes. The evaporation of water to vapour between the two standards is:
causes fractionation and concentrates the light water
molecule with 16O. This is because the 18O isotope has δ18 OSMOW ¼ 1:031 δ18 OPDB þ 30:8
greater mass and a phase change from fluid to vapour
therefore requires more energy. H16 PDB values are preferred for carbon isotopes in
2 O has higher
18 carbonate minerals but also for oxygen in carbonates
vapour pressure than H2 O. This is the reason why
while the SMOW scale is mainly used for water
rainwater and ice contain less 18O than seawater.
samples and silicate minerals.
Isotope fractionation is a function of temperature,
Hydrogen has two stable isotopes, 1H and 2H (deu-
however, and is much more effective with evaporation
terium), and an unstable one, 3H (tritium), which has a
at low temperatures than at high ones. The explanation
half-life of 12 years. The hydrogen isotopes are even
for this is that at high temperatures the energy of the
more strongly fractionated than oxygen isotopes dur-
molecules is so great that the difference in mass
ing evaporation. Water molecules with deuterium
between 18O and 16O is of less importance. At low
(heavy water) have lower vapour pressure than water
temperatures the isotopic separation evaporation is
molecules with hydrogen and during evaporation the
much more selective so that the water evaporated is
heavier deuterium atom is depleted in the evaporated
more enriched in 16O. When water vapour condenses
water and concentrated in the remaining water which
to rainwater, molecules with 18O will condense faster.
also has higher salinity.
Rain and snow becomes enriched in the heavier iso-
In meteoric (rainwater) water there is a linear rela-
tope (18O), so that the water vapour remaining in the
tion between the deuterium/hydrogen ratio (D/H) and
air becomes more enriched in 16O. Most of the evapo-
the δ18 O.
ration takes place at low latitudes and the water vapour
The isotopic composition of seawater has varied
in the air has a progressively lower 18O-content
through geological time, though not so much during
towards higher latitudes as the air cools and it rains.
the last 200–300 million years. During glacial periods,
The concentration of oxygen isotopes is expressed in
relation to a standard: seawater acquires more positive δ18 O values because
the water bound as ice has more negative δ18 O values.
δ18 O ¼ Rsample =Rstd Rainwater (meteoric water) has normal δ18 O values

1 1000
from –2 to –15. The values become more negative
R ¼ 18 O=16 O towards higher latitudes, and near the poles one can
98 K. Bjørlykke
The calcite precipitation formula is:
0 Precipitation T ¼ 16:9 4:38ð18 Ocarb 18

Owater Þþ
at low latitudes
Evaporation
–100 in closed 0:1ð18 Ocarb 18
Owater Þ2
basins
Precipitation
00
δD /
0
at high latitudes Here the values for calcite are given in PDB and for
–200 water in SMOW. We see that if the δ18O value for
calcite is 0 (PDB) and seawater has 0 (SMOW), the
temperature is 16.9 C, which may have been a typical
–300
sea temperature when the standards were
–50 –40 –30 –20 –10 0 +10 precipitated).
The above formula can be expressed graphically,
δ180 /
0
00
enabling the temperature to be read off a curve as a

Fig. 3.5 Ratio between the isotopic composition of seawater function of the isotopic composition of the calcite,
and freshwater. Evaporites will deviate from the mixing line
which is the assumed composition of the porewater
between these endmembers (modified from Faure 2005)
during precipitation (Fig. 3.7). Similar calculations
can be done for other precipitated minerals, for exam-
ple for quartz using the δ18 O fractionation as a func-
measure δ18 O values of about –50 and δD (2H) values tion of the temperature for quartz.
close to –350 (see Fig. 3.5). Minerals that form in Carbon has two stable isotopes (12 C 98:9% and
13
seawater show decreased 18 O=16 O ratios with C 1:1%). During photosynthesis a greater propor-
increased ambient temperature during formation. The tion of 12 CO2 than 13 CO2 forms organic compounds,
δ18 O=16 O ratio in carbonate-secreting marine because 12 CO2 has a smaller mass. Organic material is
organisms, for example, is thus a function of both therefore enriched in 12C relative to atmospheric CO2
temperature and salinity. The seawater changes its and HCO3 in seawater. The isotopic composition of
δ18 O values by around 1–1.5‰. Isotopes can thus carbon is expressed as δ13 C values:
provide important proxy evidence for palaeoclimate
δ13 C ¼ ½13 C=12 CðsampleÞ=13 C=12 CðstdÞ 1 1000
studies.
Cold freshwater gives strongly negative δ18 O All samples are compared against a standard of
values, whereas evaporites are enriched in 18O marine calcite, the PDB belemnite, which by defini-
isotopes (positive δ18 O). Shallow marine carbonates tion has δ13C ¼ 0‰ PDB. The isotopic composition of
that are diagenically modified by freshwater, give dissolved carbon (CO2) has been relatively constant
lower δ18 O values than marine carbonates deposited during the last 300–400 million years, but limestones
in deeper water. can nevertheless be dated and correlated using
Stable oxygen isotope analyses were first used by differences due to variation in the composition of
Urey, in 1951, to demonstrate past temperature seawater. Towards the end of the Precambrian the
changes in seawater. By taking samples through a composition of seawater seems to have been more
cross-section of a belemnite it was possible to register variable, and there this type of correlation is particu-
annual variations in seawater temperature from 150 larly valuable since there are no fossils. In large mas-
million years ago (Fig. 3.6). sive limestones the isotope composition does not
The precipitation of newly formed (authigenic) change significantly during diagenesis, because the
minerals gives an oxygen isotope composition which volume and the mass of carbon is so great. Atmo-
is a function of the composition of the porewater in spheric CO2 has δ13C ¼ –7‰. Land plants have an
which the mineral is precipitated, and the temperature. average δ13C value of –24 (–15 to –30‰), and marine
If the porewater isotope composition is known, the organisms have a similar range of values. Freshwater
temperature (T) can be calculated, and vice versa. containing CO2 released by the breakdown of organic
018 t°c
016 2.0
20
19
18
17
16
15
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 cm

Radius
Fig. 3.6 Analyses of oxygen isotopes in a Jurassic belemnite from the centre to the outermost layer. Colder water during the winter
is recorded by lower 18O/16O ratios. We can see that the belemnite lived for 4.5 years and died in the spring. (From Urey et al. 1951)
matter, and groundwater filtered through a soil profile, The strontium isotope ratio ð87 Sr=86 SrÞ in seawater
will take up CO2 with negative δ13C values from roots has varied considerably through geological time
and organic material. (Fig. 3.8). This is because there are two radically
Bacterial fermentation of organic material different sources of strontium in seawater. Continental
ð2CH2 O ¼ CH4 þ CO2 Þ forms gas (methane) which weathering supplies much 87Sr to the sea since granitic
is very strongly enriched in 12C (δ13C ¼ –55 to –90‰) rocks contains relatively high concentrations of rubid-
and CO2 which is positive ðδ13 C ¼ þ15Þ. ium which can decay to 87Sr. Dissolution of basalt at
Thermal breakdown (thermal decarboxylation) of the mid-oceanic ridges will supply strontium with a
organic matter produces δ13C values of –10 to –25 in relatively low 87 Sr=86 Sr ratio because basalt contains a
calcite.
100 K. Bjørlykke
160
8
140
6
4
120
2
0
Temperature oC
100
)=
OW –2
M
80 (S –4
O
H2
8O
60 δ1
Late Calcite
40 Cement (type I)
Late Calcite
Cement ( type II )
20
Measured range in calcites ( type I & II )
0
2 0 –2 –4 –6 –8 –10 –12 –14
δ180 Calcite (PDB)
Fig. 3.7 Relation between the isotopic composition of porewater and carbonate cement, as a function of temperature. (From Saigal
and Bjørlykke 1987)
little potassium and also rubidium which could form Detrital zirons separated from sandstones can be
87
Sr. used to indicate the source (provenance) of the
When there is rapid seafloor spreading a great deal sediments based on their U-Pb age (Davis et al. 2003).
of water passes through the mid-oceanic ridges, so that The composition of feldspar and heavy minerals
the seawater receives much Sr with a low 87 Sr=86 S may also be used to indicate the areas of erosion
ratio. During such periods, for example in the Jurassic supplying sediments to the basins.
– Cretaceous, the creation of new warm sea floor will
lead to a transgression onto the continents. This
reduces the gradients and hence transporting capacity
of rivers, limiting the supply of clastic material to the 3.5 Clay Minerals
ocean.
A number of minerals are referred to as clay minerals
Since the Jurassic, the 87 Sr=86 Sr ratio has risen
because they predominantly occur in the finest grain-
almost continuously, and by analysing marine calcitic
size fraction (clay fraction) of sediments and sedimen-
fossils such as foraminifera, one can obtain rather
tary rocks. However, this is not an accurate definition,
accurate age determinations. This applies particularly
because the clay fraction contains many other minerals
to the Tertiary period, when the rise in the 87 Sr=86 Sr than those we call clay minerals, and because the clay
ratio was particularly rapid (Fig. 3.8). minerals themselves are often larger than 4 μm
The isotopic composition of clastic sediments can (0.004 mm). By “clay minerals” we usually mean
also be used for stratigraphical correlation. Then it can sheet silicates which consist chiefly of oxygen, silicon,
be more useful to employ isotopes which do not go aluminium, magnesium, iron and water (H2O, OH–).
into solution and react with water, but retain the origi- Clay minerals in sedimentary basins are partly derived
nal age of the rocks from which they were eroded. from sheet silicate minerals occurring in metamorphic
In the North Sea and on Haltenbanken the ratio and igneous rocks (e.g. biotite, muscovite and chlo-
between the rare earth elements samarium and neo- rite), but during weathering and transport these clastic
dymium ð147 Sm=143 NdÞ was used to correlate reser- minerals are typically altered from their initial compo-
voir rocks both in-field and regionally. sition in the parent rock.
87
Fig. 3.8 87Sr/86Sr ratio in Sr/86Sr
seawater from the Cambrian
to present. Based on 0.709
MacArthur et al. (2001). This
87
ratio reflects the relative Sr from land
contribution form weathering (weathering and
of continental rocks with high erosion)
contents of 87Sr and exchange
with basaltic rocks with low
contents of 87Sr on the oceanic 0.708
spreading ridges
High rates of
seafloor
spreading
0.707
Ceno Pre-
zoic Cret. Jur. Trias. Perm.. Carbon. Dev. Sil. Ord. Cambr. cambr.
0 100 200 300 400 500 600 m.yr.
Mica (muscovite and biotite) lose some potassium using X-rays which are diffracted according to
which is replaced by water (H2O, H3O+) to form illite Bragg’s Law: nλ ¼ 2d sin φ. Here λ is the wavelength
(hydro mica). Clay minerals are also formed through of the X-ray, φ the angle of incidence and d the
weathering reactions, for example by the breakdown thickness of the reflecting silicate layers; d is thus a
of feldspar and mica. Clay minerals which are formed function of angle φ.
by the breakdown of other minerals within the sedi- Sheet silicates may also be identified by means of
ment, are called authigenic. differential thermal analysis (DTA), which records
Sheet silicates have a structure consisting of sheets characteristic exothermal or endothermal reactions as
of alternating layers of SiO4 tetrahedra and a function of temperature.
octahedra. In the tetrahedral layers, silicon or alumin- Figure 3.9 shows the structure of some of the main
ium atoms are surrounded by four oxygen atoms. In clay minerals. Illite consists of sheets with two layers
the octahedral layers the cation is surrounded by six of tetrahedra and one of octohedra, bonded together by
oxygen or hydroxyl ions. Both bi and trivalent ions potassium. This ionic bonding is relatively weak so the
can act as cations in the octahedral layer. In sheet mineral cleaves easily along this plane. The bonds
silicates with trivalent ions (e.g. Al3+) only two of the within the tetrahedral and octahedral layers are more
three positions in the octahedral layer are occupied, covalent and stronger. The potassium content in mica
and such minerals are therefore called dioctahedral. corresponding to the formula of mica is about 9%
With bivalent ions (Mg++, Fe++) all three positions K2O, while illite has a greater or lesser deficit of
must be filled to achieve a balance between the posi- potassium. Smectite (montmorillonite) has the same
tive and negative charges, so these minerals are structure except that most of the potassium is replaced
called trioctohedral. by water ðH3 Oþ Þ, other cations or organic compounds.
The main method of identifying clay minerals is X- There are strong indications that smectite consists of
ray diffraction (XRD), by which the thickness of the small particles of 10 Å, plus water.
layers in the sheet silicate structure is determined
102 K. Bjørlykke
Tetrahedron Octahedron
Fig. 3.9 Simplified
illustration of the main groups
of clay minerals. Their
physical properties can be O– – ,OH– O– – ,OH–
explained by their crystal 4+ 3+
Si , Al Al3+,Fe3+,Fe++,Mg++
structure. The chemical bonds
between SiO44 and O2 in the
{
tetrahedral structure are very Tetrahedral
strong. In the octahedral layers layer (SI,Al) (O, OH)4 K Al2 AlSi3O10 (OH)2
the bonds are weaker because
10 Å Octahedral Al3+/Mg++, Fe++ Muscovite
Mgþþ is surrounded (co- (Dioctahedral)
layer
ordinated) with 6 oxygen. In K (Mg, Fe) AlSi3 O10 (OH)2
+
the illite (mica) structure K
Biotite
potassium is co-ordinated with (Trioctahedral)
12 oxygen molecules,
resulting in weak bonds that
produce a strong cleavage
Illite (K1 – x) (Al, Mg, Fe)2–3 AlSi3 O10(OH)2
{
Tetrahedral layer (SI,Al) (O, OH)4
Octahedral layer (Al, Mg)

14–17 Å
n. H3O+ (Exchangeable cations x)
(Glycol)
Smectite × (Al2–x, Mgx) Si4O10(OH)2

(Montmorillonite)
7Å
{ Tetrahedral layer (Si3, Al) (O, OH)4
Octahedral layer (Al1)
Kaolinite – Al2SiO2O5 (OH)4
14 Å
{ Tetrahedral layer (Si3, Al)
Octahedral layer (Mg, Fe, Al)
Octahedral layer (Me, Fe, Al)

(Biotite layer)
Chlorite – (Mg, Fe, Al)6 (Si,Al)4 O10 (OH)8
Illite is most likely comprised of several layers of because there are numerous layers bonded together
these small 10 Å particles stacked on top of one with K+ or other cations, for example NH+. Smectite
another. has a very high ion-exchange capacity and to some
In an atmosphere of glycol vapour, smectite will extent can exchange ions in the octahedral layer. The
swell from 14 to 17 Å, while illite is unable to expand stability of smectite declines in aqueous solutions with
high Kþ =Hþ ðNaþ =Hþ Þ activity ratios and with their very large specific surface area they have a great
increasing temperature and K+/H+ activity ratio. We capacity for adsorbing ions, which is increased by the
see that when the porewater is basic (equilibrium with fact that clay minerals have negatively charged edges
calcite), the potassium concentration may be lower due to broken bonds. In water with a low electrolyte
before smectite converts to illite. The stability of content, clay minerals will therefore repel each other.
smectite is also a function of the activity of silica. If cations are added, clay minerals will then accumu-
When quartz precipitation starts at about 80 C, the late a layer of positive ions (a double layer), and
silica concentration in the porewater is reduced to a repulsion between negatively charged clay minerals
low level corresponding to the solubility of quartz. declines as the strength of the electrolyte increases.
Vermiculite has a structure reminiscent of the Van der Waal’s forces will therefore cause floccula-
smectites, and also undergoes ion exchange and thus tion more easily in saltwater, where repulsion due to
charge deficit in the tetrahedral layer, so that the bond- the negative charge is reduced. This is why clays
ing between each layer is too strong for much swelling transported by rivers flocculate into larger particles
to occur. Vermiculites are mostly trioctahedral, when they enter seawater and sink more rapidly to
containing mostly Mg or Fe in the octahedral layer. the seafloor.
Glauconite is a green mineral which forms on the A colloid solution of clay in water is called a sol,
seabed. It is a potassium- and iron-bearing silicate which can be regarded as a Newtonian fluid.
somewhat similar to illite and contains both di- and Flocculated clay is called a gel and has thixotropic
trivalent iron. It is therefore formed right on the redox properties. This means that the shear strength
boundary, and during periods with little or no clastic decreases with increasing deformation so that it
sedimentation this can result in relatively pure beds of changes from a gel to a sol which consists of dispersed
glauconite. colloidal particles in water (hydrosol). After a time
Kaolinite consists just of a tetrahedral layer and an without deformation, it regains its strength. This is
octahedral layer and is very stable at low temperatures. typical for smectitic clays.
There are no positions in the structure where exchange Sediments which increase their volume when they
can procede easily, which gives kaolinite a much are deformed, are called dilatant. When the original
lower ion exchange capacity than smectite. At higher packing of the grains is destroyed, the new packing
temperatures kaolinite becomes unstable and will con- may be less effective and the volume then increases.
vert to illite if K-feldspar or other sources of potassium Walking on a beach the deformation of the sand causes
are available (at 130 C) or pyrophyllite increased porosity so that water is sucked into the sand
(Al2Si4O10(OH)2) at higher temperatures. at the surface.
Kaolinite is part of the kaolin mineral group, which Norwegian clays deposited in the sea when the ice
includes dickite which tends to form at slightly higher sheet retreated about 10,000 years ago, consist of
temperatures (100 C). crushed rock fragments and have approximately the
Chlorite is a mineral which consists of two tetrahe- same composition as the parent rock. During deposi-
dral layers and two octahedral layers, totalling 14 Å. tion these clays acquired a markedly porous structure,
The octahedral layer is filled with Mg++and Fe++. with the minerals stacked like a card house. Saltwater
Magnesium-rich chlorites are typical of high tempera- helped to hold this structure together because it
ture metamorphic rocks, while iron-rich ones may neutralised the negative charges. When the clays are
form authigenically in sediments near the seafloor or elevated above the sea by isostatic recovery of the
at shallow depth. land, they are exposed to meteoric water. Even if the
Chamosite is an iron-rich chlorite mineral which clay has a low permeability, freshwater will slowly
often forms close to the sediment surface in reducing seep through it and remove the saline water. This
conditions. Together with siderite, chamosite is an reduces the strength of the clay’s structure and hence
important mineral in sedimentary ironstones which, its stability and the clay becomes “quick”. This is
especially in England, earlier were used as iron ore. caused by overpressure and weak effective stress
Clay minerals have a number of properties which between the grains and hence little friction. The card
distinguish them from most other minerals. Because of house structure may then collapse, releasing the
104 K. Bjørlykke
interstitial water to produce a low viscosity clay slurry because as the rocks near the surface are unloaded by
that will flow even down very gentle gradients. The the reduction in overlying rock, they can expand ver-
addition of salt (NaCl or KCl) binds the clay particles tically but not horizontally. In this way the vertical
so that the shear strength is increased, and this is stresses become less than the horizontal ones, and
employed when the ground has to be stabilised for joints develop normal to the lowest stresses. We see
buildings and construction. this most clearly in granites, which are homogeneous,
while expansion in metamorphic and sedimentary
rocks occurs along bedding surfaces, along tectoni-
3.6 Weathering cally weak zones with crushing, or along fractures
that were formed at great depth. Joints opened by
The composition of clastic rocks in sedimentary stress release in turn provide pathways for groundwa-
sequences depends to a large extent on the supply of ter to circulate, increasing the surface area of rock
sediments from source areas undergoing weathering exposed to chemical weathering.
and erosion. The physical properties of sedimentary In areas that experience freeze-thaw cycles, frost
rocks are controlled by the primary sediment compo- weathering becomes very important. When water
sition and changes during burial (diagenesis). We shall freezes in cracks in rock, it expands by 9% and can
here briefly look at the processes producing sediments. generate very high stresses, further widening cracks
Mechanical weathering is the physical breakdown near the surface. The surfaces of exposed rocks are
of rocks into smaller pieces which can then be also subjected to daily temperature fluctuations which
transported as clastic sediment. cause greater expansion of the outer layers relative to
Chemical weathering involves the dissolution of the rest of the rock. Desert regions in particular expe-
minerals and rocks and precipitation of new minerals rience very wide daily temperature ranges, though the
which are more stable at low temperatures and high importance of this process for mechanical weathering
water contents. Parts of the parent rock will then be has been questioned. The roots of plants and moss can
carried away in aqueous solution by the groundwater also contribute to mechanical weathering as they grow
and rivers into the ocean. into fractures, take up water and expand.
Erosion is the combined result of the disintegration
of rock and the removal of the products.
As we shall see, biological processes are not only 3.6.2 Biological Weathering
important in connection with chemical weathering but
also with respect to mechanical weathering. Both Rocks are a source of nutrients for plants, and plants
mechanical and chemical weathering, which are are capable of dissolving and breaking down the major
essentially land surface processes, are due to the rock-forming minerals. Moss and also lichen (which
thermodynamic instability of rocks that were formed consists of algae and fungi living in symbiosis), pro-
or modified under other conditions (at greater depths, duce organic acids that can slowly dissolve silicate
higher pressures or temperatures, or in different chem- minerals. Even in the earliest stages of weathering,
ical environments). They are no longer stable when we see that fungus hyphae penetrate into microscopic
exposed to the atmosphere, water and biological cracks.
activity. Plant roots produce CO2 which helps to lower the
pH and dissolve minerals such as feldspar and mica,
thus freeing an important plant nutrient, potassium.
3.6.1 Mechanical Weathering Plants also produce humic acids, which likewise
strongly influence the solubility of silicate minerals,
Igneous and metamorphic rocks which have been and also affect the stability of clay minerals. The
formed at many km depth at higher temperatures and production of humic acids is perhaps the major factor
pressures are not stable when exposed at the surface. influencing the rate of weathering. In heavily
When uplifted and unloaded the rocks expand, vegetated areas, such as rain forest in the tropics, the
mostly in the vertical direction, producing horizontal weathering rate is exceptionally high because so much
fractures (sheeting) parallel to the land surface. This is humic acid is produced.
Bacteria and fungi, which are found in almost all dominates, is called the A-horizon. Some of the
soil types, are active in breaking down minerals. dissolved salts and particularly iron hydroxide is
Animals also contribute to weathering, and certain precipitated in the B-horizon below (Fig. 3.10).
marine organisms such as mussels are able to bore These may develop into a layer of solid rock (hard-
into solid rock (see Chap. 8). Microbiology has pan) cemented with iron and aluminium oxides and
become a key area of research in the quest to under- hydroxides.
stand how minerals are dissolved and precipitated. Where precipitation is approximately equal to
evaporation, there is less leaching within the soil pro-
file. At a certain depth (about 0.5–1 m) carbonate will
3.6.3 Chemical Weathering be precipitated and form an indurated layer (calcrete)
which may be eroded and form conglomerates.
There is no sharp demarcation between biological and The organic content is greater in the B-horizon
chemical weathering, because we find biological which is brown due to less oxidation of organic matter,
activity in almost all soils and rocks near the surface. hence the term brown-earth profiles. If evaporation is
The chemical environment in water at the surface of greater than precipitation there will be a net upward
the Earth is very much affected by local biological transport of porewater, causing dissolved salts from
activity, and in most cases it is biological processes the groundwater to be precipitated high in the soil
that cause weathering to continue after rainwater has profile.
been neutralised through reaction with minerals. We
will therefore use the term “weathering” here for both
chemical and biological processes.
3.6.5 What Factors Control Weathering
Rate and Products?
3.6.4 Weathering Profiles (Soil Profiles)
Because weathering is the most important sediment-
Both chemical and biological weathering are to a large producing process, we are interested in understanding
extent controlled by climate. The crucial factor is the how the rate of weathering is related to rock type,
ratio between precipitation and evaporation in an area. precipitation, temperature, vegetation, relief etc. We
In areas where precipitation far exceeds evaporation, also try to establish correlations between weathering
podsol profiles develop in which there is a net trans- products, particularly clay minerals, and these factors.
port of ions down through the soil profile as minerals By studying sediments from older geological periods,
are dissolved. In other words, we get weathering due to we can learn something about weathering conditions
the fact that rainwater is slightly acidic (on account of at those times. Weathering products will also bear the
its CO2 and H2SO4 content) and contains oxygen. stamp of the rocks undergoing weathering. The stabil-
Rainwater is initially undersaturated with respect to ity of a mineral during weathering is largely a function
all minerals. Some minerals are only very slightly of the strength of the bonds holding the cations in the
soluble, others more soluble in this slightly acidic, crystal lattice. Potassium (K+) in mica is held by weak
oxidising water. Dissolved ions are transported down bonds because the univalent charge is co-ordinated
to the water table, but ferrous iron liberated from iron- with 12 Oxygen or OH molecules in the crystal lattice
bearing minerals will be oxidised and precipitated as of mica responsible for the pronounced cleavage. In
ferric iron (Fe(OH)3). Vegetation at the top of the soil biotite, the Mg++ and Fe++ in the octahedral layer will
profile produces CO2 from roots and organic also be weakly bonded. During weathering, cations
compounds, particularly humic acid, which will like K+, Naþ ; Caþþ ; Mgþþ and Fe++can be attacked
increase the solubility of silicate minerals. Similarly, by protons (H+) which will replace them and send
aluminium derived from the solution of feldspar and them into solution. Chain silicates like hornblendes
mica, for example, precipitates as Al(OH)3 but is less and pyroxenes will also be relatively unstable and
noticeable because aluminium hydroxide is white. The rapidly weather. In feldspars the alkali ions are
uppermost part of the soil profile, where dissolution dissolved so that the whole mineral disintegrates. Sta-
due to undersaturated rainwater and organic acids bility is lowest in calcium-rich plagioclase, while pure
106 K. Bjørlykke
Rainfall
Evaporation Evaporation
and runoff
Rainfall
Little leaching.
Podsol profile Layer A Precipitation
Precipitation of
Leaching (grey) carbonates of salts and
Precipitation (caliche) carbonates
of Fe(OH)3 (red)
Layer B
Capillary water
Water table
Phreatic (ground) water Water table

Layer C Water table
phreatic
Phreatic
(ground) water (ground) water
Layer C
Layer C
Podsol soil
Brown-earth Evaporitic soil
profile.
profile
Downwards
transport.
Fig. 3.10 Simplified representation of soil profiles as a function of rainfall (precipitation) and evaporation
albite (sodium feldspar), orthoclase and microcline the solution, K+, Na+ and silica (H4SiO4), are not
(potassium feldspar) are more stable. The breakdown removed by water flow, the reactions will cease. This
of these silicate minerals will primarily liberate alkali is why weathering is always found to begin along
cations. Silicon and aluminium have very low solubil- cracks where water can penetrate (Fig. 3.11). The
ity and form new silicate minerals, largely clay vertical and horizontal joints that often develop in
minerals, though some silicic acid (H4SiO4) goes response to pressure release when previously deeply
into solution. buried rock is exposed at the land surface provide the
1. 2KðNaÞAl2 AlSi3 O10 ðOHÞ2 ðmuscoviteÞþ initial pathways. As the weathering process spreads
2Hþ þ 3H2 O ¼ 3Al2 Si2 O5 ðOHÞ4 ðkaoliniteÞþ outward from joints, blocks of unweathered rock are
2Kþ ðNaþ Þ gradually isolated. They have rounded corners and
2. 2KðMg; FeÞ3 AlSi3 O10 ðOHÞ2 ðbiotiteÞ þ 12Hþ þ may become entirely round (spheroidal weathering)
2e þ O2 ¼ Al2 Si2 O5 ðOHÞ4 ðkaoliniteÞþ (Fig. 3.12a,b,c). In desert areas, where there is little
4SiO2 ðin solutionÞþ rainfall, weathering proceeds much more slowly. Illite
xFe2 O3 þ y4 Mgþþ þ 6H2 O þ 2Kþ and montmorillonite may be formed when the
3. 2KðNaÞ AlSi3 O8 ðfeldsparÞ þ 2Hþ þ 9H2 O ¼ porewater has higher K+/H+ and Na+/H+ ratios than
for kaolinite, and they are frequently formed where
Al2 Si2 O5 ðOHÞ4 ðkaoliniteÞ þ 4H4 SiO4 þ
there is less water percolation and the removal of
2Kþ ðNaþ Þ
potassium or sodium is slower.
We see that potassium has been replaced by
There is often also a high silica content in the water
hydrogen ions in the new silicate minerals. The same
in desert areas due to frequent silica algae (diatom)
applies to sodium in albite. The equations show that
blooms and because silica is concentrated by water
the reactions are driven to the right by low K+/H+ and
evaporation. This helps enhance the stability of
Na+/H+ ratios.
smectite.
The degree of weathering depends on how under-
Granites subjected to weathering over a very long
saturated the water is with respect to the minerals
period often develop a special topography. Fractures
comprising the rock, and on the volume of water
and fault zones weather fastest and form valleys
flowing through the rock. If the reaction products in
where the groundwater collects, which further
Rainwater
Granite
Exfoliation cracks
Weathering = Rock + H2O + H+ Weathering product + Ions in solution
Fig. 3.11 Weathering of granites along extensional fractures
accelerates the weathering. The more massive granitic derived from volcanic rocks that contain glass or solu-
areas will stand out in the terrain, and because precipi- ble silicate minerals. Biogenic sources of silica
tation runs swiftly down into the depressions between (diatoms, radiolaria) will also increase the silica con-
the elevated portions, the topographic difference will centration in porewater because amorphous silica
become more and more pronounced. Granites is much more soluble than quartz. In desert
surrounded by sedimentary rocks will, because of environments evaporation of water after rainfalls will
their high content of feldspar, normally weather faster concentrate silica in the porewater and make smectite
than the sediments which contain more quartz and other stable. Figure 3.13 shows analyses of rocks at various
stable minerals. Particularly if the sediments consist of stages of transformation due to weathering.
quartzites and shales, the granite will form a depression Weathering proceeds particularly rapidly in
in the terrain. The weathering products from granites amphibolites: Na+, Mg++ and Ca++ are quickly leached
will normally be quartz grains which form sand grains out, while A13+, Fe3+ and Ti4+ become enriched. K+ is
the same size as the quartz crystals in the granite, and however to a large extent adsorbed on clay minerals
clay consisting of kaolinite, and possibly also some and much more Na+ than K+ is therefore supplied to
illite and smectite formed from feldspars and micas. the oceans by the rivers.
We have at the outset a bimodal grain-size distribution After the alkali cations have been dissolved out of
with sand and clay, but very little silt. the silicates and kaolinite has been formed, extremely
Basic rocks (e.g. gabbro) will weather far more slow leaching of quartz commences. When the con-
rapidly than granite because basic plagioclase (Ca- centration of silica in the porewater is sufficiently low
feldspar), pyroxenes and hornblende are very unstable (see Fig. 3.14), kaolinite will be unstable and be
and dissolve faster than silica-rich (acid) minerals. replaced by gibbsite Al(OH)3. Since gibbsite cannot
During progressive weathering sodium, potassium, form as long as the porewater is in equilibrium with
magnesisum and calcium will be removed by the quartz, all the quartz must have dissolved first or
groundwater but some potassium may be adsorbed become encapsulated (e.g. in a layer of iron oxides).
on clay minerals. The weathering residue will be Clearly, gibbsite will form far more rapidly during the
enriched in elements with low solubility such as Ti, weathering of basic rocks than of granites, since the
Al, Si and Mn (Fig. 3.13). In a normal, oxidising initial silica content is considerably lower. It takes a
weathering environment, all the iron will precipitate very long time to dissolve all the quartz in a granite.
out again as iron oxide (Fe(OH)3) while the magne- The solubility of quartz at surface temperatures and
sium will tend to remain in solution. In areas with high pH 7–8 is about 5 ppm, increasing at higher pH values.
rainfall the concentration of ions like K+, Na+ and Alkaline (basic) water can therefore increase the solu-
silica will be diluted and kaolinite will precipitate. tion rate of quartz. The end product of the weathering
Where porewater circulation is slower we may process is laterite, which consists of gibbsite and iron
get a higher build-up of Mg++, Ca++ and silica oxides or hydroxides. Under atmospheric (oxidising)
concentrations in the water, so that smectite (montmo- conditions with a neutral pH, aluminum hydroxide and
rillonite) or chlorite precipitates. Smectite requires iron oxides may for practical purposes be regarded as
porewater with a relatively high silica concentration insoluble.
(Fig. 3.13) and is therefore often found in sediments
108 K. Bjørlykke
a b
Fig. 3.12 (a, b) Weathered granite (North of Kampala, Uganda) showing different stages of weathering. (c) Concentric –
spheroidal weathering in a basic intrusive rock. Weathering is much faster in the absence of quartz
At low pH values, e.g. under the influence of humic be dissolved at low (acidic) or high (basic) pH values
acids in humid tropical climates, aluminum is more and may reprecipitate as aluminum oxide, which has a
soluble than iron and selective leaching of A13+ will lower iron content and is thus a higher grade alumin-
produce iron-rich laterites. Aluminum hydroxide can ium source (bauxite).
Fig. 3.13 Chemical analyses %

200 TiO2
of changes in the chemical
composition of samples Fe2O3
representing progressive 180
weathering of an amphibolite Al2O3
compared to an unweathered 160
sample. The stable (less
soluble) elements (Ti, Fe and 140
Al) are enriched while there is
a strong depletion of Na, Mg 120
and Ca. Potassium is depleted
but is adsorbed on clay 100
minerals in the soil 90
SiO2
80
70 K2O
60
50
40
30
20 Na2O
10 MgO
CaO
J1 J2 J3 J4 J5 J6 J7
Sample numbers
Fig. 3.14 Activity diagram 8.0

showing the stability of some
minerals as a function of silica Quartz
and the activity ratio (after saturation
Aagaard and Helgeson 1982).
Weathering reactions are 6.0
characterised by reduction in
both these two parameters Illite
towards the lower left of the
diagram Log Microcline
4.0
aK+/aH+
Amorphous
silica saturation
Mixed
2.0 Gibbsite Kaolinite layer
clay
Smectite
0.0
–4.0 –3.0 –2.0 –1.0 0.0
Log aSiO2 (aq)
110 K. Bjørlykke
3.7 Distribution of Clay Minerals and short transport distance between the site of erosion and
other Authigenic Minerals as a deposition and the sediments have little time to weather
Function of Erosion and on the way. Clastic chlorite and biotite, which break
Weathering down relatively rapidly during weathering, are there-
fore likely to be preserved in addition to feldspar in rift
3.7.1 What Determines the Type of Clay basins. These unstable minerals are good indicators of
Minerals We Find in Sediments and rapid erosion and/or cold climates, since otherwise
Sedimentary Rocks? they are unlikely to survive.
An analysis of the distribution of clay minerals in
When rocks are subjected to erosion and weathering, modern sediments shows that we find clastic chlorite
clastic minerals are broken down and perhaps some- almost exclusively in high latitudes, except around
what altered, with respect to the minerals in the parent islands of basic volcanic rocks (e.g. basalts). Clastic
rocks. We can also get new minerals formed in the chlorites from metamorphic rocks or altered basic
source rock itself, and precipitation of new minerals rocks are more magnesium-rich than authigenic
through weathering. chlorites, which are iron-rich (chamositic). In temper-
The more rapidly erosion and transport take place ate areas with moderate to high precipitation,
compared to the rate of weathering, the closer the weathering proceeds relatively rapidly. In desert
composition of sediments is to the source rock. Glacial areas weathering proceeds very slowly because all
sediments represent one end of the scale in terms of weathering reactions require water. The type of
sediment composition. Because glaciations are weathering and the distribution of clay minerals are
characterised by very high rates of erosion and low clearly related to latitude (Fig. 3.15).
temperatures, chemical weathering will be very weak, Smectite and illite are typical of deserts because the
and Quaternary sediments – including clays – will weathering is much slower when water is nearly
have a composition which essentially represents the absent. When feldspar and other unstable minerals
average of the rocks which have been eroded. Quartz- are altered (weathered) gradually in a dry climate,
rich rock will produce very different clays than basic alkali ions and alkaline earth ions such as K+, Na+,
rocks and older sedimentary rocks. Sediments depos- Ca2+ and Mg2+ will not be removed rapidly enough
ited in fault-controlled basins (e.g. rift basins) have a due to little percolation of fresh rainwater.
mm
2700
Precipitation
2400
2100
1800 °C
1500 Evaporation Temperature 25
1200 20
900 15
600 10
300 5
Chlorite
Illite and smectite

Laterite
Kaolinite
Smectite (montmorillonite)
Fig. 3.15 Simplified diagram showing the distribution of characterised by a high degree of weathering, while dry areas to
weathering and common clay minerals as a function of latitude the north and south have very low rates of weathering. Temper-
and rainfall. Cold areas and deserts are characterised by little ate regions have intermediate rates
weathering and more mechanical erosion. Equatorial regions are
Rainwater will cause some dissolution of minerals acids (increased pH) will lead to precipitation of
and, during the periods of drying, the silica concentra- oxides and hydroxides which make the soil hard and
tion may be higher so that smectite becomes stable. uncultivable.
As a result we will usually get illite or smectite Volcanic ash consisting of glass and unstable vol-
formed as authigenic minerals because they are stable canic mineral assemblages may alter to smectite on
in the presence of high K+/H+ ratios in the porewater. land or on the seafloor. In deep sea sediments zeolites
Smectite (montmorillonite) is thus a common clay like phillipsite are common.
mineral in desert areas, and its ability to swell when Areas with volcanic rocks, particularly amorphous
wet renders sediments very plastic during floods. This material (volcanic glass), will often form zeolites.
expansion of smectite also lowers its permeability and These require a high concentration of both silica and
may be the reason why water can flow over the surface alkaline ions in the water, which is the situation when
for a long time before sinking into the ground. In glass dissolves. Zeolites, particularly phillipsite, are
addition, capillary forces will prevent rapid percola- formed authigenically on the Pacific Ocean bed and
tion of water through dry soil. On the ocean floor near are also found in lakes (e.g. in East Africa).
desert regions we find that illite and smectite are In summary we can say that the factors which
typical minerals, brought there by aeolian transport. determine which types of clay minerals are “pro-
In tropical areas where precipitation is relatively duced” in the various areas are:
high, the rate of weathering will be very rapid. This is 1. The rocks which are eroded/weathered (source
not only because weathering processes accelerate with rocks)
temperature, but also because vegetation produces 2. Rate of erosion
large amounts of organic acids (humic acids) which 3. Temperature
are very effective in breaking down silicate minerals. 4. Precipitation
Microbiological organisms such as fungi and bacteria 5. Vegetation
also help in the breakdown process by producing CO2. 6. Permeability of source rocks and sediments (perco-
Gibbsite (A12O3•3H2O) and iron oxides (haema- lation of water).
tite, goethite, Fe2O3•3H2O) are constituents of the Typical distribution of various minerals:
laterite which we find only in tropical areas with 1. Chlorite and biotite – high latitudes (cold climate) –
rapid weathering and slow erosion. Whereas iron rapid erosion.
oxides are also found at higher latitudes, gibbsite 2. Kaolinite – humid temperate and humid tropical
occurs almost exclusively in humid tropical areas. regions – good drainage.
Laterisation is a very slow process and takes 3. Smectite (montmorillonite) – low precipitation or
millions of years, even in tropical regions with rapid poor drainage. Typical of desert environments, but
weathering. It is therefore primarily in tropical areas also formed in impervious, e.g. basaltic, rocks in
that we find bauxite for the aluminum industry. Iron- more humid environments. Typically formed from
rich laterites may have an iron content of over 50%, volcanic rocks.
and in some areas (e.g. India) have been exploited as 4. Gibbsite – tropical humid climate – long
iron ore. Laterite forms a very hard cement-like crust weathering period.
over the weathering profile and is also virtually devoid 5. Zeolites – formed in areas with volcanic material
of nutrients, so crop cultivation is impossible. In East and restricted porewater circulation. Require a high
Africa (especially Uganda), however, erosion has concentration of silica and alkali ions.
incised through a layer of Tertiary laterites. While
the laterite cover remains on flat elevated surfaces,
fresher, more fertile, rocks and weathering material 3.8 Geochemical Processes in the
are exposed in the valley sides. The vegetation in Ocean
some tropical areas is more abundant, even if the
soils are very poor in nutrients, because the vegetation The ocean can be regarded as a reservoir of chemicals
recycles those nutrients which are available. If the dissolved in water. It looks as though the composition
vegetation is removed and organic material is no lon- of seawater has not altered radically throughout the
ger produced, oxidation and the absence of humic geological ages from the early Palaeozoic until the
112 K. Bjørlykke
Addition of Geochemical processes in the ocean

dissolved ions
Weathering
A
C Uptake of organisms in the ocean H

B
Dissolution of F Precipitation of poorly
Adsorption on organisms in soluble salts during
D Circulation of sea
clay minerals the water column evaporation, e. g. chlorides
water through
and on the sulphates, and carbonates
basalt in spreading
seafloor
ridge
E
Deposition of organic Growth of authigenic minerals on
matter in sediments the seafloor, particular by zeolites
G
E=C–D A=B+E+F+G+H
Fig. 3.16 The chemical composition of the seawater remains equal to the removal of dissolved components by precipitation
nearly constant over geologic time. The supply of ions in solu- of minerals and adsorption on clay minerals. The most soluble
tion from rivers and from spreading ridges must therefore be components (Na, Cl, KCl) are only removed by evaporation
present day, although there have certainly been some Organisms can build their own internal chemical envi-
variations. ronment, and use their energy to precipitate minerals
The supply of elements to ocean water from rivers which are not normally stable in seawater. Carbonate-
and by water circulation at the spreading ridges must secreting organisms, e.g. foraminifera, molluscs etc.,
be balanced by a removal of the same elements from will precipitate aragonite or calcite even when the
the ocean water (Fig. 3.16). The annual addition of water is cold and undersaturated with respect to these
salts dissolved in river water is about 2 109 tonnes/ minerals. Diatoms are so effective in precipitating
year. The figure was probably less in the geological silica (amorphous silicon dioxide, SiO2), that in most
past because vegetation was sparse or absent. The places the seawater near the surface in the photic zone
development of land plants that produce humic acids, (photosynthetic zone) is much more depleted with
which in turn produced more rapid weathering, has respect to silica. Organisms, when they die, will in
probably increased the supply of salts (since Devonian most cases start to break down by oxidation of organic
times). This trend was sometimes slowed by periods matter and by dissolution of the mineral skeletons. In
with higher sea level that caused widespread reducing environments much of the organic matter
transgressions and converted huge tracts of coastal will however be preserved. In shallow tropical waters,
land areas into continental shelf (e.g. during the like on a carbonate bank, the seawater may be
Upper Cretaceous), reducing the weathering and the saturated with respect to carbonate (calcite), but car-
supply of salts and nutrients to the ocean. bonate also accumulates in cold water like the Barents
Some ions like potassium (K+) are adsorbed to clay Sea because the rate of dissolution is slower than the
minerals supplied by rivers (B in Fig. 3.16). Sodium rate of precipitation.
(Na+) is so strongly hydrated that it has a tendency to The more efficient organisms are at building
remain in solution, while potassium will be far less skeletons despite undersaturation, the more rapidly
hydrated and can be more easily adsorbed onto clay they will dissolve. Diatoms, for example, dissolve to
minerals and rapidly removed from seawater. the extent of 99–99.9% before they have sunk to the
Most of the elements which the rivers bring to the seabed. Only a very small proportion is therefore pre-
ocean are precipitated by organic processes. served in sediments.
Photosynthesising organisms in the surface water because biological precipitation is more efficient in
use up CO2 and produce oxygen and organic matter: many cases and prevents the build-up of sufficient
concentrations of the elements needed for chemical
CO2 þ H2 O þ nutrients ¼ CH2 O þ O2 precipitation. Sulphate-reducing bacteria are active in
the uppermost few centimetres of the sediment, how-
This helps to keep the pH high so that carbonates ever, removing sulphur from seawater in the form of
are stable or dissolve slowly. At depth in the absence sulphate and reducing it to sulphides which are then
of sunlight, respiration and oxidation prevail, releasing precipitated (e.g. iron sulphides, FeS and FeS2).
CO2 and thus lowering the pH and increasing the In areas that are almost cut off from the open ocean,
solubility of carbonate. where evaporation exceeds the freshwater rainfall and
The depth at which the solubility of carbonate supply by rivers, we find that even very soluble salts
increases relatively rapidly is called the lysocline. are being precipitated. Under these conditions, there is
The depth where the rate of solution is greater than enrichment of elements which are otherwise
the rate of carbonate sedimentation is called the car- precipitated only to a limited degree through
bonate compensation depth (CCD). Near the equator biological or chemical processes, such as Na, Cl, S,
the CCD may be at 4–5 km, becoming shallower at Mg and trace elements such as B and Br. The amount
higher latitudes. Carbonate can therefore still accumu- of salt thus precipitated in evaporites has probably
late on the seafloor even when the water column is varied very markedly throughout the geological ages.
undersaturated, if the rate of supply exceeds the rate of Biological and chemical precipitation, as described
dissolution. This is analagous to snow accumulating above, are not sufficient alone to account for the geo-
when the temperature is above freezing so long as it chemical balance of the ocean. Important geochemical
falls quicker than it melts This solution of organic reactions are taking place in the spreading ridges in the
material liberates nutrients which can then be returned oceans (F in Fig. 3.16). Heat from the basalt that is
to the surface through upward flow. In this manner flowing up along the spreading ridge, drives convec-
they are repeatedly recycled. The annual biological tion cells which cause ocean water to flow through the
production in the ocean is therefore many times basalts and up along the ridge. The seawater reacts
greater than the supply of nutrients from the land. with hot basalt (basic rock melt) and disolves minerals
That fraction of the organic production which has containing iron and other metals. Since seawater
been removed from the ocean by preservation in sea- contains sulphur (as S04 ), this leads to precipitation
bed sediments must be replaced with nutrients, mostly of sulphides, for example iron sulphides and copper
from land. They cannot be returned to the ocean before sulphides. When hot water flows up pipe-like
the sediments are elevated and subjected to erosion chimneys (black smokers) in the ocean floor near
and weathering. The amount of organic matter which spreading ridges, it mixes with the seawater and
is deposited in sediments is a function of the rate of again sulphides are precipitated. When water is
production minus the rate of solution. oxidised, iron oxides and manganese oxides are
The growth of authigenic (newly formed) minerals precipitated around the spreading ridges.
on the seabed (G in Fig. 3.16) is also an important
process in removing elements from seawater. The
most significant are the zeolites, which can develop 3.8.1 Residence Periods for Different
where sediments on the seabed have a high silicate or Elements in the Ocean
aluminium content, particularly from volcanic mate-
rial (glass). They may remove Na+, K+ and Ca2+ from How long does an element spend in the sea after being
seawater, but growth can also proceed very much at delivered by rivers, before it is chemically or biologi-
the expense of elements already present in the sedi- cally precipitated? This residence period is an expres-
ment (e.g. in the Pacific Ocean). This applies particu- sion of how rapidly an element is removed compared
larly to phillipsite, heulandite, clinoptilolite and to its concentration in the ocean. For example, sodium
analcime. has the longest period of residence (about 55 million
Apart from this there is little direct chemical pre- years) because little sodium is removed through
cipitation from seawater with normal salinity. This is biological or chemical precipitation except in
114 K. Bjørlykke
evaporite basins. The residence time for potassium is ocean basins are isolated from this circulation, and we
about 10 million years because it is more rapidly may then have a more permanent layering of water
adsorbed onto clay minerals and thereby removed. based on temperature and salt content. Warm surface
Calcium is removed both chemically in evaporites water with low density can flow over heavier, colder
and biologically and has a residence time of 1 mill. basal water without the water masses mixing to any
yr. Rare earths have periods of only a few 100 years. major extent. The boundary between warm and cold
water masses is called a thermocline. If the density
difference is largely due to salt content, we call the
boundary a halocline. A pycnocline is the boundary
3.9 Circulation of Water in the Oceans between two water masses with different densities,
without a specified cause. Lakes in temperate and
Ocean currents are driven by: cold regions have good circulation. This is due to
1. The rotation of the Earth (Coriolis effect) water attaining maximum density at 4 C, below
2. Tidal forces which temperature density inversion causes turnover.
3. Differences in water density due to variations in The addition of freshwater to a basin (e.g. Baltic Sea,
salt content and temperature Black Sea) also leads to stratification of the water due to
4. Wind forces due to atmospheric circulation. salt concentration because brackish water flows on top
Ocean currents are extremely important for of marine water with higher salinity. Evaporite basins
redistributing heat from low latitudes to higher produce water which is heavy due to its high salt content
latitudes. This circulation is strongly dependent on and therefore forms a layer which flows along the bot-
the topography of the ocean floor and the distribution tom. If a salinity stratification becomes established, it
of ocean and continents. Bottom currents in the ocean will weaken or destroy the circulation and lead to reduc-
basins are very different from those at the surface, and ing conditions in the bottom layer. If the density contrast
often flow in opposite directions. While warm surface due to salt concentration is greater than that due to
water flows from the equator to the poles, cold surface temperature, this will impede or prevent the downward
water sinks at the poles and flows along the ocean floor flow of cold, oxygen-rich surface water (Fig. 3.17).
to the equator. Both currents are deflected by the There are indications that in previous geological
Coriolis effect, towards the right in the northern hemi- periods (e.g. the Cretaceous) the bottom water in the
sphere and towards the left in the southern hemi- ocean basins was warm, salty water (about 15 C)
sphere. While surface currents (like the Gulf Stream) compared with the present situation with cold basal
will be deflected eastwards, deeper currents will be water (2–3 ) and a normal salinity. A higher average
deflected towards the west of the oceans (e.g. the temperature in the oceans leads to reduced CO2 solu-
Atlantic Ocean). These deep-sea currents which fol- bility and a deeper carbonate compensation depth
low the depth contours may be strong enough to trans- (CCD). The volume of water welling up from deeper
port silt and fine sand, and the resultant deposits are water layers to the surface corresponds to the amount
called contourites. of downflow. If we have basal water with high salinity,
The vertical circulation of seawater is highly sensi- the reduced density contrast results in less downward
tive to variations in temperature and salt concentration. flow and consequently less upwelling, so less nutrients
In periods with glaciation at the poles, the temperature are added to the surface water.
gradient in the surface water flowing from the equator We have seen that a number of different processes
to the poles is far greater than in non-glacial periods control the geochemical equilibrium of the ocean.
(such as the Mesozoic). Oxygen-rich cold water now There must also be an equilibrium between the addi-
flowing down from the polar regions into the ocean tion and removal of chemical components for the
basins is important for maintaining oxidising ocean water composition to remain relatively constant.
conditions in the deep ocean basins. The conditions in which this equilibrium was
Animals and bacteria use oxygen from seawater maintained, however, have varied through geological
continuously (for respiration), and oxidation of dead time. During the first part of the Earth’s history, up to
organic material also requires oxygen. If we did not about 2.5 billion years ago, the atmosphere was reduc-
have this downward flow of cold surface water, the ing. Most geochemical processes acted very differ-
water in the ocean basins would be reducing. Some ently then from the way they do now. Weathering
Black Sea model Addition of

nutrients
Positive water balance Freshwater
............................................................
Low salinity Organic production
Threshold Pycnocline
Anoxic
Mediterranean and Red Sea model

Negative water balance
Evaporation > Supply of freshwater
Oxic Heavier
saltwater
Threshold
Fig. 3.17 Circulation of water in basins with a surplus of represents excess evaporation compared to freshwater input.
freshwater input compared to evaporation (Black Sea model). The surface water will then have the highest salinity and density
This results in poor vertical circulation and anoxic (reducing) and sink to the bottom. This increases the vertical circulation
bottom waters. The Mediterranean and Red Sea model and helps to maintain oxic conditions
was less efficient because of the low oxygen concen- with volcanism may produce large amounts of sedi-
tration in the atmosphere and limited biological ment compared to their area because they are tectoni-
weathering. The supply of ions to the ocean via rivers cally active, which leads to elevation and accelerated
was consequently less. On the other hand, seafloor erosion. Volcanic rocks are for the most part basic and
spreading was most probably faster with more seawa- weather quickly, forming large quantities of sediments
ter circulated through the spreading ridges. Isotope around volcanic island groups, while fine-grained vol-
studies (87Sr/86Sr) of seawater in early Precambrian canic ash becomes spread over wide areas. Submarine
rocks indicate that at that time the composition of volcanism may also produce some sediment, for
seawater was more strongly controlled by circulation example along the Mid-Atlantic Ridge, but this is
through the basalt on the spreading ridges. We can say very limited.
that the chemical composition of the ocean was buff- The main supply of clastic sediment is fed into the
ered by material from the spreading ridge, i.e. mantle ocean through deltas, then transported along the coast
material (Veizer 1982). and down the continental slope to the abyssal plains.
Around Antarctica there is a significant amount of
deposition of clastic, glacial sediments. In areas in
3.10 Clastic Sedimentation in the the middle of the Atlantic Ocean, far from land, the
Oceans rate of sedimentation is as low as 1–10 mm/1,000
years. The Atlantic receives a relatively large supply
Clastic sediments are produced chiefly on the of clastic sediment, in particular from seven major
continents and are brought to ocean areas through rivers: the St. Lawrence, Mississippi, Orinoco, Ama-
fluvial or aeolian transport. Island arcs associated zon, Congo, Niger and Rhine. Exceptionally high
116 K. Bjørlykke
sedimentation rates characterise the Gulf of Mexico, Table 3.1 Review of the ratio between mechanical and chemi-
where rapid deposition of thick sequences from the cal denudation of the different continents (After Garrels and
Machenzie 1971)
Mississippi delta has prevailed since Mesozoic times.
The South American and African continents drain Annual Annual Ratio
chemical mechanical mechanical/
mainly into the Atlantic. The water divide between the denudation denudation chemical
Atlantic and the Pacific Oceans lies far to the west in Continent (tonnes/km) (tonnes/km) denudation
South America, and that with the Indian Ocean in North 33 86 2.6
Africa is far to the east. America
The Pacific Ocean is surrounded by a belt of volca- South 28 56 2.0
nic regions and island arcs. There are relatively few America
Asia 32 310 9.7
rivers that carry large amounts of clastic sediment
Africa 24 17 0.7
directly into the Pacific Ocean, in contrast to the
Europe 42 27 0.65
Atlantic. Sediment which is eroded, for example on
Australia 2 27 10.0
the Asian continent, is deposited in shallow marine
areas (marginal seas) such as the Yellow Sea and
China Sea. The sediments are cut off from further
3.11 Summary
transport by the island arc running from Japan and
southwards. The Pacific Ocean is therefore dominated
Many of the most important geological processes are
by volcanic sediments.
controlled by chemical processes and this is also true
Volcanic sedimentation takes the form of volcanic
for those involved in sedimentary geology.
dust and glass, which may be transported aerially over
The primary sediment composition is in most cases
long distances. After sedimentation, volcanic glass
controlled by mineral dissolution during weathering
will turn into palagonite, an amorphous compound
and mineral precipitation which may be biological or
formed by hydration of basaltic tuff. Palagonite may
chemical. After deposition sediments are modified by
then be further converted into montmorillonite or zeo-
chemical processes. The chemical composition of
lite minerals. The zeolite phillipsite is very widely
sediments and sedimentary rocks provides very impor-
found in the Pacific, but is scarce in the other oceans.
tant information about the source and transport and
Pumice is also a volcanic product, and may drift float-
diagenetic changes during burial.
ing over great distances. The eruption of volcanoes in
This also provides a basis for exploration and pro-
the Pacific Ocean area in historic times has shown that
duction of natural resources from sedimentary rocks.
large eruptions produce 109–1010 tonnes of ash, and
Environmental studies of pollution related to the
much the same amount of pumice and agglomerates.
petroleum industry, mining and other industries
Submarine volcanism, by contrast, produces very
require a very strong background in sedimentary
little ash to form sediment. The lava which flows out
geochemistry.
onto the seabed will solidify as an insulating crust on
contact with the water (often forming pillow lava), so
that little volcanic matter goes into suspension.
Weathering and erosion processes are responsible
Further Reading
for the entire volume of sediment which can be depos-
ited in sedimentary basins. Material added by rivers Baskaran, M. (ed.) 2011. Treatise in Environmental Geochem-
takes the form of clastic and dissolved matter. The istry. Vol 2. Springer, 800 pp.
ratio between the quantities of these two forms of Chamley, H. 1989. Clay Sedimentology. Springer, New York,
623 pp.
sediment addition is a function of precipitation, tem-
Chester, R. 1990. Marine Geochemistry. Unwin Hyman,
perature and relief. Dry areas, like Australia, produce London, 698 pp.
mainly clastic material, while the African continent Davis, D.W., Williams, I.S. and Krogh, T.E. 2003. Historical
produces mainly dissolved material because of the development of zircon geochronology. In: Hanchar, J.M. and
Hoskin, P.W.O. (eds.) Zircon. Reviews in Mineralogy and
intensive weathering in some parts of the continent
Geochemistry 53. pp. 145-1|81.
(Table 3.1). Eslinger, E. and Pevear, D. 1988. Clay Minerals for Petroleum
Geologists and Engineers. SEPM Short Course 22, Tulsa, OK.
Faure, G. and Mensing, T.M. 2005. Isoptopes. Principles and between isotopic composition, textural development and
Applications. John Wiley, 898 pp. burial depth. In: Marshall, J.D. (ed.), Diagenesis of Sedimen-
Garrels, R.M. and Christ, C.L. 1965. Solutions, Minerals and tary Sequences. Geological Society Special Publication 36,
Equilibria. Harper and Row, New York, 450 pp. pp. 313–324.
Garrels, R.M. and Mackenzie, F.T. 1971. Evolution of Sedimen- Veizer, J. 1982. Mantle buffering and the early Oceans.
tary Rocks. W.W. Norton & Co, New York, NY, 397 pp. Naturvissenshaffen 69, 173–188.
Kenneth 1982. Marine Geology. Prentice Hall, Englewood Velde, B. 1995. Origin and Mineralogy of Clays. Springer,
Cliffs, 813 pp. Berlin, 334 pp.
Manahan, S.E. 1993. Fundamentals of Environmental Chemis- Weaver, C.E. 1989. Clays, Muds and Shales. Developments in
try. Lewis Publ., Chelsea, MI, 844 pp. Sedimentology 44, Elsevier, Amsterdam, 819 pp.
Saigal, G.C. and Bjørlykke, K. 1987. Carbonate cements in
clastic reservoir rocks from offshore Norway – Relationships
Chapter 4
Sandstones and Sandstone Reservoirs
Knut Bjørlykke and Jens Jahren
4.1 Introduction coarser grains (pebbles) we call them conglomeratic

sandstones. Most classification systems are based on
About 50% of all petroleum reservoirs are sandstones; the relationship between the relative quantity of sand-
outside the Middle East, carbonate reservoirs are less sized grains, the composition of the sand grains, and
common and the percentage is even higher. The most the clay and silt content (matrix).
important reservoir properties are porosity and perme- If we use a four-component diagram we can distin-
ability, but pore geometry and wetting properties of guish between clay, and sand grains which consist
the mineral surfaces may also influence petroleum of quartz, feldspar and rock fragments (or unstable
production. Sandstones provide reservoirs for oil and rock fragments, U.R.F.) (Fig. 4.1). Rock fragments
gas, but also for groundwater which is a fluid that consisting of microcrystalline (or cryptocrystalline)
is becoming increasingly valuable. Sandstones are quartz (including chert) are usually classified together
deposited in many different sedimentary environments with the quartz mineral grains. Sandstones with more
by marine, fluvial and eolian (wind) processes. than 25% feldspar and a low content of rock fragments
The outer geometry and distribution of sand bodies is are called arkoses. If the percentage of rock fragments
determined by the depositional environment and the resis high, we speak of lithic sandstones which are nor-
ervoir properties. The internal properties (porosity, per- mally derived from very fine-grained sedimentary
meability) are, however, critical for petroleum recovery. rocks or basalts and intrusive igneous rocks where
The properties of sandstone reservoirs are functions one sand grain often consists of several minerals.
of the primary composition, which is controlled by the Quartz arenite or orthoquartzite are the terms for
textural and mineralogical composition (provenance) pure quartz sandstones which contain less than 5%
of the depositional environment and by the diagenetic feldspar or rock fragments. Sandstones with moderate
processes near the surface and during burial. feldspar contents (5–25%) are called subarkoses.
Sand and sandstones are rocks which consist When granitic rocks and other coarse- to medium-
largely of sand grains, i.e. sedimentary particles grained rocks are broken down by weathering or ero-
between 1/16 and 2 mm in diameter. However, sion, they form sand grains consisting for the most part
sandstones also contain greater or lesser amounts of of single minerals, mostly quartz and feldspar.
other grain sizes and we find transitions to more silt- The prerequisite for forming arkose is that not too
and clay-rich rocks. Most sandstones have a well- much of the feldspar in the source rock has been
defined upper grain-size limit with variable contents weathered to clay minerals (e.g. kaolinite). Arkoses
of silt and clay. If they have a significant content of are therefore formed if granites and gneisses are
eroded rapidly in relation to weathering, and the
sediments are buried in a basin after a relatively
short sediment transport. In the great majority of
K. Bjørlykke (*) J. Jahren
Department of Geosciences, University of Oslo, Oslo, Norway cases arkose is therefore associated with sedimentary
e-mail: [email protected]; [email protected] rift basins formed by faulting in gneissic and granitic
120 K. Bjørlykke and J. Jahren
Mud
ke
Wac
ite
Aren
Quartzwacke
5 75
ke
ac
Quartz
w
Quartz arenite
sic
Subarkose 5 5 Sublith-
ko
25
Ar
arenite
25 Felds
path. µ
grayw 0
acke Lith <3
se
grayw ic rix
ko
50
acke at
Ar
t m
15 rc en
e P
Arko V
sic Volcanic
arenit
e Lithic Volcanic arenite
Felds arenit
par e 50 50
50 Ro Calclithite Phylloarenite
ck fra
gmen 0
ts
S 50 M
Sedimentary Metamorphic
Fig. 4.1 Classification of sandstones (from Dott 1964)
rocks, i.e. continental crust. In cold climates the rate of Basic rocks like gabbro and basalt and minerals like
weathering is reduced, thus preserving feldspar and amphibole and pyroxene are inherently unstable both
unstable rock fragments. mechanically and chemically. Primary sand grains of
In tectonically stable areas with mature relief there is volcanic or basic rocks may break down to become
far more time for weathering of both in situ bedrock and part of the matrix and it is then difficult to distinguish
sediment in transit. In such environments sand particles this material from the primary matrix. Greywacke is
(grains) are deposited and eroded many times before therefore typical of areas where the sand grains are
reaching their final deposition site. A greater proportion derived from volcanic or basic rocks along converging
of the feldspar grains will then break down during plate boundaries (fore-arc, inter-arc, back-arc basins).
transport and subarkoses or orthoquartzites will be Weathering of basic rocks will produce nearly exclu-
deposited. Orthoquartzites (or quartz arenites) are the sively clay since there are no quartz grains.
purest quartz sandstones, formed when weathering has We have seen that the various types of sandstone
eliminated virtually all the unstable minerals to leave a reflect different source rocks and areas with varying
concentration of quartz and some heavy minerals. This tectonic stability. Studies of different types of sand-
is particularly true of beach sand where nearly all the stone and their mineralogical maturity can therefore
clay particles have been removed. Sand deposited by be used as palaeo-indicators of relief and climate,
rivers (fluvial sand) is less well sorted and may contain and also of tectonic deformation in the geological
more clay. Sandstones with more than 15% matrix are past.
called greywackes.
Sand which is transported in suspension or through
mass flow (e.g. turbidites) can have poorer sorting, a 4.2 Prediction of Reservoir Quality
high matrix content, and grade into sandy mudstones
and form greywackes. Clay minerals in sandstones may The properties of all reservoir rocks are continuously
form after deposition during diagenesis from alteration changing, from the time the sediments are deposited
of feldspar, mica and rock fragments. This makes the through to their burial at great depth and during any
sediments less well sorted than at the time of deposition. subsequent uplift. This is a combined function of
4 Sandstones and Sandstone Reservoirs 121
mechanical compaction and of chemical processes (Vp and Vs) and the compressibility (bulk modulus,
involving dissolution and precipitation of minerals. see Chap. 11). This is also critical when predicting
At any given burial depth the properties depend on the physical rock changes during production (see 4D seis-
composition of the sandstones when at shallow depth, mic, Chap. 19).
and on their temperature and stress history during burial. The main diagenetic processes are:
Practical prediction of the porosity and permeability (1) Near-surface diagenesis. Reactions with fresh
during exploration and production is only possible if groundwater (subsurface weathering). In dry
the processes that change these parameters are environments, with saline water concentrated by
understood. evaporation. Sand may also be cemented with
It should be realised that the starting point for the carbonate cement near the seafloor.
diagenetic processes is the initial sandstone composi- (2) Mechanical compaction, which reduces the poros-
tion. This is a function of the rocks eroded (prove- ity by packing the grains closer together and by
nance), transport, and depositional environments. grain deformation and fracturing, increasing their
Diagenetic models must therefore be linked to mechanical stability. Mechanical compaction is a
weathering and climate, sediment transport, facies response to increased effective stresses during
models and sequence stratigraphy, and should be burial and follows the laws of soil mechanics.
integrated in an interdisciplinary basin analysis. (3) Chemical diagenesis (compaction), which includes
Diagenesis is often considered a rather specialised dissolution of minerals or amorphous material and
field of sedimentology and petroleum geology, but it precipitation of mineral cement so that the porosity
embraces all the processes that change the composi- and the rock volume are reduced. The clastic
tion of sediments after deposition and prior to meta- minerals in the primary mineral assemblage are
morphism. The most important factor in predicting not in equilibrium, and there is always a drive
reservoir quality at depth is the primary clastic com- towards thermodynamically more stable mineral
position and the depositional environment (Fig. 4.2). assemblages. Kinetics determine the reaction
The diagenetic changes also determine the physical rates, which for silicate reactions are extremely
properties of sandstones, such as seismic velocities slow so temperature plays an important role.
sandst Limits of
K kaolinite-
sandy mud
K illitic rich mud (K)
Carbonate ?
sandst
sandst sandy mud
sandy mud
Fig. 4.2 Schematic illustration of a sedimentary basin on a environment. Fluvial, deltaic and shallow marine sediments
continental margin. The primary composition of the sediments will be flushed by meteoric water after deposition, particularly
is a function of the provenance, transport and depositional in humid climates
(4) Precipitation of cement (e.g. quartz cement) will widespread in lakes and inland seas like the Black Sea.
increase the strength of the grain framework and However, even below well-oxygenated water, oxidising
prevent further mechanical compaction. The sand- conditions extend in most cases for only a few
stone is then overconsolidated – not due to previ- centimetres into the sediments, since oxygen is quickly
ously higher stress, but due to cementation. This is consumed by the oxidation (decay) of organic matter in
sometimes referred to as pseudo-overconsolidation. the sediment. This is for the most part facilitated bio-
Further compaction will then mostly be controlled logically by bacteria. Accumulating sediments nor-
by the rate of dissolution and precipitation. mally contain sufficient organic matter to serve as
In the following the diagenetic processes typical of reducing agents in the porewater. This organic matter
different burial depths will be discussed. is comprised of both the remains of bottom fauna and of
pelagic organisms, including algae, accumulating on
the seafloor, and also often includes terrestrial plant
4.3 Early Diagenesis debris transported into the basin.
As soon as sediments are deposited, early diagenetic

reactions start to modify the primary sediment compo- 4.4 Redox-Driven Processes on the
sition. At very shallow burial depth (<1–10 m), Seafloor
sediments have the maximum potential to react with
the atmosphere or water, both by fluid flow and diffu- Across the redox boundary there is a high gradient in
sion. Transport of dissolved solids by diffusion and the concentration of oxygen and sulphate, and of ions
fluid flow (advection) is most efficient near the surface; that have very different solubilities in oxygenated and
in the case of diffusion within about 1 m of the seabed. reduced water. The redox boundary is usually just
The potential for sediments to change their bulk com- 1–20 cm below the seafloor and represents equilibrium
position after deposition is therefore much higher at between the supply of oxygen by diffusion, and its
shallow depth than at greater burial. Near the surface consumption by the (mostly biological) oxidation of
on land, and also in the uppermost few centimetres of organic matter. The oxygen content in the porewater
the seabed, the conditions may be oxidising, while at thus decreases rapidly below the water/sediment inter-
greater depth in the basin they are always reducing. face, providing a concentration gradient for the down-
Precipitation of minerals due to porewater concen- ward diffusion of oxygen into the uppermost sediments.
tration by evaporation can only occur on land or at The rate of downward diffusion of oxygen is a
shallow depth within enclosed basins (see Chap. 6). function of the concentration gradient of oxygen in
On land, sediments are exposed to air and fresh the porewater and the diffusion coefficient in the
(meteoric) water. Weathering is partly due to reactions sediments. The diffusion coefficient in coarse-grained
with oxygen in the atmosphere and partly due to dis- sand is higher than in mud and therefore sand tends to
solution of minerals in freshwater, which is initially have a deeper redox boundary than mud.
undersaturated with respect to all the minerals present. Oxygen can also be consumed in the sediments by
These are soil-forming processes which can be consid- the oxidation of elements like iron and manganese in
ered to be examples of early diagenesis. minerals, but this is rare in marine environments and
In desert environments groundwater and occasional more common in continental deposits. In most marine
rainwater may become concentrated through evapora- environments there is enough organic matter to serve
tion, causing precipitation of carbonates and also as reducing agents and therefore little oxidation of iron
silicates. Coatings of red or yellow iron oxides and in minerals takes place, which explains why marine
clays frequently form on desert sand and this may sub- sediments do not normally acquire a red colour. A
sequently retard or prevent quartz cementation at greater notable exception is red oxidised mud which may
depth. form in marine environments characterised by slow
In the sea, the water above the seabed is normally sedimentation rates and low organic productivity.
oxidising. Only where there is poor water circulation These muds are not very common but occur in some
(poor ventilation) is the lower part of the water column deep-water facies and also in shallower water
likely to be reducing, though the phenomenom is more environments with low sedimentation rates.
Uranium is highly soluble in seawater as uranyl Bioturbation will also destroy thin clay laminae,
(UO2+) and precipitates as reduced uranium oxide which may significantly increase the vertical perme-
(UO2) on organic matter in the water column and ability and this may be very significant for reservoir
below the redox boundary. There is thus a strong con- quality. Undisturbed primary lamination may be evi-
centration gradient transporting uranium from the sea- dence of rather rapid sedimentation giving little time
water down into the sediments. The adsorption of for a burrowing bottom fauna to become established,
uranium onto organic matter settling on the seafloor or alternatively indicate strongly reducing conditions
coupled with restricted ventilation of the water above restricting the fauna. Black shales usually have well
the sediment at the time of deposition, makes source preserved lamination due to lack of burrowing
rocks like the Kimmeridge shale strongly enriched in organisms. The presence or absence of burrowing
uranium, causing peaks on the gamma ray well log curve. also influences the physical properties, particularly
Iron and manganese may be transported upwards the difference in velocity and resistivity parallel and
through the sediments in the reduced state by diffusion vertical to bedding (anisotropy) and this may be
and then precipitate on the seafloor as oxides because of important for geophysical modelling.
their reduced solubility in the oxidised state. Iron may Burrowing worms produce faecal material which
also be precipitated below the redox boundary as iron may develop into smectite-rich clays, which in turn
sulphides or iron carbonate (siderite), though iron may develop into chlorite coatings, thus improving
carbonates will not form during sulphate reduction. reservoir quality. Early diagenetic formation of
This is because all the iron will be precipitated as coatings on quartz grains is extremely important due
sulphide, which has a much lower solubility than sider- to its role in preserving porosity at greater depth.
ite. Manganese is not precipitated as sulphides because Most clastic depositional environments have some
of the high solubility of Mn-sulphides, but may be organisms producing organic matter which, at least in
precipitated as Mn-carbonate in the reduced zone. part, is incorporated within the sediments. Both
At a depth where there is practically no more free sandstones and mudstones nearly always contain sig-
dissolved oxygen in the porewater, sulphates can be nificant amounts of biogenic material from calcareous,
used by sulphate-reducing bacteria. The reduction of and sometimes also siliceous, organisms and this may
sulphates produces sulphides such as pyrite. later be an important source of carbonate and silica
The composition of clastic sediments is modified cement at greater burial depth.
by the addition of new components produced locally Marine organisms composed of aragonite dissolve
within the basin: during relatively shallow burial and calcite
(1) Biogenic carbonates and silica. precipitates either as replacements within the fossils
(2) Authigenic minerals precipitating near the seabed (neomorphism) or as cement in pore space between the
such as carbonates, phosphates, glauconite, grains.
chamosite, sulphides and iron and manganese Carbonate cement in sandstones may form layers or
minerals. concretions and in most cases is derived from biogenic
(3) Meteoric water-flushing causing leaching of feld- carbonate, particularly from organisms composed of
spar and mica and precipitation of kaolinite aragonite. Siliceous organisms composed of opal (e.g.
beneath the seafloor. diatoms or siliceous sponges) may be an important
source of microquartz coatings on quartz grains at
4.5 Importance of Biogenic Activity greater depth.
Carbonate cements in both mudstones and
Bioturbation plays an important role in changing the sandstones are mostly due to dissolution and
textural composition of the sediments after deposition. reprecipitation of biogenic carbonate or early arago-
The burrowing organisms eat mud and thereby oxidise nite cement. There are usually no other major sources
organic matter and physically destroy the primary lam- of carbonate cement. In the sulphate-reducing zone,
ination. Sediments overturned by bioturbation become carbonate concretions form, often with a negative δ13 C
more exposed to oxidation at the sea or lake bottom. due to the CO2 produced during sulphate reduction.
Bioturbation may reduce the porosity and permeability Carbonate concretions in cores may be mistaken for
of sandy laminae by mixing clay with clean sand. continuous carbonate layers but it is possible to
recognise that they are concretions (Walderhaug and must therefore be linked to sedimentary facies, more
Bjørkum 1998). Even if CO2 is generated from organic specifically to biological productivity relative to the
matter, there are few Ca2+ sources available in clastic sedimentation rate. Environments with low
sandstones or mudstones for making calcite. Leaching clastic sedimentation rates, particularly submarine
of plagioclase can supply some Ca2+, which can be highs, often have high organic carbonate production.
precipitated as calcite, but this can only account for
very small amounts of the calcite observed in such
sediments. The distribution of carbonate cement is 4.6 Meteoric Water Flow and Mineral
related to facies and sequence stratigraphy. Dissolution
The evolution of pelagic planktonic calcareous
organisms in the Mesozoic drastically increased the Meteoric water is rainwater which infiltrates the
supply of carbonate on the seafloor, including in ground. Initially this water is distilled water and there-
deeper waters. Before then most of the carbonate was fore undersaturated with respect to all minerals. The
produced by benthic organisms restricted to shallow reactions between meteoric water and the land surface
water facies. Upper Jurassic and younger sandstones are an important part of the weathering process. Rain-
often contain abundant calcite cement due to the ‘rain’ water contains carbon dioxide (CO2) and sulphur diox-
of calcareous algae, foraminifera and other planktonic ide (SO2) from the air and is therefore slightly acidic,
organisms settling on the seafloor. Silica-producing producing carbonic acid ðH2 CO3 Þ and sulphuric acid
organisms may also be important for diagenesis and ðH2 SO4 Þ.
reservoir quality at greater burial. Organisms like sili- Some of the rainwater seeps down to the ground-
ceous sponges are composed of amorphous silica water, and as long as the groundwater table is above
which at higher temperatures will be dissolved and sea level, meteoric water will flow along the most
replaced by opal CT and quartz. Diatoms and permeable beds into the basin. Meteoric water will
radiolarians may also be a major source of silica first dissolve carbonates and then slowly dissolve
which will be precipitated as quartz. Diatoms appeared unstable minerals like feldspar and mica (Fig. 4.3).
during the Cretaceous and have been a major source of Decaying organic matter in the ground produces
amorphous silica during the Cenozoic. Diatoms can CO2 which is added to the groundwater, making it
produce pure siliceous rocks like the Tertiary more acid. Humic acids generated by decaying plants
Monterey Fm of California, which is both a source also hasten the weathering reactions. At the same time
rock and a fractured reservoir rock. this acidity is neutralised by weathering reactions with
Biogenic carbonate is in most cases the main source silicate minerals like feldspar and the dissolution of
of calcite cement. The distribution of such cement carbonates which consume protons (H+). As the
Production Authigenic mineral
of carbonate precipitation: carbonates,
or silica phosphate, glauconite, Rainfall
organisms sulphides etc. Carbonate sand
from fossils
Wave
base
Met. water
flow - feldsp.
dissolution -
Accumulation Active clastic kaolinite
of carbonate and silica sedimentation precipitation
The primary clastic composition is modified by:

1) Meteoric water leaching and precipitation of kaolinite.
2) By addition of biogenic carbonate and silica.
3) By precipitation of authigenic minerals on the seafloor.
Fig. 4.3 Diagenetic processes in shallow marine environments. cement may be a very important addition to the composition of
Sandstones deposited in these environments will be flushed by the sandstones. The occurrence of siliceous organisms such a
meteoric water flow and/or from the delta top, causing dissolu- sponges can strongly influence reservoir quality at depth
tion of feldspar and mica. Calcareous fossils and early carbonate
groundwater reacts with minerals and in some cases feldspar often remains and the dissolved aluminium
with amorphous phases, it will approach equilibrium and some of the silica is precipitated as kaolinite, so
with many of the minerals present and this will happen there is a rather small increase in porosity and reduced
first with carbonates. In the case of silicate minerals permeability (Fig. 4.4a). The pores between the kao-
these reactions are very slow so the porewater may linite crystals may be too small (Fig. 4.4b, c) to be
remain under- or supersaturated for a long time with filled with oil so that the oil saturation is reduced in
respect to silicate minerals like quartz and feldspar. kaolinite-rich sandstones (see Chap. 20).
Depending on the elevation of the groundwater The silica released from feldspar dissolution can
table and the distribution of permeable layers normally not be precipitated as quartz because of the
(sandstones), the flow of meteoric water can extend low temperature near the surface, but remains in solu-
beneath the seafloor far out into sedimentary basins. tion even if the porewater is highly supersaturated with
Reactions between meteoric water and minerals occur respect to quartz. Silica must, however, also be
in the ground and are a kind of subsurface weathering removed along with potassium by the flowing water.
along the groundwater flow paths. Leaching by mete- If the silica concentration in the porewater increases
oric water is generally strong in fluvial and alluvial too much, kaolinite is no longer stable and smectite
sediments. Even within dry river beds there is a will precipitate instead. This happens in sediments rich
focused flow of groundwater. in volcanic material or biogenic silica and where the
The groundwater level represents the head (poten- flux of meteoric water is low. In a desert environment
tiometric surface) for groundwater flow and ground- evaporation of groundwater may increase the silica
water therefore has a potential to flow through concentration and make smectite more stable.
sediments or other aquifers far below sea level. The porewater does not have to be acidic for kao-
The rates of leaching of minerals like feldspar and linite to form, but the K+/H+ ratio must be low. If the
mica and the precipitation of kaolinite are functions of pH is high the K+ concentration has to be correspond-
the flux of groundwater flowing through each rock ingly lower. Authigenic kaolinite may also form in
volume per unit of time. These are in principle impure limestones as a result of meteoric water
weathering reactions similar to those which take flushing, and the porewater is then certainly not acidic.
place during normal weathering in a humid climate. Even if there is only a small amount of carbonate it
Cations like Na+ and K+ are stripped from silicate will buffer the composition of the porewater.
minerals like feldspar and mica and brought into The average groundwater flux is determined by the
solution. rainfall and the percentage of water infiltration into the
These reactions can be written as below:
2KðNaÞAlSi3 O8 þ 2Hþ þ 9H2 O ¼ Al2 Si2 05 ðOHÞ4 þ 4H4 SiO4 þ 2Kþ ð2Naþ Þ
Feldspar Kaolinite Dissolved silica and cations
2KAl3 Si3 O10 ðOHÞ2 þ 2H þ 3H2 O ¼ 3Al2 Si2 05 ðOHÞ4 þ 2Kþ
þ
Muscovite Kaolinite
+ +
We see from these reactions that low K /H ratios ground. In moderately humid climates the rainfall may
will drive the reactions to the right. Dissolution of be 1 metre/year and the infiltration in the order of
feldspar and mica and precipitation of kaolinite 0.1–0.3 metres/year. High-permeability subsurface
require that the reaction products, Na+, K+ and silica, pathways (aquifers) focus the flow. The aquifers may
are constantly removed and that there is a supply of be sand or gravel beds in muddy sediments. Meteoric
new freshwater which is undersaturated with respect to water may penetrate deeply into sedimentary basins
feldspar and mica. Without a through flow of water because of the potential created by the head of the
these reactions stop because the reaction products on groundwater table but the flux of meteoric water
the right hand side of the equations are not removed. decreases strongly in the deeper parts of basins.
Groundwater must flow into the ground and up to the The meteoric water will gradually become less
surface again or on to the seafloor. A clay coating on undersaturated with respect to minerals like feldspar
a
Kaolinite
precipitated in
the pore space
Quartz
overgrowth
Feldspar coating
Dissolved
feldspar
Fig. 4.4 (a) Scanning electron microscope picture of a sand- aluminium released when the feldspar was dissolved by mete-
stone (Brent Group) from the North Sea. The scale is 0.1 mm oric water. To the left, authigenic quartz is growing on clastic
(100 μm). In the centre of the picture we see a cavity left by a quartz. Note the relatively large pores between quartz and feld-
dissolved feldspar grain. A clay rim around the feldspar remains spar grains and the small pores between kaolinite crystals. (b)
undissolved, outlining the primary grain morphology. In the Pore-filling authigenic kaolinite. We see that the pores between
upper part, authigenic kaolinite crystals are forming small the kaolinite crystals are very small – only 1–2 μm (from T.E.
(10–20 μm) booklets. They have formed from the silica and Maast unpublished)
and mica and its leaching capacity will gradually sedimentation rates the total flow of water through
diminish. In the North Sea basin Middle Jurassic sand- the sediment will be higher because the sediments
stone has been uplifted and eroded during the late remain longer at shallow depth. If the sediment stays
Jurassic, and later overlain by Cretaceous sediments. in the zone intensively flushed by meteoric water, the
Evidence of meteoric water leaching is however lim- amount of feldspar leaching will be high. In basins
ited to a few metres below the unconformity (Bjørkum with high sedimentation rates, synsedimentary
et al. 1990). The most intense mineral leaching will faulting (i.e. growth faults) may disconnect sand bod-
therefore occur near the surface or at relatively shal- ies from the main freshwater aquifers. The degree of
low depth beneath the seafloor. In areas with low
feldspar leaching could then be used as an indication cement in sandstones and this is the best evidence
of the conductivity in the reservoir. that most of the kaolinite formed early at shallow
River water and groundwater are usually supersat- depth. The fact that kaolinite is much more abundant
urated with respect to quartz but undersaturated with in shallow marine and deltaic sandstones than those
respect to amorphous silica. About 10–30 ppm deposited on deeper submarine slopes is also evidence
dissolved silica is common in groundwater and that kaolinite forms at shallow depth.
shallow porewater while the solubility of quartz at
20 C is only 4–5 ppm, showing that quartz does not
form at low temperatures. In very alkaline water (i.e. 4.7 Consequences for Reservoir Quality
E. African lakes), quartz may precipitate at low
temperatures. Meteoric water flushing dissolves feldspar and mica
The early burial history of sandstones is not well and precipitates authigenic clay minerals, most com-
studied for the simple reason that cores are not nor- monly kaolinite. This dissolution produces holes
mally taken at depths shallower than 1–1.5 km in which are secondary pore spaces (secondary porosity)
offshore basins, while onshore, erosion may have but the precipitation of clay minerals like kaolinite
removed most of the youngest sequence. Looking at reduces the porosity, so that there is little net gain in
sandstone thin sections one can often get the impres- pore space. Authigenic kaolinite tends to occur as
sion that kaolinite is precipitated at a relatively late pore-filling minerals and this reduces the permeability.
stage because it is a pore-filling mineral that may Clean well-sorted sand may increase its specific sur-
subsequently be surrounded by quartz cement. More face and pore size distribution due to the authigenic
detailed textural studies and isotopic evidence indicate kaolinite. The smaller pores (<0.005 mm) in between
that the kaolinite is formed early and may be enclosed the authigenic kaolinite crystals may be too small to be
in quart cement. Pore-filling kaolinite must, however, filled with oil because of the high capillary entry
have been hanging to the pore wall and may have been pressure necessary to infiltrate these pores. The total
pushed aside by growing quartz cement. water saturation will consequently then be higher in
The isotopic composition ðδ18 OÞ of kaolinite the reservoir rock.
suggests that it precipitated at relatively low Authigenic kaolinite usually occurs as clusters and
temperatures, in the range 30–60 C, depending on is rarely pervasive through the sandstones, allowing
the assumptions made about the isotopic composition oil to flow between and around the most densely
of the porewater (Glasmann 1989). These kaolinite-cemented pores. However, if the kaolinite
temperatures are little higher than should be expected is altered to illite at greater depth, the damage to the
during meteoric water flushing and it is possible that reservoir may be much more severe, due to permeabil-
some of the kaolinite may be recrystallised at a higher ity reduction.
temperature, resetting the isotopic composition. Much
of what has been described or analysed as kaolinite has
turned out to be dickite, which has the same composi- 4.8 Mechanical Compaction of Loose
tion but often with thicker, more blocky crystals. Sand
Studies have shown that dickite often replaces some
of the kaolinite when temperatures exceed 100 C. During the first part of its burial history (0–2 km) well-
Another possibility is that kaolinite may form dia- sorted sand is generally still loose if it is not carbonate-
genetically from other precursor minerals such as cemented. Mechanical compaction may nevertheless
gibbsite (Al(OH)3) or amorphous aluminium be very significant. Experimental compaction of loose
compounds. Kaolinite could then form without mete- sand with an initial porosity 40–42% shows that,
oric water flushing since such reactions do not produce depending on grain strength and grain size, the poros-
any other cations like K+ which would have to be ity may be reduced to 35–25% at stresses of
removed. In the North Sea basin abundant authigenic 20–30 MPa corresponding to 2–3 km of burial for
kaolinite is found in the shallowest reservoirs normally pressured rocks (Fig. 4.5). The experimental
(1.5–2 km) where there is very little or no quartz data show that well sorted coarse-grained sand is more
compressible than fine-grained sand (Chuhan et al.
a b
Porosity (%) Porosity (%)
24 26 28 30 32 34 36 38 40 42 44
0 25 30 35 40 45 50
5 0
10
Vertical effective stress

15
10
Vertical effective stress (MPa)

Mono-quartz sand
20 Coarse-grained
30 MPa (0.63mm) 200µm
25
20 lithic sand (0.74 mm)
Coarse-graind sand (30 MPa) 30
Quartz sand (0.63 mm)
35
40 30 Quartz sand (0.36 mm)
45 Quartz sand (0.30 mm)
0.71 mm 0.1mm
50 40 Quartz sand (0.18 mm)
50
Mono-quartz sand
40 MPa (0.63mm) 200µm
Fine-graind sand (30 MPa)
Coarse-graind sand (40 MPa)
c Experimental compacƟon of quartz sand and d 5MPa 10 MPa 15 MPa

carbonate ooids 0,85
mm
Porosity(%)
25 30 35 40 45 50 55
0
5
10
15
20
25
0,35
30
mm
35
40
45
50
Mean Grain size

0,18
Microporous carbonate sand (0.75 mm)
Mono-quartzrich sand (0.71 mm) mm
Carbonate ooide sand (0.60 mm) Fig 1 K.B
Fig 1 K.B
Fig. 4.5 (a) Experimental compaction of fine-grained and become larger, reducing the grain-to-grain stress. The grain
coarse-grained sand showing that well sorted fine-grained sand fracturing and the mechanical compaction is then reduced. (d)
is less compressible than coarse-grained sand. (b) The porosity Experimental shear tests on sand applying different normal
loss as a function of grain size due to more grain crushing (from stress loading. Coarse-grained sand fractures at lower stresses
Chuhan et al. 2007). (c) Experimental compaction of bioclastic than fine-grained sand. For medium sand (0.35mm) the horizon-
carbonate, quartz grains and ooids. The bioclastic carbonate has tal stress must exceed about 15 MPa, corresponding to about 1.5
a high initial porosity but the compaction curve is similar to that km burial, before significant grain crushing occurs along faults.
of quartz grains once stresses exceed 10 MPa (from Chuhan Chuhan et al. 2003
et al. 2003). Ooids are soft and rounded so that the grain contacts
2002, 2003). This is because the stres per gain contact In sedimentary basins with normal geothermal
is higher in coarse-grained sand than in fine-grained gradients, quartz cementation will stabilise the grain
sand. Coarse sand grains have usually an irregular framework and prevent further mechanical compac-
surface so that the contact area is small and the effec- tion at about 2 km burial depth (70–80 C) (Fig. 4.6).
tive stress high. At greater depth, compaction is not primarily a func-
Most sandstones are however fine grained or tion of effective stress but of temperature. In cold
medium grained and the frequency of mechanically sedimentary basins (low geothermal gradients) quartz
crushed grains is then not so high before burial to cementation may not start before 4–6 km burial depth
about 2 km (20 MPa). Overpressure reduces the effec- and porosity loss will then occur by mechanical com-
tive stress and will then preserve porosity due to paction and severe grain crushing up to about 50 MPa
reduced mechanical compaction. effective stress.
Strain (Compaction) Porosity

25 MPa 50 MPa
Stress (MPa)
Porosity/depth
curve
Mechanical
compaction
2.5 km
x1 X1 = Overconsolidation
Rock strength due due to build-up of
to cementation overpressure.
X2 = Overconsolidation
due to cementation.
x2
5 km
Depth Effective stress (σv) at

Effective stress hydrostatic pore pressure
at overpressure
Fig. 4.6 Before sandstones become cemented (at 80–100 C) small amounts of quartz cement (2–4%?) make the sandstone
they compact mechanically as a function of effective stress stiffer and “overconsolidated” so that there is little mechanical
(depth) by grain reorientation and grain breakage. Relatively compaction (strain) at greater depth (higher stresses)
The degree of porosity loss by mechanical compac- to obtain good cores because the sandstone
tion determines the intergranular volume (IGV) at the disintegrates in the core barrel. Loose sand grains
onset of chemical compaction (quartz cementation). may then be produced with the oil (sand production).
The IGV measured in some North Sea sandstones Prior to quartz cementation or other types of
varies from about 38 to 28% (Walderhaug 1996) and cementation the sand grains compact mechanically
the net porosity after precipitation of 10% quartz by sliding and reorientation. Sand grains may also
cement will then be very different. fracture under the overburden stress; coarse-grained
sand compacts more due to grain crushing than well
sorted fine-grained sand (Fig. 4.5a,b). In basins like
4.9 Sandstone Reservoirs Buried to the North Sea the quartz cementation increases the
Intermediate Depth (2.0–3.5 km, rock strength at 2–3 km burial depth (80–100 C) but
50–120 C) coarse-grained sand may additionally show significant
compaction due to grain fracturing. Compaction
In basins where there has been mostly continuous experiments show however that carbonate grains
subsidence, reservoirs buried to depths shallower (ooids) are less compressible than quartz grains. This
than about 2.0–2.5 km are still loose or only poorly is because the grain contacts between carbonate grains
cemented, except where there is carbonate cement or will be enlarged due to mechanical and chemical
high geothermal gradients. This is well documented in deformation. The surface of quartz grains is irregular
most of the North Sea basin and parts of the Gulf Coast and so the area of grain contact is very small, also for
Basin (Sharp and McBride 1989, Bjørlykke et al. large quartz grains, resulting in very high stresses
1992). In the Statfjord Field where the Middle Jurassic which can cause grain fracturing. Grain fracturing
Brent sandstone is buried to 2.5–3 km, there are due to shear stress is also sensitive to grain size
intervals that are so poorly cemented that it is difficult
(Chuhan et al. 2003) (Fig 4.5d). During faulting coarse insufficient to precipitate quartz on the microquartz
grained sand may produce grain crushing at about surfaces which requires higher supersaturation than
1 km depth (10 MP effective normal stress). Fine normal quartz (Aase et al. 1996). Rhaxella had not
grained sand requires higher stresses (>15 MPa). evolved before the Upper Jurassic and so older
Poorly sorted sandstones and sand with rock (lithic) sandstones like the Middle Jurassic Brent sandstones
fragments lose much of their porosity at rather shallow do not have this type of microquartz.
(1–2 km) depth. Quartz cementation strengthens the At temperatures above about 100–120 C some of
rocks at a faster rate than the increase in vertical stress what we have called kaolinite has recystallised to
from the overburden. Only 2–4% quartz cement will in dickite, which has the same chemical composition.
most cases effectively shut down mechanical com- Dickite often occurs as slightly thicker crystals and
paction in sandstones, so that further compaction is can also be distinguished from kaolinite on XRD
mainly chemically controlled by the rate of mineral scans. Kaolin or kandite may be used as a common
dissolution and precipitation (Fig. 4.6). The mechani- name for these clay minerals.
cal compaction is important because it determines the Carbonate-cemented intervals may be effective
intergranular volume (IGV) which is the porosity prior barriers to fluid flow. This can be detrimental to the
to quartz cementation. This is typically 25–30% or reservoir quality, though in some cases may be useful
even more for well sorted quartz-rich sandstones. if they are laterally extensive. Such low permeability
Sand with even relatively small amounts of detrital layers may then prevent the flow of gas from below the
clay will compact more than clean sand. Lithic (rock) oil/water contact and from above the gas/oil contact
fragments also compact more readily (Pittman and into an oil-producing well. This is called coning.
Larese 1989) and this is reflected in their lower IGV. The replacement of K-feldspar or plagioclase by
Generally, in relatively well-sorted quartz arenites albite is often observed in sandstone buried to about
and feldspathic sandstones the porosity is to a large 3 km or more and is referred to as albitisation. Albite
extent destroyed by quartz cementation (Fig. 4.7a). becomes more stable than K-feldspar because Na+ is
The amount of quartz cement is mainly a function normally the dominant cation in the porewater while
of the grain surfaces available for quartz precipitation the potassium concentration is reduced due to removal
and the time-temperature integral (Walderhaug 1994). by the clay mineral reactions. Albitisation is normally
High geothermal gradients and slow subsidence rates observed as a partial replacement of K-feldspar or
will therefore tend to increase the amount of quartz plagioclase grains, which does not change the reser-
cement at a specific depth. voir properties very much. Albitisation of plagioclase
In several of the Upper Jurassic reservoir rocks will, however, release some Ca2+ that may then pre-
from the North Sea, amorphous silica from Rhaxella cipitate as calcite (Boles 1982), though the amount is
sponges dissolved to produce high supersaturation of rather limited.
silica relative to quartz. This caused precipitation of a Smectite may be present in some muddy, and par-
coating of minute quartz crystals on the surface of ticularly volcanic, sandstones which have been flushed
clastic quartz grains (Fig. 4.7b, c). This coating of with limited amounts of meteoric water. At
microquartz has prevented or retarded the precipita- temperatures from about 70 to 80 C smectite dissolves
tion of later quartz cement and is the main reason for and is replaced by mixed-layer minerals and illite.
high porosity and good reservoir quality at great depth Sandstones containing smectite normally have poor
(up to 5 km) in these reservoir rocks. The microquartz reservoir quality.
precipitated at low temperature (60–65 C), when the Dissolution of smectite and precipitation of illite
porewater was highly supersaturated with respect to and quartz will cause a sharp increase in the seismic
quartz through the dissolution of opal A or opal CT velocity and rock density and this mineral transition
and while the quartz growth rate was low. At higher may therefore show up as a horizontal reflector on
temperatures when unstable silicates like opal A, opal seismic and could be mistaken for a fluid contact i.e.
CT and smectite have dissolved, the porewater will gas/oil or oil/water contact (Thyberg et al. 2010).
only be slightly supersaturated with respect to quartz,
a Primary clastic grain Quartz cement (overgrowth)
2272 15KV 10µm WD32
b c
100 µm 5 µm
d e
100 µm
200 µm
Fig. 4.7 (a) SEM image of quartz cement which has grown on high silica supersaturatation caused by dissolution of organic
detrital quartz grains into the pore space. Note that the silica (sponge spicules). (c) is an enlargement of Fig. 4.4b
authigenic quartz crystals have smooth crystal surfaces while showing microquartz crystals that are not overgrown at greater
the primary quartz grain has an irregular abraded surface from depth. From T.E. Maast (unpublished). (d) Pictures (SEM) of
weathering and transport of the sand grain. Scale bar 10 μm the siliceous sponge (Rhaxella) which is the source of much of
(0.01 mm). Oil inclusions may be found in between the primary the early quartz cement. To the left a broken surface. The
grains and the cement, and also in the cement. Sandstone from sponges have a circular outline in cross-section (thin section),
the Brent Group, North Sea. (b) A thin layer of small quartz see Fig. (e). Upper Jurassic sandstone from the North Sea (from
crystals covering clastic quartz grains has been precipitated at T.E. Maast unpublished)
4.10 Deeply Buried Sandstones predicting the amount of quartz cement and remaining
(>3.5–4 km, >120 C) porosity. Between 100 and 140 C the rate of quartz
precipitation may double four times, i.e. increase 16-
Once quartz cementation has started and quartz over- fold. By contrast the effective stress increases linearly
growth has formed, quartz cementation does not stop with depth (under hydrostatic conditions) and thus
until nearly all the porosity is lost, unless the tempera- increases only by 30–40% through the interval from
ture drops below 70–80 C. 3 to 4 km of burial. Temperature is therefore by far the
In most sedimentary basins we find there is a rather main factor controlling the rate of quartz precipitation.
strong reduction in porosity and permeability in sand- This allows the amount of quartz cement and the
stone reservoirs from about 3–3.5 km to 4–4.5 km porosity to be modelled as an exponential function
burial depth, corresponding to a temperature range (Arrhenius equation) of the temperature integrated
from about 120 to 160 C. This is due in most cases over time and proportional with the surface area avail-
to precipitation of quartz cement and diagenetic illite. able for quartz precipitation. Commercially available
The rate of quartz cementation increases as an expo- programs (Exemplar, Touchstone) have been devel-
nential function of temperature and we estimate that oped for this purpose based on Walderhaug (1996).
the rate may increase by a factor of 1.7 for every 10 C Dissolution at grain contacts requires stress, to main-
temperature increase (Walderhaug 1996). The precip- tain the contact, so the process is often called “pressure
itation of quartz is also a function of the surface area solution”, but in the case of silicate minerals tempera-
available for quartz cementation and as quartz cement ture is the most important factor. The minerals at the
is filling the pores the surface area available for further grain contact are also important and dissolution is
quartz growth decreases (Fig. 4.8). The crystal enhanced by mica or clay minerals at the grains
surfaces of quartz have different solubilities and contacts (Bjørkum 1996). Contacts between mica or
potential for crystal growth and thereby for porosity illitic clay and quartz are preferred areas of dissolu-
reduction (Lander et al. 2008). tion. The rate-limiting process in quartz cementation
The temperature history of the sandstones at this seems to be the rate of nucleation and precipitation in
stage becomes rather critical in terms of modelling and the pore space and the reaction is then surface-
controlled (Bjørkum et al. 1998). Quartz cementation
is therefore insensitive to variations in effective stress
Compaction due to quartz cementation at
in the grain framework. If the dissolution process had
constant temperature
been rate-limiting, quartz cementation would have
Volume 100
100°C 120°C 140°C
been more sensitive to the effective stress in addition
Sandstone thickness (m)
95 to temperature and surface properties of the minerals.

100°C The silica dissolved at grain contacts or along
90 stylolites is transported by diffusion to the grain
120°C surfaces where the quartz overgrowth forms (Oelkers
85
et al. 1996) (Fig. 4.9).
80
dV/dt If the transport of silica was the rate-limiting factor
140°C
for quartz cementation (transport-controlled) we
75 would expect to observe a concentration of quartz
0 20 40 60 80 100 cement close to stylolites of thin clay laminae, where
Time (million years) dissolution occurs and where the concentration in the
Fig. 4.8 Modelling of quartz cementation and chemical com- porewater would be highest. When dissolution is
paction due to quartz dissolution and cementation as a function concentrated along stylolites with relatively large
of time and temperature (from Walderhaug et al. 2001). We see spacing (>50–100 cm), studies suggest that the
that the rate of porosity loss (compaction, dv/dt) is highest at amount of quartz cement may decrease away from
high temperatures and also when the porosity is still relatively
high. When the porosity is reduced the surface area available for the stylolites, indicating some degree of transport con-
quartz cementation becomes smaller so that the rate of cemen- trol (Walderhaug and Bjørkum 2003). The stylolites
tation slows down represent a barrier for fluid flow both during
compaction and during production, but may not be cementation. A consequence of a surface-controlled
completely continuous laterally. quartz precipitation model is that quartz cementation
We have seen that the rate of quartz cementation is will continue as long as the temperature is above the
a function of temperature, the degree of supersatura- threshold temperature for quartz growth (70–80 C)
tion and the surface area available for quartz and there is remaining porosity in the sandstone. It is
important to note that quartz cementation, and hence
sandstone compaction, will continue also during basin
Quartz
overgrowth inversion and uplift, but at a slower rate (Fig. 4.10).
Diffusion
When uplifted to shallower depths (temperature
of silica <70–80 C) the sandstone is unloaded and without
any chemical compaction there will be net extension
due to elastic expansion due to reduced stress.
Stylolite As long as the temperature is higher than 70–80 C
the cementation process will proceed but at a slower
rate, modifying the normal porosity/depth relation
found elsewhere in the basin. During progressive
burial quartz cementation must continue until all avail-
able porosity is filled and the sandstone becomes a
well-cemented hard quartzite after exposure to
Chlorite Detrital 200–300 C for several million years.
coatings clay Microquartz Asphalt/
Bitumen If the surfaces of sand grains are coated with other
minerals, or with substances like petroleum or bitu-
Fig. 4.9 Schematic illustration of a stylolite. The dissolved men, quartz overgrowth is hindered or at least stopped
silica is transported away from the clay-rich stylolite by diffu- for some time (Fig. 4.9). A thin layer of authigenic
sion. This makes more long distance and advective transport of
silica difficult. The rate of precipitation of quartz cement is a chlorite has proven to be very effective in preventing
function of the surface area available. Grain coatings such as quartz overgrowth. This has been described from
chlorite, illite, detrital clay, iron oxide (haematite), microquartz many places around the world like the Tuscaloosa
and bitumen prevent or retard quartz cementation
Diagenetic processes during burial.

Time
10
Uplift
45 Extension
Mechanical
Brittle
compaction
deformation
70
Temperature (°C)
Chemical
INTEGRATED TIME
105 compaction
TEMPERATURE
(Quartz cementation)
Illitisation of Continued
140 chemical
kaolinite +KF
compaction
during uplift
Porosity
Fig. 4.10 Diagenetic processes, mainly quartz cementation, as a function of temperature and time. Note that quartz cementation
will continue also during uplift as long as the temperature exceeds 70–80 C
sandstones of Southern Louisiana (Pittman and Larese This is the main reason for the good reservoir
1992). properties in petroleum discoveries on Haltenbanken,
Sandstones with grain coatings may remain where the porosity sometimes exceeds 25% at more
uncemented down to 4–5 km burial depth and be than 5 km depth. The Kristin Field in Haltenbanken is
subjected to 40–50 MPa effective stress, causing per- one example of reservoirs that have preserved rela-
vasive grain crushing (Chuhan et al. 2002). tively good reservoir quality at great depths (>5 km).
Under the microscope, and particularly when using Reservoirs characterised by high temperatures and
cathodoluminescence, we can see that some of the high pressure (>150 C (300 F)) and 89 MPa (10,000
sand grains have been fractured and later healed by psi). are often referred to as HTHP reservoirs. They
quartz cement (Fig. 4.11a, b). The fractures in the may represent technical challenges with respect to
clastic grains do not usually continue through the drilling and production because of the high fluid
quartz overgrowth, demonstrating that they predate pressures and the potential for gas leakage.
the quartz cementation. At temperatures of about 150 C most sandstone
At Haltenbanken, offshore mid-Norway, the Juras- reservoirs have too little porosity due to quartz cemen-
sic Tije and Garn formations have abundant chlorite tation. The preservation of porosity in these HTHP
coatings and also illite coatings (Ehrenberg 1993, reservoirs depends on clay coatings on quartz grains
Storvoll et al. 2002) (Fig. 4.12). retarding precipitation of quartz cement, as is the case
in the Kristin Field. If the reservoir had been exposed
to high temperature for a relatively short geologic time
Quartz cementation of fractured grains from
Smørbukk Field, Haltenbanken, offshore mid-Norway the porosity would be higher due to less quartz cement.
a -1mm
To predict the distribution of such high porosities at
1mm great depth we need to understand what controls the
development of authigenic chlorite. The chlorite found
in marine sandstones is most probably an alteration
product of an earlier iron silicate phase (precursor)
formed on the seafloor and which may be linked to
facies. This may be iron- and magnesium-rich
smectites coating the primary quartz grains. Illite
may also form an effective coating preventing quartz
overgrowth (Heald and Larese 1974, Storvoll 2003)
b -200µm and this may have formed from smectite.
a 5011.75 m
Natural fractures in reservoir sandstone (Tilje Fm,
Smørbukk Field). The quartz grains are chlorite-coated
but quartz cement has grown from fractured quartz. From
Chuhan et al. 2002
Fig. 4.11 (a, b) Cathodoluminescence pictures showing that

quartz grains have been subjected to fracturing prior to quartz
cementation. Quartz cementation was delayed by chlorite
coatings but grain fracturing exposed fresh quartz surfaces Fig. 4.12 Quartz grains coated with chlorite preventing quartz
from which quartz could grow into the pore space. Authigenic overgrowth. Note pore-filling authigenic kaolinite. Tilje Forma-
quartz is darker than the clastic high temperature quartz tion, Haltenbanken
a b c
Number of Number of
measurements measurements
15 Present 15 Present
temperature temperature
10 10
5 5
Fluid inclusion Fluid inclusion

temperatures temperatures
80 90 100 110 120 130 °C 80 90 100 110 °
120 130 C
Measurements from the water-saturated zone. Measurements from the water-saturated zone.
Measurements from the oil-saturated zone. Measurements from the oil-saturated zone.
Fig. 4.13 (a) Inclusions of oil in quartz cement on sand grains. Jurassic sandstones, Haltenbanken (offshore mid-Norway)
(b) Temperatures from oil inclusions in the Fulmar Field, North (Walderhaug et al. 1990b)
Sea (Saigal et al. 1992). (c) Fluid inclusion temperatures from
It is typical that clean well-sorted sandstone has a rate controlled by the temperature (Walderhaug
the best porosity down to about 3.5–4 km but then 1994b) (Fig. 4.13c).
loses its porosity rapidly due to quartz cementation. There is no evidence that quartz cementation is
Sandstones with a moderate clay content lose more episodic, controlled by the supply of silica, or that
porosity during mechanical compaction compared to quartz cementation stops after the sandstones have
the clean sandstone, but may preserve more porosity at become oil-saturated (Walderhaug 1990, Saigal et al.
greater depth due to retarded quartz overgrowth. 1992) (Fig. 4.13). In a water-wet reservoir precipita-
tion can still continue in the remaining water around
the grains. At high oil saturation, the transport of silica
4.11 Fluid Inclusions in Quartz Cement by advection as well as by diffusion becomes much
less efficient. The continued growth of quartz cements
Inclusions of small drops of oil may be trapped in after oil emplacement results from the closed system
quartz cement and show up well in fluorescent light nature of quartz cementation in sandstones.
(Fig. 4.12a). Fluid inclusion data from quartz helps to In an oil-wet system, however, quartz can not pre-
constrain the temperature for quartz cementation and cipitate on the grain surfaces and oil or bitumen may
in sandstone reservoirs it has been demonstrated quite become very effective coatings. Asphaltic oil and
clearly that quartz cementation continues after oil bitumen formed by biodegradation or other processes
emplacements in a reservoir (Karlsen et al. 1993, may preserve good reservoir quality, particularly if the
Walderhaug 1990). heavy oil only occurs as a grain coating.
The lowest fluid inclusion temperatures in quartz In summary, quartz cementation is controlled by
cement indicate an onset of quartz cementation close the slow kinetics (high activation energy) for quartz
to 70–80 C (Burley et al. 1989, Walderhaug 1994a). cementation and normally a minimum temperature of
In the Ula Field (North Sea basin), fluid inclusion 70–80 C is required. This is however also dependent
temperatures in quartz cement range from 86 C to on the pH. In sedimentary basins marine porewater
126 C (Fig. 4.13b), the highest temperature being starts out with a pH up to 7 but quickly becomes
close to the present day reservoir temperature which more acid due to the build of CO2 and other reactions
is also the maximum burial depth and temperature with the minerals present. At 3–4 km depth the pH
(Saigal et al. 1992). Fluid inclusions from 24 samples may typically be 4.5–5 but at 120 C the pH is close to
from 11 different reservoir units from the North Sea neutral (see Chap. 3). The rate of quartz cementation is
and Haltenbanken also show temperatures from about then lowered by the pH but increased by higher
80 C to values close to the present reservoir temperatures. At very high pH quartz cementation
temperatures (Walderhaug 1994a). This shows that may occur at the surface and silcrete is fine-grained
quartz cementation occurs as a continuous process at quartz formed in soils due to concentration of
a b
Fig. 4.14 (a) Pore-filling illite replacing pore-filling kaolinite preserving the kaolinite textures in Jurassic sandstone,
Haltenbanken, offshore mid-Norway. (b) Pore-filling illite probably altered from smectite. Triassic sandstone from the North Sea
porewater by evaporation, and quartz is also forming precursors for illite. However, the formation of illite
in some alkaline African lakes. requires potassium, and K-feldspar is usually the only
Authigenic illite consists of thin hair- or plate-like significant source present in the sediment.
minerals and it is fairly obvious that they would have a
detrimental effect on reservoir quality by reducing the KAlSi3 O8 þ Al2 Si2 05 ðOHÞ4 ¼ KAl3 Si3 O10 ðOHÞ2
permeability (Fig. 4.14a, b). SEM images are routinely þ 2SiO2 þ H2 O
taken from dried-out cores where the illite appearance K-feldspar þ Kaolinite ¼ Illite þ Quartz
is no longer representative of its morphology in the
reservoir. When cores are dried without destroying the High temperature detrital K-feldspar contains some
delicate illite morphology the pore space often looks sodium and the illite formed contains less potassium
like it has been filled with rockwool. Illite can often be than indicated here, where the formula for muscovite
seen to grow at the expense of kaolinite and may also is used.
form by alteration of smectite. Although authigenic The reaction between K-feldspar and kaolinite
illite may also be observed on fractures and other occurs at about 130 C and above this temperature
places where there are no obvious precursor minerals, these two minerals are no longer thermodynamically
it is most commonly found as a replacement of an stable together. In the North Sea basin and at
earlier Al-rich mineral phase. Haltenbanken this corresponds to a burial depth of
Because of the low Al-solubility in porewater, illite about 3.7–4 km. A sharp increase in the illite content
will in most cases precipitate where the source of Al in sandstone reservoirs is observed at this present day
is available locally from a dissolving mineral. depth (Bjørlykke et al. 1986, Ehrenberg 1992). From
Calculations suggest that the solubility of aluminium thin sections and SEM images the illite can be seen to
is only about 1 ppm at 150 C and that organic acids replace kaolinite, and potassium then has to diffuse
have little effect in terms of increasing its solubility from the K-feldspar to the kaolin. If the matrix is well-
(Bjørlykke and Aagaard 1992). cemented the rate of diffusion is reduced, and the
The formation of illite from smectite via mixed- minerals are then able to co-exist at higher
layered minerals is well known and occurs in temperatures if they do not occur close together.
sandstones in the temperature range of 70–100 C. In sandstones with little or no K-feldspar, however,
Sandstones with abundant smectite are poor reservoir kaolin remains stable at greater depth as it is not
rocks at the outset, and illitisation of such rocks may dissolved and replaced by illite. The formation of illite
itself slightly improve reservoir quality as illite has a can therefore be predicted from the sandstone prove-
lower specific surface area than smectite. In better- nance with respect to K-feldspar supply, and from the
sorted and potentially good reservoir rocks kaolin early diagenesis and freshwater flushing with respect
minerals (kaolinite or dickite) are the most important to the distribution of kaolinite. If a sandstone is
derived from an albite-rich gneiss the K-feldspar con- The porewater reacts with the minerals it is in
tent is likely to be too low and much of the kaolinite contact with, and with increasing temperature the
would then not be illitised. composition becomes more and more in equilibrium
Similarly, not much illite will be formed in with these minerals. This is because the kinetics of the
sandstones with little kaolinite or smectite. Both in mineral reactions become faster. The porewater in
Haltenbanken and the North Sea there are Jurassic sedimentary basins consists of solutions buffered by
reservoirs where plagioclase is the dominant feldspar the minerals present. The pH is controlled partly by the
and where the low K-feldspar content is unable to carbonate reactions (pCO2), and the pH will decrease
supply the necessary potassium for illitisation of kao- from 7.5–8 in the seawater to 4–5 at 3–4 km depth. As
linite. The low illite content in such reservoirs temperature rises above 100 C, silicate reactions
preserves better permeability. This is a direct function become increasingly important, e.g. between K-
of the provenance and could be due to erosion of albite feldspar, kaolinite and illite which will determine the
gneisses rather than granitic gneisses. K+/H+ ratio.
The distribution of authigenic illite in sedimentary Organic acids are weak acids which can not signifi-
basins like the North Sea and Haltenbanken shows that cantly change the pH in the strongly buffered
illite formation is strongly controlled by the present porewater. The buffering capacity of organic acids
day burial depth and temperature. The increase in illite has been shown to be orders of magnitude lower than
content at about 3.7–4.0 km is usually very sharp, for the carbonate and silicate systems (Hutcheon and
indicating a temperature-controlled reaction rather Abercrombie 1990). Organic acids generated in source
than a high kinetic reaction rate when the association rocks like the Kimmeridge Clay Fm in the North Sea
of kaolinite and K-feldspar becomes thermodynami- are likely to be neutralised by reactions with calcite
cally unstable. Basin loading from thick Pleistocene which is commonly present in these source rocks. The
sequences in these areas suggests that the illite formed limited effect of organic acids and CO2 on mineral
recently. dissolution and diagenesis has also been shown exper-
K-Ar dating of illite gives variable ages for the imentally (Barth and Bjørlykke 1993).
formation of illite. This is probably because even When porewater moves it will nearly always cause
very small amounts of detrital (older) mica or feldspar some dissolution and precipitation but this is rarely
will produce too-old ages (Hamilton et al. 1989). significant due to low velocities. The volume of
minerals dissolved or precipitated can be calculated
using the following equation:
4.12 Porewater
Vc ¼ F t sin α ðdT=dZÞ Δ S=ρ
In a sedimentary basin the amount of solids dissolved
in porewater is very small compared to the volume of The volume of precipitated mineral Vc is a product
solids. During burial diagenesis, significant precipita- of the fluid flux integrated over time (t), the angle of
tion of authigenic minerals must be accompanied by fluid flow relative to the isotherms (α), the geothermal
dissolution of other minerals or the same mineral, as in gradients (dT/dZ), the solubility as a function of the
the case of pressure solution. Even if the porewater is temperature (ΔS) and the density of the mineral (ρ).
supersaturated with respect to a certain mineral, only The solubility gradient (ΔS) is 1–3 ppm/ C,
very small amounts can precipitate before the depending on the temperature (Wood 1986). This
porewater attains equilibrium. Precipitation of new means that at 100–150 C about 2 ppm of quartz
minerals requires that other minerals dissolve because precipitates for each degree the porewater is cooled.
the porewater has very little capacity to store ions. This gives a solubility gradent of 2.10–6/ C. If the
This has been quantified using chemical modelling geothermal gradient is 30 C=km ð3 10 2 C=mÞ,
(Giles 1997) from which it can be concluded that porewater must move upwards more than 30 m to
extremely large fluid fluxes are required through the reduce the temperature by 1 C and the quartz cement
pores in order to dissolve or precipitate significant is distributed through these 30 m of sandstone. Assum-
amounts of cements like calcite or quartz. ing vertical flow (sin α ¼ 1) and geothermal gradients
close to 30 C=km ð3:10 2 C=mÞ, a solubility gradient precipitating calcite on the way down when the
of 2:10 6 = C and a quartz density of 2.65 g/cm3 we porewater is heated. All the calcite would then be
obtain: dissolved and quartz would be precipitated by this
process.
8
Vc ¼ F t 2:3 10
From the above equation we see that each 1 m3/m2 4.13 Effect of Oil Emplacement
will precipitate about 2:3 10 8 m3 of quartz. To pre-
cipitate 10% quartz cement ðVc ¼ 0:1Þ requires a total When oil migrates into a reservoir rock the water
flow (integrated flux over time)) of ðF tÞ of about content is reduced to a percentage of the porosity
4 106 m3 =m2 . This assumes that the porewater is in corresponding to ‘‘irreducible water saturation’ if the
equilibrium with the mineral phases which is true at rock is water-wet. This may vary from 10% water
depth with temperatures exceeding 80–100 C. At 30% content in clean sand to 50% or more in clay-rich
porosity a water column of 1,200 km must pass sandstone, the value depending on the amount and
through a sandstone layer to introduce 10% quartz type of clay present. The traditional assumption has
cement. been that the emplacement of oil stops, or at least
This is clearly impossible in sedimentary basins. In slows down, the rate of diagenetic processes and
addition compaction-driven porewater is not flowing hence the rate of porosity reduction.
upwards in relation to the surface and is therefore If the transport of silica by either diffusion or
normally not subjected to cooling which would cause advection was rate-limiting for the quartz cementation
precipiation of quartz (see Chap. 10). one would indeed expect the rate of quartz cementa-
At shallow depth the temperature is low and the tion to be very much reduced. Fluid inclusions in
fluid flow rate high, so in the zone of meteoric water quartz cement, however, clearly demonstrate that in
flushing this may not be true. The porewater may then fact quartz continues to grow after oil emplacement in
be undersaturated or supersaturated, particularly with sandstone reservoirs (Bjørkum and Nadeau 1998,
respect to silicate minerals. When temperatures Walderhaug 1990). The explanation for this is that
exceed 100 C the porewater will approach equilibrium silica is transported along the thin film of water
with the minerals, both because of higher reaction between the mineral grains and the oil phase. It is
rates and low flow rates. possible that the rate of quartz cementation could be
Small amounts of calcite are nearly always present, slower after oil emplacement but this would imply that
at least in marine sediments. Calcite has a retrograde the quartz cementation would no longer be surface
solubility meaning that the solubility normally controlled, but transport controlled. It is very difficult
decreases with increasing temperature. The solubility to prove that a higher porosity in the oil-saturated part
also depends on the pressure, but in most cases it is the of a reservoir is due to the introduction of oil. There
temperature effect which is strongest. Upwards are usually so many other variables such as facies that
(cooling) porewater flow, which should precipitate influence the final reservoir porosity. Since the oil is
quartz, will dissolve calcite at a rate which is 30–100 emplaced gradually and the oil/water contact moves
times faster (Bjørlykke and Egeberg 1993). We may downward over time, a sharp difference in porosity
therefore conclude that if calcite was present in a should therefore not be expected right at the present
sandstone very little quartz could have precipitated OWC if quartz cementation was a function of oil
until all the calcite had been dissolved. emplacement.
Thermal convection is probably not very significant In gas reservoirs the saturation may be rather high
in sedimentary basins except where there are hydro- in clean sand, perhaps reducing the water film around
thermal heat sources (Bjørlykke et al. 1988). If thermal the grains and the quartz overgrowth, but the degree to
convection did occur at a significant rate, however, which this might apply is uncertain.
quartz could precipitate because the same water could Biodegraded and asphaltic oil or bitumen will stick
be used over again, precipitating quartz and dissolving to the grain surfaces and effectively prevent quartz
calcite on the way up, and dissolving quartz and overgrowth but then some of the porosity may be lost
to the bitumen and heavy oil.
4.14 Different Types of Sandstone and some feldspar and mica is dissolved during each
cycle.
It is important to distinguish between sandstones of Palaeozoic quartzites typically occur as transgres-
different primary composition: sive sheet sands on cratons. On the North American
Volcanoclastic sandstones may vary greatly in craton there are good examples of this in the Lower
composition depending on the volcanic source and Palaeozoic sequence. Such clean shallow marine
the depositional environments. Basic volcanic rocks sandstones have extremely good reservoir properties
in particular have a very low content of stable grains at shallow to moderate burial depth. This is likewise
like quartz, but a high content of basic feldspar and the case with aeolian sandstones. Fluvial sandstones
pyroxenes which break down rapidly, both mechani- are also normally well-sorted in such environments
cally and chemically. Matrix-rich sandstones like because they are often reworked aeolian sands.
greywackes may have had a higher sand content at Carbonate cement in shallow marine sandstones is
the time of deposition because many of the grains were mostly derived by recrystallisation of calcareous
unstable during diagenesis and became effectively part organisms. Meteoric water will dissolve aragonite
of the matrix. What were deposited as grains of volca- and precipitate calcite in sandstones, producing early
nic rock fragments may be squeezed so that they cement.
become a chlorite-rich matrix. In modern environments, particularly in beach and
Volcanoclastic sandstones lose most of their poros- shoreface settings, fragments of crushed calcareous
ity at rather shallow depth (<1–2 km) and therefore organisms are quite common. We find less carbonate
make poor reservoir rocks. However, the geothermal cement in fluvial sandstones because of the lower
gradients in volcanic regions may be high, causing biogenic carbonate production in freshwater. Carbon-
source rocks to mature at shallow burial depth and ate cement has a local source in most cases, but may be
thus increase the potential for migration into shallow redistributed and concentrated by diffusion. The range
structures. of effective diffusion is generally small (<1 m)
Lithic sandstones have a high content (>10%) of because the porewater is in equilibrium with calcite
rock fragments. Normal and coarse-grained granites and there are small concentration gradients. Advective
and gneisses produce grains that mostly consist of a flow will transport dissolved carbonate but can not
single mineral while sandstones derived from finer- precipitate tight carbonate cement because the perme-
grained igneous and metamorphic rocks are mostly ability decreases as precipitation proceeds. The advec-
comprised of rock fragments. Rock fragments are tive flow will then tend to bypass the volume where
generally weaker than quartz and feldspar grains, as carbonate cementation has started. In the case of
has been demonstrated experimentally (Pittman and compaction-driven upwards-directed (cooling) pore-
Larese 1991). water, the solubility of calcite will increase, causing
Arkoses contain more than 25% feldspar and such dissolution rather than precipitation.
sandstones are typical of tectonically active basins like Aeolian sandstones and other desert sandstones gen-
rift basins where the erosion, transport and deposition erally show less evidence of meteoric water flushing
of basement derived rocks is fast, leaving little time than fluvial and shallow marine sandstones. Sandstones
for feldspar to weather. Temperature and rainfall also like the Permian Rotliegend from the southern North
play a role here. Arkoses compact more mechanically Sea have relatively low amounts of kaolinite and more
than quartzitic sandstones, leaving a smaller intergran- smectite or illite as pore-filling cement. However, even
ular volume to be cemented with quartz at greater deserts have groundwater so some leaching occurs.
depth. The area available for quartz cementation is Fluvial sediments will normally be flushed by ground-
also reduced since quartz does not grow on feldspar. water after deposition and in most cases show ample
Feldspathic sandstones and quartzites are the most evidence of feldspar leaching. Reworking of such
common sandstone reservoir rocks. The feldspar con- sediments will bring authigenic kaolinite into the clas-
tent is usually a function of the source climate and the tic clay fraction. Continental sandstones often have
relief in the drainage area. On tectonically stable haematite or manganese oxide coatings on quartz
cratons sediments are repeatedly eroded and deposited grains and this may inhibit quartz overgrowth.
Turbidites form important reservoirs in many 4.15 Predictions of Reservoir Quality

basins and although the reservoir quality may be less
favourable than within shallow marine sandstones, The oil industry has a practical need to be able to
they often form extensive vertically-stacked reservoir predict the properties of reservoir rocks ahead of dril-
sequences and this may compensate for the lower ling. When planning petroleum production the rock
porosity. properties between the wells must also be estimated.
Turbiditic sandstones generally have a higher clay Particularly in the deeper reservoirs, porosity is the
content than shallow marine sandstones. There is how- most important factor determining the economic via-
ever a wide range of clay contents in turbidites from bility of a prospect. The main diagenetic processes
rather clean, usually proximal and channel facies, to with quartz cementation are summarised in Fig. 4.15.
more clay-rich distal and overbank facies. In a relatively mature basin the porosity/depth
Sandstones with clay contents higher than 10–15% functions of the different reservoir rocks can be treated
lose their porosity rapidly with mechanical compac- statistically so that the uncertainty of the estimates can
tion, because the detrital clay acts as a lubricant in the be expressed. The estimates based on the statistical
compaction of the quartz grains. Turbiditic sandstones averages can also be adjusted up or down as a function
of Paleocene and Eocene age form very important of temperature, stress etc., depending on what the inter-
reservoirs in the North Sea. The Frigg sandstone in preter considers most significant. The Middle Jurassic
the Frigg Field in the Norwegian Sector is an example, Brent sandstone in the North Sea has been intensively
where the reservoir quality is rather good despite studied and a relatively linear trend found between
representing a distal facies relative to the Shetland burial depth and porosity (Giles et al. 1992, Bjørlykke
Platform where the sand originated. et al. 1992, Ramm et al. 1992, Wilson 1994). Porosity
Because turbidites are deposited further away from predictions will depend on the primary sediment com-
land, they are less exposed to flushing by meteoric position and the subsequent compaction processes.
water. Turbidites normally contain less evidence of Above we have discussed some of the processes
feldspar dissolution and authigenic kaolinite precipi- that cause reductions in porosity and permeability. All
tation than shallow marine and fluvial sandstones. In these processes are driven towards denser packing of
proximal turbiditic facies, however, the sandstones grains and thermodynamically more stable mineral
may have good contact with the meteoric water lens. assemblages as the stress and temperature increases
In the North Sea, Tertiary turbidites generally have a during burial. The kinetics of mineral reactions
relatively low content of authigenic kaolinite, but it determines the rate of thermodynamic equilibriation,
may be higher in proximal Cretaceous turbidities which increases as an exponential function of
which formed around islands produced by local uplift. temperature.
The total flow of meteoric water through The rate of compaction as a function of stress can
sandstones is a function of meteoric water flux and be measured experimentally using rock mechanics
the sedimentation rate. At low sedimentation rates a testing procedures. However, the reactions involved
given volume of sand spends more time in the shallow in chemical compaction are so slow, particularly in
zone of meteoric water flushing. silicate rocks, that it is difficult to reproduce them in
Cretaceous and Tertiary turbidite sandstones are the laboratory although in some cases this is now
often very tight due to pervasive carbonate cement. becoming possible.
This may be due to pelagic carbonate organisms being In the field of clastic diagenesis, petrographic
mixed in with the turbidite sand and recrystallising observations about mineralogy and textural
into carbonate cement. This makes many Tertiary relationships are used to interpret the sequence of
sandstones hard and indurated even if they have not dissolution and mineral precipitation and its relation-
been buried very deeply. ship to changes in porosity and permeability. It is
Many of the planktonic carbonate organisms devel- however important to consider the geochemical
oped during the Jurassic, so before then the sources of constraints on diagenetic reactions. During burial the
carbonate cement in deep sea sandstones were more reaction must add up so the dissolution is balanced by
restricted. precipitation because there are strong limitations with
Na+ , K+ H+. H2O Freshwater flushing in fluvial and

shallow marine environments.
SiO44–
Open system diagenesis. Dissolution of feldspar
and mica and precipitation of kaolinite.
Mechanical compaction
70–90°C Dissolution of smectite –

Quartz precipitation of illite.
cement Onset of quartz cementation.
The porosity at the onset of quartz
Normal cementation (IGV) is important for
curve prediction of porosity.
High
sed. rate
Grain 130°C Dissolution of kaolinite and

coat. K-feldspar. Precipitation of illite and quartz.
160°C, 4–5 km depth. 7–8 km in cold

basins (low geothermal gradients).
Fig. 4.15 Summary of the most important processes in clastic scale of 1–10 m distance. The composition of the dissolved
diagenesis. Dissolution of feldspar and mica requires a through material is equal to what is precipitated. Carbonate cements
flow of meteoric water removing K+ (Na+) and silica before are usually derived from biogenic carbonate or clastic carbonate
kaolinite can be precipitated. This process is most dominant in grains. Quartz cementation is controlled by temperature (geo-
fluvial and shallow marine sediments at shallow burial. At thermal gradients), subsidence rates and the presence of grain
greater depth diagenetic reactions are nearly isochemical at a coatings
respect to supply and removal of solids dissolved in Early diagenetic processes like marine carbonate
the porewater. cementation and meteoric water flushing are also
Changes in mineralogy or porosity with depth may linked to facies and they strongly influence the burial
provide useful depth trends within an area, but in a diagenesis and the porosity reduction at depth. The
sedimentary basin the initial mineral composition precursor minerals controlling the growth of chlorite
may vary laterally. This is also the case for early coating are also probably to a large extent controlled
diagenetic processes like meteoric water flushing and by facies. The distribution of silica organisms (like
marine cementation. The observed changes with depth Rhaxella) which can produce microquartz coatings is
may therefore also reflect some of these factors and linked to the environment and ecology.
not only the burial depth. A broad geological background is therefore
Based on the theory that the rate of quartz cemen- required to synthesise all the factors that have to be
tation is controlled only by temperature, time and the considered before modelling or making semi-
grain surface available for quartz precipitation, the quantitative predictions of reservoir quality. The
amount of quartz cement and consequently the poros- capacity of porewater to keep solids in solution is
ity can be modelled (Bjørkum et al. 1998, Walderhaug always rather limited and mineral dissolution and pre-
1996). The presence or absence of clay or other cipitation must therefore balance.
coatings is the most critical input for this modelling At greater burial depth (3–4 km) the solubility of
because it determines the area available for quartz silicate minerals increases but the volume of
cementation. Prediction of reservoir properties must porewater is low and the potential for supersaturation
start from sedimentological facies models. The depo- and undersaturation strongly reduced. Precipitation of
sitional environment and the provenance of the clastic new mineral phases must therefore be linked to the
sediments determine the starting composition for the dissolution of other minerals or of the same mineral.
diagenetic processes. Assuming that the burial diagenetic processes are
relatively isochemical, the reservoir properties can be
predicted from depositional facies and provenance 4.16 Porosity/Depth Trends in

studies which define the starting composition for Sedimentary Basins
burial diagenesis. Both observations and theoretical
arguments suggest that advective transport can not Data from wells in sedimentary basins which have
significantly change the rock compostion below the undergone almost continuous subsidence can be
reach of meteoric water flow. Short distance transport regarded as records of a natural compaction experi-
by diffusion may nevertheless be important. ment. We can use the log porosity and in the cored
Quartz cementation was often interpreted to occur intervals we have data from core plugs.
as events of relatively short duration (approximately At depths shallower than about 2 km (80–100 C)
<10 million years) that could start and stop late in a we can compare the log porosities with experimental
sandstone’s burial history (Robinson and Gluyas compaction of similar sands in the laboratory. There
1992). This would imply that the quartz cementation will then be a marked effect of overpressure reducing
was controlled by advective transport of silica in solu- the effective stress. Poorly sorted sands will lose most
tion and the source of silica from other reactions. Fluid of their porosity at relatively shallow depth but well
inclusion data, however, shows that quartz cementa- sorted sand may have 30–35% porosity at 2–3 km
tion occurs throughout the temperature range depth, which corresponds to experimental compaction
corresponding to the burial history. at about 20–30 MPa effective stress. This suggests that
Modelling of quartz precipitation (Walderhaug there is little creep over long geologic time.
1996, Olkers et al. 2000, Walderhaug et al. 2001) is Data from deep wells will always show a trend
based on the assumption that this is a continuous towards lower porosities with depth but there may
process controlled by the kinetics and therefore by also be intervals where this trend is reversed. This
temperature. This is the basis for the practical models does not mean that net porosity has been created by
widely adopted by the petroleum industry (e.g. Exem- diagenetic processes. Because of the very low solubil-
plar and Touchstone). ity of silica and even more so of aluminium in
The assumption that the quartz precipitation is the porewater it is very difficult to explain how minerals
rate limiting factor may not always be strictly true in in several metre thick sandstones can be dissolved
clean quartz arenites where silica sources (stylolites) without precipitation of other minerals in the same
are widely spaced, resulting in decreasing quartz sandstones. Each lithology will have a characteristic
cementation away from the stylolites (Walderhaug porosity depth trend (Fig. 4.16) and increases in poros-
and Bjørkum 2003). The modelling does, however, ity reflect variations in the primary composition.
require that the burial curve and the temperature as a As discussed above the rate of quartz cementation
function of geological time are known. It is also very can be modelled if the surface area available for quartz
sensitive to changes in the primary sediment composi- cementation and the time-temperature history during
tion, which strongly influence both the mechanical burial are known.
compaction and the chemical reactions. The porosity At about 4.0 km burial depth (120–140 C) the
loss and increased sediment density resulting from amount of quartz cement may be 10–15% so the
chemical compaction as a function of temperature remaining porosity may be only 10–15%. We do how-
cause basin subsidence (Bjørkum et al. 1998, 2001; ever find good reservoir quality (>20% porosity) at
Bjørkum and Nadeau 1998). Temperature-driven greater depths and higher temperatures but this is due
chemical compaction results in a volume reduction to grain coatings. Prediction of porosity at great depth
(shrinkage) and the strain is then independent of effec- therefore requires that the occurrence of coating of
tive stress. As a result, differential stresses in siliceous chlorite, illite, haematite or microquartz can be
rock will be relaxed by the compaction processes predicted. Such prediction of the primary sediment
during basin subsidence as long at the temperature composition must again be linked to facies and
exceeds about 80 C (Bjørlykke 2006). provenance.
3
Velocity (m/s) Bulk density (g/cm )
2000 3000 4000 1.5 40% ϕ 2.0 35% ϕ 2.5
1500 1500
Porosity (%)
2000 24 26 28 30 32 34 36 38 40 42 44
0
5
Metres below seafloor

10
15
20
2500
25
30 30
35
%
40
45
0.71 mm
3000 50
0.1 mm
Etive Formation Etive formation 15%

3500 3500
Fig. 4.16 Velocity/depth and density/depth trends for the Etive 2 km depth is mechanical and similar to experimental data
Fm (Brent Group) showing that a single lithology has a nearly inserted from Chuhan et al. (2002). At greater depth, compac-
linear trend with depth, based on Marcussen et al. (2009). The tion is mostly chemical and higher compared with mechanical
calculated porosities show that the compaction down to about compaction
Prospects at great depth called HTHP (High Tem- Reconstructions of facies and climate will provide a
perature, High Pressure) are expensive to drill and basis for predicting the degree of feldspar dissolution
involve high risks, but in many mature basins nearly and precipitation of pore-filling kaolinite. Biogenic
all the shallower prospects have been drilled already. components like calcareous and siliceous organisms
An extensive study of reservoir sandstones in the will control the distribution of carbonate cements and
Gulf of Mexico showed that temperature and time are opal A, which will be altered to opal CT and quartz.
the main factors controlling reservoir quality (Nadeau Primary aragonite may cause extensive calcite cemen-
et al. 2008). tation that occludes much of the primary porosity.
Organic silica, e.g. from siliceous sponges, may
serve as a precursor to grain-coating microquartz pre-
4.17 Practical Prediction of Reservoir serving porosity at depth.
Quality Porosity loss due to mechanical compaction can
vary greatly as a function of textural and mineralogical
The most important factor controlling reservoir quality composition. Experimental compaction of loose sands
at depth is the primary composition of the sandstones. with different grain size and sorting provides a good
Sedimentological and sequence stratigraphic analyses basis for prediction of porosity and inter-granular vol-
are normally used primarily to predict reservoir geom- ume before quartz cementation. In cold sedimentary
etry, but for diagenetic processes grain size, sorting basins (low geothermal gradient) sand may be buried
and mineralogical composition are more critical. It is to 4–5 km before there is significant quartz cementa-
imperative to establish changes in provenance since tion and in the absence of overpressure it can be
the primary mineralogy places important constraints subjected to 40–50 MPa effective stress.
on diagenetic reactions at depth.
Well log data from distinct lithologies buried to CumulaƟve volumes

different depths may also provide a useful database Oil Gas
0% 100% 0% 100%
for predicting the porosity loss due to mechanical
compaction.
60°C
Temperature
4.18 Burial Diagenesis and Reservoir
120°C
Quality
Porosity loss due to quartz cementation can be

180°C
modelled as a function of temperature, time and sur-
face area available for quartz cementation. This is
sensitive to grain size and grain coatings, which must
be predicted from primary facies evaluation. The
source of silica for quartz cementation may be unsta- Fig. 4.17 Global cumulative percent reserves as a function of
ble silica minerals like opal A, opal CT, grain contact present reservoir temperature for (a) oil, (b), gas,. From Nadeau
dissolution or clay mineral reactions derived locally. et al. 2005. Approximately 85% of the oil reserves occur
between 60 and 120 C, and significant volumes of gas are
The presence of pore-filling illite depends on found shallower than these temperatures. Many of these
precursors which may be smectite or kaolinite. Disso- resevoirs have however been buried more deeply to higher
lution of kaolinite and precipitation of illite requires temperatures.
temperatures above 130 C and the local presence of
K-feldspar. Sandstone containing mostly plagioclase 4.19 Relationships Between
does not develop pore-filling illite as there is insuffi- Depositional Environment,
cient supply of potassium. Prediction of reservoir Diagenesis and Rock Properties
quality can thus be based on provenance.
The examples of reservoir quality predictions listed Changes in the physical properties of sandstones from
above are based on the assumption that burial diage- deposition to deep burial and uplift depends on the
netic reactions are essentially isochemical. Open sys- primary sediment composition with respect to textural
tem diagenesis allowing large scale import and export and mineralogical composition. It is therefore impor-
of solids in solution violates constraints from mineral tant to study diagenesis as an integrated part of basin
solubilities and fluid flow rates and therefore provides analysis. The source (provenance) of the sediments
a poor basis for prediction. determines the starting composition with respect to
Statistical analyses show that the porosity is lost the size of quartz grains from weathering and erosion
as a function of depth and that below 4.0-4.5 km of both igneous rocks and from older sedimentary
burial depth the reservoir quality is in most cases rocks.
insufficient for economic production. There are how- The feldspar content, and the composition of
ever exceptions where relative good reservoir quality feldspars, control to a large extent diagenetic reactions
is preserved despite burial to great depth (5-6 km). in sandstones at greater depth. We have seen above
This may be due to low geothermal gradients and high that sandstones rich in plagioclase will react very
burial rates. Clay coating is also an important factor differently from those rich in K-feldspar. Changes in
retarding quartz cementation and porosity loss. High mineral composition are also very important to recon-
temperatures will also increase the degree of cementa- struct drainage and transport in relation to facies in
tion in the shales and the permeabilities of seals may sedimentary basins. It is unfortunate that many sedi-
become so low that the pore pressure may reach frac- mentological descriptions include very little about
ture pressure (Nadeau et al. 2005). See Fig. 4.17. the mineralogical and textural composition of
sedimentary sequences from outcrops or cores. This analysis which is used for basin modelling (Bjørlykke
information is critical for the prediction of reservoir 2014).
quality at greater depth.
Prediction of mechanical compaction of sediments
must be based on experimental compaction of 4.20 How Much Oil Can Be Produced
sediments with different mineralogical and textural from Sandstone Reservoirs?
composition. Grain size, shape and sorting both in
sandstones and mudstones play an important role in Petroleum exploration requires predictions about the
determining the mechanical compaction prior to reservoir properties ahead of drilling to justify the
chemical compaction. Interpretations of provenance investment that a well represents. The reservoir
and depositional environment should be used as properties determine the percentage of recoverable
input for prediction of mechanical compaction based petroleum in each volume of rock. Even after drilling
on experimental compaction curves, thus linking dia- several exploration wells and also production wells in
genesis to more general basin analysis. the development of an oil or gas field, the porosities
At shallow depth sediments may change their bulk and permeabilities are only known from the cores.
composition due to meteoric water flushing causing Data from cores, cuttings and well logs must be
leaching (dissolution) or evaporation (precipitation). extrapolated to produce a 3D model of the large
Below the reach of significant meteoric water flow the volumes of rock between the wells. This must be
porewater flow is very restricted with respect to trans- based on predictions from facies distribution, distribu-
port of solids in solution. tion of faults and fault properties. Changes in reservoir
Prediction of changes in rock properties due to properties as a function of depth require diagenetic
chemical compaction is based on thermodynamics models which can predict changes in porosity as a
and kinetics and the initial mineral composition is function of effective stress and temperature/time.
then critical. Fluid flow and transport of solids in To calculate the producible oil in a prospect the
sedimentary basins can be constrained by modelling total rock volume (Gross – G) in the structure must be
sediment compaction and also by observations of estimated as well as the percentage of sandstone which
compositional stratification of the porewater. Burial can be produced (Net Sand – N).
diagenetic reactions are nearly isochemical, and min- The oil in place (Vp) is:
eral reactions can therefore be written as balanced
equations. The burial diagenetic reactions are then Vr N=G φ Osat
functions of the primary sediment composition which
can then be related to depositional environments and Here Vr is the volume of oil between the oil/water
provenance. contact (OWC) and the reservoir cap rock or the gas/
Significant increased porosity (secondary porosity) oil contact. N/G (net/gross) is the ratio between the
is dependent on the dissolution and removal of solids fraction of the reservoir rock that can be produced and
in solution which may occur during freshwater the total volume of the reservoir rock. φ is the average
flushing. Below the reach of freshwater the pore- porosity of the producible part of the reservoir (net
water flow is limited and a geochemically closed sys- volume). Osat is the average saturation of oil; typically
tem cannot produce net increases in secondary about 80–85% of the pores in sandstone are filled with
porosity. oil. The remaining portion is water, which in
Prediction of rock properties such as porosity and siliciclastic rocks occupies the mineral surfaces and
velocity at a certain depth in a sedimentary basin must the smallest pores where the capillary entry pressure is
be based on the burial history (effective stress and too high for oil.
temperature), but the primary mineralogical and tex- If the porosity is low the permeability will in most
tural composition of the sediments is equally impor- cases also be very low so that the flow of oil from the
tant. Sedimentological studies of depositional rock formation to the well becomes too slow to be
environments and provenance should therefore be economical. About 10% porosity may be the minimum
integrated with diagenesis and be a part of basin porosity for defining the producible (net) part of the
reservoir. In fractured reservoirs the permeability may overburden causing reduced porosity (volume) at
be high even when the porosity is below 10%. temperatures below 70–80 C. At greater depth (higher
In the planning of production from an oil field, data temperatures) compaction is mostly chemical and
from cores and logs from wells must be extrapolated mineral reactions are controlled by thermodynamics
into a 3D model of the flow properties of the reservoir. and kinetics. Because of low kinetic reaction rates
The producible percentage of the oil in place is (high activation energies) silicate reactions are very
called the recovery factor, which may range from sensitive to temperature. The precipitation of quartz is
20–30% up to 40–60%. In sandstones with good res- slow and has high activation energy, and temperature
ervoir quality, improved production technology has in is the main control on quartz cementation causing
some cases (e.g. Statfjord and Gullfaks fields, offshore much of the porosity loss in well-sorted sandstones.
Norway) boosted the recovery factor to close on 70%. The dissolution of K-feldspar and kaolinite at about
Recovery is limited both by the amount of oil 130 C occur because the mineral assemblage illite and
remaining in the pores of the drained sandstones and quartz is more stable.
the presence of undrained compartments within the Because of the low solubility of silicate minerals
reservoir where oil is bypassed. Reservoir quality is a and the limited flow of porewater in the deeper parts of
very important factor in the financial risk assessment sedimentary basins, burial diagenetic reactions must
calculations for a prospect. be nearly isochemical. Significant amounts of solids
can not be exported from a sandstone and the porosity
of a single reservoir rock will only decrease and can
4.21 Conclusions not increase during progressive burial.
Prediction of reservoir quality at great burial depth
Diagenetic reactions are driven towards higher depends on the initial sediment composition (prove-
mechanical and chemical stability. Reactions in nance), sedimentary facies (Fig. 4.18), early diage-
sandstones are driven by the effective stress from the netic processes and the subsequent burial history.
Sediment input and burial diagenesis
Biogenic Detrital supply.

carbonate and Met. water
silica flushing
ks Basement
roc
Basin fill e d.
rs
de
Ol
Dissolution of feldspar
and mica, precipitation
Burial Carbonate cement Verdine(Fe) of authigenic kaolinite
depth Opal A-CT - quartz facies 2KAlSi3O8 + 2H+ + 9H2O =
0-3.5(4) Little aut. kaolinite Al2Si2O5(OH)4 + 4H4SiO4 + 2K+
km
Burial Extensive Chlorite Quartz cement,
depth > 3.5(4) km quartz cementation coatings? illitisation if kaolinite
Little illite if Little and
kaolinite and quartz K-feldspar is present
smectite are absent. cement
KAlSi3O8 + Al2Si2O5(OH)4 =
KAl3Si3O10(OH)2 + 2SiO2 + 2H2O
Fig. 4.18 Summary diagram for clastic diagenesis. The pri- controlled by the depositional environment. Burial diagenesis is
mary composition of the sand depends on the provenance, controlled by mineral stability (thermodynamics) and reaction
weathering, transport and depositional environment. Early dia- rates (kinetics)
genesis including meteoric flushing or marine cementation, is
This should therefore be linked to facies models and principal aspects of diagenetic processes in sedimentary
provenance in addition to basin subsidence and heat basins. Sedimentary Geology 301, 1-14.
Bjørlykke, K., Nedkvitne, T., Ramm, M. and Saigal, G. 1992.
flow. Modelling of quartz cementation has proved to Diagenetic processes in the Brent Group (Middle Jurassic)
be very useful for the prediction of porosity as a reservoirs of the North Sea – An overview. In: Morton, A.C.,
function of the temperature history of the reservoir Haszeldine, R.S., Giles, M.R. and Brown, S. (eds.), Geology
sandstones, but the results are sensitive to the primary of the Brent Group. Geological Society Special Publication
61, pp. 263–287.
textural and mineralogical composition of the Bjørlykke, K., Aagaard, P., Egeberg, P.K. and Simmons, S.P.
sandstones and the presence of grain coatings. 1995. Geochemical constraints from formation water analyses
from the North Sea and Gulf Coast Basin on quartz, feldspar
and illite precipitation in reservoir rocks. In: Cubitt, J.M. and
England, W.A. (eds.), The Geochemistry of Reservoirs.
Further Reading Geological Society Special Publication 86, pp. 33–50.
Bjørlykke, K., Chuhan, F., Kjeldstad, A., Gundersen, E.,
Lauvrak, O. and Høeg, K. 2004. Modelling of sediment
Aagaard, P., Egeberg, P.K., Saigal, G.C., Morad, S. and compaction during burial in sedimentary basins. In:
Bjorlykke, K. 1990. Diagenetic albitization of detrital K- Stephansson, O., Hudson, J. and King, L. (eds.), Coupled
feldspars in Jurassic, Lower Cretaceous and Tertiary clastic Thermo-Hydro-Mechanical-Chemical Processes in Geo-
reservoir rocks from Offshore Norway, II. Formation water Systems. Elsevier, Amsterdam, pp. 699–708.
chemistry and kinetic considerations. Journal of Sedimen- Bjørlykke, K. and Egeberg, P.K. 1993. Quartz cementation in
tary Petrology 60(4), 575–581. sedimentary basins. AAPG Bulletin 77, 1538–1548.
Aase, N.E., Bjørkum, P.A. and Nadeau, P.H. 1996. The effect of Bjørlykke, K. and Jahren, J. 2012. Open or closed geochemical
grain coating microquartz on preservation of porosity. systems during diagensis in sedimentary basins: Constraints
AAPG Bulletin 80, 1654–1673. on mass transfer during diagenesis and the prediction of
Ajdukiewicz, J.M., Nicholson, P.H. and Esch, W.L. 2010. Pre- porosity in sandstone and carbonate reservoirs. AAPG
diction of deep reservoir quality using early diagenetic pro- Bulletin 96, 2193–2214.
cess models in the Jurassic Norphlet Formation, Gulf of Bjørlykke, K., Ramm, M. and Saigal, G.C. 1989. Sandstone
Mexico. AAPG Bulletin 94, 1189-1227. diagenesis and porosity modification during basin evolution.
Barth, T. and Bjørlykke, K. 1993. Organic acids from source Geologische Rundschau 78, 243–268.
rock maturation; generation potentials, transport Bjørlykke, K., Mo, A. and Palm, E. 1988. Modelling of thermal
mechanisms and relevance for mineral diagenesis. Applied convection in sedimentary basins and its relevance to diage-
Geochemistry 8, 325–337. netic reactions. Marine and Petroleum Geology 5, 338–351.
Boles, J.R. 1982. Active albitization of plagioclase, Gulf Coast Bjørlykke, K. and Aagaard, P. 1992. Clay minerals in North Sea
Tertiary. American Journal of Science 282, 165–180. sandstones. In: Houseknecht, D.W. and Pittman, E.D. (eds.),
Boles, J.R. 1998. Carbonate cementation in Tertiary sandstones Origin, Diagenesis, and Petrophysics of Clay Minerals in
of the San Joaquin Basin. In: Morad, S. (ed.), Carbonate Sandstones. SEPM, Special Publication 47, Tulsa, OK, pp.
Cementation in Sandstones. International Association of 65–80.
Sedimentology, Heidelberg, Special Publication 26, Burley, S.D. 1986. The development and destruction of porosity
pp. 261–284. in the Upper Jurassic reservoir sandstones of the Piper and
Bjørkum, P.A., Mjøs, R., Walderhaug, O. and Hurst, A. 1990. the Tartan fields, outer Moray Firth, North Sea. Clay
The role of the late Cimmerian unconformity for the distri- Minerals 21, 649–694.
bution of kaolinite in the Gullfaks Field, northern North Sea. Chuhan, F.A., Kjeldstad, A., Bjørlykke, K. and Høeg, K. 2002.
Sedimentology 37, 395-406. Porosity loss in sand by grain crushing. Experimental evi-
Bjørkum, P.A. and Nadeau, P.H. 1998. Temperature controlled dence and relevance to reservoir quality. Marine and Petro-
porosity/permeability reduction, fluid migration and petroleum Geology 19, 39–53.
leum exploration in sedimentary basins. APPEA Journal 38, Chuhan, F.A., Kjeldstad, A., Bjørlykke, K. and Høeg, K. 2003.
453-464. Experimental compression of loose sands: Relevance to
Bjørkum, P.A., Oelkers, E.H., Nadeau, P.H., Walderhaug, O. porosity reduction during burial in sedimentary basins.
and Murphy, W.M. 1998. Porosity prediction in quartzose Canadian Geotechnical Journal 40, 995-1011.
sandstones as a function of time, temperature, depth, Day-Stirrat, R.J., Milliken, K.L., Dutton, S.P., Loucks, R.G.,
stylolite frequency, and hydrocarbon saturation. AAPG Bul- Hillier, S., Aplin, A.C. and Schleicher, A.M. 2010. Open-
letin 82, 637–648. system chemical behavior in deep Wilcox Group mudstones,
Bjørkum, P.A., Walderhaug, O. and Nadeau, P.H. 2001. Ther- Texas Gulf Coast, USA. Marine and Petroleum Geology 27,
mally driven porosity reduction: Impact on basin subsidence. 1804-1818.
In: Shannon, P.M., Houghton, P.D.W. and Corcoran, D.V. Ehrenberg, S.N. 1990. Relationship between diagenesis and
(eds.), The Petroleum Exploration of Ireland’s Offshore reservoir quality in sandstones of the Garn Formation,
Basins. Geological Society Special Publication 188, pp. Haltenbanken, mid-Norwegian continental shelf. AAPG
385–392. Bulletin 74, 1538–1558.
Bjørlykke, K. 2014. Relationships between depositional Ehrenberg, S.N. 1997. Influence of depositional sand quality and
environments, burial history and rock properties. Some diagenesis on porosity and permeability: Examples from
Brent Group reservoirs, northern North Sea. Journal of Sedi- Lander, R.H., Larese, R.E. and Bonell, L.M. 2008. Toward more
mentary Research 67, 197–211. accurate quartz cement models; The importance of euhedral
Ehrenberg, S.N. and Nadeau, P.H. 2005. Sandstone versus car- versus noneuhedral growth rates. AAPG Bulletin 92,
bonate petroleum reservoirs: A global perspective on 1537–1563.
porosity-depth and porosity-permeability relationships. Lichner, P.C., Steefel, C.I. and Oelkers, E.H. 1996. Reactive
AAPG Bulletin 89, 435–445. transport in porous media. Reviews in Mineralogy 34, 438.
Ehrenberg, S.N., Nadeau, P.H. and Steen, Ø. 2008. A megascale Mineralogical Society of America.
view of reservoir quality in producing sandstones from the Marcussen, Ø., Maast, T.E., Mondol, N.H., Jahren, J. and
offshore Gulf of Mexico. AAPG Bulletin 92, 145–164. Bjørlykke, K. 2009. Changes in physical properties of a
Ehrenberg, S.N., Aqrawi, A.A. and Nadeau, P.H. 2008. An reservoir sandstone as a function of burial depth - The
overview of reservoir quality in producing Cretaceous strata Etive Formation, Northern North Sea. Marine and Petroleum
of the Middle East. Petroleum Geoscience 14, 307–318. Geology 27(8), 1737–1751.
Ehrenberg, S.N., Walderhaug O. and Bjørlykke, K. 2012. Car- Nadeau, P.H., Bjørkum, P.A. and Walderhaug, O. 2005. Petro-
bonate porosity creation by mesogenetic dissolution: Reality leum system analysis: Impact of shale diagenesis on reser-
or illusion? AAPG Bulletin 96, 217–233. voir fluid pressure, hydrocarbon migration, and
Fawad, M., Mondol, N. H., Jahren, J. and Bjørlykke, K. 2010. biodegradation risks. In: Doré, A.G. and Vining, B.A.
Microfabric and rock properties of experimentally com- (eds.), Petroleum Geology: North-west Europe and Global
pressed silt-clay mixtures. Marine and Petroleum Geology Perspectives: Proceedings of the 6th Petroleum Geology
27, 1698–1712. Conference, Geological Society, pp. 1267–1274,
Giles, M.R. 1997. Diagenesis; A Quantitative Perspective. Doi 10.1144/061267.
Kluwer Academic Publ., Dordrecht, 526 pp. Oelkers, E.H., Bjorkum, P.A., Walderhaug, O, Nadeau, P.H. and
Giles, M.R., Stevenson, S., Martin, S.V., Cannon, S.J.C., Murphy, W.M. 2000. Making diagenesis obey thermody-
Hamilton, J.D., Marshall, J.D. and Samways, G.M. 1992. namics and kinetics: The case of quartz cementation in
The reservoir properties of the Brent Group. A regional sandstones from offshore mid-Norway. Applied Geochemis-
perspective. In: Morton, A.C., Haszeldine, R.S., Giles, M. try 15, 295–309.
R. and Brown, S. (eds.), Geology of the Brent Group. Ortoleva, P., Al Shaieb, Z. and Puckette, J. 1995. Genesis and
Geological Society Special Publication 61, pp. 289–327. dynamics of basin compartments and seals. American Jour-
Hamilton, P.J., Kelley, S. and Fallick, A.E. 1989. K-Ar dating nal of Science 295, 345–427.
of illite in hydrocarbon reservoirs. Clay Minerals 14, Pittmann, E.D. and Larese, R.E. 1992. Compaction of lithic
215–231. sands: Experimental results and applications. AAPG Bulle-
Heald, M.T. and Larese, R.E. 1974. Influence of coatings on tin 75, 1279–1299.
quartz cementation. Journal of Sedimentary Petrology 44, Ramm, M. 1992. Porosity-depth trends in reservoir sandstones:
1269–1274. Theoretical models related to Jurassic sandstones offshore
Hesthammer, J., Bjørkum, P.A. and Watts, L. 2002. The effect Norway. Marine and Petroleum Geology 33, 396–409.
of temperature on sealing capacity of faults in sandstone Ramm, M., Forsberg, A.W. and Jahren, J. 1998. Porosity-depth
reservoirs – Examples from the Gullfaks and Gullfaks Sør trends in deeply buried Upper Jurassic reservoirs in the
Fields, North Sea. AAPG Bulletin 86(10), 1733–1751. Norwegian Central Graben: An example of porosity preser-
Hoveland, M., Bjørkum, P.A., Gudemestad, O.T. and Organge, vation beneath normal economic basement by grain-coating
D. 2001. Gas hydrate and seeps – Effects on slope stability: micro-quartz. In: Kupecz, J.A., Gluyas, J. and Bloch, S.
The “hydraulic model”. ISOPE Conf. Proceedings, (eds.), Reservoir Quality Prediction in Sandstones and
Stavanger, 471–476, ISOPE (International Society for Off- Carbonates: AAPG Memoir 69, pp. 177–199.
shore and Polar Engineering), New York. Saigal, G.C., Bjørlykke, K. and Larter, S.R. 1992. The effects of
Karlsen, D.A., Nedkvitne, T., Larter, S.R. and Bjørlykke, K. oil emplacements on diagenetic processes – Examples from
1993. Hydrocarbon composition of authigenic inclusions – the Fulmar reservoir sandstones, central North Sea. AAPG
application to elucidation of petroleum reservoir filling his- Bulletin 76, 1024–1033.
tory. Geochemica et Cosmochemica Acta 57, 3641–3659. Storvoll, V., Bjørlykke, K., Karlsen, D. and Saigal, G. 2002.
Land, L.S. 1997. Mass transfer during burial diagenesis in the Porosity preservation in reservoir sandstones due to grain-
Gulf of Mexico sedimentary basin: An overview. In: coating illite: A study of the Jurassic Garn Formation from
Montanez, I.P., Gregg, J.M. and Shelton, K.L. (eds.), the Kristin and Lavrans Fields offshore Mid-Norway.
Basin-wide Diagenetic Patterns: Integrated Petrologic, Marine and Petroleum Geology 19, 767–781.
Geochemical, and Hydrologic Considerations. SEPM Spe- Thyberg, B. and Jahren, J. 2011. Quartz cementation in
cial Publication 57, pp. 29-39. mudstones and sheet-like quartz cement from clay mineral
Land, L.S. and Milliken, K.L. 2000. Regional loss of SiO2 and reactions during burial. Petroleum Geoscience 17, 53–63.
CaCO3, and gain of K2O during burial diagenesis of Gulf Thyberg, B., Jahren, J., Winje, T., Bjørlykke, K., Faleide, J.I. and
Coast mudrocks, U.S.A. In: Worden, R.H. and Morad, S. Marcussen, Ø. 2010. Quartz cementation in Late Cretaceous
(eds.), Quartz Cementation in Sandstones: International mudstones, northern North Sea: Changes in rock properties
Association of Sedimentologists Special Publication 29, pp. due to dissolution of smectite and precipitation of micro-quartz
183-197. crystals. Marine and Petroleum Geology 27, 1752–1764.
Thyne, G., Bjørlykke, K. and Harrison, W. 1996. Chemical reac- geometry. International Association of Sedimentologists,
tion and solute transport rates as constraints on diagenetic Special Publication 26, pp. 179–192.
models. Annual Meeting Abstract. AAPG and SEPM 5, 140. Walderhaug, O. and Bjørkum, P.A. 2003. The effect of stylolite
Walderhaug, O. 1990. A fluid inclusion study of quartz-cemented spacing on quartz cementation in the Lower Jurassic Stø
sandstones from offshore mid-Norway: Possible evidence for Formation in well 7120/6–1, southern Barents Sea. Journal
continued quartz cementation during oil emplacement. Jour- of Sedimentary Research 73, 146–156.
nal of Sedimentary Petrology 60, 203–210. Walderhaug, O., Bjørkum, P.A., Nadeau, P. and Langnes, O.
Walderhaug, O. 1994a. Temperatures of quartz cementation in 2001. Quantitative modelling of basin subsidence caused by
Jurassic sandstones from the Norwegian Continental shelf – temperature-driven silica dissolution and reprecipitation.
Evidence from fluid inclusions. Journal of Sedimentary Petroleum Geoscience 7, 107–113.
Research 64A, 311–323. Walderhaug, O., Oelkers, E.H. and Bjørkum, P.A. 2004. An
Walderhaug, O. 1994b. Precipitation rates for quartz cement in analysis of the roles of stress, temperature, and pH in chemi-
sandstones determined by fluid inclusion microthermometry cal compaction of sandstones – Discussion. Journal of Sedi-
and temperature history modelling. Journal of Sedimentary mentary Research 74, 447–450.
Research 64A, 324–333. Wilson, M.D. 1994. Reservoir Quality Assessment and Predic-
Walderhaug, O. 1996. Kinetic modeling of quartz cementation tion in Clastic Rocks. SEPM Short Course 30, 432 pp.
and porosity loss in deeply buried sandstone reservoirs. Wood, J.R. 1986. Thermal transfer in systems containing quartz
AAPG Bulletin 80, 731–745. and calcite. In: Gautier, D.L. (ed.), Roles of Organic Matter
Walderhaug, O. and Bjørkum, P.A. 1998. Calcite cement in in Sediment Diagenesis. SEPM Special Publication 38,
shallow marine sandstones: Growth mechanisms and Tulsa, pp. 181–189.
Chapter 5
Carbonate Sediments
Nils-Martin Hanken, Knut Bjørlykke and Jesper Kresten Nielsen
Carbonate sediments are a part of the carbon cycle Carbonate sediments are for the most part formed
(Fig. 1.14). CO2 in the atmosphere dissolves in water (born) within sedimentary basins even if there may be
and makes carbonic acid (H2CO3) which reacts with a clastic supply of carbonate sediments. The sedimen-
Ca2+ or Mg2+ to precipitate CaCO3 or MgCO3. This tology of carbonates therefore differs in many respects
process is an important sink for CO2. The rate of from siliceous sand and mud. Carbonates may precipi-
carbonate sedimentation globally is controlled by the tate chemically from the seawater, but most of the
supply of cations (mostly Ca2+ and Mg2+) into the limestones are composed of calcareous organisms.
ocean from rivers. This again is a function of the rate The properties of limestones are therefore closely
of weathering of Ca-bearing silicate minerals like pla- linked to their biological origin and the mineralogy
gioclase, pyroxene, amphibole and chlorite. Dissolu- of the carbonate skeletons. The primary composition is
tion of basic rock along a spreading ridge may also add very important for their alteration during burial (dia-
to the supply of calcium to the ocean water. genesis) and consequent reservoir properties (porosity
Weathering of limestones will release CO2 and and permeability), which to a large extent is controlled
therefore does not contribute as a sink for CO2. During by chemical processes. Since both biologically and
contact metamorphism when limestones are heated (to chemically precipitated carbonate sediments are com-
550–600 C), CaCO3 reacts with SiO2 to form CaSiO3 posed of minerals we will first examine the mineralogy
(wollastonite) and CO2. Large amounts of CO2 formed and geochemistry of carbonates. This is important for
this way are released through volcanoes. When understanding diagenetic reactions and prediction of
minerals like wollastonite are uplifted, exposed and reservoir quality.
weathered, Ca2+ is released into rivers and the ocean
again. Another part of the CO2 in the atmosphere is
reduced by plants (i.e. algae) and stored as carbon or
organic compounds which can generate petroleum. 5.1 Geochemistry of Carbonate
These may also be oxidised to CO2. Minerals
Carbonate minerals consist of CO23 and one or more

N.-M. Hanken (*) cations. The most common cations in carbonate
Department of Geology, UiT - The Arctic University of minerals together with their mineral names are listed
Norway, Tromsø, Norway
in Table 5.1. The common rock-forming carbonate
minerals are either rhombohedral (calcite) or ortho-
K. Bjørlykke
Department of Geosciences, University of Oslo, Oslo, Norway
rhombic (aragonite) in crystal habit. Where cations
e-mail: [email protected] with small ionic radii are incorporated into carbonates
J.K. Nielsen
a trigonal (rhombohedral) crystal lattice is formed,
North Energy ASA, New Area Exploration, Alta, Norway while larger cations result in orthorhombic unit cells.
152 N.-M. Hanken et al.
Table 5.1 Mineralogy of the most common carbonate this aragonite crystal structure are strontianite
minerals (SrCO3), witherite (BaCO3) and cerrusite (PbCO3).
Carbonate sediments formed in normal marine environments Sr in particular is an important trace element in arago-
consist of three main minerals: nite. Aragonite crystals forming in marine
Low-Mg calcite CaCO3 ð<4% MgCO3 Þ (hexagonal) environments today contain 5,000–10,000 ppm Sr.
High-Mg calcite ðCa,Mg) CO3 ð>4% MgCO3 Þ (hexagonal) Aragonite is unstable and after some time will be
Aragonite (CaCO3) (orthorhombic) replaced by calcite which still retains relatively high
Other common carbonate minerals are:
concentrations of strontium. Aragonite may occasion-
Siderite FeCO3
Magnesite MgCO3 ally be preserved, particularly in dense shales, even in
Strontianite SrCO3 Mesozoic rocks.
Rhodochrosite MnCO3 Iron is only very weakly soluble in the oxidised
Smithsonite ZnCO3 state, forming hydroxides Fe(OH)3 and oxides
Ankerite Ca(Mg,Fe)(CO3 Þ2
(Fe2O3), but in the reduced state it occurs as soluble
Dolomite CaMg(CO3 Þ2
Fe2+. Reduction of iron normally takes place within
the microbial sulphate reduction zone where high
concentrations of sulphur will cause available Fe2+ to
Ca2+ has an ionic radius close to 1 Å which is interme- be precipitated as sulphides (pyrite, FeS2), so that very
diate between small and large cations and near the little is available to enter the calcite structure. The
limit of sixfold co-ordination. Thus CaCO3 is dimor- principal environment in which Fe2+ can enter the
phous forming either rhombohedral or orthorhombic calcite lattice to form ferroan calcite is thus in the
structures. reducing porewater below the sulphate reduction
Cations smaller than 1 Å such as Fe2+, Mn2+, Zn2+ zone. The ferroan calcite may contain a few thousand
and Mg2+ can be incorporated in the calcite lattice. ppm of iron.
These metals all have an ionic radius of about Dolomite (CaMg(CO3 Þ2 Þ is a carbonate mineral in
0.6–0.7 Å, and therefore calcite can contain consider- which layers of CaCO3 alternate with layers of
able concentrations of these cations. The calcite group
MgCO3 : Fe2þ is commonly found substituting for
of minerals, such as siderite (FeCO3), rhodochrosite
Mg2+ in dolomite, and a complete series extends to
(MnCO3), smithsonite (ZnCO3) and magnesite
ankerite (Ca(Fe,Mg)(CO3 Þ2 Þ. Dolomite formed early
(MgCO3), all have the same crystal structure as calcite.
in diagenesis is fine-grained and can often have a
Two types of calcite are recognised, depending
magnesium deficit in relation to calcium [e.g.
on the magnesium content: low-Mg calcite
Ca55 Mg45 ðCO3 Þ100 ]. This is called protodolomite
ð<4 mol% MgCO3 Þ and high-Mg calcite
which during burial may be transformed into a regular
ð>4 mol% MgCO3 Þ. Much of the biologically secreted
dolomite.
calcite typically ranges between 11 and 19 mol%
MgCO3. Some organisms like coccoliths are however
composed of low-Mg calcite (or simply called calcite).
Low-Mg calcite is more stable than high-Mg calcite, 5.2 Carbonate – CO2 Systems in the Sea
and fossil fragments originally composed of high-Mg
calcite are converted to low-Mg calcite during dia- Even if most carbonate precipitation occurs biologi-
genesis. The alteration of high-Mg calcite to low-Mg cally it is important to understand the chemical
calcite takes place by a process of leaching of Mg2+ constraints on carbonate reactions. Carbon dioxide
ions, which leaves the microarchitecture of the grain concentration is the factor which has the greatest
unaffected. The exsolved Mg2+ may form micro- influence on pH and the solubility of carbonates in
dolomite rhombs that are sometimes seen as inclusions water. CO2 dissolves in water to form carbonic acid
in calcitised high-Mg calcites (quite common in (H2CO3), which dissociates into bicarbonate ðHCO3 Þ
fragments of echinoderms and calcareous red algae). and carbonate ions ðCO23 Þ.
The orthorhombic lattice has an arrangement of
CO23 anions where cations larger than 1 Å (such as CO2 þ H2 O ¼ H2 CO3 ¼ H+ þ HCO3
Sr2+, Ba2+ and Pb2+) are preferred. Analogous with ¼ CO2 þ 2H+
3
5 Carbonate Sediments 153
Calcite solubility depends on the solubility product surface layer at low latitudes, and lowest in polar areas
Ca2þ CO23 . We see that when the water is rela-

(7.5–8.0).
tively acid (high H+ concentrations) the reaction is Photosynthesis also contributes to the consumption
driven towards the left and the carbonate ions concen- of CO2, increasing the pH of the surface water. pH
tration CO23 will be low. With high pH (i.e. low H+
decreases with water depth, not only because CO2 is
no longer removed by photosynthesis, but also
activity), the bicarbonate HCO3 concentration will
because of the lower temperature and higher pressure.
be higher.
Whereas photosynthesis involves the removal of
Mineral solubility is not strictly a function of the
CO2 from the water, respiration adds CO2.
ionic concentration, but of the activity (a) which is
influenced by the temperature and other ions present.
H2 O þ CO2 ¼ CH2 O þ O2
The dissociation constants for H2CO3 and HCO3
are:
CH2O is a general formula for sugar. During pho-
tosynthesis this reaction will go to the right. The
K1 ¼ aHþ aHCO3 =aH2 CO22 reverse reaction is respiration. We see that while pho-
tosynthesis raises the pH, respiration will lower it. The
and water below the photic zone will gain CO2 from the
respiration of zooplankton, and the breakdown (i.e.
K2 ¼ aHþ aCO23 =aHCO3 oxidation) of organic matter which sinks down
through the water column will also produce CO2 and
This in turn requires that aCO23 also must be lower the pH. In shallow water, a daily variation in pH
higher in order to satisfy the dissociation equation. has been registered as a result of the fact that photo-
At low pH values the equilibrium will shift to the synthesis takes place only during the day, increasing
left, giving more free CO2 and H2CO3 (Fig. 5.1). the pH, while respiration continues at night reducing
CO2 is found in both water and the atmosphere, and the pH. Respiration by organisms in the water below
is exchanged between them. Statistically, the resi- the photic zone contributes further to the pH declining
dence time in the atmosphere is c. 8 years, while it is downwards through the water column. The recent
c. 600–1,000 years in the ocean. The CO2 in the ocean increase in CO2 content in the atmosphere (from 280
is partly removed when precipitated as organic matter to 400 ppm) will also influence the ocean water,
in the sediments, and partly by precipitation as making it slightly more acidic. Ocean water is, how-
carbonates. ever, strongly buffered and the amount of carbonate
The solubility of CO2 in water is greatest at low that can be precipitated in the oceans is primarily
temperatures and high pressures, decreasing as the dependent on the supply of cations over geological
temperature rises and pressure decreases. Since it is time, especially Ca2+ which is mostly liberated by
largely the CO2 concentration which determines the land weathering of carbonate rocks and calcium
pH of water, the pH is highest (8.0–8.5) in the warm silicates such as plagioclase. The pH in modern
HCO3–
Solubility in
milliequivalents/L
1
CO32-
6.5 7.0 7.5 8.0 8.5 9.0 9.5 pH
Fig. 5.1 Solubility of carbonate ions in seawater as a function of pH

ocean water varies from 7.9 in zones with upwelling of no supply of terrigenous sediments such as sand or
deep water to about 8.3 in parts of the North Atlantic clay which otherwise would dilute the carbonate con-
and the Barents Sea where high productivity removes tent. Many of the carbonate-secreting organisms
carbon dioxide from the sea water. Increased CO2 in require clear water because they filter out nutrients
the atmosphere and in the ocean, however, shows large from the seawater. The presence of mud will kill
local variations due to ocean water circulation, organisms like corals, severely reducing carbonate
biological activity and reactions with carbonate production. This puts an important limitation on the
minerals. occurrence of carbonate sediments in sedimentary
basins.
5.3 Skeletal Components

5.4 Plants
5.3.1 Introduction
5.4.1 Stromatolites
The evolution of organisms precipitating carbonate
skeletons has played an important role in the accumu- “Stromatolite” is the term for lamination in carbonate
lation of carbonate sediments and their properties. The rocks due to the accumulation or precipitation of car-
skeletal material is widely different in size, with bonate ascribed to bacterial growth. Stromatolites
diameters ranging from a few micrometres in have an intermediate status between fossils and sedi-
coccolithophores to more than a metre in some mentary structures. They consist of cyanobacteria
bivalves and sponges. The range and taxonomic diver- (earlier called blue–green algae) which have a growth
sity of major groups of skeletal organisms are shown form as unsegmented, micron-sized filaments or
in Fig. 5.2. However, in this text book we have limited unicells (coccooids). The filaments occur in rows or
the descriptions to those groups which are most com- strands within a sticky mucilaginous sheath. Only a
monly encountered in carbonate hydrocarbon few forms produce a biochemically precipitated
reservoirs. skeleton, usually tubiform. Other cyanobacteria may
The first condition for obtaining relatively pure generate organic films on the sediment surface, which
carbonate deposits is that there must be very little or trap and bind lime mud to form irregular laminae.
Coccolithophores
Foraminifera
Foraminifera
Brachiopods
Green algae
Gastropods
Bryozoans
Echinoids
Trilobites
Coralline
red algae
Bivalves
Crinoids
Corals
C C
M M
P P
Planktic Benthic
Dominant Important Minor sediment producers
Fig. 5.2 Diversity, abundance and relative importance of various calcareous marine organisms as sediment producers. P ¼
Palaeozoic; M ¼ Mesozoic; C ¼ Cenozoic (modified after Wilkinson 1979)
a b
Daylight
Upward growth (S. calcicola)
and sediment trapping
c
Darkness
Horizontal growth (O. submembranacea)
0.1 mm
and sediment binding
Fig. 5.3 Diagrammatic representation of the day-night accre- and around the sediment grains. (c) A sticky surface that traps
tion in stromatolites. (a) and (b) During daytime the and binds the next sediment layer is produced during the night
cyanobacteria trap and bind sediment and proceed to grow up (modified from Gebelein 1969)
Rhythmic variations in bacterial filament growth pro- the periphery. Laminae draping over the edge have
duce a laminated structure of alternating light, often accumulated more steeply than the angle of
sediment-rich laminae and dark, organic-rich laminae repose because of the sticky surface of the
(Fig. 5.3). The end result is either parallel lamination cyanobacteria. Micro-organisms in aquatic
following cyanobacterial (previously algal) mats, environments are usually in a matrix of extracellular
more complex bacterial growth structures polymeric substances, also known as exopolysac-
(stromatolites), or a concentric type of structure charides, or EPS (also called “biofilm”). The produc-
(oncoids). However, there are not always tion of EPS is a general property of prokaryotic
cyanobacterial remains to be found, and the only evi- (bacteria, archaea) and in eukaryotic (algae, fungi)
dence then is the lamination in the rock. micro-organisms. It is excreted by micro-organisms
If the laminae are flat-lying they are referred to as to keep them aggregated and attached to surfaces.
cyanobacterial laminated sediments, but if they form Cyanobacterial laminae may also form coatings on
structures with vertical relief they are called carbonate grains (e.g. skeletal material), producing
stromatolites. The different overall shapes of structures known as oncoids (Fig. 5.5). These grains
stromatolites range from: become spherical to oval as irregular, sometimes dis-
1. Laterally linked hemispheroids continuous, concentric bacterial laminations
2. Discrete, vertically stacked hemispheroids consisting of cyanobacteria growing around the
3. Discrete hemispheroids nucleus. Once again, the layer of sticky bacteria on
These main types are shown in Fig. 5.4 and Fig. 5.5, the surface of the grain will trap microscopic sediment
but combinations can occur. grains, or precipitate aragonite because CO2 is con-
The lamination commonly follows the outline of sumed by photosynthesis. In this way the grains will
the structure, or is terminated at the edge of individual “grow”. These “growth layers” will not be continuous
heads or stacks. Individual laminae are often thickest around the primary grain, as in oolites, because that
at the centre of the structure and thin laterally towards requires constant movement. Sediment grains coated
Discrete, vertically stacked hemispheroids
um
at Constant basal
tr
Su
b st radius
s
Sub
ra
tu
m
Su
bs Variable basal
tr radius
at um
um at
s tr
S ub
Fig. 5.4 Some of the main types of stromatolites (modified from Logan, Rezak and Ginsburg 1964)
a b Intertidal to supratidal stromatolites are often associated

with fenestral (birdseye) structures, mud cracks and
evaporites. The specific growth form may be used for
environmental interpretation. Laterally linked hemi-
spheroids often dominate on protected mud flats where
2 cm wave action is slight. Exposed, intertidal mud flats,
where the scouring action of waves and other interacting
Fig. 5.5 Recent oncoids from the tidal zone. (a) Top surface of
a hemisphaerical oncoid. (b) Vertical section showing the rough
factors prevent the growth of bacterial mats between
lamination. The flat base reflects stationary growth. Ras stromatolites, are characterised by vertically stacked
Muhamad Peninsula, Sinai hemispheroids. Concentrically arranged spheroids
(oncoids) may occur in low, intertidal areas that are
in this manner may reach up to 10 cm in diameter. exposed to waves, and in agitated water below the low-
Large oncoids are flattish and often called “bacterial tide mark. Oncoids are generally formed in shallow
(previously algal) biscuits”. A deposit of oncoid sand subtidal environments of low to moderate turbulence.
is called an oncolite.
5.4.1.2 Geological Range
5.4.1.1 Ecology Cyanobacteria range from the early Precambrian to
Cyanobacteria have inhabited mostly open marine, Recent. Stromatolites were very important in the
shallow water and intertidal environments throughout Precambrian, but gradually decreased during the
geological time. However, hypersaline and rare fresh- Phanerozoic due to the evolution of grazing
water forms have been recorded too. organisms, in particular gastropods. Today,
Stromatolites are most common in upper intertidal stromatolites are mostly present in stressed
to supratidal environments (Fig. 5.6), but have also environments where high evaporation and therefore
occasionally been described from subtidal environments. enhanced salinity limit grazing species.
a b
Fig. 5.6 (a) Supratidal crust composed mainly of fine-grained Cyanobacterial mat with gypsum from supratidal environment
dolomite. Bahamas (Photo G.M. Friedman). (b) Supratidal envi- (sabkha), Abu Dhabi (Photo D. Kinsmann)
ronment (sabkha), Abu Dhabi. Photo D. Kinsmann. (c)
5.4.1.3 Significance for Petroleum Geology coccolithophores, are planktonic. Most important are
Some unicellular cyanobacteria synthesise oil and red algae and green algae, but one species of brown
their remains may accumulate to form sapropelic algae (Palina) also secretes a calcareous skeleton,
deposits. These deposits may be important source which consists of aragonite.
rocks for oil. Oncolites may possess primary interpar- A wide range of growth forms exists, and the most
ticle pores, but high porosity is not as common as in important are:
oolites because of poor sorting in less turbulent water. • Encrusting calcareous algae
Oncolites also form thinner and less persistent beds These are usually red algae and form smooth or
than oolites and have relatively low reservoir and uneven crusts, nodular shapes and rigid branching
source rock potential. forms (Fig. 5.7). They vary in size from a few
millimetres to many centimetres.
• Calcareous algae with erect growth morphology
5.4.2 Calcareous Algae Usually segmented, branching algae which may
become several centimetres high (Fig. 5.8). Each
Calcareous algae form a diverse group of plants that segment is normally less than a few millimetres,
inhabit most shallow marine, clear water environ- often cylindrical with circular or oval cross-section.
ments. They prefer normal marine salinity although Since the segments generally come apart when the
some forms may be euryhaline. Most calcareous alga dies, it is very difficult to reconstruct the origi-
algae are sessile, although a few, such as the nal plant body (thallus) of fossil species.
a b c
1 cm
Fig. 5.7 Encrusting calcareous algae comprise many different growth forms including smooth and uneven crusts, nodular shapes
and rigid, branching forms. (a) Irregular crust. (b) Nodule. (c) Rigid branching form (modified from Wray 1977)
a b c
Uncalcified
joint (node)
Calcified
1 cm cellular tissue
Articulated
segments
1 mm
Fig. 5.8 (a) Segmentary growth form of the green alga algae (Corallina sp.). (c) Internal structure of calcareous
Halimeda sp. Each segment is wide and flat with a roughly segments as seen in longitudinal section. Between each joint
triangular cross section. The segments are separated by joints. there is a poorly calcified region, giving the plant some flexibility
Since there is no calcification near the joints, the algae have (modified from Wray 1977)
some flexibility. (b) Typical growth form of articulated coralline
5.4.2.1 Ecology (longwave light) being more rapidly absorbed than

Red algae are almost exclusively marine and prefer the blue (shortwave). Red algae can also use blue
shallow (<25 m) subtidal environments where the shortwave light, enabling them to live at greater depths
light is strong. They may, however, range down to a than, for example, green algae, which are dependent
depth of approximately 250 m. Some of the growth on long-wave red light (Fig. 5.9). Areas with carbonate
forms can withstand fairly strong agitation and are sedimentation normally have little clastic material in
common on shallow, rocky substrates associated with suspension, and the water is therefore very clear. The
reef and near-reef environments on platform margins. absorption of light drops off exponentially with depth
Red algae are generally less tolerant of salinity and below about 50 m there is very little light left.
variations than green algae and cyanobacteria. Most recent green algae grow in shallow marine,
During photosynthesis, algae consume CO2 and tropical waters of normal salinity, but some tolerate
produce oxygen. The absorption of sunlight by water hypersaline or brackish water. Recent green algae are
is not, however, equally rapid for the various constitu- abundant in depths of 5–10 m, but may range down to
ent wavelengths, the red portion of the spectrum 100 m and also extend into lagoons and mangrove
Intertidal skeleton of the green algae is replaced by sparry cal-

cite, whereas coralline red algae with a primary cal-
Articulated citic composition of varying magnesium content show
corallines Dasycladaceae
50 m
Cyanobacteria different degrees of alteration (often partial
Codiaceae
100 m Crustose dissolution).
corallines
150 m 5.4.2.3 Geological Range

Calcareous green algae first appeared in the Cambrian,
200 m which led to the production of large amounts of lime
mud. These algae are still an important factor in
250 m
modern shallow-water carbonate sedimentation. Red
Fig. 5.9 Generalised depth distribution of the major groups of calcareous algae also evolved in the Cambrian. In
Recent marine calcareous algae. The red algae have been drawn Recent marine environments they are commonly
in red, the green algae in green and the cyanobacteria in blue encountered both in cold and warm water deposits
(modified from Wray 1977) within the photic zone, although most of them live in
fairly shallow water (<25 m).
swamps. Some species can grow on firm substrates,
but most grow in sand or mud. They generally prefer 5.4.2.4 Significance for Petroleum Geology
environments with only moderate wave energy. The Red algae, particularly those with erect growth forms,
calcareous green algae are major producers of small may form primary framework porosity in reefs. The
aragonite crystals which form lime mud. The presence algae are susceptible to neomorphic replacement and
of green algae therefore ensures that lime mud is develop secondary porosity when exposed to exten-
deposited in the lagoons. sive freshwater dissolution.
Red algae are common associates of many oil and
5.4.2.2 Mineralogy gas producing reefs, notably stromatoporoid and coral
The skeletal material in living algae is either calcite or reefs. This includes the Middle Silurian pinnacle reefs
aragonite, never mixtures of these two minerals in the (stromatoporoid-coralgal reefs) in the Michigan Basin
same species. The extent of calcification is highly of southeastern Michigan, USA, the Upper Palaeocene
variable and may be only partial in some groups Intisar “D” Field (coralgal reef) in Libya, the Miocene
while other groups have a pervasive calcification of coralgal reefs in the Salwati Basin of Irian Jaya,
the whole plant body. Recent calcareous algae consist Indonesia and the Miocene Central Luconia Fields of
usually of small crystals (<4 μm) of precipitated cal- offshore Sarawak, Malaysia.
cite (e.g. coccolithophores produces low-Mg calcite Calcareous green algae may form important
and Lithophyllum produces high-Mg calcite) or arago- source rocks for oil.
nite (e.g. Peyssonnelia). The same is assumed to apply The biota of Late Palaeozoic warm-water
to pre-Quaternary species too. The preservation of the carbonates is often characterised by a photozoan tropi-
plant bodies depends on the extent of calcification, cal to subtropical assemblage of phylloid algae and
with the most calcified parts selectively preserved in Palaeoaplysina (if the latter is indeed a alga?). In the
the fossil record. Paradox Basin of USA some hydrocarbon reservoirs of
Modern red algae deposit either aragonite or high- phylloid algal mounds are overlain by grainstones.
Mg calcite with a Mg content up to about 30 mol% These mounds may have been formed by aggradational
MgCO3. The skeletal elements in green algae consist growth of phylloid algae and/or as wave- or tide-
only of aragonite. This primary difference in their generated current-emplaced mounds. Meteoric water
skeletal composition is an important preservation fac- may have invaded and dissolved algal plates of proba-
tor. The solubility of skeletal aragonite in pure water is ble aragonite, increasing the porosity and permeability.
about twice that of low-Mg calcite, whereas the solu- Large Palaeoaplysina–phylloid algal build-up
bility of skeletal high-Mg calcite is up to 10 times that complexes started as separate mounds and developed
of low-Mg calcite. Hence, the fine-grained aragonite into stacked build-ups with a height of tens to hundreds
of metres. The build-ups can be situated close together spheroidal or ovoidal and is covered with an external
forming elongated ridges separated by canals or ring- calcareous skeleton (coccosphere) (Fig. 5.12). The
shaped to polygonal structures surrounding a lagoon coccosphere is composed of a variable number of
(Figs. 5.10 and 5.11). Many of the build-ups and their small (a few μm), commonly round to oval, calcareous
associated lateral facies form reservoir rocks for oil plates (coccoliths). The coccoliths are made of tiny,
and gas in the Timan-Pechora Basin of northern platy low-Mg calcite crystals (0.25–1 μm) with a flat-
Russia and in the pre-Caspian basin of Kazakhstan. tened rhombic form, and are stacked in an imbricate
They are also currently of interest to the petroleum fashion that produces a spiral pattern. The
industry in the Norwegian Barents Sea, Arctic Canada coccolithophores usually disaggregate after death and
and Greenland. accumulate as individual coccoliths, sometimes
forming extensive chalk deposits.
5.4.3 Coccolithophores 5.4.3.1 Ecology

Coccolithophores are most common in the photic zone
Coccolithophores are an example of free-floating of open seas, although they may also range into
algae with carbonate shells. They are single-celled, nearshore lagoons. They swim freely in the surface
microscopic (5–20 μm) plants belonging to the waters (commonly the upper 100 m) of the ocean,
golden-brown algae. A coccolithophore is generally occurring in greatest abundance in temperate zones.
Fig. 5.10 A lagoon surrounded by a ring-shaped to polygonal species have little competition from other groups and the
Palaeoaplysina build-up complex. As a dry climate results in organic production can be very high. Hypersaline setting is
evaporation and the build-up complex restricts the exchange of also favorable in preserving organic matter, and therefore hyper-
water, hypersaline conditions develop in the lagoon. The salinity saline sediments might be potential source rocks for hydro-
increases toward the middle of the lagoon, while the biodiversity carbons. Modified from Hanken and Nielsen (2013)
decreases. Although low in number of species, the adapted
c
s.l.
s.l.
s.l.
LEGEND
Microcodium
Platform sediments lateral to build-up complex
Crinoid - Palaeoaplysina wackestone/packstone
Faunally barren dolostone containing gypsum
Fusulinid dolostone
Palaeoaplysina build-up
Fig. 5.11 (a) Palaeoaplysina build-ups initiate their growth on fauna and flora extends across the lagoon and build-up
palaeo-highs. (b) Subaerial exposure of the Palaeoaplysina complexes. On the highs, the Palaeoaplysina build-ups initiate
build-up complex and the development of vegetation on top of their growth as small disconnected structures which may
build-ups. Hypersaline conditions in the lagoon result in precip- develop into larger complexes. Based on observations from
itation of evaporites such as gypsum. (c) During transgression, Svalbard. Modified from Hanken and Nielsen (2013)
deposition of marine sediments characterised by normal marine
a b c
Coccosphere
5 µm
Coccolith
1 µm
Fig. 5.12 (a) Skeletal structure of coccospheres and coccoliths. mixture of coccoliths and loose platy calcite crystals. (c) SEM
In this example the spheroidal coccosphere is composed of many picture of Upper Cretaceous coccoliths in the Ekofisk field, North
round calcareous plates (coccoliths) which consist of tiny, platy Sea. Petroleum occurs between the small plate-like coccoliths.
calcite crystals stacked in an imbricate pattern. Modified from The limestone has 32% porosity and 1 mD permeability (from
Hjuler (2001). (b) SEM picture of Liı̀ge chalk consisting of a Bjørlykke 1989)
Coccolithophores are abundant (commonly as many as 5.4.4 Calcispheres

half a million individuals per litre of water) in the
photic zone of modern oceans. Coccoliths are impor- Calcispheres are of uncertain biological affinity, but
tant constituents of some fairly deep basinal and distal many consider them to be algae because there is a
shelf sediments. great similarity between non-ornamented fossil
Deep-sea drilling has shown that coccolithophores calcispheres and reproductive cysts of some living
were very much more widespread during the Creta- green algae (dasycladacean algae). They are generally
ceous and Lower Tertiary than they are today. The 40–200 μm in diameter and are composed of a thin
colder climate which began in the Miocene shifted the (commonly 3–30 μm), well-defined calcite wall,
northern limits for coccolithophore sedimentation enclosing a single spherical chamber (Fig. 5.13). The
southwards. Many pelagic species have very specific wall may be singly layered or have several concentric
temperature requirements and their occurrence layers distinguishable by the alternation of uniform
provides an important contribution to the micritic texture with a fabric showing radial elements.
palaeoecology of marine sediments of various ages. The wall has radial pores, but there is no aperture. The
outer surface may bear spines. The chamber is nor-
5.4.3.2 Mineralogy mally filled with cement or sediment.
The plates of the coccolithophores consist of low-Mg
calcite which makes them fairly stable during 5.4.4.1 Ecology
diagenesis. Many fossil calcispheres seem to owe their origin to
dasycladacean algae, and as such, they provide a useful
palaeo-environmental indicator. Devonian and
5.4.3.3 Geological Range Permian calcispheres are most common in shallow
Coccolithophores range from the Triassic to Recent, subtidal environments, especially in restricted or
but were not common before the Jurassic. They were back-reef environments. In the Cretaceous,
particularly abundant in the Cretaceous and Lower calcispheres are especially important in pelagic
Tertiary, forming thick, extensive chalks. They com- deposits such as Chalk sediments in the North Sea.
prise approximately 25% of present-day calcareous
oozes and up to 90% of some Cretaceous and Tertiary 5.4.4.2 Geological Range
chalks. Calcispheres are widely known from Upper
Palaeozoic limestones, especially Devonian and Car-
5.4.3.4 Significance for the Petroleum boniferous, but also occur in older and younger strata.
Industry Calcispheres have no particular significance for
Chalks made up of coccoliths may be extremely petroleum geology. However, they may be an impor-
porous. The porosity is in the form of primary inter- tant component improving the overall potential of a
particle micropores and may be as high as 40–50%, source rock.
even in deeply buried chalks. Although their porosity
may be high, permeability is generally low (less than
a few mD) because of small pore-throat diameters 5.4.5 Distribution of Algae in Modern
(generally <1 μm). Hydrocarbon-filled interparticle and Older Carbonate Sediments
pores in chalks are therefore productive only in combi-
nation with some other pore type, preferably fracture If one places a profile extending from a basin centre to
pores, which increases the permeability. Fractured a coast with reef development (Fig. 5.14), we see the
chalks of Maastrichtian (late Cretaceous) to Danian following distribution of algae:
(early Tertiary) age form major hydrocarbon A. The sedimentation basinward consists largely of
reservoirs in the Central Graben of the North Sea planktonic algae (coccolithophores), planktonic
(Ekofisk and associated fields). These are in fact the foraminifera and some other calcareous organisms.
world’s only major oilfield in such rocks, but the low B. In the reef facies we find mainly red algae which
permeability of this fine-grained lithology creates build strong, solid structures of carbonate, which in
some production problems. conjunction with the corals can resist waves which
a b c
100 µm 10 µm
100 µ m
Fig. 5.13 Sections of calcispheres. In most cases, the calcitic light, Ravnefjeld Formation (Upper Permian), East Greenland.
wall consists of only one layer (a, c), but multilayered walls are See Nielsen and Hanken (2002). a, b and c are modified from
also known (b). (d) Micrograph of the abundant calcispheres, Scoffin (1987)
partially replaced by euhedral pyrite as viewed in transmitted
Coccolithophores
Red algae
Green algae
Cyanobacteria
Reef
Open Lagoon Tidal flat
Shoal Bank
sea
Fig. 5.14 Generalised distribution of recent skeletal and non-skeletal algae across a sediment basin in a low-latitude region
(modified from Wray 1977)
break against the reef. In the lagoons landward of C. In protected parts of lagoons and in the tidal zone,
the reef we find green algae. These have bush- the sea bed is often covered with cyanobacteria
shaped skeletons and can only grow where the which lie like a gelatinous carpet over the
wave energy is moderate. They are major sediments. These algae consist of a network of
producers of small aragonite crystals which form threads which hold the sediment in place and
lime mud. The green algae therefore ensure that protect it against moderate currents and wave ero-
lime mud is deposited in the lagoon. sion. However, they may be eaten by grazing
animals, e.g. snails. In high-energy nearshore in a single row (uniserial) or in two or three alternating
environments we find red algae. rows (biserial and triserial). The shell may also be
The age range and taxonomic diversity of the major spirally coiled, either flat (planospiral) or as in a cone
calcareous algal groups are shown in Fig. 5.15. (helicoidal). The partitions (septa) between the indi-
vidual chambers are the former front wall, and the
former aperture is then often retained as an opening
5.5 Invertebrate Skeletal Fossils (foramen) between the chambers (Fig. 5.17).
5.5.1 Foraminifera 5.5.1.1 Ecology

Virtually all foraminifera are marine, but a few live in
Foraminifera (or forams for short) are a very impor- brackish or fresh water. Although they occur at all
tant group of single-celled animals that secrete cham- depths down to the abyssal zone and at all latitudes,
bered, calcareous or sometimes agglutinated tests. they have their greatest abundance and diversity in
Agglutinated tests are made up of foreign particles shallow, tropical, open marine waters. Some forami-
like mineral grains and/or shell fragments such as nifera live in symbiosis with algae living in their
coccoliths, small foraminifera and sponge spicules protoplasm, and these species are of course limited to
(Fig. 5.16a–c). These particles are attached by the photic zone. Most foraminifera are benthic (living
calcitic or ferruginous cement to a layer of tectin on the sea bed; both vagrant and sessile forms are
(an organic compound composed of protein and known), but some are planktonic (freely drifting).
polysaccharides). Many species require very specific ecological
Most foraminifera have an outer carbonate shell conditions. Some have left- or right-curved tests,
with one or more openings (apertures). A few species depending on the temperature. Benthic forms are sen-
have only one chamber, but most foraminifera add sitive to current conditions, temperature, salinity, sup-
new chambers as they grow. The individual chambers ply of nutrients etc. Foraminifera thus are useful as
are arranged in a specific pattern (Fig. 5.16d–e), either indicators of palaeoecological conditions. The amount
Cyanobacteria Green algae Red algae Coccolithophores
Cenozoic
Cretaceous Coralline
Jurassic algae
Triassic
Permian
Solenopora Older
Carboniferous
coralline
Devonian algae
Silurian
Ordovician
Cambrian
Precambrian
Fig. 5.15 Distribution of the important algal groups through algae), which was widely distributed in the Palaeozoic, died out
geological time. Cyanobacteria have existed since early Precam- during the Cretaceous. Coralline algae appeared in the Jurassic,
brian times, and the oldest stromatolites known are up to about and still exist. The evolution of the planktonic algae, mainly
3.5 billion (3.5 109) years old. Both green and red calcareous coccolithophores from the Triassic to present, has been a vital
algae first appeared in the Cambrian. The evolution of calcareous factor in global carbonate sedimentation (modified from
green algae led to the production of large amounts of lime mud Ginsburg et al. 1971)
without chemical precipitation. The Solenoporaceae family (red
a b c d e
1 mm 1 mm 1 mm 1 mm
Fig. 5.16 Examples of variations in shape, size, arrangement agglutinated with fine-grained detrital material. (d)
of chambers and mineral composition in foraminifera. (a–c) Verneuiliana sp., triserial shell. (e) Lenticulina sp., planospiral
Agglutinated forms. (d–e) Foraminifera with carbonate shells. shell. Lowermost is a median horizontal section showing the
(a) Reophax sp., agglutinated with sponge spicules. (b) (longi- individual chambers (modified from Shrock and Twenhofel
tudinal section) and (c) (external view) Hyperamina sp., 1953)
a b
1 mm 1 mm
Fig. 5.17 Partially sectioned fusulinid (a) and nummulitid (b), Permian while nummulitids have been widely used for correla-
showing the internal arrangement of chambers. The fusulinids tion of Miocene marine deposits (modified from Rasmussen
are important index fossils for the Upper Carboniferous/Lower 1969)
of planktonic foraminifera relative to benthic ones foraminifera have aragonite tests. The calcite tests
declines in shallower water, and so can be used in vary from high-Mg to low-Mg calcite, so that their
reconstructions of palaeobathymetry. solubility in seawater varies greatly. Benthic forami-
However, only a small percentage of the nifera with high-Mg calcite are found especially in
foraminiferal tests are preserved in the geological warm-water environments, where they are less soluble.
record because they are generally fragile and easily
broken by the action of scavangers and burrowing 5.5.1.3 Geological Range
organisms. The calcareous tests also dissolve as they Foraminifera range from the Cambrian to Recent.
sink through the deeper and colder waters and will not Most of the Lower and Middle Palaeozoic forms had
be preserved if they are deposited below the carbonate an agglutinated test whereas calcareous forms were
compensation depth (CCD). most common in late Palaeozoic and later times.
Planktonic foraminifera did not appear until the Meso-
5.5.1.2 Mineralogy zoic but are important constituents of Cretaceous and
Benthic foraminifera may have tests of precipitated younger deep-sea deposits. The large-sized, lens-
calcium carbonate, usually calcite, or of small sand shaped nummulitid foraminifera (up to about 15 cm)
grains cemented together (agglutinated shell). Some are widely distributed in Early Tertiary carbonate
ramp and platform deposits. They are particularly reservoir facies mainly found as banks or bars of
useful as an index fossil in Cenozoic marine deposits grainstones and packstones.
of the Mediterranean Sea region. Foraminifera provide
the basis for much of the stratigraphic subdivision of
the Upper Palaeozoic, Mesozoic and Cenozoic. 5.5.2 Sponges
5.5.1.4 Significance for the Petroleum Sponges may vary considerably in size, commonly
Industry from a few centimetres to more than a metre in dia-
Foraminifera commonly have a low reservoir potential meter. Their external morphology is also highly vari-
due to their small particle size and stable primary able, but plate-shaped, globular, vase- or bowl-shaped
mineralogy or neomorphic replacement. Some pri- forms are common. In its simplest form, the body is
mary intragranular porosity may be of local signifi- sac-shaped with a large internal cavity which opens
cance, but generally connectivity is poor, giving low upwards, the closed end being attached to the sub-
permeability. strate. The wall is perforated by numerous small
Important hydrocarbon reservoirs along the North pores leading to the central cavity.
African coastline (Tunisia and Libya) and those in the The skeleton is internal and may be spongin (an
South China Seas are largely composed of nummulitid iodine-bearing protein which is rarely preserved in the
foraminifera. The high amount of intra-skeletal porosity geological record), siliceous or calcareous. Some
and the low density of Nummulites tests and clasts make sponges may incorporate foreign particles like sand
them easily transportable by bottom and wave currents. grains in their skeleton. However, the skeleton typi-
Apparently their hydrodynamic behaviour controlled cally consists of soft tissue supported by a complex
their distribution and the predominance of them in network of spicules (Fig. 5.18). The soft tissue decays
a b d
Fig. 5.18 Examples of different forms of sponge spicules. (d) Various kinds of tetraxon spicules (modified from
(a) Monaxon spicule. (b) Part of a sponge with triaxon spicules. Rasmussen 1969)
(c) Part of a sponge skeleton built up of tetraxon spicules.
a b
G Q
S
S
0.25 mm 0.25 mm
Fig. 5.19 (a) Photomicrograph of porous spiculite consisting of cemented spiculite with very low porosity. Kapp Starostin
spicules from hyalospongiea (siliceous sponges). Note spicule in Formation (Upper Permian), Svalbard. Thin sections as seen in
longitudinal cross-section (S), quartz (Q) and glauconite grains plane polarised light. (Photographs courtesy of Sten-Andreas
(G). Pores have been impregnated with blue epoxy. (b) Well Grundvåg)
on death and spicule-bearing forms are therefore amorphous silica (opal A, SiO2·nH2O), Fig. 5.19. How-
seldom preserved intact. The spicules may occur in ever, opaline silica is relatively unstable and siliceous
two different sizes within a single individual. The sponges are commonly replaced by calcite.
large spicules (megascleres) vary from 0.1 to 1 mm Demospongiea have spicules which are made exclu-
and constitute the skeleton proper. The small spicules sively of either spongin or amorphous silica, or of both
(microscleres) vary from 0.01 to 0.1 mm and are found silica and spongin.
isolated in the sponges, especially around the pores. Sponge spicules are locally very abundant, espe-
The megascleres may occur as isolated spicules, be in cially in cherts and silicified limestones. Siliceous
close contact with each other, or be firmly intergrown sponge spicules have often been invoked as the source
to form a skeleton. of silica in such deposits.
5.5.2.1 Ecology 5.5.2.3 Geological Range

Nearly all sponges are marine although a few fresh- Sponges range from the Cambrian (Precambrian?) to
water forms are known. Sponges live mostly in rela- Recent. They are minor sediment contributors in
tively clear water, from the littoral zone down to modern settings but have been prolific in the past. Sili-
abyssal depths. Siliceous sponges are generally found ceous forms were common in the Lower Palaeozoic;
in deep water, whereas calcareous forms are more calcareous forms first appeared in the late Devonian.
common in shallow-water environments. Most Sponges are major constituents of some Upper Car-
sponges lived attached to the bottom. boniferous, Permian, Triassic and Jurassic build-ups.
Many sponges are important boring organisms Sponges have extremely high primary porosity due
which help to break down calcareous rocks and car- to their network of internal canals. However, this
bonate skeletons produced by larger organisms into porosity is rarely preserved during burial. Occasion-
finer-grained material. The holes are 0.2–1 mm in ally, high primary porosity is preserved in hollow
diameter, forming thin holes up to 0.5 cm long. spicules, but permeability is generally extremely low
and most of the porosity is ineffective.
Secondary porosity may form by dissolution of
5.5.2.2 Mineralogy
siliceous and aragonitic spicules, but rarely constitutes
The main division of the sponges is based on the
a significant part of a reservoir.
chemical composition of the spicules and the symme-
try of the skeletal elements. Calcispongiea (calcareous
sponges) have spicules of calcite or aragonite. These 5.5.3 Stromatoporoids
spicules are characterised by having a monocrystalline
microstructure where each individual spicule The systematic position of stromatoporoids has been
extinguishes as a single entity in polarised light. much debated, but they are now regarded as belonging
Hyalospongiea (siliceous sponges) have spicules of to the demosponges. They form colonies of variable
a b
c d e
Fig. 5.20 Examples of growth forms in stromatoporoids. (a) Laminar. (b) Low dome-shaped. (c) High dome-shaped.
(d) Hemispherical. (e) Branched (modified from Kershaw and Riding 1978)
size and shape. They are commonly from a few 5.5.3.3 Geological Range
centimetres to about a metre in diameter and their Stromatoporoids range from the Cambrian to Oligo-
shape may be tabular, hemispherical, cylindrical or cene, but were most important during the Silurian and
branched (Fig. 5.20). The skeleton tends to split into Devonian.
concentric layers each a few centimetres thick, and
therefore stromatoporoids often occur as larger or
5.5.3.4 Significance for Reservoir Quality
smaller fragments if they have been exposed to
Stromatoporoid reefs and associated reef detritus may
mechanical abrasion prior to their final deposition.
form important hydrocarbon reservoirs in Silurian and
Devonian rocks, e.g. Middle Silurian pinnacle reefs in
5.5.3.1 Ecology the Michigan Basin of southeastern Michigan, USA
Stromatoporoids are sessile, open-marine, colonial and Middle to Upper Devonian reefs in Alberta,
organisms. Since the growth form is to a certain degree Canada. Although stromatoporoids may form high
dependent on the environment, some forms can be primary framework porosity, this porosity is com-
used in palaeoecological analysis relating the growth monly obliterated by synsedimentary infilling and
rate vs. sedimentation rate. Stromatoporoids were cementation of the pores. Secondary porosity is gene-
important constituents of Ordovician to Devonian rally more important and may be related to
reefs along with tabulate corals, colonial rugose corals dolomitisation or freshwater dissolution.
and calcareous algae. These species mostly lived in Stromatoporoid reefs were topographically higher
agitated, shallow, warm seas, while Jurassic and Cre- than adjacent sediments and were therefore more
taceous forms were more adapted to life in muddy prone to be subaerially exposed and dissolved.
environments.
5.5.4 Corals
5.5.3.2 Mineralogy
The stromatoporoids had a very porous skeleton Most corals are colonial, although solitary forms are
(coenosta) forming a laminar structure transected by also common. Colonial corals may be from a few
fine vertical pillars. However, the primary structures centimetres to several metres in diameter, but are
readily underwent diagenetic modifications and there- commonly less than 0.5 m. The external form of the
fore were commonly lost during diagenesis, leaving colonies may be extremely variable, but can generally
only a reticulate pattern of anastomosing elements. be described as crustose, hemispherical, irregular or
This often renders primary microstructures difficult shrub-like branched (Fig. 5.21). The colonies grow in
to differentiate from secondary modifications. The size by repeated budding from the side or top of the
neomorphic fabric indicates that the original composi- individual corallites. The newly formed corallites may
tion of the stromatoporoid coenostae was aragonitic. abut against one another like the cells in a beehive,
Offset
a
1 mm
Mural pores
c d
1 cm 1 mm
1 mm
Tabulae
e Offset f
Connecting
tubule
1 cm
1 cm
Fig. 5.21 Examples of different growth forms in colonial visible. Mural pores are seen where the plane of the section passes
tabulate corals. (a) Favosites sp. Massive growth form with slightly obliquely through the corallite wall. (d) Enlarged trans-
polygonal (generally hexagonal) corallites. Part of the right side verse section showing mural pores and septal spines. (e)
has been removed to expose the corallites in longitudinal section. Syringopora sp. Open growth form with round corallites linked
(b) Enlarged section of corallites viewed in longitudinal section. by short, horizontal, connecting tubules. (f) Halysites sp. Cylin-
Note the vegetative reproduction in that one of the corallites is drical corallites which form palisade-like, dendritic rows in an
dividing into two, giving rise to an offset. (c) Enlarged longitudi- irregular, reticulate pattern with large open spaces (modified from
nal section through three corallites where tabulae are clearly Moore 1956, Rasmussen 1969)
thus forming compact colonies with polygonal, gener- flattened areas, several main types of growth forms
ally hexagonal corallites (Fig. 5.21). However, some can be recognised (Fig. 5.22).
species form more open colonies so that the individual Individual corallites are usually divided by radially
corallites are free-standing and rounded in cross- arranged vertical elements (septa), horizontal
section. partitions (tabulae) and small partitions in the lower
Solitary corals may be curved or erect, and their periphery of the corallite (dissepiments). The presence
fundamental shape appears to be a reversed cone. or absence of these structures is a basis for classifying
Depending on the apical angle of the cone and other corals into four major groups: rugosa (horn corals),
characteristics, such as the growth form of the mature tabulata, scleractinia (hexacorals) and alcyonaria
region and the occurrence of sharp angulations or (octacorals).
rods with spines and protuberances. Sclerites may

be locally relatively common in Recent
Discoid sediments, but since they are often difficult to
recognise in pre-Quaternary deposits, the fossil
Trochoid record of alcyonarians is sparse.
Patellate
5.5.4.1 Ecology
Corals are bottom-dwelling, sessile, marine organisms
which are most common in shallow, well-oxygenated,
tropical to subtropical waters of normal marine
salinity. However, some forms live in deep water
Turbinate Ceratoid Cylindrical below the photic zone and/or in cold waters. Their
growth forms are useful environmental indicators.
Fig. 5.22 Examples of different growth forms in solitary
rugose corals (modified from Moore 1965)
Scleractinian corals have been major reef builders
since the Triassic and tabulate corals were important
constituents of Silurian and Permian build-ups. Some
Rugose corals: Rugose corals consist of both solitary and
recent scleractinian corals (hermatypic corals) corals
colonial growth forms. All rugose corals have both
live in symbiosis with single-celled algae
tabulae and septa. The septa are arranged in a radial
(dinoflagellates or zooxanthella). They do best in
pattern and generally have a marked bilateral sym-
very shallow areas (1–20 m deep, but some species
metry. Both the form, size and number of septa vary
might be found down to 90 m) and in warm water
up through the corallite (Fig. 5.23). During growth,
(18–29 C, but they flourish best between 25 and
the deepest part of the horn (calice), may be cut off by
29 C). Most Mesozoic and Cenozoic reef corals are
transverse partitions (tabulae) or contracted by small,
hermatypic, and are therefore restricted to shallow
curved partitions (dissepiments).
warm water. Most of these corals are euryhaline and
Many rugose corals are colonial. There are two
do not tolerate turbid water.
main types: fasciculate where the corallites are
The other main type of scleractinian corals
cylindrical and isolated from each other, and mas-
(ahermatypic corals) do not live in symbiosis with
sive where the corallites are closely packed and
algae, and often occur below the photic zone. They
polygonal in cross-section.
are most abundant down to 500 m, but have been
Tabulate corals: All tabulate corals are colonial and
found at water depths as great as 6,000 m. They can
individual corallites may be connected by small
live in water with temperatures as low as 0 C.
pores through a common wall, porous tissue or
Ahermatypic corals can form reef-like structures at
tube-like canals (Fig. 5.21c, d). Most tabulate
depths of several hundred metres. Lophelia reefs or
corals have horizontal tabulae. Septa, as seen in
biostromes are found, for example, in deep water
rugose and scleractinian corals, are absent, but
settings off the Norwegian coast.
some species have small septal spines.
Rugose and tabulate corals were important
Scleractinia (hexacorals): Scleractinian corals consist
components of Ordovician and Silurian reef structures.
of both solitary and colonial growth forms. They
have an aragonitic skeleton in which septa form a
sixfold radial symmetry, and tabulae may form 5.5.4.2 Mineralogy
horizontal partitions. All scleractinaian corals are aragonitic and pre-
Alcyonaria (octacorals): A fourth group of corals is Holocene specimens are therefore normally preserved
the alcyonaria. All alcyonarian corals are colo- as moulds or have been more or less cemented with
nial. Most have an endoskeleton consisting of calcite spar. Rugose and tabulate corals have usually
spicules of sclerites. In its simplest form, the retained their primary microstructure, which indicates
sclerite is a rod with more or less acute ends. that the skeleton consisted of calcite. Recent
More complex forms have irregularly sculptured octacorals have sclerites consisting of calcite or
a c
Calice
Septa
Cardinal
fossula
Tabulae
d
Epitheca
Fig. 5.23 (a) The rugose coral Zaphrentites with part of the (d) Longitudinal section through Ketophyllum showing tabulae
epitheca removed, showing septa and tabulae. (b) Transverse in the middle part and dissepiments along the periphery
section showing the tabulae. (c) Ketophyllum. Solitary rugose (modified from Clarkson 1979)
coral with tentacle-like appendages producing tabular holdfasts.
aragonite, and it is believed that pre-Quaternary spe- 5.5.4.4 Significance for Reservoir Quality
cies also had the same mineralogical variations. Corals may form high primary framework porosity in
reefs. Scleractinian corals have an unstable primary
5.5.4.3 Geological Range mineralogy (aragonite) so may develop high second-
Rugose corals range from the Middle Ordovician to ary mouldic porosity during dissolution. Corals, in
the Upper Permian. They were not common until the particular scleractinian corals, therefore have rela-
Silurian, after which their numbers gradually tively high reservoir potential.
increased to a peak in the Lower Carboniferous.
Their numbers and diversity then declined until they
finally became extinct by the end of the Permian. 5.5.5 Bryozoa
Tabulate corals occur in rocks of Cambrian (Middle
Ordovician?) to Permian (and possibly Triassic to Bryozoans form colonies which are usually 0.5–2 cm
Eocene?) age. Scleractinian corals range from the in diameter, but colonies up to about 60 cm are known.
Middle Triassic to Recent, and alcyonarian corals They live permanently fixed to a substrate, which may
range from the Precambrian to Recent. Scleractinian be a stone, shell (living or dead), algae or other
corals are the dominant coral group today. objects. The colonies have a wide range of external
a b c
Zooecia
5 mm
1 mm 1 cm
Fig. 5.24 Examples of different growth forms in bryozoa. (a) Fenestrellina sp. with a cornet-shaped colony. All the zooecia
Cyclostomata. Spiropora sp. with a dendroid growth form with are placed on the radial, longitudinal branches, with their
mouths on separate polypides. (b) Trepostomata. Colony and mouths inside the cornet. (modified from Rasmussen 1969)
enlarged section showing part of the interior. (c) Cryptostomata.
morphologies, including encrusted (flat, hemispherical had a calcitic skeleton, while many Recent forms are
or irregular), branched (hollow or massive, flattened or composed of aragonite or mixed aragonite and calcite.
circular), and lacy network (Fig. 5.24). Individual
colonies may consist of tens to hundreds of circular
or polygonal tubes or boxes (zooecia) which contain
Bryozoans range from the Ordovician to Recent. They
the individual organisms.
are relatively rare in modern environments, but were
especially important as framebuilders or sediment
5.5.5.1 Ecology binders in many Ordovician to Permian build-ups.
Bryozoans are widespread, colonial, sessile, suspen-
sion feeding organisms which have a limited tolerance
5.5.5.4 Significance for Reservoir Quality
for strong wave action, soft sediment surfaces and
Bryozoans may have relatively high primary porosity
desiccation. They prefer normal marine salinity but
in the form of interparticle, intragranular and frame-
can tolerate short-term salinity variations. Most
work porosity, although this porosity is not always
bryozoans are marine, occurring in shallow to abyssal
preserved during burial. Intragranular pores are usu-
depths (most common down to 200 m water depth),
ally poorly interconnected giving low permeabilities,
although a few Recent forms live in freshwater. In
and are therefore generally of minor importance in
spite of their broad environmental distribution,
petroleum reservoirs. Framework and interparticle
bryozoans are generally most abundant in shallow to
pores are usually better interconnected and may be
moderately deep, marine settings.
important in some bryozoan build-ups, e.g. in the
Middle Carboniferous of Central Spitsbergen, occa-
5.5.5.2 Mineralogy sionally forming good petroleum reservoirs.
There are five orders of bryozoans with a fossil record. Bryozoans generally have a low potential for sec-
In the Cyclostomata, Trepostomata and ondary porosity formation due to a stable primary
Cryptostomata, the skeleton is composed of calcite mineralogy. However, bryozoan-rich sediments may
with less than 8 mol% MgCO3. Bryozoans belonging still be highly porous when they form build-ups. Bryo-
to the Ctenostomata are poorly mineralised with more zoan build-ups formed topographical highs during
than 50% organic matter. The most primitive members deposition and were therefore prone to be subaerially
of the Ctenostomata and Cheilostomata possess a exposed during sea level fluctuations. In a humid cli-
skeleton of calcite only, whereas the more specialised mate, these build-ups may be pervasively dissolved
cheilostomes have a secondary thickening of arago- with the formation of secondary, non-fabric selective
nite. However, in general, most Palaeozoic bryozoans porosity. The Lower Permian bryozoan-Tubiphytes
Crinoidea Echinoidea Asteroidea

Quaternary
Astropecten
Tertiary
Cretaceous
Woodocrinus
Jurassic Orophocrinus
Holothuria
Triassic
Permian Homalozoa Cystoidea
Ophiocomina
Carboniferous Cidaris
Devonian
Silurian Holothurioidea
Blastoidea
Ordovician
Ophiuroidea
Cambrian Dendrocystis Pleurocystites
Fig. 5.25 Time range and classification of echinoderms with representative genera illustrated (modified from Ziegler 1983)
build-ups in the Midland Basin of West Texas, USA, plates or the presence of articulated specimens
are examples of this. normally indicates deposition within the area
they inhabited.
5.5.6 Echinoderms 5.5.6.1 Crinoids

The length of crinoids varies greatly, from a decimetre
The phylum Echinodermata comprises several classes, up to several metres (even 25 m). Their general struc-
of which crinoidea and echinoidea are most common ture is shown in Fig. 5.26. In the upper part is a cup
in the fossil record (Fig. 5.25). Modern representatives (calyx), which consists of 10–15 plates in two or three
include the sea urchins (echinoids), starfish and brittle rows. Five (or a multiple of five) arms (brachia) extend
stars (stelleroids), and sea lilies and feather stars from the calyx and may have small branches
(crinoids). However, only crinoidea and echinoidea (pinnulae). The mouth is centrally located in the top
are dealt with in this book because these groups are of the calyx, and the anus is often situated on an
most commonly encountered in the field. One of the outgrowth from the calyx (anal conus). The calyx is
characteristic features of the echinoderms is that, situated on a flexible stem consisting of various types
almost without exception, various degrees of endo- of small segments (columnal plates or stem plates).
skeleton calcification are found in the body wall. A These columnals consist of variously shaped and sized
single individual may be composed of a very large plates, generally arranged in a single series
number of plates (often several hundred, or even (Fig. 5.27b). Each plate has a central round, or less
thousands). On death, the individual plates, ossicles commonly pentapetoloid, perforation, which together
and spines disarticulate, unless the organism is quickly form an axial canal (Fig. 5.27a). The surface
buried and thus avoids physical destruction of the separating the stem segments often has a regular radial
skeleton. These plates are fairly sturdy and are often pattern of fine chambers and furrows. The stem may
found whole in limey deposits. have thin lateral branches (cirri) which serve to fasten
Because of their microporous texture, echino- the crinoid to nearby objects (Fig. 5.27c). Lowermost,
derm plates have a low density and may be readily the stem may form a root-like, branched system that
transported into virtually all kinds of environ- serves as a holdfast to objects, or forms an anchor in
ments after death, though a high concentration of loose sediments. When the crinoid dies, the column is
Brachia quite likely to fall apart to a greater or lesser extent.

Some fossil crinoids have the entire stem and root
Crown system preserved, but fragmental stems and single
columnal plates are much more typical. The skeleton
in Recent echinoderms is exceedingly porous (ca.
50%) with a reticulate pore structure (Fig. 5.27d).
The pore diameter is about 25 μm.
Calyx
5.5.6.2 Echinoids (Sea Urchins)
Echinoids have a hemispherical, disc- or heart-
Column
shaped endoskeleton consisting of interlocking

plates. The outer surface is covered by spines situated
on the interambulacral plates which alternate with
the spine-free, mostly smaller, ambulacral plates
(Fig. 5.28).
Cirri
Root
5.5.6.3 Ecology
All echinoderms are marine, preferring normal marine
salinity and ranging from shallow water to abyssal
Fig. 5.26 Features of the general crinoid morphology showing depths. Crinoids inhabited shallow marine
a complete crinoid skeleton in living position (modified from environments during the Palaeozoic, sometimes
Shrock and Twenhofel 1953)
forming small bioherms or beds dominated by crinoid
a b
c d
Fig. 5.27 (a) Part of crinoid columnal showing the general the cirrate column of Comastrochinus sp. (d) Enlarged section
morphology with radial grooves for connective tissues on the of an echinoderm showing the network of large openings which
top plate. Note the round lumen in the centre of the plates. (b) transects the structure (modified from Shrock and Twenhofel
Schematic drawing of columnals viewed from the side showing 1953, Rasmussen 1969)
examples of the great variation in stem morphology. (c) Part of
Inte
ram
bula
a crum b
Tooth
Ambulacrum
1 cm
Fig. 5.28 (a) Adoral surface of the regular Archaeocidaris sp. from regular echinoids. (e) Part of Archaeocidaris sp. showing
with terminology. (b) Part of the surface showing ambulacra and a few ambulacral plates with spines preserved. (modified from
interambulacra with tubercles. (c) Interambulacra with spines Rasmussen 1969, Clarkson 1979)
preserved. (d) Examples of different growth forms of spines
debris, but since then they have been most common in beds dominated by crinoid debris. Their abundance
relatively deep waters (>100 m). In contrast, other declined in the Permian, but revived in the Mesozoic,
echinoderms have commonly inhabited shallow although they never equalled their Palaeozoic peak.
marine environments throughout their geological Crinoids remain prolific today.
time range. Echinoids appeared for the first time in the Upper
Most echinoderms are benthonic and a few are Ordovician, but are most common in Mesozoic and
pelagic. Echinoids are usually vagrant, whereas Cenozoic sediments. They remain a vital part of the
crinoids are generally sessile. invertebrate marine realm and are probably as abun-
dant now as at any time in the past.
5.5.6.4 Mineralogy
The skeleton is composed of high-Mg calcite 5.5.6.6 Significance for Reservoir Quality
(5–15 mol% MgCO3). Each plate, spine or sclerite Echinoderms are commonly overgrown by syntaxial
has a single-crystal microstructure behaving optically calcite cement which completely obliterates both
as a single calcite crystal when viewed under the intragranular and intergranular porosity in
polarising microscope. However, the optically echinoderm-rich rocks. This, along with the common
uniform crystals are really a mosaic of submicroscopic neomorphic replacement of the echinoderm skeleton,
crystals whose c-axes are aligned almost perfectly gives these rocks low reservoir potential. Dolomitised
parallel. crinoidal sediments may occasionally be oil producing,
as for example in the Silurian crinoid-rich skeletal
build-ups in the northeastern Anadarko Basin of
Oklahoma, USA (Morgan 1985).
Echinoderms range from the Lower Cambrian to
Recent, but were not common before the Ordovician.
Their greatest abundance was during the Carboniferous.
Crinoids appeared for the first time in Lower Ordovi- 5.5.7 Post-mortem Destruction
cian sediments, but did not become abundant before the
Silurian. They remained numerous in the Devonian and As shown in the preceding review, carbonate grains
Carboniferous, sometimes forming small bioherms or differ widely in size with diameters ranging from a
few micrometres in coccolithophores to more than more durable than those with a large surface area
a metre in some bivalves and sponges. Apart from per unit weight.
grain type, grain size is also dependent on mechanical • Biological destruction. As soon as any organism
abrasion and bio-erosion prior to final burial. dies, its skeletal material begins a process of deteri-
These physical and biological processes can be oration due to the activity of a great variety of
described as: organisms. In particular many species of sponges,
• Mechanical destruction. The rate of abrasion is worms, bivalves, fungi and algae have the ability to
highly dependent on the transport mechanism and sculpt or penetrate hard calcareous substrates e.g.
energy available in the environment for particle skeletal material lying exposed on the sea bottom.
movement. Shell fragments and other skeletal Skeletal material has a low hardness (e.g. low-Mg
particles are rapidly crushed and ground down to calcite has a hardness of 3 and aragonite 3.5–4 on
become structurally unrecognisable in a high- Moh’s scale of hardness) and is easily dissolved in
energy environment such as a surf zone, while in weak acids which render these substrates attractive
low-energy environments the integrity of the origi- for bioeroding organisms. The density and diversity
nal material may remain more or less intact. of bioeroding organisms may therefore be high on
Mechanical destruction of skeletal material has carbonate material which has been exposed on the
characteristic rates of destruction, depending on sea bottom for a year or more. Bioerosion is a self-
shell thickness, microstructure and possible prior destructive process, and the sculpture that is pro-
bio-erosion. A primary factor is the surface area to duced can only be preserved if covered by sediment,
weight relationship (Fig. 5.29). These differences which suffocates the endolithic community.
in abrasion rate may sometimes give rise to a fossil Bioerosion increases the vulnerability of skeletal
assemblage that deviates markedly from the origi- material to mechanical destruction, with the result
nal composition (Fig. 5.30). In general, in any that a significant part of carbonate deposits may
given abrasive environment, skeletal material with consist of a combination of bio-eroded and mechan-
a small surface area per unit weight is generally ically fragmented material (Fig. 5.31).
Percentage of original material still greater than 2 mm
100
48g
23g
17g
9g
6g
8g
80
6g
7g
60
Echinoids
Corallina
M. A. H. T. L.
Limpets
Haliotis
Mytilus
Piaster
Tegula
40
Aletes
20 E. P.
C.
0
Amounts of constituents After 2 hours
at beginning of experiment
100
80
60
40
M. A. H. T. L. M. A. H. T. L.
20
0
After 40 hours After 183 hours
Fig. 5.29 Experimental abrasion of skeletal material. Starting Pisaster sp. and the calcareous algae Corallina sp. The series
sample (upper left) contained fresh specimens of the bivalve of diagrams shows selective destruction of the assemblage by
Mytilus sp., the gastropods Aletes sp., Haliotis sp. and Tegula tumbling (modified after Chave 1964)
sp., various species of limpets and enchinoids, the starfish
a b c
Fig. 5.30 Examples of breakdown of various skeletal ostracod; (c) brachiopod; (d) calcareous red algae; (e) bryozoan;
components by mechanical abrasion. Echinoderms, coccoliths (f) bivalve. Not to scale (modified from Bromley 1980)
and many bryozoans are particularly fragile. (a) Echinoid; (b)
5.5.8 Micritisation When the bioeroding organisms die, each hole gets
filled with tiny crystals (micrite) of aragonite or high-
This process, which was first described by Bathurst Mg calcite. Which crystal type is precipitated depends
(1975), is due to the fact that exposed skeletal on the primary mineralogical composition of the shell.
fragments within the photic zone may serve as suitable Usually the precipitate is aragonite if the shell is ara-
substrates for various green algae, cyanobacteria and gonite, and likewise for fragments of high-Mg calcite.
fungi. These organisms destroy the surface by boring a However, some bacteria can secrete aragonite, and
fraction of a millimetre into the fragment by secreting therefore aragonite-fills in empty boreholes may have
acids which dissolve tubular boreholes in which the been produced by bacteria which lived on the algae. In
filaments reside. The diameter of each hole varies pre-Quaternary sediments low-Mg calcite has usually
somewhat, but usually it is around 5–7 μm (maximally replaced both high-Mg calcite and aragonite micrites.
15 μm). Repeated borings followed by precipitation and
Bored chips Cliona Infesting

(sponge) coral
(4 to 6 Φ)
Faecal waste Parrot fish

(1 to 3 Φ)
Coral fragments
Faecal pellets Diadema

(2 to 3 Φ) (sea urchin)
(0 to –2 Φ)
Relative
abundance
10 8 6 4 2 0 –2 –4 –6 –8 –10 Φ
0.001 0.004 0.016 0.63 0.25 1 4 16 64 256 1024 mm
Size
Fig. 5.31 Diagram of the dominant size fractions of grains for habitation. Both the parrot fish and sea urchins are important
produced by biological destruction of massive corals by grazers that break down the surface of calcareous substrates in
endolitic sponges, parrot fish and sea urchins. The endolitic search of epilithic and endolithic plants for food (modified from
sponges infest calcareous substrates to create a sheltered site Scoffin 1987)
Constructive Destructive
Time
Live filament
Vacant boring
Micrite-filled boring
Grain
Fig. 5.32 Micritisation of a calcareous grain. The process is confined to the outer surface. If the grain is completely
related to the activities of microboring algae, cyanobacteria and micritised, a peloid or cryptocrystalline grain may be the end
fungi. A micritic envelope is developed if the micritisation is result. Not to scale (modified from Kobluk 1977)
infilling of the empty cavities cause the fragments to sediments. Where aragonite fossils have been
be replaced by a dark micritic coating (Fig. 5.32). The micritised, the coating (now consisting of low-Mg
micritised coating is highly stable and is commonly calcite) often clearly indicates the original shape of
preserved in both Quaternary and pre-Quaternary the fossil (Fig. 5.33).
Infestation by Dissolution of Calcite

microborers remaining cementation
aragonitic of skeletal
Precipitation shell wall mould
of micrite
in holes
Fresh aragonitic
mollusc fragment
Occurs rapidly in Most commonly seen like

the marine environment this in ancient limestones
Fig. 5.33 Diagram showing the formation and preservation of of the skeletal material and immediate precipitation of calcite
a micritic envelope. Skeletal fragments of aragonite may be along a thin solution film. Right: complete solution of the
dissolved and replaced by calcite by one of two processes. skeletal fragment. During later diagenetic stage(s), the mould
Left: much of the original structure, such as organic inclusions, may be partly or completely cemented by calcite (modified from
may be preserved (albeit not perfectly) through gradual solution Bjørlykke 1989)
5.5.9 Microcrystalline Lime Mud (Micrite) In areas where strong evaporation and high salinity
approach evaporite conditions, such as in the Dead Sea
Lime mud deposited in areas with carbonate sedimen- and the Persian Gulf, this kind of chemical precipita-
tation has a grain size of about 1–4 μm. This precludes tion does occur, but only when the salinity is very
effective study under an ordinary microscope, only high, about every fifth year. A sudden proliferation
with an electron microscope can each individual of diatoms could also consume so much CO2 that the
grain really be seen. pH rises and causes aragonite to be precipitated.
Carbonate mud (micrite) was previously assumed Microcrystalline lime mud may also be formed by
to be chemically precipitated carbonate, as opposed to mechanical abrasion. Fossils which lie exposed to
fossils or fossil fragments of clearly organic origin. wave and current activity will be abraded mechani-
However, a very large amount of modern lime mud, cally, and a fine-grained lime mud formed. However,
for example from the Bahamas, has been found to this process will not produce the well-crystallised ara-
consist of aragonite needles from the breakdown of gonite needles of which modern lime muds are found
calcareous green algae, particularly Halimeda, largely to consist. We may therefore conclude that
Rhipocephalus and Penicillus. This was confirmed mechanical abrasion of fossils or carbonate fragments
partly by studies of the shape of aragonite needles is not the main source of lime mud.
(by electron microscope) and partly through isotope Through diagenetic processes, the microcrystalline
studies of the 18 O=16 O and 13 C=12 C ratios in the ara- mud will dissolve and be replaced by microsparite, but
gonite mud. The isotope values for the aragonite mud this is clearly distinguishable from other sparites, not
correspond well with the values in needles formed by least by its brownish colour due to incorporated
various types of algae. organic matter (Fig. 5.34).
In shallow marine areas near the equator, like the
Bahamas, the seawater is often saturated with respect
to aragonite. Nevertheless it has not been possible to 5.6 Non-skeletal Grains
prove conclusively that any purely chemical precipita-
tion of aragonite takes place in these areas. Sudden Carbonate grains can be classified as skeletal or non-
whitings of the seawater that are often observed in skeletal. Non-skeletal components are defined as
these areas, has been interpreted as spontaneous grains which do not appear to have been precipitated
crystallisation of aragonite, but the phenomenon can as skeletal material. However, this neither proves
also be interpreted as a disturbance of the lime mud that they could not once have been skeletal, nor
from the bottom, for example by shoals of fish which that they have an inorganic origin, it only signifies
are present in large numbers where there is whiting of that in their present state no skeletal origin can be
the seawater. ascertained.
Fig. 5.34 The original microcrystalline mud (micrite) has been trilobite and ostracod fragments. Thin section seen in plane
replaced by microsparite which consists of a mosaic of small polarised light. Silurian, Gotland
brownish crystals. The skeletal remains are dominated by
5.6.1 Intraclasts any size or shape, but have generally become well
rounded during transport. Because carbonates are vul-
Intraclasts are erosional fragments of essentially nerable to weathering and are mechanically weak, the
penecontemporaneous carbonate sediment. They are content of extrabasinal carbonate clasts declines with
of intrabasinal origin, i.e. were eroded from the sea increasing transport distance, particularly in humid
bottom or adjacent tidal flats and deposited within the climates.
area of original deposition. Early cementing, e.g. in
the beach zone, or freshwater cementation due to
regressions, may result in cemented sediments which 5.6.3 Faecal Pellets
are later broken up. We may also find early cementa-
tion in a marine environment, forming seafloor hard Faecal pellets are organic excrements generally pro-
grounds. If these are then eroded by tidal channels, for duced by mud-eating grazing animals such as
example, the resulting intraclasts may accumulate as crustaceans, worms and molluscs. They are generally
characteristic intraformational conglomerates. round, oval or rod-like in shape and commonly consist
Intraclasts can also consist of semi-consolidated of carbonate mud. Most faecal pellets are 0.1–0.6 mm
pieces of stromatolites. in length and 0.1–0.4 mm in diameter, but can be up to
Intraclasts may be of any size or shape. They are 2 mm long. Animals which eat mud deposit vast
quite common in shallow water limestone facies and quantities of carbonate mud pellets in carbonate
are good indicators of early cementation and erosion. environments, and clay and silt pellets where fine-
grained siliciclastic material forms the seafloor. Faecal
pellets have a characteristic dark brownish colour
5.6.2 Lithoclasts imparted by their organic content.
Faecal pellets are produced in all environments,
Lithoclasts, which encompass both intraclasts and often in well-sorted large quantities forming
extraclasts (or extrabasinal clasts), are produced by pelsparite/pelloidal grainstone (Fig. 5.35). Their
the erosion of exposed older or synsedimentary preservation within a sediment generally requires
lithified or partially lithified sediment. On carbonate intragranular lithification, a process occurring in shal-
platforms, lithoclasts are normally derived from ero- low waters which are supersaturated with respect to
sion of the landward margin, where exposed, calcium carbonate. The presence of a low-energy
cemented limestones are broken up by physical, chem- environment is also important to prevent physical
ical and/or biological processes. Lithoclasts may be of destruction. This implies that faecal pellets are most
a b
Pe
Pe
0.3 mm 0.3 mm
Fig. 5.35 (a) Rounded rod-shaped pellets (Pe). Upper Cretaceous, France. (b) Dark pellets (Pe) in a sandy matrix. Silurian, Oslo
Region, Norway. Thin sections as seen in plane polarised light
commonly encountered in low-energy, muddy deposi-

tional environments in the platform interior.
5.6.4 Peloids
The term peloid is used for a polygenetic group of

spherical, ellipsoidal or angular grains with diffuse
margins but no internal structure. The grains consist
of fine-grained carbonate irrespective of size or origin.
Because of the diverse origin of the particles, the term
peloid is purely descriptive and does not denote origin.
Various origins are possible: faecal pellets, micritised
grains, intraclasts, a type of abiogenic precipitate, and
a microbially mediated precipitate. Most peloids, how-
ever, are formed by extensive micritisation of pre-
existing carbonate grains which is related to the
activities of microboring algae and fungi within the
photic zone. Thus, peloids generally indicate shallow,
non-agitated waters.
5.6.5 Ooids
Ooids are round grains of carbonate formed through

chemical or biological precipitation by the rolling
action of waves and/or tidal currents in shallow Fig. 5.36 (a) Recent ooids from Bahamas. They are well
water. Sediments consisting of ooids are called oolites. sorted with grain size mostly between 0.5 and 1 mm.
(b) Ooids are composed of a nucleus overlain by successive
Ooids are by definition less than 2 mm, but most are concentric coatings of aragonite. Thin section, plane
0.3–1.0 mm in diameter (Fig. 5.36a,b). They have a polarised light. Recent, Bahamas. (Photograph courtesy of
concentric structure with layers of carbonate around a Arve Lønøy)
core which may consist of small clastic grains or a Ooids are therefore an important indicator of deposi-
carbonate fragment. Similar concentric structures tional environment and climate.
larger than 2 mm are called pisolites, and these are
usually precipitated by algae.
5.6.5.1 Mineralogy
Modern ooids consist of aragonite with a concentric
Ooids from earlier geological periods with warm
tangential structure composed of small aragonite
climates (greenhouse) have a radial structure and
crystals (<3 μm) with their c-axis parallel to the
may have been composed of primary Mg-calcite.
lamination.
During cold periods in the Earth’s history (icehouse
We find ooids only in very warm marine
conditions), as now in the Quaternary and during the
environments and in some saline lakes. The water
Permo-Carboniferous glaciation, ooids are initially
must be close to saturation with carbonate. Agitation
composed of aragonite (Fig. 5.38).
from waves or tidal currents is necessary to roll the
ooids around to so that precipitation is even and forms
concentric layers. Some studies indicate a thin organic 5.6.5.2 Significance for Reservoir Quality
membrane of bacteria on ooids which helps to Ooids may form extensive bank-margin or shoreline
precipitate aragonite, and which also may trap small sand bodies which can be cemented together to form a
aragonite particles in the sediment or suspended in the sedimentary rock called an oolite. These deposits may
bottom water. If direct chemical precipitation was have excellent reservoir qualities because of a combi-
involved we would expect the needles to orientate nation of secondary mouldic porosity related to the
themselves radially on the surface of the ooid, whereas dissolution of ooids and/or primary intergranular
snowball-type growth through the accumulation of porosity (Fig. 5.39). Permeability may vary depending
small aragonite needles would give the observed con- on the degree of cementation or dissolution, but is
centric layers. generally high.
Because ooids require warm water and frequent Examples of petroleum production from secondary
wave agitation, they are only formed in very shallow mouldic pores include the Upper Silurian oolite shoals
water, normally 2–7 m deep (Fig. 5.37). During of the northeastern Anadarko Basin, Oklahoma, USA,
storms, however, they may be carried out into greater the Lower Permian oolite shoals of the Midland Basin,
depths and deposited there. Ooids are typical of the West Texas, USA, the Upper Jurassic oolites of the
Bahamas, of many areas around the Indian Ocean and Smackover Formation of the Gulf of Mexico coastal
on islands in the Pacific Ocean. On the east side of the plain, USA and the Lower Cretaceous oolite shoals of
Atlantic Ocean – along the coast of Africa – the water offshore Angola. Petroleum production from primary
is too cold to permit sufficient carbonate saturation. intergranular pores is exemplified by Lower Carboni-
ferous oolite shoals in the St. Genevieve Formation,
Illinois Basin, USA.
High-Mg calcite + aragonite

Low-Mg calcite
? ?
Aragonite threshold
Climate
Cold Warm Cold Warm Cold
PC C O S D C P TR J C T
Fig. 5.37 Aerial photograph of ooid sand bar prograding over Fig. 5.38 A comparison between the mineralogy of ooids and
carbonate mud. The mud is covered with blue–green algae climate. During warm climate conditions ooids consist of low-
(cyanobacteria) which is cohesive and protects the sediments Mg calcite while during cold periods they consist of high-Mg
from erosion to certain extent calcite or aragonite (modified from Sandberg 1983)
a b
Oo Oo
0.7 mm 0.7 mm
c d
Py Py
Ve Ve
Ca
0.7 mm 0.7 mm
Fig. 5.39 (a) and (b) Oomoulds (Oo) formed by dissolution of oomoulds infilled with calcite cement (Ca) and pyrobitumen
ooids as viewed in transmitted light and plane polarised light (Py), viewed in transmitted light and plane polarised light
respectively. The matrix consists of dolomite. Roker, England. respectively. Upper Ordovician, Oslo Region, Norway
(c) and (d) Calcite vein (Ve) cross-cutting partially cemented
5.7 Carbonate Environments equal to the total amount of carbonate which

becomes dissolved in the oceans plus the supply of
5.7.1 Introduction Ca2+ released by weathering on land and transported
to the ocean by rivers.
Most carbonate deposits are due to the production Accumulations of carbonate deposits require:
of skeletal material rather than being precipitated • A supply of nutrients to feed the carbonate-
chemically. The composition of the sediments is precipitating organisms.
therefore largely a function of the type of organisms • Clear water with a low content of suspended clastic
which produced them. Biological precipitation material.
removes CaCO3 from seawater at a sufficient rate The effect of suspended fine-grained clastic mate-
that the seawater does not become supersaturated rial is to dilute the carbonate content, but it may also
with respect to calcite, except locally. Shallow sea- retards the growth of filter feeding organisms like
water may be supersaturated with respect to calcite corals. These sessile organisms require a through-
and in the tropics also with respect to aragonite and flow of water to obtain enough planktonic organisms
high-Mg calcite. It is mostly in evaporite basins that for nutrition, and too much suspended clastic material
chemical precipitation is dominant but even there it tends to clog up their filtering system.
is often assisted by algae. However, a very large Carbonates accumulate along coastlines with little
fraction of the carbonate precipitated by organisms clastic supply from rivers. This is often the case when
is dissolved in the water column or soon after depo- the adjacent land has a dry climate with little runoff.
sition at the sea bottom and will therefore not be Carbonates also accumulate on submarine highs and
incorporated in the geological record. The total carbonate platforms which are surrounded by deeper
biological production of carbonate is approximately water that traps the clastic sediments. The Bahamas
platform is a good example of such a setting. Reefs 50

Chloralgal
commonly grow at the edge of deep waters where they field Chloralgal
are exposed to light, high wave energies and a steady extends
supply of nutrients through upwelling. 45 to at ?
least 56‰
Minimum salinity ‰
During periods of global transgressions (high- Chloralgal-
stand) more carbonate is deposited on the continental extended
chlorozoan
shelves leaving less carbonate to be deposited in deep- 40 ?
water environments. In the stratigraphic record an
Chlorozoan
abundance of carbonate rocks is seen from the Ordo-
vician to Devonian periods and from the Cretaceous. 35
Foramol
This is not because more carbonate sediments were
precipitated during these periods, but because more of
what was precipitated was preserved on the cratons. 30
0 10 20 30
During transgressions there is little supply of clastic Maximum temperature °C
material that could dilute the carbonate precipitated.
Fig. 5.40 Relationships between salinity-temperature annual
These carbonate deposits therefore generally contain
ranges and occurrences of skeletal grain associations in modern
little clastic material. However, during sea level low- shelf carbonate environments (modified from Lees 1975)
stand more of the carbonate becomes mixed with
clastic sediments to produce carbonate deposits with
It is important to remember that the distribution of
a fairly high terrigenous content.
warm-water carbonate facies is not only governed by
latitude, but also depends to a large extent on
ocean circulation patterns. On the western side of the
5.7.2 Major Controls on Carbonate Atlantic Ocean, warm-water carbonate facies with
Sedimentation reefs and oolites are found up to about 30 N from
the Equator, while on the eastern side, off West Africa,
Temperature and salinity are the main factors affecting
the ocean is generally too cold, even close to the
shallow marine carbonate-secreting organisms. On
Equator. This pattern is due to the east-west equatorial
this basis it is possible to distinguish between three
winds producing upwelling on the eastern side of the
principal skeletal grain associations; chlorozoan,
ocean, and accumulation of warm surface water on the
foramol and chloralgal associations (Lees and Buller
western side.
1972) (Fig. 5.40). We may also distinguish between
warm-water carbonate - photozoan, and colder water
carbonates - heterozoan (James 1997). 5.7.2.2 Foramol Association
Where the seawater temperature range is 0 C up to
5.7.2.1 Chlorozoan Association about 15 C, the sediments are dominated by benthic
The chlorozoan association is found in warm shallow foraminifera and molluscs, together with a somewhat
seas at low latitudes as in the Bahamas and the Persian minor contribution from echinoderms, barnacles,
Gulf. Such warm-water carbonates are mostly derived bryozoans, calcareous red algae and ostracods. This
from algae, or from benthic organisms living in sym- association is referred to as foramol or heterozoan
biosis with algae. Warm-water carbonates are formed (James and Clarke 1977). The foramol association
in areas with clear water and little clastic supply. The thus comprises temperate and cold water carbonates
reef-building corals in warm areas (hermatypic corals) currently being deposited e.g. the North Sea and the
live in symbiosis with algae and can only grow in Spitsbergen Bank northwest of Bear Island in the
shallow water with abundant sunlight. These corals Barents Sea. The Spitsbergen Bank is 30–100 m
are very vulnerable to cold (and very warm) waters deep and surrounded by deeper channels so receives
and tolerate only a quite narrow salinity range. There- little clastic matter (Bjørlykke et al. 1978). The area
fore the chlorozoan association does not exist where lies on a cold oceanic front where cold currents from
the minimum surface temperature falls below 15 C the north and east mix with the warmer water of the
and the salinity ranges lies outside 32–40‰. Atlantic Ocean. This results in powerful currents down
to 80–100 m and high productivity of molluscs, ben-

a
thic foraminifera, bryozoans and barnacles. The sea-
floor sediments are characterised by a Holocene
80–90% pure carbonate sand deposited across the top
of Pleistocene moraines (Bjørlykke et al. 1978). Simi-
lar carbonate deposits are encountered on other banks
off North America, for example Grand Bank, and
along the coasts of western Scotland and Norway.
The biological precipitation of carbonate is not
dependent on the water being warm or saturated with
calcium carbonate. On the contrary, we often find the
highest productivity in cold areas because the water b
there is richer in nutrients, particularly in areas of
upwelling. Cold seawater is undersaturated with
respect to carbonate minerals so that the skeletal mate-
rial begins to dissolve as soon the organisms die. Net
accumulation of skeletal material on the sea bottom is
possible because carbonate deposition merely requires
that the biological production rate of carbonate
skeletons exceeds the rate of dissolution. Skeletal
material with a primary aragonite or high-Mg calcite
composition will dissolve faster than that composed of
low-Mg calcite, because of the lower pH due to Fig. 5.41 Corals from the modern Great Barrier Reef. Mainly
increased CO2 in the cold water. brain corals and stag corals
In the interpretation of ancient limestones we
should not always assume that we can use the well- function of the ecological environments of the reef-
known warm water carbonate models, but also con- building organisms:
sider the possibility that we are dealing with cold 1. Reef-building corals require warm surface water
water carbonate sediments at higher latitude. and their distribution today is therefore limited to
the lower latitudes. However, palaeogeographical
reconstructions must take into account that surface
5.7.3 Modern Environments of Carbonate water temperature is not a simple function of lati-
Sedimentation tude. This is illustrated by the coasts of West Africa
and along the western side of the American conti-
5.7.3.1 Reefs nent. Where there is upwelling of cold water the
From a sedimentological point of view a reef may be surface layers are in most cases too cold for reefs to
defined as a laterally restricted body of carbonate rock form, even near the equator. Off the coast of East
whose composition and relationship with the Africa and in the Caribbean, on the other hand, the
surrounding sediments suggest that the bulk of its water is warm because of prevailing onshore winds,
biota were bound together as a framework during and coral reefs are abundant.
deposition, maintaining and developing a positive 2. The hermatypic corals which build reefs live in
topographic structure on the sea bottom. Modern reef symbiosis with algae and require sunlight. The
structures are dominated by hermatypic corals and reefs are therefore sensitive to changes in sea
calcareous red algae (Fig. 5.41), but the biotic compo- level since reef organisms tolerate neither exposure
sition of the reef structures through the Phanerozoic nor “drowning” below the photic zone. The Holo-
has been fairly variable (Fig. 5.42). Generally, all cene transgression about 10,000 years ago raised
these different reef types show similar facies patterns sea level about 100 m in a few thousand years, but
to modern coralgal reefs. Recent reef formation, and most reefs managed to grow quickly enough to
the shape of the reef complexes, is essentially a keep pace with the rising sea level.
Periods Biotherms Major skeletal elements

Corals
Tertiary
Rudists
Bryozoa
Cretaceous Rudists Corals, stromatoporoids
Sponges
Corals stromatoporoids
Jurassic
Corals Stromatoporoids
Triassic Tubiphytes Corals, sponges
Sponges, Tubiphytes, calcareous algae
Permian Calcisponges, fenestellid bryozoa, corals
Phylloid algae Tubular foraminifers
Pennsylvanian Reefs Palaeoaplysina Tubiphytes
Bryozoa
Mississippian Fenstrate bryozoa
Devonian
Stromatoporoids Corals
Silurian
Stromatoporoids Bryozoa
Ordovician & corals
Sponges Calcareous algae
Reef
mounds Calcareous algae
Cambrian Archaeocyathids
& calcareous algae
Precambrian
Fig. 5.42 Simplified stratigraphic column showing how the dominating biota in reefs and reef mounds has changed through the
Phanerozoic. Gaps indicate times when there appear to be no reefs or reef mounds (modified from James 1983)
3. If sea level rises faster than the coral reef can grow, temperature is combined with low mud content
the reef may drop below the photic zone and and good water circulation.
“drown”, but this is very rare because they can 7. The reef facies is built up as a wave-breaking struc-
gain height rapidly. ture, with powerful networks of corals which are
4. Reefs provide a favourable ecological environment braced by an encrustation of coralline algae. Red
for animals which are not part of the reef structure algae are massive and strong and have an important
itself, but which live on other organisms. function in supporting the reef structure so that it can
5. Many organisms in coral reefs live by filtering withstand the strong waves. However, parts of the
water to trap planktonic organisms and organic most exposed parts of the reef structures can be
material. If the water contains too much siliciclastic mechanically destroyed by wave action during
mud, the clay minerals can choke the filtering heavy storms. In this way reef fragments of various
organs so that the organisms die. Corals are parti- sizes may fall down the fore-reef slope, be
cularly sensitive to the clay content in the water and transported up onto the reef flat or be deposited in
can only live in clear water. The addition of clay, gaps and cavities in the reef framework. This
for example from a delta, will kill a coral reef. mechanical destruction is aided by the boring and
Pollution will have the same effect. grazing action of different bioeroding organisms.
6. Clear sea water, however, is usually very poor in The bio-erosion can be so extensive that many of
nutrients, so the organisms in a reef are dependent the primary structures of fossil reefs are destroyed.
on good water circulation to obtain enough food. Boring mussels are particularly effective, excavating
Consequently, reefs tend to grow on the edge of elongated cavities about 1 cm in diameter in corals,
ocean basins, or as structures projecting high up algae etc. Boring sponges, worms and algae are
from deeper waters. In this way, high water also important reef-damaging organisms, typically
Fig. 5.43 Section through a reef which is developing out into a basin. Note the distribution of various reef facies and their
diachronous nature (modified from Bjørlykke 1979)
forming minutely thin but densely spaced holes

which destroy the skeletal material. Coral reefs pro-
vide a very favourable environment for a variety of
fish, and the cavities in the structure offer protection
against predators, but some species of fish also live
by literally eating the reef. They take small bites,
and obtain nutrients from the surface. In this manner
they contribute to the bioerosion of the reef and the
deposition of lime sediment on the sea bed.
A reef grows by virtue of the ability of its reef-
building organisms to keep pace with the biological
and mechanical processes which are continually
breaking it down. Reefs will grow fastest on the out- Fig. 5.44 Aerial photograph showing reef facing the deep
side, where wave energy and nutrient supply are ocean and the protected lagoon, Bahamas. (Photo G.M.
greatest. In the case of barrier reefs, lagoons form on Friedman)
the landward side and contain a mixture of carbonate
sand and fragments which have been transported from most important source of carbonate ooze on the
the reef facies during storms, together with skeletal Bahama Bank; chemical precipitation is of minor
material derived from organisms living in the lagoon importance. Other characteristic elements in the
(Fig. 5.43). The lagoons are usually only 2–6 m deep lagoon fauna are bivalves, gastropods, echinoderms
and protected from the highest waves by the reefs and annelids. Of particular sedimentological impor-
(Fig. 5.44) and contain a rich flora of calcareous tance are crabs and shrimps (Callianassa) which
green algae such as Halimeda and Penicillus. These dig tunnels and churn up the sediments, destroying
are low, bush-like forms which cannot withstand high the primary structures. Their tube-like trace fossils
energy waves and require the protected environment (e.g. Ophiomorpha) are typical of shallow marine
of the lagoon. However, they secrete small (c. 1 μm) sediments. The stability of the bottom sediments in
needle-like crystals of aragonite. When the algae die, the lagoons is enhanced by the presence of plants, e.g.
the aragonite needles are released and form carbonate Thallassia grass which covers much of the sea bottom
ooze. Green algae are therefore considered to be the on the Bahama Bank.
5.7.3.2 Reef and Carbonate Platforms subjected to pollution from clay minerals produced by
Surrounded by Deep Oceans weathering. Most of the small amounts of clastic
Volcanic sea mounts reaching up to sea level are ideal sediments deposited on Bahamas are aeolian dust
places for carbonate sedimentation since all the clastic blown all the way from the Sahara region in Africa.
sediment will be trapped in the surrounding deeper Carbonate platforms like the Bahamas represent the
water (Fig. 5.45). Smaller sea mounts may host circu- more or less continuous build-up of carbonate on the
lar atolls while larger areas of volcanic rocks may seafloor from Jurassic, Cretaceous and Tertiary times
form platforms. Carbonate production may then keep right up to the present. As the seabed has subsided,
pace with the seafloor subsidence over long periods of carbonate sedimentation has built up the area so that it
geological time. Even during the Pleistocene most of has remained in the photic zone. This has led to it
the carbonate platforms have been able to survive sea being surrounded by deeper areas of sea which have
level changes of more than 100 m without drowning. acted as traps for clastic sediments from neighbouring
Nearly all the carbonate production occurs in the pho- continents. East of Andros Island the edge of the
tic zone (<100 m) and most of it occurs within the carbonate platform gives way to a very steep sub-
upper 20 m. marine slope (20–30 in many places). At the Tongue
Also the Maldive Islands in South Asia are car- of the Ocean it slopes down to 2,500 m, and east of the
bonate platforms. They have grown to form a submarine little Bahamas Bank that lies to the north, it descends
mountain 3 km above the surrounding seafloor and its to 4,500 m in the Atlantic Ocean. Throughout its
growth has kept pace with the subsidence and changes existence, the eastern side of the Bank has been
in sea level. reinforced by reef structures which are solid carbonate
The Bahamas Bank is one of the best researched rock capable of forming steep submarine slopes with-
areas of modern carbonate banks in the world and out submarine slides. Diving with submersibles has,
covers a very large area, about 700 km N–S and however, revealed great blocks of limestone at the
200–300 km wide (Fig. 5.46). The greater part of this foot of the slope. The stresses in the rock in a steep
area is less than 10 m deep. There is deep water on the slope such as this may have contributed to fracturing,
west side towards the Mexican Gulf, so there is no releasing the blocks from the rock wall. Near the base
clastic sediment supply from land in that direction. of the slopes there are very high differential stresses due
Sediments on the Bank are almost pure carbonate, to the high vertical stress from the overlying carbonate
and the carbonate-producing organisms are often not rocks compared to the water pressure in the horizontal
Lagoon and
carbonate bank
Prevailing wind
Ooids
Sediment
input
Upwelling
Warm water Cold water
Reef
Bioclastic
Clastic trap carbonate
4500m 4500m
Carbonate
turbidites Pelagic Pelagic
carbonate carbonate
Dissolution of carbonate
Gulf of Mid-Atlantic North-West

Bahamas
Mexico Ridge Africa
Fig. 5.45 Distribution of carbonate sediments across the Atlantic Ocean (modified from Bjørlykke 1989)
a 80° 78° 76°

b 79°
Northw
78°
0 km 100 26° est Prov

200 m idence C
hannel
200
m
200 m
Little Bahama Bank

Bimini Islands
Florida Berry Islands
Great Abaco
Grand Bahama
Pr
ov
ide
nc
eC 26°
ha
nn Joulters Cays
Bimini Islands el
New Providence
Eleuthera
New providence 25°
To
ts
ai
20
ngu
r
20
0m
St
200 m
0m
a
Andros
e
id
To n
of t
or
Island
200
Cat Island
Fl
200 m
he
ue o
m
Gr
Ex Andros Island
O
S an
um
ea
cea
f th
aS
t
tar
ou
n
e Oc
nd 24°
en
Cay Sal Bank

Ba
Ch
e an
ha
ann
Nic 20
a
e
hol 0 Great Exuma

l
m
as C
han Long Island
nel
Ba
nk
24°
20 0 m Lithofacies
Reef
Coralgal
Oolitic
Great Ragged Island Grapestone
Cuba Oolite
22°
Pellet mud 0 20 km
Mud
c W Windward E
Barrier rim Shelf lagoon lagoon
Sea level
Oolitic Grapestone Mud facies Grapestone Oolitic Reef
facies facies facies facies
Coralgal Coralgal
facies facies
Fig. 5.46 (a) Map of the Bahamas Platform with surrounding require high energy, are distributed along the edge of the basin,
areas. Note the deep ocean basin on the landward side. (b) Part where wave power is greatest. Reef facies, which require the
of the Bahamas platform showing the distribution of recent highest wave energy, only develop on the east side against the
carbonate sediments. (c) Simplified E–W section across the prevailing winds (modified from Bathurst 1975)
Bahamas platform. Note that the coral and oolite facies, which
direction. This may cause fracturing and the release of the last 10–12,000 years are not more than 3–4 m thick
carbonate slabs from these submarine slopes. A 2 km and rest unconformably on well-cemented, Pleisto-
thick carbonate sequence represents a load (stress) of cene carbonate rocks.
about 40–50 MPa assuming an average density of Sedimentation on the Bahamas Bank reflects the
2.0–2.5 g/cm. In the water column at the same depth climatic and bathymetric conditions. The temperature
the water load is only 20 MPa and the differential stress of the surface water varies from about 20–22 C in
is then about 25 MPa. winter to 30–32 C in summer. Because of the limited
When the Bahamas Bank, and other carbonate exchange between water overlying the Bahamas plat-
banks in the world, was exposed by the >100 m sea form and the surrounding ocean, the water in the
level drop during the glacial periods, it became interior of the Bank, particularly in summer, has a
lithified by percolating meteoric water which higher salinity than normal – up to 40‰. In winter
dissolved the aragonite and precipitated it as calcite the salinity is reduced by increased water circulation
cement. The loose carbonate sediments deposited over with the surrounding ocean, by rainwater and by
runoff from Andros Island. Since the prevailing winds

are from the east, where we also have deep water,
wave energy is strongest on the east side of the plat-
form. We therefore find reef facies along the east side
of the platform, but not along the west side. However,
corals and algae (coralalgal facies) are also present on
the west side of the bank, but they do not form proper
reef structures there.
Oolite banks form in shallow areas with constant
wave agitation and strong tidal currents. In water
deeper than the normal wave base, we find large
bedforms of oolites and other carbonate sand which
are only mobilised by major hurricanes. Dunes or Fig. 5.47 Arial view of supratidal environments with tidal
sandwaves with a wavelength of about 50–100 m, channels. Bahamas. The tidal range is reduced in the inner part
mapped from aerial photograph surveys, show no of the platform
sign of having moved in 20–30 years. Extensive
areas of seabed where the wave energy is less than
pellets. They form small grains (0.1 mm) consisting of
on the oolite banks are covered with algal mats. The
clay and silt particles so that the “recycled” lime mud
algal threads forming this mat help to protect the
now behaves like fine sand when transported by
sediments from erosion, inhibiting bed transport and
currents. The faecal pellets consist of mud which has
current ripple development. They also contribute,
passed through the alimentary canal of marine
through photosynthesis, to the creation of a chemical
organisms such as gastropods, bivalves and annelids.
environment with low CO2 concentrations which
Crustacea, such as the shrimp Callianassa, are also
may cause local precipitation of carbonate within the
important pellet producers.
mat.
In the middle areas, sedimentation of carbonate
mud facies prevails. This is because wave energy is 5.7.3.3 Carbonate Platforms and Reef
reduced along the edge of the Bank, and the whole of Environments Along the Coast;
the shallow area within is a low-energy environment. the Persian Gulf
Tidal currents are also reduced in the shallow water Attached carbonate platforms can only form if there is
because most of the tidal energy is dissipated in little sediment supply from land. In dry areas, in par-
overcoming the friction against the bottom and the ticular, there will be very little runoff from land which
tidal range drops from about 0.7–0.8 m along the could add clastic sediments. The drainage pattern also
edge to practically zero in the centre. The most impor- plays a major role. During transgressive periods a
tant sediment transport mechanism is probably shallow sea will invade the low-lying land, raising
hurricanes. These can mobilise normally stable the base level of the rivers. In this way the clastic
sediments, forming large dunes or sandwaves. Large sediment supply is trapped in the lower valley reaches
oolite banks may migrate along the edge of the plat- and much of the shelf will have clear water dominated
form, and storms also inflict some erosion of the reefs. by carbonate sedimentation.
Towards the centre of the platform wind stress can The Persian Gulf (Fig. 5.48) provides an important
cause changes in sea level of up to 3 m, and create present day model of a carbonate-producing environ-
great turbulence which brings large quantities of mud ment which is fundamentally different from the
and sand into suspension. Some of this material will be Bahamas Bank. Whereas the Bahamas Bank is
transported off the platform and down the submarine surrounded by deeper water, and therefore represents
slopes. a pure carbonate environment, the Persian Gulf lies
The areas west of Andros Island, which are the in a fold zone between the alpine mountain chain of
shallowest and best-protected against storms from Iran to the north and the stable Arabian shield to the
the east, consist mainly of lime mud (Fig. 5.47). south and southwest. The Gulf is at the most only
Over wide areas the carbonate mud is dominated by 80–90 m deep, and a delta is being built out into the
a Iraq
Iran
Ku
wai
t
Iran
O MA N
Saudi arabia
Qatar
t
as
Co
United
Arab Oman
Emirates
Trucial Abu Dhabi
b Oolitic sand Aeolian dune-beach

(tidal delta) (mainly oolitic)
Holocene sabkha
Pelletoid, Gastropod sand
(supratidal)
Muddy skeletal sand Tertiary outcrop
(channel)
Lagoonal mud with foraminifera
and gastropods
Organic reefs
n
Algal mat
retio
(intertidal) acc
and
rt
spo
tran
iment rier
Sed bar
l
sta
Coa
s
delta
Tidal
Sed
s
on
ime
o
A
ag
nt t
l
H
ail
tal
K
Coas A
B
A
S
C O A S T A L
Ca. 10 km
Fig. 5.48 (a) The Persian Gulf showing the facies distribution of Recent bottom sediments. (b) Map of the carbonate facies in the
Abu Dhabi area (modified from Wagner and van der Togt 1973, Purser and Evans 1973)
northwestern end by the Euphrates and Tigris. Clastic the continental shelf, biogenic carbonate accumulates in
siliceous sediments are also being added from the the form of shells of foraminifera, molluscs, ostracods,
north, from the mountains in Iran with abundant bryozoans and echinoderms. Especially in shallow water
limestones and dolomites. Much of the sediment there- environments the skeletal material can be heavily
fore consists of impure carbonates (marl). Only on the affected by post-mortem micritisation if exposed on the
south (Trucial Coast) side are there purer carbonate sea bottom for some time. Along the outer edge of the
sediments, because there is very little runoff from the shelf are reefs, built of coral species that are particularly
deserts on the Arabian shield. Here too, though, there well adapted to the high salinity.
is some supply of clastic material, particularly through Tidal channels connect the lagoons with the gulf.
aeolian transport from the desert. Sedimentation has At the mouths of the channels there are ebb-tidal deltas
been proceeding in a marine basin/embayment in this with oolite banks in the shallowest parts (<2 m). In
area since Mesozoic times. addition to oolites there are grains of bioclastic mate-
The sea temperature in the Persian Gulf varies from rial. Tidal deltas of this sort would form good oil
20 C in winter to 34 C in summer, and in the shallow
reservoirs because of the high primary porosity, and
areas it can be even hotter. The salinity in the open in the Mesozoic series we find similar reservoir rocks.
water is around 39–42‰, and may be higher due to Shoreface deposits are not as well sorted as we would
evaporation in the lagoons. The tidal range varies from normally expect as they consist of fine sand and lime
0.5 to 2 m. The innermost supratidal area consists of a mud. This is possibly because the shoreline is
10–15 km broad marginal belt covered by stabilised by plants which bind the sediments.
cyanobacteria and salt deposits. This belt is called Gastropods are particularly important in the
sabkha in Arabic, and the word has become a geologi- lagoons and the intertidal environment. On the
cal term. beach, shrimps and crabs burrow and disturb the
Because of the high temperature and salinity few lamination in the sediments. The lagoons are
organisms can live on the sabkha, which is dominated surrounded by swamps with bushes and mangroves,
by algal mats (stromatolites) that form a crust of or algal flats. Landward of the algal mat facies,
precipitated carbonate and gypsum. The stromatolites which is most typically developed at the high water
may grow in layers parallel to bedding or develop mark, the sabkha facies continues. The degree of
dome-shaped, columnar or irregular structures. The evaporation is higher here in the supratidal zone
stromatolite structures are often broken into polygons and dilution with seawater rarer. The sulphates,
which are separated from each other by sediment- especially gypsum or anhydrite, are precipitated
filled cracks. The cracks are the posthumous reflection within the sediment near the groundwater table.
of shrinkage cracks as a result of drying. The sabkha flats pass landwards into drier areas
The slope of the sediment surface here is only 0.4 dominated by aeolian sand.
m/km and during powerful storms water is driven in The facies distribution we have just described, from
over the sabkha and later evaporates. Anhydrite, dolo- the open marine lagoon to the supratidal environment,
mite, magnesite and halite may be precipitated, though will form a characteristic vertical sequence in the
the halite is easily dissolved again. The precipitation event of relative sea level fall and progradation.
of gypsum and anhydrite leads to the water developing Periods with transgressions and subsequent regressive
a higher Mg2þ =Ca2þ ratio, which favours the forma- outbuilding will result in a series of sequences
tion of dolomite and magnesite. (cycles), starting with shallow marine (subtidal)
In the Persian Gulf green calcareous algae such as sediments and culminating with evaporites (anhydrite)
Halimeda and Penicillus, which are important producers at the top. These cycles are also found in the Mesozoic
of lime mud in the Bahamas, are for the most part absent. and the impervious anhydrite beds forms an ideal cap
Chemical precipitation is therefore probably the most rock above oil reservoir rocks. Carbonate sand facies,
important process, particularly during periods when pellet facies, and particularly oolites, make good
diatoms proliferate and raise the pH by consuming reservoir rocks, while the marine muds are source
CO2 and thereby reducing the solubility of calcite. On rocks.
5.7.3.4 Evaporite Basins Therefore, carbonate sediments commonly become

In ocean areas with normal salinity, practically all concentrated on top of submarine highs where the
carbonate deposition is through biological precipita- siliciclastic sedimentation rates are low. This is often
tion. This is because the biological precipitation is observed on seismic lines where the reflections
rather efficient so the sea water does not become become stronger on the top of submarine positive
saturated with respect to the most common carbonate tectonic structures.
minerals like aragonite and calcite. In areas with In the deep ocean much of the pelagic carbonate
somewhat higher salinity, however, for example in production is dissolved as the particles settle through
the Persian Gulf, there may also be chemical precipi- the water column. The sedimentation rate is a function
tation of calcium carbonate. This is because there are of productivity in the upper water layers minus solu-
fewer organisms to precipitate carbonate and because tion as the dead organisms sink towards the seafloor.
of evaporation concentrating the sea water. Here too The dissolution of skeletal material is due to
this precipitation may nevertheless be linked to undersaturation of CaCO3 in deep ocean water
biological factors. Periods with algal blooms in the because this cold water can dissolve more CO2 than
surface waters involve photosynthesis and consump- warm surface water. The combination of low
tion of CO2. This raises the pH, creating oversatura- temperatures and increasing hydrostatic pressure
tion and thus favourable conditions for chemical with depth involves an increase in the pCO2 and
precipitation. decrease in the pH. The CO2 is produced by respira-
Carbonates make up only a small percentage of tion and decay of pelagic organisms in the deep ocean.
the salt precipitated when seawater evaporates to Below the depths where the rate of dissolution is equal
dryness. However, they are among the least soluble to the rate of sedimentation of carbonate, no carbonate
of the common salts in restricted ocean basins where sediments accumulate. This is called the carbonate
the salinity is too low for the more soluble salts to compensation depth (CCD) and varies from 1–2 km
precipitate (e.g. NaCl). Carbonates together with in cold water at higher latitude to 4–5 km in the warm
sulphates like gypsum or anhydrite often form thick water equatorial regions.
evaporite sequences. In highly saline basins with more Since the planktonic carbonate organisms are very
intense evaporation and restricted circulation, the small, pelagic carbonate deposits form a fine-grained
water will become enriched in Mg2+ and dolomite ooze of clay- and silt-size particles, with occasional
will form, perhaps also magnesite. larger fossil fragments. Large areas of the South
Atlantic and Pacific are covered by sediment
5.7.3.5 Carbonate Turbidites on Slopes containing more than 50% CaCO3 from planktonic
Skirting Carbonate Platforms organisms. Foraminifera and coccolithophores form
Carbonate sand and mud which is stirred up during the most important deep-sea carbonate deposits
storms over shallow water may be transported over the (Fig. 5.14).
shelf edge to continue downslope as turbidites or Coccolithophores live mainly in the photic zone.
debris flows. Carbonate platforms may rise 2–3 km In areas of high productivity, for example in the fjords
above the surrounding seafloor and the slope may be of Norway, the concentration of coccolithophores may
very steep (20–30 ) because of the hard and stable be several millions per litre, but 50,000–500,000 is a
carbonate rocks. more normal level. Although they consist of low-Mg
calcite, their size makes them relatively soluble in
5.7.3.6 Pelagic Carbonate cold water. In consequence, although production is
Pelagic carbonate deposits consist largely of plank- greatest at high latitudes, it is only at lower latitudes
tonic organisms (coccolithophores, foraminifera, that large quantities of coccolithophores are able to
pteropods etc.) which live in the upper part of the accumulate on the seafloor.
water column, sinking to the bottom when they die. Shallow, warm seas with little other carbonate pro-
How clean the carbonate deposits are depends on how duction provide particularly favourable conditions
much other biological sedimentation there is, e.g. for the deposition of purer coccolith deposits. The
from siliceous organisms like diatoms and radiolaria, seas of northwest Europe in Cretaceous times were
and how rapidly clastic sedimentation takes place. a good example. The climate in the Mesozoic was
undoubtedly considerably warmer than today, and carbonate. Thicker beds of carbonate below lake floors
northwest Europe also lay further south. Chalk forms can be associated with longer term climate variation.
a characteristic rock which is exposed in Denmark, In Lake Victoria, the sediments older than about
South England and France, and continues under the 12,000 years have high carbonate content. This is
southern and middle sections of the North Sea. It is because the lake nearly dried up during the last glacial
missing in the north, possibly for palaeoclimatic period, when rainfall in the region was lower than
reasons, although it is found in northeast Ireland. now.
Chalk sediments were probably deposited in water
depths of not more than 100–300 m, mostly below 5.7.3.8 Calcareous Tufa Deposits
the photic zone. Since chalk is a micritic limestone, and Travertine
one would not expect it to form a suitable reservoir Tufa is the name of a porous and spongy calcareous
rock. The Ekofisk and associated fields are in fact the deposit common in limestone areas, usually at the base
world’s only major oilfield in such rocks, and the low of slopes where groundwater emerges at the ground
permeability of this fine-grained lithology creates surface. Accreted in thin layers which often incorpo-
problems for production. rate vegetation, it can build up into deposits several
Pelagic calcareous algae such as coccolithophores metres thick. Travertine is a more massive, relatively
did not become common until the late Jurassic and dense and sometimes finely banded and laminated
early Cretaceous. Consequently we do not have chalk carbonate deposit associated with freshwater springs
deposits from older periods. This is an example of or precipitated as speleothems in caves. Both varieties
where the rock type is totally dependent on which are precipitated from freshwater supersaturated with
organisms were precipitating carbonate. During the respect to calcium carbonate. Groundwater flowing
Palaeozoic most carbonate production took place in through carbonate rock contains CO2 at a higher par-
shallow water, as there were no planktonic calcareous tial pressure than at the surface because of the lower
algae to form deepwater pelagic carbonates. ambient temperature and the higher pressure that is
due to the weight of the overlying water column. The
5.7.3.7 Lakes and Inland Seas increased partial pressure of CO2 promotes increased
Carbonate sedimentation in lakes depends on the rate dissolution of calcium carbonate from the host
of weathering and supply of Ca2+ from older calcareous rock. When the water flows out of the rock at the
sediments and calcium-bearing silicate rocks in surface, or reaches a cave, it will regain equilibrium
the drainage basin. If there is an ample supply of Ca2+ with atmospheric pressure, causing degassing of CO2.
The loss of CO2 entails that the water may become
and CO23 also lacustrine sediments may contain
supersaturated with regards to calcite, followed
considerable amounts of carbonate, and particularly
by precipitation of this mineral. This is particularly
at lower latitudes we find pure carbonate deposits. In
effective in summer when the water will quickly warm
cold lakes the solubility of carbonate is high and the
up once it emerges. Exposure to light will also cause
carbonate content in the bottom sediments will mostly
biogenic precipitation (photosynthesis) by algae.
be limited to a few species of bivalves and gastropods.
Temperate lakes often accumulate calcareous muds
called marls, which in some lake sediments contain 5.7.4 Meteoric Water Flow
dolomite. Algae and certain higher plants often play an and Diagenesis
important role in carbonate production in lakes.
The Dead Sea is a good example of an inland sea The introduction of meteoric water has a profound
where carbonate is precipitated chemically due to effect on carbonate sediments and their potential as
strong evaporation. In Africa and other tropical reservoir rocks.
areas, lakes will be subject to seasonal evaporation A proportion of the rainwater (meteoric water) fall-
to dryness, and we may find alternating layers of ing on land infiltrates into the groundwater. The flow
biogenic carbonate and chemically precipitated of groundwater ultimately is limited by the rate of
carbonate. Here, too, algal blooms in the surface recharge by rainwater, which determines the water
water layer play a major role in precipitating table gradient. As long as the water table is above
Water table
Vadose zone
Sea level
Fresh (metoric) water flow
Freshwater
phreatic zone
Mixing zone
Marine phreatic
Fig. 5.49 Cross-section of an ideal permeable carbonate sand groundwater table are permanently water-saturated. The fresh
island showing the distribution of major diagenetic interstitial waters float on the denser saline waters beneath.
environments in the shallow subsurface. The vadose zone is There is a mixing zone between the two water phases (modified
situated near the surface above the water table, and the pore from Scoffin 1987)
spaces are occupied by water and air. The pore spaces below the
sea level the groundwater has the hydrodynamic undersaturated with respect to calcite, in which case
potential to flow beneath the beach and out into the caverns are likely to develop.
basin beneath the seafloor, floating on top of the more When meteoric water flows through recent
saline basin porewater (Fig. 5.49). The freshwater lens carbonate sediments it will be undersaturated with
is floating like an iceberg in the sea. With a ground- respect to aragonite but become rapidly supersaturated
water density of 1.00 g/cm3 and the more saline water with respect to calcite. Aragonite will therefore
1.025 g/cm3, the ratio between the groundwater head dissolve first and calcite will precipitate (Fig. 5.50a).
and the depth of freshwater penetration is theoretically Gradually the meteoric water will reach equilibrium
1=ð1:025 1:00Þ ¼ 1=40. A groundwater head of just with low-Mg calcite, and calcite cement in the form of
10 m can drive freshwater to a depth of up to 400 m large crystals (block-shaped cement) may be
below sea level. Shallow water carbonates deposited precipitated (Fig. 5.51). This cement is very different
in coastal environments and around islands will thus in from marine cements precipitated from modified
most cases be flushed by fresh groundwater after depo- seawater (without sulphate).
sition. The diagenetic effects are greatest at shallow On land, the sediments above the water table are
depth where the flow rates are highest. located in the vadose zone, where the pores are alter-
However, coastal carbonate environments are usu- nately filled with water and air as a consequence of
ally rather dry and while carbonate platforms may intermittent meteoric water percolation. Partial desic-
have more rainfall, the islands on them may be small cation results in an unequal distribution of the
compared to the size of the platform. Both these porewater with it primarily held near grain contacts,
factors tend to reduce the flux of meteoric water into by capillary forces; as a result there will be a preferen-
marine carbonate sediments, although it may still be tial cementation of pore throats giving a rounded pore
very significant, particularly when the sedimentation geometry. This cement type is called meniscus cement
rate is low. More distal and pelagic facies may avoid (Figs. 5.50b and 5.51). Porewater will also collect on
this flushing altogether. On land and in the nearshore the underside of grains as pendant droplets and precip-
parts of the basin meteoric water may also be itate cement in this form, called pendant cement. Both
the Holocene (postglacial) deposits, which normally

are less than 2–3 m thick.
5.7.5 Shallow-Marine Diagenesis
5.7.5.1 Introduction
Lithification is the process which transforms loose
sediment into solid rock. It occurs through new
minerals (cement) being precipitated in the pore
spaces binding together the primary particles. To
cause carbonate cement to be precipitated, we must
have porewater which is oversaturated with respect to
a carbonate phase.
In general diagenetic processes are driven by a pro-
gression towards more mechanically stable grain
packings and more thermodynamically stable mineral
assemblages. This is also the case for carbonate
sediments. During progressive burial the increasing
overburden stress causes denser packing of grains so
that the porosity is reduced. The reduction in porosity is
then a function of the effective stress that may be
expressed as the compressibility of the rock. Mechanical
compaction is in principle instantaneous but there is
usually some additional compaction with time at the
same effective stress. This is called creep.
The mechanical compaction of carbonate sand and
Fig. 5.50 (a) Mould after a dissolved mollusc (large, green- mud follows the same principles as for terrigenous
stained pore in the middle). After dissolution of the aragonitic sand and clay. In the case of carbonate sediments
shell the void is outlined by a micritic envelope. There is only a however, chemical processes involving dissolution
thin veneer of calcite spar precipitated in the mould at this early
diagenetic stage. Thin section, plane polarised light. Quaternary,
and cement precipitation are much more important at
Rhodes, Greece. (b) Meniscus carbonate cement is low temperatures. This is because the kinetics of car-
concentrated at grain contacts resulting in pore rounding. This bonate reactions are much faster at low temperature
cement is diagnostic for freshwater vadose cementation. Pri- than is the case for silicate reactions. The prediction of
mary porosity is stained blue. Thin section, plane polarised
light. Quaternary, Rhodes, Greece
porosity and permeability in carbonate rocks therefore
also depends to a very large extent on chemical dia-
genesis at shallow depth. Porosity reduction in
meniscus cement and pendant cement are characteris- carbonates can therefore not be predicted solely on
tic of partial cementation in the vadose zone. the basis of effective stress because it also depends
At sea level lowstands, particularly in the Quater- on the primary mineralogical composition of the
nary, marine carbonate sediments were directly grains and textural relationships.
exposed to freshwater that caused rapid cementation. Cement formed in a marine environment is arago-
The sea level drop of more than 100 m during the nite or high-Mg calcite, which forms needle-shaped
glaciations exposed and cemented all carbonate crystals. High-Mg calcite can also precipitate as
sediments that had been in the photic zone during the micritic cement. Early marine aragonite cement may
preceding interglacial (highstand) periods. On modern grow as evenly distributed layers of aragonite needles
carbonate banks soft sediments are therefore limited to perpendicular to the surface of the grains. This is
a Finely crystalline
calcite cement Carbonate sands,
ooids, fossils etc.
Grain
Needles of aragonite Pore
or high-Mg calcite space
re
d po
nde
Rou Meniscus
cement c
Pendant
cement
Vadose
zone
Early marine
cement
(beachrock)
Water
table Reef Sea level
Meteoric
environment
Marine porewater
b
d
Block-shaped calcite cement
Mixture of freshwater
and marine porewater
(Dorag dolomitization)
Fig. 5.51 Types of carbonate cement as a function of diage- and precipitation a cement with pendant form will be
netic environment. (a) Water is held near grain contacts in the precipitated (pendant cement). (b) Phreatic calcite cement is
vadose zone by capillary forces, and as a result cements form in typically clear and blocky shaped. (c) Early marine cements
these places (meniscus cement). In this way the pores will commonly consist of an isopachous fringe of acicular aragonite
become more rounded due to the preferential cementation of or high-Mg calcite. (d) Dolomite can be precipitated from a
the pore throats. Water will also collect at the underside of mixture of fresh and marine porewater
grains as pendant droplets, and after multiple phases of drainage
called isopachous fibrous cement because a layer of rate will be. Calcite cement that grows out from echi-
uniform thickness is formed (Fig. 5.52). Isopachous noderm fragments will have a fast growth rate because
calcite cement may also be precipitated in meteoric both will have an identical lattice structure. The
(phreatic) porewater (Fig. 5.51). cement precipitates as syntaxial rims in optical conti-
The relative growth rate of the cement will depend nuity with the single calcite crystals of echinoderm
on the chemical composition of the substrate and its plates, such that twin lamellae and cleavage planes
microstructure. The greater the accordance between transgress skeletal plates and cement overgrowths
the substrate and the cement, the higher the growth (Fig. 5.53). This style of cementation, called syntaxial
When calcite replaces earlier aragonite or high-Mg

calcite by neomorphism, we sometimes see “ghosts”
of the earlier crystals. Radiaxial fibrous cement
mosaics frequently form as a result of replacement of
high-Mg calcite cement but also may form by direct
precipitation of aragonite. These are mosaics of calcite
crystals with optic axes that converge away from the
substratum (cavity wall) on which they grew. They
contain twin laminae which are convex towards the
substratum.
Calcite cement formed early in oxidised porewater
will contain practically no iron, since Fe3+ is not
soluble in the oxidised state. Only Fe2+ can substitute
Fig. 5.52 Beach rock dominated by foraminifera and for Ca2+ in the calcite structure, and the formation of
lithoclasts cemented with an isopachous rim of acicular arago-
ferroan calcite therefore requires reducing conditions.
nite cement. The pink areas are voids filled with stained epoxy.
Thin section as viewed in plane polarised light. Recent, Calcite precipitated in the sulphate-reducing zone
Bahamas is free of iron (non-ferroan), however, since all the
available Fe2+ will form sulphides. Calcite formed at
greater depths under reducing conditions will nor-
mally contain some iron and manganese, depending
on the availability of such ions in the porewater at the
time of formation availability of such ions in the
porewater at the time of formation (Fig. 5.54a,b) At
temperatures of about 100 C and above, iron-rich
carbonates like ankerite ðCaðMg; FeÞðCO3 Þ2 Þ become
increasingly stable and are often found in minor
quantities.
When a particle, e.g. a fossil, which consists of
aragonite, is dissolved and replaced by calcite, this
can occur by means of two different processes
(Fig. 5.55):
A. By complete dissolution of the particle leaving a
mould (secondary cavity) showing the outline of
Fig. 5.53 Crinoidal limestone. Dusty looking crinoid ossicles
the original grain. The mould can subsequently
with syntaxial calcite cement. The cement is in optical continu- become filled with low-Mg calcite, either entirely
ity with the echinoderm plates, with twin lamellae transgressing or partially depending on the supply of pore
skeletal plates and cement (modified from Greensmith 1978) fluid supersaturated with calcite. This secondary
precipitate has a highly characteristic texture. A
or epitaxial cementation, often completely obliterates narrow zone of small prismatic crystals develops
porosity in echinoderm-rich rocks. The volume of along the periphery of the cavity. The crystals
syntaxial overgrowth on echinoderm fragments is gen- become larger and more equidimensional towards
erally greater than the volume of cement crystals that the centre of the cavity; this is called a “drusy
have started their growth on multicrystalline fossil mosaic” (Bathurst 1975).
fragments or terrigenous material. In extreme cases, B. Through gradual dissolution of aragonite and
so much syntaxial cement may surround the echino- simultaneous precipitation of low-Mg calcite.
derm fragments that other grains will also become This reaction mechanism, neomorphism, will to
incorporated (poikilotopic crystal growth). some extent preserve primary structures.
5.7.5.3 Hardground
A combination of dissolution and precipitation of car-
bonate sediments close to the seabed may produce a
well-cemented surface called a “hardground”. This
can be recognised by evidence of borings and/or
encrustations of hard bottom faunal elements such as
calcareous algae, sponges, corals, serpulids, crinoids
and oysters. Hard grounds are best developed in areas
of slow sedimentation and high current activity. Long
exposure on the seafloor can also lead to impregnation
with minerals such as iron hydroxides, phosphorite
and glauconite.
5.7.5.4 Cementation in Modern Reefs

Modern reefs have a very high porosity, made up of
everything from small cavities to large caverns. It was
previously believed that this primary porosity was
preserved in older reefs. Boring and blasting into
modern reefs, however, has revealed that the primary
porosity is rapidly reduced because the hollows are
rapidly filled up with fossil fragments and lime mud,
and/or become more or less cemented by early marine
cement. Reefs can therefore only make good reservoir
rocks if secondary porosity develops through the dis-
solution of fossils or cement. This may happen by the
reef being exposed above sea level to groundwater
Fig. 5.54 (a) shows a well cemented limestone with an articu- percolation and also by flow beneath the seabed of
lated ostracod in the middle. From this transmitted light view it meteoric water from land. Aragonite or high-Mg cal-
is not possible to differentiate between the three different cite will then dissolve particularly easily and low-Mg
cement generations which are clearly visible with cathodolumi- calcite will precipitate. This diagenetic process may
nescence microscopy (b). Cement generations 1 and 2 have been
precipitated before compaction of the fossil, because these not significantly increase the overall porosity, but
cements are missing on the broken surfaces which are due to merely redistribute it.
early burial compaction of the sediment. The third cement Marine cementation commences early in reefs. The
generation (3) is a post-compaction precipitation filling the parts most exposed to waves, where water flux is
remaining primary porosity and cracks in the fossil. Thin section
as seen in plane polarised light and cathodoluminescence. Upper greatest, will undergo rapid marine cementation. In
Tertiary, Iran. (Photographs courtesy of Torleiv Torgersen) the interior and landward parts there are less marine
cement and more secondary porosity due to subaerial
exposure and freshwater flushing. Skeletal material of
5.7.5.2 Beach Rock primary aragonite may dissolve as calcite precipitates
The tidal zone on sandy beaches in tropical areas is in the open framework porosity. Thus the porosity to a
commonly cemented by aragonite or less commonly large extent becomes mouldic due to dissolution of
by high-Mg calcite to produce “beach rock”. On modern many of the reef-building organisms.
beaches the lithified zone may be about 0.5 m in thick- Early cementation of aragonite or high-Mg calcite
ness containing embedded bottles or other artifacts reduces porosity but the cementation produces a
indicating that cementation has been very rapid by geo- mechanically much stronger rock. The potential for
logical standards. Cementation takes place in situ preserving the remaining porosity is therefore higher
beneath a thin sediment cover and is due to a combina- than in uncemented carbonate sand. Furthermore,
tion of high flux of water due to wave action and rise in pressure solution may be reduced because the cement
ionic concentration due to evaporation of seawater as it increases the contact area which reduces the stress per
drains through the beach at low tide. grain contact.
Aragonite Dissolution and

Sparry calcite formation of cavity
(mould)
Indistinet
Gradual dissolution Incipient filling
primary
and precipitation of cavity
organic
lamination
Neomorphic calcite
Complete cementation
of the mould
Fig. 5.55 Two types of conversion of aragonite to calcite. Left: neomorphic replacement of the grain. Right: complete solution of
the grain and later precipitation of sparry calcite. If oil or gas is introduced before the moulds are cemented up the porosity may be
preserved (modified from Bjørlykke 1989)
5.7.5.5 Reefs as Reservoir Rocks burial depth (Schmoker and Halley 1982), although
The properties of reefs as reservoir rocks vary greatly the porosity/depth function varies greatly with the
and depend very much on the type of reef building lithology. The loss of porosity in carbonate sediments
organisms. These have varied through time and is partly a result of mechanical compaction of the
depend also on the environment. The fossils determine carbonate mud and grains, just as in terrigenous mud
the initial mineralogical composition of the reef and sand, but with an important difference. Unlike
structures and hence the burial diagenesis. silicate reactions, carbonate reactions are relatively
The coarse carbonate block deposits (talus) down fast even at low temperatures and chemical compac-
the slope in front of the reef (the fore-reef facies) and tion is therefore important both at shallow depth and
bioclastic sand behind the reef experience less water low temperature, as well as at deeper burial depth and
flow and tend to develop less marine cement. These higher temperature.
facies constitute better reservoir rocks in terms of Aragonite and high-Mg calcite are metastable
primary porosity. minerals in the marine environment. They precipitate
Reefs are often surrounded by organic-rich terrige- because low-Mg calcite, which is the thermodynami-
nous mud which may form a good source rock during cally stable calcium carbonate phase, fails to precipi-
burial. The reefs themselves may be good hydrocarbon tate. This is probably due to the poisoning (inhibiting)
traps, because they rise up from the seabed and thereby effect of sulphate on calcite precipitation (Kastner
constitute structures which may become sealed if the 1984). During burial both aragonite and high-Mg cal-
reef drowns and is covered by terrigenous mud cite dissolve and are replaced by low-Mg calcite, but
(Fig. 5.56). we do not know how fast this reaction is in marine
sediments. This thermodynamic drive for dissolution
and compaction is almost independent of stress. The
5.7.6 Burial Diagenesis dissolution of aragonite fossils or cements may cause
grain framework to collapse, whereas high-Mg calcite
Early cementation by calcite cements reduces the pri- is replaced by calcite that retains much of the original
mary porosity but the cement may strengthen the grain texture so that the grain framework is conserved.
framework and thus reduce compaction. Statistically, The solubility of carbonate grains is also a function
however, porosity clearly decreases as a function of of effective stress and thus burial depth. Pressure
a Island
Reef
Mean sea level Reef Lagoon proper
Water Reef
frame Water
Reef
talus or
Lagoonal deposits reef
Circum reefal rock
calcareous oozes
Reef rock
Lagoonal patch reefs
Platform from which reef-complex grew up
b
Shale
Sandstone Traces of oil from shale below
Impervious shale prevents
loss of oil upward
Ancient platform rock – impervious or water soaked
Much oil Less oil Little or no oil
Fig. 5.56 (a) Vertical cross-section through a carbonate plat- a steep submarine slope. The red algae is a strong reefbuilding
form from which a reef complex grew up. (b) Micritic organism while the more delicate green algae live in the
limestones in the lagoon have low porosity and permeability protected lagoonal environments. In the supratidal environment
and oil and/or gas will mainly be present in reef and reef talus we find blue-green algae and offshore carbonate is precipitated
facies. The reef has been building into the deep ocean producing by planktonic algae (coccolithophores)
solution will often be concentrated at particular

horizons where there has been some enrichment of
clay particles which increases the rate of solution.
Stylolites, which are surfaces where a considerable
amount of solution has taken place, will then form.
The horizon will be enriched with finely divided sili-
cate minerals and other insoluble material in the
limestones (Fig. 5.57). Where carbonate, usually cal-
cite, is dissolved due to pressure solution, the
dissolved material tends to be precipitated in the
Fig. 5.57 Cross-section of a stylolite. The dark colour is due to
neighbouring pore spaces. When the pores between
insoluble material within the limestone (such as clay minerals,
stylolites are filled with cement, dissolution along the pyrite and organic matter) which is concentrated in the seam.
stylolite ceases. Stylolites may have a relief from The amplitude (the distance between top and bottom) may be
1–2 mm up to several tens of centimetres and a taken as a minimum measure of the thickness of the dissolved
carbonate layer
spacing of typically 1–5 cm. As the rate of solution
along the surface varies, a very irregular relief sedimentation of carbonate sediments is comparable to
develops which may be taken as a minimum measure that of terrigenous material. However, today most
of the thickness of the carbonate layer dissolved. workers prefer to use the classification by Dunham
The carbonate dissolved at grain contacts and along (1962) because it is not based on the composition of
stylolites is transported to the adjacent sediments by the matrix but on the nature of the framework, which is
diffusion. Here the carbonate will usually re- more applicable in revealing the depositional processes.
precipitate in the pores, with a possible reduction in
porosity of up to about 20%. Investigation of
interbedded layers of limestone and dolomite (the 5.7.8 Folk (1959, 1962) Classification
rock is also called dolostone) show that the former is
more susceptible to pressure solution than the latter. Disregarding admixture of terrigenous material, Folk
Dolostone usually has a strong framework of (1959, 1962) distinguished between three basic
interconnected dolomite rhombs. Stylolites are there- components of limestones:
fore especially well developed in carbonates that are 1. Sediment grains (allochems). The principal
almost exclusively calcitic. Some late diagenetic dolo- allochems are: skeletal grains, ooids, peloids and
mite rhombs may precipitate along developing fragments of carbonate rocks (intraclasts or
stylolite surfaces as a result of a dynamic system extraclasts).
with pressure gradients on a microscopic scale. 2. Microcrystalline lime mud (micrite) comprising
CO2 and organic acids generated from clay-size particles (grain size <4 μm). In modern
decomposing organic matter are rapidly neutralised. carbonate environments such as the Bahamas most
The pH decreases with depth because of the increasing of the mud consists of micron-sized needles of
amounts of CO2 that can be dissolved in the porewater aragonite produced by green algae like Halimeda
as the pressure increases. The solubility of calcite and Penicillus. These are transported as clay frac-
increases with increasing pressure but at normal tion material.
hydrostatic pressure gradients temperature is the 3. Sparry calcite cement (sparite) which is carbonate
overriding factor determining the solubility gradients. crystals that have been diagenetically precipitated
Calcite precipitates only very locally during upwards in the pore space after deposition. While the car-
flow, where there are abrupt pressure drops. bonate mud and some of the grains have a brown
Compaction-driven flow is normally directed stain due to organic material, the cement stands out
upwards and since the solubility then is reduced, dis- as clear and transparent in thin section.
solution rather than precipitation will occur. The Almost all carbonate deposits contain more than
capacity of compaction-driven porewater to transport one type of material, and Folk’s classification is
carbonate is in any case rather limited because of the based on the relative proportions of the three
low solubility gradients and moderate fluid fluxes at endmembers: allochems, microcrystalline lime mud
greater depth. and sparry calcite cement (Fig. 5.58). Allochems rep-
resent the framework of the rock making up the bulk of
most limestones. The matrix between the allochems
5.7.7 Classification of Carbonate Rocks may consist of lime mud if there is little bottom cur-
rent. In limestones this mud may have recrystallised
Early petrologists often subdivided limestones according into small calcite crystals (microcrystalline mud),
to the size of the dominant mechanically deposited which is called micrite and is thus an indicator of a
grains: Calcilutite (grains <63 μm), calcarenite (grains low-energy environment.
between 63 μm and 2 mm) and calcirudite (grains Deposits consisting of well-sorted allochems have
>2 mm). Later Folk’s (1959, 1962) classification was primary porosity which may be filled with cement during
widely accepted because of its applicability to a wide diagenesis. The cement is precipitated from aqueous
range of carbonate rock types and the ease with which its solutions and consists of clear, transparent crystals
terms could be utilised and understood. Folk’s classifica- (spar) which are easy to distinguish from micrite,
tion is based on the idea that, in principle, the which tends to be brownish because of the organic
Intraclasts
Sparry allochemical Micritic allochemical
limestones limestones
Fossils
Pellets
Ooids
Sparry
calcite
Intraclasts
Allochem grains
Intrasparite Intramicrite
Microcrystalline Sparry allochemical
allochemical limestones
limestones Ooids
(Cleanly washed)
II I Oosparite Oomicrite
Bioclasts
Microcrystalline
limestones
(Micrites) Biosparite Biomicrite
III
Carbonate Sparry calcite
mud cement
Peloids
Fig. 5.58 Triangular diagram showing the three major textural

types of limestones. Shaded areas depict the most commonly Pelsparite Pelmicrite
encountered limestones. I. The limestones consist chiefly of
allochemical constituents cemented by sparry calcite. The rela- Fig. 5.59 Somewhat simplified schematic representation of the
tive proportions of sparry calcite cement and allochems varies constituents that form the basis for Folk’s classification of car-
within rather restricted limits depending on the packing and bonate rocks (modified from Tucker and Wright 1990)
shape of the allochems, which are important factors for the
pore volume. II. These limestones also consist of a considerable
proportion of allochems, but in these cases the currents were not of oolites or pellets with a carbonate mud matrix.
strong enough or persistent enough to winnow away the micrite.
In these deposits the restriction of packing imposes a certain
Intrasparite consists of carbonate fragments eroded
maximum on the amount of allochems, but there is no mini- inside a basin to form well-sorted carbonate deposits.
mum. Therefore the content of allochems may vary from about Sediments with a considerable percentage of grains
80% down to almost nothing in clean carbonate mud. III. larger than 1.0 mm are called rudites. A coarse-grained
Limestones consisting almost entirely of micrite with little or
no allochems or sparry calcite (modified from Folk 1962)
bioclastic sparite would thus be classified as
biosparrudite.
If the allochems are found in a matrix of carbonate
content. Sparite is thus a term for well-sorted carbonate mud (micrite) the rock is termed biomicrite, oomicrite,
sand, originally with high primary porosity, which has intramicrite or pelmicrite. These sediments represent
later been filled with calcite cement (spar). As such it low-energy environments where neither currents nor
normally represents a high-energy environment. Thus wave action separate the mud from the grains. Micrites
the relative proportions of micrite and sparry calcite normally have too low porosity and permeability to be
cement indicate the degree of “sorting” or current reservoir rocks. While the carbonate mud may initially
strength of the environment during deposition. have high porosity this is strongly reduced by mechan-
As illustrated in Fig. 5.59 there are two parts to the ical and chemical compaction. Micrites may, however,
rock name classification. The first part is contributed become tectonically fractured, thus increasing their
by the abbreviated allochems’ name based on the porosity and particularly their permeability. Second-
dominating type of allochem. The second part reflects ary porosity due to dissolution of aragonite and high-
the void-filling material (micrite or sparry calcite). Mg calcite fossils may also improve the reservoir
Biosparite, for example, means that the carbonate quality of micrites.
rock is dominated by fossil fragments without any Carbonate sediments deposited in environments
matrix, but cemented with sparry calcite cement. with sufficient wave or current energy to separate the
Oomicrites and pelmicrites are sediments consisting grains from the muddy matrix can be well-sorted with
Fig. 5.60 Dunham (1962) classified carbonate rocks according autochthonous and allochthonous types of reef limestones by
to depositional texture (modified from Tucker and Wright Embry and Klovan (1971, see also James 1984). See text for
1990). The classification was later expanded with different explanation
an initial pore space of about 40% between the grains matrix, the latter a greater or lesser amount of calcite
(or even more if the skeletal material shows irregular cement in the intergranular pores. The term
growth forms). These sparry deposits are potential boundstone is used where the fabric indicates that the
reservoir rocks if they only become partly cemented original components are bound together during depo-
with calcite spar. The presence of petroleum may sition (e.g. as in reefs). It is equivalent to the term
retard the precipitation of carbonate cement. biolithite in the classification of Folk (1959).
The textures of carbonate deposits correspond in
many ways to those we observe in clastic terrigenous
5.7.9 Dunham (1962) Classification sediments. Mudstone corresponds to clay, and
wackestone and packstone correspond to greywackes.
Working for Shell Oil, Dunham (1962) developed a Grainstones are well sorted with little matrix. We
classification which was more oriented to the needs of distinguish between grain-supported and matrix- or
the oil industry, and this simple classification has mud-supported sandstones. Mud-supported
become the most widely used today. His classification sandstones will be subject to compaction of the matrix
is also based on the depositional texture of the lime- between the grains, and have many of the plastic
stone, but the fundamental criterion is not the compo- properties of fine-grained sediments (muds). The
sition on the matrix as in the Folk (1959, 1962) degree of compaction will depend on the proportion
classification, but the nature of the framework of grains.
(whether it is mud-supported or grain-supported), see Grain-supported carbonates have a framework of
Fig. 5.60. Mud is defined as carbonate particles with a grains which rest upon one another. Compaction can-
grain size <20 μm. not take place without the grain framework being
Mud-supported deposits imply deposition in a low deformed. This may take place by the grains being
energy environment. Limestones with very few grains packed more tightly through mechanically crushing
(<10%) floating in a mud matrix are classified as due to the force exerted by the overburden, or being
mudstones. Mudstones with more than 10% grains, chemically dissolved, particularly at the contacts
but still mud-supported, are classified as wackestones. between grains (pressure solution). The mechanical
Grain-supported limestones indicate deposition in a strength of carbonate fragments from fossils may be
high energy environment. They are classified as relatively low, but that of thick-shelled fossils and
packstones or grainstones, the former having a mud ooids is high. The contacts between grains will
initially be very small, for example round ooids will

only have very limited areas of contact. As the over-
burden increases, the pressure per unit area at the
contact points will be very great. Pressure solution
will then occur at the contact points, so that the contact
area expands and the pressure per unit area decreases.
The distinction between grain-supported and mud-
supported rocks is not simply a function of the ratio of
grains to mud, because carbonate grains have widely
different and often highly irregular shapes. Rocks
composed of spherical grains (e.g. ooids) may need a
grain content of about 60% to achieve grain support,
whereas rocks with highly irregular grain shapes may
form a self-supporting framework with a grain content
of only 20–25%.
Mudstones, wackestones and packstones would all
Fig. 5.61 Partly dolomitised oolite. The dolomite rhombs (dark
be poor reservoir rocks because of the carbonate brown) transect the oolite grains (light brown) and the sur-
mud in the matrix. In the case of packstones the grains rounding spar crystals (blue), indicating a late diagenetic
form a grain-supported fabric but the matrix is still dolomitisation (modified from Greensmith 1978)
carbonate mud. Grainstones represents well-sorted
carbonate deposits with good porosity and permeability, The mineral dolomite ðCaMgðCO3 Þ2 Þ, consists of
which may serve as a good reservoir if cement
layers of CO23 groups separated by alternating layers
precipitation is not too advanced. This is the carbonate
of Mg2+ and Ca2+. This is a highly organised structure
equivalent of well-sorted siliceous sand. Boundstones
(trigonal rhombohedral) and the organisation of more
are rocks bound together by organisms (fossils) as in
or less pure layers of Mg2+ and Ca2+ leads to high
reefs, and they may also have high porosity and
kinetic energy being required for the crystallisation of
permeability.
dolomite. This is particularly true at low temperatures,
Embry and Klovan (1971) extended the classification
and so far it has not been possible to synthesise low
of Dunham (1962) with different types of autochthonous
temperature dolomite (<100 C) in the laboratory.
and allochthonous reef limestones (Fig. 5.60). The
Dolomite is in most cases not formed directly, but
autochthonous types are bafflestone that contains in-
as a secondary mineral as a result of reactions between
place stalked fossils that trapped sediment by baffling
different forms of CaCO3 and Mg2+. The reaction
the currents, bindstone that contains in-place large
lamellar or tabular fossils that encrust or bind loose
2CaCO3 þ Mg2þ ¼ CaMgðCO3 Þ2 þ Ca2þ
sediment together, and framestone that contains in-
place large frame-builders forming the rigid framework
of the reef. The autochthonous types contain more than is dependent on the Mg2þ =Ca2þ ratio. The
10% particles larger than 2 mm; the matrix-supported dolomitisation process will only proceed with a supply
type is floatstone while the clast-supported is termed of magnesium maintaining a high Mg2þ =Ca2þ ratio.
rudstone. Mg2+ is more strongly hydrated than Ca2+ in sea-
water. The Mg2+ ion together with its surrounding
water molecules, MgðH2 OÞ2þ 6 , cannot easily enter a
5.7.10 Dolomitisation crystal position at surface temperatures. However, the
hydration decreases with increasing temperature and
The term “dolomite” is used to designate both a min- this is a factor often cited in favour of explaining late
eral and rocks in which this mineral is the main con- burial dolomitisation.
stituent (Figs. 5.61 and 5.62a,b). To avoid confusion, Experiments show that in the absence of sulphate,
the term “dolostone” has been introduced for the rock, dolomite forms rapidly in solutions with
but has not been widely adopted. MgCl2 þ NaCl þ CaCl2 . The main reason that
bacteria. The sulphate is reduced to hydrogen sulphide

(H2S), which may react with dissolved reduced iron
(Fe2+) to form metal sulphides such as pyrite (FeS2).
NH4þ can replace Mg2+ adsorbed on clay minerals. In
this way magnesium is liberated for dolomitisation.
5.7.10.1 Models
Dolomitisation means that CaCO3 is dissolved and
dolomite precipitated. The conditions for this are:
1. That calcium carbonate becomes unstable, and that
the solution is supersaturated with respect to
dolomite.
2. That Mg2+ is added to the solution so that
dolomitisation can continue.
3. That an inhibitor such as sulphate is absent or at
least in low concentration.
Solution of calcium carbonate takes place most
easily if we have aragonite at the outset. Fine-grained
carbonate mud has a large specific surface which
enables it to react more rapidly than massive
carbonate.
The dolomitisation process is accelerated if the
sediment concerned has a high permeability and a
high rate of percolating porewater containing magne-
sium, though if carbonate sediments already contain a
good deal of magnesium, dolomitisation will be able
to proceed without any addition of Mg2+. This
applies, for example, to carbonate sediments rich
Fig. 5.62 (a) Partly cemented secondary cavity in a limestone in high-Mg calcite as in reefs. We often find thin
as seen in plane polarised light. The remaining porosity has been dolomite beds or finely divided dolomite in shales,
impregnated with blue coloured epoxy. (b) The same area and this may be due to a supply of Mg2+ from clay
viewed with cathodoluminescence microscopy. The cement minerals.
consists of brown luminescent calcite (1) and zoned dolomite
crystals (2). The zoned dolomite crystals have both euhedral and Seawater is a highly complex solution. We cannot
irregular cores (arrow) with a brown luminescence, while the simply predict the way in which it will react from the
periphery is characterised by thin zones of reddish lumines- concentrations found through chemical analyses.
cence. Upper Tertiary, Iran. (Photos courtesy of Torleiv Some of the Mg2+ and Ca2+ is associated with Cl
Torgersen)
through ion pairing, and must be excluded from
calculations regarding activities involving carbonates.
dolomite is not common in modern marine Theoretical calculations indicate that
environments is that the sulphate ion ðSO24 Þ is very dolomitisation should occur when the Mg2þ =Ca2þ
efficient at preventing dolomitisation. Consequently activity ratio is about 0.6. Although these figures are
dolomitisation takes place more easily when there somewhat uncertain, it is clear that seawater, in which
are few sulphate ions (Baker and Kastner 1981). For the Mg2þ =Ca2þ ratio is 5.6, is oversaturated with
this reason dolomite is formed in a number of lakes respect to dolomite. The fact that dolomite does not
and in brackish zones with mixing of salt- and fresh- form is ascribable to kinetic reasons, one probably
water (Dorag model). In a microbial sulphate reduc- being the high SO24 concentration, and only when
tion zone, the SO24 concentration will be lower and the ratio is over about 7 will dolomitisation take
ammonium ðNH4þ Þ can be formed by nitrate-reducing place in seawater. In freshwater the ion strength is
lower, and here dolomite can be formed at lower enriched in magnesium, and dolomite will only be
Mg2þ =Ca2þ ratios because there is very little sulphate. precipitated at high Mg2þ =Ca2þ ratios. In addition
However, freshwater contains little magnesium, so magnesite (MgCO3) may form. This explains why
Mg2+ must be supplied by mixing with seawater and dolomite is important in most evaporite sequences.
large amounts of seawater must circulate through the However, evaporite minerals are very soluble, parti-
limestone. A steady percolation of freshwater alone cularly chlorides, and often are not preserved; gypsum
will not lead to any great degree of dolomitisation, may also dissolve and be replaced by carbonate. Dolo-
because freshwater will displace the magnesium-rich mite might therefore have been deposited in an evapo-
salt water, and the mixing zone will be too small. rite environment despite the fact that we do not find
Seawater is the only source of magnesium for large any of the original highly soluble salts preserved. For
scale dolomitisation, except in evaporite sequences this reason it is important to look for indirect evidence
where Mg-rich evaporite minerals can dissolve and of evaporite conditions and solution, including
add Mg2+ to the porewater. The seawater which replacement of evaporite minerals.
circulates inside atolls and reefs is reducing and there- The most important indicators of evaporite
fore low in sulphate, at the same time having abundant conditions are:
magnesium. Reef and atoll facies therefore offer 1. Absence of ordinary marine fossils, apart from
favourable conditions for dolomitisation. stromatolites which can tolerate high salinity. In
Dolomitisation of large volumes of limestones would Palaeozoic and younger deposits stromatolites are
require vary large sources of magnesium and it is only typical of evaporites, because under normal marine
very near to the surface that seawater can flow with conditions cyanobacteria have too much competi-
sufficient fluid fluxes to supply enough Mg2+ to cause tion from other organisms.
pervasive dolomitisation. Freshwater has a low 2. Breccias which may have been formed through
sulphate content which makes it easier to precipitate solution of underlying salt deposits so that beds
dolomite but it does not explain the supply of such collapse and form a collapse breccia. Such breccias
large quantities of Mg2+. are characterised by angular fragments from an
Hydrothermal circulation of seawater will cause overlying bed, for example of carbonate, which
dolomitisation but that will be limited to the flow have fallen down into a solution cavity.
path controlled by fractures and permeable layers. 3. Pseudomorphosis (replacement) of evaporite
Very little of the flow will penetrate tight rocks, i.e. minerals, e.g. halite (NaCl) and gypsum
micritic limestones. ðCaSO4 2H2 OÞ. Evaporite minerals, which are
Massive and extensive dolomites are likely to be disseminated through a matrix of less soluble
formed at shallow depth near the surface or seafloor in minerals such as carbonates, are often replaced
marine basins. This is favoured by high salinity and through pseudomorphosis so that the crystal form
evaporitic conditions. may reveal the original mineral. The cubic halite
Dolomite showing fluid inclusion and textural evi- crystals are typical, and the characteristic swallow-
dence of having formed at higher temperatures may tail twins of gypsum crystals are easily recognised
have been recrystallised at greater depth. even if they have been replaced by other minerals.
4. Chickenwire structure. Anhydrite often forms very
characteristic nodules or continuous layers which
5.7.10.2 Evaporite Model for Dolomitisation look like chicken wire. Even if the anhydrite layers
Dolomite is often associated with evaporite are converted to calcite, these structures may be
environments. The first definite example of dolomite preserved.
being formed today was found in evaporating 5. Authigenic quartz and feldspar. Evaporites are often
sediments in a supratidal environment in the Bahamas associated with microcrystalline quartz (chert) and
in about 1960. It is clear that when seawater chalcedony. A high content of authigenic feldspar
evaporates, and aragonite and also gypsum and zeolites is also typical of many evaporites. With
ðCaSO4 2H2 OÞ are precipitated, the composition of low-grade metamorphism (200–300 C) zeolites will
the fraction still in solution will become increasingly dissolve and be replaced by feldspar.
5.7.10.3 Constraints on Diagenetic Dolomite term “dedolomitisation” should be dropped, and the
Formation positive term “calcitisation” be used instead.
While early diagenetic dolomite is normally relatively
fine-grained, late-diagenetic dolomite usually forms 5.7.10.4 The Significance of Dolomitisation
larger crystals, often well-defined dolomite rhombs. For many years there has been intensive research into
The smaller crystals are formed by rapid crystal the processes which lead to dolomitisation. A great
growth under conditions of high supersaturation. The deal remains to be learnt, however, before we really
larger crystals are formed by slow crystal growth from understand the precise conditions for dolomitisation so
a few nucleation centres at a very low degree of that we can predict the extent of dolomite in sedimen-
supersaturation combined with deep burial. tary basins. The reason for this great interest is that
As mentioned previously, hydration of Mg2+ dolomitic carbonate rocks are very important
decreases with increasing temperature, making Mg2+ reservoirs for oil and gas. The dolomitisation process
more readily available for the dolomite structure. may create secondary porosity because dolomite has a
Dolomite might then be formed at very low greater density than calcite so if an identical number of
Mg2þ =Ca2þ ratios; at about 80 C the ratio can be as mol dolomite is precipitated as in the original calcite,
low as 0.1. Although dolomite can precipitate in we would get approximately 12% smaller volume and
solutions with low Mg2þ =Ca2þ ratios, the question an equivalent increase in porosity. The dolomitisation
remains of how sufficient magnesium is added for process assumes, however, that calcium is removed
the dolomitisation to take place. The amount of mag- and magnesium introduced, and there is then no reason
nesium in solution in the deeper part (>2–3 km burial why there should be an increase in porosity propor-
depth) of sedimentary basins is in most cases very low tional to the difference in density, since there is no
and is far from adequate as a source for large-scale reason why the same number of mol dolomite should
dolomitisation. Water from compaction of mudstones be precipitated as were removed by the calcite
and shales is probably also insufficient to supply much dissolving. Since dolomitisation involves large-scale
magnesium for dolomitisation. Primary high-Mg cal- percolation of porewater, we may also have net
cite will be a source of Mg to form dolomite at depth. increase in the porosity associated with this process.
There is only a little magnesium in the porewater of Micritic limestones have in most cases too low
sedimentary basins apart from in the vicinity of porosity and permeability to be regarded as reservoir
evaporites. Probably most of the dolomitisation occurs rocks except when fractured. Tectonically fractured
near the surface where the magnesium comes from limestones may become cemented in relatively short
seawater. Fine-grained dolomite, formed at an early geologic time before the migration of petroleum. Frac-
stage, may be dissolved at depth and recrystallise as tured dolomites are, however, more stable and likely to
coarser-grained dolomite, in which case there is no need remain open longer because of the lower solubility of
to postulate a magnesium supply deep in the basin. dolomite.
We often find dolomite enriched along stylolites,
probably because dolomite is less soluble than calcite 5.7.11 Formation of Carbonate Sediments
and the solution and precipitation round a stylolite will Rich in Siderite and Chamosite
concentrate clay minerals which, in turn, may release (Ironstones)
some magnesium.
If the composition of the porewater later shifts Iron carbonates like siderite (FeCO3) are stable car-
towards a low Mg2þ =Ca2þ ratio, dolomite may dis- bonate phases with rather low solubility (Berner
solve and calcite reprecipitate. There are a number of 1981). In the presence of sulphides, however, most
cases where distinctly dolomite-type rhombs are found of the iron will be precipitated as iron sulphides (i.e.
to consist of calcite. One common cause of reversed pyrite or marcasite). Precipitation of siderite and other
dolomitisation – often called dedolomitisation – is iron carbonates is therefore restricted to settings where
porewater coming from gypsum which is dissolving. the porewater has a low content of reduced sulphur
This gives the porewater a high Ca2+ concentration. (sulphides), as in freshwater and below the sulphate-
However, many people have recommended that the reducing zone.
Glauconite and Weathering

nd,
chamosite facies Quartz sa Fe+ org.
,
kaolinite
Well-sorted sand (high energy)

partly siderite cemented
Calcite mud Kaolinite and
and clay siderite mud
SO 42– SO 2–
4
Glauconite/chamosite
H 2S H 2S
Sulphate reduction
Aragonite mud Aragonite mud

and clay and clay
Mud with aragonite and
high-Mg calcite + iron
Dissolution of aragonite and high-Mg calcite,
precipitation of siderite
Calcite mud Siderite mud
and clay and clay
Sideritic mudstones and
siderite cemented sandstones
Fig. 5.63 Formation of sideritic limestones. During weathering reduced to Fe2þ and when aragonite becomes unstable, siderite
large amounts of iron are released and transported as red fine (FeCO3) is formed instead of calcite because siderite is most
grained iron oxides. Reacting with organic matter iron oxide is stable (lower solubility)
There is almost no iron in solution in oxidised water clastic iron-rich sediments are trapped in the deep
and iron can therefore not be taken from seawater. water around the carbonate platform and there is
During weathering on land, however, large amounts practically no iron in the seawater covering the plat-
of iron may be released and transported as small form. Only small amounts of aeolian dust with some
particles of iron oxides, hydroxides or adsorbed on iron are transported to the Bahamas from Africa
clay and organic matter. These are often concentrated (Sahara).
in the distal parts of deltas where clastic sedimentation
is low (Fig. 5.63).
Most of the iron supplied by rivers will be reduced 5.8 Carbonate Reservoir Rocks
just below the seafloor by small amounts of organic
matter. It may precipitate at the redox boundary as iron 5.8.1 Introduction
oxides or as glauconite or chamosite. In carbonate
sediments, aragonite may be partly replaced by sider- Carbonate reservoir rocks are mostly limestones and
ite instead of calcite if iron is present. Because of the dolomites and these rock types contain about 50% of
lower solubility of siderite all the available iron will be the oil reserves in the world. They are particularly
exhausted and precipitated as siderite before calcite common in the Middle East accounting for many of
can begin to precipitate. the giant reservoirs, the majority of which have been in
Iron-rich sediments are typical of mixed carbonate production for several decades.
and clastic sedimentary sequences. On carbonate Carbonate reservoir rocks differ in several impor-
platforms like the Bahamas there is little supply of tant aspects from sandstone reservoirs:
iron because it is not connected to a source on land. All
1. The sediment particles are in most cases produced It is commonly formed by the decay of organic
locally within the basin by biological production material within carbonate skeletons (e.g. fora-
(fossils) or by chemical precipitation. minifera, corals and bryozoans).
2. The mineralogy and the textures depend very much 3. Growth-framework porosity (porosity inside a
on the organisms precipitating the carbonates. rigid framework produced by fossils)
3. Carbonate minerals have higher solubility than sili- Growth-framework porosity is developed
cate minerals in porewater, and also higher reaction between and within organisms during the in-
rates at low temperature. Dissolution of grains and place growth of a carbonate framework. Typical
precipitation of cements may therefore be impor- examples are reef corals with uneven and patch-
tant also at shallow depth (<2 km). ily developed pores between the branches or
4. The reservoir quality is highly dependent on the between different colonial organisms.
porosity, which may be of different types. The 4. Shelter porosity
porosity in carbonate rocks may be pore space Shelter porosity is often found below large
between grains (intergranular porosity) or porosity plate-like grains which have acted as umbrellas
within grains (intragranular porosity), commonly protecting the pore space beneath from being
fossils. These primary porosity types are usually filled with finer material. Typically porosity
strongly modified even at shallow depth during may be preserved below the convex side of
burial. Well-cemented carbonate rocks may be brit- molluscs and brachiopod shells. Organic mate-
tle, particularly during uplift, and tectonic fractur- rial which is later decomposed may have pro-
ing may produce important fracture porosity and duced shelter porosity which can be preserved if
permeability. there is early cementation.
5. Fractured carbonate reservoirs are very important B. Secondary porosity is pore space produced by dis-
both in limestones and dolostones (a rock of solution of grains or cement after deposition. Sec-
dolomite). ondary porosity may therefore be regarded as an
The most important porosity types in carbonate addition to the primary porosity. This requires a net
rocks are the following: transport of carbonate in solution out of the rock. If
A. Primary porosity is the pore space that existed in aragonite dissolves and the same amount of calcite
the sediment just after deposition prior to diage- precipitates inside the rock, no porosity is gained,
netic alteration. During burial the primary porosity though the distribution of pore space and the per-
is reduced by compaction and cementation. meability may change drastically. Meteoric water
1. Primary intergranular porosity (pore space is capable of dissolving both aragonite and calcite
between carbonate grains) and transporting the dissolved material out from a
Intergranular porosity (inter-particle porosity) is limestone so that the net porosity is increased.
the preserved primary pore space between At greater burial depth however the porewater
grains. Pore shapes may vary considerably and is nearly always in equilibrium with calcite which
are homogeneous only when the enclosing is present in most marine sediments. The potential
particles are of uniform size and shape (e.g. for significant net dissolution and formation of
well-sorted ooids). The porosity increases with secondary porosity is therefore very limited.
better sorting and more irregular grain shapes, 1. Mouldic porosity
attaining values up to approximately 60%. Also Mouldic porosity is formed by the selective
carbonate mud, just after deposition, has high dissolution of grains, particularly skeletal mate-
intergranular porosity (70–80%) between the rial with a primary aragonitic composition.
small mud particles, but the permeability is low Whole fossils or grains like ooids may dissolve
and in most cases this porosity is rapidly reduced after the precipitation of cement in the primary
by mechanical compaction during early burial. pores. These secondary pores thus become
2. Intragranular porosity (porosity inside grains) moulds of the dissolved structures. Mouldic
Intragranular porosity (intra-particle porosity) porosity may also form by dissolution of fossils
consists of pore space that occurs within grains. from a matrix of carbonate mud which has been
lithified. In the Jurassic Smackover Formation 5.8.2 Some Examples of Carbonate

of Arkansas, USA, oomouldic porosity was cre- Reservoirs
ated by the dissolution of ooids during fresh-
water flushing (More 1989). 5.8.2.1 Grainstones
2. Fenestral porosity Well-sorted carbonate sands (grainstones) are good
Fenestral porosity consists generally of small reservoirs if the porosity is not reduced too much by
elongate to equant pores which are typically cementation. The sand grains are usually ooids or
1–10 mm in diameter. These characteristic pores fossil fragments. If the grains are mainly of low-Mg
are often arranged in layers within the sediment calcite, most of the porosity tends to be primary
and are most commonly encountered in algal mats porosity which gradually fills with cement during
facies where they have been produced by burial. The source of most of the cement is then
decaying organic matter or by desiccation. normally pressure solution along stylolites or grain-
3. Breccia porosity grain contacts.
Breccia porosity may form by the collapse of a Precipitation of early carbonate cement reduces the
rock due to dissolution (e.g. dissolution of stresses at grain contacts and strengthens the grain
evaporites or limestone dissolution during framework. The mechanical compaction may therefore
karst weathering) or tectonic deformation. be reduced. When many of the grains consist of arago-
4. Fracture porosity nite they may dissolve, producing secondary porosity
Fracture porosity commonly forms due to fold- while the dissolved carbonate fills the primary porosity.
ing, faulting, salt doming, differential compac- Dissolution of aragonitic ooids and fossils produces
tion, salt dissolution or fluid overpressure. moulds which have very little communication between
Carbonate sediments often become cemented them and such mouldic reservoirs are characterized by
at shallow depth and may fracture and dilate rather high porosity but low permeability.
so that fracture porosity is produced. Open
fractures will gradually be filled with cement 5.8.2.2 Fractured Reservoirs
but fractures in dolomite may remain open lon- The term fracture is used for any break in a rock and
ger than those in limestones. includes cracks, joints and faults. In many cases the
Shrinkage porosity is a special type of fracture fracture is less permeable than the matrix, but some
porosity formed during early diagenesis when may also be partly or totally open. Open fractures
sediments may shrink i.e. due to dewatering. formed by extension are common during uplift and
Septarian fractures in limestone concretions folding. Well-cemented limestones and dolomites
are also a kind of shrinkage fracture. tend to have brittle properties and faulting may pro-
5. Vuggy, channel and cavern porosity duce a breccia with good permeability.
Vuggy, channel and cavern (karstic) porosity The rock fragments produced by the brecciation
are not fabric selective, i.e. they cut across take the stress and prevent the fault plane from closing.
grains and/or cement boundaries. The pores Movements (off-set) along the fault plane have a sim-
are of irregular size and shape and may or may ilar effect. The porosity and permeability produced in
not be interconnected. Many vugs are solution- carbonates is, however, temporary because it may
enlarged moulds where evidence of the precur- relatively rapidly be filled with carbonate cement.
sor grain has been destroyed by dissolution of Filling of open fractures by oil or gas will retard or
the neighbouring matrix. Vuggy pores are com- inhibit calcite cementation and thus contribute to the
monly 1 mm–1 m in diameter. Cavern porosity conservation of the porosity. This is particularly true if
commonly relates to meteoric (karstic) leaching the rocks are relatively oil-wet.
and is differentiated from vuggy porosity by the Fractures in limestones are commonly cemented by
larger pore size (man-sized or larger). Karst is diffusion of carbonate from the adjacent matrix. Flow
formed by dissolution related to underground of water upward along the fault plane will cause dis-
river systems and can therefore be very exten- solution of calcite rather than precipitation because of
sive and form large oil reservoirs. the retrograde solubility with respect to temperature.
In carbonates where the fractures are mostly open and has produced fracturing in the chalk and these
forming a three-dimensional fracture network, the fractures have been essential in increasing the overall
effective permeability can be very good. However, reservoir permeability.
fractures are often formed during several phases of Early filling of oil and gas in carbonate rocks like
tectonic deformation and then only some of the the chalk will retard calcite cementation and tend to
fractures may remain open. The overall drainage dur- preserved porosity compared to water-saturated chalk.
ing production is then more difficult to predict.
Fractured carbonate reservoirs often consist of
5.8.2.4 Carbonate diagenesis and reservoir
dolomite. Well-cemented massive dolomite is rather
properties
brittle forming large fracture networks when subjected
The initial mineralogical and textural composition
to tectonic forces.
determines the loss of porosity by mechanical and
Fractures in dolomite may be preserved longer than
chemical compaction during burial. Since most car-
in limestones. This is because dolomite is both
bonate sediments have been precipitated biologically
mechanically stronger and less soluble than calcite so
both the mineralogy and the textural properties are
that it will take longer for the fractures to be cemented
primarily controlled by the organisms and later
up. For this reason we may find dolomite reservoirs at
reworking. The content of unstable minerals, mainly
great depth (5–6 km) and also in uplifted basins. Frac-
aragonite and Mg calcite, and also the grain size rang-
tured reservoirs may have very high permeability so
ing from mud to very coarse material, reflects the
that the reservoirs can be produced even if the average
organisms that precipitated the carbonate.
porosity is low.
In carbonate mud, aragonite will in most cases be
replaced by calcite, producing a dense microcrystal-
5.8.2.3 Chalk
line limestone. High contents of Sr are evidence of
Chalk is a pelagic sediment where calcareous algae
primary aragonite content.
(coccolithophores) are dominant. Coccolithophores con-
Sediments containing almost only calcite, like the
sist of low-Mg calcite and are therefore mineralogically
coccolliths making up chalk, will be rather stable
stable during early diagenesis. The presence of
chemically and also mechanically because the over-
aragonitic skeletal material is fairly meagre, so these
burden stress in these fine-grained sediments is
mineralogically unstable fossils are not an important
distributed on so many grain contacts. This is why
calcite source for early cementation. During periods
the chalk in the Ekofisk Field may preserve 30%
with little pelagic sedimentation seafloor cementation
porosity at 2.5–3.0 km depth (Fig. 5.64).
may form hardgrounds. These are often encrusted by
Carbonate sand representing a coastal facies is likely
bivalves, boring organisms, bryozoa and sponges and
to be flushed by freshwater, dissolving aragonite and
may be mineralised by glauconite and phosphate.
precipitating calcite. This process may create karstic
For the most part the chalk is rather pure CaCO3 but
porosity and also enhanced porosity in sandstones.
more clay-rich intervals like the Plenus Marl occur in
At greater depth however the bulk chemical com-
the North Sea reservoirs, forming a tight low-
position of carbonates does not change significantly
permeability zone. Thin clay laminae enhance pres-
except when influenced by hydrothermal activity
sure solution and often develop into stylolites. The
(Bjørlykke and Jahren 2012, Ehrenberg et al. 2012).
dissolution of calcite by pressure solution is the most
Below the reach of meteoric water flow the porewater
important source of carbonate cement.
flow is very low, limited by the rate of compaction
During diapirism of the underlying Zechstein salt
(See Chap. 4).
(Upper Permian) the chalk became unstable and was
In marine sedimentary basins there is nearly always
reworked by gravity flows (turbidites and debris
some calcite, also in sandstones and shales and the
flows). The eroded sediments were already somewhat
porewater is therefore in equilibrium with calcite. This
cemented and the debris flow units have proved to be
reduces the potential for dissolution, transport of
better reservoir rocks than the primary chalk deposits.
solids in solution and precipitation both by advective
The early cementation may have strengthened the
flow of porewater and by diffusion, because the con-
grains or aggregates of grains so that compaction
centration gradients are so small.
during burial was reduced. In addition, the salt doming
Fig. 5.64 Compaction curves for carbonates in sedimentary basins .The rate of porosity loss as a function of depth varies greatly
with the initial textural and mineralogical composition. Chalk with little aragonite and high-Mg calcite preserve much porosity due
its stable mineralogy (calcite) and fine grained composition.
Most black shales (source rocks) contain some car- Nearly all the dissolution will occur in the proximal
bonate layers or concretions which would neutralise parts where the meteoric water is recharged.
organic acids and CO2 generated during maturation of Early calcite cementation due to dissolution of
kerogen. In the absence of carbonate, black shales aragonite may produce hard and mechanically stable
would generate acid porewater but this would be limestones at shallow depth. Mechanical tests of
neutralised rapidly in contact with sediments shallow carbonates offshore Australia show that they
containing calcite. would undergo very little mechanical compaction
To increase the porosity of a 100 m thick limestone (strain) if subjected to stress of 40–50 MPa,
by 1%, 1 m3 of calcite would have to be removed in corresponding to burial depth of 4–5 km (Croizé
solution (Fig. 5.65). This would require 27 000 et al. 2010). This means that most of the porosity
volumes of porewater assuming 100 ppm) reduction would be chemical. In the case of
undersaturation with respect to calcite. Even if the sandstones, chemical compaction is mostly controlled
porewater contained relatively high concentrations of by temperature due to the kinetics of quartz precipita-
organic acids, very large fluxes would be required to tion. In carbonates temperature is less important and
produce a significant net increase in (secondary) the compaction rate is a very complex function of the
porosity (Barth and Bjørlykke 1993). distribution of stress on grain contacts and along
The potential for increasing the porosity of carbon- stylolites.
ate sediments is therefore very low except due to In conclusion prediction of porosity in carbonate
meteoric water flushing at rather shallow depth. Even rocks must be based on the primary sediment compo-
if meteoric water can flow in porous carbonate sition and facies relationships, controlling chemical
sediments far offshore the porewater will rather and mechanical compaction.
quickly approach saturation with respect to calcite.
Dissolved
calcite
Mudstone with small amounts of calcite
Porewater in
Flow equilibrium with
Limestone Flow parallel with calcite
perpen-
bedding
dicular to
bedding
Concentrated
dissoluƟon
Mudstone with high organic Porewater
content and no carbonate Under saturated
with respect to
calcite
Fig. 5.65 Most marine sediments including sandstones and some organic acids but these will be neutralized at the base of
shales contain some carbonate and the porewater is nearly a limestone because of the high reaction rates of carbonates.
always in equilibrium with calcite. Most source rocks also Porewater undersaturated with respect to calcite can not create
contain some calcite. Black organic rich shales will produce significant secondary porosity inside a limestone.
5.9 Summary Dolomite is both mechanically and chemically

more stable (less soluble) than calcite and therefore
Limestones are mostly organically precipitated except preserves more of its porosity during burial. Exten-
in evaporitic environments. The evolution of calcare- sional fractures formed tectonically will tend to stay
ous organisms determines to a very large extent the open longer in dolomite than in limestones and may
initial grain size and mineralogy of the sediments, therefore form important reservoirs.
which in turn strongly influences the properties of Carbonate sediments may compact both mechani-
carbonate reservoirs. The depositional environment is cally and chemically also at shallow depth and low
very important in controlling primary sorting of car- temperatures. Early cementation due to dissolution of
bonate sand and the distribution of framework builders aragonite and precipitation of calcite cement will
like reefs. increase rock strength and prevent further mechanical
Limestones compact mechanically as a function of compaction. Carbonates are very complex with
effective stress, but mineral dissolution and precipita- respect to reservoir properties. This is because of the
tion (chemical compaction) may also be important at great variation in the primary composition of the
shallow depth (0–1 km). Thermodynamically unstable sediments, which may be precipitated chemically or
aragonite dissolves at rather shallow depth which may biologically by many different types of organisms.
cause the formation of secondary mouldic porosity Expertise is required in disciplines ranging from
and early cement. Sediments consisting of mostly cal- palaeontology to mineralogy and complex diagenetic
cite will compact mechanically until pressure solution reactions.
becomes an effective compaction process. Grain-to-
grain dissolution and stylolites then provide the
sources of cement reducing the primary porosity. Further Reading
Dolomite forms in most cases near the surface
under evaporitic conditions, and under reducing con- Audet, M.D. 1995. Modelling of porosity evolution and
dition in contact with seawater (as in reefs and atolls) mechanical compaction of calcareous sediments. Sedimen-
where the sulphate content is low. Dolomitisation of tology 42, 355–374.
Barth, T. and Bjørlykke, K. 1993. Organic acids from source
calcite requires dissolution and reprecipitation and a
rock maturation; Generation potentials, transport
large-scale supply of magnesium and there is no rea- mechanisms and relevance for mineral diagenesis. Applied
son that this in itself should cause increased porosity Geochemistry 8, 325–337.
even if dolomite is denser than calcite. Bathurst, R.G.C. 1975. Carbonate Sediments and their Diagene-
sis. Elsevier, Amsterdam, 658 pp.
Berner, R.A. 1981. A new geochemical classification of the influence of climate, salinity and sea level. In:
sedimentary environments. Journal of Sedimentary Research Ga˛siewicz, A. and Słowakiewicz, M. (eds), Palaeozoic Cli-
51, 359–365. mate Cycles: Their Evolutionary and Sedimentological
Burchette, T.P. 2012. Carbonate rocks and petroleum reservoirs: Impact. Geological Society Special Publication 376, pp.
geological perspective from industry. In: Garland, J., 269–305.
Neilson, J.E., Laubach, S.E. and Whidden, K.J. (eds.), James, N.P. 1983. Reef environment. In: P.A. Scholle, D.G.
Advances in Carbonate Exploration and Reservoir Analysis. Bethout and C.H. Moore (eds.). Carbonate Depositional
Geological Society Special Publication 370, pp. 17–37. Environments. AAPG Memoir 33, 345–440.
Bjørlykke, K. 1989. Sedimentology and Petroleum Geology. James, N.P. 1984. Reefs. In: Walker, R.G. (ed.), Facies models.
Springer-Verlag, Berlin, Heidelberg. New York, London, Geoscience Canada, pp. 229–244.
Paris, Tokyo, Hong Kong, 363 pp. James, N.P. 1997. The cool-water carbonate depositional realm.
Bjørlykke, K., Bue, B. and Elverhøi A. 1978. Quaternary In: James, N.P. and Clarke, J.A.D. (eds.), Cool-Water
sediments in the northwestern part of the Barents Sea and Carbonates. SEPM Special Publication 56, pp. 1–20.
their relation to the underlying Mesozoic bedrock. Sedimen- James, N.P. and Clarke, J.A.D. 1997. Cool-water Carbonates.
tology 25, 227–246. SEPM Special Publication 56, 440 pp.
Bjørlykke, K. and Jahren, J. 2012. Open or closed geochemical Kastner, M. 1984. Control of dolomite formation. Nature 311,
systems during diagenesis in sedimentary basins: Constraints 410–411.
on mass transfer during diagenesis and the prediction of Kobluk, D.R. 1977. Calcification of filaments of boring and
porosity in sandstone and carbonate reservoirs. AAPG Bul- cavity-dwelling algae, and the construction of micritic
letin 96, 2193–2214. envelopes. In: R.C. Romans (ed.). Geobotany. Springer-
Buchen, van F.S.P., Gerdes, K.D. and Estebahn, M. 2010. Meso- Verlag US, pp. 195–207.
zoic and Cenozoic Carbonate Systems of the Mediterranean Lees, A. 1975. Possible influences of salinity and temperature on
and Middle East: Stratigraphic and Diagenetic Models. modern shelf carbonate sedimentation. Marine Geology 19,
Geological Society Special Publication 329, 424 pp. 159–198.
Chave, K.E. 1964. Skeletal durability and preservation. In: J. Lees, A. and Buller, A.T. 1972. Modern temperate water and
Imbrie and N.D. Newell (eds.). Approaches to Paleontology, warm water shelf carbonate sediments contrasted. Marine
pp. 377-387. John Wiley & Sons, Inc., New York. Geology 13, 1767–1773.
Clarkson, E.N.K. 1979. Invertebrate Palaeontology and Evolution. Logan, B.W., Rezak, R. and Ginsburg, R.W. 1964. Classifica-
George Allen & Unwin, London, Boston, Sydney, 323 pp. tion and environmental significance of algal stromatolites.
Croizé, D., Bjørlykke, K., Jahren, J. and Renard, F. 2010. Experi- Journal of Geology 72, 68–83.
mental mechanical and chemical compaction of carbonate Lucia, F.J. 1999. Carbonate Reservoir Characterization.
sand. Journal of Geophysical Research: Solid Earth 115 pp. Springer, Berlin, 226 pp.
Dunham, R.J. 1962. Classification of carbonate rocks according Lønøy, A. 2006. Making sense of carbonate pore systems.
to depositional texture. In: Ham, W.E. (ed.), Classification of AAPG Bulletin 90, 1381–1405.
Carbonate Rocks. AAPG Memoir 1. Tulsa, OK, pp. Moore, C.H. 1980. Porosity in carbonate rock sequences. AAPG
108–121. Continuing Education Course Note Series from the Fall
Ehrenberg, S.N., Walderhaug O. and Bjørlykke, K. 2012. Car- Education Conference 1980. AAPG Tulsa, OK, 124 pp.
bonate porosity creation by mesogenetic dissolution: Reality Moore, C.H. 1989. Carbonate Diagenesis and Porosity. Elsevier
or Illusion? AAPG Bulletin 96, 217–233. Science Publishers, Amsterdam, 338 pp.
Embry, A.F. and Klovan, J.E. 1971. A late Devonian reef tract More, C.H. 2004. Carbonate Reservoirs. Porosity Evolution and
on northeastern Banks Island, Northwest Territories. Bulle- Diagenesis in a Stratigraphic Framework. Developments in
tin of Canadian Petroleum Geology 19, 730–781. Sedimentology 46. Elsevier, Amsterdam, 444 pp.
Folk, R.L. 1959. Practical petrographic classification of Nielsen, J.K. and Hanken, N.-M. 2002. Late Permian carbonate
limestones. AAPG Bulletin 43, 1–38. concretions in the marine siliciclastic sediments of the
Folk, R.L. 1962. Spectral subdivision of limestone types. In: Ravnefjeld Formation, East Greenland. Geology of
Ham, W.E. (ed.), Classification of Carbonate Rocks. AAPG Greenland Survey Bulletin 191, 126–132.
Memoir 1. Tulsa, OK, pp. 62–84. Purser, B.H. and Evans, G. 1973. Regional sedimentation along
Garland, J., Neilson, J., Laubach, E.E. and Whidden K.J. 2012. the Trucial Coast. SE Persian Gulf. In: Purser, B.H. (ed.).
Advances in carbonate exploration and reservoir analyses. The Persian Gulf. Holocene Carbonate Sedimentation and
Geological Society Special Publication 370, 1–15. Diagenesis in a Shallow Epicontinental Sea. Springer-
Gebelein, C.D. 1969. Distribution, morphology and accretion Verlag, Berlin, Heidelberg, New York, pp. 211–231.
rate of recent subtidal algal stromatolites, Bermuda. Journal Purser, B., Tucker, M. and Zenger, D. 1994. Dolomites. Interna-
of Sedimentary Petrology 39, 49–69. tional Association of Sedimentologists Special Publication
Ginsburg, R.N., Rezak, R. and Wray, J.L. 1971. Geology of 21, Blackwell, Oxford, 451 pp.
calcareous algae (notes for a short course). Sedimenta I. Rafaelsen, B. and Nielsen, J.K. 2005. Introduction to
Comparative Sedimentology Laboratory, University of Carbonates. European Association of Geoscientists and
Miami, 61 pp. Engineers (EAGE), Learning Geoscience Journal, 57 pp.
Greensmith, J.T. 1978. Petrology of the Sedimentary Rocks. Rasmussen, H.W. 1969. Palæontologi: Fossile invertebrater.
George Allen & Unwin, 241 pp. Munksgaard, 420 pp. (in Danish)
Hanken, N.-M. and Nielsen, J.K. (2013). Upper Carboniferous- Roehl, P.O. and Choquette, P.W. 1985. Carbonate Petroleum
Lower Permian Palaeoaplysina buildups on Svalbard: Reservoirs. Springer, New York, 622 pp.
Sandberg, P.A. 1983. An oscillating trend in Phanerozoic non- Tucker, M.E. and Wright, V. P. 1990. Carbonate Sedimentology.
skeletal carbonate mineralogy. Nature 305, 19–22. Blackwell, Oxford, 482 pp.
Schmoker, J.W. and Halley, R.B. 1982. Carbonate porosity Wagner, C.W. and van der Togt, C. 1973. Holocene sediment
versus depth: A predictable relation for south Florida. types and their distribution in the southern Persian Gulf. In:
AAPG Bulletin 66, 2561–2570. Purser, B.H. (ed.). The Persian Gulf. Holocene Carbonate
Scholle, P.A. 1978. A Color Illustrated Guide to Carbonate Sedimentation and Diagenesis in a Shallow Epicontinental
Rock, Constituents, Textures, Cements, and Porosities. Sea. Springer-Verlag Berlin, Heidelberg, New York, pp.
AAPG Memoir 27, 241 pp. 123–155.
Scholle, P.A., Bebout, D.G. and Moore, C. 1983. Carbonate Wilkinson, B.H. 1979. Biomineralization, paleooceanography
Depositional Environments. AAPG Memoir 33, 708 pp. and the evolution of calcareous marine organisms. Geology
Scholle, P.A. and Ulmer-Scholle, D.S. 2003. Color Guide to 7, 524–527.
Petrography of Carbonate Rocks. AAPG Memoir 77, 474 pp Wray, J.L. 1977. Calcareous Algae. Elsevier, Amsterdam, 185
(2nd printing). pp.
Scoffin, T.P. 1987. An Introduction to Carbonate Sediments and Zenger, D.H. and Dunham, J. B. 1988. Dolomitization of Siluro-
Rocks. Blackie, Glasgow and London, 274 pp. Devonian limestones in a deep core (5,350 m), southeastern
Swart, P.K., Eberli, G. and Mckensie, J. 2009. Perspectives on New Mexico. In: Shukla, V. and Baker, P.A. (eds.), Sedi-
Carbonate Sedimentology. International Association of mentology and Geochemistry of Dolostones. SEPM Special
Sedimentologists. Wiley-Blackwell, Hoboken, NJ, 400 pp. Publication 43, pp. 161–173.
Chapter 6
Mudrocks, Shales, Silica Deposits and Evaporites
Knut Bjørlykke
6.1 Mudrocks and Shales sedimentary processes of erosion, transport and deposi-
tion since most clastic sediments contain significant
Mudrocks and shales are the most abundant lithologies amounts of clay. The properties of clays are not only
in most sedimentary basins. They are important controlled by the mechanical strength of the grains but
because shales include source rocks for oil and gas, also by the composition of the pore fluid. This is also
and recently large reserves of gas have been found in true during sediment compaction. Kaolinite has much
shales. Shales may therefore be reservoir rocks lower surface charge than smectite and illite.
because they may have significant porosity and some Mudrocks and shales are often treated as one lithol-
(although small) permeability to flow gas. The rising ogy, but they vary greatly as a function of both mineral
production of oil and gas from shales (shale gas and composition and grain-size distribution. A relatively
shale oil) in recent years has led to a new interest in the large fraction of grains may be larger than clay size, but
primary composition and the properties of mudstones as long as the larger particles are floating in a finer matrix
and shales as functions of the burial depth and temper- the properties are dominated by the clay-sized particles.
ature. The seismic image for sandstones also depends Here we will discuss siliceous (i.e. non-carbonate)
on the properties of adjacent shales. mudrocks and clay. The clay minerals in mudrocks
There is however no precise definition for mud or may have different origins:
mudrocks. It is used to describe fine-grained rocks with (1) Clay minerals formed by weathering of igneous
a relatively high content of clay-sized particles, mostly and metamorphic rocks.
clay minerals but also other minerals. Carbonate mud is (2) By erosion of older shales and mudrocks.
discussed under carbonate sediments. The upper limit (3) From volcanic ash.
for clay particles is 0.004 mm in the geological litera- (4) By diagenesis on the seafloor and during burial.
ture, but in the engineering literature (soil science) The clay mineral assemblage produced by
0.002 mm is commonly used. Because clay minerals weathering depends on the composition of the rocks
are essentially flat flakes, they have large surface areas, that are being weathered, and the climate. A humid
some minerals like smectite having several hundred m2/ climate will favour the formation of kaolinite. In a gran-
g. Kaolinite minerals are much larger (>0.01 mm) and ite or gneiss, feldspar, mica and most other silicate
have much smaller surface area. There is a cohesion minerals will dissolve and the aluminum and silica will
between small particles, and clay minerals also have a precipitate as kaolinite, and after long periods of
surface charge due to broken bonds in the mineral weathering, as gibbsite (Al(OH)3. The quartz grains
structure. This cohesion plays an important role in will be weathered out as sand and their grain size will
reflect the quartz crystal size range in the parent rock (see
Chap. 3). The result is a bimodal distribution of sand and
kaolinitic clays.
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway In areas with mostly basic rocks like anorthosite,
e-mail: [email protected] gabbro and basalts, weathering will only produce clay
218 K. Bjørlykke
minerals, since there is no quartz which in more acid

rocks forms most of the sand fraction. Minerals like
basic plagioclases, pyroxenes and hornblende will
quickly break down to clay minerals such as kaolinite,
and after long periods of weathering gibbsite (Al(OH)3.
6.2 Supply of Clay Minerals to

Sedimentary Basins
While kaolinite is derived from humid climate

weathering, smectite and illite are more typical of deserts
because there is less flow of fresh (meteoric) groundwa-
ter. Chlorite is mostly derived from erosion of metamor-
Fig. 6.1 Satellite photograph showing the distribution of clay
phic rocks in relatively cold climates where weathering outside the Mississippi delta. The limited extent of the delta mud
is slow. In warmer and wetter climates chlorite will is due to flocculation of the clay particles when freshwater is
break down, but may occur near basalts and basic volca- mixed with ocean water
nic rocks, particularly in the marine environment where
chlorite is more stable. Volcanic ash consisting of glass
and unstable volcanic mineral assemblages may alter to Overburden stress
smectite, both on land and on the seafloor. In deep sea
sediments, zeolites like phillipsite are common.
Erosion of older mudrocks and shales can produce --
Na+
nearly all types of clay minerals. Glacial clays are K+
essentially mechanically ground down sedimentary,
metamorphic or igneous rocks. Since there is very little
chemical weathering, their chemical composition is
nearly the same as the rocks eroded. Chlorite and illite
Fig. 6.2 Poorly compacted clays have a “house of cards”
are formed by disintegration of mica and metamorphic structure. Clay minerals have negative surface charges due to
chlorite, and most of the quartz and feldspar is pre- broken chemical bonds in their silicate structure. In freshwater
served. We find such clays accumulating in front of there is repulsion between the clay minerals, keeping them in
modern glaciers that terminate in lakes or the ocean. suspension. In seawater, cations like sodium and potassium help
to neutralise these charges and the clays will form denser
When the continental ice sheet withdrew from aggregates (flocculation)
Scandinavia 10,000–9,000 years ago, thick glaciomarine
clays were deposited in the fjords and sea. Subsequent
postglacial isostatic uplift elevated the inland portions content in the water will cause the clay particles to
by up to 200 m above the present sea level. flocculate, which makes them sink faster to the bottom
Clay transported by rivers into lakes will remain near to the river mouth. This is because clay minerals
suspended for some time (due to slow flocculation) have a negative charge which prevents them from stick-
and very finely laminated clayey sediments will be ing together in freshwater. In seawater these negative
deposited. In lakes there is usually less bioturbation charges are neutralised by the cations in seawater, such
to destroy the lamination than in marine environments. as Na+, K+, Mg++ and Ca++. K+ is most effective
Freshwater sediments tend to be finely laminated com- because it is not so strongly hydrated (surrounded by
pared to most marine sediments. water molecules) as Mg++, Na+and Ca++ (Fig. 6.2).
Clay from rivers is not transported far offshore and Clay minerals transported into marine sedimentary
we see from satellite pictures that there is clear water basins will also be subjected to sorting by grain size.
not so far from the delta front (Fig. 6.1). When the clay Recent sediments in the Gulf of Mexico contain about
minerals come into contact with seawater, the salt 30–40% silt in addition to clay and compact rather
6 Mudrocks, Shales, Silica Deposits and Evaporites 219
efficiently from about 80% porosity on the sea floor to (seawater) is usually undersaturated with respect to
about 35–40% at about 500–600 m depth (Day-Stirrat quartz. This is because the organisms can precipitate
et al. 2012). This may be due partly to poor sorting. silica from seawater even if the concentration of SiO2
Kaolinite is the coarsest grained of the clay minerals is far less than 1 ppm. While amorphous silica has a
and will therefore be deposited in the most proximal solubility of about 150 ppm, cristobalite and tridymite
parts of deltas and shorefaces while illite and smectite have a solubility of 6–15 ppm, depending on the
will be transported further out into the more distal parts degree of order in the crystals. The most important
(Fig. 6.1). As sedimentation progrades into deep water silica-producing organisms are:
on a continental margin illitic and smectitic clay will Phytoplankton: diatoms and silicoflagellates
form the base and grade upwards into more kaolinite- Zooplankton: radiolaria
rich sediments (Bjørlykke 2011). Silica sponges
From deltas and the shelf edge, clayey sediments as These organisms are built up of amorphous silica
well as sand may be transported down the slope by (opal A). The total production of organic silica in the
turbidity currents or debris flows. In the most distal oceans has been estimated to be from 2 1010 up to
shelf or deep-water facies, sedimentation rates can be 1011 tonnes/year. The largest contribution comes from
very low and aeolian dust may make up much of the diatoms, and also a very large percentage (50–70%) of
sediment deposited. This is particularly true offshore the total primary production of carbon (2 1010
dry areas like the Sahara in West Africa. This mud is tonnes/year) is ascribable to diatoms consisting of
rich in fine-grained smectite and illite, often with iron about 60% silica and 40% carbon.
oxides which may serve as an important nutrient for The flux of silica into the oceans from the continents
organic production in the South Atlantic Ocean. Volca- is far less than the organic production of siliceous
nic ash may be deposited over large areas. During organisms. The approximately 4 108 tonnes/year
transgressions, extensive layers of mud may be depos- brought by rivers only represents about 1% of all
ited on shelves along the basin margin. This mud may organic silica precipitation. Some silica is added to the
later be eroded and supplied to the basin during periods sea through submarine volcanism along the oceanic
of uplift, so that the prograding regressive sequences spreading ridges. Seawater is being circulated
may be rich in clay. Weathering of basic rocks like (convected) through hot basalt, dissolving silica and
anorthosites and gabbroid rocks proceeds rapidly and precipitating sulphides from the sulphate in the seawa-
produces almost only clay minerals and very little sand, ter. Even with this source of silica, the supply to the
because of the lack of quartz. ocean water can not balance the amounts precipitated
biologically. Since we must assume that the composi-
tion of seawater has been relatively constant, this means
6.3 Silica (SiO2) Deposits that only about 1% of the overall organic silica produc-
tion is retained in sedimentary deposits. Most of the
Silica which is liberated through weathering goes into silica from plankton redissolves in the undersaturated
solution as silicic acid (H4SiO4). Even if quartz and seawater before it reaches the bottom, and some also
also feldspar have relatively low solubility, large dissolves on the seafloor and diffuses up into the water
quantities of silica are transported by rivers out into column. Consequently it is only when the rate of
the sea. The total amount of dissolved silica added organic precipitation is higher than the rate of solution
annually to the sea is estimated to be about 4 108 that we find deposition of organic silica.
tonnes. Some silica is also introduced from the mid- Diatoms dissolve because seawater is undersatu-
oceanic ridges, but is probably of only very modest rated with respect to silica, and large quantities of
significance compared with the fluvial input. An equal organic material can thereby be released without oxi-
amount of silica must be removed for the seawater dation taking place. Organic matter produced by the
composition to remain constant. Silica is removed solution of diatoms thus constitutes a large part of the
biologically as biogenic silica, and through the forma- total organic matter accumulated. Phosphates, nitro-
tion of silicate minerals in the sea. Radiolaria and gen and various trace metals are also released through
diatoms are particularly efficient at removing silica. the disintegration of plankton as they sink through the
As a result, even though quartz is only very slightly water column. These recycled nutrients can once more
soluble in seawater (3–6 ppm SiO2), surface water provide the basis for organic production when water
220 K. Bjørlykke
Prevailing wind
Little supply of
Sea level clastic sediments
Upwelling – supply
of nutrients (N, P)
Silica Precipitation of silica
Opal A and phosphates,
siliceous sponges,
Opal CT diatoms, radiolaria
(50–60°C) Increasing
burial
Quartz
(>60–80°C)
Fig. 6.3 Upwelling of water rich in nutrients causes biological precipitation of silica and phosphates. The silica deposits (opal A)
will, when buried, be altered to opal CT and then to quartz
wells up to the photic zone (Fig. 6.3). The upwelling

currents also prevent supply of clastic sediments from
land, so that nearly pure silica can be deposited.
Opal A will decompose to opal CT which may
consist of bladed crystals forming small spheres called
lepispheres. Because more energy (temperature) is
required to precipitate quartz, minerals like cristobalite
and tridymite are formed. These are minerals which are
stable at very high temperatures (1,000–1,500 C), but
precipitate out instead of quartz at low temperatures,
even though quartz is thermodynamically more stable.
This phase is called opal CT, sometimes also
porcellanite. Opal CT will, when subjected to higher
temperatures, slowly dissolve and the silica will be
precipitated as quartz.
Fig. 6.4 Diatoms and radiolaria from Oligocene siliceous
Amorphous silica (opal A) dissolves and is
sediments from the North Sea basin (1,430 m depth). From
replaced by opal CT, usually at a temperature of Thyberg et al. (1999)
around 50–70 C which corresponds to about
1.5–2 km of overburden at average geothermal
gradients. The reason why amorphous silica (opal A) silica but there is little net accumulation. In the Ter-
can exist so long despite being thermodynamically tiary sequence in the North Sea there are major silica
unstable, is that quartz does not crystallise at low beds. Those with the greatest extent are associated
temperatures. Opal CT is transformed into quartz at with ash layers from Eocene volcanicity related to
temperatures around 60–80 C. The change in acoustic the opening of the Norwegian–Greenland Sea and
impedence which accompanies the transition from consist of radiolarians and diatoms together with
opal A to opal CT and then to quartz (chert) may altered volcanic sediments (ash) with abundant smec-
produce a significant seismic reflection. Because tite. Well-cemented silica beds of Eocene age are
these reactions are controlled by temperature they called Moler in Denmark, and Balder Formation in
tend to occur as horizontal zones that may be mistaken the North Sea where it generates a very prominent
for a fluid contact (gas/water or oil/water). seismic reflector.
Vast amounts of diatoms are found today round In the Oligocene there are also nearly pure silica
Antarctica, and the thick sediment accumulations beds, and where they are buried to less than
there have a very high content of amorphous silica. 1,500–1,600 m, fossils of opal A such as diatoms are
In the North Sea too there are now large amounts of exceptionally well preserved (Fig. 6.4). The silica is
still opal A. When altered to opal CT most of the

primary structures are gone.
When opal A and opal CT dissolve, the water
becomes oversaturated with respect to quartz which
then begins to crystallise at a large number of points
(nucleii). It thus forms microcrystalline quartz, known
as “chert”. The dark colouration is due to organic
material. In Europe, flint is the best known variety of
chert which is abundant as concretions in the Upper
Cretaceous Chalk. Cherts are formed of silica which
has been finely distributed in the sediment, largely as
sponges and radiolaria often concentrated along spe-
cial horizons. Because small particles with a large Oil seep in chert. Santa Maria, north of Santa Barbara
specific surface are highly unstable, they will dissolve Fig. 6.5 Monterey chert with oil seeps, California
and silica will be precipitated in nodules which are
massive structures with a small specific surface and in
consequence thermodynamically more stable. similar chert of Upper Miocene age is found. The
Novaculite is more or less synonymous with light, Monterey Formation is both an important source and
laminated chert which contains less organic material. reservoir rock for oil in California, especially in the
Porcellanite is a term applied to more contaminated Ventura Basin. It is a source rock on account of the
silica (chert) which has the appearance of unglazed organic matter produced by the diatoms. After the
porcelain. Laminated chert is very common in transformation to opal C and quartz, the rock fractured
Palaeozoic and Mesozoic sequences immediately due to tectonic folding, so that it became a fractured
above oceanic seafloor (ophiolites) near spreading reservoir (Fig. 6.5).
ridges, sourced from water circulating through the Subjected to 1–2 km of overburden, amorphous
basalts. This chert may be white, grey or dark, silica is converted to quartz and the rock becomes
depending on the content of organic matter or traces brittle and fractures during folding and faulting to
of iron, magnesium etc. form good fractured reservoirs.
Chert containing oxidised iron has a reddish The Monterey Formation contains organic-rich
colouration and is called jasper. Jasper may form on shales and phosphate deposits. This is a very common
the seafloor near the spreading ridges. When water association when sediments have been deposited in
circulates through the spreading ridges it carries upwelling areas with a limited clastic supply. Trans-
away iron which then oxidises in the seawater and formation of amorphous silica to quartz appears to
produces the red colour. When mixed with silica depend on the chemical environment (temperature,
which is precipitated mostly by organisms, chert is pH, ionic strength, Mg concentration). Clay minerals
formed. Jasper is found on top of obducted ocean reduce the conversion rate. Because of its high organic
floor basalts in nappes of Ordovician age in western content the Monterey formation has recently been
Norway. considered to have a potential for production of shale
Chert has also been found during ocean drilling, oil. It is however not clear if this will be permitted for
overlying present ocean floor of Cretaceous or Ter- environmental reasons.
tiary age. Although all amorphous silica has been In soils, however, quartz may precipitate at low
converted to quartz, it is often possible in chert to temperature near the surface, forming silcrete
detect the remains of organic particles, e.g. radiolaria, resulting in very hard ground.
especially in Palaeozoic chert deposits. Younger silica Precambrian chert deposits typically occur in con-
deposits are chiefly precipitated by diatoms. In areas junction with banded iron formations (BIFs). It is
with upwelling, silica may also be deposited in rela- uncertain whether there were organisms which could
tively shallow water. The Monterey Formation in precipitate silica in Precambrian times, or whether
California is a well-known example of chert formed Precambrian chert was chemically deposited. There
in an upwelling zone. Around the Pacific Ocean are no definite indications of biological precipitation
222 K. Bjørlykke
of Precambrian chert, and since silica must also have Sulphates

been added to the oceans, we must assume that the Alkaline carbonates
ocean was saturated with respect to silica, and that Ca-Mg carbonates
there may have been inorganic precipitation of silica. Borates
Silica deposits may also be formed in lakes by Nitrates
freshwater diatoms. In lakes along the rift systems of Silica deposits
East Africa, there are thick deposits of diatomite Iron deposits
which are exploited for use in insulating materials. A
great deal of CO2 is taken from the lake water by
diatoms and other algae for photosynthesis, and the
6.4.1 Marine Evaporite Environments
water therefore becomes strongly basic (pH 9–10).
This increases the solubility of silica, thus increasing
Although the salt content of seawater varies somewhat
the corrosion of silicate minerals. Examples of this are
in the different parts of the world’s oceans, the com-
found, for instance, at Lake Turkana in Kenya. Volca-
position of seawater is relatively constant. The table
nic rocks and water from hot springs are also impor-
below shows the percentage composition of dissolved
tant sources of silica.
salts in seawater which add up to a salinity of 35‰. To
Chemical precipitation of silica may also take place
the right are the percentages of the various salts
in evaporite basins, and in ephemeral lakes which dry
obtained through evaporation.
up between rainy seasons.
Percentage by weight of salts in seawater
Percentage in Percentage by weight of common
6.4 Evaporites Ion seawater Salt salts after evaporation
Na 30.64 NaCI 77.76
Evaporites consist of minerals which have crystallised Mg 3.76 MgCl2 10.86
out through evaporation of water. This can happen in Ca 1.20 MgSO4 4.74
many ways: K 1.09 CaSO4 3.60
1. Evaporation of seawater in completely or partly C1 55.21 K2SO4 2.47
cut-off marine basins. SO 7.70 MgBr2 0.22
CO 0.21 CaCO3 0.35
2. In lakes which have little or no outlet and a high
Br 0.19
evaporation rate.
100.00 100.00
3. Through evaporation of seasonal precipitation
which collects in topographical depressions with- By evaporating seawater one can therefore deter-
out outlets (playas). mine the relative amounts of different salts. An evap-
4. In soil profiles or sandy sediments, through evapo- orite basin will often have some supply of seawater,
ration of groundwater. such that the evaporation is not total. As a result,
5. In arctic areas, the sublimation of ice and the freez- evaporite deposits may accumulate carbonates and
ing of seawater to ice both increase the salt concen- sulphates, while the chlorides remain in solution.
tration of sea water, and evaporite minerals such as When seawater evaporates, carbonates are among
gypsum may be precipitated. the first salts to precipitate, but the amount of carbon-
6. Through solution and precipitation of salts from ate in solution is very small. When the volume of
older evaporite deposits. seawater is reduced to 1/3–1/5, both CaCO3 (arago-
The formation of evaporites is therefore not an nite) and CaSO4 2H2 O (gypsum) have precipitated.
unambiguous indication of a high temperature, but Only when the volume is down to 1/10 will NaCl
most evaporite deposits (salt) form in the climatically (halite), quantitatively the main constituent, be
dry belts about 20–30 from the equator. Evaporites precipitated (Fig. 6.6). MgSO4 and MgCl2 will be
contain a number of salt minerals which are too solu- precipitated at the same time. Polyhalite

ble to be precipitated in normal marine or continental Ca2 K2 MgðSO4 Þ5 2H2 O commonly precipitates
environments. The most important are: when the seawater has been reduced to 1/20. KCl
Chlorides (sylvite) and bromides are among the most soluble
80˚C
Anhy-
drite
stable
60˚C Anhydrite (CaSO4)
stable
Gypsum
soluble
40˚C
Sylvite (KCl)
Gypsum
Halite and Mg-salts
20˚C (CaSO4.2H2O)
(NaCl) stable
stable
stable
1/1 1/3 1/5 1/7 1/10 1/20 Remaining volume
Increasing evaporation of seawater
Fig. 6.6 Stability of salt minerals as a function of progressive evaporation of seawater and temperature
salts, and are hence the last to precipitate. If the salt Gypsum is the mineral which normally forms in
concentration increases with evaporation, this is the marine evaporites, but anhydrite is also observed in
sequence of salt deposition. During periods with modern evaporites in supratidal zones. This is true, for
greater circulation and supply of normal seawater the example, of the sabkha deposits in the Persian Gulf.
salt concentration will fall, and there will be cyclic salt These are black algal muds in the supratidal zone,
deposition and solution. Chlorides (NaCl, KCl) will where temperatures may be up to 80 C, and anhydrite
dissolve in moisture from the air unless the humidity is is deposited in the sediment. In patches of open water
very low. where the temperature is lower, gypsum forms instead.
Periods of increased evaporation and hence high Evaporite sediments have formed (or preserved) in
salt concentration may alternate with influxes of sea- the sedimentary record throughout geological history,
water with a more normal salt concentration, produc- but appear to have formed more abundantly during
ing cycles representing changes in salinity. These are certain periods, particularly the Permian. In Northern
called evaporite cycles. In many evaporite basins the Europe and the North Sea we find thick evaporites
salt concentration has been sufficiently high for gyp- from this period in what is called the Zechstein Sea,
sum to be precipitated, but not high enough for which serve as a cap rock for the gas in the underlying
chlorides (see Fig. 6.7). Permian aeolian sandstones.
The stability of the various salts during evapora- The lighter salt layers will flow upward and gradu-
tion can be determined experimentally, or estimated ally form large mushroom-shaped or columnar
through physical chemistry calculations. However, it structures. The prerequisite for initiating this salt
is clear that certain metastable mineral phases can dome formation, however, is that the salt beds must
also be formed. Calcium sulphate may precipitate be at least 100–200 m thick. Salt domes are common
both as a hydrated mineral, CaSO4 2H2 O (gypsum), in Northern Germany, Denmark and the southern part
and as a non-hydrated mineral CaSO4 (anhydrite) of the North Sea, where they may form oil or gas traps.
(see Fig. 6.8). Which of these two phases forms as During the Permian, NW Europe was in the dry belt
a result of oversaturation of calcium sulphate 20–30 N, similar to the Sahara today.
depends on the temperature, salinity and water Surveys of the bottom of the Mediterranean Sea
vapour pressure. In a solution of CaSO4 alone, anhy- have revealed the existence of considerable evaporite
drite forms only at temperatures of over 60 C, but as deposits in Upper Miocene sequences. Adjacent land
the concentration of other salts increases, anhydrite areas also have numerous localities where salt beds of
may form at temperatures down to 25–30 C this age (Messinian salt) have been preserved. This
(Fig. 6.6). can indicate that the Straits of Gibraltar were closed
224 K. Bjørlykke
Fig. 6.7 Layers of gypsum in sabkha
for large parts of the Tertiary and that the Mediterra-

nean Sea was a vast desert inland sea 2,000–3,000 m
below sea level at that time.
6.5 The Sabkha Model
“Sabkha” is the Arabic term for the large flat areas

around the Persian Gulf. This area extends from the
tidal zone, forming an 8–10 km wide belt with a very
low seaward gradient of only about 0.4 m/km. The
sabkha flats were created by carbonate sediments
infilling lagoons, chiefly during the Holocene. For Fig. 6.8 Gypsum crystals growing on the sabkha surface
the last 4,000 years there has been a gradual regression
in the area. The climate is very dry (30–100 mm
annual precipitation), but this is nevertheless sufficient will dissolve easily the next time salt water flows in
to cause groundwater from surrounding areas to flow over the sabkha. Moisture in the air will also help to
out to the sea. The groundwater flow is extremely dissolve halite on the surface, so that it is not preserved
slow, only a few centimetres a year, but is nevertheless in the bedding series.
responsible for the diagenetic environment containing The sediments which are deposited alternate
porewater of continental origin. During storms and between gypsum-carbonate and some clastics. A con-
high tides, marine water floods the sabkha. Some spicuous lamination develops because of algal growth
runs off, some evaporates and some filters down into (cyanobacteria), which does not get destroyed because
the sediment and mixes with the groundwater. Rela- there is no bioturbation in this salt environment.
tively large amounts of gypsum are then formed. The Although it is hot, up to 50–80 C on the sabkha,
remainder of the water is highly enriched in magne- evaporation is limited where there is no open water
sium, and dolomite is formed. Evaporation will because of the low permeability within the sediments.
increase the salt concentration of the porewater, and When the water table sinks, the evaporation rate rap-
if it is intense enough, chlorides (halite) are idly declines. Thick layers of sediments with gypsum
precipitated in addition to gypsum. However, halites make up much of the Sabkha sediments (Fig. 6.8).
Whereas the rate of evaporation from open water in horsts or by volcanoes and lavas. The Red Sea used
the Persian Gulf is about 124 cm/year, it is only about to be an active evaporite basin but gradually the con-
6 cm/year in the sabkha. We therefore have only nection with the Indian Ocean has become too large.
moderate “evaporite pumping” where chlorides are Rifting, and spreading of the ocean floor in connec-
precipitated above the water table and sulphates tion with the formation of the Atlantic Ocean, led to
below the water table. The chlorides deposited will, the accumulation of thick evaporite series in the dry
however, redissolve as the overburden increases and regions and freshwater lakes in the wetter parts. In
the relative position of the water table rises. In some Jurassic and Lower Cretaceous times early seafloor
semi-arid regions, e.g. the Coorong region of spreading resulted in the formation of a series of
Australia, dolomite lakes form which evaporate to evaporite basins in the area between Africa and
dryness in the summer (ephemeral lakes). Evaporite South America south of the equator, and in the Gulf
minerals deposited in summer will redissolve during of Mexico and North Africa north of the palaeo-
the winter rains, and will not be preserved in the equator. Where they were sufficiently thick, they
bedding series. formed diapirs which greatly influenced further sedi-
mentation and the structural development of these
parts of the continental shelf. As seafloor spreading
6.6 Marine Evaporites continued, such that the Atlantic Ocean widened and
the ocean floor basalts cooled and subsided more rap-
Evaporation from the surface of the sea will cause the idly, the opportunities for forming closed basins
salinity of the water in a basin to increase. If there is diminished. After the mid-Cretaceous no major evap-
little wave or current action, the warm surface water orite basins formed in this region. The salt is an ideal
will not mix quickly with the underlying, colder water. cap rock and both in the Gulf of Mexico and offshore
If the salt concentration increases, the density will Brazil large reservoirs of “sub-salt” oil have been
increase and the surface water will sink to greater found.
depths and mix with the water there, despite being
warmer. To create a high salinity basin we must have
physical barriers reducing or totally blocking the con-
nection to the open ocean, and the evaporation must be 6.8 Evaporites in Lakes and Inland Seas
greater than the total amount of freshwater added to
the basin by precipitation, rivers and groundwater. If Basins with no outlet are formed particularly in tec-
the evaporation is lower than the supply of freshwater tonically active areas. In East Africa we find numerous
it will develop into a freshwater basin. freshwater evaporites in the rift basins where the cli-
An evaporite basin with a limited connection to the mate is sufficiently arid, and at the end of the
open sea may not dry out; it can then accumulate large Cainozoic during the Mio-Pliocene there were many
amounts of gypsum, but not more soluble salt like landlocked lakes and inland seas in connection with
halite (NaCl). The Mediterranean has enough seawater rift valleys. Since the chemistry of river water is quite
exchange through the Straits of Gibraltar to prevent different from that of seawater, such “continental”
the formation of evaporites in the Mediterranean evaporites are quite different from marine series. The
today. composition will vary according to the types of rocks
and the weathering in the drainage area around the
lake. Lake evaporites normally contain large amounts
6.7 Tectonic Control of the Formation of carbonate, particularly sodium carbonate and a
of Evaporite Basins number of other salts, hence the name “soda lakes”.
The mineralogical composition of these soda
Partial or complete severing of marine basins occurs as deposits is very complex. Two important minerals
a result of tectonic uplift of barriers. Rifting and incip- are trona Na2 CO3 NaHCO3 2H2 O and gaylussite,
ient spreading of the ocean floor provide ideal CaCO3 Na2 CO3 5H2 O. If there is volcanism in the
conditions for the formation of evaporite basins, as area, as is often the case around rift valleys, this will
marine basins may be partly or entirely cut off by modify the composition of the water, both by
226 K. Bjørlykke
weathering of ash and lava and by the direct addition come into contact with groundwater due to erosion
of volcanic water (springs) rich in dissolved salts. of overlying sediments they will gradually, depending
In playa lakes there may be total evaporation of on the water circulation, hydrate to gypsum which is
seasonal rainfall resulting in thin layers of carbonates, then the stable phase. The loss of water resulting from
including dolomite and sulphates. Signs of dessication the transition from gypsum to anhydrite leads to a
and wind reworking will be common. Salt and clay volume reduction (compaction) of 38%, and the tran-
particles in the dried-out lakes will tend to be sition from anhydrite to gypsum leads to a
transported by the wind, to form small dunes. Water corresponding increase (expansion) of 60%. The vol-
from occasional rains in deserts will often collect ume increase can produce near-surface deformation
between large aeolian dunes and form interdune and faulting in basins where evaporites have been
lakes and sabkhas. uplifted due to erosion or diapirism. This compaction
and expansion creates serious geotechnical problems
in those areas of Europe, e.g. Switzerland and
6.9 Evaporation of Groundwater Germany, where evaporite deposits are common.
Sulphates are rather stable and are not easily reduced
Where groundwater evaporation exceeds rainfall there to sulphides inorganically; this requires sulphate-
is net upward transport to the surface of the soil and reducing bacteria. However in evaporite deposits
salts have to precipitate out in the soil profile. This can which have been buried to depths where the tempera-
happen in the vadose zone near the sea, or inland ture has exceeded 80 C, sulphate-reducing bacteria
where evaporation is high and the groundwater has have not survived and the rocks are then pasteurised.
flowed from another area. Gypsum and chlorides The reduction of sulphides like gypsum and anhydrite
exert great crystallisation power, and can push aside will then be very slow even at shallow depth.
the sediment matrix so that large euhedral crystals
form. In brown soil types such as prairie soils, where
the rainfall is high enough to prevent the accumulation 6.11 Iron- and Manganese-Rich
of more soluble salts, a layer of carbonate (caliche) is Sediments
commonly formed and in drier areas also gypsum.
When such soils are eroded the broken up caliche Iron and manganese share many similarities in their
will form pebbly conglomerates. geochemical behaviour in sedimentary environments.
Evaporites from older geological periods are easily Both elements are poorly soluble in the oxidised state
dissolved by groundwater or surface water. The Dead because they form hydroxides and oxides: Fe(OH)3,
Sea is an example of an evaporite basin where water Mn(OH)4, Fe2O3 and MnO2. In the reduced state they
flowing into the basin is already rich in dissolved salts. occur as Fe2+ and Mn2+ and are then much more
This is because the Jordan river runs through evaporite soluble. Both iron and manganese can be precipitated
sediments of Cretaceous and Tertiary age. In addition, as carbonate (FeCO3, MnCO3), either as separate
water of volcanic origin enters the rift valley through minerals, or as part of calcite, dolomite or ankerite.
faults and fractures. They are therefore not very soluble in basic solutions
with low redox potentials, but are quite soluble at low
pH values and in the reduced state. Consequently we
6.10 The Stability of Gypsum and can precipitate iron and manganese in two ways: (1)
Anhydrite During Burial Diagenesis through oxidation, (2) by changing the solution from
acid to basic.
Higher pressure and temperature will favour the sta- Fe2+ forms iron sulphides with very low solubility
bility of anhydrite, which has a more compact struc- while manganese sulphides are far more soluble and
ture than gypsum. Gypsum formed in evaporite therefore more soluble in porewater which is reducing
environments will therefore turn into anhydrite when and acid. Based on these geochemical considerations,
there is sufficient overlying sediment, usually at we can predict a great deal about the deposition of iron
depths ranging from just under 1,000 to 3,000 m. and manganese in sediments. As silicates and other
When anhydrite-bearing sediments are uplifted and minerals dissolve in the oxidising environment during
weathering, the concentration of insoluble iron oxides precipitated in the reducing zone below the redox
(Fe(OH)3) and manganese (Mn(OH)4) will increase, boundary.
along with aluminium (Al(OH)3), and constitute a Manganese will not be as easily trapped in the
major part of laterite. sulphate-reducing zone, as Mn2+ does not form such
Bogs, such as we find in the Scandinavian stable sulphides as Fe2+, and it will therefore have a
highlands, have conditions ideal for such precipitating greater tendency to be precipitated in the oxidised
iron (“bog iron”). Water seeping through bogs will zone.
have low Eh and pH, and iron-containing minerals During breaks in sedimentation, or slow sedimen-
from underlying rocks will be dissolved by humic tation, porewater expelled by compaction will cross
acids, transported in the reduced state and precipitated the redox boundary and this may cause the precipita-
through oxidation where the water flows out of the bog. tion of iron and manganese on the seafloor. Manga-
Porewater in sediments and sedimentary rocks is nese nodules are concretions of manganese and iron
normally reducing. This is because most sediments hydroxides and oxides found on the seafloor, particu-
contain reducing agents, particularly organic material. larly in the ocean basins (South Atlantic and Pacific).
Only the upper 0–20 cm below the sea (lake) floor are They also contain relatively high concentrations of
oxidised. The depth to which the groundwater will be metals like Ni, Cu, Zn and Co and there have been
oxidising depends on the supply of oxygen, i.e. the plans to mine these deposits. The nodules grow by
oxygen content and rate of flow of groundwater com- very slow concentric accretion in a pelagic ooze.
pared to the consumption of oxygen due to oxidation. Plankton is capable of accumulating very high
Desert sediments contain little organic material which concentrations of metals from seawater and when the
can function as a reducing agent, and the groundwater organisms dissolve, the planktonic ooze becomes very
will therefore remain oxidising longer. This can rich in these metals, which will be precipitated
explain the red colouration of desert sediments which together with the manganese hydroxides in the
is due to trivalent iron. concretions.
The redox boundary, which normally lies just Iron and manganese are virtually insoluble in
below the sediment/water boundary on the ocean bot- oxygenated sea water, and consequently cannot be
tom, represents an important geochemical trap. The transported in ordinary solution. However, rivers can
concentration of elements on each side of this bound- carry a good deal of iron and manganese adsorbed
ary is very different because of the different solubility onto organic particles or as clay-sized iron oxides
of elements in the two separate chemical which in many cases may produce a red colour.
environments. A high concentration gradient of Fe2+ Iron can be precipitated not only by oxidation, but
and Mn2+ on the reducing side of the redox boundary also in the reduced state in a basic environment. Thin
leads to a diffusion and precipitation of oxidised iron carbonate laminations in shales represent a local high-
and manganese immediately above the boundary. At a pH environment, where reduced iron which is nor-
depth where there is little oxygen in the porewater, mally soluble can be precipitated:
sulphate-reducing bacteria consume the oxygen in the
sulphate ions to form H2S. The sulphate concentration Fe2þ þ CaCO3 ¼ FeCO3 þ Ca2þ
in the porewater declines rapidly downwards causing a
downwards diffusion. H2S is an acid which can release This reaction occurs at relatively low Fe2þ =Ca2þ
(leach) iron bound in various clastic minerals to form ratios because iron carbonate (siderite) is less soluble
FeS (machinawite), which is black and easily than calcite. If there is sulphur present, iron will first
oxidised, and then pyrite (FeS2) which is a little form the sulphide, however, and iron carbonate will
more stable. If the sediments have a high silica con- not be stable. Siderite is therefore most typically
tent, chamosite may also be formed. Glauconite must formed in freshwater basins, and not directly in the
be formed right at the redox boundary, since it sulphate-reducing zone in marine sediments. Siderite
contains both bivalent (Fe2+) and trivalent (Fe3+) can also be formed below the sulphate-reducing zone,
iron. Uranium and vanadium, which have low solubil- for example by reaction between aragonite and iron in
ity in the reduced state and are more soluble when the sediment.
oxidised, will be able to diffuse downwards and be
228 K. Bjørlykke
6.12 Phosphorite (or “Phosphatic which is an enrichment of 140,000 compared to

Deposits”) dissolved phosphorus in ordinary seawater.
Organisms with an amorphous silica skeleton (e.g.
Phosphorites are sedimentary layers in which phos- diatoms and radiolaria) will readily dissolve in sea-
phate minerals are the main components, mainly apa- water, and in some cases carbonate (calcite and par-
tite Ca5 ðPO4 Þ3 ðF; Cl; OHÞ. These crystals may contain ticularly aragonite) will also dissolve, resulting in
fluor (F ), chlorine (Cl ) or hydroxyl (OH ). further phosphate enrichment. Phosphate minerals
Fluorapatite Ca5 ðPO4 Þ3 F is the most important in then crystallise out of the phosphate-rich sediments
marine sediments. We also have francolite and often replace other minerals, e.g. carbonate. Apa-
ðCa; NaÞ5 ðF; OHÞðPO4 ; CO3 Þ3 , where some carbonate tite is a heavy mineral (specific weight 3.18) and may
has substituted for phosphate and where sodium also be enriched mechanically by weak traction
substitutes for calcium. Iron apatites such as strengite currents.
ðFePO4 2H2 OÞ are also relatively common minerals Apatite will often crystallise out as concretions in
in secondary (weathered) phosphatic deposits. bottom ooze, and erosion by traction currents may
Phosphorite rocks are an important source of fertil- concentrate these nodules into a conglomerate. On
izer. Large deposits such as those found in Morocco, the continental shelf, phosphate forms at depths of
Spanish Sahara and Senegal in West Africa, and in between 100 and 400 m, i.e. below the photic zone.
Florida, are of great commercial value. Understanding However, phosphate nodules and massive beds of fine-
how phosphate deposits form is therefore a matter of grained phosphate mud (phosphate micrite) may also
considerable economic interest. In most sedimentary form in lagoons where the water is less clear. These
rocks phosphorus is a trace element, and very special phosphate deposits will be easily eroded even as a
conditions are necessary for phosphate enrichment to result of minor regressions, and form conglomerates
take place. These deposits therefore tell us something of phosphate mudstone (micrite). Today phosphate
important about the environment of deposition. deposits are forming only in a few areas with strong
Guano deposits are formed on land from bird or bat upwelling, but in previous geological periods we find
excrement in which the phosphate gradually becomes very extensive phosphate beds, often associated with
concentrated as the other organic components are transgressions. The transgressions will hold clastic
leached out. For the phosphates to be preserved, the sediment back for a while, enabling biogenic matter
rainfall must not be too high, because it will dissolve to be concentrated. Phosphate beds are often
phosphate sediments on land. We also have freshwater associated with other authigenic (formed in situ)
phosphatic deposits, but it is the marine ones that are minerals which take a long time to form, particularly
the most important. glauconite and also manganese deposits.
The first prerequisite for marine phosphate deposits In Florida there are large phosphate deposits of
is that sedimentation must proceed very slowly, i.e. mid-Tertiary age, and the same beds are exposed on
there must be virtually no clastic sedimentation. In the floor of the continental shelf off South Carolina
consequence we find phosphate beds associated with (Blake Plateau), where the manganese-rich phosphatic
major or minor breaks in sedimentation, or with deposits total 109 tonnes.
periods of very slow sedimentation. It has long been Marine phosphate deposits may also be formed by
known that phosphate is formed in areas with strong fossils with a phosphatic skeleton, for example fish.
upwelling and high organic productivity. We have These deposits may also be extensive, and are often
good examples of this along the edge of the continen- called “bone beds”.
tal slope off Chile and Peru. Water welling up from Phosphate minerals such as apatite may contain
great depths brings with it nutrients which are considerable amounts of uranium and rare-earth
liberated when marine organisms disintegrate in metals which substitute for calcium. Weathering of
deeper water. When the water flows up to the surface phosphate deposits will lead to oxidation of uranium
the nutrients are consumed by organisms, mostly plant to UO2þ 2 , which is soluble in the form of uranyl ions,
plankton, which provide a high primary production. and uranium may be precipitated again in the reduced
These organisms contain about 1% P (dry weight), state (UO2) when it comes into contact with organic
matter. This is why black shales and source rocks tend Bjørlykke, K. 2014. Relationships between depositional
to have rather high radioactivy. environments, burial history and rock properties. Some prin-
cipal aspects of diagenetic processes in sedimentary basins.
Weathering of phosphate sediments will initially Sedimentary Geology 301, 1–14.
cause solution of carbonate and carbonate-containing Chamley, H. 1989. Clay Sedimentology. Springer, New York,
apatite so that the sediments become enriched in 623 pp.
fluorapatite. As calcium is removed through solution Day-Stirrat, R.J., Flemming, P.B., You, Y., Aplin, A.C. and
Pluijm, van der, B.A. 2013. Modification of mudstone fabric
of calcite by rainwater, the sediment will become and pore structure as a result of slope failure; Ursa Basin,
richer in iron and aluminium, and iron and aluminium Gulf of Mexico. Marine Geology v 341, 56–67.
phosphates will form. Phosphate deposits are also Day-Stirrat, R.J., Flemming, P.B., You, Y., Aplin, A.C. and
formed through weathering of phosphate-bearing Pluijm, van der, B.A. 2012. The fabric of consolidation in
the Gulf of Mexico mudstones. Marine Geology. v. 295–298,
igneous rocks. Carbonatites from the East African 77–85.
rift system contain a high percentage of apatite, and Delisle, G., Teschner, E., Faber, E., Panahi, B., Guliev, I. and
when carbonate is dissolved during weathering, apa- Alliev, C. 2010. First approach in quantifying fluctuating
tite becomes concentrated and forms phosphate gas emission and methane and radon from mud volcanoes
in Azerbaijan. In: Wood, L.J. (ed.), Shale Tectonics. AAPG
deposits which are also rich in iron minerals. There Memoir 93, pp. 211–224.
are many types of apatite deposits and even if there is a Schieber, J., Eier, W. and Sethi, P. 1998. Shales and Mudstones.
high demand for phosphate in fertilizers it is unlikely Schweizerbartsche Verlagsbuchhandlung, Stuttgart, 384 pp.
that they would be exhausted in the next few centuries. Thyberg, B.I., Stabell, B., Faleide, J.I. and Bjørlykke, K. 1999.
Upper Oligocene diatomaceous deposits in the northern
North Sea – Silica diagenesis and paleogeographic
Further Reading implications. Norsk Geologisk Tidskrift 79, 3–18.
Thyberg, B., Jahren, J., Winje, T., Bjørlykke, K., Faleide, J.I.
and Marcussen, Ø. 2010. Quartz cementation in Late
Abtew, W. and Melesse, A. 2013. Evaporation and Evaoptran- Cretatceous mudstones, northern North Sea. Changes in
spiration. Meaurements and Estimations. Springer Verlag, rock properties due to dissolution of smectite and precipita-
206 pp. tion of micro-quartz crystals. Marine and Petroleum Geol-
Aplin, A.C., Fleet, A.J. and Macquaker, J.H.S. 1999. Muds and ogy 27, 1752–1764.
mudstones: Physical and fluid-flow properties. Geological Weaver, C.E. 1989. Clays, Mudrocks and Shales. Developments
Society Special Publication 158, 190 pp. in Sedimentology 44, 818 pp.
Bjørlykke, K. 2011. Open-system chemical behaviour of Wilcox Wood, L.J. 2010. Shale Tectonics. AAPG Memoir 93, 238 pp.
Group mudstones. How is large scale mass transfer at great Yang, Y. and Aplin, A.C. 2010. Permeability-porosity re-
burial depth in sedimentary basins possible? A discussion. lationships for mudstones. Marine and Petroleum Geology
Marine and Petroleum Geology 28, 1381–1382. 25, 1692–1696.
Chapter 7
Stratigraphy
Jenø Nagy and Knut Bjørlykke
7.1 Introduction and Concepts evolution throughout geological history. (3) Content
of radioactive fission products in minerals or rocks
Stratigraphy is the study of the succession and time- which may be used for age dating. (4) Magnetic
related architecture of rock strata. Stratification is not properties of rock strata. (5) Geochemical features of
limited to sedimentary deposit, but is also found in sediments. (6) Geophysical expression. These correla-
igneous rocks, particularly volcanic developments,
tion methods are so different that it has been found
and in certain plutonic rocks. All bedded rocks can be
useful to work with three forms of stratigraphy which
treated stratigraphically, i.e. to establish age relations
can be used in parallel:
between beds. However, the term “stratigraphy” is used
1. Lithostratigraphy: Classification of sedimentary
primarily in the context of sedimentary research.
rock types on the basis of their composition,
Stratigraphy involves the study, subdivision and
appearance and sedimentary structures.
documentation of sedimentary successions, and on
2. Biostratigraphy: Classification of sedimentary
this basis to interpret the geological history they rep-
rocks according to their fossil content.
resent. In order to reconstruct an environment or spe-
3. Chronostratigraphy: Classification of rocks on the
cial events in the geological past, regionally or even
basis of geological time. Geochronology is the
globally, it is necessary to characterise and correlate
actual subdivision of geological time.
sedimentary horizons from different areas. It is impor-
The first two are thus based on rock relationships
tant to establish which sedimentary units were depos-
which can be described, and are often collectively
ited at the same time or by the same or similar
referred to as rock stratigraphy.
sedimentological or biological processes, even if they
Despite radiometric dating, geological time is not
are not quite contemporary. Correlation is generally a
absolute. Even this dating method gives different ages,
question of what is possible to achieve with the avail-
able amount and quality of data. depending on which half-life is used for calculating
We usually have no possibility of determining defi- radioactive decay, and is encumbered with several
nitely which beds were deposited at the same time, but other uncertainty factors. Chronostratigraphy is there-
attempt to use all the information available in the fore a theoretical and abstract concept which describes
rocks. This information falls into five main categories: a time scale we cannot measure exactly. Rock stratig-
(1) Rock composition and structures resulting from raphy, on the other hand, takes the rocks themselves as
sedimentological processes. (2) Fossil content, which its starting point and is based on boundaries which are
is a result of biological, environmental and ecological identified as potentially suitable for correlation across
greater or lesser areas. For rules for stratigraphic
nomenclature, see stratigraphy.org. The different
J. Nagy (*) K. Bjørlykke types of stratigraphic units are listed below in hierar-
Department of Geosciences, University of Oslo, Oslo, Norway chic arrangement. See also Stratigraphy.com.
e-mail: [email protected]; [email protected]
232 J. Nagy and K. Bjørlykke
Lithostratigraphic Biostratigraphic Chronostratigraphic Geochronological

Supergroup Assemblage zone Eonothem Eon
Group Range zone Erathem Era
Subgroup Acme zone System Period
Formation Interval zone Series Epoch
Member Biozone Stage Age
Bed Chronozone Chron
7.2 Lithostratigraphy consists of two or more groups. This unit is used

when a common name is needed to encompass a
The formation is the fundamental lithostratigraphic thick sedimentary package.
unit defined in a bedded succession. An important In sedimentary basins which are partly or entirely
property of a formation is that it should be easily buried in the subsurface, for example the North Sea
recognised in the field or a borehole due to its compo- basin, lithostratigraphic units like formations and
sition (lithology). The formation is thus a mappable groups are defined on the basis of records from
unit, which can be visualised on an ordinary geologi- wells, such as well logs. In offshore areas where it is
cal map (scale e.g. 1:50,000) or recognised in the difficult to find enough geographical names to give the
description of a bedded succession. There is in princi- stratigraphic units, other names are then also used,
ple no limit to the thickness of a formation, but it including historical names, names of animals etc.
usually varies from a few tens to several hundred
metres. A 50–300 m thick sandstone, over- and
underlain by quite different rocks such as shale or 7.2.1 Lithostratigraphic Terminology
limestone, would constitute a natural formation. How-
ever, a formation will seldom be homogeneous, and As research into sedimentary successions expanded, it
for more detailed mapping it will be useful to divide required a standardised lithostratigraphic nomencla-
the formation into smaller units. Parts of a sandstone ture that could be applied anywhere. Therefore, a set
formation containing shale or conglomerate beds, for of international rules was established for naming
example, can be defined as members of the formation. lithostratigraphic units and are summarised below:
The smallest unit in the lithostratigraphic classifi- 1. A stratigraphic unit should preferably be defined
cation is a bed which is assumed to have been depos- with reference to a type section (stratotype) present
ited by a single depositional process without a break in in a good exposure, or in a well where the unit is
sedimentation, and is distinguishable from the beds adequately represented.
above and below. When sections are logged, beds 2. Each stratigraphic unit ought to be named after a
and other lithological entities are often recorded as geographical site, located preferably near the type
“units”. Units which are <1 cm thick are called section if exposed on land. Offshore stratigraphic
laminae, but this is a purely size-descriptive term. units may be named after marine features (e.g.
Most beds/units contain, or may even be entirely fishing banks), although in the North Sea basin
comprised of, laminae. Only when a lamina lies other names have also been used.
isolated between significantly different sediments 3. The same name ought not to be used for more than
will it be named, e.g. if it is a thin tephra (ash bed) one stratigraphic unit. The unit which is first
or bentonite in a shale succession. defined has priority.
For some purposes it is expedient to group several 4. Stratigraphic names may consist of a unique name, i.e.
formations together into a larger unit called a group. A from a locality, and a stratigraphic unit hierarchy
group normally consists of three to six formations and designation, e.g. Kimmeridge Formation or Kimme-
can be divided into two or more subgroups. The larg- ridge Clay. In the US many stratigraphic names are
est lithostratigraphic unit is a supergroup, which well established even if they do not strictly follow the
international rules for stratigraphic nomenclature.
7 Stratigraphy 233
Fig. 7.1 Application fields of micropalaeontological methods in petroleum exploration and appraisal activities (based on
Copestake 1993)
7.3 Biostratigraphy microscopic remains commonly occur in large

amounts, and can be extracted easily both from drill
Biostratigraphy is based on the fossils in sedimentary cuttings, sidewall cores as well as conventional cores.
rocks. Its tasks are to group strata into units The rapid evolution of many microfossil groups makes
characterized by fossil content, and apply these to them valuable tools in subsurface stratigraphic work
date and correlate sedimentary successions. Biostra- (Fig. 7.1). In addition to the general tasks of age
tigraphy has made it possible to correlate sedimentary determination and correlation, their applications
rocks on a global scale, and forms the foundation for include: unconformity identification, characterisation
global stratigraphic classification of sedimentary and correlation of seismic (depositional) sequences,
successions in combination with radiometric and fingerprinting of formations, reservoir zonation and
other age dating methods. The biostratigraphic appli- palaeoenvironmental modelling. These applications
cation of fossils is based on the fact that during the are crucial in petroleum exploration, contributing sig-
course of geological time, biological evolution took nificantly to assessments of reservoir distribution,
place whereby some species died out and new ones source rock evaluation, trap evaluation, estimation of
appeared. Any particular species will therefore be reserves, field development studies and impacts on
represented only in sediments deposited within a lim- drilling problems. The fast evaluation potential of
ited time span, and can be used to recognise this time microfossil samples is also an important factor in any
span or part of it. subsurface work.
7.3.1.2 Distribution of Microfossils

7.3.1 Nature of the Fossil Record Practically all rocks of sedimentary origin contain
microfossils, although their abundance, diversity and
7.3.1.1 Fields of Application state of preservation are highly variable. These
Industrial application of biostratigraphy is mainly features are strongly influenced by the age, deposi-
based on microfossils, owing to the fact that these tional environment, composition and diagenetic
Microfossils
Botryococcaceans
Coccolithophores
Prasynophyceans
Silicoflagellates
Dinoflagellates
Spores, pollen
Calpionellids
Chitinozoans
Foraminifera
Radiolarians
Conodonts
Acritarchs
Ostracods
Diatoms
Rocks
Clays and shales
Limestones and marls
Flints and cherts
Coal, lignite, peat
Sands and sandstones
Dolomites, ankerites Abundant

Common
Evaporites: gypsum, halite
Rare
Metamorphic rocks:
slates, phyllites, marbles Sporadic
Fig. 7.2 Distribution trends and importance of major microfossil groups in different types of rocks
history of the sediments. Marine mudstones, marls and (2) Arenaceous microfossils: agglutinated forami-
limestones usually have a particularly rich microfossil nifera, single-celled heterotrophs.
content of high diversity (Fig. 7.2). Well sorted (3) Siliceous microfossils: radiolarians, single-celled
sandstones have a generally low microfossil content, heterotrophs; diatoms, single-celled algae;
as most microfossils are both lighter and smaller than silicoflagellates, single-celled algae; ebridians,
the average sand grains. Dolomites are also poor in single-celled heterotrophs.
microfossils because most of the fossil organisms with (4) Phosphatic microfossils: conodonts, primitive
primary calcite have been dissolved during vertebrates.
dolomitisation, or because the beds represent an evap- (5) Organic-walled microfossils: acritarchs, uncertain
orite environment. Coals and related organic origin; prasinophytes, cysts of green algae;
sediments are rich in terrestrial microfossils (spores chitinozoa, uncertain origin; chlorophytes, fresh-
and pollen). Evaporites are characterised by small water algae, algal colonies; dinoflagellates, single-
quantities of microfossils, mainly of terrestrial origin. celled algae.
Microfossils occur in Precambrian rocks but rapid
development of many central groups started in the
Cambrian (Fig. 7.3). Every part of the stratigraphic
column from Cambrian to recent time contains one or
7.3.2 Factors Controlling Stratigraphic
several microfossil groups which are potentially useful
Application
for biostratigraphical or palaeoecological purposes.
According to skeletal composition, microfossils can
There are several factors influencing the applicability
be arranged into five groups, which comprise several
of microfossils to stratigraphic purposes. The most
taxonomic divisions:
important ones having a positive effect on the applica-
(1) Calcareous microfossils: calcitic and aragonitic
tion potential are: (1) Presence of hard parts with high
foraminifera, single-celled heterotrophs; ostracods,
preservation potential. (2) High evolution rate of
microscopic crustaceans; calpionellids, uncertain
taxa. (3) Extensive regional distribution, or high
origin; carophytes, algal reproductive organs; cal-
potential for dispersal. (4) High degree of facies
careous nannoplankton, single-celled algae.
independence. (5) Small redepositional potential.
7 Stratigraphy 235
Micro-fossils
Phosphatic
Ostracods
Organic-walled Siliceous Calcareous
Silicoflagellates
Dinoflagellates
nannoplankton
Calpionellids
Foraminifera
Tasmanitids
Calcareous
Chitinozoa
Conodonts
Acritarchs
Radiolaria
Ostracods
calc algae
Ebridians
Diatoms
Bryozoa
Skeletal
Eras Periods
(Extant) ?
Quaternary
Cenozoic
Tertiary
Cretaceous
Mesozoic
Jurassic
Triassic
Permian
Carboniferous
Palaeozoic
Devonian
Silurian
Ordovician
Cambrian
Precambrian
Fig. 7.3 Outline of the stratigraphic distribution of microfossil groups from Precambrian to present time, arranged according to
skeletal composition (modified after Haq and Boersma 1978)
(6) Presence of many distinct morphological features, sediment/water interface), and endobenthos (living
increasing the precision level of taxon identification. buried in the sediment). Epibenthic forms are freely
moving or are attached to firm surfaces. Foraminifera
7.3.2.1 Mode of Life as a group reveal the widest environmental range by
The mode of life (ethology) of organisms is crucial for being represented in all the ethological categories
the regional distribution of their fossil remains. Living except nektonic. Ostracods and diatoms also have a
and fossil biota can be arranged into two ethological relatively wide range as they occur both in the benthic
groups (Fig. 7.4). Pelagic organisms live freely in the and pelagic habitats. Conodonts are regarded to have a
water column and comprise the planktonic forms (pas- nektonic origin. Acritarchs, dinoflagellates, calcareous
sively floating) and nektonic forms (actively swim- nannoplankton, radiolarians and several minor groups
ming). The benthos comprises bottom dwelling are exclusively planktonic. Spores and pollen occupy a
organisms belonging to epibenthos (living at the special position because they are the reproductive
Microfossils
Coccolithophores
Prasynophyceans
Silicoflagellates
Botryococcace.
Dinoflagellates
Calpionellids
Chitinozoans
Foraminifera
Radiolarians
Conodonts
Acritarchs
Ostracods
Mode of life
Diatoms
Spores
Pollen
and habitats
Pelagos Planktonic
Living in the Floating
water above
Nektonic Not plankton,
dispersed by
the sea bed Swimming air and water
Epi- Free
Benthos benthos
Living on the On the
sediment- sediment Attached
water contact surface
or below Endobenthos
Buried in the
sediment
Fig. 7.4 Mode of life and habitats of major microfossil groups in aquatic environments
organs of terrestrial plants, but are dispersed princi- The members of the ubiquitous group of organic-
pally as plankton (by air or water). walled microfossils are produced in a wide range of
It is obvious that the pelagic group contains contrasting habitats (Fig. 7.4). Pollen and spores are
microfossils with particularly high stratigraphic appli- produced by land and freshwater flora. They are
cability, owing to their relative independence of transported far out into the marine realm and show
facies, e.g. planktonic foraminifera, calcareous nanno- decreasing frequency with increasing distance from
plankton, dinoflagellates and the nektonic conodonts. source. They are commonly used in the stratigraphy
Benthic biota usually have less stratigraphic potential, of terrestrial and marginal marine deposits in the
although they are also commonly used for this pur- absence of other fossils, although they provide a low
pose, particularly in marginal marine deposits where age resolution. Chlorophyte freshwater algae (as
planktonic biota are rare. On the other hand, benthic Botryococcus) are also prone to be transported to
forms are commonly excellent facies indicators, e.g. marine areas. Prasionophyte marine algae and
benthic foraminifera and ostracods. acritarchs are common in brackish marginal marine
and shallow shelf settings. Dinoflagellates are most
7.3.2.2 Environmental Distribution common in marine shelf and oceanic waters, and are
The life habitat of biota is of prime importance both widely used for stratigraphical purposes.
for stratigraphical purposes and biofacies analysis. In Foraminifera occur in a wide variety of
non-marine subaquatic environments important fossil environments ranging from open ocean to estuarine
groups include diatoms and ostracods (Fig. 7.4). In waters, and are extensively used in stratigraphic and
marginal marine brackish environments low diversity facies analyses. Marginal marine environments (such
assemblages of foraminifera, ostracods, and diatoms as tidal marshes, estuaries and lagoons) are typified by
are dominant. Shallow shelf to bathyal environments low diversity benthic assemblages composed of
are typified by their high diversity assemblages of agglutinated and mixed calcareous-agglutinated
benthic and pelagic biota. In abyssal environments, components. The low species diversity restricted
below the compensation depth of calcium carbonate, nature of these assemblages reduces their stratigraphi-
microfossils are exclusively siliceous, arenaceous or cal applicability. Normal marine shelf areas are
organic-walled. typified by high diversity calcareous faunas and
7 Stratigraphy 237
Microfossil groups
Spores & pollen

Palaeoenvironment
reservoir
Dinoflagellate
Agglutinating
foraminifera
nannofossils
foraminifera
foraminifera
Radiolaria
Planktonic
source
Calcareous
Main
types
Ostracoda
Calcareous
rocks
Main
benthonic
Diatoms
cysts
Water
depth
(metres)
“Red beds”
Non marine
Terrestrial
sands
Lacustrine/
fluvial
0
Marginal marine
coastal sands
Coals & organic

Deltaic &
(gas prone)
rich shales
Deltaic/
Stromatolites (oil prone)

estuarine
Inner
Lagoonal
30
Shelf
Middle
100
bar sands
& reefs
Barrier
Outer
200
Marine
Upper
Submarine fan sands, limestones

600
Bathyal
Basinal organic-rich
Middle
shales (oil prone)

1000
Lower
2000
Abyssal
4000
Fig. 7.5 Depth-related palaeoenvironmental distribution of major microfossil groups in the North Sea basin (Triassic to Tertiary),
and position of the main petroleum reservoir and source rock facies (from Copestake 1993)
increasing frequency of planktonic species toward the and Cenozoic succession is well known owing to the
shelf break. prevailing oil and gas exploration (Fig. 7.5). The
The foraminiferal planktonic component reaches its microfossil distribution in this area reflects the general
maximum in bathyal areas. These shelf and bathyal bathymetric distribution pattern of the various groups.
assemblages form an excellent basis for age determi- Abyssal deposits below the carbonate compensation
nation and stratigraphic correlation. The amount of depth are dominated by radiolarians and cosmopolitan
calcareous benthic and planktonic foraminifera agglutinated foraminifera. Bathyal and outer shelf
decreases downslope through the lysocline and they sediments are characterised by high dominance of
disappear below the compensation depth of calcium planktonic foraminifera and calcareous nannoplank-
carbonate. Below this level the sea bed is populated by ton. Mid-shelf sediments contain abundant calcareous
agglutinated species, usually of a cosmopolitan nature, benthic foraminifera and dinoflagellates. Deltaic
which provide an adequate basis for stratigraphic and deposits are typified by spores, polen and agglutinated
environmental application. foraminifera. In non-marine deposits the most com-
The environmental distribution of the various mon fossils are spores and pollen.
microfossil groups in the North Sea basin Mesozoic
7.3.2.3 Redeposition biozones (local ranges) and integration with

Erosion of fossil-bearing sediments with subsequent magnetostratigraphy and chemostratigraphy is a use-
transport and redeposition of microfossils can lead to ful approach to this type of correlation.
serious problems for dating and correlation.
Palynomorphs are particularly liable to redeposition
owing to their small size and weight. Conodonts also 7.3.3 Biozones
have a high redeposition potential because their cal-
cium phosphate skeletons are strongly resistant to The basic unit of biostratigraphy is the biozone which
corrosion by weathering and transport. Redeposition is a body of strata defined by its fossil content. During
is suspected if microfossils appear at higher strati- the history of biostratigraphy, a variety of names and
graphic levels than expected. Reworked microfossils definitions of units have been proposed, but there are
can be distinguished by changes in colour and traces of only three basic types of zones: interval zones, assem-
wear. blage zones and abundance zones. Zones defined in
sedimentary successions are usually arranged into a
7.3.2.4 Facies Barriers and Biostratigraphy zonal scheme, which can be developed for an exten-
Sediment cores from deep oceanic settings usually sive region and even for global application.
provide undisturbed sections containing a continuous
fossil record, commonly of high stratigraphic resolu- 7.3.3.1 Interval Zones
tion. Based on this type of material, biostratigraphy These are the most commonly used type of zones, with
combined with magnetostratigraphy and chemostra- boundaries defined by particular fossil events, specifi-
tigraphy contributed heavily to development of a cally the first and last occurrence of taxa (usually
global biochronozonal framework. Although with species). Thus, the interval zone comprises the strata
marked reductions, this framework has been extended deposited between these two events. There are five
to formations deposited in marine shelf, deltaic and types of interval zones (Fig. 7.6).
coastal areas, where it forms an important basis for 1. Taxon range zone is defined by the first and last
dating and correlation. occurrence of a particular taxon.
The basic unit of biostratigraphic correlation is the 2. Concurrent range zone is the interval between the
taxon chronozone, also called biochronozone. As defined first occurrence of a taxon and the last occurrence
below, a taxon chronozone represents all sediments of another taxon.
deposited between the evolutionary inception and 3. Partial range zone is the interval between the last
extinction (total range) of its marker species (Fig. 7.8). occurrence of a taxon and the first occurrence of
In contrast, the biozone represents only those sediments another taxon partitioning the range of a third
at a given place that contain one or several marker taxon.
species. In the case of a single marker, the biozone can 4. Successive first appearance zone represents the
be defined by the local appearance and disappearance interval between the successive first appearance of
(local range) of the species. The time interval represented two taxa.
by the biozone may differ from place to place. 5. Successive last appearance zone is the interval
In deep oceanic regimes, the latitudinal distribution between the successive last appearance of two taxa.
of each species is generally defined by the water mass
temperature and chemistry. Therefore, the regional The successive last appearance zone is the type
(lateral) extent of biozones and taxon chronozones is most commonly used in industrial biostratigraphy,
of restricted nature. Outside its optimal stratigraphic because the samples are mainly derived from well
distribution area, the range of the marker species can drilling cuttings. In such sample sets, the actual first
vary considerably owing to environmental factors. At appearances are unreliable owing to downhole
such sites the chronozone cannot be defined and the (caving) contamination. Precise positions of first
term biozone is used. Correlation between tropical and appearances are required in order to recognise the
temperate oceanic regions is difficult, because of the first four interval zone types. Taxon range zone and
low number of biota that include reliable zonal indica- successive first appearance zone are the most applied
tor species common to both areas. Interfingering of in zonation of onshore sections.
7 Stratigraphy 239
Time b Assemblage biozones

Taxon
1 X
range zone
Time a Assemblage zone 3
Assemblage zone 2
2 Y X Concurrent
range zone Assemblage zone 1
Partial Abundance biozones

3 Z
range zone
Acme zone 2
X
X
Acme zone 1
Successive
4 Y first appearance
zone
Fig. 7.7 Outline of the biostratigraphic assemblage and abun-

dance zones
Successive
5 Y last appearance restricted taxon range zones. One or several of such
zone
local taxon range zones can be included in the total
X range of the taxon.
Fig. 7.6 Outline of the five types of biostratigraphic interval

zones shown by occurrences of microfossils X, Y, and Z. Verti-
cal lines are taxon ranges, horizontal bars are first and last
occurrences 7.3.4 Biostratigraphic Correlation
7.3.3.2 Other Biozones Biostratigraphic correlation is based on the fact that

during the course of geological time, biological evolu-
tion has taken place, whereby some species and higher
1. Assemblage zone is a biozone characterised by the taxa have died out and new ones have appeared
association of three or more taxa (Fig. 7.7). This (Fig. 7.3). Any particular species will therefore be
zone is recognised without regard to range limits of represented only in sediments deposited within a lim-
taxa, and is usually of regional importance. ited time span. However, this method also has many
2. Acme zone is based on the numerical maximum of limitations. Some fossil groups developed rapidly,
one or several taxa, and is also termed abundance with individual species only existing for a short
zone. Taxon maxima are strongly influenced by period. Other groups evolved slowly and persisted
local environmental conditions, and therefore this throughout long periods of geological time. Some
zone is usually of local stratigraphic importance. fossil organisms are found only in rocks deposited in
3. Taxon chronozone represents the interval between particular environments or facies. Many were depen-
two unique bioevents: the first evolutionary incep- dant on a certain salinity, oxygen content and water
tion of a taxon, and its extinction. Thus, this zone depth. Bottom conditions played a major role for epi-
comprises the total range of the taxon, and is of faunal (attached) and infaunal (burrowing) animals.
global nature (Fig. 7.8). The distribution of taxa, Biozones of sedimentary successions are arranged
however, is to a varying degree facies-controlled, into zonal schemes, which are the main tools of fossil-
which leads to the development of geographically based correlation. The ideal index fossil for defining a
Shelf
Time
Open ocean
ma
Inner Middle Outer
Extinction
Local range zone
Total range
Disappearance
Biozone boundary Taxon

chronozone
Appearance
Evolutionary
appearance
Fig. 7.8 Facies-related distribution of an imaginary fossil taxon exemplifying the difference between a taxon chronozone (global)
and a biozone (local) taxon range zone
biozone ought to belong to a fossil group which to put it on a section with optimal potential for corre-
underwent rapid biological evolution and had a global lation with other areas and the global time scale
distribution. It should not have had special environmen- (Fig. 7.9).
tal requirements, and should have been cosmopolitan Geologists used to define a stratigraphic boundary
with respect to habitat. One such fossil group is the at a break (hiatus) in the sedimentary record. However,
graptolites, which are common in Cambro-Silurian sediments later found elsewhere that had been depos-
shales. They were planktonic and drifted in the surface ited in the period represented by the hiatus, could not
seawater. However, even this type of plankton may be assigned to either unit. For example, the boundary
show a distribution pattern which is influenced by between the Tertiary and Cretaceous was defined in
ocean currents and temperatures etc. At the present, England where a hiatus is developed between these
microfossils are being used more and more for biostrati- systems. As a result it was difficult to reach agreement
graphic correlation particularly by the petroleum indus- on whether sediments which were deposited for exam-
try. Several animal and plant groups (e.g. conodonts, ple in Denmark during this period (Danian), should
planktonic foraminifera, radiolarians and spores and belong to the Cretaceous or the Tertiary.
pollen) have wide distributions and have the advantage
that even small samples, e.g. from boreholes, provide
adequate material for statistical treatment. 7.4 Time Stratigraphy
Although fossils are always helpful in stratigraphic
correlations, we now avoid defining important geolog- Chronostratigraphy is an attempt to correlate rocks
ical boundaries, for example that between the Ordovi- deposited at the same time, across larger areas. The
cian and the Silurian, by means of fossil occurrences. accuracy achievable with chronostratigraphic correla-
Such a boundary would have to be moved whenever tion depends on whether the sediments contain evi-
one found a new occurrence of a certain fossil. The dence of well-defined geological events which were
geological periods are at the present defined by inter- simultaneous across the region. These events may be
national committees which select a type section with biological (e.g. appearance of particular species), sed-
continuous sedimentation and preferably fossiliferous imentological (e.g. deposition of ash layers) or geo-
record, and the boundary is physically marked as a physical (e.g. reversals of the Earth’s magnetic field).
fixed point in the section. The boundary is then unam- Stratigraphical research has a long tradition of
biguously defined. All available means can then be correlating synchronous events in geological history.
used, including fossils, to correlate the boundary Such correlations are independent of an absolute time
with other areas. This is the principle of the arbitrary scale. Only after the development of radiometric dating
boundary. In reality it is not entirely arbitrary. We try methods did it become possible to set up a series of
7 Stratigraphy 241
GEOLOGIC TIME SCALE

PHANEROZOIC PRECAMBRIAN
CENOZOIC MESOZOIC PALEOZOIC
Quat. Period
Period
Period
Eon
Age Epoch Age/Stage Age Age Epoch Age/Stage Age Age Epoch Age/Stage Age Age Era Period Age
0 Holocene Lopin- Changhsingian
Mesoproterozoic Neoproterozoic
0.78 254.2 Ediacaran
Pleisto- Calabrian 70
Maastrichtian gian Wuchiapingian 600
635
1.81 259.8
cene Gelasian 2.59
72.1
Capitanian 700
Guada- 265.1
Plio- Piacenzian Wordian Cryogenian
Permian
3.60
Campanian lupian Roadian 268.8 800
5 cene Zanclean 80 275
272.3 850
5.33 Kungurian 900
Messinian Late 83.6 279.3 Tonian
Santonian
7.25 86.3 Cis- Artinskian 1000 1000
Coniacian uralian
90 89.8 290.1 Stenian
Neogene
Turonian 1100
Tortonian Sakmarian
10 93.9 295.5
Asselian 298.9 1200 1200
Cenomanian 300
Pennsylvanian
L Gzhelian
Cretaceous
11.63 303.7
100 100.5 Kasimovian 1300 Ectasian
Miocene
Serravallian 307.0
Proterozoic
M Moscovian 1400 1400
13.82
Carboniferous
Langhian Albian 315.2
15 E Bashkirian 1500 Calymmian
15.97 110
323.2
113.0 325 1600
L Serpukhovian 1600
Mississippian
Burdigalian 330.9
1700 Statherian
120 Aptian
M Visean
Paleoproterozoic
20 20.4 Early 1800 1800
126.3 346.7
Aquitanian 1900
Orosirian
130
Barremian 350
23.0 130.8 E Tournaisian 2000
Hauterivian 2050
133.9
358.9 2100
25 Valanginian
Chattian 139.4 Famennian Rhyacian
140
Oligocene
2200
Berriasian Late 372.2
28.1 145.0 375 2300 2300
Frasnian
Devonian
150
Tithonian 2400 Siderian
30 382.7
152.1 Givetian
Rupelian Late Middle 387.7 2500 2500
Kimmeridgian Eifelian
archean
157.3 393.3
Neo-
2600
160 Oxfordian
33.9 400 Emsian 2700
163.5
35 Callovian 166.1 Early 407.6
Priabonian Bathonian Pragian 2800 2800
Jurassic
Middle Bajocian 168.3 410.8

170 170.3 Lochkovian
archean
Aalenian 2900
Meso-
37.8
174.1 419.2
Pridoli 423.0 3000
Bartonian
Silurian
425 Ludlow Ludfordian

Paleogene
40 Toarcian Gorstian 425.6

180 Homerian 427.4
Wenlock Sheinwoodian 430.5 3100
Archean
41.2 182.7 433.4

Pliens- Llando- Telychian 3200 3200
Aeronian 438.5
Early bachian very 440.8
Eocene
Rhuddanian
archean
190 443.8 3300

Paleo-
Hirnantian 445.2
Lutetian 190.8
45 450 Katian
Late 3400
Sinemurian 453.0
Ordovician
Sandbian 3500
200 199.3 458.4
47.8 Hettangian
201.3 Middle Darriwilian 3600 3600
467.3
Eoarchean
Rhaetian Dapingian
50 470.0 3700
210 209.5 475 Floian
Ypresian Early 477.7 3800
Tremadocian
485.4 3900
220 Late Norian Furon- Age 10 489.5
55 gian Jiangshanian 494.0
4000 4000
Triassic
56.0 Paibian
497.0
500 Guzhangian 500.5 4100
Epoch Drumian 504.5
Thanetian
Cambrian
228.4
230 3 Age 5 509.0 4200 Hadean
Paleocene
59.2 Carnian Age 4 (informal)

60 Epoch 514.0 4300
Selandian 237.0 2 Age 3
61.6 240 Ladinian 521.0 4400
Middle 241.5 525 Age 2
Anisian 4500
Danian Terre- 529.0
Olenekian
247.1 neuvian Fortunian
65 250.0
250 Early
66.0 Induan 252.2 541.0
This chart was drafted by Gabi Ogg. Copyright © 2012 Geologic Time Scale Foundation
Fig. 7.9 Geological time scale for the Phanerozioc Aeon (Geological Time Scale Foundation 2012)
datings approaching an absolute time scale, but radio- time period. This may make it possible to map
metric datings are not absolute with respect to time. variations in sedimentation rate and the ratio of sand-
We thus distinguish between two types of time stone to shale. Maps showing the sediment thickness
stratigraphy: (1) Geochronology, which is the subdivi- between two marker beds are called isopach maps.
sion of the Earth’s history into finite time units. A This is very useful for reconstructing sedimentary
geochronological unit is a specific interval of geologi- facies on the basis of borehole data, a standard method
cal time. (2) Chronostratigraphy or time-rock stratig- in connection with oil prospecting.
raphy is the subdivision of sedimentary successions, The best chronostratigraphic horizons (marker
and their correlation, on the basis of time. Chronostra- beds) are bentonite (ash) layers; on a smaller regional
tigraphic units are by definition synchronous. scale also coal beds, phosphate beds, thin limestone or
A geochronological unit represents a specific inter- sandstone beds, or particular fossil horizons. When
val in the geological time scale. For example, the analysing stratigraphic records from wells (logs) one
Jurassic period is a geochronological unit. A tends to use beds which produce a distinctive log
geochronological unit may define the time span pattern. Seismic reflectors can also in certain instances
between two specific geological events. We can say, be used as datum beds (horizons).
for example, that some of the rocks in the North Sea A stage is a chronostratigraphic unit which includes
were deposited during the Jurassic period. The one or more chronozones but which nevertheless
corresponding chronostratigraphic unit (system) covers a limited period of time, usually 3–10 million
signifies the rocks which were formed during the years (Fig. 7.9). This is the smallest unit in the
same period. We therefore say that some of the rocks chronostratigraphic hierarchy which is used for corre-
in the North Sea belong to the Jurassic system. lation all over the world. A stage is defined in a type
The basic chronostratigraphic unit is a chronozone. section and usually designated by a geographical name
A chronozone includes all the deposits formed during near the type profile. For example, the Kimmeridgian
a particular, relatively short, geologic time interval stage is well exposed on the Dorset coast at
and which are defined by a geological phenomenon Kimmeridge. The correlation of a stage is usually
or by a particular interval of a rock succession. In most based on biostratigraphy. An age is the period of
cases a chronozone is a taxon chronozone, which is time (geochronological unit) which corresponds to a
defined as the period between the first appearance and stage.
last occurrence of a particular fossil taxon (Fig. 7.8). A series is a chronostratigraphic unit larger than a
Whereas a biozone can only be defined where the stage. For example, the Late Jurassic is a series
fossil is present, a chronozone represents all the constituting part of the Jurassic system. The
rocks that were formed during this period, regardless geochronological unit which corresponds to a series
of whether they contain fossils. A chron is the interval is an epoch. We can say that a certain limestone was
of time during which the rock in a chronozone was deposited during the Late Jurassic epoch.
formed. It is thus the geochronological equivalent of a A geochronological period varies in duration from
chronozone. about 20–30 million years (Silurian) to about 60–70
A chronozone may be named after a biostrati- million years (Cretaceous). The Quaternary period,
graphic unit, e.g. a Didymograptus extensus however, is much shorter, only about 2.5 million
chronozone, or after a lithostratigraphic unit, which years. The rocks formed during a period constitute a
can be recognised over large areas. Such characteristic system. An era is comprised of two or more periods.
strata are often called marker beds, key beds, or datum The Palaeozoic era had a duration of about 300 million
beds. Chronostratigraphic horizons are important for years, but the Cenozoic era did not last longer than the
correlation within sedimentary basins, and form the longest Palaeozoic periods (65 million years).
framework for all facies reconstructions. If we have The largest units in the chronostratigraphic scale,
two chronohorizons or datum beds, we can measure erathem and enothem, are not used much, since it is
the variation in thickness and composition of the seldom relevant to group rocks which were deposited
sediments which were deposited during a particular over such long periods of time. However, when we
7 Stratigraphy 243
discuss geological history in relation to the biological deltas will tend to shift over long periods of time,
development on the Earth, for example, it can be useful and the sandstones deposited will therefore have an
to speak about the Phanerozoic aeon which covers the upper and a lower boundary which are not parallel to a
Palaeozoic, Mesozoic and Cenozoic eras. The Precam- theoretical time plane, i.e. they are time-transgressive.
brian is divided into two aeons: the Proterozoic In practical geological work (mapping, well and
(542–2,500 million years ago) and the Archean borehole studies, etc.) one is primarily interested in
(2,500–4,000 million years ago). correlating rock types, in other words lithostra-
tigraphy. To construct a facies map that gives an
overview of the distribution of sediments at a particu-
7.5 The Relation Between lar time, it is necessary to establish a time correlation.
Lithostratigraphy, Biostratigraphy Good index fossils or marker beds will provide an
and Chronostratigraphy important framework for portraying depositional
conditions, for example during regressive or transgres-
These three types of stratigraphy are based on different sive phases. If we study a modern coastal area, we find
criteria, and geological experience has shown that it is different fauna in different environments. We can
useful to maintain this tripartite division. Stratigraphi- easily see that the distribution of biotas within a lim-
cal boundaries defined in these three ways may some- ited geological time period is not controlled by strati-
times nearly coincide, but usually show considerable graphic time, but by facies. Many animal groups
divergences particularly over larger distances. This provide good indices for sedimentary facies
can be demonstrated schematically by the occurrence (Figs. 7.4 and 7.5).
of a sandstone formation between two shale units
(Fig. 7.10). Lithostratigraphically, this sandstone for-
mation is unambiguously defined by the boundaries
between shale and sandstone. However, if we find
7.5.1 Correlation
good index fossils, X and Y, in both the shale and
Rocks from the Phanerozoic Aeon were originally
the sandstone, it may turn out that the top of the
classified on the basis of fossils. Later, committees
sandstone in area B corresponds chronostratigra-
for each period have been set up, responsible for
phically to the bottom of the sandstone in area A.
selecting type sections where the lower and upper
Sedimentation of sand has thus moved from B to A
boundaries are physically defined by a bolt (“golden
during the course of a measurable period of geological
spike”). Correlation with other areas can then be car-
time. Sandstone sedimentation along coasts and on
ried out with the aid of fossils or other age indicators,
anywhere in the world.
Although biostratigraphic correlation is still the
most important method, magnetostratigraphy and
radiometric dating methods are gaining importance.
Within the confines of a sedimentary basin, however,
certain types of lithostratigraphic correlation can turn
out to be the most accurate. The various types of well
logs provide good opportunities for lithostratigraphic
correlation. Seismic profiles have perhaps to an even
greater degree made it possible to correlate lithostra-
tigraphy over great distances as a basis for sequence
stratigraphy (see next chapter).
It was realised relatively early that the lateral
Fig. 7.10 Occurrence of a time-trasgressive sandstone forma-
tion between shale units. X and Y are good index fossils which changes in facies corresponded to the vertical changes
are little affected by facies. Consequently, they indicate that the in sedimentary basins if there are no breaks in sedi-
sandstone is younger in area A than in B, and has been deposited mentation (¼ unconformity). Walthers law states that,
by progradation from B to A. This could be a shallow marine “In a conformable succession the only facies that can
sandstone or a sandstone deposited by a prograding delta.
occur together in vertical successions are those that material we measure the ratio between the fissionable
can occur side by side in nature”. nuclide (parent nuclide) and the daughter nuclide.
7.6 Radiometric Dating Methods 7.6.2 Potassium-Argon Method

7.6.1 Concept of Dating One of the most commonly used dating methods is the
potassium-argon method. 40K is an unstable nuclide
The geological time scale is mainly founded on age
which decays mainly to 40Ca with the emission of beta
determinations based on radioactive processes. There
particles, but about 12% is transformed into argon-40
are a number of other indirect methods of measuring
through capture of electrons and emission of X-rays
geological time, but only radiometric datings give
(gamma-rays). Any mineral which contains potas-
satisfactory quantitative results. The measurements
sium, for example biotite, will also contain some 40K
are based on the fact that radioactive nuclides undergo
which decays to 40Ar. The amount of argon in the
decay at a certain rate. The rate of fission can be
mineral is thus an expression of the mineral’s age,
determined with some accuracy, and there is no reason
which can be calculated if we know the half-life (T1/2)
to believe that it has varied through geological time,
or the decay rate (λ) (half-life is 1.31 109 years).
even though this is difficult to prove. The rate at which
The amount of argon gas in the minerals is analysed
radioactive decay proceeds is proportional to the
with a sensitive mass spectrometer. Argon is a very
assumed number of fissionable atoms present. The
volatile gas, however, and when minerals containing
number of atoms remaining which can undergo fission
argon are heated up or deformed, for example during
decreases according to the following formula,
metamorphism or folding, the argon will escape and the
however:
radioactive “clock” will indicate an age which
corresponds more to the time of metamorphism than
dN=dt ¼ λN which is the rate of decay at to the formation of the mineral. Therefore, a radioactive
any time t: age determination does not necessarily give a figure for
N ¼ No eλt the age of the rock, but will tell us something about the
geological processes which have affected the rock and
where thus help us to reconstruct its geological history.
No ¼ number of atoms at time To
N ¼ number of atoms at time T
λ ¼ decay (disintegration) constant, which is the frac- 7.6.3 Rubidium-Strontium Method
tion of the total number of atoms which will decay
in a given time. Another commonly used dating method is the
rubidium-strontium rato, 87Rb to 87Sr. Rubidium has
A convenient measure of the rate of fission is the a radioactive isotope, 87Rb, which decays to 87Sr with
half-life (T1/2) i.e. the time it takes for half of the a half life of 4.89 1010 years. The amount of 87Sr is
original atoms to decay. That is to say then a function of the time since the formation of the
mineral or rock. Instead of measuring absolute
λT
N=No ¼ 1=2 ¼ e1=2 amounts of 87Sr, it is simpler to measure the
87
Sr=86 Sr ratio on the mass spectrometer. Then the
The half-life is then: 87
Rb=86 Sr and 87 Sr=86 Sr ratios are plotted as axes on
a graph. Analyses of minerals or rocks which have the
T1=2 ¼ ln 2=λ ¼ 0:693=λ same age will then form points on a straight line
(isochron). The slope of this line will be an expression
Radioactive decay processes result in new nuclides of the age of the rock.
called daughter nuclides. In order to date geological
7 Stratigraphy 245
14
Using this method we can analyse both minerals C from the atmosphere also becomes mixed with
and whole rocks, since we assume that the 87Sr formed sea water and freshwater. Carbonate-secreting
does not escape from the rock as easily as argon (in the organisms which live in the sea and in lakes will take
K-Ar method). When dating sediments one should up CO2 from the water, and the amount of 14C in their
choose the finest-grained clay sediments with low CaCO3 skeletons can be measured. Molluscan and
permeability. It can then be assumed that after deposi- foraminiferal shells are particularly suitable for age
tion the sediment was homogeneous with respect to dating. Chemically precipitated carbonates can also
strontium isotopes, and will provide an isochron which be dated in this way. Living organisms have a ten-
dates any diagenesis occurring a relatively short time dency to fractionate the lightest isotopes from the
after deposition. Larger clastic fragments, however, heaviest. The ratio between two stable carbon
will contain a strontium isotope ratio which isotopes, 12C and 13C, can therefore be used to correct
14
corresponds to the age of the source rock, and the C determinations.
obtained date will be intermediate between the age of
the source rock and the time of deposition of the
sediments. 7.6.5 Other Radiometric Methods
There are numerous other radiometric dating methods,

7.6.4 Carbon 14 Method of which the uranium-lead method and the lead-lead
method (relationship between different lead isotopes)
The carbon 14 method is the one most commonly used are the most important.
for dating the youngest sediments, from about 50,000 The methods based on fission processes mentioned
years old up to the present. 14C is formed in the above, involve long half-lives, 109–1010 years, so for
atmosphere by cosmic rays when a 14N atom absorbs young sediments the quantity of decay products avail-
a neutron and gives off a proton. 14C is unstable, and able to be analysed will be small and the accuracy
decays to 14N. The production of 14C thus occurs only poor. An exception is the protactinium method
in the atmosphere, at altitudes over 10,000 m. The 14C

231
Pa=230 Th , which has given age determinations
then mixes with the lower air layers and the seawater.
The 14C method is based on the assumption on younger sediments showing good agreement with
(discussed below) that the 14C content of the atmo- the 14C method.
sphere has been constant for a long time, due to an The fission-track method is especially well suited
equilibrium between the 14C added from the atmo- for younger rocks. It is applied to glass or minerals
sphere and the 14C which decays to 14N. The half-life which contain a sufficient amount of uranium 238. The
of 14C is 5,730 years. This means that after 5,730 years material will show tracks from fission products which
half of the 14C atoms will have changed to 14N. 14C are observed as deformations. By counting the number
enters the carbon dioxide (CO2) in the air and is taken of such tracks, an expression of the age is obtained. In
up by plants through photosynthesis. If we assume that apatite the tracks disappear (aneal) at about 100 C,
the carbon dioxide in the air in the past had as much which is called the blocking temperature. The fre-
14
C as now, the 14C content of older plants or plant quency of tracks is therefore an expression of the
remains is an expression of their age. This age can be time elapsed since rocks have been uplifted and cooled
determined analytically with relatively great accuracy, to below the blocking temperature.
for example 10,200 100 years, depending on the
nature of the sample. When 50–60,000 years has
elapsed since plant material formed (i.e. since it 7.7 Chemostratigraphy
ceased to take up CO2), the 14C content will be so
small that we will be approaching the limit of detec- 7.7.1 The 87Sr/ 86Sr Method
tion. This is then the upper limit to the age of material
we can analyse. We now know that the concentration The 87 Sr=86 Sr ratio is particularly useful for
of 14C has varied in the last 10,000 years and more due carbonates and other minerals precipitated in the
to variation in cosmic radiation. ocean. In the ocean water the 87 Sr=86 Sr ratio has varied
greatly over geologic time (Fig. 3.8). By analysing this the orientation of this remanent magnetism can be
ratio in minerals precipitated in marine environments measured. The data will reveal the position of the
the ages can be constrained. During Tertiary time, the geographic pole, and indicate the palaeogeographical
87
Sr=86 Sr ratio increased and analyses of calcareous latitude and longitude.
organisms will give a unique age, in some cases with a Palaeomagnetic measurements are of great help in
resolution of 1–2 million years. This method can be reconstructing the positions of the continents during
applied also to calcareous microfossils such as forami- the geological past, and are important for plate tec-
nifera and coccoliths. In carbonate the Sr/Ca ratio may tonic and palaeoclimatic reconstructions. From the
be useful also in detecting depositional facies, because Palaeozoic onwards there is quite good agreement
the data will reflect the primary aragonite/calcite ratio. between palaeomagnetic determinations of latitude
and palaeoclimatic indications like biogeography, gla-
cial deposits and evaporites, but this is not the case
7.7.2 Bulk Chemical Composition Analyses with Precambrian deposits, which are more difficult.
Chemical analyses of bulk samples of cuttings and

cores may be useful in correlation between wells, 7.8.2 Magnetic Field Polarity Changes
because the chemical composition reflects changes in
the composition of sediments supplied to the basin. It has become evident that the polarity of the Earth’s
Particularly good results are expected in correlation of magnetic field has been reversed for numerous time
reservoirs. For this purpose major elements, trace periods of varying duration. A number of
elements and isotopic composition can be used. The measurements of magnetism in rocks of known age
samples can be analysed by XRF, by energy dispersive have provided us with a time scale based on periods of
systems in a SEM, or in mass-spectrographs. Analyses normal and reversed magnetic field. We thus find that
can be performed for bulk chemical composition. we can divide geological time into periods with domi-
The bulk chemical composition varies as a function nantly normal or reversed polarity (Fig. 7.11). Within
of grain size but the ratio between major elements such these we also find several shorter intervals when the
as Na/Al and K/Al may be characteristic for the prov- magnetic field switched between the two polarities.
enance rocks of sediments. Both sandstones and shales Since we must assume that the switching between
may have very different ratios between Na-feldspar normal and reversed magnetism has taken place simul-
and K-feldspar. Trace elements may be useful also in taneously and suddenly all over the world, such physi-
provenance studies. cal changes offer an ideal basis for correlation.
There are often major practical problems, however,
since many periods of the Earth’s history are
7.8 Magnetostratigraphy characterised by a predominance of successions with
rapidly changing reversals. Where we have many
7.8.1 Palaeomagnetism measurements and continuous profiles, e.g. in deep-
sea cores, we will be able to correlate with relative
When clastic sediments are deposited or when volca- certainty on the basis of the longer periods of magnetic
nic rocks solidify, minerals orientate themselves in the field stability. Volcanic rocks and sediments deposited
prevailing magnetic field. Magnetic minerals in clastic in fluvial or shallow-water environments, however,
sediments can orient themselves according to the mag- will have numerous hiatuses between beds, hampering
netic field during deposition, while diagenetic registration of continuous variations in the residual
minerals will become oriented in the magnetic field magnetism. During the last 700,000 years we have
during diagenesis. The magnetic minerals in rocks had apparently normal polarity, possibly with the
thus define a magnetic vector, which indicates the exception of a short period about 200,000–300,000
direction and strength of the magnetic field during years ago. If we find sediments or volcanic rocks
formation. By compensating for later magnetic effects,
7 Stratigraphy 247
with reversed magnetism, we know that they are
Period
Series
Geomagnetic
Stage
polarity highly probably more than 700,000 years old.
C30
Maastrichtian
C31
C-sequence
C32
Campanian C33
7.9 Sequence Stratigraphy
Late
Santonian
Coniacian
7.9.1 General Aspects
Turonian
C34
Sequence stratigraphy is defined as the subdivision of
Cretaceous normal super-chron

(“Cretaceous quiet zone”)
Cenomanian
sedimentary basin fills into genetically related stratal
Cretaceous
M″–3”r
(C34n)
packages bounded by unconformities and their correl-
Albian
M″–2”r ative conformities. Thus, the concept of sequence
(C34n) stratigraphy is closely related to that of allostra-
Aptian M″–1”r
tigraphy. Allostratigraphic analysis applies bounding
discontinuities, e.g. erosional surfaces and marine
Early
(C34n)
M0r
M1 flooding surfaces, for recognition of sedimentary
Barremian M3
M5 entities independent of any genetic model. Sequence
Hauterivian –
M10
M10N stratigraphy applies allostratigraphic features to inter-
M-sequence
Valanginian M11
–
M15
pret the depositional origin of sedimentary packages
M16
145.5
Berriasian M17
M18
by analysing these in a framework of transgressive-
M19
Tithonian
M20
M21 regressive developments producing base level
M22
M23 changes.
Late
Kimmeridgian M24
–
M27
The infilling of a basin is controlled by the interac-
Pre-M25 Marine
M28
anomaly series
Oxfordian M29
magnetic
Callovian
–
M38 tion of tectonics and eustasy, producing base level
M39
Bathonian
–
M41 changes which define the accommodation space for
Middle
S. Switzer- Toarcian S. Switzer- S. Spain

series
Bajocian sediments to be deposited. This is the space or height

Jurassic
Aalenian between the seafloor and the sea level. The evolution
series
land
of basin architecture is controlled by the balance

composite
Toarcian
series
between accommodation space and sediment supply.

Early
Pliensbachian
series
If sediment supply exceeds the accommodation space

land
Sinemurian available, progradational geometries will be the

Austrian
series
199.6 Hettangian
result. If accommodation space and sediment supply
Rhaetian
Rhael-N
Basal-I are roughly balanced, aggradational geometries
Newark basin (E. USA) with cycle scaling
E24
– result. When sediment supply is less than the creation
E18
Norian E17
E16
of accommodation space, retrogradational geometries
Late
E15
E14
are formed. Tectonics and climate are the most impor-
E13
E12 tant factors determining the production and supply of
E11
Carnian sediments.
Triassic
E10
E9
E8
E7 The major bounding and subdividing surfaces of an
E6
Ladinian
–
E1 Exxon-type depositional sequence are commonly
Cycle-scaled in Albania, Greece, Italy
Austria,
Turkey
Fass-N
Middle
Fass-R
IIIy–3
represented by: the lower and upper sequence
IIIy–2
Greece,
Germanic Basin Poland
Anisian Pelson-R
boundaries, transgressive surfaces and maximum
Bithy-N
sn10 flooding surface (Fig. 7.12). These surfaces principally
Early Olenekian –
251.0 Induan sn1 define three sediment bodies: lowstand systems tract,
transgressive systems tract and regressive systems
Fig. 7.11 Reversals of the Earth’s magnetic field from Triassic tract. The building blocks of the systems tracts are
to Cretaceous time. Normal polarity: black bands, reversed
polarity: white bands (From Gabi Ogg 2010, http://stratigra- parasequences, forming parasequence sets.
phy.science.purdue.edu/charts/educational.html) The boundary of a depositional sequence is defined
as a subaerial unconformity and its correlative
Fig. 7.12 Depositional sequence portrayed in a depth-distance marginal to basinal areas. Abbreviations of sequence elements:
and a time-distance diagram. It shows the regional extent of the SB = sequence boundary, LST = low stand systems tract, TS =
condensed section from deep to shallow water, with its time transgressive surface, MFS = maximum flooding surface, HST =
span strongly expanding from coastal to basinal areas. The time high stand systems tract (modified after Loutit et al. 1988)
span of the sequence boundary unconformity decreases from
conformity. This can be recognised as a change in The volume of water in the ocean basins is also a
lithology and facies. The sediments between two primary factor influencing eustatic sea level changes.
unconformities or their correlative maximum regres- Continental and mountain glaciations and
sive surface have been called a transgressive-regres- deglaciations decrease and increase, respectively, the
sive (T-R) sequence. volume of ocean water. These developments are
exemplified by the well-documented regressions and
transgressions accompanying Quaternary glacial and
interglacial periods (Fig. 7.13). Other factors
7.9.2 Accommodation Space
influencing the volume of oceanic waters are the
expansion of water with increasing temperature and
Local accommodation space for sediments represents
the amount of continental groundwater and surface
simply the water depth in the depositional area. A
water.
global factor influencing the accommodation space is
Tectonic movements result in transgressions and
the eustatic sea level stand which is defined by a
regressions of local to regional nature. These
combination of several factors. One of these is the
movements are driven by factors that affect how con-
volume of the ocean basins, which is believed to be
tinental lithosphere floats on the asthenosphere, which
mainly controlled by the rate of seafloor spreading.
is controlled by three main mechanisms: (1) Stretching
Young lithosphere is warm and buoyant, therefore it
of continental lithosphere resulting in replacement of
will rise and decrease the volume of the world ocean.
continental lithosphere by denser and thinner
On the contrary, old lithosphere is colder and denser,
astenosphere which sinks. (2) During stretching, con-
and sinks. Another factor contributing to ocean vol-
tinental lithosphere is heated, becomes less dense and
ume changes is the sedimentation rate.
7 Stratigraphy 249
Fig. 7.13 Schematic illustration of transgressive-regressive cycle, sedimentation takes place only during short episodes
cycles in the Quarternary, resulting from glacially-induced (note the three types of depositional architecture). During most
changes of the volume of ocean water. During a complete of the time there is non-deposition or erosion (hiatus)
tends to uplift. Subsequent cooling results in subsi- encroaching over the supratidal zone, and contain
dence. (3) Weight of tectonic load added to lithosphere shallowing-up subtidal, intertidal to supratidal strata
can produce subsidence e.g. wedges of fold and thrust (Fig. 7.15).
belts pushed into foreland basins. Stacked parasequences form parasequence sets,
When analysing sedimentary successions, it is usu- which show stacking patterns according to their posi-
ally extremely difficulty to distinguish between the tion in the sequence stratigraphic architecture
effects of eustatic sea level changes and tectonic (Fig. 7.16). If sediment supply exceeds the available
movements. Therefore, in sequence stratigraphic stud- accommodation space, progradational stacking geom-
ies, the phrase “relative sea level” is used to express etry will be the result. If accommodation space and
the local sum of global sea level and tectonic sediment supply are roughly balanced, aggradational
movements. geometries occur. When sediment supply is less than
the creation of accommodation space, retrogradational
stacking geometries are formed.
7.9.3 Parasequences
Parasequences are sediment packages of genetically 7.9.4 System Tracts

related beds representing a single minor transgression
event followed by sediment progradation. In the The lowstand systems tract is bounded by the
marine realm parasequences are bounded by flooding sequence base and the first transgressive surface
surfaces. Most of the marine parasequences are (Fig. 7.12). It was deposited when the sea level was
formed in offshore shelf to shoreface environments, located at or below the shelf break. The shelf was
and show an upwards-coarsening development exposed to erosion and crossed by rivers eroding
(Fig. 7.14). In coastal areas upwards-fining incised valleys and transporting sediments to the
parasequences are formed during transgressions, shelf margin. The lowermost part of the lowstand
Facies Environment
Facies Environment
Coal Swamp
Flooding
surface Offshore
Burrowed and Supratidal
Seaward-inclined Foreshore rooted
laminae mudstones
Trough cross- Shoreface
stratified Burrowed
Meters to tens of meters

sandstones mudstones and
trough cross-
laminated Intertidal
Meters to tens of meters
sandstones
Proximal storm with flaser to
beds with Transition zone lenticular
mudstone bedding
interbeds
Sigmoidal
cross-bedded
sandstones Subtidal
Burrowed
mudstones with Offshore
distal storm
beds
Flooding
surface Flooding Supratidal
surface
Fig. 7.14 Example of marine parasequence showing coarsen- Fig. 7.15 Example of coastal parasequence showing fining-
ing- and shallowing-upwards from offshore shales to foreshore and shallowing-upwards from subtidal sandstones to supratidal
sandstones (from University of Georgia web site, 2009) swamp deposits (from University of Georgia web site, 2009)
systems tract is the basin floor fan, dominated by now filled up by sediments, therefore the systems tract
gravity-driven deposition of turbidite lobes and in is characterised by outbuilding of deltas and shorelines.
feeder channels. The upper part of the lowstand
systems tract is the lowstand wedge, which consists
of progradational parasequences. Uppermost, estua-
7.9.5 Condensed Sections
rine sediments are deposited in drowned incised
valleys (Fig. 7.17).
The transgressive surface and maximum flooding sur-
The transgressive systems tract is bounded by the
face were formed during periods of transgression
transgressive surface and the maximum flooding sur-
characterised by sediment starvation, which leads to
face. It is composed of retrogradational parasequence
development of condensed intervals forming impor-
sets showing a generally upward-fining and upward-
tant marker horizons in the sequence stratigraphic
deepening development. This systems tract is gener-
architecture of a basin (Fig. 7.12). Condensed sections
ally thin because much sediment is trapped in
are thin marine horizons commonly showing several
estuaries, and the advancing transgression succes-
of the following features: pelagic to hemipelagic sedi-
sively creates new accommodation space.
mentation; presence of apparent hiatuses; borrowed
The highstand systems tract is developed between
horizons and hardgrounds; presence of authigenic
the maximum flooding surface and the upper sequence
minerals such as glauconite, phosphate, siderite; high
boundary. This systems tract represents a relatively
faunal abundance (if not anoxic); high faunal diversity
thick succession of progradational parasequences
(if oxic), low diversity (if hypoxic). Deposition of
showing an upward-coarsening and upward-shallowing
organic-rich black shales with high gamma readings
development (Fig. 7.17). The parasequences downlap
is typical of many condensed sections.
to the maximum flooding surface. The estuaries are
7 Stratigraphy 251
1 km
Foreshore Shoreface Flooding
Transition zone
surface
20 m bedsets
Flooding
surface
Outcrop A Outcrop B Outcrop C

Mud sand Mud sand Mud sand
Seaward-dipping planar beds
Trough cross bedding
Hummocky beds
Bioturbated marine mudstone
Fig. 7.16 Lateral and vertical relationships between parasequences forming a parasequence set, which progrades from foreshore to
offshore shelf (from University of Georgia web site, 2009)
Flooding surfaces Maximum flooding surface
HS T
TST
HS T
Fluvial or estuarine L ST
sandstones in incised valleys
Coastal-plain sandstones
and mudstones
Transgressive surface
Shallow-marine sandstones
Shelf and slope mudstones
and thin sandstones
Sequence boundary
Submarine-fan and levee-
channel sandstones
Fig. 7.17 Depositional sequence (type 1) portrayed as a down- tracts and parasequences. LST ¼ low stand systems tract, TST
dip section across an idealised passive continental margin with ¼ transgressive systems tract, HST ¼ high stand systems tract
shelf break. The diagram shows sequence boundaries, systems (from University of Georgia web site, 2009)
Condensed sections have a large regional extent, sea level fall can continue to a position below the
and show an expanding time span from coastal to shelf-slope break, or stop on the shelf. According to
basinal areas (Fig. 7.12). They are particularly well these developments, two types of sequences are
developed at the maximum flooding surface and help distinguished.
to identify this horizon. Thus, condensed sections rep- Type 1 sequence: This is formed if the sea level
resent important levels of correlation within sedimen- falls below the shelf break at the sequence boundary
tary basins. They are also applicable to recognition of (Fig. 7.17). The shelf is exposed for erosion and a low
global transgressive-regressive cycles. stand systems tract is formed. The boundary reveals a
sea level fall below the previous shoreline. A type 1
sequence can be formed in two basinal developments:
7.9.6 Sequence Types by deposition in a basin having a shelf-slope break; or
by deposition in a basin that lacks a shelf-slope break
During development of the sequence boundary the but has an evenly sloping ramp-like margin.
relative sea level fall is of varying magnitude. The
Type 2 sequence: The sea level position above the Embry, A.F. 2009. Practical Sequence Stratigraphy. Canadian
sequence boundary corresponds to the position of the Society of Petroleum Geologists, Online at www.cspg.org,
81 pp.
previous shoreline below the boundary. Consequently, Gradstein, F.M. and Ogg, J.G. 2004. Geological time scale 2004
the boundary reveals an aggradational development. – Why, how and where next! Lethaia, 37, 157–181.
During the sea level low stand a shelf margin systems Hallam, A. 1992. Phanerozoic Sea-level Changes. Columbia
tract is deposited. University Press, New York, NY, 266 pp.
Haq, B.U. and Boersma, A. 1978. Introduction to Micropaleon-
tology. Elsevier, North-Holland, 376 pp.
Hedberg, H.D. (ed.). 1976. International Stratigraphic Code
Guide. Wiley, New York, 200 pp.
7.9.7 Carbonate Sequence Stratigraphy Jones, R.W. 1996. Micropalaeontology in Petroleum Explora-
tion. Oxford University Press, Oxford, 432 pp.
Sequence stratigraphy was originally developed for Jones, R.W. 2006. Applied Palaeontology. Cambridge Univer-
sity Press, Cambridge, 434 pp.
siliciclastic systems, but the same principles can be
Kautsakos, E.A.M. (ed.). 2005. Applied Stratigraphy. Springer,
applied to carbonate systems as well. However, the Dordrecht, 488 pp.
special features of carbonates make the appearance of Lindberg, D.R., Lipps, H.J. and Hazel, J.E. 1993. Micropaleon-
stratigraphic elements somewhat different in the two tology. In: Lipps, H.J. (ed.), Fossil Prokaryotes and Protists.
Blackwell Scientific Publications, Boston, MA, pp. 31–50.
system types.
Lipps, J.H. (ed.). 1993. Fossil Prokariotes and Protists.
Carbonate production takes place mainly in situ, Blackwell Scientific Publications, Boston, MA, 342 pp.
and is particularly high during moderate sea level Loutit, T.S., Hardenbol, J., Vail, P.R. and Baum, G.R. 1988.
rise, which typically results in extremely thick trans- Condensed sections: The key to age determination and cor-
relation of continental margin sequences. In: Wilgus, C.K.,
gressive systems tracts. Parasequences formed in
Hastings, B.S., Kendall, C.G.St.C., Posamentier, H.V., Ross,
peritidal environments are particularly thick. The C.A. and Van Wagoner, J.C. (eds.), Sea Level Changes – An
highstand systems tracts in carbonate successions are Integrated Approach. SEPM Special Publication 42, pp.
much thinner than in siliciclastic sequences. During 183–213.
Nagy, J., Hess, S., Dypvik, H. and Bjærke, T. 2011. Marine shelf
transgression, extremely rapid sea level rise stops in
to paralic biofacies of Upper Triassic to Lower Jurassic
situ carbonate production, which leads to the forma- deposits in Spitsbergen. Palaeogeography, Palaeocli-
tion of condensed sections. Carbonates are prone to matology, Palaeoecology 300, 138–151.
dissolution, leading to sequence boundaries com- Nystuen, J.P. 1998. History and development of sequence stra-
tigraphy. In: Gradstein, F.M., Sandvik, K.O. and Milton, N.J.
monly marked by solution surfaces, karst development
(eds.) Norwegian Petroleum Society Special Publication 8,
and palaeosols. Elsevier, Amsterdam, pp. 31–116.
Ogg, J.G. 1995. Magnetic polarity time scale of the Phanero-
zoic. Global Earth Physics – A Handbook of Physical
Constants. American Geophysical Union, Washington, DC,
Further Reading pp. 247–270.
Ortiz, S., Alegret, L., Payros, A., Orue-Etxebarria, X.,
Apellaniz, E. and Molina, E. 2011. Distribution patterns of
Armstrong, H.A. and Brasier, M.D. 2005. Microfossils, 2nd benthic foraminifera across the Ypresian-Lutecian
edition, Blackwell Publishing, Oxford, 296 pp. Gorrondatxe section, Northern Spain: Response to sedimen-
Basov, V., Nikitenko, B. and Kupriyanova, N. 2008. Lower and tary disturbance. Marine Micropaleontology 78, 1–13.
Middle Jurassic foraminiferal and ostracod biostratigraphy Rodrı́guez-Tovar, F.J., Uchman, A., Payros, A., Orue-
of the eastern Barents Sea and correlation with northern Etxebarria, X., Apellaniz, E. and Molina, E. 2011. Sea-level
Siberia. Norwegian Journal of Geology 88, 259–266. dynamics and palaeoecological factors affecting trace fossil
Coe, A.L. (ed.). 2005. The Sedimentary Record of Sea-level distribution in Eocene turbiditic deposits (Gorrondatxe sec-
Changes. Cambridge University Press, Cambridge, 287 pp. tion, N Spain). Palaeogeography, Palaeoclimatology,
Copestake, P. 1993. Application of micropaleontology to hydro- Palaeoecology 285, 50–65.
carbon exploration in the North Sea. In: Jenkins, D.G. (ed.), Setoyama, E., Radmacher, W., Kaminski, M.A. and Tyszka, J.
Applied Micropaleontology. Kluwer Academic Press, 2013. Foraminiferal and palynological biostratigraphy and
Dordrecht, pp. 93–152. biofacies from a Santoniane Campanian submarine fan sys-
Dera, G., Neige, P., Dommergues, J-L. and Brayard, A. 2011. tem in the Vøring Basin (offshore Norway). Marine and
Ammonite paleobiogeography during the Pliensbachian- Petroleum Geology 43, 396–408.
Toarcian crisis (Early Jurassic) reflecting paleoclimate, Sluijs, A., Ursula R., Schouten, S., Brumsack, H.J., Sangiorgi,
eustasy and extinctions. Global and Planetary Change 78, F., Jaap S., Sinninghe, J., Damste, S. and Brinkhuis, H. 2008.
92–105. Arctic late Paleocene–early Eocene paleoenvironments with
special emphasis on the Paleocene-Eocene thermal
7 Stratigraphy 253
maximum (Lomonosov Ridge, Integrated Ocean Drilling overview of the fundamentals of sequence stratigraphy and
Program Expedition 302). Paleoceanography, vol. 23, key definitions. In: Wilgus, C.K., Hastings, B.S., Kendall, C.
Pa1S11, doi:10.1029/2007pa001495. 17 pp. G.St.C., Posamentier, H.V., Ross, C.A. and Van Wagoner, J.
Van Wagoner, J.C., Posamentier, H.V., Mitchum, R.M., Vail, P. C. (eds.), Sea Level Changes – An Integrated Approach.
R., Sarg, J.F., Loutit, T.S. and Hardenbol, J. 1989. An SEPM Special Publication 42, pp. 39–45.
Chapter 8
Seismic Stratigraphy, Sequence Stratigraphy

and Basin Analysis
Knut Bjørlykke
8.1 Seismic Stratigraphy expelled will also give higher velocities, not only
because the water content decreases, but because more
Seismic records are based on measurements of the time numerous and larger contacts are formed between the
sound waves (seismic waves) take to travel through clastic grains. Velocities in moderately consolidated
rock. The sound or signal is produced by explosives or sediments, such as the Tertiary sediments of the North
compressed air (air guns). Rock is an elastic medium Sea, are 2–3 km/s. In more consolidated (compacted and
and the velocity of sound conveys a lot of information cemented) sedimentary rocks which have not been
about the properties of the rock. Normal sound waves subjected to metamorphosis, velocities are mostly
(P-waves) travel through both the solid phase, which for between 3 and 5 km/s. This is the case for many of the
the most part consists of minerals or rock fragments, and Mesozoic sediments in the North Sea. Metamorphic and
the liquid or gas in the pores. Shear waves (S-waves) on eruptive rocks will have velocities of about 5–6 km/s.
the other hand can only go through the solid phase. Limestones will often have higher velocities than
The most important parameters influencing the sandstones at the same depth because they often are
velocity of sound are: porosity, mineral composition, more cemented and because carbonate cement has a
and the degree of cementation. These factors determine high degree of rigidity and low compressibility. Carbon-
the stiffness of the rock (Bulk modulus, see Chap. 11). ate reefs may be strongly cemented and have high
The velocity of sound waves in water is about 1,500 m/s, velocities at shallow depth. Sandstone in turn provides
but depends on temperature and salt concentration. a more rigid medium for sound waves than shale at the
Sound passes through unconsolidated sediments at same depth, because of its grain-supported structure.
velocities which are only slightly higher than the veloc- If the rocks do not contain oil and gas we can assume
ity in water (1,500–2,000 m/s, and sometimes even that their porosity is identical with the water content in
lower) because they have high water content and the rock. Velocity will then be a function of porosity (φ),
because the framework on which the sediment grains and if we know the velocity of sound in the rock matrix,
are based does not offer any real strength (stiffness) as a we can calculate the porosity using Wyllie’s equation:
medium for the seismic waves.
Cementation of sand with carbonate or siliceous 1=Vr ¼ ð1 φÞ=Vm þ φ=Vf
cement will bind the grains together in a framework
which will increase the stiffness and velocity consider- where
ably even if the porosity is relatively high. Compaction Vr ¼ velocity in rock when saturated with liquid, i.e.
due to overlying sediments which causes water to be the measured velocity
Vf ¼ velocity in the fluid
Vm ¼ velocity in the rock matrix.
The inverse values of the velocities are expressions
K. Bjørlykke (*)
of the time the signals take to travel through a layer of
e-mail: [email protected] certain thickness.
256 K. Bjørlykke
If the pores are filled with gas instead of liquid (water where ρ1 and ρ2 are the densities of the two rocks, and
or oil), the velocity reduction will be even greater v1 and v2 their respective velocities (Fig. 8.2). We see
because the sound travels much more slowly through that the greater the difference in density and velocity,
gas than through a liquid. Whyllie’s equation is however the greater the amount of energy which will be
not an accurate description of the relation between poros- reflected. Sandstone will often have significantly dif-
ity and velocity. ferent acoustic impedance from shale, and a consider-
Vr thus approaches Vm at zero porosity, and for able amount of sound energy will be reflected from the
sandstone Vm is about 5.5–6 m/s. boundary between a sandstone bed and a shale bed.
If we know the velocity of the rock matrix and the This is however not always true and the contrast
fluid we should be able to calculate the porosity as a depends on the type of clays and their clay mineral
function of velocity, but the Wyllie equation is very composition. Limestones will tend to have both high
much a simplification. Rocks with the same porosities velocities and high densities. The result will be even
can have rather different velocities depending on the greater contrast in acoustic impedance between
type of grain contacts and the distribution of cement. limestones and, for example, shales. However, this
When sound waves move between sedimentary contrast will always depend on the porosity of the
beds with different velocities, they will be refracted limestone in question.
according to Snell’s law (see Fig. 8.1): On a seismic section, beds which have greatly
contrasting acoustic impedances stand out as strong
sin x1 = sin x2 ¼ v1 =v2 reflectors. This makes it possible to map characteristic
rock boundaries, e.g. the top of a limestone or the
Here x1 is the angle of incidence of the waves where boundary between shales and sandstones, using seis-
they meet the boundary plane between two strata, and x2 mic sections (Fig. 8.3).
is the angle of refraction of the emergent waves. v1 and As we have seen, the critical parameters determin-
v2 are the velocities through the respective rock strata. ing the reflection coefficient are the velocities and
If the two beds have different velocities, they will densities of the different lithological units. Using a
as a rule also have different densities, and part of the well, we may measure a velocity log (sonic log) and
acoustic energy will not be refracted, but reflected. a density (ρ) log which record how these properties
How much of the energy is reflected depends on the change through the sequence (see Fig. 8.3a). The
difference in the acoustic impedance, which is the product of velocity and density may then be computed
product of velocity and density (Fig. 8.2). and presented as an acoustic impendance (p·v) log.
The coefficient for reflection (R) is then:
R ¼ ð ρ2 v 2 ρ1 v1 Þ=ðρ2 v2 þ ρ1 v1 Þ
Reflection
Snells’s refraction law
Layer 1,
Velocity: V1
Layer 1,
Density: D1
X1 Velocity: V1
Layer 2,
Velocity: V2
Layer 2,
Density: D2
Velocity: V2
X2
Sin X1 V1
Sin X2 V2
Fig. 8.2 Diagram for the reflection of waves in a layered
sedimentary sequence. The amount of energy reflected is a
function of acoustic impedance, which is the product of the
Fig. 8.1 Snell’s law for the refraction of sound waves density of the beds (ρ) and the velocity of the sound waves (v)
8 Seismic Stratigraphy, Sequence Stratigraphy and Basin Analysis 257
a Lith. log V log D log AI log RC log horizontal and does not always follow the other rock
strata. This is called a “flat spot” and exemplifies
direct indication of hydrocarbons (usually gas)
through seismic methods. At greater depth and higher
pressure the contrast between gas and oil and also oil
and water will be lower.
Reflections which are multiples of a relatively flat
sea-bottom reflection are also near-horizontal and may
be confused with “flat spots”, but these may be
removed by filtering the data during processing.
b Synthetic seismograms with Temperature-dependent diagenetic reactions may also
Lith. log RC log different pulse shapes produce horizontal reflections, e.g. the transformation
of amorpheous silica (opal A) and opal CT to quartz
which produces a strong increase in velocity and den-
sity. If the geothermal gradients are rather uniform
(horizontal isotherms) this diagenetic transition will form
horizontal reflections which may crosscut the bedding.
It may also be possible using AVO (see Chap. 17)
to detect contacts between oil and water. Oil has a
lower velocity than water and if the oil has a high gas
content, the difference is even greater.
Fig. 8.3 (a) Illustration of a reflection coefficient log based on Changes in the oil/water contact during production
velocity (v) and density (ρ) (from Anstey 1982). (b) Synthetic can be monitored by shooting seismic in an oilfield at
seismogram of the different reflection coefficients (RC log) in a
intervals of several years (see 4D Seismics). In deeper
bedded sequence. The resolution of the seismogram varies with
the width of the seismic pulse (from Anstey 1982). Normally, reservoirs with lower porosities it is more difficult to
sediments have to exceed 20–30 m in thickness to be distinctly detect fluid contacts.
recorded on a seismic section, depending on the wavelength of High pore-pressure causing reduced effective stress
the seismic signal. The upper sand contains some gas in the
and stiffness (elasticity) results in lower seismic
upper part, causing lower velocities
velocities in sandstones and clays, particularly at
depths less than 3 km. Limestones can have relatively
Note that at the boundary between limestones and high velocities even at shallow depth.
shales there is a very significant drop in both velocity At greater depths chemical compaction is the most
and density, resulting in a marked change on the important factor, and in clastic sediments it is gener-
acoustic impedance log. Sandstones usually have ally a function of temperature and less dependent on
higher velocities than shales, but they may not differ the effective stresses. Nevertheless, seismic velocities
much in density, so the difference in acoustic imped- are often observed to fall in overpressured rocks even
ance will be small. The reflection coefficient, which is if the porosity is not significantly reduced. This may be
an expression of differences in acoustic impedance, is related to reduced stress at grain contacts.
a synthetic seismic trace such as we would have seen As the quality of seismic data has improved, one
on a seismic cross-section through the sequence. has been able to use seismic profiles for detailed inter-
If we have gas instead of water in a rock, the pretation of stratigraphic relations and even deposi-
velocity will be considerably reduced. The velocity tional environments. The basis for this is that seismic
of sound in gas is much lower than it is in liquid, reflections usually follow time lines in a sedimentary
depending on composition, temperature and pressure. sequence. In other words, seismic reflections follow
The boundary between gas-bearing and water-bearing surfaces which constituted the seafloor surface at the
rocks may produce a strong reflection because there is time when the sediments were deposited.
a large difference in impedance between the two Seismic reflections can be followed from a sandy
layers. For this reason the boundary between gas and facies into a clay/siltstone shale facies. We can, for
oil is often revealed as a strong reflector because it is example, follow seismic reflections from the fluvial
258 K. Bjørlykke
part of a delta out into the pro-delta muds. Relatively package of sediments deposited during a definite
thin transgressive sandstones may be deposited on the period of time, defined by unconformities above and
delta top, and are useful for correlation. This is below.
because the relief on land may be very low so that a The unconformities may be due to a break in sedi-
few metres sea level rise causes a large transgression. mentation due to relative changes in sea level or other
Transgressions will then result in a wide shelf with causes such as changes in sediment supply.
little clastic supply, causing deposition of carbonates. A seismic profile through a sedimentary succession
Thin calcareous sediments and limestones may there- has reflectors that show layers of contemporary
fore also be relatively close to time lines. Limestones deposits. Terminolgy has been established to describe
typically have rather high velocities and so we find the geometry of seismic reflections (Fig. 8.4).
strong reflectors. On the delta slope, however, fluvial Baselap is the term for gradual deposition above
sediments and marine delta front sand may prograde the lower boundary of a depositional series and
into a marine basin over considerable geological time, represents a small unconformity.
depending on the depth of the basin and the sediment If a sequence progrades out across an unconfor-
supply. The seismic reflections will follow the surface mity, depositing successively younger beds
from delta front sand to delta slope, where we have basinwards, we call this type of contact downlap.
sand and mud (shale) beds lying parallel with the The building of beds out into the basin like this is
slope. Even though we can often follow the reflector called offlap. We are thus dealing with a bed which
a little further into the basin, it will be less marked has a primary depositional slope with respect to the
there because there is less contrast in lithology and unconformity surface.
hence in acoustic impedance. Different types of clay
may however also produce differences in seismic
response and particularly smectitic clays are
characterised by low density and velocity.
BASELAP
The shifting of sedimentation input from one part
of the delta to another through channel switching Offlap
T2
(distributary abandonment as part of delta-lobe
shifting), also contributes to the formation of T1
lithological contrasts on the delta slope. Progradation Downlap
of new delta lobes results in deposition of sheets of
sand over mud near time-stratigraphic boundaries. The
inactive delta lobes will be compacted and often High
develop a thin carbonate or transgressive sandstone Distal onlap Proximal onlap
layer, while sedimentation takes place in the active
TOPLAP Toplap
lobe. The small unconformities produced in this way
also tend to produce lithological contrasts which may
be recorded on the seismic record.
Seismic sections through prograding deltas provide
information about the water depth, the rate of sediment Erosion toplap
input and the wave energy in the basin.
8.2 Different Types of Seismic

Signatures
A stratigraphic unit which is composed of a conform- Concordance
able bedding series, genetically linked together at the Fig. 8.4 Different types of seismic stratigraphic relations. The
top and bottom by unconformities, is called a deposi- seismic reflectors represent time lines as a rule, i.e. rocks depos-
tional sequence. A depositional sequence is thus a ited at the same time
D
H Coastal onlap
Unconformities
Unconformity
H - Relative sea level rise during period of onlap

D - Relative sea level fall
Fig. 8.5 Coastal onlap followed by a sea level drop and a renewed onlap. The coastal onlap indicates a relative sea level rise of
H metres but this may be due to local tectonic movements and loading by water and sediments
Onlap is the term for primarily almost horizontal REFLECTION FREE:

beds against a sloping unconformity which may be a Homogeneous medium (salt domes, igneous rocks,
mud, overpressured shales).
submarine or a subarial slope. It occurs most com-
monly as a result of sedimentation gradually burying CHAOTIC:
Reflecting layer folded or contorted.
an unconformity during a transgression onto a land
surface that provides the unconformity (Fig. 8.5). We PARALLEL REFLECTIONS:
call this proximal onlap or coastal onlap. If sedimen- Shelf sediments.
If large amplitude: Shale or carbonate
tation covers a positive relief structure in the basin, for If low amplitude: Shale/siltstone
example a horst or a salt dome, we get distal onlap
SIGMOID REFLECTIONS:
(Fig. 8.4). Low energy, progradation into a sedimentary
Toplap is the contact between the seismic reflectors basin or rapid subsidence.
and an upper unconformity. An erosion surface will OBLIQUE REFLECTIONS:
truncate the reflections sharply, forming an erosional Progradation simultaneous with erosion.
truncation (erosional toplap). Shallow shelf, beach or delta.
Transgression may form a coastal onlap across an SHINGLED REFLECTIONS:

unconformity (Fig. 8.5). Later the relative sea level Progradation in a shallow environment.
fell, and a prograding offlap sequence formed. Finally
Fig. 8.6 Classification of internal structures in seismic units.
the sea level rose again and an onlap sequence formed.
Layering in the sedimentary sequences causes changes in the
H and D are not to be regarded as absolute values for acoustic impedance
sea level changes. They must be adjusted for isostatic
responses to loading and unloading and to tectonic
uplift or subsidence. H is the relative rise in sea level contemporaneously, it is also possible to a certain
during the period with onlap. D is the relative fall in extent to interpret the depositional environment. The
sea level which led to a downward shift in the most important parameters we use are:
prograding downlap. 1. Reflection amplitudes – the strength of the
reflections. As we saw above, the proportion of
the energy reflected at the boundary between two
8.3 Interpretation of Lithology and beds is a function of the difference in the acoustic
Sedimentary Facies by Means of impedances (velocity multiplied by density). If we
Seismic Profiles have an alternating series of different beds, the
distance between the bed boundaries in relation to
In addition to structural data, a seismic profile the wavelength of the transmitted seismic waves
provides us with information about the properties of will play a major part (Fig. 8.3).
sedimentary rocks. The internal properties of seismic 2. Reflector frequency. The distance between the
units provide important information about the lithol- reflectors will indicate the thickness of the bed,
ogy (Fig. 8.6). but there will be a lower limit to the thickness that
Because seismic reflections mainly represent time can be detected, which should correspond to about
lines, i.e. sedimentary beds which were deposited the half wavelength of the seismic waves.
260 K. Bjørlykke
3. Internal velocity in the beds. The internal velocity Primary seismic reflections will be deformed
of the bed can provide information about lithology through tectonic deformation so that they become
and porosity. tilted or folded. Folded beds will only be realistically
4. Reflector continuity. The continuity of reflectors depicted if the folds are sufficiently gentle that the
will be a function of how continuous the sediment beds have a low angle of dip.
beds are, information which is essential for We can distinguish faults where good reflections
reconstructing the environment. suddenly stop, suggesting an abrupt lateral change in
5. Reflector configuration. If we take the compaction lithology. Faults are generally too steep to reflect the
effect into account, the shape of the reflecting beds sound wave straight back again, and the fault plane
gives us a picture of the sedimentation surface as it itself will not appear as a reflector on the seismic
was during deposition. The slope of the reflectors, profile. Because of the special “edge” effects near
for example, represents the slope of prograding faults, the ends of the reflecting layers which should
beds in a delta sequence with later differential define faults will not be quite correctly located on the
compaction and tilting superimposed. Erosion seismic profile, and it may therefore be difficult to
boundaries with unconformities will in the same trace the fault entirely accurately. The termination of
way show the palaeo-topography during erosion. beds against faults may produce diffraction from a
Seismic profiles provide information about the fill- point source, giving a curved alignment. Special treat-
ing of the basin in response to the type of subsidence ment of seismic data (migration) will rectify a fair
and sedimentary facies. number of these errors and give a more correct picture.
When interpreting lithology and depositional envi- In recent years seismic lines have been shot with
ronment from seismic profiles, it is important to use smaller and smaller grid spacings to obtain a better
sedimentological models as aids. If there are well data map of the reservoir structure. A three-dimensional
on lithologies, these must also be integrated. The seismic data set is then produced and seismic sections
information we gain from seismic profiles is often can be constructed at any angle relative to the grid.
not sufficient by itself for an unambiguous interpreta- This method also allows us to construct horizontal
tion. There may be several lithological compositions time-slices through the structure. This is almost like
and environments which could give similar seismic a topographic or geological map which is a horizontal
signatures. Only by looking at the whole basin in a projection of the geology. It is also a very powerful
sedimentological context do we have a good basis for method of delineating faults and other important struc-
selecting the interpretation which seems most tural elements.
reasonable. Another relatively new development is borehole
seismics, particularly vertical seismic profiles
(VSPs). This method involves firing shots near the
8.4 Seismic Interpretation of seafloor close to a well and recording signals at regular
Sedimentary Basins depth intervals in the well. The main advantage of
VSPs is that they produce a very good profile of the
Seismic profiles present a picture of the way the basin seismic velocity as a function of depth, better than a
has been filled in. This is a result of an interaction synthetic seismic log.
between the rate of subsidence, rate of deposition and
the energy of the depositional environment. The seis-
mic signatures can be used to interpret facies and basin 8.6 Changes in Sea Level
infilling.
It has been clear for a long time that there are
unconformities in sedimentary sequences which can
8.5 Faults and Tectonic Boundaries be correlated over long distances, and that there were
periods in geological history with a high sea level and
It must be remembered that the principle we use for others when it was low.
calculating the depth to a reflecting boundary assumes Proximal onlaps are due to sedimentation moving
that the layering is relatively horizontal. landwards over an unconformity surface. If we are
dealing with a coastal deposit, a proximal onlap will During the Quaternary, cyclic changes in sea level
mean that the sea level has risen in relation to the land of up to 120 m accompanied the growth and decay of
surface which forms the top of the unconformity. On continental ice sheets. These changes were rapid and
seismic profiles we can see onlaps onto the land, of large magnitude and can be traced throughout much
measure the height range between the lowest and of the world. Nevertheless it is often the local
uppermost onlaps, calculate the difference in seismic conditions which count most. In the areas which had
time and convert this into approximate thickness. supported ice sheets, such as Scandinavia, the melting
However, we must remember that the thickness of of the ice led to uplift due to unloading, and this
the sediments deposited is due not only to a rise in exceeded the rise in sea level so that there was a
eustatic sea level, but also to local subsidence of this regression. A 2,000 m thick ice sheet will cause
part of the basin. The weight due to increased water about 600–700 m of isostatic depression of the crust
depth will cause further subsidence, and sedimentation under the ice because the density of ice (0.9 g/cm3) is
will increase the load, resulting in further subsidence about one third of that of rocks (2.7 g/cm3).
to attain isostatic equilibrium. Local tectonic subsi- When seismic stratigraphy was established (Vail
dence may produce a relative change in coastal onlap et al. 1976), it was assumed that most of the variations
in a seismic profile. Regressions are defined as the in sea level that could be interpreted from seismic
boundary between land and sea being displaced out records, were attributable to eustatic changes and
into the basin. They may be caused by a fall in sea thus could be employed for correlation across great
level which will shift the coastline to further out on the distances. There was a problem in that one did not
shelf or to the edge of the continental slope. Here know of any other processes than the accumulation of
unloading of some of the water plus erosion of ice on the continents capable of producing large and
sediments leads to isostatic uplift of the area landward rapid global sea level changes. However we only know
of the coast line, so that the measured regression is of such ice ages from the Quaternary and late Tertiary,
greater than the real lowering of the sea level. the Carboniferous-Permian, late Ordovician and the
The definition of a transgression is that the sea end of the Precambrian.
encroaches over what was previously land, while a The general consensus now is that the rapid sea
regression is a situation where the boundary between level changes outside the ice ages were mainly caused
sea and land (“shoreline”) moves seaward so that by tectonic activity. Nevertheless, transgressions and
seabed becomes land. regressions can be correlated over greater or smaller
Transgressions and regressions are not always distances, depending on the type of tectonic
directly related to sea level changes. When a delta movements. In particular large-scale plate tectonic
builds out into the sea, there is a local regression on movements involving changes in seafloor spreading
the delta even if the sea level has not fallen. If sedi- rates and subduction rates cause global sea level
mentation is sufficiently rapid, we can have a local changes.
regression on a delta even with a rising sea level. In order to prove that a transgression is eustatic, we
Along a coastline we can have transgressions in need rather accurate age determinations of the trans-
some areas and regressions in others at the same gressive deposits from many parts of the world, but it
time, depending on the rate of sedimentation or ero- is difficult to get high resolution datings. However
sion with respect to sea level change. seismic reflectors correlated with well data can be
The term “forced regression” is used to indicate used to obtain good stratigraphic control and a picture
that there is a primary lowering of the sea level. of onlap and offlap can be interpreted in terms of sea
Changes in sea level can be due to: level variation. They may represent relative sea level
a. Local tectonic movements, for example uplift of a changes for a part of a sedimentary basin, or eustatic
horst or subsidence of a graben structure. (global) transgressions or regressions.
b. Plate-tectonic movements which can be of great Although one might perhaps expect that the sedi-
extent, but are not global. mentation within a basin would primarily be
c. Sea level changes. These are global and are called characterised by local tectonic conditions, drainage,
eustatic sea level changes. depositional conditions etc., studies of thousands of
seismic profiles and wells from many sedimentary
262 K. Bjørlykke
basins indicate that there have been simultaneous mechanism can explain the great fluctuations in sea
transgressions and regressions in completely different level through geological time. Note too that periods
parts of the world. This has been documented by with major transgressions can be correlated with
correlating seismic profiles with oil wells in the same periods of rapid seafloor spreading, for example in
area where the age of seismic unconformities and the Cretaceous and Carboniferous periods. The deep
depositional sequences can be dated by means of bio- channels formed in connection with subduction are, in
stratigraphy. It turns out that characteristic seismic fact, small in relation to the width of the spreading
reflectors which represent falls in sea level are of ridge. In Permian and Triassic times we had one big
approximately the same age, for example in the supercontinent and little seafloor spreading. This was a
North Sea, South China Sea, Mexican Gulf and regressive period with a large land area.
Alaska. It has become clear that many areas, espe- Drying out of cut-off ocean basins may also lead to
cially continental margins, have a rather similar tec- eustatic changes in sea level. There is much to indicate
tonic history which is ascribable to global seafloor that the Mediterranean Sea was cut off from the Atlan-
spreading. tic and dried up in Upper Miocene (Messinian) times.
Much of geological time is not represented by This reduced the world’s total volume of ocean basins,
deposits, though; there are usually greater or lesser increasing the sea level by about 5–6 m.
breaks in deposition (hiatuses). This becomes very Crustal thickening and thinning. An increase in the
clear when one looks closely at the continental depth to the Moho (seismic discontinuity separating
sequences. Ocean floor sequences are more continu- the Earth’s crust and mantle) will lead to elevation of
ous, but also there one finds clear breaks in deposition. the land. The greatest land elevation results from con-
The time represented by breaks (hiatuses) varies tinental collision, when the thickness of the continen-
greatly because there will always be some sedimenta- tal crust may be doubled (to 70–80 km), as in the
tion somewhere, and correlation with sequences with Himalayas.
continuous sedimentation entails trying to find signs of Stretching and thinning of the continental crust will
rapid changes in facies and deposition depth. move heavy mantle rocks upwards and increase the
Microfossils can be helpful in indicating water depth, average density of the rocks down to a compensation
even though they are not always reliable. depth of 100 km. This will lead to subsidence and
Mapping unconformities in the field can be difficult more low density sediments can be accumulated. We
in the absence of abundant exposures. Many of the see this at the transition between continental and oce-
best examples have therefore been found from desert anic crust, and where we have rift formation the conti-
areas in the USA and elsewhere. nental crust thins below the rift, causing graben
formation.
Variations in the temperature gradient affect the
8.7 Changes in the Volume of the density of the rocks and thereby the isostatic equilib-
Ocean Basins rium. Rifting causes elevation of the areas along the
margin of the rift where the crust is not thinned (e.g.
As we have seen, large-scale ocean floor topography is East Africa) and subsidence of the whole area when
a function of the age of the seafloor, i.e. how long it the rifting ceases and the crust cools (e.g. the North
has had to cool down since it was formed, and of the Sea).
overlying sediment thickness. Periods with rapid sea- The major transgressions in the Cambrian, Ordovi-
floor spreading will result in relatively broad spread- cian and Cretaceous, can be explained fairly satisfac-
ing ridges which cause the volume of the ocean basins torily by means of plate tectonic models. A relatively
to decrease, and seawater then will spread further onto low sea level at the end of the Palaeozoic
continent margins. If all seafloor spreading ceased, the (Carboniferous-Permian) and the beginning of the
spreading ridges would slowly sink and within about Mesozoic (Triassic) can be explained as being due to
100 million years would have disappeared almost limited seafloor spreading, for example along the
completely, The volume of the oceans basins would Atlantic Ocean. With fewer and smaller spreading
then be greater, as there would be a sea level drop ridges the oceans could accommodate more water
corresponding to the volume of the ocean ridges. This and consequently less seawater would flood the
continents. On the continents there was active rifting For the most part, then, we can apply the classical
without seafloor spreading and this increased the geo- Airy isostasy model to sedimentary basins, which
thermal gradient and elevated areas which were previ- enables us to draw a number of interesting
ously covered by shallow epicontinental seas, to above conclusions.
sea level. Uplift and subsidence are also linked to variations
The short-term variations in sea level are more in geothermal gradients and heat flow in sedimentary
difficult to explain, however. For geological periods basins.
which experienced major continental glaciations Movement of rocks in relation to the surface affects
(Quaternary-Upper Miocene, Permian-Carboniferous, the geothermal gradients. Erosion removes the upper-
Upper Ordovician and late Precambrian) we can resort most, colder strata so that warmer strata come closer to
to glacioeustatic changes in sea level (changes in sea the surface, and the geothermal gradient increases. As
level due to glaciation), which are very rapid in geo- a result of subsidence of a sediment basin and sedi-
logical terms. Widespread sea level changes in the mentation, heat flow upward will be partly offset by
Lower Tertiary and Mesozoic must be due to plate rock subsidence, giving lower geothermal gradients.
tectonic factors. Sedimentary basins with high rates of sedimentation
Subsidence of the seafloor and continental crust of are therefore often characterised as “cold basins”.
various thicknesses is a function of time after the Tectonic elevation and erosion will increase geother-
rifting stage. The spreading ridges with hot basalt mal gradients.
(oceanic crust) often lie at about 2 km water depth or The geothermal gradient in seafloor rocks is consis-
above sea level (island). Seafloor basalt gradually tently greater than it is over the continents, and on the
subsides to almost 6 km due to cooling over 100 ma continents it is highest in areas of volcanic activity.
without sediment loading (Fig. 8.7a). Sedimentation Areas with a high geothermal gradient due to vol-
on top of the sea floor basalt will gradually reduce the canism will slowly cool down to normal gradients
average density of the crust down to the compensation when volcanic activity ceases. It may take about 100
depth (about 100 km) and the water depth will then be million years before a normal geothermal gradient is
reduced. Great water depth is found on old continental re-established, due to cooling and contraction in accor-
crust where there is little sediments (Fig. 8.7b). dance with the crust’s coefficient of expansion and
isostatic subsidence due to increased density. Seafloor
basalt is hot and flows at relatively shallow depths in
8.8 Sedimentation and Isostatic the Earth’s crust, and the spreading oceanic ridges are
Equilibrium only about 2.2 km below the surface of the sea. After
about 180 million years of cooling, the water depth at
8.8.1 Why Do Sedimentary Basins isostatic equilibrium is about 5.7 km without sediment
Subside? loading. With sediment loading the oceanic crust may
subside to about 17 km (Fig. 8.7a). This is the theoret-
Sedimentary basins can be assumed to be in isostatic ical maximum thickness of a sedimentary sequence
equilibrium in relation to the Earth’s crust. However, overlying oceanic crust. Sediment basins on the conti-
the crust has a certain rigidity, and it takes time before nental crust have a sedimentary thickness which is a
equilibrium is attained after loading. Studies of uplift function of the thickness of the crust and the density of
curves, for example from Scandinavia, show that dur- the sediments and the basement. The thinner the con-
ing the course of 10,000 years the crust has undergone tinental crust beneath a sedimentary basin, the more
major adjustments in the form of uplift to compensate sediments can accumulate while maintaining isostatic
for the unloading of ice after the last glaciation. This is equilibrium (Fig. 8.8). Along continental margins sed-
geologically a very fast response and we can assume imentary basins have formed on thin (stretched) conti-
that major sedimentary basins are in approximate iso- nental and oceanic crust.
static equilibrium with regard to most geological pro- The formation of sedimentary basins clearly
cesses. Nevertheless, even the filling of water requires that the density of the rocks below the basin
reservoirs for hydroelectric plants causes local is greater than that of the rocks which surround it. The
subsidence. sediment and water which fill the basin are lighter and
264 K. Bjørlykke
a 2
30 km
Thickness of continental crust

sea level 0
25 km
Ocean depth without sediment

20 km
–2
–4 10 km
loading (km)
Oceanic
crust
–6
0 20 40 60 80 100 mill.
Post-rift age
c
SEA LEVEL
SPREADING RIDGE
CARBONATE
COMPENSATION MANGANESE
SILICA TURBIDITES
TURBIDITES NODULES
F LOO R BA
SILICA
N SA
EA LT
OC
Fig. 8.7 (a) Subsidence of the sea cooling in the bed and conti- High spreading rates are therefore associated with transgressions.
nental crust of various thicknesses as a function of post-rifting age (c) Schematic presentation of an Atlantic type spreading ridge. The
(after Kinsman 1975). Basaltic spreading ridges (oceanic crust) are shallow part may be above the carbonate compensation depth
often at 2 km depth, then the seafloor sinks to almost 6 km depth while the deeper part will be without carbonate sediments and
after 100 million years. (b) The rate of seafloor spreading control contain little clastic sediments, being far away from land. The sea
the average water depth which is a function of cooling of the floor with then be sediment-starved and concentrate silica from the
oceanic crust. The basalt shrinks and becomes heavier with time. volcanism and manganese nodules
Average density of sediments (g/cm3) be due to a transgression increasing the weight of

1.9 2.1 2.3 2.5 2.7 the water column. Sediment loading occurs with
0
basin infilling.
30 km 4. Subsidence along subduction zones. This results in
lower geothermal gradients in the downward-
25 km
deflected crust, so that the density also increases.
5
5. Tectonic loading. Thrusting of tectonic plates leads
to increased loading on the part of the Earth’s crust
Maximum sediment thickness (km)
20 km
Thickness of continental crust

in question and we have subsidence, especially in
front of nappes (foreland basins).
10
15 km
8.8.2 Changes in Sea Level and
15 Sedimentation and Isostatic
10 km Compensation
Variations in sea level due to eustatic transgressions or

20 tectonic subsidence will represent loading or
Oceanic crust unloading of the crust which will reinforce the primary
change in sea level. If the sea rises 100 m, for example
due to ice sheet melting, this will increase the isostatic
25 loading on the seafloor. We can calculate that there
will be a further 43.5 m of subsidence, so that the total
Fig. 8.8 Potential loading capacity to isostatic equilibrium on
oceanic crust and of various thicknesses of continental crust as a
increase in the depth of water at equilibrium will be
function of sediment density (after Kinsman 1975). If the aver- 143.5 m.
age sediment density is 2.5 g/cm3, the maximum thickness of If a sedimentary basin with this depth of water is
sediments can be 17 km on top of the oceanic crust. If the filled with sediment, there will be further isostatic
sediments have lower density (higher porosity) the sediment
thickness must be lower to main isostatic balance with respect
subsidence because of the weight of sediments,
to the crust and the sediments down to about 100 km depth. If providing accommodation for deposition totalling
there is 20 km of continental crust and the sediment density is 250–300 m depending on the density of the sediments.
2.3 g/cm3, there is room for 4 km of sediments A 100 m rise in sea level will thus lead to deposition of
almost 300 m of sediment. In the same way, primary
have to be compensated for with heavier rocks below tectonic subsidence due to cooling of the ocean floor
the basin. We can assume a depth of compensation of will lead to further subsidence due to increased water
about 100 km. This means that the weight of a column and sediment loading.
of rock plus any water present down to a depth of Using stratigraphic data in the form of measured
100 km must be the same everywhere. Subsidence of profiles or oil wells as our starting point, we can
a sedimentary basin may be due to several crustal calculate backwards to the primary tectonic or eustatic
processes: changes in sea level. This method is called
1. Stretching and thinning of the continental crust. “backstripping”.
Heavier mantle rocks then make up a greater per-
centage of the rock column down to a compensa- Z ¼ Yððρm ρs Þ=tðρm ρw ÞÞΔHρw =ðρm ρw Þ
tion depth of about 100 km. þ ðH ΔHÞ
2. Cooling, i.e. lower geothermal gradients in the
crust, lead to contraction and a higher rock density where Z is the primary tectonic subsidence, F is a
(thermal contraction), and this will result in factor which is a function of the rigidity of the Earth’s
subsidence. crust, ΔH is the change in sea level and H is the water
3. Increased loading by water, sediments or other depth (Watts 1983). Y is the sediment thickness
rocks will cause subsidence. Water loading could compensated for compaction, i.e. the sediment
266 K. Bjørlykke
thickness (and density ρ), shortly after deposition sedimentary basin because the continental crust is
(Stickler and Watts 1978b, Bally et al. 1981). If we thin and because heavy rocks from the mantle push
know the variations in sea level and the depth of the their way up, thus increasing the average density of the
water from environmental interpretations, for exam- rocks. Those parts of the rift valley system which have
ple, these can be inserted into the equation so that the much volcanism will have less space for the
primary tectonic movement can be worked out. sediments. Along the margins of a rift system, where
The backstripping technique which is used to recon- the continental crust is not stretched, uplift occurs due
struct the subsidence history of different parts of a to the higher geothermal gradient. This causes the
sedimentary basin lends itself very well to computer basement rock at the surface to slope away from the
modelling. The data obtained on the depth and tempera- rift valley as the geothermal gradients are reduced.
ture history of the source rock in particular has allowed This will to some extent be compensated for by the
much better assessments of kerogen maturity and the formation of erosional valleys which cut backwards
times of oil expulsion and migration. One parameter into the raised shoulders on the sides of the rift valley.
which is crucial but difficult to estimate is the variation Because of the high relief around these basins and
of the geothermal gradient as a function of geological the short transport distance for the sediment eroded
time. It is also often difficult to estimate the palaeodepth from the bedrock, these basins will be characterised by
during the deposition of different sedimentary mineralogically very immature sediments, largely
formations. Whether a formation was deposited at a arkoses and conglomerates deposited in fan deltas
depth of 200 or 1,000 m will make a very significant along the active faults. In the central and deeper
difference, and it is often difficult to make accurate parts of the basins we find deposition of finer-grained
palaeoenvironmental estimates of the palaeobathymetry. sandstones and clayey sediments. Rift valley basins
To form a sedimentary basin with a thick infill of may be continental lacustrine basins as in East Africa,
sediments requires a crustal depression large enough or marine as those offshore East Africa and the Juras-
to provide the accommodation space for the deposits. sic basins of the North Sea. Horsts, which are
This may be a result of large-scale crustal movements unstretched (thick) pieces of continent crust, may
like seafloor spreading and crustal thinning (exten- become topgraphically very high due to high heat
sion). The supply of sediments is also critical and flow. The Ruwenzori Mountains of East Africa,
clastic sediments have to be supplied from adjacent reaching more than 5,000 m, are one example. Both
land areas which are being uplifted and eroded. marine and lacustrine rift basins will tend to have
Chemical and biogenic sediments are formed reducing conditions in the deeper part due to limited
locally from seawater and thus are not reliant on sedi- circulation of oxygenated water.
ment supply from land areas. This typically applies to Rift basins formed in areas with wet climates will
limestones, which accumulate where there is little be occupied by large lakes. Lacustrine basins often
clastic sedimentation. Sedimentary basins that are have an even better potential for producing source
more or less isolated from the sea in arid regions can rocks than marine rift basins, because water stratifica-
be filled up with evaporites at a rather high sedimentation (density stratification) is usually more marked in
tion rate. lakes. We will therefore often find black, organic-rich
shales in these basins.
Rift basins formed in arid zones are characterised
8.9 Continental Rifting by evaporite deposits. Block faulting will readily lead
to isolated basins or horsts which cut off contact with
Stretching and thinning of the continental crust takes the open sea. Evaporites are typical of rift deposits
place by tensional tectonics in connection with rifting. today, e.g. in East Africa, and were widespread in
Crustal stretching may be a result of tensional forces Europe and North America during the Permo-Triassic,
and uplift due to the high geothermal gradients before the ocean-floor spreading which created the
associated with rifting. This leads to a thinning of the Atlantic Ocean started in the Mid-Jurassic. The
continental crust, causing isostatic subsidence. Zechstein salt deposits in Germany and the North
Fracture zones in the continental crust (rift valleys) Sea are typical examples. Jurassic and early Creta-
produce subsidence which often is located in a ceous rifting during the early phase of the opening of
the South Atlantic Ocean, evaporite basins were oceanic and continental crust. With cooling, the rigid-
located in the arid regions of the time. They can ity of the crust increases, so that the bending of the
form almost perfect cap rocks that are not likely to continental crust near the continental margin broadens
leak. If they are thick enough they may form salt and there is overall subsidence of the continental shelf,
domes which produce structural traps in overlying and in consequence transgression and onlap. It will
rocks, as at Ekofisk and in the Gulf Coast basin. Rising subside isostatically and make accommodation space.
salt domes also strongly influence the clastic sediment During high sea level the clastic supply is pushed back
distribution in a basin. onto the continent and during low stand a
Salt has high conductivity causing the temperatures progradational sequence is formed (Fig. 8.10).
above the salt to be higher than normal and the In other places the drainage discharge from land
sediments below the salt to be cooler than normal. focuses the sediment in large delta areas. This is the
This must be taken into account when modelling the situation with the Mississippi delta in the Mexican
maturation of source rocks associated with the salt. Gulf, which has been receiving sediment from large
These temperature anomalies may also be important expanses of the North American continent since Meso-
when modelling diagenesis and reservoir properties. zoic times.
Large subsalt discoveries have been made in recent The sediment supply is controlled by the drainage
years both in the Gulf Coast and offshore Brazil. system on the continents, which may follow old rift
The subsidence due to rifting renders the adjacent systems because most of the sediment is produced on
rocks unstable and promotes gravitational sliding of the continents and has plenty of space to be deposited
blocks in the crust, in towards the rift structure. There along the margin of the deep ocean.
is a tendency for listric faults to form, i.e. parallel, The Niger delta represents a similar focusing of
curved fault planes which start as normal faults and sedimentation on the African side of the Atlantic.
curve round with depth until they are almost horizon- Sediment basins along passive margins are deposited
tal. The blocks then become rotated so that they tilt above old rift basins which have cooled. If basin
over and slope away from the rift. During the initial subsidence is rapid, this will also contribute to a low
part of the spreading phase, basins with limited circu- geothermal gradient because the sediments have to be
lation will be formed so that evaporites and carbonates heated during burial (20–30 C/km). A relatively great
are often deposited. Upper Jurassic and Lower Creta- thickness of sediment (4–5 km) must therefore be
ceous deposits of this sort are found extensively along deposited for the underlying source rocks to achieve
the margins of the Atlantic Ocean. maturation. That is to say, attain 100–150 C,
depending on the subsidence rate and hence the time
available for the heating. This kind of sedimentary
8.10 Subsidence Along Passive Margins basin situated along a passive margin is called a
“cold basin”.
Passive margins develop from a rift phase to a spread- Foreland basins form in front of mountain chains
ing phase. The transition between continental crust when they are uplifted and eroded. The advancing
and oceanic crust therefore consists of a thinned con- nappes help to depress the continental crust and pro-
tinental crust with listric faults and horsts formed vide accommodation space for the sediments shed
during the rifting phase (Fig. 8.9). As ocean-floor from the mountains. Foreland basins tend to be broad
spreading progresses, the geothermal gradients in this and gently folded into giant structures in the distal
part of the continental shelf will decline, resulting in parts. The Middle East is a large foreland basin in
cooling and thermal subsidence of the continental front of the alpine mountain chain running along Iran
margins. The oceanic crust will also experience ther- and Turkey. The Persian Gulf is a continuation of this
mal subsidence, as a function of the age of the seafloor. basin collecting sediments coming from the mountains
Subsidence flexure will develop where the subsidence of Iran. The Jurassic and Cretaceous sequences in Iraq,
is most rapid, on the outer parts of the continental shelf Saudi Arabia, Kuwait and The Emirates contain some
and slope nearest the oceanic crust. Further in from the of the largest oil fields in the world.
passive margin the subsidence rate will be slower. On the east side of the Rocky Mountains we have
Eventually sedimentation fills the prism between the similar foreland basins from the Denver Basin all the
268 K. Bjørlykke
Passive margin
Mountain chain
Hinge zone
ρ = 1.03 g/cm3
3
Syn ρ = 1.7–2.6 g/cm Post-rift
-rif
t
ρ = 2.9 g/cm3
ρ = 2.8 g/cm3
Mantle ρ = 3.3 g/cm3
Overthickened Continental Thinned Oceanic crust

continental crust 35–80 km crust 30 km continental crust
Fig. 8.9 Simplified cross-section of a passive continental mar- along the passive margin. The thinner the crust is, the more
gin. The sediments that were deposited during the initial phase sediments can accumulate and the maximum thickness is
of rifting lie beneath the younger sequence which was deposited reached when the progradation reaches cold oceanic crust
way to Alberta, Canada, where there are very large pronounced by sediment loading and thermal subsi-
deposits of heavy oil and tar sand. The heavy oil in dence. When two plates move parallel to one another,
Venezuela (and Columbia) is located in a similar tec- we have strike-slip faults of the San Andreas type in
tonic position in front of the Andes. California, but there is no new plate formation, nor any
subduction. We distinguish between “right-lateral” or
“dextral” faults, where the opposite side of the fault
8.11 Strike-Slip Faults and Pull-Apart has moved to the right, and “left-lateral” or “sinistral”
Basins faults. The San Andreas is a dextral fault, and the
western part of California (Salina block) has moved
We find “strike slip” faults in both the oceanic and the northward in relation to the North American continent.
continental crust. Transform faults in the oceanic crust If the fault plane follows a completely straight struc-
result in ridges of younger basaltic material which can ture in the rock parallel with the direction of move-
help to limit the extent of sedimentary basins, particu- ment there will be neither tension nor compression
larly in the early phases of the opening (see Chap. 6). along the fault plane. However, faults usually follow
Strike-slip faults in the continental crust can lead to older structures in the basement rocks which may be
both compression, i.e. thickening, forming small curved and strike slip movements may then produce
mountains, and stretching of the continental crust both compression and tension along the fault plane.
forming deep sedimentary basins. Calculations show Compression will lead to folding and the formation of
that relatively modest stretching of the continental small mountain ranges, while tension will lead to the
crust will cause considerable thinning and subsidence. opening of deep sedimentary basins.
Bends in the strike-slip fault system like the San We find the most typical examples of this in
Andreas Fault open up deep holes in the continental California. When a fault branches we may also see
crust which can be filled in with very thick sequences both compression and tension, with elevation and sub-
of sediments like in the Ventura and Los Angeles sidence respectively of blocks, depending on their
Basins. The sedimentation rate may be very high orientation. If the fault shifts to another parallel fault
because of the large uplifted land areas and the rela- plane, we have crustal tension in the area in between,
tively small basins. and often also basalt flows. A so-called “pull-apart”
Here, too, relief which has developed through basin forms, which is a “hole” in the continental crust
stretching of the continental crust will be made more formed by the strike slip movement. The Salten
Highstand Sea level
Shelf
sand/carbonate Sea level
Submarine
canyons
Offlap Onlap
Sand-rich turbidites and

traction current
deposited sands
Lowstand Sea level
Sea level
Slope prodelta
Muds and turbitites
Progradational onlap
sequence
Fig. 8.10 A modal for transgression with sediment starvation and renewed progradation across the shelf onlapping (downlapping)
onto an unconformity
Trough in Southern California is a good example and young sediments (<4–5 million years) which have
much of the Los Angeles and Ventura basins been folded and which contain large amounts of oil.
originated in this way during the Pliocene. Basins of this type provide almost ideal conditions
The basins are very deep holes in the continental for oil accumulation and are amongst the richest oil
crust, and are sometimes floored with volcanic rocks. regions in the USA. Very few basins can compete with
Because of the limited size of the basins and the the California strike-slip related basins in terms of
continued uplift of the areas surrounding them, we barrels of oil per square kilometre. This is because
get a very high rate of sedimentation. In the these deep basins often have reducing conditions in
Miocene-Pliocene basins in California, 5–8 km or their bottom water due to limited circulation of
more of sediments have been deposited in roughly as oxidised water. We therefore get organic-rich
many million years. This corresponds to an average of sediments deposited. Turbidite sandstones, particu-
1 mm/year, which is a very high figure in this context. larly the more proximal parts, often have sufficiently
The sediments consist largely of proximal turbidites high porosity to become reservoir rocks. Even if the
along the edge of the basin and distal turbidites and porosity is not especially high, turbidite deposits often
fine-grained clay sediments in the more central, deeper form thick sequences that extend over large areas. The
parts. In the Ventura Basin which continues into the reservoir quality can also be controlled by turbidite
Santa Barbara Channel, there are about 15 km of lobes or channels with levees.
270 K. Bjørlykke
Sediments deposited in basins of this type tend to • Back-arc basin (sedimentary basin on the oceanic
undergo folding, which will result in good structural crust).
traps. Geothermal gradients will be relatively high, • Retro-arc basin (formed if continental blocks are
and allow thorough maturation. upthrust and faulted, creating a secondary basin in
the continental crust).
Trench basins seldom contain large amounts of
8.12 Converging Plate Boundaries sediment if separated from the continent by an island
arc preventing sediment supply from the continents.
When plates collide (converge), there can be One of the reasons why they are the deepest oceanic
boundaries between: areas is because they do not fill up with sediment. This
• Oceanic crust and continental crust, such as along is because the ocean floor beneath trench basins is part
the west side of South America (Andes type). of a dynamic system and is being consumed through
• Continental crust against continental crust (Alpine subduction as sedimentation occurs. The downward
or Himalayan type mountain chain). movement of the oceanic crust helps to maintain a
• Oceanic crust against oceanic crust, with formation low geothermal gradient in the trench since heat
of volcanic islands. must flow up through rocks that are descending.
The oceanic plate has a higher density than the Because of the continual lateral displacement of the
continental crust, so in a subduction zone the oceanic rock floor under the trenches, the sediment thickness
plate will be carried downwards under the continental will be limited. The island arcs are the main source of
one. The oceanic plate is relatively cold and its geo- sediment supply to the trench basins.
thermal gradient is low because since it is descending, Most of the sediment coming from the continents will
the upward heat flux is reduced as a function of the rate be trapped in back-arc basins. The sediments in trench
of subduction. Where there is active subduction, ocean basins consist of deep-water conglomerates, turbidites
trenches are formed, such as the 10–11 km deep ones and pelagic sediments. Trench basins lie mostly below
outside the Philippines and Japan. This is because the the carbonate compensation depth and therefore contain
underlying rocks are cold and therefore dense. little carbonate and are also usually starved with respect
In subduction zones the downward-moving part of to clastic sediment supply. Sedimentary transport of very
the oceanic crust is characterised by extremely low fine-grained pelagic sediments tends to occur along the
geothermal gradients. This is because the heat flow axis of the trench. The inner slope up to the island arcs is
must move counter to the direction of movement of the steeper than the outer one, and the subduction zone
plate undergoing subduction which is therefore colder (Benioff zone) is at the foot of this slope (Fig. 2.48).
and denser than the average old ocean floor crust. The When the oceanic plate descends into the Benioff
subduction zones form the deepest parts of the ocean zone, sediments deposited on the sea bed may be
(up to 10–11 km deep) where there is not a great deal scraped off, forming wedges of sediment. This is
of sedimentation. Ordinary, cold, 100–150 million called an accretionary prism and consists of a compli-
year old ocean floor is in isostatic equilibrium at a cated pattern of folded and brecciated sedimentary
depth of 5–7 km without sediment loading. rocks exhibiting overthrusts and underthrusts (Fig. 2.
Where we have subduction of oceanic crust beneath 49). The sedimentary sequence is the right way up
continents we find the following types of basin: within each thrusted unit, but the sequence of thrusted
• Trench basin (oceanic sedimentation within the units is inverted. The youngest sediment wedges are at
trench). the bottom because they are the last sediments to have
• Accretionary prism. been scraped up from the seafloor.
• Fore-arc basin (relatively shallow basin in front of On the inside of the trench, slope basins are formed
island arcs). with sediment supplied by the island arcs. These
• Inter-arc basin (sedimentary basin between island deposits are folded, deformed and reworked by the
arcs of continental crust). subduction processes.
Although sediments in trench basins may be rich in very strong meteoric water drive which flushed out
organic material, their oil potential is poor due to a many of the oil fields.
lack of reservoir rocks and the intense tectonic defor-
mation. In addition the geothermal gradient will also
be lower than in other basins, at least at the time of
8.13 Conclusions
deposition.
Stratigraphy is a response to changes (movements) in
Fore-arc basins form between the actual island arcs
the Earth’s crust and sediment supply. Changes in the
with volcanoes, and the subduction trench. They are
environments and the biological production of
not subjected to the intense tectonic deformation
sediments are also important. The accommodation
which occurs in trench basins. Fore-arc basins often
space for sediments is also controlled by the rate of
transgress onto the island arcs as the belt of volcanic
compaction of the underlying sediments.
activity gradually withdraws towards the continent.
Geophysical methods, mainly seismics, are used to
These basins will contain fluvial and deltaic sediments
map out sedimentary basins on a regional scale and also
closest to the island arcs, then shallow marine conti-
on a small scale for detailed exploration and production.
nental shelf sediments. In the outer parts sediments are
It is also important to reconstruct the evolution of sedi-
deposited in deeper water towards the oceanic slope.
mentary basins though time and the changes with
There is good potential here for organic-rich
respect to the sediment supply and the environment.
sediments to accumulate, but the geothermal gradient
This must to a large extent be based on reconstruc-
may be low despite the proximity to a volcanic island
tion of plate tectonic movements and basin modelling.
chain, and maturation consequently slow.
Large parts of California in front of the Rocky
Mountains (Sierra Nevada) are a fore-arc basin (San Further Reading
Joaquin Basin) and there are numerous oil fields here,
not least in the area around Bakersfield.
Intra-arc basins are a result of tensional tectonics in Berg, O.R. and Woolverton, D.G. 1985. Seismic Stratigraphy II.
An Integrated Approach to Hydrocarbon Exploration.
the island arcs within an area otherwise characterised AAPG Memoir 39, 276 pp.
by convergent plate movement. Here a system of fault- Brown, A.R. 1996. Interpretation of Three-dimensional Seismic
controlled basins (graben) is formed. Sediments Data. AAPG Memoir 42, 424 pp.
deposited in these basins will for the most part be Catuneanu, O. 2006. Principles of Sequence Stratigarphy.
Elsevier, 375 pp.
immature, and volcanic material will be common. Emery, D. and Myers, K. 2009. Sequence Stratigraphy. Wiley,
Volcanic sandstones will often lose their primary 304 pp.
porosity rapidly due to unstable minerals and diage- Frish, W., Meschede, M., and Blakey, R.C. 2010. Plate Tecton-
netic transformation. The geothermal gradient is high, ics: Continental Drift and Mountain Building, Springer
Verlag, 212 pp.
but conditions for formation of source and reservoir Sclater, J.G. and Christie, P.A.F. 1980. Continental stretching:
rocks are not very good. an explanation of the post-mid Cretaceous subsidence of the
Intra-arc basin sediments will frequently be Central North Sea Basin. Journal of Geophysical Research,
subjected to intense tectonic deformation which may 85, 3711–3739 [Appendix A, 3730–3735].
Tearpock, D.J. and Bischke, R.E. 1991. Applied Subsurface
deform potential reservoirs. Geological Mapping. Prentice Hall, Englewood Cliffs, NJ,
Back-arc basins develop on the oceanic crust due to 648 pp.
seafloor spreading behind the island arcs. There is Vail, P.R. et al. 1977. Seismic stratigraphy and global changes
likely to be an abundant supply of sediment from the of sea level. I. In: Payton, C.E. (ed.), Seismic Stratigraphy –
Applications to Hydrocarbon Exploration. AAPG Memoir
continent, and back-arc basins may also fill with del- 26, pp. 49–212.
taic and shallow marine sediments. Because of later VanWagoner, J.C. et al. 1990. Siliciclastic Sequence Stratigra-
tectonic deformation, back-arc sedimentary basins are phy in Well Logs, Cores and Outcrops. AAPG Methods in
not the most promising oil prospects. Exploration Series 7, 55 pp.
Payton, C. 1977. Seismic Stratigraphy – Applications to Hydro-
The Jurassic and Cretaceous seaway through North carbon Exploration. AAPG Memoir 26, 516 pp.
America east of the Rocky Mountains was a back-arc Suppe, J. 1985. Principles of Structural Geology. Prentice-Hall,
basin during subduction of the Pacific plate. The area Inc., Englewood Cliffs, NJ, Chapter 1, pp. 17–24.
is a good oil region but Tertiary uplift resulted in a http://strata.geo.SC.edu
Chapter 9
Heat Transport in Sedimentary Basins
Knut Bjørlykke
The temperature increases downwards in the crust and The heat flow (Q) is most commonly expressed by
there is therefore a transport of heat upwards, referred W/m2 but can also be expressed as cal/cm2s or joule/
to as the heat flow. Most of the flow is by conduction cm2s. A heat flow of 70 mW/m2 corresponds to
(thermal diffusion). Flow of porewater will also trans- 1.4 μcal/cm2s. This is the heat flow unit (mW/m2)
port heat in the subsurface but the flow rates in sedi- which may be referred to as HFU. The temperature
mentary basins are normally so small that we can of a volume of rock is a function of the heat flux and
ignore the contribution from fluid flow. Around igne- the conductivity of rocks and fluids. The increase in
ous intrusions there is usually thermal convection with temperature with depth (temperature gradient) is
high flow rates and heat transport. In shallow areas called the geothermal gradient (dT/dz). Typical geo-
with high flow rates of meteoric water, advective heat thermal gradients may also be written as T and in
transport is also significant. sedimentary basins are usually 25–45 C/km.
The source of the heat is mainly radioactive pro-
cesses which are particularly important in the conti- kg m2
nental crust due to the enrichment of uranium, thorium Joule ¼
s2
and potassium in granitic rocks. 1 Joule ¼ 0:239 cal
Heat is transported through sedimentary basins 1 Watt ¼ 1 joule=s
mostly by conduction following the heat flow
equation:
In a sedimentary basin there is the background heat
flux from the underlying basement, and granitic rocks
Q ¼ c dT =d z (9.1) have higher heat production and temperatures than
basic rocks. The sedimentary sequences overlying
The thermal conductivity (c) is expressed as the basement also produce heat by radioactive reaction
Wm 1 C 1 or Wm 1 K 1 which is the heat (W) and black shales with a high content of organic matter
transported over a given distance (m) with a certain often have a relatively high uranium content. This
drop in temperature ( C). Temperature is expressed as additional heat source may be significant in terms of
Celsius ( C) or Kelvin (K) but Fahrenheit was and still increasing the heat flux and the geothermal gradients
is commonly used in the USA. Conductivity may also in sedimentary basins.
be expressed in terms of calories (cal), which is an The temperature distribution (geothermal
alternative unit for energy/heat. 1 W equals 1 J/s or gradients) in sedimentary basins can vary regionally
0.239 cal/s. and over geologic time. This determines both the
generation and expulsion of petroleum, and it also
strongly influences the reservoir quality. It is therefore
K. Bjørlykke (*) important to understand the processes that control heat
Department of Geosciences, University of Oslo, Oslo, Norway transport in sedimentary basins.
274 K. Bjørlykke
The temperature gradient rT ðdT=dzÞ quartz to 1.8 for illite and smectite. The conductivity
(9.2) is therefore to a large extent a function of the quartz
¼ Heat flowðQÞ=Conductivity(cÞ:
content and the water content (porosity).
The conductivity of shales from the North Sea
In rocks with low conductivity (like mudstones and
shales) the geothermal gradients will be high and in ranges from about 0.8 to 1:1 Wm 1 C 1 (Midttømme
highly conductive rocks (like salt) the geothermal et al. 1997) so they are not very much more conductive
gradients will be low (Fig. 9.1). The thermal conduc- than water. The conductivity parallel to bedding may
be up to 70% higher than perpendicular to bedding.
tivity of salt (halite and anydrite) is 5:5 Wm 1 C 1 ,
Most of the heat transport is vertical except around
while shales may have conductivities between 1.0 and
hydrothermal or igneous intrusions, but in the case of
2:5 Wm 1 C 1 . Sandstones and limestones have
steeply dipping beds the conductivity would be higher.
values between shales and salt.
Over a limited vertical interval of the sedimentary
In a situation where a rock is filled with stationary
section the heat flow may be relatively constant and
porewater the total heat flux (Q) is the sum of the heat
we see from Eq. (9.2) that the geothermal gradient is
conducted through both the rock’s matrix and pores
inversely related to the conductivity.
(porosity φ filled with fluids):
When there are rocks with low conductivity near
the surface the geothermal gradient will be higher so
Q ¼ C r ð1 φÞ þ φCf (9.3)
that the underlying sediments will be warmer. This is
called a blanketing effect. Mudstones with low ther-
Cr is the conductivity of the solid rock and Cf is the
mal conductivity on top of granites or older sedimen-
conductivity of the fluids (usually water) in the pore
tary rocks will have this effect.
space. Water (fresh) at 20 C has a conductivity of
Salt with high conductivity has the opposite effect.
0.6 W/mK while seawater and saline brines are much
Because the temperature gradient through salt is low,
more conductive. The conductivity of common sedi-
the temperature will be relatively high at the top of the
mentary minerals ranges from 7:7 Wm 1 C 1 for salt and low at the bottom. This has consequences for
maturation of source rocks. This is a very important
Heat f low
effect for petroleum prospects below thick salt layers,
F Thermal conductivity C Temperature T i.e. in the Gulf of Mexico, offshore Brazil, West Africa
and the North Sea. The temperatures below the salt
Sandstone
will be significantly lower than normal at this depth.

This means that the reservoir quality of sandstones
reservoirs will be better due to less quartz cement.
Lower temperatures will also preserve more petroleum
as oil or condensate as there will be less cracking to
Mudstone
gas.
The heat flux is only constant in an equilibrium
situation. When sediments subside they are heated
and a part of the background heat flux is used to heat
Evaporites
the subsiding rocks (Fig. 9.2). This is equal to the heat

capacity of the rocks and the subsidence rate. In basins
with high sedimentation rates the heat flow is strongly
reduced and in the Plio-Pleistocene depocentres the
geothermal gradients are down to 20–25 C/km
Geothermal gradient dT/dZ = Q/C
Z (Harrison and Summa 1991).
During subsidence and sedimentation the
Fig. 9.1 Relationship between heat flow (Q), conductivity (C) sediments must be heated, and this heat is taken from
and geothermal gradients. The geothermal gradients have an
inverse relation with the conductivity for the same heat flux the background heat flow and the geothermal gradient
(F). Thick salt layers or domes cause higher geothermal is reduced.
gradients above the salt and low temperatures below the salt
9 Heat Transport in Sedimentary Basins 275
Temperature
Fast deposited
+ ΔZ new sediment layer (Δ Z )
Temperature T
Fast erosion of a sedimentary
– ΔZ layer with thickness Δ Z
Low geothermal gradient Isotherms -

due to high sedimentation rate (+Δ Z ) Salt equal
High geothermal gradient temperatures
due to fast uplift erosion (–Δ Z )
Geothermal
gradients
through
Geothermal gradient salt
dT/dZ Depth
Compared with
an average
geothermal
gradient
Z Heat flow
from deeper Fig. 9.3 Geothermal gradients are strongly influenced by
Depth
rocks (mW/m2) layers of salt or salt domes. Since the heat flow is relatively
constant the geothermal gradient must be low through the highly
Fig. 9.2 Geothermal gradients as a function of rapid uplift conductive salt. As a result the overlying sediments will be
(erosion) or subsidence (sedimentation). During subsidence warmer than normal while the underlying sediments will be
some of the heat flow is used to heat the subsiding sediments cooler
and underlying basement and this will reduce the geothermal
gradients, forming cold basins. During uplift the heat from
as in the water phase. During the first period of advec-
cooling rocks will add to the heat flux, producing steeper geo-
thermal gradients tive flow along a fault or through permeable sandstone
beds a high percentage of the advected heat will be lost
by conduction to the mineral matrix.
During subsidence the heat flux is:
Qs ¼ Qb Fs CðhcÞ dT=dZ 9.1 Heat Transport by Fluid Flow

The heat capacity of rock including porewater is When fluid, usually water, is transported in a sedimen-

CðhcÞ . tary basin there is also heat transport unless the trans-
Here Qb is the background heat flux from the base- port is parallel to the isotherm (Fig. 9.3).
ment, Fs is the subsidence rate (downward flux of The advective heat transport Qt is proportional to
rocks) and Chc is the heat capacity of the rocks. the flux of water (Darcy velocity F ¼ m3 =m2 =s), the
When rocks are uplifted and cooled the heat given heat capacity of water (Chw), and the geothermal gra-
off from the cooling rocks adds to the background dient (T).
heat flow:
During uplift the heat flux becomes: Qt ¼ Fh Chw rT sin α
Qs ¼ Qb þ Fur Ch dT=dZ Here α is the angle between the direction of fluid

flow and the isotherms which are lines with equal
Here Fur is the rate of uplift. The heat capacity of temperature. h is the length along the direction of
the mineral matrix may be estimated at about fluid flow.
8–900 J/kgK. The heat capacity of water (Chw) ¼ During compaction-driven flow the flow rates are in
4,200 J/kgK. most cases too small for this heat transport to be
In terms of volume the heat capacity of rocks is significant. Focused compaction-driven flow may
however close to 2,500 J/dm3K. cause a significant heat flow by advection, but only if
This means that at 20% porosity about 2.5 times as the rate of porewater flow is very high. The average
much heat is stored in the mineral matrix (density 2.7) water flow upwards relative to the sediments is very
276 K. Bjørlykke
low, but relative to the seafloor the porewater is in producing transient thermal heat flows (Hermanrud
most cases sinking. et al. 1991).
Meteoric water fluxes along aquifers into sedimen- In the Mississippi Valley, USA, compaction-driven
tary basins are many orders of magnitude faster than in flow has been shown to be quite insufficient to generate
compaction-driven flow; in some cases the downwards hot fluids capable of precipitating ores (Bethke 1986).
flow of cool meteoric water from mountains into sedi- Compaction-driven flow from thrust belts may produce
mentary basins may cause a significant reduction in significant thermal perturbations on a relatively local
the geothermal gradients. scale, but modelling suggests that such flow is insuffi-
cient to cause large-scale thermal anomalies in the
adjacent foreland (Deming et al. 1990). In continental
9.2 Heat Transported by Conduction rifts like the Rhine Graben, where the rift margins are
and by Fluid Flow (Advection) exposed and elevated topographically, groundwater
flow can to a large extent explain the observed thermal
The relative contribution from these types of heat anomalies (Person and Garven 1992).
transport can be expressed by the Peclet number (Pe):
Pe ¼ ρf Cf Qz L=Cr ð1 φÞ þ φCf 9.3 Importance of Heat Flow and

Geothermal Gradients
Here ρf is the fluid density, Cf the heat capacity of
the fluid, Qz the vertical component of the Darcy Heat flow is a very important parameter, which
velocity, L the length of the flow path, and Cr and Cf strongly influences geothermal gradients and rates of
the respective thermal conductivities of the solid petroleum generation. It also strongly influences rates
phases (minerals) and the fluids (water) (Person and of quartz cementation and other types of chemical
Garven 1992). Sedimentary sequences with permeable compaction in siliceous sediments.
sandstones and limestones normally include low per- Heat flow and geothermal gradients are also impor-
meability shales and siltstones The distance between tant for the utilisation of geothermal energy and heat
them tends to control the length of the flow path and pumps in the ground or in rocks.
the height of the convection cells. In thick permeable In sedimentary basins we have a heat flow from the
sandstones and limestones the vertical flow and trans- basement into the overlying sedimentary sequence. The
port of heat is faster, resulting in lower geothermal composition of the basement rock determines the rate.
gradients. Bjørlykke et al. 1988. See Chap. 10. Granitic rocks with high potassium and uranium con-
The conductivity of water depends on temperature tent will produce more heat than rocks like anorthosites
and salinity but is much lower than that of the mineral and gabbros which are very low in potassium. Offshore
matrix (0.6 W/m C and 2.5–3.5 W/m C, respectively). Norway the background heat flow varies significantly
Hot porewater therefore rapidly loses its heat to the depending on the basement rocks. Organic-rich shales
mineral matrix. Convection is driven by the primary like the Upper Jurassic source rocks from the North Sea
temperature gradients. Porewater convection does basin may also contribute significant heat because of
change the temperature field but temperature the radioactivity (high uranium content).
perturbations due to this flow are not very large As we have seen above, increasing sedimentation
(Ludvigsen 1992). rates will reduce the geothermal gradient because
Numerical calculations of fluid flow in modern some of the heat flux is used to heat new layers of
sedimentary basins like the Gulf of Mexico basin subsiding sediments. Cold basins with rapid subsi-
show that compaction-driven porewater flow is insig- dence and low geothermal gradients (<20–25 C/km)
nificant in terms of advective transport of heat require deep burial of the source rocks before they can
(Harrison and Summa 1991). In the North Sea basin, generate petroleum, both because of low temperature
too, the geothermal gradients only vary within rather and the short geologic time (<2–3 million years) for
narrow limits (35–40 C/km). The occurrence of petroleum generation. The small time/temperature
locally higher values offshore Western Norway has integral will also result in little quartz cement in reser-
been attributed to the effect of recent glacial erosion voir sandstones. The amount of quartz cement can also
9 Heat Transport in Sedimentary Basins 277
be used as a measure of the time/temperature index. Bethke, C.M., Deming, D., Nunn, J.A. and Evans, D.G. 1990.
The temperature history is an important parameter in Thermal effects of compaction-driven ground water flow
from overthrust belts. Journal of Geophysical Research 95,
basin modelling because it influences the sediment 6669–6683.
density and therefore the rate of subsidence and the Bjørlykke, K., Mo, A. and Palm, E. 1988. Modelling of thermal
generation of hydrocarbons, and also the reservoir convection in sedimentary basins and its relevance to
quality. The sediment density is increased by porosity diagentic reactions. Marine and Petroleum Geology 5,
338–351.
reduction during diagenesis and mineral dehydration. Demongodin, L., Pinoteau, B., Vasseur, G. and Gable, R. 1991.
Heating however causes a slight expansion of the Thermal conductivity and well logs: A case study in
minerals and density reduction. the Paris Basin. Geophysical Journal International 105,
The heat flow and the geothermal gradients may 675–691.
Evans, D.G. and Nunn, J.A. 1989. Free thermohaline convection in
change over geologic time and that complicates basin sediments surrounding a salt column. Journal of Geophysical
modelling. In subsiding basins, however, it is the geo- Research 94, 12413–12422.
thermal gradients during the last part of the subsidence Harrison, W.J. and Summa, L.L. 1991. Paleohydrology of the
which are most important. In the North Sea Basin we Gulf of Mexico Basin. American Journal of Science 291,
109–176.
may have had relatively high geothermal gradients in Hermanrud, C., Eggen, S. and Larsen, R.M. 1991. Investigations
late Jurassic times, but both the source rocks and the of the thermal regime of the Horda Platform by basin
reservoir rocks were then only buried to rather shallow modelling: Implication for the hydrocarbon potential of the
depths and the temperatures were still relatively low, Stord basin, northern North Sea. In: Spencer, A.M. (ed.),
Generation, Accumulation and Production of Europe’s
except close to volcanic intrusions and hydrothermal Hydrocarbon. European Association of Petroleum
activity. Geoscientists, Special Publication 1, Oxford University
Press, Oxford, 65–73.
Midttømme, K., Roaldset, E. and Aagaard, P. 1997. Thermal
Further Reading conductivities of argillaceous sediments. In: Mcann, D.M.,
Eddleston, M., Fenning, P.J. and Reves, G.M. (eds.), Modern
Geophysics in Engineering Geology. Geological Society
Special Publication 12, pp. 355–363.
Allen, P.A. and Allen, J.R. 2013. Basin Analysis. Principles and Person, M. and Garven, G. 1992. Hydrologic constraints of
Play Assessment. Wiley. 619 pp. petroleum generation within continental rift basins: Theory
Bethke, C. 1985. A numerical model of compaction-driven and application to the Rhine Graben. AAPG Bulletin 76,
groundwater flow and heat transfer and its application to 466–488.
the paleohydrology of intracratonic sedimentary basins.
Journal of Geophysical Research 90, 6817–6828.
Chapter 10
Subsurface Water and Fluid Flow in Sedimentary Basins
Knut Bjørlykke
The pore spaces in sedimentary basins are mostly determine if there is a breakthrough of the injected
filled with water. Oil and gas are the exceptions and water to the production well.
most of the information we have about fluid flow in (4) The water composition determines the density of
sedimentary basins is derived from the composition of the water column which is required to interpret
water and the pressure gradients in the water phase. It and calibrate the fluid pressure data from wells.
is therefore important to characterise and understand (5) Water composition and its resistivity are critical
the variations in the composition of these waters. All for the calibration of well logs. The resistivity is
the porewater may be referred to as subsurface water crucial for calculating the formation factor F,
but the water that is analysed from exploration wells or which is the resistivity of the formation water
is produced during oil production is usually called relative to the resistivity of the whole fluid-
formation water or oil field brines. saturated rock.
The composition of subsurface water can provide Water buried with the sediments has long been
vital information for several different practical referred to as connate water and thought to represent
purposes: the original seawater. This is an unfortunate term
(1) The water composition may give information because the origin of such water is very complex and
about the origin of the water and the pattern of meteoric water has been found to have a much stron-
fluid flow in the basin. ger influence in sedimentary basins than earlier
(2) Differences in water composition with respect to assumed. In addition, seawater changes its compo-
salinity or isotopic content may provide important sition significantly as soon as it is buried; the first
information about communication (permeability major change is the removal of sulphate ions in the
and fluid flow) across barriers like shale layers or sulphate-reducing zone (Figs. 10.1 and 10.2). Then the
faults. porewater will gradually approach equilibrium with
(3) The composition of water produced together with the minerals present in the sediments as the tempe-
oil may give information about the rock units from rature increases.
which the water is being drained. When water Subsurface waters are mainly derived from:
injection is applied to a reservoir, the composition (1) Seawater buried with the sediments.
of the injected seawater is different from that of (2) Meteoric water, which is groundwater (originally
the formation water and this can be used to rainwater) that can flow from land to far offshore
along permeable beds, mostly sandstones and
limestones.
(3) Water released by dehydration of minerals i.e.
from gypsum, or clay minerals like smectite and
kaolinite.
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway (4) Hydrothermal water introduced by igneous acti-
e-mail: [email protected] vity. This is also referred to as juvenile water.
280 K. Bjørlykke
Composition of porewater
NaCl SO42– CO2 PH
Seawater
Met. water flow
Clastic sediments
Diffusion of salt
Evaporites
Transport and precipitation of elements near the seafloor occurs mainly by diffusion across the redox boundary.
Fig. 10.1 Simplified illustration of how the composition of porewater is influenced by reaction near the seafloor and supply of
freshwater into sedimentary basins from land areas
Evolution of the concentrations in water

2–
O2 SO4 U Mn Fe
Sea level
2+
O2 UO2
2–
SO 4 2+
UO 2
Mn(OH)4 Fe(OH)3
Seafloor
FeS UO2
Sediments UO2
(Fe) Mn FeS2
FeCO3
Mn2+ FeS2
The redox boundary is a function of the supply of free oxygen or sulphate from the seawater and the consumption of oxygen
mainly by oxidation of organic matter.
In the absence of H2S from sulphur reduction, Fe2+ is more soluble but may precipitate as FeCO3.
Fig. 10.2 Only a few centimetres below the seafloor there are concentrated below the redox boundary while elements like
important reactions between the seawater, which is normally iron and manganese are precipitated above the redox bound-
oxidising, and reducing porewater. Elements that are most ary on the seafloor
soluble in the oxidised state (sulphur and uranium) are
10 Subsurface Water and Fluid Flow in Sedimentary Basins 281
10.1 Composition of Formation Water minerals are normally negatively charged on the sur-
face and particularly at the ends, where there are
In most cases the formation water is saline and Cl is broken silicate bonds. Negatively charged ions are
by far the dominant anion, so that the Na+=Cl ratio is repulsed and held back by the membrane due to the
less than 1. The rest of the positive charge is mostly negative charges of the clay minerals. In order for the
made up of Mg2+and Ca2+. The composition of charges on both sides to equalise, the small ions,
porewater is a function of its primary origin, modified particularly H+, must move in the opposite direction,
by the mineral composition of the sediments, the and as a result there is a higher H+ concentration
temperature and the quantity of dissolved gases, parti- (lower pH). This process, which will concentrate salt,
cularly CO2. is called salt sieving. Membranes also discriminate
Pure water of meteoric origin derived from fresh selectively amongst different cations, depending on
groundwater may gradually mix with more saline their size and charge. Those alkali ions which are
water and become more brackish. Meteoric water usu- strongly hydrated (Na+, Li+) and also Mg2+, will to a
ally has less than 10,000 ppm dissolved material com- lesser degree be adsorbed onto the surface of the clays
pared to seawater, which has 35,000 ppm and contains and will be more mobile than for example K+ and Rb+,
more bicarbonate (HCO3 ) and small amounts of mag- which are less hydrated. At higher temperatures the
nesium, sodium and calcium. Because meteoric water hydration becomes less effective and ions like Mg2+
comes from a land surface, it brings with it oxygen and are more available to be adsorbed on clay minerals and
bacteria which can break down hydrocarbons (see form carbonate minerals like dolomite.
“Biodegradation”). In sediments with even small At 70 100 C, smectite will dissolve and form
amounts of organic matter the porewater quickly illite and water. At 120–140 C kaolinite will become
becomes reducing as oxygen is consumed by break- unstable and form illite, quartz and water. This pro-
down of the organic matter. The composition of mete- cess binds cations, particularly K+, and releases pure
oric water alters as it reacts with the more readily water, thus reducing the salinity of the porewater. As
soluble minerals in the sediments. Small amounts of a result the porewater in shales often has a lower
carbonate makes the meteoric water basic and will act salinity than that in sandstones at the same depth.
as a buffer with respect to pH. Meteoric water will The composition of porewater in sandstones varies
flow from land areas where the groundwater table is greatly, depending on whether they contain meteoric
elevated above sea level through sandstones serving as water or not.
aquifers far offshore, depending on the pressure head The quantity of dissolved solids in porewater
of the groundwater table. increases as a function of depth in most cases.
Drilling on the continental shelves, for example off Concentrations of 100,000–300,000 ppm of dissolved
the east coast of the USA and also offshore Africa, matter (total dissolved solids, TDS) are typical for
shows that meteoric water aquifers are widespread and basins with evaporite beds. With increasing tempe-
that water which is virtually fresh is sometimes found rature, the kinetic obstacles to mineral solution and
below the seabed as far out as 100 km from the coast. precipitation reactions are reduced. At temperatures
In the northern North Sea there is also isotopic evi- above about 80 C the porewater will tend to be in
dence that the formation water in shallow reservoirs is equilibrium with most of the minerals present. In sedi-
partly of meteoric origin. This also shows that the mentary basins with evaporites, these will greatly
porewater is often stratified and there is very limited influence the composition of the porewater in
vertical mixing of porewater with different salinities. overlying formations. Dissolved salts move upwards
Marine porewater, which is present in sediments through compaction-driven porewater flow; diffusion
when they are deposited, will initially have an approx- due to high concentration gradients also plays a major
imately normal salinity but sulphate is removed by role. Sedimentary basins along the Atlantic Coast in
sulphate reduction just below the seafloor. Clay areas where Mesozoic evaporites are deposited have
minerals act as ion exchangers, absorbing cations porewater compositions which are essentially differ-
like K+ and Mg2+ from the porewater. Compacted ent from those in areas north of this palaeoclimatic
clay and mud may function as a membrane. Clay belt. The South American continental shelf, the Gulf
282 K. Bjørlykke
Coast and the area off the East Coast of the USA are that it has very low sulphate content. The chlorinity,
characterised by Jurassic/Cretaceous evaporites. In the however, remains practically unchanged because Cl
North Sea the extension of Zechstein evaporites forms is not consumed by any significant diagenetic process.
an important boundary which is reflected in the Steep concentration gradients and a strong drive for
porewater composition of overlying Mesozoic transport by diffusion exist across the redox boundary
sediments. In addition to chemical analysis, isotope because of the difference in solubility of many ions
composition (13 C=12 C and 18 O=16 O) can provide valu- between the seawater, which is normally oxidised, and
the porewater below. Sulphate SO24 is transported

able information concerning the origin of the
porewater. downwards and is reduced to sulphides below the
Many of the dissolved ions in porewater are in redox boundary. Reduced sulphur reacts with iron-
equilibrium with the minerals present. They are then bearing minerals including iron oxides (haematite)
not very useful as indicators of the origin of the water. to form pyrite (FeS2). Reduced manganese and
Cl and Br are better tracers for fluid flow as they do iron (Mn2+, Fe2+) will be transported upwards and
not react very much with the minerals present. precipitated above the redox boundary. While much
The water produced from drill stem tests may be of the iron will be trapped in the reduced state as
strongly contaminated by the drilling mud filtrate and sulphides, manganese sulphides are rather soluble and
the composition of the mud should be considered very little Mn will therefore be trapped below the redox
when using the analyses of such waters. Water pro- boundary. Ferrous iron (Fe2+) may also be trapped as
duced during production is less likely to be strongly carbonate such as siderite (FeCO3). Manganese is
contaminated because of the larger volume involved. therefore transported upwards more efficiently than
The results of the analyses may be expressed as % or iron and may form large deposits (as manganese
ppm. nodules) in deepwater environments where the sedi-
mentation rate is low.
mg=l ¼ ppm/density of water The composition of subsurface water is strongly
influenced by the dissolution of evaporites where
meq=l ¼ mg=l valence/mol.weight these are present but the effect can often be shown to
be rather local (radius <1 km). Meteoric water can
The main anions in subsurface waters are Cl , dilute the chlorinity of porewater but mixing of
HCO3 and SO24 , and the main cations are Naþ ; K þ porewater is not very efficient in sedimentary basins.
and Ca++. This is because the flow is slow and laminar and
Meteoric water is characterised initially by low pore waters with different salinities have different
ionic strength but it then reacts with minerals and densities, which also tends to inhibit mixing. Trans-
also amorphous phases like opal A. Unless meteoric port by diffusion from high to lower salinity may
water flows through evaporites the chlorinity will nevertheless be significant over distances of a few
remain very low and the main anions will be bicarbon- hundred metres, depending on the diffusion constant
ate ðHCO3 Þ or carbonate CO23 . Sulphate SO24 of the sediment matrix. Low permeability shales also

may form in meteoric water due to oxidation of have low diffusion constants. Dehydration of minerals
sulphides in rocks and of sulphur in organic matter, such as gypsum or clay minerals like smectite, kaolin-
but the sulphate will tend to be reduced to sulphides by ite and gibbsite also causes reductions in salinity
sulphate-reducing bacteria. because pure crystal-bound water is released into the
Porewater of marine origin naturally starts with the porewater.
composition of seawater but only a few centimetres Higher salinities than seawater are in most cases
below the seafloor most of the free oxygen is removed due to the dissolution of evaporites.
from the porewater due to oxidation of organic matter High salinity porewater is found in the central
in the sediments. A few metres below the seabed North Sea above the Permian evaporites in the Central
nearly all the sulphate which was in the seawater has Graben, while in the northern North Sea the porewater
been consumed by sulphate-reducing bacteria. It is is of normal salinity or brackish composition. In the
therefore a characteristic of so-called connate water northern North Sea where there are no evaporites,
formation water salinity is in most cases close to that with respect to the silicate minerals because of the
of seawater, or less saline (brackish) due probably to a slow reaction rates. Meteoric water is usually highly
component of meteoric water. Near evaporites the supersaturated with respect to quartz because it does
porewater is also often rich in calcium due to gypsum not precipitate at low temperatures (<70–80 C). Sea-
or anhydrite, with a corresponding reduction in water, on the other hand, is usually very much under-
sodium, so that CaCl2 is an important dissolved salt. saturated with respect to quartz because silica is taken
Dissolved NaCl may be transported away from the out of seawater by siliceous organisms, mainly
evaporites by diffusion or by porewater flow (advec- diatoms.
tion). In both cases the salinity will be reduced away Amorphous silica from organisms (opal A) may
from the salt deposit and it is unlikely that the survive during burial down to 1.5–2 km (60–80 C)
dissolved salt would have become sufficiently before dissolving. Opal A is replaced first by opal
concentrated for halite to be precipitated again. CT, which is unstable and will be replaced by quartz.
The halite and also gypsum commonly observed in As long as opal A and opal CT exist in the sediments
small amounts in sandstone cores is due to evaporation the porewater is supersaturated with respect to quartz.
in the core store; they are not diagenetic minerals as At higher temperatures the porewater becomes closer
has sometimes been reported. During tectonic uplift to equilibrium with quartz and other minerals present.
and erosion, hydration of minerals like anhydrite The isotopic composition of subsurface water
causes increased salinity but this is normally diluted reflects to a large extent the initial composition.
with meteoric water. Near the surface the salinity can Porewater with a marine origin has characteristic
of course increase by evaporation. In the seawater at values close to Standard Mean Ocean Water
shallow water depth the pH is relatively high because (SMOW). More negative values may indicate the
the water is often at least nearly saturated with respect introduction of meteoric water into a marine basin.
to calcite and the pCO2 is low. At greater water depth At greater burial depth the composition of porewater
in the ocean and also below the seafloor the solubility is more influenced by the reactions of the minerals. As
of CO2 increases, lowering the pH. Carbon dioxide is a general rule dissolution of minerals formed at low
consumed near the ocean surface by photosynthesis temperature, i.e. during weathering (kaolinite, smec-
and released by oxidation of organic matter sinking tite), causes the pore waters to become more positive
towards the ocean floor. Just below the seafloor in the when they dissolve at greater depth (higher
sulphate reduction zone more CO2 is produced during temperatures). This is because the minerals that
fomentation of organic matter. Then at greater burial precipitated at a higher temperature, in this case illite,
depth, thermal maturation of kerogen releases CO2. will contain less 18O. Dissolution and precipitation of
Organic acids are also generated from the source rocks clastic quartz, which was originally precipitated at
and probably also from the oil. Porewater is, however, high temperature, will shift the porewater in a negative
like seawater a buffered solution and organic acids are direction when low temperature quartz, with more
18
weak acids. Compared to the total buffering system of, O, is precipitated.
firstly, the silicate mineral system and, secondly, the We now have considerable data on the composition
carbonate system, the addition of relatively small of porewater in sedimentary basins from exploration
amounts of comparatively weak acids (organic acids) and production wells. It is clear that the porewater
will not change the pH of the porewater significantly composition varies greatly over distances of a few
(Hutcheon 1989). These different types of CO2 have hundred metres and this is evidence of very limited
characteristic ranges in δ13 C composition. mixing by advection. Except around salt domes the
Formation analyses from sedimentary basins show porewater seems to be crudely stratified with respect to
both salinity and isotopic composition δ18 O . This

that the pore waters are frequently crudely stratified
with respect to salinity and this puts constraints on puts important constraints on the transport of solids in
porewater flow. Also the oxygen isotope compositions solution by fluid flow. The saline porewater is not
may vary with depth, probably at least partly due to transported very far from the salt. With increasing
diagenetic reactions. At shallow depth and low tempe- temperature the density of porewater is reduced,
rature the porewater is normally out of equilibrium while it is increased by increasing salinity.
284 K. Bjørlykke
10.2 Composition of Porewater in an Oil of flow. This part of the chapter will discuss the
Field factors controlling the flow of fluids in sedimentary
basins.
Analyses of porewater that is present in the oil field The fluids are mostly water, but may also be oil and
along with the oil provide information about fluid gases including air, filling the pores between the grains
flow and migration of oil. Differences in salinity in the sediments. The grains are in most cases minerals
and chemical composition between parts of a reser- but some may be amorphous (e.g. silica – opal A, or
voir may indicate lack of communication by fluid organic matter and kerogen). Porosity is the percent-
flow and also diffusion between compartments. The age (or fraction) of the rock volume which is filled
isotopic composition of the formation water may also with fluids. This may be also expressed by the void
help to indicate degrees of communication in a res- ratio which is the ratio between the volume of voids
ervoir in addition to changes in the composition of (porosity) and solids. What is called void is not strictly
oil. Strontium isotopes have been used for this pur- void but filled with fluids and is the same as porosity.
pose. If there are primary differences in the compo- The relation between porosity (φ) and void ratio (Vr) is
sition of oil across an oil field this can be used to thus:
monitor the contribution from the different parts
during production. During onshore production fresh- Vr ¼ 1=ð1 φÞ
water may be used for water injection into the reser-
voir. Offshore, seawater is injected, but this has a In the oil industry porosity is mostly used while
composition which is distinctly different from the rock and soil mechanics literature tends to use void
formation water. ratio.
At any given time the amounts of solids in solution Porewater flow in sedimentary basins can be classi-
are very small except in highly saline porewater near fied according to the origin of the water and the
evaporites. The solubility of most silicates and also driving mechanism for the flow:
carbonate minerals is sufficiently low that the (1) Meteoric water flow is sourced by groundwater
porewater composition of sedimentary basins is (originally rainwater) and in most cases the
almost totally controlled by the solid phases. groundwater table is above sea level, providing a
Porewater in sedimentary basins often shows evidence drive downwards into the basin. This water is
of stratification with respect to both salinity and iso- normally fresh (low salinity) except in very arid
topic composition, which precludes large scale mixing regions where meteoric water may dissolve
of porewater. The composition of the formation water evaporites.
in reservoirs may provide useful information about the (2) Compaction-driven water is driven by the effec-
source of the water and communication within the tive stress and thermally-driven chemical compac-
reservoir, in the same way as the composition of the tion which causes a reduction in available pore
oil gives information about the source of the oil. space. The upward component of this flow is a
function of the rate of porosity reduction (compac-
tion) in the underlying sediments.
10.3 Fluid Flow in Sedimentary Basins (3) Density-driven flow is driven by gradients in the
fluid density due to differences in salinity or
Fluid flow in sedimentary basins is important because temperature. Thermal convection is driven by
it determines the distribution of pore pressures in the the thermal expansion of water. As the tempe-
water phase, and also in oil and gas. High pore rature increases downwards in sedimentary
pressures may also be a hazard when drilling wells. basins the density is reduced, creating a density
The fluid phases have the capacity to transport solids inversion with depth. Flow driven by thermal
in solution, and heat by fluid flow (advection) and by convection differs from the two other types of
diffusion. The flow of fluids may be through the pore flow in that the same water is used over again
network in the rock matrix or along fractures, and and is not dependant on an external supply of
this is a principal difference between these two types porewater.
Hydrodynamic (fluid) potential

Reference level
(i.e. sea level)
Fluid density ρf
h (Fluid column height) h

P = g ∫ ρf dh
h=0
Fluid pressure P
Fluid potential Fp = P - ρf gh. When Fp = 0, the pressure is hydrostatic
The fluid density (ρf) is a function of both temperature and salinity

and we should try to estimate an average density over the depth interval (h).
Fig. 10.3 Illustration of fluid potential which is the difference between the fluid pressure at a certain depth and the weight of the
overlying column of porewater and seawater
10.4 Fluid Potentials σ ev ¼ ρr gh P
The flow of fluids in sedimentary basins follows sim- Differences in hydrodynamic potentials can also be
ple fluid dynamics laws. Fluids do not necessarily flow expressed in terms of potentiometric (or piezometric)
from higher to lower pressures, but from higher to surfaces which are the heights to which water would
lower fluid potentials. The fluid potential is defined as: rise above sea level (or some other reference datum
like the groundwater level) in an open pipe from a rock
Fp ¼ P ρgh: in the subsurface. Pore waters with a higher potentio-
metric surface than sea level are defined as
Here P is the fluid pressure, ρ is the density of the overpressured while those that have a potentiometric
fluid, g the acceleration of gravity and h is the distance (piezometric) surface close to sea level or the ground-
up to some reference level, which in a sedimentary water table are normally pressured.
basin could be the sea level or the water table It is often stated that fluids flow from high to lower
(Fig. 10.3). The fluid potential is thus the potential pressure but this is obviously not always true. In the
for fluid flow, so if the fluid potential is zero there ocean the pressure increases from the surface down
can not be any flow. towards the bottom but water does not flow from the
The fluid potential is an expression of the deviation bottom to the surface because there is in most cases no
from the pressure gradient due to the density of the potentiometric head. The pressure gradient in the
fluid column. In the case of water the hydrodynamic ocean water is close to the density gradient (ρrgh)
potential expresses the deviation from the hydrostatic and ocean currents are driven by very small differ-
pressure gradient (ρwg), which is defined by the weight ences in fluid potential due to changes in temperature.
of the water column. In the case of groundwater flow To maintain significant pressure (potentiometric)
the fluid potential is usually referred to as the hydrau- gradients, there must be a resistance to flow; in the
lic potential, which is the mechanical energy per unit case of flow in porous rocks this is measured as per-
volume of groundwater. The lithostatic stress, which meability. Fluid flow is a function of permeability (k)
may also be referred to as the total stress, is the weight and viscosity (μ), and the flow of water and other
of the rock column saturated with fluids (ρr), for the fluids, can be described by the Darcy equation:
most part water. The difference between the total
stress (ρrgh) and the pore pressure (P) is the effective F ¼ rP k=μ:
stress (σ ev) which is transmitted by the sediment
particles or the rock:
286 K. Bjørlykke
Permeability
The permeability is the resistance to flow in any material.
If the permeability is 1 Darcy (1 D) the fluid flux is 1 cm3 . cm–2 . s–1, when the pressure
gradient is 1 atm.cm–1 (100 kPa.cm–1) in the direction of flow and the viscosity of the
fluids is 1 centipoise (as for water at 20°C).
Pressure gradient
∇P = 100 kPa.cm–1
Flux: 1 cm3 . cm–2 . s–1
SI unit:
1 D = 10–12 m2
1 mD (one millidarcy) = 10–3 D (10–15 m2)
1 nD (one nanodarcy) = 10–9 D (10–21 m2)
Fig. 10.4 Illustration of the Darcy equation for fluid flow in sedimentary basins
The flux F can be expressed as volumes of fluids absolute or intrinsic permeability to make it clear that
passing through a certain cross-section in a given time, it only relates to the characteristics of the rock, as
i.e. cm 2 cm 2 s 2 (Fig. 10.4). This is also called the opposed to the relative permeability which is the per-
Darcy velocity. The absolute velocity of the flow is meability of one immiscible fluid in the presence of
vðcm=sÞ ¼ fluxðcm3 cm 2 s Þ divided by the poros- another fluid, compared to the permeability with 100%
ity (φ). So if the porosity is 0.1 (10%) the velocity is 10 saturation of one fluid.
times the flux. In reality the velocity is a little higher The hydraulic conductivity (K) or transmissibility
because the fluid pathway is not along a straight line. is an expression of the ability of the rocks to conduct
Since the porosity varies greatly along the flow path it or transmit fluids of a certain viscosity (μ). The
is more useful to use the flux (cm3/cm2) or the Darcy conductivity has the dimension of m/s or ft/s and can
velocity (cm/s) as a measure of fluid flow rates. rP is be calculated from the permeability if the viscosity is
the potentiometric gradient, i.e. the change in potential known ðK ¼ k g=μÞ. At about 20 C, the kinematic
over a certain distance. In the horizontal direction this viscosity of water is 1 : 10 6 m2 =s and at 100 C it is
is the same as the pressure gradient. 0:210 6 m2 =s which is one centipoise. 1 Darcy (per-
Given a pressure P1 in a point X1 at a depth h1, and a meability) ðkÞ ¼ 10 5 K (conductivity) for water
pressure P2 in a point X2 at the depth h2, the potentio- (strictly 9:66 10 6 K), 1 Darcy ¼ 10 12 m2 or 10 8
metric gradient (rP) is: cm2. Well sorted and poorly cemented sand may have
permeabilities between 1 and 10 Darcy. In tight shales
rP ¼ ððP1 ρgh1 Þ ðP2 pgh2 ÞÞ=X2 X: the permeability is typically below 1 nanodarcy (10 9
ð X2 X1 is the distance between P1 and P2: Þ Darcy) or even much lower (Fig. 10.4).
Water is not very compressible. The compressi-
The permeability k is an expression of the resistance bility is close to 4:4 10 10 =Pa 1 . The pressure of a
to flow and is a constant in the Darcy equation which 1 km water column (10 MPa) causes a compression of
relates solely to the properties of the rock. Permeability water of about 0.4% or 4 m. The expansion of water
has the dimension of m2 ð1 Darcy ¼ 10 12 m2 Þ. The during tectonic uplift or release of overpressure is
permeability of a rock may be referred to as the therefore relatively small. Usually the cooling of
Groundwater
table
Hydrodynamic
potential Potentiometric surface Head
H2 H1
H4 H3
Seawater
Shale
Aquitard
Sandstone
Aquifer
Shale
Fig. 10.5 Fluid flow along a sandstone which is a confined aquifer overlain by a shale which is an aquitard
water associated with uplift will cause a thermal con-

traction which is greater than the expansion due to Recharge area
pressure reduction, so that the porewater becomes Precipitation Potential lines
denser. Water table
Porewater flow is oriented perpendicular to points
Stream lines
of equal hydrodynamic potential (isopotentional lines) Sea level
because that will represent the steepest potentiometric
gradient. That will be the case in relatively homo-
geneous sediments but generally the flow is very
much controlled by the distribution of highly per-
meable aquifers such as poorly cemented sandstone
layers, and of shales which serve as low permeability
aquitards (fluid barriers) (Fig. 10.5).
The distribution of very high permeability aquifer
sandstones and low permeability aquicludes (mostly Fig. 10.6 Flow of meteoric water in massive sand, perpendic-
shales, salts and cemented layers in salts) that control ular to the pressure distribution (potential lines)
most of the flow in sedimentary basins is closely
linked to primary sedimentary facies, tectonic into the small pores, due to capillary forces. Fractures
developments and diagenesis. These depositional and may be open and provide highly permeable pathways,
diagenetic processes determine the “plumbing sys- but others may be almost impermeable barriers to fluid
tem” in the basin and this must be the starting point flow because they are closed or cemented up. In a
for fluid flow modelling. massive sand the flow is very different and controlled
Aquicludes are beds of such low permeability that by the orientation of the isopotential lines which may
they can not transmit significant quantities of fluids be controlled by the groundwater table (Fig. 10.6).
under normal hydraulic gradients. These are ground- Fluid flow modelling that uncritically applies the
water hydraulics terms but over geological time even Darcy equations to fluid flow in sedimentary basins
low permeability shale will transmit some fluid. Some can lead to quite unrealistic results. Frequently the
shales may have extremely low permeability and be porewater flux is calculated from an observed pressure
practically impermeable, thus capable of maintaining gradient and from assumed or modelled permeabilities
overpressures on this timescale. The term seal is often for the different lithologies. One of the main problems
used as equivalent to aquiclude in the oil industry, and with such calculations is that the permeability can not
may describe a seal for both oil and water. Shales be determined very accurately. It typically varies by
which are not completely sealing with respect to several orders of magnitude from bed to bed up
water may nevertheless prevent oil from entering through a sequence and there may also be large
288 K. Bjørlykke
variations along the bedding due to sedimentary distribution of permeable sediments (usually sand-
structures. stones and limestones) will to a large extent dominate
In the case of flow perpendicular to the bedding the the pattern of fluid flow. At depths greater than 3–4 km
average permeability is the harmonic mean of the (100 C) quartz cementation may be more extensive
permeability of the individual beds. The harmonic along the fault planes than in the adjacent sandstones.
mean of the permeabilities of a number of beds (n) is The importance of fault planes as conduits for fluid
defined by: flow is greatest in well-cemented uplifted sedimentary
rocks or in metamorphic rocks, because the matrix
"
n
# 1 permeability is so low. These are rocks that have
X
kh ¼ i=n 1=ki been subject to unloading and usually some extension
i¼1 (fracturing), and they possess high rock strengths
which can prevent fractures from closing. Deforma-
The harmonic mean is strongly influenced by the tion of well-cemented rocks may produce rock
bed with the lowest permeabilities such as a tight shale fragments (brecciation) which by wedging the faults
or a carbonate-cemented interval. For flow parallel to may help to resist horizontal stress, keeping the
bedding the arithmetic mean is relevant. Even if we fractures open. In more porous sedimentary rocks
measure the permeability at relatively short intervals like sandstones and limestones, which are not well
in a cored section it is very difficult to come up with a cemented and have higher matrix permeabilities,
good average permeability, partly because relatively fractures are less critical. In subsiding sedimentary
thin, low permeability, layers affect the value to such a basins the sediments do not usually have sufficient
degree. strength to resist the horizontal stress that is trying to
In exploration we want to make predictions ahead close any open faults and fractures. Faults in this
of drilling, but without core data it is very difficult to setting are therefore more likely to be barriers than
provide assumptions about the permeability distri- conduits for fluid flow (see Chap. 11 on rock
bution with the degree of confidence needed for mechanics).
modelling fluid flow. In most cases, however, the Prediction of fluid pressure ahead of drilling in the
fluid flux (F) is primarily constrained by the supply basin depends on the permeability distribution in three
of fluids. The flux of meteoric water (groundwater) dimensions over distances of several kilometres. Even
flow is limited by the infiltration of rainwater into the if the geology is known in great detail, it would still be
ground. The potentiometric gradient near the surface is difficult to specify sufficient details about the per-
the slope of the groundwater table (or the potentio- meability to obtain a realistic fluid flow model based
metric surface). If the rainfall is 1 m/year the infiltra- on sedimentology and structural geology. During
tion may be 0.3 m/year and this is the initial flux in the exploration we normally have insufficient data to
recharge area. If this flux is continuous for 1 million model fluid pressure. During production much more
years the total flow is then 3 105 m3 m 2 which is data is available on the distribution of permeabilities
very significant, several orders of magnitude larger and the model can be constrained by the pressure
than the average compaction-driven flow. However response to production, and to water injection.
the meteoric water flux is greatest near the surface
and decreases rapidly with depth. We must remember
that the meteoric water flowing into the basin also 10.5 Meteoric Water Flow
must flow up to the surface. Sandstones that pinch
out in mudstones or shales will support only a very If the permeability is homogeneous in all directions,
low flux despite the high permeability of the the porewater flow can be calculated from the eleva-
sandstones. tion of the groundwater table and the fluid densities
Fluid flow resulting from differences in hydro- alone. The flow is perpendicular to lines with equal
dynamic potential can be treated mathematically potentials. In an isotropic rock matrix with constant
using Darcy’s equation and mass conserving equations fluid density, the flow of meteoric porewater follows a
during flow (continuity equations). Modelling flow for curved pattern perpendicular to the isopotential lines
whole basins is very complex. The orientation and (Fig. 10.6). Sedimentary rocks are generally very
inhomogeneous and the contrasts in permeability ρsw ¼ 1:025 g=cm3 and ρmw ¼ 1:0 g=cm3 ; D ¼ 40 H,
caused by clay layers and sand or gravel beds will meaning that the depth of the freshwater wedge is 40
generally totally dominate the flow of water times the height of the groundwater table underneath
(Fig. 10.5). While sand and gravel beds may have an island or within a confined aquifer along the coast.
permeabilities of 1–10 Darcy, the permeability of If the meteoric water becomes brackish by mixing
poorly-compacted mud may be 0.01 mD or lower. In with saline waters, the density difference is reduced
such cases, mathematical modelling is of little value if and the depth of penetration can be deeper because
the stratigraphy, sedimentology and structural defor- ðρsw ρmw Þ becomes smaller. The depth of meteoric
mation of the sedimentary sequences are not inter- water flow into highly saline porewater is very much
preted correctly. Modelling can nevertheless help to less. Relative sea level changes serve as a pumping
constrain some of the interpretations by calculating the mechanism, driving meteoric water into sedimentary
consequences of different alternatives. basins at low sea level stands due to the increase in
Meteoric water (freshwater) has a potentiometric hydrodynamic head. In this way early diagenesis may
head defined by the groundwater table, which is nor- be linked to sequence stratigraphy.
mally higher than sea level. Fresh porewater will thus During the last glaciations sea level was lowered
flow into marine sedimentary basins from the more than 100 m several times. This increased the
coastlines (Fig. 10.7). Along the coast and underneath head of the groundwater table on the land area, push-
islands, lenses of freshwater float on more saline ing meteoric water deep into sedimentary basins.
porewater like an iceberg in the sea. The depth to Under completely hydrostatic conditions, a 100 m
which freshwater will penetrate is a function of the head (10 MPa pressure) should theoretically corres-
density difference between the freshwater and the pond to a penetration down to about 4 km following
saline water: the above calculations. However, the flow will mostly
follow permeable beds (aquifers) and may extend a
D ¼ Hρmw =ðρsw ρmw Þ great distance out from the coastline. Well log
analyses from offshore Georgia, USA, suggest that
Here, D is the depth of penetration below sea level, freshwater extended for up to 100 km offshore just a
H is the height of the groundwater table above sea few metres below the seafloor (Manheim and Paull
level and ρswand ρmw are the densities of saline water 1981). In the modern Gulf of Mexico Basin, the
and meteoric pore waters, respectively. For depth of meteoric water penetration is estimated to
Rainfall
Fluvial
Shallow
Shelf marine
Sea level
Turbiditic
slope
Sandy sediments
facies
Fig. 10.7 Flow of meteoric water into sedimentary basins. The fluid flux will decrease away from the coastline and with increasing
depth.
290 K. Bjørlykke
be about 2 km (Harrison and Summa 1991). The depth feldspar dissolution (secondary porosity) and contain
of penetration does not depend only on the head of the more authigenic kaolinite than sandstones repre-
meteoric water but also on compaction processes in senting turbidite facies (Bjørlykke and Aagaard
the sediments, which can generate overpressures that 1992). Sediments in sedimentary basins like the
may exceed the meteoric water head. Even slight North Sea may be intensively flushed by meteoric
overpressures due to compaction will strongly reduce water immediately after deposition and also after
the depth of meteoric water penetration. uplift episodes and erosion. When tectonic uplift
The rainfall, the catchment area and the percentage results in subaerial exposure and the formation of
of infiltration into the groundwater determine the islands, meteoric water is collected on land and driven
upper limit of meteoric water flow. Meteoric water into the subsurface around the islands and adjacent to
which flows down into a sedimentary basin must even- other land areas. The sediments most strongly affected
tually flow up to the surface, in order to maintain a by meteoric water leaching are constantly being
continuous flow. Permeable sandstones are hydrauli- removed by erosion, however. This may explain why
cally almost dead ends if they pinch out into very low there is not always much evidence of feldspar leaching
permeability mudstones. Pressure will then build up in and high kaolinite contents immediately below
the aquifer and reduce the meteoric water inflow, unconformities.
forcing water to flow through the overlying Good examples of clay mineral diagenesis related
mudstones. We must remember that even if the fluid to modern groundwater systems have been observed
flux is small, the large area of contact between a sand down to 3–400 m depth in the Mississippi Gulf coastal
layer which serves as aquifer and the overlying mud plain (Hanor and Mcmanus 1988). In Canada there is
will allow relatively high volumes of porewater to isotopic evidence of recent meteoric water diagenesis
escape upwards through the mud. Even if the flow extending several hundred metres below the land sur-
per area (flux) is small through the mudstones the face (Longstaffe 1984). It must be stressed that the
area for vertical flow is large compared to the vertical isotopic composition of the porewater acquires a mete-
cross-section. At shallow depths (<500 m) prior to oric signature as soon as a volume of meteoric water
severe compaction, overpressure is rarely developed has displaced the marine (connate) porewater. New
as the permeability is much higher than in compacted minerals (like calcite, kaolinite and quartz) precipi-
mudstones and shales. tated in this porewater will reflect that isotopic
The degree of meteoric water flushing is highly signature.
dependent on climate and facies. The land surface
and vegetation determine the percentage of rainfall
which infiltrates down to the groundwater. Fluvial, 10.6 Porewater Flow Driven by Thermal
deltaic and nearshore shallow marine sediments will Convection
be flushed by meteoric water shortly after deposition.
The flux is then likely to be high and the porewater is Thermal convection is an effective mechanism for the
still very much undersaturated with respect to feldspar mass transfer of dissolved material in sedimentary
and mica. The total volume of water flowing through basins, because the same water can be used over and
each volume of sediment is inversely related to sedi- over again (Wood and Hewett 1982, Davis et al. 1985).
mentation rates. At high sedimentation rates the The limitation of fluid (water) supply, which constrains
sediments spend less time in the zone of meteoric compaction-driven flow, is then eliminated. Thermal
water flushing. Sands deposited in more distal shelf convection may occur because the density of water
facies (nearer the shelf edge) and turbidites (on the decreases with depth in a sedimentary basin as the
slopes) are normally less well connected to the main temperature increases, due to the thermal expansion
groundwater wedge, so that the flux is lower and the of water. This creates an inverse density gradient
porewater is closer to equilibrium with respect to the which may be unstable. If the isotherms (lines of
mineral phases. equal temperature) are horizontal, the density as a
In the North Sea basin it has been demonstrated that function of temperature will not vary horizontally and
reservoir sandstones deposited in fluvial and shallow there is no flow unless the water overturns. The denser
marine environments have been subjected to more upper layers of porewater may start to overturn and
Sandstone
k = 1 Darcy
Siltstone Shale
1000 m
h = 1.0 m h = 0.1 m
k = 10–3 Darcy k = 10–5 Darcy
Sandstone
k = 1 Darcy
∆h= 0.001 ∆h = 10–4

n= 0.001 n = 10–5
n = ratio between the permeabilities in
the thin layers and the thicker layers
K=k
∆T KH2 K = nk
R = 1.2 •10−2 • H h
H
K=k
Fig. 10.8 A modelling of the low permeability layers on vertical Rayleigh convection in sedimentary basins (Bjørlykke et al. 1988)
sink into the less dense water below. The condition are rarely uniform and the vertical permeability typi-
required for such overturning can be expressed in cally changes abruptly in a sequence of sandstones and
terms of a critical Rayleigh number. In the case of shales. Thin layers (0.1 m) of low permeability shales
thermal convection of water the Rayleigh number can or cemented layers in sandstones may cause almost
be defined as follows (Bjørlykke et al. 1988): complete flow separation, producing smaller convec-
tion cells instead of potentially larger ones (Fig. 10.8)
R ¼ gβΔTHk=κμ (Bjørlykke et al. 1988). Each of the convection cells
may then have insufficient height (small H) to exceed
Here, g is acceleration due to gravity, β is the the critical Rayleigh number. The low vertical perme-
coefficient of thermal expansion of the fluid (water), ability in layered sequences suggests that Rayleigh
ΔT is the temperature difference between the upper convection is probably not very important in sedimen-
and lower boundaries of the convection cell, H is the tary basins. Several hundred metre thick sandstones
thickness of the layers, k is the permeability, κ is the with no thin shales or cemented intervals are rarely
thermal diffusivity and μ is the viscosity. Assuming encountered.
reasonable values for the properties of water the equa- Non-Rayleigh convection will always take place
tion can be expressed in a simpler form (Bjørlykke when the isotherms are not horizontal. This is because
et al. 1988): the temperature and the fluid density are then not
constant in the horizontal direction. This situation is
R ¼ 1:2 10 2 krTH always unstable because there is a potentiometric
drive for the waters to overturn, which will produce
The critical Rayleigh number which must be some fluid flow without the need to exceed a critical
exceeded for Rayleigh convection to occur is about Rayleigh number. The velocity for non-Rayleigh con-
40. We see that the most critical factors are the height vection is:
of the water column and the permeability of the rocks.
If the permeability is 1 Darcy and the geothermal v ¼ g k β sin αrT=μ
gradient 30 C/km, the thickness (H) of the permeable
layer (sandstone) must exceed about 300 m for the When the geothermal gradients vary only moder-
critical Rayleigh number to be exceeded so that ther- ately within a basin, the slope of the isotherms (α) will
mal convection can occur. Sedimentary rocks, though, be small and the flow velocity very low. Sloping
292 K. Bjørlykke
isotherms will also result from sloping beds because The increase in density due to the salinity may
the heat flux is reflected when the conductivity of the totally or partly offset the density reduction due to
beds varies, but this effect is also normally quite small. the thermal expansion of water. At a salinity gradient
From the equation above we see that the flow rates are of 30 ppm/m, and an average geothermal gradient, the
a function of the isotherm slope, as well as a function effects of the thermal expansion of water are more
of the height of the convection cell which often is than offset, so that the water becomes denser with
equal to the thickness of a sandstone bed. The thick- depth (Fig. 10.9). This effectively removes any drive
ness of the sandstone beds (H) or the distance between for convective flow (Bjørlykke et al. 1988). When
the low permeability shales, will normally define the such trends are recorded in formation water analyses
height of the convection cells and in most sedimentary or well logs it provides strong evidence that vertical
sequences the distance between thin shales or even mixing is not taking place, because convection would
clay laminae is only a few metres or less. Although have destroyed the salinity gradients (Gran et al.
non-Rayleigh convection nearly always occurs to 1992). Around salt domes the permeabilities are
some degree it is probably rather insignificant in often rather low, further reducing the potential for
most cases in terms of the transport of solids in fluid flow.
solution in sedimentary basins. This is because the The salinity in sediments surrounding salt diapirs
low slopes of the isotherms and the low vertical per- can be used to trace fluid flow, since Cl– is not con-
meability result in very low flow rates, mostly inside sumed to any significant degree by diagenetic
rather thin sandstones separated by shales. These are reactions. Analyses of the salinity distribution around
probably not very significant in terms of solid trans- salt domes from offshore Louisiana show some evi-
port in connection with diagenetic processes. Around dence of convection, but the observed salinity stratifi-
igneous or hydrothermal intrusions, however, the lat- cation and the lack of more mixing and dilution of the
eral change in geothermal gradients and the slope of saline porewater suggest this convection is very slow
the isotherms may be very high and then thermal (Ranganathan and Hanor 1988, Evans and Nunn
convection is very important. Also around salt domes 1989). Inverted salinity gradients are only likely to
geothermal gradients may change due to the higher develop around salt diapers or underneath salt layers.
conductivity. Inverse salinity gradients will increase The inverse salinity gradients can only sustain flow on
the drive for thermal convection, while normal salinity the down-going limb of a convection cell. Unlike
gradients will make the porewater more stable. Even thermal convection there is no mechanism to make
moderate salinity gradients strongly influence fluid the water flow up again and the water would also
flow in sedimentary basins (Fig. 10.9). gradually homogenise with respect to salinity. In
basins like the North Sea, porewater analyses
Density of porewater
3
1.00 g/cm3
3
0.90 g/cm 1.10 g/cm
40°C Salinity
grad: 1.5%/km
70°C
Temperature
100°C
130°C
Temperature/ Salinity
density grad: 0.3%/km
160°C
Fig. 10.9 Density of water as a function of salinity and temperature (from Bjørlykke et al. 1988). The increase in temperature
causes a thermal expansion and a density inversion while salinity gradients may have the opposite effect
Porosity (%)
10 20 30 40 50
0 F = kvP/u
The compaction-driven flux
2 Porosity/depth is relatively constant over
time.
10% porosity
Sediment
4 Porosity loss/km subsidence: 1.5%
Km
at 4 km depth
Integrated flux (from 4–8 km
depth) pr km subsidence at 4 km
6
depth:
4% porosity
8 at 8 km depth 60 m3/m2 over 33 million years
Fluid flux = 60 m3/m2/33 × 106yr =
Base-
ment
2 × 10–5 m3/m2yr
Fig. 10.10 Illustration of a porosity/depth function in a sedi- (the derivative) is an expression of the compaction-driven water
mentary basin. The integrated area defined by this curve is an flux from each layer
expression of the total volume of water in the basin and the slope
demonstrate that the porewater is at least stratified in a from the seafloor though the sedimentary sequences to
crude way with respect to its composition. The salinity the underlying basement we obtain an area A below
increases downwards towards the Permian salts, effec- the porosity/depth curve. This is an expression of the
tively ruling out large scale convection and excursions total volume of water in the basin per unit area
of compaction-driven flow from the deeper parts of the (Fig. 10.10). The total compaction-driven flow of
basin into the overlying sequence (Gran et al. 1992). water in the basin is a function of the changes in the
There is also a trend towards more positive δ18 O porosity/depth curve. As new layers of sediment are
values with depth, which again confirms some degree deposited the underlying sediments compact and the
of porewater stratification (Moss et al. 2003). The porewater is forced upwards.
vertical salinity gradients and the isotopic composition It is possible to show that at a constant sedi-
of the porewater confirm that the porewater is not mentation rate the average upward component of the
undergoing convection on a large scale and that there compaction-driven flow is always equal to or lower
is no large flow of porewater from the deeper part of than the subsidence rate (Caritat 1989). The porewater
the basin to shallower depths. This has important is therefore moving upwards through the sedimentary
consequences for diagenetic models in connection sequence but nearly always downwards relative to the
with fluid transport of solids in solution. seafloor. We may say that the sediments are sinking
through a column of porewater. Typical sedimentation
rates in sedimentary basins are 0.1–0.01 mm/year and
10.7 Compaction-Driven Porewater the average rates of upwards porewater flow are lower
Flow than these values. Assuming there is no flow of water
from the basement there is practically no upward flow
As sediments compact they lose porosity and the in the basal layer of sediments. The porewater in this
excess porewater has to be expelled. This is the driving layer subsides at almost the same rate as the basement.
force for compaction-driven flow. The rate of porosity Higher up in the sequence the compaction-driven flow
loss is a function of effective stress, lithology, temper- receives contributions from more and more layers. In
ature and time. The porosity/depth functions observed the uppermost layer the porewater flux is equal to the
in sedimentary basins may be very complex, total integrated porosity loss over time in the under-
depending on the lithologies. At the transition between lying sequence. The upwards-flowing porewater is
two lithologies the porosity may increase with depth filling the pore space of new layers deposited on the
but for a uniform lithology the porosity will decrease seafloor, and during continued subsidence there is
with depth. If we integrate the porosity/depth function normally no porewater flow up though the seafloor
294 K. Bjørlykke
into the water column except when there is local near tectonically uplifted areas, may contain
focusing of the flow. During periods with no sedi- porewater of meteoric origin with very low salinity.
mentation (hiatus), the deposited sediments continue The isotopic composition of the porewater often shows
to compact and porewater flows across the redox a very clear stratification. Oxygen isotopes are often
boundary just below the seabed and dissolved ions negative in the shallow parts of the basin due to the
like Fe2+ and Mn2+ may precipitate as Fe(OH)3 and inflow of meteoric water, while they may be positive at
Mn(OH)4. greater depth due to diagenetic reactions.
Very high degrees of focusing are required to During subsidence there are generally no open
obtain high flow rates. It is then easy to understand fractures because of the ductile properties of sub-
why compaction-driven flow is several orders of siding sediments (Bjørlykke and Høeg 1997).
magnitude lower than meteoric water flow, at least in The permeabilities in shales are very low, probably
the shallow parts of the basin near land. The slow less than a nanodarcy (Leonard 1993). Samples
porewater flow rates also imply that this water has measured in the laboratory may show erroneously
time to approach thermal and chemical equilibrium high values because of fracturing resulting from
with the minerals. Porewater will always transport unloading during core retrieval. Increasing the effec-
some heat but, compared to the background heat flow tive stress during laboratory measurements to compen-
by conduction, this is very small and in most cases can sate for this has been shown to lower the
be ignored (Bethke 1985, Ludvigsen 1992). permeabilities by two orders of magnitude (Katsube
Modelling compaction-driven flow in the Gulf et al. 1991). In sediments which have been subject to
Basin, Harrison and Summa (1991) found that the uplift, however, fracture permeability may be impor-
maximum rate of the vertical component is 2 mm/ tant. Laboratory measurements of shale permeabilities
year (2 km/million years), which is approximately may be several orders of magnitude larger than the
equivalent to the maximum sedimentation rate. This results of well tests in the field (Oelkers 1996).
flow rate is too slow to contribute significantly to the In subsiding basins the effective permeabilities on a
heat flow. The temperature distribution in the Gulf large scale must be low, not above about 10 9 Darcy,
Basin at the present day does not reflect compaction- to maintain overpressures over time. Calculating the
driven porewater flow. Most of the heat transport is by porewater flux due to compaction and using observed
conduction. The average compaction-driven flow pressure gradients, the effective permeability of thick
within a basin is more or less independent of the shale sequences can be calculated. Using data from
permeability because it is a function of the rate of Haltenbanken offshore Norway this method gave
loss of porosity in the underlying sediments. During permeabilities between 10 9 and 10 10 Darcy (Olstad
mechanical compaction, low permeability sediments et al. 1997). These calculations are based on one
may result in overpressure and a certain reduction of dimensional models and the results are most probably
fluid flow, but at greater depth where the compaction is minimum values for the permeability values of the cap
mostly chemical, the fluid flux is independent of the rock since lateral drainage is not included. Lateral flow
permeability although variations in permeability may would reduce the vertical flow and the permeability
focus the flow. would have had to be lower to maintain the observed
vertical pressure gradients. Increasing the vertical
fluid flux by focussing the flow would imply that the
10.8 Constraints on Water Flow in permeability was higher.
Sedimentary Basins by Porewater
Chemistry
10.9 The Importance of Faults
Many sedimentary basins contain evaporites, often
occupying the basal part, having formed during the Faults may greatly affect fluid flow in sedimentary
initial rifting. Very high salinity is then typically found basins and may serve either as conduits or barriers,
in a zone of a few hundred metres adjacent to the salt. depending on the situation. A clear distinction must be
This is the case with the Zechstein salt in the North Sea made between flow along, and across, the fault plane.
basin. The shallower parts of a basin and in particular Faults are also important because they may offset
Pressure
Water
Lithostatic pressure
∇P = ∆P/H
Tight (low permeability)
Depth
shale (cap rock).
∆P
Reservoir sandstone. F = Fluid flux

(m3/m2/s)
Hydrostatic pressure Fracture pressure

(ρwgh)
Fig. 10.11 Illustration of fluid pressure and rock pressure fracture pressure which is equal to the horizontal stress, as the
(lithostatic). If the permeability in shales (seals) is low enough seal will then leak
overpressure will build up. The pressure can not exceed the
porous sandstones (aquifers) against shales, rendering Brecciated fault planes may therefore be important
permeable sandstones a dead-end in terms of fluid conduits for fluid flow but over time the permeability
flow. Faults that cut through sandstones may have a will gradually be reduced by cementation.
clay smear from adjacent shales or from mica or The cements will in most cases not be due to
authigenic kaolinite inside the sandstone, and this precipitation from advective flow but form by diffu-
may significantly reduce the permeability, in some sion from the adjacent rock matrix. Carbonate and
cases sufficiently to form an oil trap. silicate minerals next to the faults are under lithostatic
Flow along fault planes requires that they are kept stress and therefore more readily soluble than when
open to some degree. A force equivalent to the hori- unstressed. In the case of carbonate cement this is
zontal stress acts on the fault plane, trying to close it; fairly clear because upwards (cooling) flow will dis-
an equivalent overpressure is required to counteract solve calcite rather than precipitate it, due to its retro-
this in order for the fracture to remain open grade solubility. Brecciated faults contain broken rock
(Fig. 10.11). However, this corresponds closely to fragments and quartz grains that are good nucleation
the fracture pressure and even without the presence sites for quartz cement formation. A thermodynamic
of a fault the rocks would fracture. At such high drive towards dissolution of the minerals under stress
overpressure there are very low effective stresses and is therefore likely, with ensuing precipitation in the
the sediments are unable to compact mechanically. fault plane. Renewed fracturing is therefore required
We must also consider the source of the fluids. When for the faults to remain permeable.
there is no compaction (porosity reduction) in the
adjacent sediments, water can not flow from the rock 10.10 Seismic Pumping
matrix into the fractures. Well-cemented sedimentary
rocks and basement rocks have high shear strength and In crystalline and well-cemented sedimentary rocks
may produce rock fragments (brecciation) during tectonic shear may result in an extensional shear fail-
faulting. These rock fragments may wedge the fault ure which opens up fractures. Increases in the tectonic
plane, helping to resist the horizontal stress. stress may reduce the opening of such fractures and
296 K. Bjørlykke
produce fluid flow that is often referred to as seismic pressure changes related to fracturing. Compaction
pumping. This mechanism may work in metamorphic will therefore slowly build up the pore pressure again
and well-cemented sedimentary rocks where the rock unless there is continued flow from the overpressured
strength is high enough to keep fractures from being section.
closed by tectonic stress. In these types of brittle rocks As we saw from the calculations, the flow resulting
most of the water is present in the fractures and little in directly from the pressure release is rather limited.
the rock matrix. Softer sediments are by contrast more The main expulsion of porewater is due to sediment
ductile in their response to tectonic stress and fractures compaction, which is an indirect consequence of the
will normally not stay sufficiently open to transmit reduction of overpressure and increase in effective
fluid rapidly. In compacting (normally consolidated) stress (effective stress ¼ overburden stress minus
sediments most of the water is in the sediment matrix pore pressure). Compaction and expulsion of
and even if fractures should remain open the rate- porewater resulting from increased effective stress is
limiting step is the flow of porewater from the matrix, a gradual and rather slow process and this strongly
which often consists of low permeability mudstones influences the rate of water supply to the faults from
and shales. mudstones. The permeability of the surrounding
If the low permeability seal overlying or mudstones further limits the rate of flow into the
surrounding an overpressured part of the basin is bro- fault zone. If a fault plane does not extend up to the
ken, i.e. through faulting, a rapid pulse of porewater surface (or seafloor), the upwards-moving porewater
flow upwards may follow. However due to the low will have to be accommodated in shallower strata.
compressibility of water (4:3 10 10 Pa 1 ) the Large volumes of porewater can not suddenly be
upward flow of porewater necessary to reduce the injected into shallower sandstones even if the latter
pressure is relatively small. For an overpressure of are normally pressured.
107 Pa (potentiometric surface 1 km above sea level), Flow into shallow aquifers of limited extent will
the average expansion of the water would be result in a temporary pressure build-up before the
4.310 3 (Leonard 1993). In a vertical column water can be displaced, and this will reduce the flow
through an overpressured sequence this would result rate. The highest flow rates can be expected when the
in an average upwards flow of 4.3 m for each km of fault extends all the way to the surface so that the
sequence, if all the overpressure was released at one porewater can escape into the water column or onto
time. If the flow was focused through a smaller cross- the land surface. In many sedimentary basins like the
section, this figure would increase proportionally. North Sea basin, most of the faults extend only into the
It is very unlikely that large volumes of shales Cretaceous or lower Tertiary section. These faults were
would reduce their overpressure over a very short therefore not very active during the early Tertiary, and
time, given their low permeabilities, so the potential certainly not during the later Tertiary and Quaternary.
for episodes of rapid flow of compaction water is The timing of faulting must be taken into account when
limited. If the water is saturated with respect to gases faults are called upon to explain fluid flow and dia-
like carbon dioxide or methane the compressibility of genetic reactions in sedimentary basins. In a sequence
the pore fluid will increase significantly. A reduction with a high clay/sand ratio, faults may be sealing
in the pressure then will cause gas to come out of between two sandstones due to the clay smear on the
solution and form a separate phase, which has a high fault plane.
compressibility. The degree of overpressure is reduced
by porewater flow through a leaking seal, but only by a
small amount. The pressure will not drop much below 10.11 Episodic Flow
fracture pressure before the fine fractures close. In the
relatively shallow section this small increase in effec- When the source of the fluids is hydrothermal the flow
tive stress may result in some mechanical compaction. may be episodic, at least when considering individual
Chemical compaction (>100 C), however, occurs at a fractures or a limited area, because new fractures
very slow rate which is mostly a function of tempe- develop and close. Igneous intrusions (sills and
rature and time and is relatively independent of the dykes) may cause boiling and episodic flow. For a
cooling batholith as a whole the fluid flow may be more During compaction of sediments there must always
uniform, depending on the rate of cooling and on the be a slight overpressure because there must be suffi-
circulation of groundwater by thermal convection. It cient pressure gradients for the excess porewater to
has also been argued that compaction-driven flow may flow out so that the porosity can be reduced. The
be episodic when the fracture pressure is reached. Darcy equation shows that the pressure gradient must
Fluids flow through the fractures produced by hydro- be an inverse function of the effective permeability of
fracturing but as they escape, pressure will be reduced the rocks forming the seal. When we have very low
and the fractures will be reduced or close. It is not so permeabilities a high pressure gradient will build up,
clear if the fracture will stay closed for some time and which will drive the water out. The term disequili-
then leak again or whether it will continue to transmit brium compaction has been used to describe the devel-
fluids at a variable rate corresponding to the rate of opment of overpressure because the permeability is
compaction. At depths greater than about 3 km where too low for the water to be expelled at lower pressure
most of the compaction is chemical, the rate of com- gradients. A disequilibrieum is always required for
paction will be slow and relatively independent of compaction to take place, but if the permeablities are
changes in the stress field. When considering larger not very low, only a slight overpressure is required for
volumes of rocks they can not be heated or cooled the explusion of porewater during compaction. The
rapidly, at least not in a non-hydrothermal environ- build-up of overpressure reduces the effective stress
ment, because of the high specific heat capacity of the with the effect of stopping or at least reducing
rocks. The rate of compaction and the resulting fluid mechanical compaction (Fig. 10.11). Overpressure
flow will then be relatively uniform over a limited time. thus provides a negative feedback on mechanical com-
paction. Chemical compaction in siliceous rocks
involving quartz cementation will still continue as a
10.12 Formation of Overpressure function of temperature also during uplift as long as
(Abnormal Pressure) the temperature exceeds 70-80 C but at a lower rate.
The pore pressure can then build up to fracture
Overpressure is a term used for subsurface pressures that pressure.
significantly exceed the hydrostatic pressure. This From the Darcy equation we see that the pressure
implies that the flow of porewater to the surface during gradient (P) is:
compaction is resisted to a considerable degree, so that
the pressure gradients are increased in the least perme- rP ¼ F μ=k
able part of the sediments. Overpressure may be pro-
duced by different mechanisms. The simplest type of It is clear that increases in the fluid flux (F) could
overpressure is due to the pressure (head) of an elevated cause high overpressure, but the pressure gradient is
groundwater table connected to the basin through an very sensitive to variation in permeability.
aquifer. Rainwater infiltration into the ground will then High sedimentation rates and basin subsidence will
help to maintain the pressure even if the aquitards (seals) increase the compaction-driven fluid flux (F) and will
do not have very low permeability. The overpressure contribute to the build-up of overpressure if the per-
will nevertheless decrease away from the area of meability is considered to be constant. As shown
recharge. This is also expressed by decreasing piezo- above, the average upwards flow of porewater is
metric surfaces along the direction of flow. For over- equal to the integrated change in the porosity/depth
pressure to develop due to compaction the permeability curve in the underlying sediments.
in the seal must be many orders of magnitude lower In addition there is a fluid flux driven by the release
because the fluid flux is very much lower than in the case of crystal-bound water which in sediments with high
of meteoric water flow (Bjørlykke 1993). In a meteoric contents of water-bearing minerals may be significant.
water aquifer the flux may be up to 0.1–1.0 m/year, When porewater is heated the thermal expansion of
while the compaction-driven flux is usually less than the water will also add to the fluid flux but calculations
sedimentarion rate which may typically be 0.1 mm/year. show that this is not very significant.
298 K. Bjørlykke
This is partly because the porewater is not heated higher than 80–100 C, will reduce the porosity and
very much during basin subsidence since it is moving permeability not only in the sandstones but also in the
upwards as the basin and the sediments are sinking. shales. High sedimentation and subsidence rates
When solids (e.g. kerogen) are transformed into will increase the compaction-driven flux but the rate
fluids like oil and gas, a volume expansion may of permeability reduction in the shales (seals) due
occur. Cracking of oil and the formation of gas also to chemical compaction will be slower so the per-
involves a phase change, which will cause increased meability of the sealing shales will be higher. This
pressure because of the expansion of gas. could compensate for the higher fluid flux. Many
Generation of oil from kerogen gives an increase in shales are almost impermeable.
volume but calculations suggest that it is not very large Extremely low permeabilites are required to main-
because this leaves some solid material (coke) tain overpressures in basins that are uplifted and no
remaining. Generation of gas will cause a higher vol- longer undergo compaction. Overpressured reservoirs
ume increase, especially at relatively shallow depths. in onshore basins in North America have, for the most
Even if the total increase in volume is moderate the part, not subsided since the early Tertiary and it is
convertion of solid kerogen to fluid petroleum is a remarkable that the overpressures have been retained
very efficient mechanism for increasing the pore pres- without more recent compaction. In the Anadarko
sure. This is because the fluid/solid ratio is changed Basin the Missisippian and Pennsylvanian sequence
and this may cause source rocks to hydrofracture so is overpressured close to fracture pressure, while the
that the petroleum is expelled. An illustration of such underlying Ordovician rocks are normally pressured.
a phase change can be observed in the spring when In the Powder River Basin, Cretaceous shales are
lenses of ice frozen in the ground melt. The conver- highly overpressured but not to fracture pressure.
sion from solid ice to water represents a reduction in Widely distributed free gas (Surdam et al. 1994) may
overall volume but may generate overpressure also reduce the permeability for water in the shales. It
because the fluid/solid ratio has been very much is not clear if these pressures are maintained by active
increased. gas generation at depth. Reservoirs flanking mountain
When considering larger compartments in sedi- chains like the Rocky Mountains may also be
mentary basins, however, the pressure is mainly con- overpressured, here due to meteoric water flow from
trolled by the water phase which is much more the mountains.
abundant. In sedimentary basins the permeabilities are very
Development of overpressure depends on the fluid much higher parallel to bedding than perpendicular to
flux in relation to the permeability of the rocks. The bedding, and overpressure usually depends more on
permeability of shales which may serve as seals for the lateral drainage than on variations in the vertical
overpressure compartments varies greatly and is diffi- permeabilities. Modelling 1 D vertical flow is there-
cult to predict. A change from 1.0 nD (10 9D) to 0.1 fore not very realistic. Faulting that offsets permeable
nD will increase the pressure gradient 10 times. We sandstones against tight shales may contribute to the
must also remember that for vertical flow perpendi- development of overpressure. Synsedimentary growth
cular to bedding the effective permeability is the har- faulting in particular is very common in basins with
monic average of the permeabilities in the different high sedimentation rates like the Gulf Coast basins.
layers. Thin layers with very low permeability may This is one of the reasons for the widespread
therefore control the flow rate and the build-up of overpressuring in such basins.
overpressure. It is therefore very difficult to model Models for the prediction of overpressure are often
and predict overpressure. High overpressures will based on changes in fluid flux with less emphasis on
reduce the effective stress and make the sediments the permeability which is more difficult to constrain.
less consolidated. There will then be very little mecha- This is because the vertical flux depends on the har-
nical compaction and compaction-driven fluid flow to monic average of the permeabilities in all the layers
build up an overpressure. Fluid transfer from greater also within shales. If the permeability is kept constant,
depth may cause higher overpressure at shallower it is possible to model overpressure as a function
levels, though. The onset of chemical compaction, of other variables such as rates of compaction, hydro-
and in particular quartz cementation at temperatures carbon generation and thermal expansion of the
porewater. These variables must, however, be com- phases and approach equilibrium with the mineral
pared with the range of permeability values that are phases present at a rate which is kinetically controlled.
likely to exist in a sedimentary basin. Modelling of Highly soluble ions like chlorides, however, are not in
fluid pressures and the build-up of pore pressure are equilibrium with the minerals except within evaporite
often based on permeability distributions derived from deposits with halite (NaCl). Fluid flow in the deeper
porosity distributions, which are also poorly parts of sedimentary basins is constrained both by the
constrained. Smectitic mudstones have very low pressure gradients and the supply of fluids by compac-
permeabilities even at shallow depth and are parti- tion (reduction in porosity) and by mineral dehydration.
cularly effective seals and may be a significant factor Meteoric water is the most important supply of fluids
causing overpressure. In the North Sea smectite-rich and because it is renewed by rainfall the flow can be
Eocene and Oligocene mudstones and associated maintained for a very long time at shallow depth in
sandstones are frequently overpressured at just sedimentary basins. If other factors are constant the
1–2 km depth. total meteoric water flow through sediments are the
The permeability (k) of fine-grained sediments may inverse of the subsidence rate.
be related to porosity (φ) and will therefore decrease Compaction-driven flow is limited by the volume
during compaction. This relationship may be expressed of water buried in the basin and fluids produced in situ
as k ¼ cφ5 (Rieke and Chillingarian 1974). It is clear by mineral dehydration and petroleum generation. The
that rather small variations in porosity can produce upwards flow of porewater is usually lower than the
large variations in permeability. The most important sedimentation rate so that the porewater is moving
factor in the sediment composition is the surface area downward relative to sea level. High flow rates can
(s), which is closely linked to grain size (d). This is therefore not be sustained except by extreme focusing
expressed in the Konzeny-Carman equation: of the flow.
k ¼ cφ3 =ð1 φÞ 2 s 2
Further Reading
The effect of the surface area can also be expressed
in terms of tortuosity (t):
Audet, D.M. and McConnell, J.D.C. 1992. Forward modelling
3 2 of porosity and pore pressure evolution in sedimentary
k ¼ φ d =72 tð1 φÞ: basins. Basin Research 4, 147–162.
Berner, B.A. 1980. Early Diagnesis, A Theoretical Approach.
In the case of smectitic clays the specific surface Princeton University Press, Princeton, NJ, 141 pp.
Bethke, C.M. 1985. A numerical model of compaction-driven
may be several hundred m2/g while kaolinite and illite
groundwater flow and heat transfer and its application to
typically have about 10 m2/g (Skjeveland and Kleppe the paleohydrology of intracratonic sedimentary basins.
1992). The specific surface of mudstones rich in smec- Journal of Geophysical Research 90, 6817–6828.
tite may be more that 10 times that of mudstones with Bethke, C.M. 1986. Hydrothermal constraints on the genesis of
the Upper Mississippi valley mineral district from Illinois
mostly kaolinite, chlorite and illite. According to the
Basin brines. Economic Geology 81, 233–249.
Konzeny-Carman equation, the permeability in smec- Bethke, C.M. 1989. Modelling subsurface flow in sedimentary
tite-rich layers may thus be lower by a factor of 10 2 basins. Geologische Rundschau 78, 129–154.
compared to other mudstones. The validity of this Bethke, C.M., Harrison, W.J., Upson, C. and Altaner, S.P. 1988.
Supercomputer analysis of sedimentary basins. Nature 239,
equation is however not clear for such fine grained
261–267.
sediments. Bjørlykke, K. 1993. Fluid flow in sedimentary basins. Sedimentary
Geology 86, 137–158.
Bjørlykke, K. and Aagaard, P. 1992. Clay minerals in North Sea
10.13 Summary sandstones. In: Houseknecht, D.W. and Pittman, E.D. (eds.),
Origin, Diagenesis, and Petrophysics of Clay Minerals in
Sandstones. SEPM Special Publication 47, Tulsa, OK,
The origin of porewater in sedimentary basins may be pp. 65–80.
seawater, meteoric water (freshwater) or water released Bjørlykke, K. and Høeg, K. 1997. Effects of burial diagenesis on
from minerals by dehydration. Pore waters change their stresses, compaction and fluid flow in sedimentary basins.
composition by reacting with minerals and amorphous Marine and Petroleum Geology 14, 267–276.
300 K. Bjørlykke
Bjørlykke, K., Jahren, J., Aagaard, P. and Fisher, Q, 2010. Role Longstaffe, F.J. 1984. The role of meteoric water in diagenesis
of effective permeability distribution in estimating over- of shallow sandstones: Stable isotope studies of the
pressure using basin modelling. Marine and Petroleum Geo- Milk River Aquifer and Gas Pool, southeastern Alberta.
logy 27, 1684–1691. In: McDonald, D.A. and Surdam, R.C. (eds.), Clastic Dia-
Bjørlykke, K., Mo, A. and Palm, E. 1988. Modelling of genesis. American Association of Petroleum Geologists
thermal convection in sedimentary basins and its relevance Memoir 37. AAPG, Tulsa, OK, pp. 81–98.
to diagentic reactions. Marine and Petroleum Geology 5, Ludvigsen, A. 1992. Thermal convection and diagenetic pro-
338–351. cesses in sedimentary basins. PhD Thesis, University of
Buhrig, C. 1989. Geopressured Jurassic reservoirs in the Viking Oslo.
Graben: Modelling and geological significance. Marine and Manheim, F.T. and Paull, C.K. 1981. Patterns of ground water
Petroleum Geology 6, 31–48. salinity changes in a deep continental-oceanic transect off
Caritat, P. de. 1989. Note on the maximum upward migration of the Southeastern Atlantic coast of the USA. Journal of
pore water in response to sediment compaction. Sedimentary Hydrology 54, 95–105.
Geology 65, 371–377. Mudford, B.S., Gradstein, F.M., Katsube, T.J. and Best, M.E.
Cathles, L.M. and Smith, A.T. 1983. Thermal constraints on the 1991. Modelling 1D compaction driven flow in sedimentary
formation of Mississippi Valley-Type Lead-Zinc Deposits basins: A comparison of the Scotian Shelf, North Sea and
and their implications for episodic basin dewatering and Gulf Coast. In: England, W.A. and Fleet, A.J. (eds.), Petro-
deposit genesis. Economic Geology 78, 983–1002. leum Migration. Geological Society Special Publication 59,
Chapman, R.E. 1987. Fluid flow in sedimentary basins: a 65–85.
geologist’s perspective. In: Goff, J.C. and Williams, B.P. Nadeau, P. H., Bjørkum, P. A. and Walderhaug, O. 2005. Petro-
(eds.), Fluid Flow in Sedimentary Basins and Aquifers. leum system analysis: impact of shale diagenesis on reser-
Geological Society Special Publication 34, pp. 3–18. voir fluid pressure, hydrocarbon migration, and
Chester, R. 1990. Marine Geochemistry. Unwin Hyman, biodegradation risks. In: Doré, A. G. and Vining, B. A.
London, 698 pp. (eds.), Petroleum Geology: North-West Europe and Global
Davis, S.H., Rosenblat, S., Wood, J.R. and Hewett, T.A. 1985. Perspectives—Proceedings of the 6th Petroleum Geology
Convective fluid flow and diagenetic patterns in domed Conference. Published by the Geological Society, London,
sheets. American Journal of Science 285, 207–223. pp. 1267–1274.
Evans, D.G. and Nunn, J.A. 1989. Free thermohaline convection Oelkers, E.-H. 1996. Physical and chemical properies of
in sediments surrounding a salt column. Journal of Geo- rocks and fluids for chemical mass transport calculations.
physical Research 94, 12413–12422. In: Lichtner, P.C., Stefel, C.I., and Oelkers, E.H. (eds.),
Evans, D. et al. 2001. The Millennium Atlas. Petroleum Geology Reactive Transport in Porous Media. Reviews in Mineralogy
of the Central and Northern North Sea. Geological Society. 34, pp. 131–191.
Giles, M.R. 1987. Mass transfer and problems of secondary Olstad, R., Bjørlykke, K. and Karlsen, D.K. 1997. Pore water
porosity creation in deeply buried hydrocarbon reservoirs. flow and petroleum migration in the Smørbukk Field area,
Marine and Petroleum Geology 4, 188–201. offshore Norway. In: Møller-Pedersen, P. and Koestler, A.G.
Gran, K., Bjørlykke, K. and Aagaard, P. 1992. Fluid salinity and (eds.), Hydrocarbon Seals – Importance for Exploration
dynamics in the North Sea and Haltenbanken basins derived and Production. Norwegian Petroleum Society 7. Elsevier,
from well log data. In: Hurst, A., Griffiths, C.M. and Amsterdam, pp. 201–216.
Worthington, P.F. (eds.), Geological Application of Wireline Ortoleva, P. 1994. Basin Compartments and Seals. AAPG
Logs II. Geological Society Special Publication 66, pp. Memoir 61. AAPG, Tulsa, OK, 459 pp.
327–338. Person, M. and Garven, G. 1992. Hydrologic constraints of
Hanor, J.S. and Mcmanus, K.M. 1988. Sediment alteration and petroleum generation within continental rift basins: Theory
clay mineral diagenesis in a regional ground water flow and application to the Rhine Graben. AAPG Bulletin 76,
system, Mississippi Gulf Coastal plain. Transactions-Gulf 466–488.
Coast Association of Geological Societies 38, 495–502. Ranganathan, V. and Hanor, J.S. 1988. Density driven ground
Harrison, W.J. and Summa, L.L. 1991. Paleohydrology of the Gulf water flow near salt domes. Chemical Geology 74, 173–188.
of Mexico Basin. American Journal of Science 291, 109–176. Rieke, H.H. and Chillingarian, G.V. 1974. Compaction of Argil-
Hautshel T. and Karuerauf, A.I. 2009. Fundademntals of Basin laceous Sediments. Elsevier, New York, NY, 424 pp.
and Petroleums Systems. Springer, New York, NY, 476 pp. Surdan, R.C., Boese, S.W. and Crossey, L.J. 1984. The chemis-
Hutcheon, I.E. 1989. Application of chemical and isotopic try of secondary porosity: Part 2. Aspects of porosity modi-
analyses of fluids to problems in sandstone diagenesis. In: fication. In: McDonald, D.A. and Surdam, R.C. (eds.),
Hutcheon, I.E. (ed.), Short Course in Burial Diagenesis. Clastic Diagenesis. AAPG Memoir 37. pp. 127–149.
Mineral Association of Canada, Québec, pp. 270–310. Warren, E.A. and Smally, P.C. 1994. North Sea Formation
Katsube, T.J., Mudford, B.S. and Best, M.E. 1991.Petrophysical Water Atlas. Geological Society Memoir 15, 208 pp.
characteristics of shales from the Scotian Shelf. Geophysics Wood, J.R. 1986. Thermal transfer in systems containing quartz
56, 1681–1689. and calcite. In: Gautier, D.L. (ed.), Roles of Organic Matter
Leonard, R.C. 1993. Distribution of subsurface pressure in Nor- in Sediment Diagenesis. SEPM, Special Publication 38,
wegian Central Graben. In: Parker, J.R. (ed.), Petroleum Tulsa, OK, pp. 181–189.
Geology of N.W. Europe. Proceedings of the 4th Confer- Wood, J.R. and Hewett, T.A. 1982. Fluid convection and
ence. Geological Society, pp. 1295–1393. mass transfer in porous sandstones – A theoretical model.
Geochimica et Cosmochimica Acta 46, 1707–1713.
Chapter 11
Introduction to Geomechanics: Stress and Strain in

Sedimentary Basins
Knut Bjørlykke, Kaare Høeg and Nazmul Haque Mondol
At shallow depths in sedimentary basins there are soft secondary compression or creep. Compaction in
clays and loose silts and sands, unless there are deep sedimentary basins has occurred over geologic
carbonates or carbonate cemented layers. At greater time scales at very low strain rates, and at higher
depths they are transformed by diagenetic processes to temperatures than shallow sediment compaction.
hard claystones, shales, siltstones and sandstones. Sed- Therefore, in addition to mechanical compaction,
imentary rocks continuously undergo physical and there are important effects of mineral grain dissolu-
chemical changes as a function of burial depth, tem- tion, precipitation and cementation. This process is
perature and time, and important hydro-mechanical called chemical compaction. Here the rate of compac-
parameters change during burial, erosion and uplift. tion is controlled by the rate of chemical reactions
An understanding of these processes is important involving dissolution and precipitation of minerals.
in order to predict the magnitude and distribution Compaction determines the porosity, density and per-
of sediment properties and stresses in the basin. The meability of the sediments which are essential input
in situ stress condition affects the rock response for basin modelling; petroleum reservoir quality is
(strain) to changes in the stress field due to drilling dependent on the porosity and permeability. The pro-
and petroleum production. cesses of mechanical and chemical sediment compac-
Soil and rock mechanics (geomechanics) have tion (diagenetic processes) determine the physical
mainly been developed to solve engineering problems properties and are also important for understanding
in relation to landslides and surface and underground seismic velocity records and seismic attributes in sed-
construction. These are usually at very shallow depths imentary basins.
compared to that of a petroleum reservoir. We will
here focus on some aspects of geomechanics of partic-
ular relevance for the petroleum geologist. 11.1 Subsurface Fluid Pressure and
Porosity loss (volumetric compaction) with time Effective Stress Condition
due to increased effective stress is referred to in the
engineering literature as consolidation, while compac- A distinction should be made between total stress,
tion at constant effective stress is usually called effective stress and fluid pore pressure. This is not
always done in technical reports and publications
related to petroleum geology.
K. Bjørlykke (*) K. Høeg
e-mail: [email protected]; [email protected] 11.1.1 Total and Effective Stress
N.H. Mondol
Department of Geosciences, University of Oslo, Oslo, Norway In general, stress (σ) is defined as force per unit area.
Norwegian Geotechnical Institute (NGI), Oslo, Norway The overburden weight of the sediment including the
e-mail: [email protected] weight of the fluid in the pore space produces a
302 K. Bjørlykke et al.
vertical stress (σ v). For a sedimentary basin with a

fairly horizontal surface, and without major lateral
variations in the sediment compressibility, the vertical Overburden stress
stress at any point can simply be computed as:
σ v ¼ ρb gh (11.1)
Fluid pressure
where ρb is the average sediment bulk density (solids +
fluids) of the overlying sequence, h is the sediment
thickness and g is the acceleration of gravity. This is
the vertical total stress or the lithostatic stress. It may
be calculated more accurately by integrating the vary- Grain contact stress
ing density over the depth of the sediment column. The
Fig. 11.1 The total vertical stress from the overburden (σ 0v ) is
sediment density varies with the density of the grains, carried by the mineral grain framework (solid phase) and the
mainly minerals, the porosity and the density of the pore pressure (fluid phase). The effective stress is defined as the
pore fluids which could be water, oil or gas. The overburden vertical stress minus the pore pressure and it is
transmitted through the grain contacts
effective vertical stress (σ 0v ) is defined as the difference
between the vertical total stress (σ v) and the pore
pressure (u): The solid phase may also have variable density due
to different mineral composition, and in some cases
amorphous phases also play a role. Usually the density
σ 0v ¼ σ v u (11.2)
of the mineral matrix in sandstones and shales is close
to 2.65 2.70 g/cm3. If there are significant contents of
This is the effective stress which is sometimes
denser minerals such as siderite or pyrite the bulk
called the average intergranular stresses because it is
density will be higher. Smectite and mixed-layer
transmitted through the grain framework. It is the
minerals have variable but generally lower densities.
effective stress that governs the mechanical compac-
The fluid density also varies with the composition of
tion of sediments where little chemical compaction
water and petroleum. In the case of gas-saturated rocks
(cementation) has taken place. Stress (σ) is force (F)
the bulk density becomes significantly lower. The
divided by the area of contact (A). In coarse-grained
increase in total vertical stress per metre of depth is
sediments like sandstones the area of contact may be
commonly called the lithostatic stress gradient
very small and the stresses between the grains rather
(Fig. 11.2). At about 10% porosity (and assuming
high. It should be noted that the local intergranular
pores filled with water) the lithostatic stress gradient
particle-to-particle contact stress is many times higher
is typically 25 kPa/m (25 MPa/km) corresponding to a
than the effective stress as defined here, due to the
mineral density of about 2.66 g/cm3 as in quartz and
small area of contact. The total overburden weight is
illite. At 30% porosity the bulk density of sediments is
carried by the mineral grain framework and the pore
typically 2.1 g/cm3.
pressure (Fig. 11.1).
The rock density is critical for modelling isostasy
The effective stress in the horizontal direction is
and backstripping and it is mostly a function of the
defined as total horizontal stress minus the pore pres-
degree of compaction since mineral densities normally
sure. The horizontal stress will in general not be equal
do not vary greatly even if the mineral composition
to the vertical stress as discussed in Sect. 11.3 below.
does. Carbonates, particularly dolomite, are however
However, the pressure in the pore fluid (pore pressure)
significantly denser than shales and sandstones.
is the same in all directions.
The bulk density (ρb) of sedimentary rocks varies as
a function of the porosity (φ), the density of the fluid
(ρf) in the pore space, and the density of the solid 11.1.2 Fluid Pressure
phase (ρm) which is comprised mainly of minerals:
In general, the pressure in the pore fluid at any given
ρb ¼ φρf þ ð1 φÞρm (11.3) point may be equal to the weight of the water column
to sea level or groundwater table. The porewater may
11 Introduction to Geomechanics: Stress and Strain in Sedimentary Basins 303
a Stress ( ) b 10 20 30 40 50 60 70 80 90 MPa Stress

Sea
level
Water
Lithostatic stress (σv)
(20–25 kPa/m) 1 km Sea floor
Fracture pressure
Sedimentary
(18–20 kPa/m)
σ′v rocks
P1 Hydrostatic pressure 2 km
Depth (H)
(10–11 kPa/m) LithostaƟc stress

Overpressure curve.
The hydrostatic pressure
3 km
(blue curve) must be
σ′v
subtracted from Gas pressure Gas-saturated
P2 the red curve. gradient sand
GWC/OWC
4 km
HydrostaƟc
ρsgH Depth Pressure (water
ρbgH pressure)
At hydrostatic pressure (P1) the effective stress at a depth

H is σ′v = Hgρb–ρw but at overpressure (P2) the effective
stress is σ′v = Hgρb–P2.
Fig. 11.2 (a) Simplified diagram showing the increase in ver- pressure in a basin with 1 km water depth. The lithostatic stress
tical total stress (lithostatic) and hydrostatic pressure as a func- is equal to the weight (density) of the overlying sediments and is
tion of depth. In reality these lines are not strictly straight not a straight line. The hydrostatic pressure is the weight of
because the total vertical stress varies as a function of the the water column and the porewater under normal pressure
sediment bulk density (ρb) which tends to increase with depth. conditions. In a layer saturated with oil or gas the pore pressure
The hydrostatic pressure curve is a function of the density of the is reduced because of the buoyancy relative to water. At the gas/
formation water (ρw) which varies with temperature and salin- water contact or oil/water contact the pressure in these fluid
ity. (b) Diagram showing the distribution of stress and fluid phases is the same
then be said to be in a hydrostatic state. When signifi- (Fig. 11.2). Overpressure is sometimes also called
cantly higher pressure is measured it is called over- abnormal pressure. If u is equal to the hydrostatic
pressure, and underpressure also exists in some basins. pore pressure, then Δu is zero, and the sediment
There may also be a pressure gradient in the porewater sequence is said to be normally pressured. There is
which is different from the hydrostatic, and then there then no tendency to fluid flow.
will be a fluid flow which is a function of the perme- The overpressure (Δu) can be expressed as the equi-
ability and the viscosity. The flow may be relatively librium height (ΔH) of a hypothetical water column
constant over long time, but if the pore pressure above the level corresponding to hydrostatic pressure.
changes over a relatively short time the flow is said This is the level that water would rise to in a pipe from
to be transient. This would be the case with fluid flow the formation to the surface. It is called the piezometric
related to earthquakes. or potentiometric surface level. As discussed in Chap.
In offshore sedimentary basins the pore fluid pres- 10 on fluid flow, differences in fluid potentials or
sure (u) measured in oil or water is a result of several piezometric surfaces are expressions of the driving
contributions: forces for fluid flow in sedimentary basins.
The fluid densities referred to in the equations
u ¼ ρp gHp þ ρw gHw þ ρsw gHd þ Δu (11.4) above are the average densities for a certain fluid
column. The density of water decreases with depth
where ρp gHp is the pressure contribution from a petro- due to the temperature increase, but near evaporites
leum column of height Hp with the density of petro- the salinity gradient can offset the thermal expansion
leum ρp, ρw gHw is the pressure due to a water- so that the water becomes denser with depth. The fluid
saturated sequence (Hw), and ρsw gHd is the pressure pressure can be calculated more accurately by
contribution of the seawater column (Hsw) density ρsw. integrating the fluid density over the height of the
Δu is the overpressure (or sometimes underpressure) fluid column. If the water density is constant and
which is any deviation from the hydrostatic pressure equal to 1.0 g/cm3, the hydrostatic pressure gradient
is 10 kPa/m (0.1 bar/m). Expressed in psi (pounds per caused by the creation of new fractures will pro-
square inch) the equivalent gradient for freshwater is duce a significant lowering of pressure. During
0.434 psi/ft. In basins like the North Sea and the Gulf uplift the sediments no longer compact and water
Coast the water density varies significantly, and typi- can therefore not flow in from the rock matrix to
cal Gulf Coast pressure gradients are 0.465 psi/ft or fill the fractures without lowering the pressure.
10.71 kPa/m (Dickey 1979). Extension during uplift will thus tend to draw in
If a standpipe (well) is installed in a normally meteoric water from above, but if the fractures are
pressured sediment (no overpressure) the water not connected so that the water can flow up to the
would rise in the pipe to the sea level or groundwater surface, the flow will be rather limited. Compres-
table. This is called a piezometric or potensiometric sional tectonics or strike slip tectonics may pro-
surface. Artesian overpressures may be due to mete- duce episodes of rapid fluid flow along fractures,
oric water flow from, for instance, a mountain lake often referred to as seismic pumping.
into a sedimentary basin. If fluid pressure is higher (2) Condensation of gas to liquid petroleum may
than the fluid pressure corresponding to the weight of cause reduced fluid volume and lower pore pres-
the fluid above, the water in the standpipe would rise sure. This may, however, often be compensated
ΔH m above the local water table depending on the for by the expansion of dry gas and the release of
degree of overpressure. gas from porewater.
During burial and basin subsidence (compaction, (3) Cooling and contraction of water (the opposite of
compression) the pore pressure is above hydrostatic aquathermal pressuring) may cause lowering of
(transient overpressure) and the water flows out of the fluid pressure below hydrostatic.
sediments as they compact. Unless the permeability
in the sediments is very low, only very small
overpressures are required for the expulsion of water 11.2 Normally Consolidated Versus
during compaction. Then the rate of porewater flow is Overconsolidated Sediments
a direct function of the rate of compaction (porosity
reduction). If there are low-permeability barriers to A layer in a sediment sequence that never before in its
flow in all directions, high overpressures may develop geological history has been subjected to higher verti-
during burial because it takes a long time for the pore cal effective stress than at present, is called normally
pressures caused by the added overburden to dissipate/ consolidated (NC). If, on the other hand, the sediment
drain. There are also other processes that may lead to has been subjected to higher effective stresses, e.g. by
overpressure. High sedimentation rates will cause previous glacial loading, by higher overburden that
higher rates of compaction and compaction-driven subsequently has been eroded, and/or by pore
flux. Overpressure will retard mechanical compaction pressures in the past that were lower than at present,
because the effective stress is reduced. The porosity the sediment is called overconsolidated (OC) as it has
reduction is however very much a function of time and been preloaded. The ratio between the past maximum
temperature in the case of chemical compaction. Since effective vertical stress and the present stress is com-
chemical compaction in siliceous sediments is mainly monly called the overconsolidation ratio (OCR).
a function of temperature, compaction will continue At relatively shallow depths in a sedimentary basin
even at high overpressures and reduced effective (less than 2 3 km, <70 90 C), the mechanical com-
stress. paction processes dominate over the chemical com-
Lower than hydrostatic pressures (underpressure) paction in siliceous sediments. At higher temperature
can also develop but are less common and are usually (deeper burial) chemical compaction processes
formed during uplift in the sedimentary basin. Below become dominant in controlling the rate of compac-
are listed three ways in which underpressure may tion. Carbonate sediments may, however, become
develop: cemented and highly consolidated at shallow depth.
(1) Tectonic extension may slightly increase porosity Since this consolidation (compaction) is not due to
and create fractures which need to be filled with mechanical compaction but to chemical processes it
fluids, thus lowering the fluid pressure. As water is sometimes referred to as “pseudo overconso-
with no gas bubbles has low bulk compressibility lidation”. Since we refer to the maximum effective
(4.10 4 MPa 1), a small increase in the porosity stress sediments can not be undercompacted. Poorly
compacted sediments have either been subjected to 11.3.1 Theoretical and Semi-empirical
low effective stress or they have a low compressibility. Relationships
The hydro-mechanical properties at shallow depths
may be very different for a normally consolidated In a basin which is wide compared to its thickness, the
sediment sequence compared with an over- compaction process due to added overburden may be
consolidated one, depending on the magnitude of the modelled as a one-dimensional deformation situation
OCR. For the overconsolidated sediment, the com- (i.e. only strain in the vertical direction, no strain
pressibility and permeability are usually much lower horizontally). This is often denoted as a uniaxial strain
and the shear strength significantly higher. As compaction situation. If the sediment mineral skeleton
discussed below, the lateral stresses in (framework) may be assumed to behave in a linearly
overconsolidated sediments may be higher than in elastic and isotropic manner (see Sect. 11.4), the hori-
normally consolidated sediments. zontal stress which is built up as the vertical overbur-
den is increased, is defined by:
ν
11.3 Horizontal Stresses in Sedimentary σ 0H ¼ σ 0v (11.5)
1 ν
Basins
where ν is the Poisson’s ratio for the sediment mineral
Knowledge of the magnitude and distribution of hori- skeleton. The same relationship would hold for
zontal stresses in sedimentary basins is important in unloading if the material really exhibits linearly elastic
relation to petroleum exploration, drilling and produc- behaviour. Furthermore, for isotropic material, the
tion. Their magnitude is also important in the interpre- magnitude of horizontal stress would be the same in
tation of seismic signals used in field exploration and all directions. If one assumes anisotropic behaviour,
in reservoir production management. In a sedimentary the equations corresponding to Eq. (11.5) would be
basin the geomechanical properties vary from those of somewhat more complicated, and the horizontal
loose cohesionless sediments at shallow depths to stresses would be different in the different directions.
dense and cemented sedimentary rocks at greater The horizontal stress coefficient for a uniaxial deforma-
depth. This affects the horizontal (lateral) stress distri- tion situation is commonly called K0 in geomechanics.
bution with depth. For an assumed Poisson’s ratio ν ¼ 1/3, K0 becomes
While the vertical stresses are determined by verti- 0.5 from Eq. (11.5).
cal equilibrium (Eq. 11.1), the magnitude of lateral As discussed in Sect. 11.4, linearly elastic
stresses cannot be determined by equilibrium behaviour may be an acceptable approximation for a
equations and is statically indeterminate. Their cemented sediment (sedimentary rock) undergoing
magnitudes are governed by a number of factors, minor deformation. However, during the initial grad-
including the overburden/erosion (loading/unloading) ual build-up of loose sediments in a basin, it is not
and uplift history of the basin and the deformation realistic to assume linear elastic behaviour of the sedi-
characteristics of the sedimentary rocks. These are a ment framework. Its behaviour is very non-linear and
result of gravitational and tectonic forces, and also of inelastic, undergoing mainly permanent deformation.
stress changes caused by chemical compaction and the In soil mechanics one uses the following semi-
accompanying volume change. Their magnitude is empirical relationship for normally consolidated
determined based on an understanding of the geo- (NC) sediments. It is based on idealised theoretical
logical history, theoretical and semi-empirical considerations and on laboratory and field
relationships, and field measurements. Horizontal measurements:
stresses can be measured in wells, and particularly in
area of uplift and erosion they may exceed the vertical K0nc ¼ 1 sin φ0 (11.6)
stresses. At high pore pressures this will result in
horizontal fractures rather than vertical which is the where φ0 is the angle of shearing resistance (friction
case when the vertical stress is highest. angle) used in the Mohr-Coulomb failure criterion
expressed in terms of effective stresses, and it is Assuming φ0 ¼ 30, Kext and Kcom would be 1/3 and
assumed that there is no cementation (cohesion c ¼ 3, respectively. If a cohesion intercept (c) is included,
0). The K0 value defined this way refers to the ratio of the value for Kextwould be smaller and Kcom higher
effective horizontal and vertical stresses (not total than given by Eqs. (11.8) and (11.9), respectively. For
stresses). For φ0 ¼ 30; K0nc becomes 1/2. the case of horizontal compression the maximum hor-
When such a sediment is unloaded (erosion) and izontal effective stress is:
thus becomes overconsolidated (OC), the horizontal
stress does not decrease proportionally with the reduc-
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 þ sin φ0 0 1 þ sin φ0
tion in vertical effective stress because the sediment σ 0H ¼ σ þ 2c0 (11.10)
1 sin φ0 v 1 sin φ0
skeleton does not exhibit elastic behaviour. Horizontal
stresses are “locked in” in the sediment, and the K0-
value increases. The following semi-empirical rela- This is the linear Mohr-Coulomb failure criterion
tionship is used, based on laboratory experiments and expressed in terms of the major and minor principal
field measurements: stresses. In this case σ H is the major principal stress
and σ v the minor principal stress.
K0oc ¼ K0nc ðOCRÞn (11.7) Zoback et al. (1985) and other investigators have
measured high horizontal stresses in basement rocks.
where OCR is the overconsolidation ratio and n is a These stresses probably reflect compressional plate
coefficient experimentally determined to usually be tectonic movements. However, basin sediments are
between 0.6 and 0.8, depending on the sediment much more compressible than the basement rocks
properties. K0 may well reach values above 1 (2–3 (uncemented sands have been found as deep as
have been measured), which means that the horizontal 1.5–2 km in the North Sea). Therefore, the external
stress is significantly larger than the vertical stress. plate tectonic and regional tectonic stresses that are
During burial and compaction and erosion (uplift), transmitted through the underlying basement and the
the horizontal stresses may change due to tectonic deeper well-cemented sedimentary rocks will have
movements, and with time due to combinations of little effect on the horizontal stresses in the compress-
mechanical loading and unloading and also chemical ible sedimentary basin above, unless the lateral tec-
compaction which may be independent of stress. tonic movements (compressive strains) are very large
If extension occurs in a sediment with friction angle (Bjørlykke and Høeg 1997, Bjørlykke et al. 2005,
(φ0 ) but no cohesion intercept (c), the effective hori- 2006).
zontal stress coefficient would decrease from K0 to a In the North Sea the horizontal stress has been
lower limiting value of: found to increase with depth faster than the vertical
stress, and at 4 km and deeper the total horizontal
stress is usually 0.8 0.95 of the total vertical stress,
Kext ¼ 1 sin φ0 =1 þ sin φ0 (11.8)
approaching unity. The magnitude of horizontal stress
at these depths is influenced by the effects of chemical
At this low lateral effective stress, shear failure
compaction and creep. It should be noted that the ratio
would occur and a shear plane (normal fault) would
between the total horizontal and vertical stresses is not
form. Such extension may occur due to general basin
the same as the ratio between the effective stresses at
extension, or more locally over a salt dome, over an
the same location. In a sediment with high pore
elevated fault block of sedimentary rock with softer
pressures (overpressure), and a ratio between total
sediments on either side, or at the top of a slope.
stresses of 0.9, the corresponding ratio between effec-
If on the other hand, lateral compression occurs in
tive stresses may be about half that value, depending
the same sediment, the lateral effective stress coeffi-
on the magnitude of overpressure. It is the ratio
cient would increase to a limiting value (reverse
between effective stresses that indicates how close
faulting):
the sediment may be to a local shear failure, and it is
the magnitude of the minimum effective stress that
Kcom ¼ 1 þ sin φ0 =1 sin φ0 (11.9)
governs whether a tension fracture may occur.
The strain rates are important for the laboratory deter- tangential stresses around the borehole vary. When
mination of K0 because deformation by creep is a func- the fluid pressure is equal to the minimum effective
tion of time. Some rocks behave very differently at low stress plus the rock tensile strength (τ) at the most
strain rates and at high strain rates. For further discussion critical location around the borehole, i.e. the location
on brittle and ductile behaviour see Sect. 11.4.4. with the smallest initial tangential compression stress,
When the effective stress in the reservoir is increased a fracture opens in the wall of the borehole. This fluid
due to reduced fluid pressure during petroleum produc- pressure level is called the fracture pressure. By low-
tion, the strain rates are fairly high. Therefore, the ratio ering the fluid pressure after the crack has opened, and
between the horizontal and vertical stress will be con- then increasing the fluid pressure again, one may
trolled by mechanical compaction, and the K0 values determine the tensile strength of the rock as the differ-
determined in laboratory tests may be applied. ence between the fracture pressure during the first and
second load cycles. The tensile strength of a sedimen-
tary rock is only a small fraction of the compressive
11.3.2 Field Measurements of Horizontal
strength, and tensile strength can often be neglected.
Stress
Thus the fracture from the first load cycle may be used
to determine the horizontal stresses.
As it is difficult to predict the horizontal in situ stress
The best way to measure the fracture pressure dur-
condition in the basin, and as the horizontal stresses
ing the drilling phase of exploration is to perform a
may be very different in two orthogonal directions, it
series of “minifrac” tests. However, this is not nor-
is common to resort to field measurements. The maxi-
mally done, and it has become common practice in the
mum horizontal stress is termed σ H while the mini-
petroleum industry to perform a simpler measurement
mum horizontal stress is called σ h. If the vertical stress
that is called a “leak-off” test (Fig. 11.3). This is a
is the major principal stress σ 1, the two horizontal
pressure test in the well which is closed using the
principal stresses are σ 2 and σ 3, respectively.
blow-out preventer valves. After the string of drill
To measure these two horizontal stresses and their
casing is set and cemented in the well, a leak-off test
orientation one may use so-called hydrofracturing
is normally run after a few metres of hole are drilled
tests in a borehole (e.g. Goodman 1989, Fjaer et al.
below the drill shoe. Mud is pumped into the well
2008). For a typical situation where the major princi-
through the string using the cement pump of the drill
pal stress in the sediment is vertical, a vertical radial
rig. Return flow is prevented by cementing the casing,
fracture will open in the wall of a vertical borehole
and the mud pressure is recorded as a function of time
when the fluid pressure in the borehole is increased.
and injected volume. Before any fracture is opened,
This is because the fracture will be oriented perpen-
little or no mud is leaked into the sediment formation
dicular to the direction of lowest effective stress which
and the pressure simply increases as a linear function
is in the horizontal direction. If the maximum and
of the volume of mud injected. When the first fracture
minimum horizontal stresses are different, the
σv
σh + τ
Fracturing Mud pressure

Pressure
τ = tensile strength
FP Well
LOTP
Opening of very small fractures increasing the permeability to accommodate
the fluid flux. These very thin fractures are probably able to close again
almost perfectly. The permeability produced by microfracturing is a dynamic
variable controlled by the fluid flux.
Rate of mud injection
Fig. 11.3 Principle of leak-off test (LOT). The leak-off pres- subsiding sedimentary basins the horizontal stress is usually
sure must be higher than the lowest stress; fractures develop lowest and the fractures will be vertical
perpendicular to the direction of minor principal stress. In
is produced the pressure increase is reduced relative to

the injected volume of mud, thus changing the slope of
σH σh
the curve. This leakage of mud into the formation is an
indication that a thin fracture(s) has formed and that
the fracture pressure has been reached (hence the name
leak-off test). When repeated, the leak-off test fractur-
ing starts a little earlier because now τ ¼ 0 across the
fracture that was created during the first load cycle.
The results are often very reproducible which
indicates that the rock was not seriously damaged σH
during the first test, and that the small fractures pro-
duced probably closed again rather efficiently. The σh
location of the fracture(s) may be determined by mod-
ern formation imaging tools called FMI. The location Fig. 11.4 Borehole break-out and the orientation of principal
stresses. σ H is the highest horizontal stress and σ h is the lowest
of the radial fracture gives the direction of the mini-
horizontal stress
mum principal stress as the latter is normal to the
fracture (tangential to borehole wall at that location).
The small fractures produced during a leak-off test
may resemble the fractures formed during natural may escape from the pores and allow the volume
hydrofracturing at the top of an overpressure compart- change to occur.
ment in a sedimentary basin.
The directions of the maximum and minimum hor-
izontal stresses may also be determined from borehole 11.4.1 Concepts from the Theory of
break-outs as shown in Fig. 11.4. The maximum com- Elasticity
pressive stress in the wall of the borehole occurs at
each end of a diameter normal to the maximum hori- Elastic material behaviour, linear or non-linear, means
zontal stress direction. If the tangential stress is high that all strains (volume change and shear distortion)
enough to cause a compressive failure, a break-out caused by a stress change are recovered when the
occurs. By locating the break-out, one can determine stresses return to their original condition. If the grain
the direction of the maximum horizontal stress. In skeleton (framework) of a sedimentary rock may be
practice the location may be determined by calliper considered linearly elastic and isotropic for very small
measurements. The calliper measures the width of the deformations, the deformational characteristics may
borehole by recognising an oval shape. Modern for- be defined by the theory of elasticity.
mation imaging tools (FMI) can also be used to see the Youngs modulus (E) is the ratio between the
borehole break-outs. increase in normal stress and the resulting strain in
the stress direction, when there is no change in the
orthogonal normal stresses:
11.4 Deformation Properties of
Sedimentary Rocks E ¼ σ z =εz (11.11)
The fluid in the pores of the sediment compresses, but where σ z is the applied stress and εz is the compressive
if there is no gas in the pore fluid, this effect is very strain in the z-direction, and Δσ x ¼ Δσ y ¼ 0.
small and insignificant when computing volumetric Poisson’s ratio (ν) is defined as εx =εz in this situation.
compaction in sedimentary basins. It is the compres- x is the strain in the x-direction and is equal to y if the
sion bulk modulus of the grain structure that will material is isotropic. For a linearly elastic, isotropic
govern the volumetric deformations, and the perme- material, only two constants (for instance E and ν) are
ability and neighbouring drainage boundary required to fully define all the deformation
conditions that will govern how quickly the fluid characteristics.
The bulk modulus (K) is defined as the ratio coupling between the effects of normal stresses and
between the increase in equal-all-round stress and the shear stresses and between normal strains and shear
resulting volumetric compression (compaction): strains. For instance, if an element is only subjected to
shear stresses, there will not be any normal strains. As
K ¼ Δσ=εvol (11.12) discussed below, the behaviour of sediments and sedi-
mentary rocks is not that simple, and there can be
For an isotropic material the bulk modulus is significant volume expansion (dilatancy) or contraction
K ¼ E=3ð1 2vÞ. The shear modulus (G) is the ratio even if only shear stresses are applied. This phenome-
between the increase in shear stress and the resulting non is also related to the degree of overconsolidation, as
shear strain (angular change due to deformation). For the higher the overconsolidation ratio, the more signifi-
an isotropic material it may be shown that cant is the degree of shear dilatancy.
G ¼ E=2ð1 þ vÞ.
For a uniaxial strain compaction situation with
strain only in the vertical direction and no lateral strain
11.4.2 Non-linear, Inelastic Behaviour in
allowed, the compaction (compression) modulus (M)
Uniaxial Strain Compression
may be expressed as M ¼ Eð1 vÞ=ð1 þ vÞð1 2vÞ.
As stated in Sect. 11.3.1, the compaction process in a
The stress–strain results from a saturated sediment
fairly homogeneous and wide sedimentary basin may
tested in uniaxial strain compression are shown in
be considered uniaxial. This is not the case in a narrow
Fig. 11.5a. The starting point (I) for the curve
basin or where there are abrupt changes in depth of
represents the initial state of stress, and it is assumed
basin, lithology and material compressibility, for
that initially the specimen is normally consolidated.
instance due to faulting and block rotations.
The uniaxial compression modulus increases with the
The strains induced by the transmission of seismic
level of applied vertical effective stress. We therefore
signals are so small that linear elastic behaviour may
use the term tangent modulus (Mt) and/or secant mod-
be assumed. However, anisotropic deformation
ulus (Ms) to represent the behaviour. The tangent
characteristics should be allowed for when the seismic
modulus gives the slope of the curve at any specified
signals are used to derive deformation characteristics
stress level, while the secant modulus gives the slope
of the sedimentary rocks. A special type of anisotropy,
of the secant between two stress levels, commonly
which is a useful extension of the isotropic material
between the initial point I and point A. At stress
theory, assumes the sedimentary rock to be trans-
level A the specimen is unloaded to the initial stress
versely isotropic. This implies that the elastic
level (point B). As may be seen from the figure, a
properties are equal for all directions within a plane,
significant irrecoverable strain has been accumulated,
but different in the other directions. Transverse isot-
given by the horizontal distance IB. The specimen is
ropy may be considered to be a representative symme-
then reloaded back to A and up to a higher level, point
try for horizontally layered sedimentary rocks. The
C. It is found that the curve from A to C is a natural
properties are assumed isotropic, but different, in the
elongation of the curve from I to A. From I to A and A
vertical and horizontal planes. Five elastic constants
to C the specimen is normally consolidated, while
fully define all the deformation characteristics for such
during the unloading/reloading sequence it is
a material.
overconsolidated. It should be noted that linearly elas-
In general, sedimentary rocks cannot be treated as
tic loading and unloading behaviour would be
linearly elastic materials because the normal and shear
represented by only one common straight line in this
strains are not recovered when the element is
diagram, from I to C.
unloaded. However, the modulus concepts from elas-
Extensive laboratory testing of different types of
ticity theory, as outlined above, are very useful even
sediments has shown that the tangent modulus may be
when analysing the more realistic non-linear
determined by the following general expression:
behaviour of such rocks, including permanent (plastic)
deformation. 0 1 a
It should be pointed out that for a linearly elastic σ
M t ¼ m p0 v (11.13)
ideal material like the one described above, there is no p0
σv
C
σv
A
Laterally confined.
I No lateral strain allowed,
σv
B i.e. εH = 0
εv
b
P
σ1
A
R
σ3 σ3
I
σ2 = σ3
B σ1
ε1
Fig. 11.5 Stress–strain behaviour of sedimentary rocks. (a) sediment under triaxial compression with no restraint of lateral
Stress–strain behaviour of sediment during uniaxial loading, strains; P is the peak strength and R is the residual strength
unloading and reloading. (b) Stress–strain behaviour of
where p0 is a reference stress to make the ratio inside the reservoir chalk at the Ekofisk Field in the North
the parenthesis non-dimensional (often p0 is set equal Sea. When the effective vertical stress was increased
to 0.1 MPa in the published literature), and “m” and by reducing the fluid pressure in the reservoir, it
“a” are non-dimensional coefficients depending on the reached a level at which the coccolith structure
type of sediment and its geological history. To fit the (framework) of the chalk collapsed and caused large
different experimental curves, the exponent “a” is vertical strains, reservoir compaction and seafloor sub-
found to lie between 0 and 1. For a normally sidence. Subsequent injection of seawater maintained
consolidated clay sediment a ¼ 1 and this fits quite the fluid pressure in the Ekofisk reservoir and avoided
well, and the effective stress σ 0 v gives an Mt further increase of effective stresses. The rate of com-
value which is directly proportional to σ v. For paction was reduced, but not stopped, because it was
overconsolidated clay sediments (weak claystones) later found that the seawater weakened the chalk
an a-value of 0 gives a fair approximation, and that framework and increased the compressibility.
corresponds to an Mt which is constant as given by the
almost straight lines for the unloading and reloading
stress–strain curves in Fig. 11.5a. 11.4.3 Non-linear, Inelastic Behaviour
When the effective vertical stress is increased much When Approaching Shear Failure
beyond the C-level, one may find that the stress–strain
curve bends rather sharply over to the right (see dotted An element in a state of true uniaxial strain compres-
curve in Fig. 11.5a) before it again starts to rise for a sion (Sect. 11.4.2) can undergo large vertical compres-
further increase in the vertical effective stress. In a sion (compaction), but it cannot fail in shear by
sand this is caused by crushing of the coarser sand creating failure planes and fractures. This is prevented
grains because the intergranular contact stresses by the lateral confinement of the element (one-
become so high (Chuhan et al. 2002, 2003); this is dimensional compression).
further discussed in Sect. 11.5. A similar phenomenon However, consider now a cylindrical specimen of a
happens in sediments with a cemented but open and saturated sediment/sedimentary rock as shown in
porous structure. This was clearly demonstrated for Fig. 11.5b. The specimen is first subjected to an axial
stress (σ 1) and a radial stress (σ 3) representing the strain test. This is done by adjusting the horizontal
initial in situ stress condition. The initial pore pressure stress at all stages of the test such that no horizontal
in the specimen is also the same as the one in situ and strain is allowed to occur. This is called a K0 test.
thus the initial effective stresses. Then the axial stress
is increased while the lateral stress is kept constant.
The loading is performed so slowly that any tendency
11.4.4 Brittle Versus Ductile Stress–Strain
to overpressure in the pore fluid is avoided by allowing
Behaviour
the fluid to drain out of the specimen (dissipate), so
that the pore pressure remains at its initial value. In
For some sedimentary rocks, and depending on the
geomechanics this is called a drained test, as compared
magnitude of the lateral effective stress, the
to an undrained test in which the fluid is not allowed to
stress–strain curve in Fig. 11.5b may drop abruptly
drain and overpressures (positive or negative) build-up
after it has reached a peak at point P (the peak
during the loading.
strength). This is accompanied by a drop in shear
The recorded stress–strain curve for this axial load-
resistance with further strain down to a level R
ing is shown in Fig. 11.5b. Loading occurs from the
which is called the residual strength after the failure.
initial point I to point A. The initial section of the
The behaviour from the peak down to residual is
curve is fairly straight (linear), but as the axial stress
termed strain-softening or strain-weakening.
increases the curve starts to bend. The slope of the
For other sedimentary rocks there is no strain-
curve at any point is called the tangent Young’s mod-
softening, and the stress–strain curve is fairly horizon-
ulus (Et). The secant from point I to A gives the secant
tal or even slightly climbing as the strain increases.
modulus (Es) up to that stress level. If the axial stress
This is called ductile behaviour. A sedimentary rock
at point A is reduced down to the original axial stress
found to behave as a brittle material at low horizontal
level, the unloading curve goes down to point B and an
effective stress, may well behave as a ductile material
irrecoverable strain given by the distance IB has been
when the effective horizontal stress becomes suffi-
accumulated. Upon reloading the curve climbs back
ciently high, i.e. the horizontal stress has reached the
up to point A. The slopes of the unloading-reloading
brittle-to-ductile transition stress level (e.g. Goodman
curves are very similar and close to the initial slope of
1989).
the curve I to A. When the axial stress is increased
Other factors also affect the degree of brittleness.
beyond point A, the curve continues to bend as the
The higher the temperature and the lower the stain
specimen approaches a shear failure condition. Both
rate, the less tendency for brittle behaviour. In this
the tangent and secant modulus decrease significantly.
connection it should be pointed out that the rate of
As the stress difference (σ 1 σ 3 ) becomes even
stress change in and around a reservoir during petro-
larger, so does the maximum shear stress in the speci-
leum production is orders of magnitude higher than the
men. The stress difference cannot exceed a certain
geological stress changes, except for earthquake
level (strength), as the specimen is not able to carry
occurrences. Some rocks behave very differently at
any more load, and large axial strains (and shear
low strain rates compared with at high strain rates.
strains) ensue.
Rock salt is brittle when loaded at a very high strain
If one were to start the test described above with a
rate, but flows like a viscous fluid over geological
higher horizontal effective stress level, the
time. Also carbonate sandstones and shales yield to
stress–strain curve would be steeper and climb higher.
stress in a ductile manner by mechanical as well as
For loose sediments the moduli and shear strength are
chemical compaction if the strain rate is low enough.
strong functions of the horizontal effective stress level,
therefore it is so important to be able to estimate the
effective stresses. For sedimentary rocks (shales and
sandstones) with strong cementation caused by chem- 11.5 Compaction in Sedimentary Basins
ical processes, the modulus and strength are also
influenced by the horizontal effective stress, though The deformation characteristics for sedimentary rocks
to a much smaller extent. Note that the triaxial test are complex as shown above. The strain rate in sedi-
shown in Fig. 11.5b may also be run as a uniaxial mentary rocks is normally rather low except during
tectonic deformation like faulting. During production With increasing compaction due to grain rearrange-
of reservoir rocks the stain rates may become high. ment or breakage, and cementation, the rock becomes
In sedimentary basins down to depths of 1.5–2 km less porous, less compressible and stronger. This pro-
mechanical compaction caused by the increase of cess increases both the number and area of grain
effective vertical stresses is the dominating compac- contacts.
tion process and has the greatest influence on the Loose uncemented sands subjected to shear defor-
sediment’s hydro-mechanical properties. However, at mation may develop thin deformation bands. Such
greater depths where temperatures are higher shear bands may be composed of densely packed
(>70–80 C), it is mainly chemical compaction that grains where smaller silt-sized grains have been
contributes to the volume change and to the hydro- packed between larger grains during the shearing. If
mechanical properties through the effects of dissolu- there is little clay the shear strength of the shear band
tion, precipitation and cementation. will exceed the shear strength of the matrix and the
shear deformation will shift laterally to an area where
there has been no strain (deformation). This is called
11.5.1 Sands and Sandstones strain-hardening and results in a network of shear
bands which have only been subjected to small offsets.
The effective stress of the overburden is transmitted Large displacements recorded on seismics may in
through a framework of load-bearing grains. These reality consist of a broad zone of deformation bands.
grain-to-grain contact stresses may become very Shear deformation in sand may also result in grain
much higher than the average effective stress. Not all crushing. Experimental deformation of sand suggests,
grains will be subjected to high stresses because they however, that the effective stress normal to the shear
may be shielded by other grains within the load- band must be at least 10 MPa for coarse sand to be
bearing grain framework. The stress from the overbur- crushed. In the case of normal faults the horizontal
den will thus be concentrated on these other grains, stress must have been 10 MPa and the vertical stress
which then may fracture. The small areas of grain 20–25% higher, which corresponds to burial depths of
contact make the stress at these contacts very high, 1–1.5 km.
even at moderate burial depth. Precipitation of quartz or other cements increases
Natural sand grains of quartz and feldspar are the stiffness of sand and reduces its compressibility,
mostly blocky rather than spherical and have an irreg- transforming loose sand into indurated sandstones.
ular surface. This means that the area of contact is Only relatively small amounts of quartz cement, prob-
likely to be very small even for the larger grains, ably only 2–4%, are required to effectively stop the
resulting in much higher contact stresses than for mechanical compaction that is due to rearrangement of
smaller grains. If sand grains had been perfectly spher- grains. As a result the velocity, and particularly the
ical the contact would be controlled by the elasticity, shear velocity, will increase sharply for a modest
and the stress would then be independent of the grain reduction in porosity.
size. Sandstones may behave as if they were
Experimental compaction of well sorted sand overconsolidated due to cementation (chemical com-
reveals that coarse-grained sand aggregates are paction) and will only compact following the stress-
subjected to significant grain fracturing and compac- strain curve for overconsolidated rocks. We must dis-
tion at 20–30 MPa effective stresses while fine-grained tinguish between overconsolidation due to previously
sand does not fracture, and compacts much less at the higher effective stresses and “pseudo overconso-
same stress levels (Fig. 11.6) (Chuhan et al. 2002, lidation” caused by cementation and chemical com-
2003, Bjørlykke et al. 2004). paction. This because the highest effective stress must
Well sorted sand (e.g. beach sand) with an intitial be estimated from the burial curve and the pore pres-
porosity of 40–45% may compact mechanically to sure while the chemical compaction may be rather
30–38% porosity, depending on the stress level and insensitive to changes in stress.
the grain size. Poorly sorted sand and sand with high Chemical compaction involving the dissolution and
mud contents will compact much more at even lower precipitation of quartz is controlled mainly by temper-
stresses. ature because the rate of quartz cementation seems to
be the rate-limiting step rather than the effective stress. During production of a reservoir, reduced pore
The mineral cement growing as overgrowth on detrital pressure and higher effective stress will cause com-
quartz grains into the pore space between the grains is paction of the reservoir rocks. In the case of well-
not subjected to stress at the time of cement precipita- cemented reservoir sandstones this effect is very
tion and is not influenced by the stress. However, after small, but in shallow reservoirs with loose sand such
further compaction it may become a part of the load- compaction can be significant.
bearing structure. Therefore the grain-to-grain contact
stresses in a sandstone may be lower at depths of
3–4 km than at 1–2 km because the stress is distributed 11.5.2 Compaction of fine grained
over a larger grain contact area and is also supported sediments
by quartz cement. Most mechanical grain crushing
therefore occurs at depths shallower than about 11.5.2.1 Clays and Mudstones
2–3 km where there is little quartz cement present. In Clays, mudstones and shales have very different phys-
the case where quartz cementation (overgrowth) is ical properties compared to coarser-grained sediments
prevented by coatings (e.g. chlorite), grain fracturing like siltstones and sandstones. The physical properties
may occur at 3–4 km at about 35–40 MPa effective of clays depend not only on the strength of the sedi-
stresses. Grain fracturing will then expose uncoated ment particles (mostly clay minerals), but also on their
fresh quartz surfaces for quartz cementation which surface properties and chemical bonds which are con-
gradually will reduce the porosity (Fig. 11.6). trolled by the composition of the pore fluids. Clays,
Porosity (%)
24 26 28 30 32 34 36 38 40 42 44
0
5
Mono-quartz sand
30 MPa (0.63mm) 200 μm 10
Coarse-grained sand (30 MPa)
15
Vertical effective stress
20
25
30
35
40 Fine-grained sand (30 MPa)
45
Mono-quartz sand 0.71 mm 0.1 mm
40 MPa (0.63mm) 200 μm 50
Coarse-grained sand (40 MPa)
Fig. 11.6 Experimental compaction of loose sand grains (after contacts and more stress per grain contact in coarse-grained
Chuhan et al. 2002). Coarse-grained sand is more compressible sand, resulting in more grain fracturing
than fine-grained sand. This is because there are fewer grain
mudrocks and shales may vary greatly and have very several million years. We can determine the resultant
different physical properties, depending on the clay rock properties by analysing natural rock samples in
mineral composition and on the content of silt and the laboratory. It is nevertheless still difficult to esti-
sand. Mudstones and shales may have high contents mate the effective stress and temperatures to which
of silt and sand and also carbonate and may from these rocks have been subjected. This is particularly
transitions into poorly sorted sandstones and impure true in the case of samples exposed on land after
limestones. substantial uplift. Samples from offshore wells in sub-
The most common clay minerals are smectite, illite, siding basins are much better constrained with respect
chlorite and kaolinite. When the silt and sand content to the burial history but representative mudstones are
exceeds 40–50% there may be a grain-supported struc- rarely cored. Cuttings can be analysed mineralogi-
ture, and this marks the transition into clay-rich cally, but it is difficult to test their mechanical
siltstones and sandstones. A sediment of sand and silt properties without reconstituting the samples.
grains floating in a matrix of clay has for many purposes There is often a need to predict the compaction of
the same properties as the clay fraction, but the density sediments (soils) including clays in an engineering
is higher and so is the seismic velocity. This is because context and they are then tested in the laboratory to
the stiffness is determined by the load-bearing grains. measure the strain (compaction) as a function of effec-
Poorly sorted clays like glacial clays compact read- tive stress and other soil and rock mechanical
ily at relatively low effective stresses because the clay parameters. To simulate natural burial in sedimentary
particles are packed in between the silt and sand basins we use rather high stresses, up to 50 MPa or
grains. Data on compaction of clay and mudstones more, corresponding to 4–5 km of overburden. In most
may be obtained from measuring the degree of com- cases, though, chemical compaction becomes domi-
paction where the burial history and maximum effec- nant at shallower depth.
tive stress can be estimated, or by experimental By testing artificial mixtures of clays we can mea-
compaction in the laboratory. sure their physical properties as a function of clay
In fine-grained sediments like clays and mudstones mineralogy and silt and sand content. Kaolinitic
the total overburden stress is distributed over a very clays compact much more readily than smectite,
large number of grain contacts and the stress per grain which is the most fine-grained clay mineral and has
contact may be quite low. Relatively small amounts of very low compressibility (Mondol et al. 2007). At
minerals precipitated as cement between the primary about 20 MPa effective stress, corresponding to
grains can then cause a very significant increase in about 2 km of burial, pure smectite has more than
stiffness and seismic velocity even at shallow burial. 40% porosity while kaolinite has less than 20%
Most commonly this involves carbonate cement, (Fig. 11.7). Even at 50 MPa corresponding to
which makes soft clay grade into marls and calcareous 4–5 km burial depth at hydrostatic pressure the poros-
mudstones with higher bulk moduli. Quartz cementa- ity may still exceed 40%. Clay minerals compact more
tion requires higher temperatures (>80 C) when wet than dry (Mondol et al. 2007). This is
corresponding to 2–2.5 km in basins with normal probably because the friction between the grains is
geothermal gradients. higher in dry clays (Fig. 11.7).
In the case of smectite the large surface area and the
11.5.2.2 Experimental Compaction of Clays water which is bound to these clay surfaces make it
Experimental compaction of clays is difficult. It difficult to define the proportion of free water, and
requires very careful sample preparation, and compac- hence determine the exact porosity, which will also
tion tests up to 50 MPa stress may take 5–6 weeks for depend on the composition (electrolytic strength) of
smectite-rich clays. This is because the permeability is the porewater. Clay minerals tend to have a negative
so low that it takes a long time for the excess water to charge, causing repulsion between the clay particles.
drain. Time is also required to allow for the slight However, the negative charges will adsorb cations like
compaction at constant stress (creep) which may also Na+ and K+, thus neutralising this repulsion. This is
be referred to as secondary compaction. what causes flocculation when river-borne clays enter
Compaction of mud to mudstones and shales is the the sea. Marine clays therefore have a more stable clay
result of natural processes during burial, usually over mineral fabric and higher shear strength than
Porosity (%)
0 20 40 60 80
0
10
Vertical effective stress (MPa) Porosity

BS at 20 MPa
20
Kaolinite: 20% (BS)/ 38% (dry)
Smectite: 42% (BS)/ 53% (dry)
30
Kaolinite (100%)
Kaolinite (80%), Smectite (20%)
40 Kaolinite (60%), Smectite (40%)
Smectite (100%)
50
Fig. 11.7 Experimental mechanical compaction of dry (in at 20 MPa effective stress of dry and brine-saturated pure
grey) and brine-saturated (in colour) clay aggregates under smectite and kaolinite mixtures is shown
uniaxial compression strain (after Mondol et al. 2007). Porosity
freshwater clays. Slow weathering and leaching of 11.5.2.3 Chemical Compaction of Clays and
saltwater (Na+, K+) from marine clays uplifted by Mudstones
glacial unloading in Scandinavia may therefore lead Clays compact mechanically at shallow depth and
to slope instability and “quick clay” slides. Experi- at temperatures below 70–80 C, but at higher
mental compaction of clays confirms that their com- temperatures compaction may be controlled by chemi-
pressibility and shear strength are a function of the cal reactions. Chemical compaction must have thermo-
salinity of the porewater. dynamic drive so that less stable minerals dissolve and
Smectitic clays are less compressible than kaolin- more stable minerals precipitate. Clay minerals like
itic clays and this may be partly due to chemical bonds smectite become unstable and are replaced by mixed-
and partly because of the fine grain size. The negative layer minerals and illite. The silica released by this
charge of the clay minerals also causes water to be process must be precipitated as quartz cement for this
bound to the mineral surfaces by the positive charge of reaction to proceed and this causes marked stiffening
the dipole of the water molecule. This is important in and higher velocities. Compaction is then mainly con-
the case of smectite which has a surface area of several trolled by temperature rather than effective stress.
hundred m2/g but is not so significant for coarser clay Amorphous silica from volcanic sediments, and
minerals like illite, chlorite and kaolinite which have from amorphous silica (opal A) from fossils like
much lower surface area bound water. The total stress diatoms and siliceous sponges, will be a silica source
is divided by the number of grain contacts and the for precipitation of quartz cement even at low temper-
stress per grain is then less in smectitic clays. It has ature. Carbonate cements from calcareous fossils will
been shown experimentally (Mondol et al. 2008a) that also result in a marked increase in the stiffness and
fine-grained kaolinite is less compressible than coarse- velocity. In the absence of thermodynamically unsta-
grained kaolinite (Fig. 11.8). Similarly, well sorted ble minerals like smectite, mudstones may remain
fine-grained sand is less compressible than coarse- nearly uncemented to greater depth. At about 130 C
grained sand (Chuhan et al. 2003). Smectitic clays kaolinite becomes unstable in the presence of K-
are characterised by low velocities and low density feldspar and causes precipitation of illite and quartz.
because they have high porosity (Fig. 11.9) (Mondol Gradually, however, mudstones become harder and
et al. 2008b). develop a schistocity, becoming a shale. The cleavage
0 σv = pow(ϕ, −3.79) * 6540017.18

R2 = 0.87
100
Cumulative (%)
10
80
60
20 40
< 2 micron 20
2–8 micron 0
30 > 8 micron 0.1 1 10 100
Composite mixture Grain size (mm)
Best fit line
40 < 2 μm
Composite mixture
2−8 μm
50 > 8 μm
10 20 30 40 50 60
Porosity (%)
Fig. 11.8 Experimental mechanical compaction of brine- mixtures. The maximum porosity reduction is observed in the
saturated kaolinite aggregates sorted by grain size (after Mondol composite mixture containing all the grain sizes, demonstrating
et al. 2008). The sample containing less than 2 μm sized kaolin- the importance of both grain size and sorting for the rock
ite aggregates retained higher porosity compared to all the other properties
a c
2400 6
Vp (m/s)
Vp/ Vs
2000 4
1600
2
0 10 20 30 40 50 0 10 20 30 40 50
Vertical effective stress (MPa) Vertical effective stress (MPa)
b 1200
Kaolinite (100%)
Smectite (20%), Kaolinite (80%)
Vs (m/s)
800
Smectite (100%)
Generalized best fit lines
400 Best fit lines for specific sample
0 10 20 30 40 50
Fig. 11.9 Crossplots of ultrasonic velocities vs. vertical effec- compared to pure kaolinite. The Vp =Vs ratio is higher in pure
tive stress of brine-saturated smectite-kaolinite (modified after smectite than in pure kaolinite (c)
Mondol et al. 2008b). Pure smectite has lower Vp (a) and Vs (b)
is produced by pressure solution of quartz and other the reorientation of clay minerals during diagenesis so
minerals in contact with layers enriched in clay that the velocity parallel to bedding will be higher than
minerals. perpendicular to bedding. In addition the resistivity
The transition from mudstones to shales does not will to a large extent be controlled by the orientation
only involve a marked increase in stiffness and veloc- of the clay minerals and the quartz cementation in
ity but also an increase in the anisotropy. This is due to mudrocks and shales.
11.6 Summary fracturing of grains. This is more pronounced in

coarse-grained rather than fine-grained sands and
In subsiding sedimentary basins, mechanical compac- may account for a significant component of the poros-
tion caused by the increase of effective vertical ity reduction. After the grain crushing and permanent
stresses is the dominating compaction process and collapse deformations have occurred, the grain size is
has the greatest influence on the sediment’s porosity reduced and the reservoir regains stiffness. With
and hydro-mechanical properties down to depths of increasing depth such grain crushing is less likely
2–2.5 km. At greater depths where temperatures are due to increased cementation of the grain structure
higher (>70 C), it is mainly the chemical compaction caused by chemical processes. In most sandstone
that contributes to the volume change and to the reservoirs quartz cementation starting at 2–2.5 km
hydro-mechanical properties due to the effects of dis- (70–80 C) will stabilise the grain framework and pre-
solution, precipitation and cementation. Carbonate vent further mechanical compaction. Sandstone
rocks, though, may undergo chemical compaction at reservoirs with a critical content of quartz cement
shallower depths. (>2–3%) will therefore experience very little compac-
The magnitude and distribution of stresses in sedi- tion even if the effective stress is increased during
mentary basins are important in relation to petroleum production. Similar compaction by grain breakage
exploration, production and reservoir management. may also occur for high effective stresses in a reservoir
Knowledge of in situ stress is also important in con- where the framework of the sedimentary rock is very
nection with drilling, particularly during deviation and porous with little cement (e.g. the chalk in the Ekofisk
horizontal drilling. The magnitude and orientation of reservoir, North Sea). Here this led to very large res-
stresses affect the propagation and interpretation of ervoir compaction and subsequent seafloor subsidence
seismic signals, particularly the S-waves, through sed- (c. 10 m).
imentary rocks. Smectitic clays are characterised by very high Vp =Vs
The total vertical stress (σ v) of a rock sequence is ratios when compared with other clays (Fig. 11.9c).
carried partly by transmission of stress in the solid This implies that in a sequence of mudstones, smectitic
0
grain framework (effective stress σ v) and partly by clays will stand out with very different characteristics
the pressure in the fluid phase (porewater or compared with kaolinite and probably also illite-rich
petroleum). sequences, which have much lower velocities and
Determination of effective stresses depends on reli- Vp =Vs ratios. At temperatures above 70–80 C smectite
able estimates of the fluid pore pressure which often is in will no longer be stable and mudstones will be more
excess of hydrostatic (overpressure). There exist semi- influenced by chemical compaction and cementation. In
empirical relationships for the ratio between horizontal cold basins, however, mechanical compaction can be
and vertical effective stresses, but reliable estimates of dominant down to 4–5 km burial depth. Every mud-
horizontal stresses depend on field measurements like stone has a unique compaction curve which will depend
hydraulic fracturing tests. It is common practice in the on a number of factors such as mineralogy, grain size,
petroleum industry to use a simplified procedure (leak- pore fluids, pore pressure, pore aspect ratio, etc.
off tests) to determine the magnitude and orientation of Well log data from well sorted Jurassic sandstone
the minimum horizontal stress. like the Etive Fm in the North Sea basin show that
The virgin (in situ) distribution of stresses in sedi- natural compaction curves agree well with experimen-
mentary basins is the result of both mechanical and tal mechanical compaction in the laboratory (Fig 4.16).
chemical compaction, usually over geological time. The physical properties of silt and clay mixtures
Changes in stresses during petroleum production from depend strongly on both the clay mineralogy and the
a reservoir are much more short term and are mainly content of silt and sand. (Manzar et al. 2010). This
mechanical. In a carbonate reservoir the chemical pro- includes porosity and velocity and also anisotrophy as
cesses may be so fast that also chemical compaction measured in the laboratory. The properties of
may become significant at that time scale. mudstones are also influenced by carbonate and silica
The effective stresses in sand may cause grain-to- cement.
grain contact stresses which are so large that compres- The effect of time on mechanical compaction will
sion (compaction) may occur due to crushing and always be difficult to evaluate in the laboratory but
prolonged compaction tests suggest that compaction porosity reduction during burial in sedimentary basins.
due to creep is rather small. At depths shallower than Canadian Geotechnical Journal 40, 995–1011.
Dickey, P.A. 1979. Petroleum Development Geology. Petro-
about 2 km, high porosity (>30%) is preserved in leum Publishing Co., Tulsa, OK, 398 pp.
Upper Palaeozoic sandstones despite long burial time. Fjær, E., Holt, R.M., Horsrud, P., Raaen, X. and Risnes, R. 2008.
In spite of the limitations, laboratory porosity/den- Petroleum Related Rock Mechanics 2nd ed. Developments
sity/velocity-stress relations and their comparison in Petroleum Science 53, 491 pp.
Gueguen, Y. and Palciauskas, X. 1994. Introduction to Physics
with data found in well logs will provide important of Rocks. Princeton University Press, Princeton, 294 pp.
constraints on evaluation of burial depths, pore pres- Goodman, R.E. 1989. Introduction to Rock Mechanics. Wiley,
sure prediction and/or the amount of uplift and erosion New York, 562 pp.
found in sedimentary sequences (Mondol et al. 2007, Manzar, F., Mondol, N., H. Jahren, J. and Bjørlykke, K. 2010.
Microfabric and rock properties of experimentally com-
2008b, c,). pressed silt-clay mixtures. Marine and Petroleum Geology
27, 1698–1712.
Mavko, G., Mukerji, T. and Dvorkin, J. 1998. The Rock Physics
Further Reading Handbook. Cambridge University Press, Cambridge, 327 pp.
Mondol, N.H., Bjørlykke, K., Jahren, J. and Høeg, K. 2007.
Experimental mechanical compaction of clay mineral
aggregates – Changes in physical properties of mudstones
Barton, N. 2007. Rock Quality, Seismic Velocity, Attenuation and
during burial. Marine and Petroleum Geology 24(5),
Anisotrophy. Taylor and Francis/Balkema, London, 729 pp.
289–311.
Bjørlykke, K. and Høeg, K. 1997. Effects of burial diagenesis on
Mondol, N.H., Bjørlykke, K. and Jahren, J. 2008a. Experimental
stresses, compaction and fluid flow in sedimentary basins.
compaction of kaolinite aggregates: Effect of grain size on
Marine and Petroleum Geology 14, 267–276.
mudrock properties. 70th EAGE Conference & Exhibition,
Bjørlykke, K. 2006. Effects of compaction processes on stress,
Extended Abstract I037, June 2008.
faults, and fluid flow in sedimentary basins. In: Buitersand,
Mondol, N.H., Bjørlykke, K. and Jahren, J. 2008b. Experimental
S.H.J. and Schreurs, G. (eds.), Analogue and Numerical
compaction of clays. Relationships between permeability
Modelling of Crustal-Scale Processes. Geological Society
and petrophysical properties in mudstones. Petroleum Geo-
Special Publication 253, pp. 359–379.
science 14, 319–337.
Mondol, N.H., Fawad, M., Jahren, J. and Bjørlykke, K. 2008c.
Synthetic mudstone compaction trends and their use in pore
compaction during burial in sedimentary basins In:
pressure prediction. First Break 26, 43–51.
Stephansson, O., Hudson, O. and King, L. (eds.), Coupled
Mondol, N., Jahren, J., Bjørlykke, K. and Brevik, I. 2008d.
Thermo-Hydro-Mechanical-Chemical Processes in Geo-
Elastic properties of clay minerals. The Leading Edge 27,
systems. Fundamentals, Modelling, Experiments and
758–770.
Applications. Geo-Enginering Book Series, vol. 2, Elsevier,
Peltonen, C., Marcussen, Ø., Bjørlykke, K. and Jahren, J. 2008.
London, pp. 699–708.
Mineralogical control on mudstone compaction; a study of
Bjørlykke, K., Høeg, K., Faleide, J.I. and Jahren, J. 2005. When
Late Cretaceous to Early Tertiary mudstones of the Vøring
do faults in sedimentary basins leak? Stress and deformation
and Møre basins, Norwegian Sea. Petroleum Geoscience 14,
in sedimentary basins: Examples from the North Sea and
127–138.
Haltenbanken Offshore Norway – A discussion. AAPG Bul-
Voltolini, M., Wenk, H.-R., Mondol, N.H., Bjørlykke, K. and
letin 89, 1019–1031.
Jahren, J. 2009. Anisotropy of experimentally compressed
Bjørlykke, K. 2003. Compaction (consolidation) of sediments.
kaolinite-illite-quartz mixtures. Geophysics 74(1), 1–11.
In: Middleton, G.V. (ed.), Encyclopedia of Sediments and
Wiprut, D. and Zoback, M.D. 2000. Constraining the stress
Sedimentary Rocks, Kluwer Academic Publishers,
tensor in the Visund field, Norwegian North Sea: Applica-
Dordrecht, pp. 161–168.
tion to wellbore stability and sand production. International
Chuhan, F.A., Kjeldstad, A., Bjørlykke, K. and Høeg, K. 2002.
Journal of Rock Mechanics and Mining Sciences 37,
Porosity loss in sand by grain crushing. Experimental evi-
217–336.
dence and relevance to reservoir quality. Marine and Petro-
Zoback, M.D., Moos, D., Mastin, L. and Andersen, R.N. 1985.
leum Geology 19, 39–53.
Well bore breakouts and in situ stress. Journal of
Chuhan, F.A., Kjeldstad, A., Bjørlykke, K. and Høeg, K. 2003.
Geophysicsl Research 90, 5523–5530.
Experimental compression of loose sands: Relevance to
Chapter 12
The Structure and Hydrocarbon Traps of Sedimentary Basins
Roy H. Gabrielsen
12.1 Tectonic Regimes and Stress or purely structural (e.g. an anticline). To start with the
purely structural trap, this depends on the tectonic
Old knowledge told the early oil explorationist to focus environment (contraction, extension or strike-slip) in
drilling on structurally high-standing areas because which it was created (see Sect. 12.1.2).
hydrocarbons are trapped in positive structures like On a worldwide scale, the structural trap is by far the
folds and rotated fault blocks that, as such, define posi- most common hydrocarbon trap type, but there is a
tive structural features and associated topographic tendency for the two other trap types to be more relevant
highs. Principally, however, three categories of traps in mature hydrocarbon provinces, because exploration
are recognised, namely the structural trap, the strati- moves into a more detailed approach in such areas.
graphic trap and the hydrodynamic trap. Structural traps are generated by tectonic, diapiric,
Structural geology is significant in all stages of the compactional or gravitational processes. They must
“value chain” of the exploration for and the exploita- contain a reservoir constrained in three dimensions, a
tion of hydrocarbon resources, including the initial cap, and be in communication with a reservoir unit. The
exploration screening phase, the targeted exploration, structure has a spill-point defining the volume available
reservoir description and production (Gabrielsen & in the trap, and commonly develops a gas/oil contact at
Møller-Holst 1997). This involves studies that include the top and oil/water contact at its base.
the basin-scale tectonic evolution, interaction between Geometrically speaking, the simplest structural trap
structural evolution and sedimentation, maturation, may be the anticline, or modifications hereof (Fig.
hydrocarbon migration, characteristics of trap type, 12.1a). Compressive tectonic regimes commonly
sealing potential and leakage. But generally, the most include a host of contractional folds and thrusts.
common questions to ask a structural geologist in an Good examples of such areas that are also hydrocarbon
oil exploration environment concerns the trap itself. provinces are the Zagros Mountains of Iran, the Carib-
Many types of traps exist, some of which are deter- bean of South America and the North American
mined purely by the sedimentary development, Cordillera. To be of interest for the entrapment of
whereas the most common are due to structural devel- hydrocarbons, such structures must have a closure as
opment alone. Hence the purely stratigraphic trap is seen along the fold axis, implying that they should not
due solely to sedimentary processes, such as the pri- be strictly cylindrical. This is often the case for natural,
mary pinch-out of sedimentary units. Others may be contractional folds, particularly in situations where
due partly to sedimentological and partly to structural deformation took place above an uneven basement
developments (e.g. some types of inconformity traps), topography and particularly so in areas where double-
folding (folding with to sets of cross-cutting fold-axes)
and halokinesis have occurred (Fig. 12.1d). The anti-
cline must be thoroughly mapped in three dimensions.
R.H. Gabrielsen (*)
Department of Geosciences, University of Oslo, Oslo, Norway The complexity and communication between con-
e-mail: [email protected] tractional structures are, to a great extent, dependent on
320 R.H. Gabrielsen
3 4
1 2
6
5 7
9
8
8
10
11 12 13
3
1 4 5
Fig. 12.1 Most common structural trap types. (a) Trap types close genetic relation between contractional structural traps so
(folds and thrusts) associated with contraction. There can be a that one type (e.g. a symmetrical fold) may be an early stage of
12 The Structure and Hydrocarbon Traps of Sedimentary Basins 321
the deformation. Thus, a mildly deformed province may complex structural pattern and dismembering of the
contain sets of symmetrical buckle-folds (Figs. 12.1a and structural traps are typical for the more advanced
12.16b), whereas a province of advanced contractional stages of deformation, including diapirism.
deformation may be dominated by systems of complex, In the following sections, trap types and risks
overturned fault-propagation folds and thrusts associated with the different trap types are investigated.
(Figs. 12.1a and 12.15).
Complex folds are also common in strike-slip
regimes, developed either as fold-trains formed orthog- 12.1.1 Principal Stress Regimes and Types
onally to the contractional axis of the strike-slip sys- of Stress
tem (Figs. 12.1c and 12.14) in areas of constraining
bends (Fig. 12.14c), or in connection with transpressional The concept of plate tectonics offers a useful frame-
segments of the strike-slip system. Here, folds may work for structural geological analysis on all relevant
grade into flower structures. (Figs. 12.1a and 12.14c). scales in petroleum geology, from regional in the
Many petroleum provinces occur in extensional basin exploration stage, to local in the reservoir evaluation
settings, where the rotated fault block and affiliated and production stages. This is natural, because the
accommodation structures provide the most common principal geological stress systems are ruled by pro-
trap type (Figs. 12.1b, 12.8, 12.10 and 12.11). For exam- cesses in the deep Earth like mantle convection and
ple some of the major hydrocarbon fields in the Norwe- lithosphere subduction, the secondary effects of which
gian continental shelf, like the Statfjord and the Oseberg are manifested at the base of the lithosphere and along
fields, occur in such settings. The rotated fault block plate margins. Based on these concepts, the basic
itself becomes tilted according to the amount of exten- dynamics of the lithosphere can be quantified, which
sion and the geometry of the principal fault plane (planar is a prerequisite for the evaluation and calculation of
or listric). Secondary traps may occur due to elastic the state of stress at any point. As seen in the perspec-
uplift of the footwall, roll-overs in the hangingwall due tive of the petroleum structural geologist, understand-
to a strongly listric fault plane, or forced folds associated ing and quantifying the stress situation at the plate
with ramps in the detachment (Fig. 12.1e). margins is a prerequisite for understanding the state
Halokinesis (the processes associated with the salt of stress in any basin system and in any reservoir.
movements) occurs because salt has a specific gravity We term principal stresses originated at plate
of 2.2 g/cm3; as opposed to fully consolidated (low margins far-field or contemporary stresses. The stress
porosity) sedimentary rocks that have specific situation in a basin or a reservoir may be a sum of
gravities of 2.5–2.7 g/cm3), creating a buoyancy several far-field stresses combined with a local stress,
effect. This is further supported by the fact that on a which may be related to burial, erosion, geothermal
geological timescale, salt flows plastically. A number gradients, topography, basement relief and structural
of pronounced structural types result, from pillows to inhomogeneities in the substratum. In other words, the
diapirs of salt. Strong structuring may also be plate tectonic framework provides a basic and general
associated with mud diapirism. In a classical study of concept on which any structural geological analysis of
mild to extensive salt movements, Halbouty (1979) a sedimentary basin rests, but it must be supplied with
identified no less than nine different structural trap information on the local stress system that is
types (Figs. 12.1 and 12.17). The general experience superimposed on it. In the context of the far-field
is that larger structures are affiliated with early-stage plate tectonic stress, we distinguish between the plate
salt structures, like pillows, whereas a much more boundary and the intra-plate component.
Fig. 12.1 (continued) the development of fault propagation in strike-slip regimes. (c) Extensional, rotated fault-blocks
fold and a fully developed thrust. See Figs. 12.2 (for structural) above a decollement (floor fault). See Figs. 12.11 and 12.12
and 12.15 (for stratigraphic) traps associated with contractional for structural, stratigraphic and unconformity traps in such
or inverted terranes. (b) Typical structural trap type associated systems. (d) Different types of salt structures and associated
with strike-slip (positive and negative flower structures and trap types. (See also Fig. 12.16)
associated folds. See Fig. 12.13 for possible stratigraphic traps
322 R.H. Gabrielsen
The plate boundary stress is subdivided into three and destroyed throughout large parts of the history of
basic plate margin settings, namely constructive our planet, so that many may be preserved inside the
boundaries where adjoining plates are moving away present plates and hence do not coincide with present
from each other and new crust is formed by magmatic continent/ocean margins. The three basic types of
activity, destructive boundaries where lithosphere is plate margin coincide with the three principal stress
consumed by subduction or obduction, and conserva- configurations, namely the tensional, the compres-
tive boundaries where plates are moving past each sional and the strike-slip regimes. To describe these
other in a strike-slip sense and where lithosphere is regimes, we rely on the concept of principal stress
neither created nor consumed (Fig. 12.2a). It is impor- configurations and their definition in the context of
tant to note that plate boundaries have been generated the principal axes of stress.
a
North American
Plate
Eurasian Plate
Juan de Fuca North American

Iran
Plate Plate
Plate
Caribbean
Plate Philippine
Cocos Arabian Plate
Pacific
Plate Plate
Plate
Nazca
Plate Indo Australian
African Plate
South American
Plate
Plate
Antarctic Plate Scotia Plate Antarctic Plate
Convergent boundary Divergent boundary Transform fault
b σ1
σ3
σ1
σ3
σ2 σ2
σ2
σ3
σ1
Fig. 12.2 (a) The major tectonic plates and their boundaries (from Nystuen in Ramberg et al. 2008). Reprinted with permission
from the Norwegian Geological Society and the author. (b) Principal stress configurations (from Fossen and Gabrielsen 2005)
12.1.2 Tectonic Stress in the Earth’s

σ H > σ v > σ h ; strike-slip; (12.3)
Lithosphere
σ H > σ h > σ v ; contraction; (12.4)
As we have seen already, in addition to stresses
generated at tectonic plate contacts, several other pro-
where σ H and σ h are the maximum and minimum
cesses contribute to the generation of stresses. We
horizontal stresses, respectively. By using σ v as a
may, for example, have residual stress, which is
reference, further calculations become dependent on
inherited from previous deformation, where a rock
the reference system applied. In the contractional
body may have become bent with the elastic stress
regime there will be a tectonic component σ t in

component remaining unreleased, and thermal stress
addition to the reference stress, so that the greatest
which is related to expansion of rocks during heating
(horizontal) stress is:
or contraction during cooling and stress related to local
gravitational gradients. Accordingly, the total stress
σ H ¼ ρgz þ σ t (12.5)
consists of several components:
Assuming uniaxial stress, which implies that the
Total stress ¼ reference stress þ residual stress
reference stress is a function of the elastic properties
þ thermal stress þ tectonic stress of the rock, and ignoring the thermal expansion, we
can describe the stress as:
where the reference stress refers to the stress inside the
plate, devoid of plate tectonic stresses, and thus the h υ i
σH ¼ ρgz þ σ t (12.6)
1 υ
Tectonic stress ¼ contemporary stress þ local stress:
where υ is Young’s modulus (Chap. 11). Because
We will not go further into the analysis of residual ðυ=1 υÞ < 1, σ t > σ t . Accordingly, the reference
and thermal stress here, but concentrate on stress stress condition selected also influences the calculated
generated by primary interactions at plate margins or total tectonic stress as long as the buried rock is
forces derived thereof. compressive. For a non-compressive rock ðυ ¼ 0:5Þ,
Taking into consideration the reference stress lithostatic and uniaxial reference systems are equal.
conditions, one can describe the stress situations for
the three principal conditions of deformation, namely
extension, contraction and strike-slip. This was done 12.1.3 Deformation Mechanisms and
by Anderson in two influential works in 1934 and Analogue Models
1951, in which the framework for all modern tectonic
and structural geological analysis is defined. Anderson Our daily contact with the physical world tells us that
used the vertical lithostatic stress (σ v) as a reference: materials deform in many ways, depending on the
applied stress and type of material and its physical
σ v ¼ ρgz (12.1) state. Thus, a liquid reacts to outer stress very differ-
ently from a piece of rock, and one type of rock like
(where ρ is the specific weight, g is the constant of chalk has very different physical properties as
gravity and z is the height of the rock column). This compared to another like granite. Furthermore, one
applies for the three principal stress conditions material may change its mechanical properties dra-
because σ v can be considered similar for one rock matically by change of temperature and pressure.
type for a constant h, assuming that the burial history These contrasts are founded on processes occurring
has been similar. Thereby the three principal stress on the scale of the grains (of the rock) and on the
systems can be defined, each corresponding to a molecular and atomic scales. We have given these
regime of deformation (Fig. 12.2a, b): processes and their associated meso- and macroscopic
physical expressions names like brittle, elastic, plastic,
σ v > σ H > σ h ; extension; (12.2)
324 R.H. Gabrielsen
viscous and ductile, and sometimes combinations like settings. Thus, extensional stress regimes either are
elastico-plastic. associated with subsidence and basin formation (in
Unfortunately, these terms are not always used in a intra-plate settings) or characterise active break-up of
consequent manner and are therefore liable to cause continents (along constructive or passive margins).
confusion when taken out of context or not precisely Although the conditions for development of petro-
defined. This particularly concerns the term “brittle”, leum systems in areas of active spreading may be
because it is used in a double sense, namely as a meagre, the remnants of the earlier stages of break-
deformation mechanism and a deformation style. up, now situated in passive margins settings, fulfil all
When applied in the context of deformation the requirements that characterise productive petro-
mechanisms, brittle deformation implies that existing leum provinces. This is because such tectonic regimes
bonds are physically broken between mineral grains, have undergone crustal thinning and associated subsi-
or that fracturing of the individual grains themselves dence, which involves all the processes essential for
takes place. As a consequence, the rock loses its petroleum to be generated, trapped and accumulated in
cohesion and (potentially) physically falls apart. In sufficient volumes and concentrations for petroleum
contrast, the plastic deformation mechanism implies fields to be commercially interesting. Accordingly,
that deformation takes place by the transfer of such settings frequently display an attractive combina-
dislocations on the atomic scale. This means that the tion and distribution of source, reservoir and cap
mineral can change its shape without loss of cohe- rocks, structural and stratigraphic traps and the
sion. Generally, plastic deformation occurs at higher conditions for maturation, expulsion, migration and
p,T-conditions than those accompanying brittle accumulation of hydrocarbons.
deformation.
Concerning deformation style, the term brittle is
used about localised strain, like that associated with
12.2.1 Extensional Basins
jointing and faulting, and particularly in cases when
the rock loses its cohesion and where the deformation
The formation of extensional basins may be seen as
occurs at lower p,T-conditions (though not necessarily
the first stage of the Wilson Cycle, which begins with
so). The ductile deformation style characterises strain
thinning, stretching and rifting of the continental crust
also by low stress, which is homogeneously distributed
followed by continental break-up and mid-oceanic
over a wider area, as commonly observed in connection
spreading. The concept of the Wilson cycle predicts
with folding and meso- and mega-scale shear-zones. A
that this process becomes reversed, causing closure of
ductile style of deformation is predominant at high p,T-
the ocean, collision between the adjacent continental
conditions, but may also occur under very low p,T-
plates, and hence the construction of a mountain chain
conditions, if it involves weak materials like sand and
along the zone of collision (Fig. 12.3). The junction
clay. In such cases, however, displacement takes place
between the continental plates defines the suture
along grain boundaries or along borders between rock
between the two.
bodies and not by dislocation creep or other atomic-
If we use the present North Atlantic as one exam-
scale mechanisms that characterise the plastic deforma-
ple, the highly hydrocarbon-rich northern North Sea
tion mechanism.
basin system is situated in a passive continental mar-
gin configuration, where the extensional basin system
developed during continental break-up. In contrast,
12.2 Petroleum Systems in Extensional Iceland, where petroleum resources are less abundant,
Regimes is situated on the top of the mid-oceanic spreading
ridge. However, if one looks more closely at the struc-
Areas of extension are affiliated with horizontal diver- tural configuration at depth, one finds that the northern
gent stress and are found in association with construc- North Sea basin system, which includes the Viking
tive or passive plate boundaries and in intra-plate Graben that developed in Jurassic-Cretaceous times, is
Rift valley
Ancient contine
Rifting phase nt Heat flow from
Ancient contine the mantle
nt
Thrusting
direction for
nappes and small
thrust sheets
Continental lithospheric Asthenosphere
plate Continental volcanic zone
Spreading phase Continental shelf

Ocean
Active spreading ridge
Oceanic lithospheric plate
Island arc phase

Volcanic
Island arc
Epicontinental ocean
Island arc core with

gabbro and diurite
Closing phase Passive spreading ridge

Deep-sea trench Accretionary
wedge
Island arc mountain chain
Foreland
Accretionary wedge
Folding and thrusting
Mountain chain on a new continent

Collision phase Marginal zone
Marginal zone
New lithospheric plate
Fig. 12.3 The major stages in the Wilson Cycle (from Nystuen in Ramberg et al. 2008). Reprinted with permission from the
Norwegian Geological Society and the author
326 R.H. Gabrielsen
underlain by an older (Permo-Triassic) basin system in geosciences in modern times. This model assumes
(Fig. 12.4). The Permo-Triassic basin system is in turn thinning of the weak lower crust/lower lithosphere by
superimposed on the even older Caledonian orogeny, pure shear, and hence is characterised by the develop-
which was subsequently affected by gravitational col- ment of a symmetrical configuration (Fig. 12.5a). The
lapse in Devonian times, representing the last stage in pure-shear extension of the ductile lower crust is
a previous Wilson Cycle. accompanied by thinning of the upper crust by brittle
There are several models for the lithospheric faulting and subsequent development and rotation of
configurations that accompany extensional crustal fault blocks.
thinning, the end members of which are the “pure In this context it is possible to separate the active
shear” (symmetrical) and “simple shear” (asymmetri- stretching stage, which is associated with fault-
cal) models (Fig. 12.5). It should be noticed that these controlled thinning of the upper crust, and later subsi-
models are not necessarily mutually exclusive; we can dence controlled by thermal processes. As a response
find basin systems that display elements from more to extension, the crust and upper mantle lithosphere
than one model, such as the “delamination model” becomes thinned and, promoted by extensional
(Fig. 12.6). faulting, the basin floor will subside quickly. This
The “pure-shear model” for extensional crustal implies that deeply seated warm rocks are transferred
thinning was suggested by Dan McKenzie in 1978, upwards in the lithosphere so that the isotherms in the
in a paper that has become the most frequently cited thinned area become elevated and the thermal
Fig. 12.4 The architecture of the North Sea continental shelf. Note the deep structure, showing a Permo-Triassic rift system buried
beneath the younger Jurassic – Cretaceous Viking Graben (from Fossen 2002)
gradients become steepened accordingly. These deep contraction, sediment compaction and sediment load-
processes influence the relief of the basin floor because ing. This sounds complex, but luckily these processes
heating causes rock volumes to expand and elastic, are well understood and can be modelled with good
quasi-plastic and isostatic adjustments to occur simul- accuracy on the basis of the algorithms proposed by
taneously at lithospheric, basin (e.g. by uplift of the McKenzie and supplied with additional modelling
basin margins) and fault block scales. In the next stage tools, developed particularly in the late 1990s.
of development (the post-rift stage), the basin will For modelling purposes and for the analysis of
continue to subside due to a combination of thermal extensional basins with respect to petroleum explora-
tion, three stages of development can be distinguished
(Fig. 12.7).
a
Pure shear The pre-rift stage is characterised by gentle flexuring
and fracturing of the lithosphere. In some rifts we see
Moho the development of a gentle bulge, caused by mantle
doming and associated warming – and hence expansion
– of the lithosphere. In other cases, a gentle subsidence,
defining a broad, shallow basin is seen, caused by mild
b extension of the cold (not-yet-heated) lithosphere. In
Simple shear
both cases, the lithosphere is prone to develop steep
Moho fractures on a crustal or even lithospheric scale. These
fractures have the capacity to accommodate magma,
generating dikes. Regarding hydrocarbon reservoir
potential characterising the pre-rift stage, sand deposits
Fig. 12.5 Basic configuration of (a) pure shear and (b) simple are likely to be sheet-like and relatively thin, with few
shear extensional basins (modified from Fossen and Gabrielsen structural traps developing at this stage. Sediment trans-
2005) port is mainly transverse to the basin axis, but quite
Conceptual model - northern North Sea

NW East Shetland Basin Viking Graben Horda Platform Øygarden F.Z. SE
Base Cretaceous
0
Top
10 basement D1
D2 Top
Depth (km)
basement
20 ? ?
30 Moho High velocity body

(+6 km/s)
Mantle fault
40
250 200 150 100 50 0
Distance (km)
Moho Moho
Localised lower crustal shear zones? Distributed lower crustal deformation?
Fig. 12.6 Model of the Viking Graben, displaying elements of pure and simple shear. Modified after Odinsen et al. (2000)
328 R.H. Gabrielsen
a produce a variety of lithofacies due to the increasing

topographic relief associated with high fault activity.
The marine transgression that commonly follows the
increased subsidence of the basin floor also contributes
to this variety in sedimentary facies. Sand that is
eroded from the high-standing parts of the basin (e.g.
b
basin shoulders and crests of rotated fault blocks) may
be trapped in lows in various structural positions and
these units are likely later to be covered by transgres-
sive marine sediment accumulations. The sediment
transport system in the active stretching stage is likely
c
to be dominated by complex transverse and locally bi-
directional fluvial systems that are strongly influenced
by the elongated, rotated fault blocks, generating axis-
parallel transport in segments along the basin margin.
The central part of the basin may be less complex and
axial-parallel sediment transport would prevail there
Fig. 12.7 Three major stages in the devlopment of extensional (Fig. 12.8b).
basins. After Gabrielsen (1986) In the thermal subsidence stage, thermal contraction
of the lithosphere dominates the basin subsidence pat-
homogeneous due to lack of pronounced gradients in tern. Because solids typically contract during cooling,
the basin. The marginal sediment transport system is the parts of the basin that have experienced the stron-
prone to act in concert with the axial transport system, gest extension (i.e. those that have been thinned the
feeding the latter with sediments. This may consist of most and hence heated the most) will contract and
braided or meandering river systems, depending on subside more than other parts. In a pure-shear configu-
factors like axial basin gradient and climate. Since ration this is most likely to be the central segment
most rifts are generated by break-up of continents, a running along the basin axis. This means that the rota-
terrestrial depositional environment would be most tion of strata upwards away from the basin axis
common for the pre-rift stage, so that source rocks becomes reversed so that strata begin to rotate down-
and cap rocks, which are mostly of marine depositional wards towards the basin axis (Fig. 12.9b). This rotation
origin, may be scarce (Fig. 12.8a). There are, however, is strengthened by sediment loading and compaction
numerous examples of both source rocks and cap rocks (thickest sequence in the central part of the basin).
of terrestrial origin. The transverse sediment transport will persist dur-
In the active stretching stage extension, and hence ing the thermal subsidence stage, while the basin floor
also subsidence, accelerate. Simultaneously, heat becomes gradually smoothed. An axial transport sys-
input increases due to upheaval of hot layers of the tem may also still be active, but is likely to become
mantle lithosphere. The steep fractures generated in less pronounced through this stage of development
the pre-rift stage will not be able to accommodate the (Fig. 12.8c). Depending on the balance between sub-
extension and a new set of low-angle planar or listric sidence and sediment input, the water depth will vary
faults will be activated, separating fault blocks that are from one basin to another, but the depositional envi-
detached from the lower crust by a subhorizontal zone ronment is likely to be marine and the central part of
of weakness. Gliding on the system of detachments, the basin may attain great water depth (thousands of
the fault blocks and their internal beds will rotate away metres). The fault systems that dominated the basin
from the basin axis (Figs. 12.7b and 12.9a). From the floor geometry during the active stretching stage are
view of the petroleum explorationist, the active now quiescent, and stratigraphic hydrocarbon traps
stretching stage deserves particular attention because rather than structural ones are likely to be the most
of the variety of structural and stratigraphic traps that common.
may develop. This stage is also characterised by a Syn-rift to post-rift transition. The pure-shear
complex sediment distribution system that is likely to model predicts that a simple geometrical change of
a b c
Fig. 12.8 Principal sedimentary transport systems associated with the three mains stages in the development of extensional basins.
After Gabrielsen et al. (1995)
the outline of extensional basins will accompany the during the syn-rift stage to tilting directed towards the
transition from the syn- to the post-rift stage. In this basin axis during the post-rift development (Fig. 12.9).
model the margins of the relatively narrow, steep- This change is due to a shift from bulk thermal expan-
walled rift, which traps the syn-rift sediments, become sion to bulk thermal contraction of the lithosphere and
overstepped at the syn- to post-rift transition. This is in most cases clearly distinguishable in reflection
implies that the basin becomes wider and the rate seismic data.
of subsidence decreases asymptotically during the fol- It needs to be emphasised that the syn- to post-rift
lowing post-rift stage. Thus, one defines the beginning transition is unlikely to occur simultaneously through-
of the post-rift development as the stage by which out the entire basin. This is due to differences in
the syn-rift faults become inactive and subsidence structural configurations, e.g. the existence of graben
becomes controlled dominantly by thermal contrac- units, and thermal inhomogeneities associated with
tion and sediment loading. variable stretching both along and transverse to the
In practical terms, the identification of this stage in basin axis. For reasons discussed below (Chap. 22),
the basin development is not trivial, because the tran- the entire Cretaceous sequence of the northern North
sition is frequently not synchronous all over the basin, Sea is included in the post-rift development sensu
and the criteria for identifying the transition in reflec- stricto. Furthermore, analysis of the basin topography
tion seismic data are not always well constrained. To permits three sub-stages to be identified within the
overcome this problem, the syn- to post-rift transition framework of the post-rift development: the incipient,
should be defined more precisely as the point in time the middle and the mature post-rift stages. The config-
when net heat out of the system is greater than net heat uration at the syn-rift/post-rift transition is treated
into the system. It is recognised that a lateral heat flow separately in the present analysis (Section 12.3.1).
gradient commonly exists perpendicular to the basin In the analysis of basin subsidence it is important to
axis. This implies that the area closest to the basin remember that in addition to the effects of fault-related
axis, which coincides with the area of greatest thin- subsidence and thermal expansion and contraction, the
ning, is also the part of the basin displaying the highest basin’s subsidence is affected by elastic deformation
heat flux at the end of the syn-rift stage. The litho- and isostasy, and in many cases also by extra-basinal
sphere beneath the central part of the basin will stress.
accordingly undergo the greatest vertical contraction The simple-shear model for extensional basins is in
during the post-rift stage. The enhanced subsidence at considerable geometrical and mechanical contrast to
the basin axis is further enhanced in cases where the the pure-shear model for extensional basins in that the
basin is filled by sediments, creating an extra load and simple-shear model assumes that extension is
also a greater total compaction. Hence, the syn- to concentrated along one or several inclined fault zone
post-rift transition coincides with a regional shift in (s) affecting the entire crust (Fig. 12.5b). Still, where
tilt from fault block rotation away from the graben axis thermo-tectonic and isostatic responses are concerned
330 R.H. Gabrielsen
Fig. 12.9 Pattern of rotation of sedimentary units in the (a) syn- and (b) post-rift stages. After Gabrielsen et al. (1995)
the principles are similar to those of the pure-shear development. However, the delamination model
model. The simple-shear model is based on makes it necessary to take into account an additional
observations in the Basin-and-Range of North Amer- variable parameter, namely that the two parts of the
ica and was formulated by Brian Wernicke in 1981. lithosphere situated above and beneath the delamina-
The Basin-and-Range basin system displays a particu- tion surface have undergone different amounts of
lar geometry in that the lithosphere is extended to the extension.
degree that the lower crust, described as a metamor-
phic core complex, has become uplifted and exposed
in the central part of the basin. The asymmetrical 12.3 The Structural Architecture
configuration of the basin particularly influences the of Extensional Basins
pattern of isostatic response to extension. An impor-
tant factor is the relative thickness of the upper mantle/ Comparison of many extensional basins reveals that
lithosphere. This is because the lower crust commonly they have many architectural elements in common.
is denser than the upper astenosphere, causing large- These include the position and geometry of the domi-
scale contrasts in differential subsidence and uplift nant fault systems, the position of the most prominent
across the basin. Superimposed on this are more terraces or platforms, and the position of structural
local isostatic effects, associated with contrasting highs. This does not imply that all basins are similar
thicknesses of layers with different densities and the or that all contain the same types of structures. Never-
topography of the basin. theless, a systematisation suggested in the following,
Since the same tectono-thermal principles that is useful in the analysis of extensional basins and their
apply for the pure-shear basin also are valid for exploration for hydrocarbons. Before looking in detail
simple-shear basins, the main basin stages and the at the exploration leads that are typical for extensional
conditions for hydrocarbon generation and entrapment basins, it is therefore instructive to examine the prin-
are also the same. Even though the simple-shear model cipal structural building blocks of the extensional
was inspired by analysis of the Basin-and-Range basin basin (Fig. 12.10a, b).
system it has proved relevant for many other basins The extra-marginal fault complex. The distal part
too, suggesting that simple shear is a common compo- of the basin is separated from the foreland by an
nent in the formation of basins. enhanced fault frequency (as compared to the foreland
The delamination model can be seen as a combina- of the basin) and a set of planar to listric faults com-
tion of the simple- and pure-shear models. In this case monly arranged in an en echelon geometry and some-
the upper and middle crust extends by simple shear. At times generating a horst-and-graben-system with a
depth, the master fault flattens and merges with the moderate relief. In some cases dike systems affiliated
lower crust, which becomes thinned by pure shear. with the initial stage of extension are filling in some of
The Viking Graben of the northern North Sea the faults of the extra-marginal fault complex. In many
seems to have a configuration that fits the delamination cases, a horst is seen to separate the extra-marginal
model (Fig. 12.6). Also in this case, the thermo- fault complex from the platform, defining a topo-
mechanical pure-shear model applies and, with some graphic threshold between the foreland and the basin
modifications, can be used to model the basin itself.
a Sub-platform
Platform
Extra-marginal
Internal graben Internal platform fault complex
Marginal
platform fault complex
high
Outer master
fault system
Inner master Axis of rotation

Interior fault system and area of faulting
graben Deep during cooling stage
Graben planar faults? Limit of intense
axis faulting during
Deep extension
planar faults?
Fig. 12.10 (a) Principal sketch of major structural elements in configuration as seen along the strike of the margin. Redrawn
graben systems. (b) Graben margin with its main structural from Gabrielsen (1986)
elements as seen in three dimensions. Note the change of
The platform is a relatively flat tectonic unit that is transition from the tectonically quiet platform to the
less intensely faulted. The extension may be much more heavily faulted sub-platform or the inner
concentrated on a limited number of faults, but even marginal fault complex. The platform marginal horst
those display only moderate throws as compared to the is asymmetrical in outline in that its platform-facing
major fault complexes that delineate the interior basin. margin is defined by a few faults with moderate
Hence, the platform is characterised by a few broad, throws, whereas its basinward border consists of a
and only slightly tilted, structural traps. Where age complex system of faults with great throws, some-
determinations are possible, the fault systems of the times in the order of several 100 m. The fact that the
platform area are seen to have been activated at an marginal platform horst is such a common feature in
early stage of the basin development, but the activity many mature extensional basins suggests that it has an
slowed down or became arrested when the subsidence important mechanical significance, the position of its
accelerated in the inner part of the basin system. The distal (antithetic) fault perhaps being determined by
platform may also be delineated on its basinward side the mechanical strength of the platform rocks.
by a horst. The sub-platform is delineated by the marginal
A platform marginal horst is sometimes developed platform high on its distal side and the interior graben
at the basinward side of the platform, defining the on the other. The inner marginal fault system that
332 R.H. Gabrielsen
separates the subplatform from the interior graben, 12.3.1.1 The Initial Stage
together with the extra-marginal fault complex, are According to the model presented above, sedimenta-
the most profound fault zones of the basin, and the tion during the initial stage will occur in a broad,
two are likely to be linked along the principal detach- shallow basin with moderate surface gradients. Due
ment found within the lower crust. The subplatform is to the moderate stretching at this stage, sedimentation
heavily faulted and encompasses a number of second- will generally keep pace with subsidence. Local
ary rotated fault blocks that again may be criss-crossed depocentres would be related to few, steep normal
by a third order of faults. There are examples that the faults with attached accommodation structures and
second order faults flatten along local detachments at shallow catchments. Because of the moderate surface
shallower levels than both the primary master faults of gradients and only few and minor fault escarpments, it
the marginal fault complex and the inner and outer is reasonable to assume that source areas for sand must
marginal fault systems. The inner marginal fault sys- be sought outside the marginal platform fault system.
tem also coincides with the axis of basinward rotation Traps generated at this stage can be buried to very
as activated during the post-rift stage. great depths (several kilometres) during the total
The interior basin is the unit of the basin which is graben subsidence. In addition to potential over-
underlain by the most extensively thinned crust and maturation due to deep burial, the scarcity of source
where the maximum post-rift subsidence occurs. In rocks may be a general problem for prospectivity
the case of a symmetrical (pure-shear) basin, it is of traps related to this stage, because terrestrial
delineated on both margins by inner margin fault conditions are likely to prevail. However, exceptions
systems, whereas in the case of an asymmetrical to this are for example the West African rift basin
basin (simple shear) only one of the margins has this lacustrine shales.
status. Even the deepest part of the basin is underlain In cases where later extension has caused signifi-
by rotated fault blocks, initiated during the active cant rotation of the pre-rift succession, the burial prob-
stretching stage of the basin formation. lem may be locally avoided, with pre-rift strata riding
It should be noted that the structural elements structurally high, for example along the marginal areas
described above are not likely to be present along the of the Viking Graben tilt blocks. As a consequence of
entire basin margin. Thus, in some segments, the plat- their early establishment in the subsidence history,
form may be in direct contact with the interior basin, traps of this type may become faulted and fragmented
whereas in other segments the platform or the platform by later movements. Even though the topography is
marginal high may be missing. This inconsistency influenced by active faults to a limited degree, traps
may reflect the influence of structural or lithological related to such structures still may occur.
inhomogeneities in the basement, varying strain rates The general three-stage model suggests that axial
or uneven bulk extension along the basin axis. Indeed, transport dominates in the initial graben stage and
it is common for large rifts that the basin is divided stratigraphic traps might be generated by the axial
into several sub-basins or basin units, each distin- fluvial system (Fig. 12.11, trap type 4). The modest
guished by its particular geometry and even polarity. tectonic subsidence characteristic of this stage would
favour large lateral extension and good continuity of
the sand sheet. The stratigraphic traps would be related
12.3.1 The Structural Influence on Reservoir to meandering or braided river systems and shallow
and Source Rock Distribution in lakes, and the geometry of the system would to a large
Extensional Basins extent be ruled by the subsidence rates in the incipient
central graben. Examples of axially transported
The types of traps related to the different stages in sandstones of this type in the primitive Viking Graben
graben formation are illustrated in Fig. 12.11. The include the Statfjord Formation and Lomvi Formation.
numbered trap types mentioned in the following These sheet sands contrast with the more lenticular
sections refer to this figure. sand bodies of the Lunde and Teist formations,
12
The Structure and Hydrocarbon Traps of Sedimentary Basins
Fig. 12.11 Variations in trap types associated with the pre-rift (orange), syn-rift (yellow) and post-rift (green) stages. The trap types associated with the syn-rift/post-rift transition
are marked in red. See text for detailed explanation. From Gabrielsen et al. (1995)
333
334 R.H. Gabrielsen
deposited under the control of greater subsidence enhance the internal graben relief by rotation and
rates. upheaval of fault-block crests and by isostatic and
In the evolving graben which should be flexural adjustments of the edges of the fault blocks.
characterised by increasing fault activity and eventual The fault pattern, which is initially relatively sim-
magmatism, the continuity of those sand bodies would ple and stable and consists of isolated fault strands,
be later broken. develops into linked structures, and the relief is
Units related to transverse sediment transport enhanced. This gives the possibility of accumulation
encompass both pure stratigraphic and mixed traps of considerable thicknesses of sediments, and with
(Fig. 12.11, trap type 5). Alluvial fan systems, which potentially good strike-continuity in the hangingwall
offer pure stratigraphic traps with potentially great fault blocks of the master faults. Such catchments may
lateral extension and considerable thickness, might be fed with sediments which are transported across the
represent the most important type. The general fault scarp from the foot-wall hinterland, and also
model for this stage in the graben development from the eroded crests of the neighbouring
suggests that axial transport systems dominate, but hangingwall dip slope. With continued subsidence,
that most of the axial transport comes ultimately the initiation of accommodation structures
from transverse systems upstream. (hangingwall anticlines, antithetic faults and forced
Clastic fans related to primary synthetic fault- folds) will restrict both the area of deposition and the
growth represent a mixed trap type, which depends continuity of sands. Simultaneously, fault complexity
on proximal sealing faults and a distal pinch-out will increase with the development of different types
(Fig. 12.11, trap types 1 and 2). Taking the likely of fault-related transfer zones and steps, which in turn
moderate relief, low surface gradient and simple may act as foci for drainage systems.
fault geometry into consideration, it is probable that Finer-grained pelagic and other mature sediments
the reservoirs in this type of trap are characterised by derived from the hinterlands would dominate over the
good lateral continuity in the dip direction. The thick- coarser sandy or conglomeratic deposits eroded from
ness of the source rock may be considerable in stable local structural highs, although the latter may domi-
basins. nate in isolated basins. Initially, the sediment transport
Clastic fans related to accommodation structures is still essentially axial, but due to enhanced relief in
along master faults (Fig. 12.11, trap type 1b) classify the central parts of the basin will be branched to some
as mixed palaeomorphic/structural traps. Because of extent. A growing local influence from relief caused
the steep geometries of the master faults in the initial by rotated fault blocks is expected. This may cause
rifting stage, the accommodation structures will be sediment transport parallel to the graben axis also in
laterally restricted transverse strike, and to constitute the more distal parts of the graben system. Locally this
a trap of some size accordingly demands good conti- pattern will be broken by transverse transport systems
nuity along strike. This is especially the case in late cutting across highs related to rotated fault blocks, or
stages of development where channel amalgamation in connection with relay ramps and bridges between
and progradation lead to a sheet-like geometry. fault blocks. In broader tilted fault-blocks or in
grabens immediately adjacent to the hinterlands, the
volume of transversely-transported coarse sediment
12.3.1.2 The Active Stretching Stage may be great.
During this stage the area of subsidence narrows, and Compared to the initial stage, the final burial depth
the structural elements typical for mature grabens start attained by traps generated in the active stretching stage
to appear. The subsidence, and hence the sedimenta- will of course be shallower. Due to the likely marine
tion pattern, which to a large extent will be influenced influx and sub-basinal restricted conditions, the chances
by the faulting within the graben area, is characterised for generation of source rocks will also be higher, and
by numerous fault-bounded depocentres that are only semi-regional or local areas of enhanced subsidence
partly interconnected. The geometry of faults will may promote the presence of local pockets of source
change from steep planar to low-angle, either by rocks in addition to those of regional significance.
fault-plane rotation or by development of listric faults. Fans related primarily to synthetic growth along
Both low-angle rotational and listric faulting will faults give potentially a mixed trap type, but may also
Drainage system
of footwall Marginal Hangingwall
uplands high Axial dip slope
Relay drainage
ramp system
Fig. 12.12 Drainage systems associated with rotated fault sand is also derived from the rotated hanging wall fault block.
block and associated fault scarps. Note that the fan systems Modified from Nøttvedt, Gabrielsen and Steel (1995)
associated with the fault scarps may be discontinuous and that
include stratigraphic trap elements (Fig. 12.11, trap scarps, the foot-wall fault block is likely to be the
type 5). The trap depends on the sealing of the fault major source for the hangingwall catchment, even
zone. This trap type is initiated before accommodation though the hangingwall dip slope may represent the
structures are activated, and may therefore be more extensive surface of erosion. Examples of strati-
characterised by wide extension towards the basin graphic traps in the hangingwall dip-slope position are
centre. The trap size and complexity depend upon documented several places in the Norwegian shelf,
sediment influx relative to basin floor subsidence rate and are well described from the western margin of
and the stability (potential for reactivation) of the the Viking Graben.
master fault. The first stage of hangingwall deformation in listric
In these syn-rift traps the width of the belt of reser- faulting may be normal drag, followed by development
voir sand/conglomerate out from the master fault zone of a hangingwall anticline (“roll-over”) which mirrors
is directly related to the subsidence rate and the size of the curvature of the master fault. A structural low is
the crestal area being drained. Abundant sediment defined on the proximal side of the hangingwall anti-
supply and moderate subsidence rates will cause larger cline, and opens for entrapment of sediments (Fig. 12.11,
radius fans to develop, whereas high subsidence rates trap type 6). Traps in this position are primarily mixed
cause all clastic sediment to be trapped in much stratigraphic/palaeotopographic and may be related to a
narrower belts in the immediate vicinity of the fault. local unconformity. Since displacement rates along
This trap type is well known from the late Jurassic faults are likely to increase with progressive displace-
interval in the North Sea, where the most prominent ment, it is possible also locally for subsidence to outpace
examples are the Brae Field and the Magnus Field. sedimentation with time, to create considerable space for
Because the coarse-sediment gravity-flow sands and deposition along active faults.
conglomerates are restricted to narrow fault-parallel A special type of palaeotopographic trap related to
belts, they commonly have problematic low continuity hangingwall anticlines may be expected where forced
in this same direction (Fig. 12.12). folds or fault-bend folds above flats in an irregular
Fans associated with accommodation structures fault plane have caused surface deformation. As in
may develop where extended fault activity takes regular hangingwall anticlines, there is room for sedi-
place. Such structures are likely to develop in the ment entrapment between the master fault and the
hangingwall close to the master fault. The type and forced-fold anticline (Fig. 12.11, trap type 9). This
geometry of the accommodation structure depend on trap type is in principle equivalent to that of the con-
the geometry of the master fault as well as on the ventional roll-over described above.
amount of subsidence. Because of development of a In addition, sediments may be trapped along the
fault scarp and the instability associated with such flanks of the forced fold (Fig. 12.11, trap types 8a and
336 R.H. Gabrielsen
8b). Where fault-bend folds are associated with a large transport mechanisms would be block sliding, rock
flat with a significant inclination to the regional dip of fall, or mass flow. In all these cases large volumes of
the master fault, and where sediment supply is good, reservoir rocks may become resident in the
the trapped volumes might be considerable. In fact, the hangingwall in proximal or distal positions relative
total volume may further be increased since the to the master fault.
depocentre will move as the hangingwall is Different types of gravity slides have been reported
transported across the flat, causing stacking of the in this position in front of rotated fault-blocks from the
units. The major disadvantage foreseen for this type North Sea, but have not so far been deliberately
of stratigraphic trap – especially in view of its distal drilled.
position – would be the difficulty for sediments to
bypass the hangingwall anticline. On the other hand, 12.3.1.3 Unconformity Traps Related to
the model allows the possibility of eroding the top of Transition from Active Stretching to
the anticline, thus utilising a local source area. Large- Thermal Cooling
scale fault-bend folds are not uncommon (e.g. Njord The transition from active stretching to thermal
Field), but we are not aware that stratigraphic traps cooling is manifested by a change in style of subsi-
related to such features have been reported in the dence pattern. In the active stretching stage, fault
literature from the North Sea. blocks rotate away from the graben axis, causing sim-
By continued subsidence along the master fault, ilar tilting in the sedimentary cover. The thermal
antithetic faults are likely to form on the distal flank cooling stage, however, causes tilting of strata towards
of the hangingwall fold, and a graben develops parallel the graben axis, mainly because the most rapid subsi-
to the master fault. If the sedimentation rate does not dence takes place along the graben centre (Fig. 12.9).
keep up with subsidence the graben may influence the In strongly asymmetric grabens, this picture will of
drainage pattern, and generate a sediment trap as well. course be modified accordingly.
This eventually heralds the shift from an open to a Overall transgression and onlap of the crestal areas
closed system. At starved basin margins the accom- may be expected towards the end of the active
modation graben may further restrict development of stretching stage because of the overall subsidence of
the primary fan, and in systems where the graben floor the rifted area, and because sediment starvation is not
is close to the marine level, the top of the accommo- uncommon when subsidence outpaces sediment yield.
dation structure may be eroded, and there may be local On a regional scale relief may be enhanced by
sediment transport towards the master fault upheaval of graben shoulders due to isostacy and
(Fig. 12.11, trap types 7a, 7b). elastic response to faulting. The spatial distribution
Examples of this type of trap are found in the “Ula and the magnitude of elevated (possibly eroded) and
trend”, which locally defines the margin of the Central subsiding areas is influenced by the geometry and
Graben. In block 2/2 a system of antithetic faults, depth of the detachments, and by whether the crustal
partly triggered by halokinesis, has developed a local thinning happens during simple or pure shear. Accord-
high, contributing to the initiation of a graben where ingly, these factors will be of importance to the devel-
the sands of the Ula Formation have been trapped. opment, distribution and preservation of stratigraphic
Gravity slides across fault escarpments may occur traps.
along the crestal areas of the rotated fault blocks The accelerated axial subsidence, which is likely to
(Fig. 12.11, trap type 7c). These are the most unstable occur during the early stages of cooling because of the
areas of the graben system and have their instability exponential nature of the thermal decay, will normally
enhanced by the flexural cantilever effect, which be associated with marine transgression, and at the
predicts accentuation of uplift in this area. Eventually, stage where earlier graben walls are onlapped and
slight inversion may further contribute to the uplift of drowned, a break-up unconformity will develop.
fault-block crests along basin margins. Depending on the nature of the subsidence, the uncon-
In reflection seismic data, erosion of uplifted and formity will be diachronously onlapped both along the
rotated fault block crests may be the most easily axis and transversely in the graben system. Depending
detectable effect, but deformation of the escarpment upon the relation between subsidence and sedimenta-
by gravity sliding should not be neglected. The basic tion rate, isostatic stability of the graben margins and
individual fault blocks, the unconformity may be very As the graben is expected to be filled by water,
complex. In this situation a number of unconformity transport agents may be gravity mass flows and turbid-
traps may develop. ity currents (Fig. 12.11, trap type 11b). Large transport
Truncational unconformity traps are formed by distances are therefore possible and the submarine
erosion of units deposited during the active stretching fans may be completely separated from the delta
stage. Fault block rotation and incipient compaction of systems along basin margins, particularly in periods
the sediment package are likely to cause slightly tilted with low-stand of sea level (Fig. 12.11, trap type 12).
units which may be eroded and sealed by the thermal Examples here include some of the main reservoirs in
subsidence shales (Fig. 12.11, trap type 10a). In prin- the North Sea, like the Frigg, Forties and Bruce fields.
ciple, this trap type may form in all parts of the graben The platform-vergent fans are closely related to the
system, but it is most likely to develop along the graben-vergent fans, but occur between crests of
graben shoulders. rotated fault-blocks (Fig. 12.11, trap type 13). During
On-lapping unconformity traps are found above the infilling of the graben, sedimentary packages in areas
unconformity, and are dependent upon whether ero- with the thicker sedimentary fill will tend to suffer a
sion on nearby highs has taken place or not stronger compaction than areas with thin packages.
(Fig. 12.11, trap type 10b). The relief across master This results in development of a hangingwall compac-
faults may be considerable at this stage, and it is likely tion syncline, which, if it has surface expression, may
that there is significant erosion and reworking of sands act as a local sediment trap.
from the marginal highs at this stage. Since the fault The platform-vergent fans will, however, be more
activity is retarded, the graben relief will diminish restricted than the basin-vergent ones, and are more
during this process, and the traps may be of consider- dependent on a local sediment source. These
able lateral extent and contain large volumes of sand. circumstances make this trap type less attractive due
to small potential sediment volumes.
12.3.1.4 Stratigraphic Traps Related to
Thermal Subsidence and Sediment
Loading 12.4 Strike-Slip Systems
At this stage sediment transport may be both axial and
transverse within the graben system, but it is likely that The strike-slip structural regime is characterised by
the transverse systems will dominate in the basin mar- horizontal orientations of σ 1 and σ 3, whereas σ 2 is
gin areas close to major hinterland relief. Minor vertical (Fig. 12.2). Hence, the orientation of the
continued fault activity should still be expected along plane of τmax is vertical, which is also the orientation
the master faults of the graben margins due to isostatic of the master faults. The displacement along the mas-
adjustments. These areas will also act as pivots during ter fault will be in the horizontal plane (parallel to
the shift in subsidence pattern. Altogether this implies strike) and the faulting includes initiation of a complex
relatively smooth graben slopes, with an increased system of secondary fractures.
possibilty to develop thick and extensive sandsheets The general development of shear systems can
with axes oriented transversely to the graben axis. conveniently be analysed by the use of analogue
Because of the high rate of subsidence of the graben mechanical experiments. Strike-slip systems are
floor at this stage, deposition is likely to take place in a highly dynamic, and the geometry of the initial stages
marine environment and the axial basin may be is very different from that of the mature stages of
starved of sediment. As the thermal gradients of the development. Figure 12.13a shows the relation
system approach equilibrium, the basin will fill in and between the structural elements at the initial stage of
finally level out the relief completely. strain for a right-lateral (dextral) shear system. To
Basin-margin fans (Fig. 12.11, trap type 11a) rep- analyse the shear system, one decomposes the shear-
resent a well-described trap type. These are true strati- forces into compressional and tensional vectors by
graphic or palaeotopographic in type, and will constructing a vector parallelogram. The dominant
normally have great lateral extent. Their thickness features define a system of conjugate fractures
will depend upon the degree to which the graben relief (Riedel- and Riedel’-shears) that are related to the
was levelled out when deposition took place. compressive component of the shear. Of these, the
338 R.H. Gabrielsen
Riedel-shears are synthetic to the major shear (mean- along a strike-slip fault. In the ideal case, where the
ing that they are sub-parallel and have similar shear- fault trace is planar and the movements of the oppos-
sense), whereas the Riedel’-shears are oriented at a ing fault blocks are absolutely parallel, the trace would
large angle to the regional orientation of the fault zone be one vertical plane. But since this is the case only for
and also display an opposite (antithetic) shear-sense. very restricted segments of strike-slip faults, there will
In addition to the Riedel- and Riedel’-shears, contrac- be segments where material is squeezed up and out of
tional structures (reverse fault, thrusts and folds) with the fault zone, and cases where slivers of the footwall
their axes oriented 90 to the compressional stress and hangingwall fall into the fault zone. In both cases
component may develop. Due to their favourable ori- the faults are likely to have a steeply dipping root,
entation and synthetic sense of shear, with continued creating diagnostic geometries for strike-slip faults
movement the Riedel-shears tend to become dominant called (positive and negative) flower-structures
at the expense of the antithetic Riedel’-shears. In (Fig. 12.13c). In cases where distinct fault segments
accordance with the orientation of the tensional overlap, but are not in direct contact, zones with pull-
vectors, a set of tensional fractures (T-fractures) may apart basins or turtle-back structures will occur,
be initiated with orientation parallel to the compres- whereas extensional and contractional duplexes will
sional component of the system. develop where the fault-segments are in contact in
By continued displacement there will be a tendency zones of releasing or restraining bends.
for both Riedel- and Riedel’-shears to rotate and for Movements in shear-zones are in many cases not
the tips of Riedel-shears to become joined to constitute entirely parallel, so that a contractional or extensional
a system of linked fractures striking parallel to the component adds to the shear. These situations are
main shear trend. These Y-shears require that some called transpressional and transtensional, respectively,
displacement has taken place and are accordingly not and contribute to exaggerating the morphological
present at the initial state of shear. Another set of shear expressions of the structures described above. In
fractures, P-shears, also occurs at a more advanced such cases there is a tendency for forces to decompose
stage of development. These are oriented at an angle along weak beds in the deforming units so that strain is
of 60 to the contractional component (Fig. 12.13a) taken up in separate systems. Thus a transpressional
and are accordingly symmetrical with the Riedel- stress can be decomposed into a pure contractional
shears, with the Y-shear direction as the plane of regime and a pure strike-slip regime. The process is
symmetry. Dynamically, the P-shears nucleate at the called strain partitioning and is well known for e.g.
profound, principal fault trace and develop up-section the transform delineating the western Barents shelf,
to create an array of fractures with an en echelon where a dextral transpression is decomposed into col-
geometry. lision in the West Spitsbergen Fold- and Thrust Belt
In total, the interaction between the sets of second- and shear along the Hornsund Fault Zone.
ary fractures (R, R’, T, P and Y) contributes to the
complex geometry of the strike-slip fault and
generates an uneven and step-like morphology. The 12.4.1 Hydrocarbon Prospectivity
arrangements of the steps in left-stepping and right- in Strike-Slip Regimes
stepping arrays generate contrasting stress-
configurations along the strike of the strike-slip fault, Large-scale strike-slip systems may be highly
depending on the relative shear-sense (Fig. 12.13b). dynamic depositional systems for sediments and also
Thus a right-lateral (dextral) shear that affects a right- offer a great variety of structural and sedimentary
stepping system of strike-slip fault branches causes traps. However, compared to extensional basins,
extension in the overlap-zones (ramps or bridges) there are two significant differences. Firstly, the ther-
between the individual fault branches, whereas a left- mal development is different in that the steep dips and
stepping system generates overlap zones of contrac- deep roots of the master faults are likely to cause very
tion for the dextral system. For a left-lateral (sinistral) significant and fast thinning. This involves the substra-
shear-sense, the relations are opposite. tum of the basin down to the level where the faults
This implies that a variety of structures, and hence a detach, which may be top of the lowermost crust or
variety of hydrocarbon trap types, are likely to develop even the Moho. This implies that the thermal gradient
a
Antithetic Extension
shear fracture
Master
fault
Secondary Synthetic
synthetic shear
shear
En echelon
folds
Riedel (R) Conjugate

shear Riedel (R’) shear
Y shear
Tension (T)
P shear
fracture
Fig. 12.13 (a–b) Fracture types typical for a dextral strike-slip common hydrocarbon traps in strike-slip systems. Types of
system. By advanced stages of strain, synthetic shears (Riedel structures are indicated in the figure. (a) and (b) are redrafted
shears) and fractures associated with the master fault (Y-shears) from Crowell (1974)
tend to dominate the geometry of the fault zone. (c) Most
340 R.H. Gabrielsen
may very rapidly steepen during early stages of the deformation takes place under high to extremely high
basin formation and that significant leakage of heat p,T-conditions. Such settings are not optimal for the
may start before the syn-rift stage is passed. This may generation and accumulation of petroleum resources
be accompanied by fast burial and maturation (and and will therefore not be considered further here.
perhaps over-maturation) of the source rock, and also However, in the upper (shallow) parts of an orogen
make basin modelling difficult. Secondly, the geome- as well as along its frontal parts, sedimentation and
try and tectonic position of strike-slip systems are such structuring take place during p,T-conditions that are
that the likelihood for accumulation of marine source compatible with the generation and accumulation of
rocks is less than for extensional basin systems now hydrocarbons. Here basins related to the interior
situated in passive margin settings. development and collapse of the central part of the
On the other hand, the structural complexity and orogen will occur, particularly in the later stages of the
variability may develop structural and stratigraphic mountain building, whereas foreland basins may be
trap types that are not found in extensional basin active throughout the entire life of the orogen.
systems. Examples are flower-structures and arrays
of anticlines that may be found along the strike-slip
fault at regular intervals. 12.5.1 The Architecture of Thrust Systems
The most important building blocks of orogens are

12.5 Contractional Regimes folds and contractional faults. Both these types of
structures affect rocks volumes, so that they may be
When reading the following, it is important to remem- constrained from their surroundings by a certain
ber that compression characterises the stress system, geometry and intrinsic style of deformation. Thrust
whereas contraction describes the physical process faults tend to climb up-section, because this is the
of shortening. Contractional regimes are associated direction of σ 3. Still, contractional faults are
with large-scale orogenic processes (formation of characterised by shallow dip and a tendency to flatten
mountain chains), but areas of shortening also may over greater distances, particularly where they follow
occur on local scales within regional extensional beds of low mechanical strength. They are also fre-
or strike-slip realms. Compression is characterised quently seen to merge with other faults along horizon-
by the principal stresses being oriented so that tal fault strands and surfaces between mechanically
σ 1 > σ 2 > σ 3 ¼ σ h max > σ h min > ρgz. In other weak beds, and they regularly intimately affiliate
words, the smallest stress acts in the vertical plane so with folds. This is so because the folds and faults
that the most energy-efficient way of shortening is by frequently are seen to develop in concert. One exam-
transferring excess mass towards the surface. Also, ple can be initial buckling of a bed followed by a fault
due to the orientation of the principal stresses, the breakthrough along the fold hinge. In other cases folds
dip of the plane of maximum shear is 30 , promoting can develop in front of an advancing fault (fault-
the development of thrust faults (Fig. 12.2). propagation folding). In both these examples, the
Orogens are extremely mobile tectonic zones and folds would be asymmetrical with the longer fold
such systems may accommodate displacements that limb dipping at a shallow angle away from the trans-
are several orders of magnitude greater than that port direction, and the shorter fold limb at a steeper
which is typical for extensional basins. This implies angle. The fold axes would be oriented transverse to
that the pattern of deformation may be very complex the direction of transport and may constitute structural
and involve a number of stages or “phases”, each traps of considerable magnitude. By continued
phase representing a unique set of stress conditions shortening the faults may link up, trapping isolated,
and p,T-relations. Still, even the most intricate pattern lens-shaped rock bodies, which commonly incorporate
can be analysed by the utilisation of relatively simple folded beds. The lenses are referred to as horses and
geometrical methods and modelling techniques. where they are stacked between a horizontal floor fault
The major strain in a large-scale contractional sys- and a roof fault, they constitute a duplex (Fig. 12.14a).
tem like an orogen is generally associated with plate In cases where the faults propagate systematically
margins. In the deep and central parts of the oregen, in the direction of the front of the contractional
Age (in Ma)

80 60 40 20 0
a I II III IV V VI VII
150
Temperature (in °C)

PAZ 100
1
50
5
3 4 6
∆T (in °C)
b 0
Transient heat effect –10
–20
c
(a) st st
hru thru
wsw Lewis thrust M nt. t gstone ENE
2 Femie basin 6-14-8-5W5 Turtle Livin
2
0 0 km
–2 –2
–4 –4
–6 –6
–8 –8
0 20 40 60 80 km
(b)
t
wsw
hrus
t hrus
am t ald t hrus
t Turtle Mountain
gsto
ne ENE
gw a c Don is t 7-20-6-3W5 07-27-6-3W5 Livin ust
Win M Femie basin Lew r
2 07-14-6-3W5 th
2
0
0 km
–2 –2
–4 –4
–6 –6
–8 –8
0 20 40 60 80 km
(c) Flathead Cate
wsw 8-20-4-1W5
Valley Creek Lewis thrust ENE
MacDonald Range Howell Creek 2-1-5-2W5
structure graben Window 07-34-3-1W5 2-16-5-2W5 6-22-5-1W5 3-27-6-27W5
2 6-15-5-2W5 2
0 0 km
–2 –2
–4 –4
–6 –6
–8 –8
–10 –10
0 20 40 60 80 km
Fig. 12.14 (a–b) Temperature-history curves for different (blue: low maturation, dark red: over-maturation). Light blue,
parts of the Canadian Cordillera at different positions relative green and orange colours indicate oil – gas maturation for the
to the apatite annealing zone (PAZ). (c) Structural cross- same profile. From Hardebol et al. (2009). Reproduced by
sections with corresponding organic maturity indications permission of the American Geophysical Union
system, detachments become linked by climbing fault on top of the older ones, and an antiformal stack is
branches. Together these faults may generate duplexes generated. Changes in overburden and friction may
developed by foreland-directed in-sequence thrusting also cause fault activity to switch from one place to
(Fig. 12.14b). In cases where the system halts e.g. due another in an unsystematic way. This is termed out-of-
to increasing friction, the younger horses may pile up sequence thrusting (Fig. 12.14b). Thus, it is not
342 R.H. Gabrielsen
uncommon that shortening is accommodated by This implies that reservoir rocks and hydrocarbon
hinterland-directed thrusting. In cases where the single traps of all kinds are abundant. Because the mountain
faults do not become joined along a roof fault, a chain necessarily is uplifted, however, organic-rich
system of “blind” faults may develop, terminating at marine deposits of the kind that would produce the
a tip-line. Alternatively, the faults may break the sur- source rock, would be rare. An exception to this would
face. In rarer cases the faults climb down-section in the be cases where a source rock deposited before the
direction of transport. This is an indication that the contraction started becomes involved in the orogen.
strain rate is greater along the roof-fault than it is along In such cases, the critical factors would be the depth of
the floor fault. tectonic burial of the source rock, the geothermal
The regions where contractional faults climb up- gradient of the greater orogeny and the positioning of
section are termed ramps and the total geometry of the the source rock relative to the reservoirs. In the
fault is that of a ramp-flat-ramp. In such cases, the pure dynamic environment of a nappe pile, it must be
geometry of the fault planes forces the strata inside the taken into consideration that units now separated by
horses to become folded. The folds reflect the geome- tens of kilometres may have been juxtaposed at the
try and steepness of the fault plane because the front of time of maturation and migration.
the horse depends on the cut-out angle of the original Three principally different basin settings can be
ramp. Ramps parallel to the transport direction may distinguished. Intramontane basins are collapse
also influence the development and the geometry of structures or structural lows generated by folding and
the thrust system. Such features may potentially sepa- thrusting inside the realm of the mountain chain.
rate subunits of contrasting deformational style. In Those active at the peak tectonic activity are likely
cases where strain rates are not similar across the to trap large amounts of coarse clastics over a short
ramp, shear and strong rotation occur, and when the period of time. They are in most cases of restricted size
structures propagate to affect the surface topography and source rocks are rarely associated with them.
they may strongly influence the depositional systems Foreland basins are far more interesting from a hydro-
associated with the mountain chain. carbon exploration point of view. These are stabilised
It is obvious that the tectonic processes described as accommodation areas for sediments due to the
above produce a variety of structural traps, among gravity load of the progressing orogen, and may trap
which anticlines with along-strike closure and horses the bulk of the sediments eroded from the rising
delineated by sealing faults may be the most obvious. mountains and transported towards the orogenic
Because the subsurface structuring also per definition front. The central parts of foreland basins may reach
affects the topography during mountain building, dif- thousands of metres in depth and constitute deep
ferent types of subtle and stratigraphic traps are also marine depositional systems. Thus, the foreland
likely to be generated (Fig. 12.14a). basin may offer a whole range of sedimentary
environments from fluvial, via shallow marine to
deep marine. The structuring in the foreland basin
12.5.2 Hydrocarbon Prospectivity of position is moderate, but increasing during the pro-
Contractional Regimes gressive development of the orogen. Thus, the basin
will eventually become overrun by the advancing
The very dynamic character of contractional systems deformation front and cannibalised. In this process, a
obviously produces a variety of sedimentary systems variety of structures are developed from gravitational
and structural and stratigraphic traps. The relief extensional to contractional complexes of folds, thrust
associated with orogens normally is measured in sheets and duplexes. In basins related to subduction
kilometres. Strong erosional forces, gravitational zones, different types of accretionary prisms may pro-
instability and climatic influences are important vide source rocks and reservoir rocks, as well as traps,
parameters in the development of mountain chains. both stratigraphic and structutral. However, these are
The system is flooded with a variety of clastic ero- very dynamic systems, sometimes too dynamic to
sional products, the mineralogical composition of provide low-risk exploration targets. Also, the
which reflects the types of rocks that are involved in sediments in such systems are likely to be too fine-
the orogen in the first place. grained to provide good reservoirs.
Although perhaps more rare than in extensional • Uplift of basin margins

regimes, source rocks may be involved in orogens, • Uplift of central parts of basins.
such as in the Canadian Cordillera. Detailed studies
here performed by Hardebol and his co-workers reveal The most common configuration at an early stage
a complex maturation history, where both tectonic and of inversion of a fault is shown in Fig. 12.15a. In this
sedimentary burial have to be taken into account. case, the accommodation space generated in the
Although some mountain chains may reveal surpris- hangingwall during extension is completely filled by
ingly homogeneous bulk geothermal gradient patterns, sediments. During inversion of the master fault, these
the individual tectonic units may have undergone sediments will be squeezed out of their position in the
contrasting histories of burial and uplift, making a hangingwall and onto the footwall, accompanied by
full tectonic restoration necessary in the evaluation uplift and folding. For inversion without any oblique
of the hydrocarbon maturation of the system component, the fold axes will be oriented parallel to
(Fig. 12.14). the strike of the extensional fault and, accordingly,
orthogonal to the new σ h max . By continued deforma-
tion, the pre-existing fault may be squeezed against the
12.6 Structural Inversion hangingwall and become steepened as a consequence,
whereas new, low-angle faults generated in the foot-
By the term “structural inversion”, or simply “inver- wall may create local thrusts (Fig. 12.15b).
sion”, we generally mean a system of extensional If one looks at the entire basin, the response on
structures that has subsequently undergone contrac- inversion will depend on the geometry of the basin
tion. This implies that the principal axes of stress and the mechanical properties of the crust and litho-
have been changed from being orientated such that sphere. In the case of a basin that has already been
affected by thinning and thermal weakening, the cen-
σ1 > σ2 > σ3 tral part may become overdeepened and the basin
shoulders uplifted. In contrast, in the case of a
¼
mechanically strong basin fill, the central basin may
σ v > σ h max > σ h min be uplifted, forming an inverted eye-shaped basin
¼ geometry. Alternatively, the basin fill may be folded
ρgz > σ h max > σ h min and squeezed out of the basin, as described above for
switches to faults.
σ h max > ρgz > σ h min
This implies that the dip of the plane of maximum 12.6.1 Hydrocarbon Prospectivity in Basins
shear will switch from 60 to 30 (Fig. 12.15). Thus, with Structural Inversion
although an established zone of weakness will repre-
sent a potential zone of reactivation when structural From a petroleum exploration point of view, structural
inversion occurs, it is unlikely that the already inversion is an effect that comes on top of and
established faults will be able to accommodate much subsequent to the development of a regular exten-
strain, meaning that new faults with lower angles of sional basin, and particularly affects the basin
inclination will be initiated. The most common margins. Inversion structures may provide additional
characteristics of an inverted system are: structural traps as very well exemplified in the mid-
• Reverse reactivation of (extensional) faults Norwegian margin by the Ormen Lange and Helland
• Generation of new, low-angle fault traces Hansen structures. On the other hand, inversion
• Development of secondary contractional structures invokes an additional risk for breaking of the seal
(folds, reverse faults, thrusts) and leakage though reactivated faults. Finally,
344 R.H. Gabrielsen
a Extensional half-graben
Post-rift
Syn-rift
Pre-rift
Basement
Incipient footwall
Normal fault shortcut fault
Inverted half-graben
Fault Footwall
reactivated as thrust shortcut fault
b
3
1
2
2
4
Old planar
fault zone
Fig. 12.15 (a) Typical geometry of an inverted extensional 1: Topographic low inside a pop-up. 2: Topographic low in the
fault. Note the low-angle footwall shortcut fault (modified hangingwall of an inverted fault. 3: Incipient syncline between a
from Cooper et al. 1989). (b) Surface expressions of deforma- train of emergent anticlines. 4: Syncline between two emergent
tion in an inverted terrane. Elevated areas, prone to yield ero- and eroded anticlines
sional products, are marked in red, sand accumulations yellow.
inversion is commonly associated with uplift and ero- properties of halite is well established (the deforma-
sion, that in most cases add to the complication of the tion mechanism changes from dislocation creep to
geological history and reservoir pressure. diffusion creep at a water content as low as 0.05%),
there is still not enough data available on the rheologi-
cal properties of evaporites to predict the detailed
12.7 Basins with Evaporites strain path and geometry that large evaporite bodies
develop when exposed to gravity and loading from a
Rock salt is strictly speaking a crystalline aggregate of clastic sedimentary overburden. In addition, many
the mineral halite (NaCl), which is one of more than evaporite sequences contain a high proportion of clas-
20 evaporite minerals formed by precipitaion from tic material that may be involved in the deformation,
saturated brines, most commonly from solar evapora- and the rheological effect of these clastic
tion. Although salt deposits may contain large portions “contaminations” is not easily predictable.
of other evaporite minerals like anhydrite and gypsum, Although one tends to associate the problem of
most studies of the mechanical properties, and hence seismic imaging in areas with extensive salt deposits
the dynamics, of evaporites, consider rock salt as the to be affiliated with complex salt structures, one
dominant mineral. Although the very strong influence should not overlook that many basins contain
of water on the mechanical strength and flow evaporites that are tabular, flat-lying and stable, and
where the problem of seismic imaging is restricted to migration of contraction can then reach the exten-
seeing through the salt as such. It is still fair to say that sional domain and squeeze the diapirs.
the structural geology of evaporites has attracted much Analogue experiments show that the overall struc-
attention, partly because the high mobility of such tural zoning is mainly controlled by the initial condi-
deposits poses intriguing structural geological tion (salt basin) and the basal slope angle, whereas the
problems, and even more so because salt structures type of structures in the structural domains strongly
are associated with a variety of structural and strati- depends on sedimentation rate (Fig. 12.16b).
graphic traps of significance for the petroleum indus- An early attempt to classify salt structures system-
try. The structural geology of salt very much reflects atically and to set this into a dynamic context was
the local tectonic environment, be it extensional, con- made by Trusheim in 1960. He suggested that salt
tractional or strike-slip. This implies that salt bodies impiercements grow from elongated low-profile ridges
come in an almost infinite variety of shapes, some of (anticlines and rollers) triggered by gravitational
which even challenge the limits of the imagination. contrasts, developing into rows of pillows, diapirs
This variety of shapes, combined with the acoustic and eventually into walls and sheets of salt. The
properties of salt, poses great challenges for reflection diapirs come in a variety of shapes from regular mas-
seismic imaging of salt bodies and the strata beneath sive stocks, via irregular masses to elegant
and close to them. mushrooms. This geometric classification is undoubt-
Gravity-driven deformation at continental margins edly valid for a tectonically stable, evenly subsiding
is generated by the regional gradient of the margin basin. But even this relatively predictable kinematic
itself and is characterised by upslope extension and growth pattern of salt structures causes great problems
downslope contraction. Salt and its overlying sedi- in seismic imaging due to the complex pattern of
mentary pile spread in a seaward direction due to internal flow in the salt structure itself, including hori-
regional tilting in response to lithosphere cooling, zontal displacement, affiliated with the development
whereas synkinematic sedimentation induces loading of overhanging or even horizontal walls.
instabilities. At basin scale, thin-skinned deformation
may induce extreme upslope salt thinning, leading to
the formation of salt welds, as well as massive down- 12.7.1 Hydrocarbon Prospectivity in
slope salt thickening. Most often, the extensional Salt-filled Basins
domain can be divided into three sub-domains, which
are, in a seaward direction, the sealed tilted block, Deep parts of basins where salt structures tend to be
growth fault/rollover, and diapir domains situated may be excellent sediment traps. The growth
(Fig. 12.16a). The upper domain is characterised by of diapirs contributes to the development of local
tilted blocks that are sealed early by synkinematic depocentres and the areas around salt diapirs may
sedimentation. The rollover domain displays a large accumulate large volumes of reservoir rocks of good
amount of extension, whereas the domain of diapirs is quality and be associated with excellent structural and
generally considered as gently translating, accommo- stratigraphic traps. However, due to the capacity of
dating small amounts of extension. Diapirs correspond salt to flow horizontally at shallow levels and develop
to weak zones and are easily and often squeezed. overhanging bulges and sheets, and even to become
Downslope of the margin, contractional structures bal- detached from its deeper sources, the detailed geomet-
ance the amount of upslope stretching. ric configuration around the stem of the salt structure,
The domain of shortening is also divided into three including its diameter, is commonly disguised and the
sub-domains. In a seaward direction they are com- diameter of the stem itself may be impossible to deter-
posed of diapirs squeezed at late stage, polyharmonic mine from reflection seismic data.
folds and thrust faults developed at early stage, and In addition to the bulge above the salt
folds and thrusts developed at late stage. Contractional empiercement itself, which will reflect the geometry
structures are initiated in a domain located at a dis- of the upper layers of the empiercement, the main
tance from the initial salt edge. Compression remains types of features that may constitute structural hydro-
localised in this domain during the initial stages of carbon traps adjacent to salt diapirs are the rim
evolution by continued deformation. The upslope syncline system (sometimes several generations),
346 R.H. Gabrielsen
a Seaward Landward
- Migration of contraction
Extensional domain
1.0
2.0
TWT (s)
3.0
4.0
5.0
6.0
25 km
Sealed tilted
Growth fault/Rollover domain block domain
Diapir domain
25 cm
Wall 5
6
5
3 1
2 7
1 4
1
Pillow Stock
Fig. 12.16 (a) Analogue experiment showing salt structures different structural positions. 1: Rim syncline. 2: Rim syncline
and related structuring of sediments between salt ridges in an between a stock and a pillow. 3: Rim syncline along a salt wall.
extensional, inclined slope (from Brun and Fort 2008). (b) 4: Salt-induced graben. 5: Graben on top of collapsed salt
Surface expressions of deformation in a terrane affected by pillow. 6: Salt-induced rotated fault block. 7: Stratigraphic trap
halokinesis. Elevated areas, prone to yield erosional products, covered by overhang in salt pillow
are marked in red, sand accumulations given in yellow for
anticlines associated with the rim syncline system, relatively straightforward. For mushroom-shaped
faults generated due to volume reduction during verti- diapirs, however, this is much more challenging and
cal transport of salt, and drag structures close to the several parameters have to be taken into consideration
stem of the diapir (Fig. 12.16b). Due to the circular in the structural analysis: the distance of the rim syn-
nature of the diapir, all these structures are likely to be cline system from the centre of the diapir and its
closed when seen in three dimensions. In addition, amplitude and wavelength depend on the thickness
numerous types of stratigraphic trap may be related of the original salt sequence, on the diameter of the
to any of these structural features. For salt anticlines diapir, and the salt flow rate relative to sedimentation
and simple walls, which have not developed rate. In many cases however, this structure is situated
overhangs, seismic imaging of the structures is usually sufficiently far away from the diapir for seismic
imaging to be straightforward. When it comes to the 12.8 Faults and Fault Architecture
fault systems and the drag structures, these occur close
to the stem of the diapir and are likely to be covered Fractures (faults and joints) are found in practically all
and completely obscured by the overhanging diapir hydrocarbon reservoirs and are crucial elements
bulb. because they both influence the migration of
One particular difficulty in seismic interpretation hydrocarbons within the reservoir and contribute to
may occur in cases where dynamic salt interacts with the entrapment of fluids. Due to great variation in
faulting. At the crests of salt diapirs and stocks a fault rocks and fracture types and their distribution,
combination of ring-shaped and radial fault systems the influence of fractures on reservoir communication
is commonly found, but these are unproblematic in is not easily predictable. It is therefore natural that the
seismic imaging and interpretation. In addition, analysis of single faults and fracture systems is receiv-
numerous examples exist of fault activity promoted ing increasing attention.
by the underlying active salt, providing a substratum For the assessment of the architecture of faults and
for detachment. Again, such structural relations are fracture systems one can, in principle, chose between
also clearly displayed in reflection seismic data. stochastic and deterministic or a combination of the
Finally, though, due to transfer of larger volumes of two. Given the complexity and generalised architec-
salt towards the basin axis, smaller pillow-shaped ture of larger faults, however, stochastic methods are
volumes of salt are frequently left and trapped along less favourable in analysis of such features. This also
basin margins, where they interact with the basin mar- seems to be the case for fracture systems generated in
gin fault system. In such cases different configurations stress situations where σ 1 is distinctly different from
are sometimes developed in the hangingwall and the σ 3. To provide input to a deterministic reservoir model
footwall of the salt-involved fault, causing complex that aims at taking the complexity of larger faults into
and contrasting sedimentary conditions across the consideration, we have performed field studies in
fault so that significant problems arise in sequence order to constrain realistic characteristics for the
correlation. Where salt has intruded along the fault- units that commonly can be defined within the realm
plane, interpretation of seismic data may be hampered of a fault zone.
by reduction of the general data quality. Since mesoscopic and macroscopic faults affect
In more complex tectonic environments (strike-slip volumes of rock, and accordingly should be described
and contraction), the complexity and variety of salt as composite rock bodies that include a complex sys-
body configuration is commonly much greater, tem of structures (fault rock, folds and fractures), the
because the final geometry of the salt body will be term “fault zone” is applied here. In the following we
determined by directed flow reflecting varying differ- use “fracture” as a general term that includes faults,
ential stress and strain. For example, the importance of deformation bands and joints, and we distinguish
evaporite sequences in the development of many between shear fractures (microscale) and faults
thrust belts like the Pyrenees and the West Spitsbergen (meso- and megascale). Also, we use the term “high-
thrust-and-fold belt is well documented. In such strain zone” for the parts of the fault core where shear
settings seismic imaging may be complicated by salt is concentrated.
being involved as an extensive, continuous or
disrupted unit during the thrusting, and also because
it may have accumulated unevenly and become
integrated in contraction structures like fold cores 12.8.1 The Structural Elements of the Fault
and duplexes, and as intrusions along fault planes. Zone
The quality of seismic imaging performance in
areas of salt has been greatly improved in recent It is well established that faults are commonly zoned
years. Still, the days of surprises are not yet over and composed of several units with distinct deforma-
when results from drilling become available. The tion styles (Fig. 12.17). These include the fault core,
effort in improving seismic imaging techniques must where most of the displacement is accommodated, and
therefore continue. And it should go hand in hand with its associated damage zones that are geometrically and
field study and analogue modelling. mechanically related to the development of the fault.
348 R.H. Gabrielsen
Footwall
master Hangingwall
Fault core master
branchline
Distal Central Distal branchline
A: Host rock horse

F B: Fault rock horse
B C: High strain zone
D: Outer fault core
E: Fault core collapsed to slip surface
D A
F: Clay lens
H: Slip surface
III I - III Fracture sets in hangingwall
C
IV - VI Fracture sets in footwall
VI I
IV V E II
Fig. 12.17 The structural elements of an extensional fault
The fault core is in general separated from the suggested that the lenses are close to symmetrical
footwall and hangingwall damage zones by distinct with reference to both the central a- and b-axes. The
fault-branches. Lozenge-shaped rock bodies fre- available data also suggest that these relations are
quently dominate the cores of extensional faults. roughly valid also for the higher-order lenses (2nd,
These are commonly referred to as fault lenses or 3rd, 4th order etc.), although there is a tendency for
horses, which may occur in isolation, as en echelon the a:c- and b:c-values to become slightly reduced for
trains, or be stacked to constitute duplexes. The fault- the higher-order lenses.
rock lenses may consist of relatively undeformed The high-strain zones separating individual or
country rock derived from the footwall or the groups of fault lenses may include deformed units
hangingwall of the fault core. In faults with greater that can be recognised as country rock in the footwall
displacement, the fault lenses may represent and hangingwall, as well as several types of fault
lithologies exotic to that of the observable footwall rocks, the host rock of which cannot be determined.
and hangingwall or be completely reworked fault For example the most intensely deformed zone of the
rocks like cataclasites and breccias. The geometry of fault core is easily distinguishable and this zone may
the lenses, their relative arrangement and their relation represent the latest area of deformation, indicating that
to intervening high-strain zones are important for the strain softening has occurred.
fluid communication along and across faults in cases The damage zones define a halo of fractures on both
where contacts between units of high or low perme- sides of the fault core. The fractures of the damage
ability control the fluid flow. zones are associated with the dynamic development of
Field study of the shape of fault core lenses the fault and may encompass remnants of the propa-
suggests that such features have relatively regular gation of the incipient fracture, commonly termed the
shapes and that the a:c-ratio (relation between length process zone. The strain intensity in the damage zones
measured in the dip-direction and maximum thick- is generally modest compared to that of the fault core,
ness) in extensional faults is in the order of 10:1 and and in sedimentary rocks bedding and other primary
that the b:c-ratio (relation between length measured in features can commonly be recognised. The fracture
the strike-direction and maximum thickness) is some- distribution, frequency and orientation in the
what less than this, perhaps 9:1 or 8:1. It is also hangingwall and the footwall are generally different,
and it is therefore convenient to distinguish the foot- Cunningham, W.D. and Mann, P. 2007. Tectonics of Strike-Slip
wall damage zone from that of the hangingwall. Restraining and Releasing Bends. Geological Society Spe-
cial Publication 290, 482 pp.
Field investigations show that damage zones are Dahlstrom, C.D.A. 1970. Structural geology in the eastern mar-
not symmetrically distributed around the fault core. gin of the Canadian Rocky Mountains. Bulletin Canadian
Accordingly, it is commonly observed that the fracture Petroleum Geology 18, 332–406.
frequency curves build up towards their maximum Fossen, H. and Gabrielsen, R.H. 2005. Strukturgeolgi.
Fagbokforlaget Vigmostad & Bjørke AS, 375 pp (in
values more steeply in the footwall than in the Norwegian).
hangingwall. Also, there are differences where orien- Gabrielsen, R.H. 1986. Structural elements in graben systems
tation of the major fracture populations is concerned. and their influence on hydrocarbon trap types. In: Spencer,
The footwall damage zone is characterised by a set of A.M. et al. (eds.), Habitat of Hydrocarbons on the Norwe-
gian Continental Shelf. Norwegian Petroleum Society.
fractures oriented subparallel to the footwall master Graham & Trotman, London, pp. 55–60.
fault-branch. This fracture population may interfere Gabrielsen, R.H. and Møller-Holst, J. 1997. The influence of
with fracture sets that dip both more and less steeply present and future development in information technology
than the master fault-branch. The hangingwall damage on the Exploration and Production phase model. Petroleum
Geoscience 3(3), 193–202.
zone is influenced to a greater extent by antithetic Gabrielsen, R.H., Steel, R.J. and Nøttvedt, A. 1995. Subtle traps
fractures to the master fault and a wider area is in extensional terranes: A model with reference to the North
affected by these fractures. The fractures of the dam- Sea. Petroleum Geoscience 1, 223–235.
age zones may be of different kinds, depending on Halbouty, M.T. 1979. Salt Domes: Gulf Region, United States
and Mexico, 2nd edition. Gulf Publishing, Houston, Texas.
lithology, depth of burial at the time of deformation, 561 pp.
and strain intensity. Hancock, P.L. (ed.). Continental Deformation. Pergamon Press,
Faults that are in their early stage of development Oxford, 421 pp.
do not possess the complexity described above. In Hardebol, N.J., Callot, J.P., Bertotti, G. and Faure, J.L. 2009.
Burial and temperature evolution in thrust belt systems:
sandstones they develop from single or arrays of Sedimentary and thrust sheet loading in the SE Canadian
deformation bands that with increasing strain coalesce Cordillera. Tectonics 28, TC3003, 1–28, doi: 10.1029/
into one zone of focused shear. This is generally also 2008TC002335.
the case for incipient faults in carbonates. Jackson, M.P.A. and Talbot, C.J. 1991. A glossary of salt tec-
tonics. Geological Circular 91-4, University of Austin,
Texas, Bureau of Economic Geology, 44 pp.
Kearey, P., Klepeis, K.A. and Vine, F.J. 2009. Global Tectonics,
3rd edition, Wiley-Blackwell, New York, NY, 482 pp.
Further Reading Lister, G.S., Etheridge, M.A. and Symonds, P.A. 1986. Detach-
ment faulting and the evolution of passive continental
margins. Geology 14, 246–250.
Allen, M.R., Goffey, G.P., Morgan, R.K. and Walker, I.M. McKenzie, D.P. 1978. Some remarks on the development of
(eds.). 2006. The Deliberate Search for the Subtle Trap. sedimentary basins. Earth and Planetary Science Letters 40,
Geological Society Special Publication 254, 304 pp. 25–32.
Biddle, K.T. and Christie-Blick, N. (eds.). 1985. Strike-slip Nøttvedt, A., Larsen, B.T., Gabrielsen, R.H., Olaussen, S.,
Deformation, Basin Formation and Sedimentation. Society Brekke, H., Tørudbakken, B., Birkeland, Ø. and Skogseid,
of Economic Palaeontologists and Mineralogists (Tulsa J. (eds.), 2000. Dynamics of the Norwegian margin. Geolog-
OK), Special Publication 37, 386 pp. ical Society Special Publication 167, 472 pp.
Boyer, S.E. and Elliot, D. 1982. Thrust systems. American Nøttvedt, A., Gabrielsen, R.H. and Steel, R.J. 1995. Tectono-
Association of Petroleum Geologists Bulletin 66, stratigraphy and sedimentary architecture of rift basins, with
1196–1230. reference to the northern North Sea. Marine and Petroleum
Brun, J.-P. and Fort, X. 2009. Entre Sel et Terre. Structures et Geology 12, 881–901.
Méchanismes de la Tectonique Salifière. Société Géologique Nystuen, J.P. 2008. The changing face of the Earth, Chapter 2 in
de France, Paris, 154 pp. Ramberg, I.B., Bryhni, I., Nøttvedt, A. and Rangnes, K.
Butler, B.W.H. 1982. Terminology of structures in thrust belts. (eds.) The Making of a Land – Geology of Norway, Norsk
Journal of Structural Geology 4, 239–245. geologisk forening. The Norwegian Geological Association,
Cooper, M.A., Williams, G.D., de Graciansky, P.C., Murphy, R. Oslo, pp. 20–61.
W., Needham, T., de Paor, D., Stoneley, R., Todd, S.P., Turner, Odinsen, T., Christiansson, P., Gabrielsen, R.H., Faleide, J.I.
J.P. and Ziegler, P.A. 1989. Inversion tectonics – A discussion. and Berge, A.M. 2000. The geometries and deep structure of
Geological Society Special Publication 44, 335–347. the northern North Sea rift system. In: Nøttvedt, A. et al.
Coward, M. 1994. Continental collision. In: Hancock, P.L. (ed.), (eds.), Dynamics of the Norwegian Margin. Geological
Continental Deformation. Pergamon Press, New York, Society Special Publication 167, pp. 41–57.
55–100 pp.
350 R.H. Gabrielsen
Ramberg, I., Bryhni, I., Nøttvedt, A. and Ragnes, K. 2008. The Twiss, R.J. and Moores, E.M. 1992. Structural Geology. W.H.
making of a land. Geology of Norway. Published by the Freeman & Co., New York, NY, 532 pp.
Geological Society of Norway, 624 pp. Wernicke, B. 1981. Low-angle normal faults in the Basin and
Ranalli, G. 1995. Rheology of the Earth, 2nd ed. Chapman & Range province: Nappe tectonics in an extending orogen.
Hall, London, 413 pp. Nature 291, 645–647.
Tchalenko, J.S. 1970. Similarities between shear zones of dif- Wernicke, B. 1985. Uniform-sense normal simple-shear of the
ferent magnitudes. Geological Society of America Bulletin continental lithosphere. Canadian Journal of Earth Science
81, 1625–1640. 22, 108–125.
Chapter 13
Compaction of Sedimentary Rocks: Shales, Sandstones

and Carbonates
Knut Bjørlykke
The physical properties of sedimentary rocks change rates of these reactions are controlled by the kinetic
continuously during burial as a response to increasing parameters such as the activation energy and thereby
stress and temperature; they also change to a certain the temperature.
extent during uplift and cooling. There is an overall The main lithologies in sedimentary basins are
drive towards lower porosity with depth, which shales, sandstones and carbonates, and they respond
increases the sediment density and sonic/seismic very differently to increased stress and temperature
velocity. during burial.
Increased effective stress from the overburden or This is important both for basin modelling and in
from tectonic stress will always cause some mechani- seismic data interpretation.
cal sediment compaction (strain), expressed by the There are no precise definitions for mud, mudrock
compressibility and the bulk modulus (see Chap. 11) and shale. The term mud is used to describe fine-grained
which can be measured in the laboratory. During sediment with a relatively high content of clay-sized
mechanical compaction the solids, mainly minerals, particles, chiefly clay minerals. Carbonate mud will be
remain nearly constant so that the reduction in bulk discussed separately, under carbonate compaction.
volume is equal to the porosity loss. Sediments may The compaction (porosity loss) as a function of
also contain solid amorphous phases which are chem- burial depth varies greatly because each primary
ically unstable (e.g. opal A) The transformation of lithology has a different compaction curve. While
solid kerogen to fluids (petroleum) and dehydration porosity may increase with depth through an interval
of minerals involves however some changes in the due to changes in lithology, for each individual lithol-
volume of solids. ogy the porosity will nearly always be reduced with
Chemically, the mineral assemblage will be driven depth (Fig. 13.1a).
towards higher thermodynamic stability (Lower Gibbs
Free Energy) (Fig. 13.1b). These reactions involve the
dissolution of minerals or mineral assemblages that 13.1 Compaction of Mudrocks
are unstable, and precipitation of minerals that are and Shales
thermodynamically more stable with respect to the
composition of the porewater and the temperature. Mudrocks and shales are often treated as one lithology
Higher temperatures will favour minerals with lower in connection with basin analyses, seismic interpreta-
water content, for example by dissolving smectite and tion and well log analyses, but in reality they span a
kaolinite and precipitating illite (see Chap. 4). The wide range of properties determined by the diversity
of mineral composition and grain-size distribution.
Furthermore, the composition of mudstones and shales
changes during progressive burial due to diagenesis,
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway which includes both mechanical and chemical
e-mail: [email protected] compaction.
352 K. Bjørlykke
a Physical properties
Generalised porosity-
depth trend φ = φ0 e–cz
Real porosity-
depth trend
Depth
Permeability
Density/velocity
b Burial diagenesis – Compaction of siliceous sediments

Initial sediment
composition?
Density – velocity (ampl.)
Surface Strain
Porosity
Mechanical Prediction of rock properties
Compaction can be based on observations,
Effective stress experiments and modelling
70−100 °C
Measurements from logs
Chemical or cores provide a good
compaction basis for prediction of
Thermo- D rock properties at deeper
dynamics e and shallower depths
and kinetics p
t
h
Stress/Temperature
Fig. 13.1 (a) The porosity/depth trends will be different for sediment composition the porosity is reduced and the density
different lithologies (primary mineralogical and textural com- and velocity are increased. Mechanically the compaction is a
position). A simple exponential function may be rather far off strain due to effective stress. Chemical compaction resulting
from the real porosity/depth function. (b) Principal aspects of from dissolution and precipitation of minerals is controlled in
sediment compaction (burial diagenesis). During burial, siliceous rocks by thermodynamics and kinetics and is therefore
sediments are subjected to changes in physical properties as a a function of temperature and time. The strain (compaction) is
function of increasing stress and temperature. From an initial here independent of stress
Just after deposition the porosity of the mud near Clay minerals, which usually account for the bulk
the sea or lake bottom may be extremely high, up to of the finest fractions, have an impressive size range.
70–80%. After about 1,000–2,000 m burial depth most Kaolinite particles are sheets where the longest dimen-
of the mechanical compaction has taken place even if sion is typically 1–20 μm, while smectite particles may
the porosity may still be relatively high (20–40%). be smaller by a factor of 1,000 (only a few nm). Illite,
Muddy sediments can become very compact because chlorite and most other clay minerals have grain sizes
silt and clay can occupy much of the pore space that are intermediate between these end members.
between the larger grains, resulting in a densely Smectite has a very high specific surface area (several
packed mass. hundred m2/g) because of the small grain size and is
13 Compaction of Sedimentary Rocks: Shales, Sandstones and Carbonates 353
very sensitive to the chemical composition of the

porewater. Additions of salt (NaCl or KCl) are used
to stabilise soft clays for engineering purposes (con-
struction), increasing their compressive and shear
strengths. Addition of KCl in the drilling mud is also
used to stabilise clays when drilling.
Sediments are often highly anisotropic and
parameters like seismic velocity and resistivity can
vary greatly with the orientation of the measurement
relative to the bedding. Mudstones may also become
increasingly anisotropic with burial depth, giving
higher velocity parallel to the bedding than in the
vertical direction. Experimental compaction shows,
however, that the degree of grain reorientation varies
markedly with the clay mineralogy and the content of
sand and silt (Voltolini et al. 2009).
The composition of mudstones and shales with
respect to their clay mineralogy and their content of
silt and sand can provide not only important informa-
tion about the environment both in and around the
basin, but also about the rock properties controlling
bulk compressibility, density, seismic velocity and
Fig. 13.2 Jurassic mudstone from the North Sea basin buried to
resistivity. These parameters are important in the 2.5 km depth. Note that many of the grains are of silt-sized
interpretation of seismic data and also electromagnetic quartz and that mica grains have a parallel orientation (scale ¼
surveys. This is clearly seen in some of the silty 0.06 mm). The white spherical structures are framboidal
Jurassic shales from the North Sea basin (Fig. 13.2). pyrite. The velocity in this shale is about 3 km/s (Vp 3,019–Vs
1,665 m/s)
In the North Sea basin and Haltenbanken, poorly
sorted, clayey, partly glacial sediments of Pleistocene
and Pliocene age fall on a nearly linear compaction In the laboratory the velocites (Vp and Vs) can be
trend reaching velocities up to 2.8 km/s near the base measured as a function of stress for mixtures of differ-
of this sequence (Fig. 13.3). Glaciomarine clays over- ent clay minerals (see Chap. 11). Smectitic clays have
run by glaciers can become very hard and compact; in very much lower velocities than kaolinitic clays but
the Peon gas field in the North Sea they have devel- additions of silt increase the velocity. The Vs =Vp ratio
oped sufficiently low permeability to trap gas at just also varies as a function of clay mineralogy. This is
160 m below the seafloor. very important since this ratio is used to determine the
Eocene and Oligocene smectitic clays of volcanic fluid content in sand and siltstones. Mudstones and
origin have much lower velocities (<2 km/s) and shales also may have a significant content of gas
densities (Fig. 13.3). Kaolinitic clay is far more com- which changes the Vs =Vp ratio.
pressible because it is very much coarser-grained so The primary composition of the mud deposited on
that the stress per grain contact is higher. Experimental the seafloor depends on the clay mineralogical com-
compaction also shows that fine-grained kaolinite is position and the amount of silt- and sand-sized grains.
less compressible than coarse-grained kaolinite (Fig. Carbonate and silica from biogenic debris are critical
11.7). Illite and chlorite are much more difficult to components with respect to burial diagenesis. Rela-
characterise. The clay mineral illite as determined by tively moderate amounts of carbonate cement in
XRD includes both relatively coarse-grained detrital mudstones result in high velocity at shallow depth.
mica and very much finer-grained diagenetic ilitte, i.e. The source of the carbonate cement is in most cases
formed from smectite. Chlorite also varies consider- biogenic carbonate. Fossils composed of aragonite are
ably, from detrital chlorite from metamorphic rocks to particularly important because they dissolve and
authigenic, usually Fe-rich, diagenetic chlorites. become a source of calcite cement.
354 K. Bjørlykke
0 0
Haltenbanken,
western region
17 wells
1000 1W 1000
2W
3W Haltenbanken,
2000 2000 western region
Depth (mTVD)
Depth (mTVD)
4W 17 wells
3000 3000
5W
6W
4000 4000
Black = Shale
5000 Red = Sandstones 5000 Black = Shales
Green = Spekk Formation Red = Sandstones
= Trend line, published data Green = Spekk Formation
1000 2000 3000 4000 5000 1.6 1.8 2 2.2 2.4 2.6 2.8
Velocity (m/s) RHOB (g/cm3)
Storvoll et al., AAPG Bull. (2005)
Fig. 13.3 Compaction trends are a function of burial depth and the sandstones may preserve high porosity (low density) due to
primary (initial) composition (from Storvoll et al. 2005). The grain coatings and reduced quartz cementation. The Upper
poorly sorted glacially influenced Pliocene and Pleistocene Jurassic Spekk Formation which is the main source (in green)
sediments (1 W) compact readily while the Eocene and Oligo- is characterised by low densities and velocities (low acoustic
cene smectite-rich sediments of volcanic origin (4–3 W) have impedance, AI). When source rocks become mature at 3-4 km
low compressibility. The underlying Cretaceous and Jurassic depth the generation of petroleum, particularly gas, contributes
sediments (5 W) show increases in density and velocity which strongly to this
probably are caused mostly by chemical compaction. Some of
Mudstones may contain significant amounts of when the concentration (activity) of silica in the
organic (amorphous) silica (opal A), particularly radio- porewater is high. The precipitation of quartz
larian and diatoms, in areas with high organic produc- provides a sink for the silica and lowers the silica
tivity. Siliceous sponges may also be an important concentration in the porewater so that the reaction
source of silica in fine-grained siltstones and can continue. Therefore the rate of quartz cementation,
sandstones. which is a function of temperature, controls this
Biogenic silica will react to form opal CT and reaction.
microcrystalline quartz at about 60–80 C and this It has been suggested that the silica released from
will also produce a strong stiffening of the mudstones the above reaction could be transported by diffusion
(Thyberg et al. 2009a, Peltonen et al. 2008, Marcussen into adjacent sandstones and precipitated as quartz
et al. 2009). Smectite becomes unstable and dissolves cement there. Recently, small authigenic (grown in
at temperatures above 70–100 C and mixed layer place) quartz crystals have been identified in
minerals and illite precipitate. smectite-rich mudstones which have been heated to
more than about 80–85 C (Fig. 13.4). This shows that
Smectite þ Kþ ¼ illite þ quartz the silica is conserved locally in the mudstones. Even
if the silica concentration in the mudstones were
In the case of iron-rich smectite, chlorite may form higher than in sandstones, diffusive transport in shales
as well. This reaction releases excess silica which would be very inefficient. In mudstones without smec-
must precipitate as quartz. Smectite is only stable tite or amorphous silica there are no obvious sources
a c
Fig. 13.4 (a and b) Authigenic microquartz precipitated in are found in mudstones which have been buried deeply enough
smectite-rich Late Cretaceous mudstones from the northern to reach temperatures (>70–80 C) which make illite replace
North Sea (from Thyberg et al. 2010). They can be distinguished smectitic, providing excess silica which is then precipitated as
from clastic quartz grains by their cathode luminescence micro-sized quartz crystals
responses (c and d) and their chemical composition (e). They
of silica to be precipitated as early quartz cement. At and releases silica which is precipitated as quartz
greater depth most of the quartz cement is probably (Bjørlykke 1981, Bjørlykke et al. 1986):
derived from pressure solution of detrital quartz.
In sandstones the quartz cement is sourced by solu- Al2 Si2 05 ðOHÞ4 þ KAlSi3 O8 ¼ KAl3 Si3 O10 ðOHÞ2
tion of quartz grains at grain contacts or along þ 2SiO2 þ H2 O
stylolites, but it is not clear to what extent silt and Kaolinite þ K-feldspar ¼ illite þ quartz þ water
sand grains floating in a matrix of clay will dissolve
and cause precipitation of quartz as cement or as This reaction is driven towards increased density
overgrowth on the grains. While quartz grains dis- (lower water content).
persed in a clay matrix may dissolve in contact with Kaolinite is however stable up to more that 200 C
clay minerals, the surrounding clay may prevent or if there is no K-feldspar or other source of potassium
retard overgrowth. available locally in the rock. It may then be replaced
At greater burial and temperatures (>130 C) kao- by pyrophyllite (AlSi205(OH)) which contains less
linite becomes unstable in the presence of K-feldspar water.
356 K. Bjørlykke
Mud containing mostly quartz, illite and chlorite Sediment compaction - rock shrinkage
will be chemically stable up to high temperatures Bulk modulus = Stress/strain(ΔV)
because these are metamorphic minerals. With If the strain ΔV is 0.001 or 0.1% and the bulk
increasing overburden and temperature, massive modulus is 50 GPa the effective stress is
mudstones develop a more pronounced cleavage typi- reduced by 50 MPa
k volume (VR) = Solids (V) + Fluids (porosity)
cal of shales.
Void ratio = VS/ Vf = ϕ/(1-ϕ)
This is due to a higher degree of parallel orientation
of the sheet silicate minerals, particularly mica, illite For isochemical reactions VS = const.
and chlorite. ΔV = Δϕ , dV/dt = dϕ/dt
During folding, high horizontal stresses may pro-
duce an axial plane cleavage. This is controlled by a
reorientation of clay minerals (sheet silicates) and also
a flattening (elongation) of quartz grains by stress- Fig. 13.5 Some definitions related to sediment compaction.
driven dissolution and precipitation. During mechanical compaction the strain is produced by an
Mudstones and shales which have a high organic increase in the effective stress. Chemical compaction in
sandstones and other siliceous sediments produces strain with-
content contain kerogen which often occurs as thin
out stress. The strain will however reduce differential stresses
lamina. Before the source rock becomes mature the
kerogen is a part of the solid phase and can carry some
of the overburden stress. When most of the kerogen is occurs preferentially in contacts with mica and clay
altered to oil and gas it becomes part of the fluid phase, minerals which favour the development of stylolites.
thus changing the fluid/solid ratio so that the pore Transport distance between the dissolution and precip-
pressure reaches fracture pressure which makes expul- itation sites is very short and is driven by diffusion,
sion more efficient. Shales with a lower organic con- and will be limited by the distance between the
tent also will generate petroleum and gas which may stylolites.
migrate into the most porous and permeable layers. Compaction-driven porewater can not explain sig-
Some of the oil and gas, though, will be retained in the nificant transport of silica. At normal geothermal
small pores as shale gas. There is now considerable gradients 3·109 volumes of water are required to pre-
interest in shale gas, particularly in onshore basins cipitate one volume of quartz. In addition, porewater is
where drilling costs are moderate. generally not moving upwards in relation to the sea-
floor so there is little cooling of the porewater (see
Chap. 4).
13.2 Sandstones As in mudstones, chemical compaction in
sandstones is mostly controlled by temperature and
Compaction of sand and sandstones has been both sandstones and mudstones compact chemically
discussed in Chap. 4. Mechanical compaction of during burial. Overpressure reduces the effective
sand is very sensitive to primary grain size and sorting. stress but has then little effect on compaction.
This is a function of the depositional environment. The loss of porosity results in higher density and a
Mechanical compaction is dominant at shallow burial reduction in the total rock volume or shrinkage
down to about 2 km (70-80 C) and at greater depth (Fig. 13.5). Even a very small loss of porosity (strain)
compaction is mainly chemical, involving mineral by chemical compaction will reduce the bulk volume
dissolution and precipitation. Dissolution at grain so that the stress is reduced.
contacts (pressure dissolution) is driven by the This shrinkage will contribute to a reduction in
increased solubility due to stress causing a slight horizontal stress because some of the compaction
supersaturation of silica with respect to quartz and a may occur in the horizontal direction. This is indicated
precipitation of new authigenic quartz (overgrowth). It by the leak-off pressures at greater depth (Fig. 13.6).
is now generally assumed that the precipitation, which This may reduce horizontal tectonic stresses.
is a function of temperature, is the rate-limiting step In the upper parts of sedimentary basins
and that this chemical compaction is therefore rather (<70–80 C) the compaction of siliceous sediments
insensitive to the stress. Dissolution at grain contacts follows the laws of soil and rock mechanics. At greater
Leak-off pressure data from Central Graben, North Sea

Pressure (psi)
0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
0
Leak-off pressure data
2000 Hydrostatic gradient
Lithostatic gradient
1 km
4000
Sub-sea depth (feet) 6000 σv = lith.grad. 2 km

σh = Leak-off
8000 pressure + τ
10000 3 km
12000
r = 2.25 4 km
14000
g/cm3
16000
5 km
r = 2.0 g/cm3
18000
10 20 30 40 50 60 70 80 90 100 MPa
Fig. 13.6 Leak-off pressure is an indication of the horizontal stress. At very slow strain rates a sandstone may respond nearly
stress and at 3–4 km this is nearly equal to the vertical stress. as a fluid where the stress is equal in all directions. The horizon-
This suggests that during chemical compaction the rocks com- tal stress is close to the vertical overburden stress. From The
pact both vertically and horizontally, thus reducing differential Millennium Atlas. Geological Society of London, 2003
Stress in passive margin basins with mostly siliceous sediments
Spreading ridge
Water
Mech. compaction
Sedimentary Rocks
Chemical
compaction Continental
basement
Oceanic crust
Ridge push Horizontal stress

Temperature
80–120°C from basement
relief Bjørlykke 2006
Fig. 13.7 Simplified cross-section through a sedimentary basin In the case of ice loading, the strain rates are relatively high and
on a passive margin. Most of the tectonic stress is transmitted the response in the sediments will be mostly mechanical com-
through the basement and the well-cemented sedimentary rocks. paction. Gravitational stress may also be important
depth compaction is mainly chemical and controlled Chemical compaction due to quartz cementation in
by temperature (Fig. 13.7). siliceous sediments will continue during uplift as long
When the compaction is mechanical any reduction as the temperature is higher than 70–80 C. Even a
in the effective stress due to uplift or the build-up of small shrinkage (porosity loss) due to this compaction
overpressure will cause the sedimentary rocks to will result in a reduction in the in situ stress controlled
become overconsolidated, and the deformation does by the bulk modulus. The mechanical extension due to
not follow the virgin loading curve (see Chap. 11). unloading will then at least partly be compensated for
358 K. Bjørlykke
by chemical compaction, and open fractures will grad- may have porosities exceeding 30% at nearly 3 km
ually be healed by quartz cement. burial depth due to high overpressure, because the
effective stress only corresponds to about 1 km with-
out overpressure.
13.3 Carbonate Compaction The processes controlling porosity loss in carbonate
sediments are still poorly understood.
Compaction of carbonates is controlled by principally The dissolution rate may be more important com-
very different processes than in siliceous sediments. pared to sandstones. At the contact between two cal-
Because the kinetics of carbonate dissolution and pre- cite grains there is probably only a very thin layer of
cipitation are so much faster than for siliceous rocks water, while clay minerals have a double layer due to
(mudstones and shales), temperature is not the main the negative surface charges. The transport of calcium
control on carbonate compaction. Cementation of along the calcite grain contacts may also be rate-
carbonate sediments into hard solid rocks may occur limiting. Carbonate grains, particularly of fossils,
right near the surface. In addition the presence of may have an organic coating which may influence
aragonite which is thermodynamically less stable precipitation.
than calcite provides a strong drive for cementation. Early porosity reduction in carbonate at shallow
The sediments then become mechanically over- depth may help to preserve the resultant porosity dur-
consolidated and may be unable to undergo further ing deeper burial. Carbonates have generally lower
mechanical compaction even when subjected to porosity than sandstones at the same depth
40–50 MPa (4–5 km depth). Carbonate sediments (Fig. 13.8) but there is a wide range of porosity/depth
like the Chalk, composed almost entirely of low-Mg values, particularly for carbonates.
calcite, undergo little dissolution and cementation also Near the surface where there may be meteoric
along stylolites at 2-3 km depth. Overpressure is very water flow the system is relatively open and net poros-
important in reducing both mechanical compaction ity may be created by dissolution. There is, however,
and pressure dissolution. In the Ekofisk Field, Chalk limited potential for mass transport of carbonate in
Sandstones (siliceous) versus carbonate reservoirs

0
2
Depth (km)
P90 Carbonate Reservoirs, Ehrenberg and Nadeau, 2005

5 P50 Carbonate Reservoirs, Ehrenberg and Nadeau, 2005
P10 Carbonate Reservoirs, Ehrenberg and Nadeau, 2005
P90 Siliciclastic Reservoirs, Ehrenberg and Nadeau, 2005
P50 Siliciclastic Reservoirs, Ehrenberg and Nadeau, 2005
6 P10 Siliciclastic Reservoirs, Ehrenberg and Nadeau, 2005
0 5 10 15 20 25 30 35
Porosity (%)
Fig. 13.8 Compaction trends for carbonates and sandstones indicate that 90, 50 and 10% of the reservoirs’ values have
(from Ehrenberg and Nadeau 2005). Average porosity versus higher porosity than this value
top depth for global petroleum reservoirs. P90, P50, and P10
solution during burial and the reactions must be nearly Bjørlykke, K. 2003. Compaction (consolidation) of sediments.
isochemical. This is because the porewater will always In: Middleton, G.V. (ed.), Encyclopedia of Sediments and
Sedimentary Rocks. Kluwer Academic Publ., Dordrecht, pp.
be closely in equilibrium with the carbonate minerals 161–168.
that are present. This leaves little potential for trans- Bjørlykke, K. 2006. Effects of compaction processes on stress,
port by diffusion, or by advection, because the flow faults, and fluid flow in sedimentary basins. In: Buiters, S.H.J.
rates are so small, particularly in relation to the and Schreurs G. (eds.), Analogue and Numerical Modelling of
Crustal-Scale Processes. Geological Society Special Publica-
isotherms. Focused flow as along faults will cause tion 253, pp. 359–379.
some dissolution because of the retrograde solubility Bjørlykke, K. 2014. Relationships between depositional
of carbonates like calcite. The solubility is also a environments, burial history and rock properties. Some prin-
function of pressure but flow across pressure barriers cipal aspects of diagenetic processes in sedimentary basins.
Journal of Sedimentary Geology 301, 1–14.
is rather limited. Bjørlykke, K., Aagaard, P., Dypvik, H., Hastings, D.S. and Harper,
Much of the compaction of carbonate rocks occurs A.S. 1986. Diagenesis and reservoir properties of Jurassic
along stylolites because the dissolution and transport sandstones from the Haltenbanken area, offshore mid – Norway.
along grain contacts are enhanced by the presence of In: Spencer, A.M. et al. (eds.), Habitat of Hydrocarbons on the
Norwegian Continental Shelf. Norwegian Petroleum Society,
sheet silicates (see Chap. 5). Graham & Trotman, London, pp. 275–286.
13.4 Summary compaction during burial in sedimentary basins. In: O.
Stephansson, J. Hudson and L. King (eds.), Coupled
Thermo-Hydro-Mechanical-Chemical Processes in Geo-
Shales, sandstones and carbonates follow different systems. Elsevier, London, pp. 699–708.
compaction trends and they are controlled by princi- Chuhan, F.A., Kjeldstad, A., Bjørlykke, K. and Høeg, K. 2002.
pally different processes. Porosity loss in sand by grain crushing. Experimental evi-
Both shales and sandstones compact mechanically dence and relevance to reservoir quality. Marine and Petro-
as a function of effective stress until chemical com- Chuhan, F.A., Kjeldstad, A., Bjørlykke, K. and Høeg, K. 2003.
paction takes over and further compaction is mainly a Experimental compression of loose sands: Relevance to
function of temperature and time. The initial mineral- porosity reduction during burial in sedimentary basins.
ogical and textural composition is very important both Canadian Geotechnical Journal 40, 995–1011.
Croizé, D., Renard, F., Bjørlykke, K. and Dysthe, D. 2010.
for sandstones and mudstones (shales). This is clearly Experimental calcite dissolution under stress: Evolution of
shown by experimental compaction of sediments with grain contact microstructure during pressure solution creep.
different mineralogical and textural composition. Journal of Geophysical Research Solid Earth 115, B09207,
Smectitic clays are very different from clays with doi:10.1029/2010JB000869, 15 pp.
Ehrenberg, S.N. and Nadeau, P.H. 2005. Sandstone vs. carbon-
other types of clay minerals. ate petroleum reservoirs; a global perspective on porosity-
Carbonate sediments may compact chemically at depth and porosity-permeability relationships. AAPG Bulle-
very shallow depth and low temperature and the com- tin 89(4), 435–445.
paction process is driven by a complex interaction Ehrenberg, S.N., McArthur, J.M. and Thirlwall, M.F. 2006.
Growth, demise and dolomitization of Miocene carbonate
between stress and chemical compaction, but the tem- platforms on the Marion Plateau, offshore NE Australia.
perature is less important than in siliceous sediments. Journal of Sedimentary Research 76, 91–116.
One of the main factors controlling compaction Ehrenberg, S.N, Nadeau, P.H. and Steen, Ø. 2008. A megascale
and rock properties in carbonates is the primary view of reservoir quality in producing sandstones from the
offshore Gulf of Mexico. AAPG Bulletin 92, 145–164.
content and distribution of aragonite, causing early Hesthammer, J., Bjørkum, P.A. and Watts, L. 2002. The effect
cementation. of temperature on sealing capacity of faults in sandstone
reservoirs – Examples from the Gullfaks and Gullfaks Sør
Fields, North Sea. AAPG Bulletin 86(10), 1733–1751.
Hovland, M., Bjørkum, P.A., Gudemestad, O.T. and Orange, D.
Further Reading 2001. Gas hydrate and seeps – Effects on slope stability: The
“hydraulic model”. ISOPE Conference Proceedings,
Stavanger, pp. 471–476, ISOPE (International Society for
Bjørlykke, K. 1981. Diagenetic reactions in sandstones. In: Offshore and Polar Engineering), New York.
Parker, A. and Sellwood, B.W. (eds.), Sediment Diagenesis. Marcussen, Ø., Thyberg, B.I., Peltonen, C., Jahren, J.,
NATO Advanced Study Institute, Reading, UK, Reidel Publ. Bjørlykke, K. and Faleide, J.I. 2009a. Physical properties
Co., pp. 169–213. of Cenozoic mudstones from the northern North Sea: Impact
360 K. Bjørlykke
of clay mineralogy on compaction trends. AAPG Bulletin 93 and Møre basins, Norwegian Sea. Petroleum Geoscience 14,
(1), 127–150. 127–138.
Marcussen, Ø., Thyberg, B.I., Peltonen, C., Jahren, J., Bjørlykke, Storvoll, V., Bjørlykke, K. and Mondol, N.H. 2005. Velocity-
K. and Faleide, J.I. 2009b. Physical properties of Cenozoic depth trends in Mesozoic and Cenozoic sediments from the
mudstones offshore Norway: Controlling factors on sediment Norwegian Shelf. AAPG Bulletin 89, 359–381.
compaction and implications for basin modeling and seismic Teige, G.M.G., Hermanrud, C., Wensås, L. and Nordgård Bolås,
interpretation. AAPG Bulletin 93(2), 1–24. H.M. 1999. Lack of relationship between overpressure and
Marcussen, Ø, Faleide, J. I., Jahren, J. and Bjørlykke, K. 2010. porosity in North Sea and Haltenbanken shales. Marine and
Mudstone compaction curves in basin modelling: a study of Petroleum Geology 16, 321–335.
Mesozoic and Cenozoic sediments in the northern North Sea. Thyberg, B., Jahren, J., Winje, T., Bjørlykke, K. and Faleide, J.I.
Basin Research 22, 324-340. February 2009. From mud to shale: Rock stiffening by
Mondol, N.H., Bjørlykke, K. and Jahren, J. 2008. Experimental micro-quartz cementation. First Break 27, 27–33.
compaction of clays. Relationships between permeability Thyberg, B., Jahren, J., Winje, T., Bjørlykke, K., Faleide, J.I.
and petrophysical properties in mudstones. Petroleum Geo- and Marcussen Ø. 2010. Quartz cementation in Late
science 14, 319–337. Cretatceous mudstones, northern North Sea: Changes in
Mondol, N.H., Bjorlykke, K., Jahren J. and Høeg, K. 2007. rock properties due to dissolution of smectite and precipita-
Experimental mechanical compaction of clay aggregates – tion of micro-quartz crystals. Marine and Petroleum Geol-
Changes in physical properties of mudstones during burial. ogy 27, 1752–1764.
Marine and Petroleum Geology 89, 289–311. Voltolini, M., Wenk, H.-R., Mondol, N.H., Bjørlykke, K.
Peltonen, C., Marcussen, Ø., Bjørlykke, K. and Jahren, J. 2008. and Jahren, J. 2009. Anisotropy of experimentally com-
Mineralogical control on mudstone compaction; a study of pressed kaolinite-illite-quartz mixtures. Geophysics 74,
late Cretaceous to early Tertiary mudstones of the Vøring D13–D23.
Chapter 14
Source Rocks and Petroleum Geochemistry
Knut Bjørlykke
As discussed in Chap. 1, petroleum is generated from The transformation of amino acids, carbohydrates,
organic matter which accumulates in sedimentary humic acids and other compounds into kerogen is
basins. Only a small fraction of the organic matter achieved by the removal of functional groups such as
produced in the photic zone in the ocean becomes acid groups, aldehydes and ketones. This involves a
trapped in sediments (Fig. 14.1). Most of the organic loss of oxygen from the organic material, also of
matter is formed by photosynthesis producing algae nitrogen, water and CO2.
(bacteria) and higher organisms which feed on algae. Kerogen therefore has higher H/C, and lower O/C,
There there may also be a supply of organic matter ratios than the initial compounds (Fig. 14.3).
from land. Most of the organic matter is oxidised in the Kerogen may also include organic particles of mor-
water column or on the seafloor and the nutrients are phologically recognisable biological origin such as
released into the water and become available for new vitrinite (derived from woody tissues and liptinite
organic production near the surface during upwelling. materials, e.g. algae spores, cuticles, etc.). Because
Most source rocks are black shales like the Upper of its resistance to strong oxidising acids kerogen can
Jurassic Kimmeridge Clay and its equivalents in the be recovered from sedimentary rocks by dissolving
North Sea basin (Fig. 14.2). Formation of rich source most of the rock away with HCl or HF.
rocks requires that the organic matter is not diluted too It is also possible to separate kerogen by a density
much by deposition of clastic material or by precipita- method, using heavy liquids, because kerogen is ligh-
tion of carbonates. ter than minerals. The resulting concentrate of kerogen
The organic matter is transformed into kerogen can be studied microscopically using transmitted and
which consists of very large and complex molecules. reflected normal light, to identify the biological origin
We do not usually apply the term kerogen to fresh and the degree of thermal alteration. These phases of
organic material, but to material which is somewhat altered organic material are called macerals. Algal
dehydrated after burial to about 100 m or more. Kero- material has a dull appearance, while wood and mate-
gen is formed gradually within the upper few hundred rial from higher plants is called vitrinite. Vitrinite
metres of the sediment column after deposition from becomes increasingly shiny when exposed to higher
precursor products like humus, and humic and fulvic temperatures and by measuring the amount of
acids. The organic matter may be derived from marine reflected light in the microscope we obtain an expres-
organisms, mostly algae, or from plants derived from sion for the degree of thermal alteration (Vitrinite
land. index).
It is also useful to use ultraviolet light microscopy,
since certain components, i.e. liptinites, display char-
acteristic fluorescence colours. Infra-red spectroscopy
(IR) or nuclear magnetic resonance (NMR) spectros-
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway copy can be used to investigate the chemical composi-
e-mail: [email protected] tion and structure of kerogen.
362 K. Bjørlykke
Photosynthesis phytoplankton, zooplankton

Sea level
(consume phytoplankton)
Plankton eaten by higher marine organisms
Formation of pellets
Breakdown of organic material

oxygen minimum
Adsorption of dissolved organic material

onto clay minerals
Accumulation of pellets and other organic matter

Sea floor
Oxidation
Ox.
Red. Breakdown by burrowing organisms and bacteria
Preserved organic matter in sediment
Fig. 14.1 Formation of source rocks. Only a small fraction of the organic matter is preserved. The formation of organic-rich source
rocks requires restricted water circulation and oxygen supply
Fig. 14.2 Draupne shale (Kimmeridge shale) cores from northern North Sea (Block 34/7)
Being a complex of very large molecules (poly- classified into 3 main types which may be plotted as a
mer), kerogen is difficult to analyse, but upon heating function of the H/C ratio and the O/C ratio (Fig. 14.3)
to 350–450 C in an inert atmosphere (pyrolysis) it will Type I sapropelic kerogen is formed from organic
break down into smaller components which can then material with a high content of lipids with long ali-
be analysed by means of gas chromatography and phatic chains. It consists of spores and planktonic
mass spectrometry. algae, as well as animal matter, which have been
Kerogen has a wide range of compositions, depen- broken down microbially after deposition in the sedi-
dant on the original organic composition, but may be ment. Sapropelic material which consists of fats, oils,
14 Source Rocks and Petroleum Geochemistry 363
H/C Type III humic kerogen is derived from organic

2.0 matter from land plants, such as lignin, tannins and
elic cellulose. This type has a low initial H/C ratio, and a
rop
Sap gen high initial O/C ratio, reflecting the composition of the
H2O I ke r o
precursor plant matter. In maturing (through the effect
CO2
of temperature) this kerogen, which is often called
1.5
II Type III (Tissot and Welte 1984), generates abundant
water, CO2 and methane (CH4). Most coals have a
Diagenesis
composition and structure similar to Type III
4
CH
kerogens. Coal generates mostly gas but some coals

1.0 Humic may also generate some oil.
III kerogen
Katagenesis
14.1 Transformation of Kerogen with

Burial and Temperature Increase
Oi
0.5
l
Meta
W
et
genesis
ga
Dry Source rocks compact mechanically as a function of

s
gas
O/C effective stress similar to mudstones and the porosity
0.1 0.2
and void ratio which is defined by the volume of solids
Fig. 14.3 Diagram (Van Krevelen diagram) showing the pri- (Kerogen and minerals). When the kerogen gradually
mary composition of the different types of kerogen and the becomes mature some of the solid kerogen becomes
changes as a function of heating (maturation) during progressive fluid petroleum, thus changing the void ratio. This is
burial
an important factor in the expulsion of oil and gas. The
transformation of kerogen is mainly a function of
waxes, etc., has a high H/C ratio, usually between 1.3 temperature even if the pressure may have some
and 1.7. This kind of kerogen is often called Type I, effect.
and contains little oxygen (O/C <0.1). It will provide With increasing temperature the chemical bonds in
mainly oil, with less gas (CH4 and CO2). Type I these large molecules (kerogen) are broken and kero-
kerogen is typical of oil shales, especially in freshwa- gen is transformed into smaller molecules which make
ter basins like the Green River basin in Colorado, up oil and gas. This requires that the temperature must
Wyoming and Utah, but is also found in marine basins. be 100–150 C over long geological time (typically
Type II kerogen is a mechanically and chemically 1–100 million years).
complex mixture of algae and other marine organisms The conversion of kerogen to oil and gas is thus a
and plant debris. The composition varies considerably, process which requires both higher temperatures than
depending on the initial organic precursor materials one finds at the surface of the earth and a long period
which again may be linked to depositional facies. of geological time. Only when temperatures of about
Type II kerogen represents a composition midway 80–90 C are reached, i.e. at 2–3 km depth, does the
between types I and III but it does not represent a conversion of organic plant and animal matter to
mixture of these end members. It has relatively high hydrocarbons very slowly begin to take place. About
H/C, and low O/C, ratios, but contains more oxygen- 120–150 C is the ideal temperature range for this
containing compounds (ketones and carboxyl acid conversion of kerogen to oil, which is called matura-
groups) than Type I. Esters and aliphatic chains are tion. This corresponds to a depth of 3–4 km with a
also common. This is the usual type of kerogen found normal geothermal gradient (about 30–40 C/km). In
in marine basins where mixtures of phytoplankton, volcanic regions organic matter may mature at much
zooplankton and micro-organisms have accumulated lesser depths due to high geothermal gradients (e.g.
under reducing conditions, sometimes along with high heat-flow areas). In large intracratonic sedimen-
land-derived plant material. This type of kerogen is tary basins or along passive margins, however, the
the most common source of oil (Fig. 14.3). geothermal gradient may be only 20–25 C/km and
364 K. Bjørlykke
the minimum overburden required to initiate petro- of petroleum changes the solid/fluid ratio and particu-
leum generation will be correspondingly greater larly gas will reduce velocity and density. Mapping
(4–6 km). out source rocks and their maturity has become a new
In general one can say that petroleum can not be and important exploration technique in some sedimen-
generated near the surface except locally through the tary basins (Løseth et al. 2011). Seismic data may also
influence of hydrothermal and igneous activity. Shal- be used as indications of the degree of maturity.
low deposits of oil and gas which we find today were
actually formed at great depths and either the overbur-
den has been removed by erosion or the hydrocarbons
have migrated upwards considerable distances. As we 14.2 What Factors Influence the
have already seen, however, large amounts of natural Maturation of Kerogen?
gas, chiefly methane (CH4), may be formed near the
surface by biochemical processes. The term “maturity” refers here to the degree of ther-
Temperature increases with increasing overburden, mal transformation of kerogen into hydrocarbons and
causing the carbon-carbon bonds of the organic ultimately into gas and graphite. The conversion of
molecules in the kerogen to rupture. This results in kerogen into hydrocarbons is a chemical process
smaller hydrocarbon molecules. When kerogen matu- which takes place with activation energies of around
ration reactions are completed, the kerogen’s 50–60 kcal/mol. This energy is required to break
“organic” components, which may be derived from chemical bonds in the kerogen which consists of
lipids, fatty acids and proteins, have been converted very large molecules (polymers) so that smaller hydro-
into hydrocarbons. carbon molecules can be formed.
As the temperature rises, more and more of the It has been assumed that formation of oil is a first
bonds are broken, both in the kerogen and in the order reaction, the rate of which is an exponential
hydrocarbon molecules which have already been function of time. Understanding the factors which
formed. This “cracking” leads to the formation of influence the rate of this reaction is of great interest.
lighter hydrocarbons from the long hydrocarbon Four factors are thought to contribute:
chains and from the kerogen. The removal of gas, 1. Temperature
mainly CH4, leaves the residual kerogen relatively 2. Pressure
enriched in carbon. At the outset kerogen (Type I 3. Time
and II) has an H/C ratio of 1.3–1.7. Humic kerogen 4. Minerals or other substances which increase the
(Type III), which has high initial oxygen contents, rate of reaction (catalysts) or which inhibit
gives off mostly CO2 gas and so its oxygen/carbon reactions (inhibitors).
ratio gradually diminishes. This diagenetic alteration Temperature is clearly the most important factor,
begins at 70–80 C and as water and CH4 are removed and hydrocarbons can be produced experimentally
the H/C ratio will fall to about 0.6 and the O/C ratio from kerogen by heating it (pyrolysis). This reaction
will become less than 0.1 at about 150–180 C. is time-dependent and in laboratory experiments,
In the North Sea basin, most of the oil is generated where time is more limited than it is in nature, fairly
at temperatures around 130–140 C, which equates high temperatures (350–550 C) have to be used in
with a depth of about 3.5 km. If temperatures higher pyrolysis. This is the case when oil and gas is pro-
than 170–180 C persist for a few million years, all the duced from oil shales by pyrolysis of immature kero-
longer hydrocarbon chains will already have been gen in ovens after quarrying (see chapter 21). Pressure
broken (cracked), leaving us only with gas – mainly appears to play a minor role but increasing pressure
methane (dry gas). The kerogen composition will should reduce the rate of the reaction because of the
gradually be depleted in hydrogen and move towards increase in volume involved in the formation of
pure carbon (graphite) (H/C!0). hydrocarbons (Le Chatelier’s rule).
Source rock can in some cases be mapped out on There is a relatively small volume increase when
seismic. Source rocks have in most cases lower density kerogen becomes oil, even though oil is lighter than
and velocity (low AI) than sandstones, limestones and kerogen. This is due to the residual (coke) which
shales. When the kerogen becomes mature, generation remains unaltered.
When kerogen is converted directly into gas, or During this transformation of organic matter, water
from oil which has been formed first, there is a marked and oxygen-rich compounds are liberated first, then
volume increase. This should lead to slower reaction compounds which are rich in hydrogen. This conver-
rates under high pressure in a closed system and retard sion results in enrichment of carbon and the colour of
generation of gas. Generation of petroleum, particu- the residual kerogen changes from light yellow to
larly gas, may contribute to the formation of overpres- orange, brown and finally black. These gradations
sure but in a sedimentary basin the pressure will for the can best be registered by measuring light absorption
most part be controlled by the flow of water which is of fossil pollen and spores (palynomorphs). It is also
the dominant fluid phase. In the source rock however possible to analyse colour changes in other kinds of
overpressure is likely to develop, causing fossils, for example conodonts.
hydrofracturing which helps to expel the generated For application in exploration a rapid semi-
petroleum. The main cause for overpressuring is, how- quantitative method has been developed whereby
ever, not only the increase in fluid volume but the these colour changes are estimated from smooth
transformation of solid into fluids. When solid kerogen spores examined under transmitted light and compared
is transformed into fluid oil or gas the ratio between with a standard colour scale. This parameter is called
the solid phase and the fluid phase is changed, as the “Thermal alteration index” (TAI) and will give a
expressed by the porosity and the void ratio. rough idea of the thermal maturity of the sediments
Temperature is however the most important factor and their temperature history.
controlling petroleum generation. Another way of assessing palaeotemperatures at
It has long been suspected that minerals, particu- which alteration took place in sedimentary rocks is to
larly clay minerals, might affect the rate of hydrocar- record the degree of carbonisation of other plant
bon generation. A number of laboratory experiments remains which are usually present. Vitrinite, which
have been carried out in which kerogen is mixed with originally was fragments of woody tissue, is a com-
various minerals but the results have not been mon component of coal but is also found in smaller
conclusive. amounts in marine source rocks. This material is
The conversion of organic matter begins at analysed by measuring the amount of light it reflects.
70–80 C, given long geological time. Between 70 It becomes shinier and reflects light better as the
and 90 C the transformation of kerogen proceeds degree of carbonisation increases. By measuring the
very slowly, and it is only in ancient, organic-rich reflectivity of vitrinite particles under a reflected light
sediments that significant amounts are formed. Most microscope an exact value is obtained for this maturity
of the maturation process occurs between 100 and parameter, expressed by the reflectivity coefficient R0
150 C. Here the degree of kerogen transformation is (% vitrinite reflectance). If R0 is less than 0.5% in a
also a function of time. This means that rocks which shale it can not have generated much oil and is classed
have been subjected to 100 C for 50 million years are as immature. Shales with R0 ¼ 0.9–1.0 have been
more mature than rocks which have been exposed to exposed to temperatures corresponding to maximum
this temperature for 10 million years. As the organic- oil generation. R0 ¼ 1.3 represents the upper limit for
rich sediment (source rock) is buried in a sedimentary oil generation, above which the shale will only pro-
basin, it will normally be subjected to increasing tem- duce condensate (light oils) or gas.
perature as a function of increasing burial depth. If we For certain source rocks the ratio between extract-
know the stratigraphy of the overlying sediment able alkenes (paraffins) with an even number of carbon
sequence and the geothermal gradient and the subsi- atoms per molecule and those with an odd number
dence curve, we can calculate the temperature as a may also be an expression of maturity. In plant mate-
function of time. rial and in marine algae, one finds a higher abundance
At low degrees of maturity we find more of the of alkenes with an odd number of carbon atoms than in
alkenes (olefins) and cykloalkenes (naphtenes), transformed organic matter like waxes and fatty acids.
which have high H/C ratios, while with greater matu- The decrease in this predominance of odd over even in
rity there is an increase in the proportion of aromates source rocks with increasing maturity is due to the
and polyaromates (low H/C ratio). Oil thus acquires dilution of the original biologically derived n-alkanet
increasing gas content with increasing maturity. mixture with a newly generated mixture which has a
366 K. Bjørlykke
regular carbon number distribution. This odd/even mol. A is an exponential constant dependant on the
ratio is normally expressed by means of an index type of kerogen.
called the Carbon Preference Index (CPI): Since the temperature changes during the burial
history the effect of temperature has to be integrated
CPI ¼ n-alkenesðoddÞ=n-alkenesðevenÞ over the range of temperatures that the kerogen is
exposed to. This may be expressed by the Time Tem-
It is based on analyses of alkenes with carbon perature Index (TTI)
number between 25 and 33 (C25 and C33) in oils.
Z tx
However, organisms also start off with different
ðTTIÞ ¼ 2FðT Þ dt
CPI index values, and land plants have high ratios t0
between odd and even carbon numbers. Bacteria
have a predominance of even carbon numbers. This is the integrated temperature (T) over time
The maturation process will cause a shift in the from the time of deposition (t0) to the present day
carbon number distribution towards smaller (tx). F is a factor dependant on the activation energy
molecules, particularly in the range C13–C18. for the reactions. The reaction rates may approxi-
Oil that comes from carbonate source rocks often mately double for each 10 C increment. The matura-
has a low CPI index, while oil derived from plants has tion of kerogen to petroleum can be successfully
a high index value. With increasing temperature the modelled (Makhous and Galushkin 2005) but the
CPI index goes towards 1, that is to say, equal even results depend very much on the input data with
and odd carbon numbers. respect to the temperature history. The activation
The isotopic ratio also changes because the bonding energy may also vary significantly for different types
between hydrogen and 13C, and between hydrogen and of source rocks. If the geothermal gradient can be
12
C, are not equally stable. Light gases such as meth- assumed to have been constant throughout the relevant
ane, are enriched in the light isotope 12C, and the period, this is relatively simple. However, if the geo-
hydrocarbons that remain will therefore have an thermal gradient has varied considerably through time
increasing 13C/12C ratio with increasing temperature. it is a lot more complicated.
There is isotopic fractionation of carbon, and when A theoretical maturity parameter (P) can be calcu-
kerosene is releasing petroleum, this phase is some- lated by integrating temperature with respect to time:
what enriched in 11C corresponding to the precursor
Z t
kerosene. Gases, particularly methane, normally have
lighter carbon isotopes than kerosene and oil. When P ¼ ln 2T=10 dT
0
methane is formed from larger hydrocarbon molecules
by thermal cracking, the 12C–12C bond is less stable t, geological time (million years); T, temperature ( C).
than the 13C–12C bond and the product becomes We see that a doubling of the reaction rate for every
enriched in 12C (low δ13C). 10 C is built into this expression (Geoff 1983). This is
an expression which is very similar to Lopatin’s Time-
Temperature Index (TTI) (Waples 1980) which
14.3 Modelling of Petroleum Generation integrates the temperature the source rock is subjected
to with respect to the burial time.
The rate of petroleum generation can be calculated and When the temperature rises above about
modelled. 130–140 C, maturation proceeds very rapidly, and
We assume that the rate (ki) of petroleum genera- then the time factor is less crucial.
tion follows an Arrhenius function: There are differing views as to how much emphasis
should be placed on time in relation to temperature in
ðki Þ ¼ A expð Ei =RT Þ the matter of maturation. Oil companies use different
formulae for calculating these temperature factors and
where R is the gas constant, T is the temperature and Ei the 10-degree rule is now found not always to be valid,
is the activation energy. The activation energy mostly particularly for very young sedimentary basins with
varies between 50 and 80 kcal/mol or about 200 kJ/ high geothermal gradients.
The rate of sediment heating, i.e. temperature Time

increase per unit time, can be expressed as dT/dt.
The rate of subsidence is dZ/dt, and the geometrical
gradient dT/dZ. We then get: A
C
Rate of heating ðdT=dtÞ ¼ rate of subsidence B
ðdZ=dtÞ geothermal gradient ðdT=dZ Þ:
In basins like the North Sea the rate of heating is

typically only 1–2 C/million years but during rapid
subsidence and sedimentation in the late Cenozoic
the heating rate was higher. During the deposition of
1–1.5 km of upper Pliocene and Pleistocene sediment
Depth
in 2–3 million years, the gradient was lowered but the Temperature
heating rate must have been about 10 C/million years.
The maturity of source rocks can now be calculated Fig. 14.4 The maturity of a source rock is a function of the
time – temperature index (TTI). This can be calculated from the
with the help of basin modelling integrating tempera- burial curve if the geothermal gradients are known. The geo-
ture over time. The subsidence curve for the source thermal gradients are most critical during the deepest burial
rock is determined from the stratigraphic age and because the maturation is an exponential function of the tem-
thickness of the overlying sequence. perature. At a certain depth and temperature the maturity may
vary greatly and burial curve C will produce the highest maturity
By estimating a geothermal gradient, depth can be at this depth. Source rocks buried following curve A and B are
converted to temperature. This way we obtain a curve less mature because their exposure to greater burial depth
that shows the temperature history of the kerogen (higher temperatures) has been much shorter
through geological time. Integration of this curve
enables the maturity to be calculated (Fig. 14.4).
The sample of shale is crushed and heated to
If we know the stratigraphy of the overlying
300 C, at which point one measures the amount of
sediments and the geothermal gradient, temperature
hydrocarbons that are already formed in the source
can be calculated as a function of time.
rock but have not migrated out. The content of hydro-
Calculating the maturity is important not only for
carbon with carbon numbers between C1 and C25, is
predicting where the source rocks are sufficiently
called S1. It is measured as the area beneath the peak
mature to produce oil and gas (or perhaps over
S1 (Fig. 14.5).
mature). It is also important for estimating the timing
On further heating from 300 to 550–600 C, new
of the generation and migration of oil and this type of
petroleum is formed in the laboratory from the kero-
basin modelling has become an important part of oil
gen by heating (pyrolysis), and this amount is called
exploration. The most important input into the calcu-
S2. This is a measure of how much oil and gas could
lation is the subsidence history as derived from the
have been generated if the source rock and been buried
stratigraphic record, and the estimated geothermal
deeper. The reason it requires such high temperatures
gradient as a function of geological time. The success
is that the heating in the laboratory lasts just a few
of the basin modelling depends as always very much
minutes or hours, instead of some millions of years.
on the quality of the input data.
During heating from c. 300 to 550 C, CO2 is also
formed and is collected and measured separately as the
S3 peak (Fig. 14.5). Most of the CO2 groups dissociate
14.4 Rock-Eval Analyses from the kerogen between 300 and 390 C. The gener-
ation of petroleum varies with temperature and
Rock-Eval is a standard routine analysis of source reaches a peak corresponding to S2 (Fig. 14.5). This
rocks, usually shales, to establish how much of the temperature, which gives the maximum petroleum
kerogen has been transformed into petroleum and how generation, is called Tmaks and is typically in the
much can be transformed at a higher temperature. range 420–460 C.
368 K. Bjørlykke
Generation of migration with respect to the formation of traps.

hydrocarbons Once we know with reasonable certainty when oil
S3
generation and migration took place, we can attempt
to construct maps to show the structures and faults at
that time. For example, if a trap was formed by folding
S2 after the oil had migrated, it can not have captured any
S1 of the oil, but perhaps the gas which is formed later.
The generation of oil and gas lends itself to mathe-
matical modelling, and there are good programs which
make this easier. The most important data which have
to be fed in are the subsidence rate, the geothermal
300 400 500 600 C
gradient and the activation energy of the kerogen.
Fig. 14.5 Rock-Eval analyses. A rock sample representing a The activation energy varies according to the type
possible source rock is heated gradually to about 550 C while of kerogen and is an important factor when it comes to
the amount of hydrocarbons generated is measured. At about calculating the timing of the oil generation.
300 C oil and gas which has already been generated in the
In nature natural processes cause fractionation of
source rock is expelled and measured as the S1 peak. The peak
at about 400–460 C represents the amount of hydrocarbons carbon which is reflected in the isotopic composition
generated from the kerogen in the sample. The temperature of of CO2 (Fig. 14.6). Note that the CO2 released by
peak HC generation is called the Tmax. The Hydrogen Index bacterial fermentation has a positive δ13C while the
(HI) ¼ S2 (S)/TOC (total organic carbon) is a measure of the
carbon from maturation of kerogen (thermal decarb-
potential of the source rock to generate petroleum. The total
amount of CO2 generated is measured as the S2 peak. The oxylation) has negative values.
Oxygen Index (OI) ¼ P3/TOC Re-Os dating of petroleum has made it possible to
date the age of the source rock for oil found in
rerservoirs. Rhenium (Re) and Osmium(Os) are metals
The ratio between the amount of petroleum that are concentrated under reducing conditions in
generated (S2) and the total content of organic material organic matter and also in sulphides. They are released
(TOC) is known as the Hydrogen Index (HI). under oxygenated conditions during weathering. Their
The quantity of CO2 which is formed (S3), is lim- relative concentrations change as a function of geo-
ited by the oxygen content of the kerogen. The S3/ logic time relecting different conditions in the hydro-
TOC ratio is the Oxygen Index (OI). sphere and lithosphere. The isotope 187Re decays to
187
The ratio between the quantity of free oil already Os and as a result the ratio of 187Os /188Os increases
formed (S1) and the total amount of petroleum (S1 + with geologic time. The age of of organic matter and
S2), is an expression of how much petroleum is still sulphides in black shale can the be determined from an
left in relation to how much has already been isochron defined by the 187Re/188Os ratio and the
187
generated. The S1/(S1 + S2) ratio is the Production Os/188Os ratio (Hannah et al. 2012). This method
Index (PI). Good source rocks have a high production is now being developed to include dating the age of the
index. source rocks for the oil in oil reservoirs.
When we analyse source rocks, we must however
take into account that some of the oil which has been
generated has migrated out of the source rock: the S1 14.5 Composition of Petroleum
peak represents the remaining petroleum.
Good source rocks are the first prerequisite for Naturally occurring petroleum has a very complicated
finding oil and gas in a sedimentary basin. If they are chemical composition. The most important
not present, one can save oneself the trouble of further hydrocarbons occurring in oil are alkanes (paraffins,
prospecting. However the quality of the source rock CnH2n+2). Important gases are methane (CH4), ethane
can vary through the basin and the type of source rock (C2H6), propane (C3H8) and butane (C4H10). Paraffins
determines the composition of the oil. from carbon numbers 5, pentane (C5H12) to 15,
The timing of the oil and gas generation is also very pentadecane (C15H32) occur mainly as liquids at
important. It also determines the timing of oil room temperature. These paraffins are an important
Rate of supply of CO2 Carbon isotope composition

Concentration (mM/L) derived from organic (δ13 C PDB) of CO2
10 20 30 40 50 carbon (and CH4)
Marine
carbonate 0 (approx.)
0
Bacterial
oxidation –25
HCO3 –
SO42–
CH2+O2→ CO2 +H2O
10–2
Bacterial
Depth in m below seafloor
sulphate
HCO3–
→
reduction –25
2CH2O + SO42– → →
SO42–
2CO 2 + 5 2– + 2H2O
101
CH4
Bacterial
+15
HCO3 –
fermentation → (–60 to –90)
2 CH2O → CH4 (methane)
CH4O + CO 2 CH4
2–3.103
Thermal –10 to –25

decarboxylation
Due to oxidation by Fe3+ reduction Isotopically light CO2

Due to thermal decarboxylation Isotopically heavy CO2
Fig. 14.6 Natural fractionation of carbon isotopes in CO2 while bacterial fermentation produces positive values. Thermal
generated in sediments during burial. During oxidation by oxy- decarboxylation also produces negative δ13C values
gen or sulphate reduction the δ13C values are highly negative
part of the gasoline fraction of oil. Higher paraffins are hydrocarbons have a strong odour and a lower boiling
waxy and almost a solid phase at room temperature but point than aliphatic hydrocarbons with the same num-
become less viscous at higher temperature. ber of carbon atoms. Aromatic compounds make up
Naphthenes (cycloparaffins) are a series of cyclic, 10–39% of crude oil. Refinement of crude oil with a
saturated hydrocarbons with the general formula high content of aromatic compounds results in a high
(CH2)n. Cyclopentane (C5H10) and cyclohexane octane product.
(C6H12) are important members of the naphthene Asphalt is brown to black, solid or highly viscose
series which are found in all types of petroleum. petroleum consisting of hydrocarbons with high
Naphthenes with lower carbon numbers, C3H6 and molecular weights. Asphalt is rich in aromatic
C4H8, are gases. The percentage of naphthenes in compounds and naphthenes, and is enriched in sul-
petroleum varies from 7 to 31. Petroleum rich in phur, nitrogen and oxygen.
naphthenes is called asphalt-based crude, because The sulphur content of crude varies from 1 to 5–6%
many of the more complex naphthenes have high and occurs as pure sulphur, hydrogen sulphide (H2S)
boiling points and high viscosity. and organic sulphur compounds. Components in oil
Olefins (alkenes) are hydrocarbons with the same that contain sulphur, nitrogen and oxygen, are called
chemical composition as alkanes, but contain one or NSO compounds.
more double bonds. Hydrocarbons of this type do not The density of crude oil is measured using the unit
normally occur in crude oil. Aromatic hydrocarbons API gravity where “gravity” is expressed as an inverse
are unsaturated, cyclic, and have the general formula value in relation to density as defined by the American
CnH2n 6. For example benzene, C6H6, and toluene, Petroleum Institute.
C6H5CH3, are important ingredients of car petrol on
account of their high octane numbers. Aromatic API gravity ¼ 141:5=ρ 131:5
370 K. Bjørlykke
Here ρ is the density of the petroleum at 60 F 2. High shrinkage oil: Oil which after production and
(15 C). separation at ordinary temperature and pressure
We see from the formula that a light oil with a contains a smaller percentage of liquid (less than
density (ρ) of 0.8 g/cm3 has a gravity close to 45, 70%) and a large amount of gas.
whereas a heavy oil with a density of 1.0 g/cm3 has a 3. Condensate: Gas which turns into a liquid with
gravity of 10. Oil with API gravity of about 15–20 or pressure reduction. Condensates have low densities
lower is considered heavy oil. (about 40–60 API).
Crude with low API gravity, contains more sulphur 4. Dry gas: This is gas that does not form a liquid
than the light crudes. phase at normal temperature. This is especially
Pure (native) sulphur also occurs in great quantities methane and other molecules with low carbon num-
in some oil fields, and most of the world’s sulphur bers (less than 6, hexane).
production comes from sources of this type. In most 5. Wet gas: This is gas that in addition to methane also
oil reservoirs, however, sulphur is an unwelcome com- contains significant amounts of alkanes with high
ponent, since it causes corrosion and problems in the carbon numbers. It may be in a gaseous phase in the
refining process. Burning oil of this type releases SO2 reservoir, but after production and separation may
into the atmosphere. Much of it can be removed during form a good deal of liquid at normal pressure and
refining, but that increases the cost. temperature.
6. Undersaturated crude: Oil which contains less gas
than is potentially soluble at reservoir temperature
Elemental composition of hydrocarbons in percent (after Hunt and pressure.
1979)
7. Saturated crude: Oil which contains as much gas
C H S N O as can be dissolved at reservoir pressure and
Oil 84.5 13 1.5 0.5 0.5 temperature.
Asphalt 84 10 3 1 2 Oil saturated with gas will form gas bubbles if the
Kerogen 79 6 5 2 8 pressure is reduced due to tectonic uplift or during
production in a reservoir.
The NSO compounds include resins and
asphaltenes which are enriched in oxygen, nitrogen
and sulphur. In addition hydrocarbons contain a num- 14.6 Gas
ber of trace metals derived from the organic matter in
the source rock. The most important are Ni and V. Methane is a thermodynamically stable compound,
even at temperatures of 500–600 C or more. As a
API gravity and density of some hydrocarbons
result we may find methane gas at very great depths
Density in sedimentary basins and also deep in the Earth’s
Molecule H/C gravity g/ crust. The limiting factors for economic exploitation
Hydrocarbon type Formula ratio API cm3
of gas of this type are the low porosity and permeabil-
Hexane Paraffin C6H14 2.3 82 0.6594
ity we normally find in reservoirs at these depths
Cyclohexane Naphthene C6H12 2.0 50 0.7786
Benzene Aromatic C6H6 1.0 29 0.8790
(6–8 km), and the expense of deep drilling. The main
component of natural gas is normally methane (CH4),
but ethane (C2H6), propane (C3H8) and butane (C4H10)
Hydrocarbons with a high H/C ratio tend to have
may also be important. In addition we have varying
lower densities.
amounts of CO2, H2O, nitrogen, hydrogen and inert
Classification of crude oil:
gases such as argon and helium.
1. Low shrinkage oil: Oil which after production and
Methane produced by bacterial breakdown of
separation at ordinary pressure and temperature
organic matter at low temperatures (<60 C) is
contains a high percentage of fluid (>80%) and
characterised by a light carbon isotope composition.
little gas.
This kind of gas is often referred to as “shallow bio- production from reservoirs containing heavy,
genic gas”. biodegraded oil. Increasing oil prices have however
made production of heavy oil and tar sand more prof-
itable. Heating the oil may however require up to 30%
14.7 Biodegradation of the energy recoved thus increasing the CO2 emis-
sion (see tar sand and oil shales, Chap. 21). Excavation
Hydrocarbons can be broken down by micro- and separation of oil from sand also require much
organisms (bacteria, yeast and fungi). This is a form more energy than production of conventional oil.
of biological oxidation whereby hydrocarbons are Reservoirs buried to greater depth (>80–100 C)
oxidised to alcohols, ketones and various acids. The are sterilised by the heat and may remain without
biological breakdown of hydrocarbons proceeds far biodegradation also after uplift to lower temperatures.
more rapidly with smaller molecules (carbon number It is difficult to introduce new bacteria. Biodegration
<20). When it comes to molecules with the same and the formation of heavy oil and tar sand occurs
carbon number, n-paraffins will break down first, most readily when oil migrates into reservoir rocks
and then isoparaffins, naphthenes and aromatics. which have not been sterilised at temperatures exceed-
Isoprenoids, steranes and triterpanes show the greatest ing about 80 C. When oil forms seeps near the surface,
resistance to biodegradation. Relatively rapid biodeg- aerobic biodegradation occurs rather rapidly.
radation depends on a supply of oxygen, usually In recent years it has become possible and econom-
dissolved in water or in air. If sulphate is present ical to produce oil and gas directly from shales that are
hydrocarbons may also be broken down by sulphate- source rocks. This is the petroleum which has not been
reducing bacteria which use the oxygen in sulphates expelled from the source rocks. This is shale oil
(e.g. gypsum) to oxidise some of the oil. Minerals and shale gas which has been generated at greater
containing trivalent iron (Fe3+), such as haematite, depth to become mature but is produced by drilling
can also contribute to the biodegradation and oxida- at shallow depth in uplifted areas, mostly on land (see
tion of oil deeper in the basin when ferric iron is Chaps. 1, 23).
reduced to ferrous (Fe2+). The content of haematite
and other minerals with trivalent iron in sediments can
thus also be an oxidant. Further Reading
Anaerobic biodegradation may occur under reduc-
ing conditions deeper in a sedimentary basin and the
supply of nutrients such as phosphorus in the Geoff, J.C. 1985. Hydrocarbon generation and migration from
sediments may then be critical. Jurassic source rocks in the East Shetland Basin and Viking
Graben at the North Sea. Journal of the Geological Society
The temperature must however be lower than about 140, 445–474.
80 C. The bacteria eat the lighter compounds so that Hannah, J.L. and Stein, H.J. 2012. Re-Os geochemistry. In:
the larger, more asphaltic, compounds are enriched. Melezhik, V.A., Kump, L.R., Fallick, A.E., Strauss, H.,
Biodegraded oils (tar) have therefore high viscosity. Hanski, E.J., Prave, A.R. and Lepland, A., (eds), Reading
the Archive of Earth’s Oxygenation, Volume 3: Global
Biodegradation occurs as soon as the oil flows out Events and the Fennoscandian Arctic Russia – Drilling
at the surface as oil seepage and the lighter compounds Early Earth Project: Springer-Verlag, Berlin, Heidelberg,
will evaporate. Bacteria can then use oxygen from the pp. 1506–1514. DOI 10.1007/978-3-642-29670-3_10.
air and the biodegradation is relatively fast. Hunt, J.M. 1996. Petroleum Geochemistry and Geology. Free-
man and Co., New York, 743 pp.
Biodegradation may occur in relatively shallow Karlsen, D.A., Nedkvitne, T., Larter, S.R. and Bjørlykke, K.
reservoirs (<1,500–2,000 m, <70–80 C). Very close 1993. Hydrocarbon composition of authigenic inclusions –
to the surface and in contact with groundwater flow, Application to elucidation of petroleum reservoir filling his-
the biodegradation may be oxic but for the most part is tory. Geochemica et Cosmochemica Acta 57, 3641–3659.
Løseth, H., Wensaas, L., Gading, M., Duffaut, K. and Springer,
not possible to supply oxygen from the surface and we M. 2011. Can hydrocarbon source rocks be identified on
must assume that the biodegradation is anoxic. As a seismic data? Geology 39, 1167–1170.
result the oil is less valuable and difficult to produce. Makhous, M. and Galushkin, Y. 2005. Basin Analysis and
Injection of steam to heat the oil and reduce its viscos- Modeling of the Burial, Thermal and Maturation Histories
in Sedimentary Basins. Editions Technip, Paris, 379 pp.
ity is the most effective method of increasing
372 K. Bjørlykke
Nadeau, P.H., Bjørkum, P.A. and Walderhaug, O. 2005. Geology Conference, Geological Society, London, pp.
Petroleum system analysis: impact of shale diagenesis 1267–1274.
on reservoir fluid pressure, hydrocarbon migration, and Tissot, B.P. and Welte, D.H. 1984. Petroleum Formation and
biodegradation risks. In: Doré, A.G. and Vining, B.A. Occurrence. Springer, Berlin, 669 pp.
(eds.), Petroleum Geology: North-West Europe and Wapples, D.W. 1980. Time and temperature in petroleum explo-
Global Perspectives—Proceedings of the 6th Petroleum ration. AAPG Bulletin 64, 916–926.
Chapter 15
Petroleum Migration
Knut Bjørlykke
The transport of petroleum from the source rock to the Kerogen is a solid compound of very large
reservoir rocks is called migration which occurs along molecules (polymers) formed from organic matter
permeable carrier beds. It is important to understand and may occur as dispersed particles in the sediments
this process so that the direction of migration and or as laminae in a claystone. Kerogen may be load-
trapping of petroleum can be predicted. Many differ- bearing and capable of transmitting stress before it
ent theories have been proposed in the past but it is generates petroleum. As the kerogen matures, much
now clear that petroleum is mainly transported as a of this solid matter breaks down to generate oil or gas
separated phase and that the process is mainly driven and is thus transformed into fluid phases. If the fluids
by the buoyancy of petroleum relative to water. The are not expelled immediately, this process increases
porewater is normally not moving much in relation to the volume of the fluid phase (porosity) compared to
the rocks. The upwards flow of the porewater is the original volume of the solid phase in the source
constrained by the rate of compaction and it is rock.
moved downwards by the basin subsidence during The ratio between the volume of the fluid phase
sedimentation. The solubility of oil in water is very (porosity) and the solid phase is often referred to as the
low for most compounds. The solubility of gas, partic- void ratio: Vr ¼ φ=ð1 φÞ.
ularly methane, is much higher both in oil and water Not all kerogen is transformed to fluids during
and increases with depth (pressure). There is, how- maturation. There is a residue of solids which is called
ever, very limited flow in sedimentary basins to trans- coke.
port petroleum. It has usually been assumed that there is a volume
Considerable amounts of gas can bubble out of expansion during maturation of kerogen because the
water or oil if the pressure is reduced due to uplift or density of the oil and gas and the remaining solids in
due to pressure reduction in a reservoir during the kerogen may be smaller than the density of the
production. primary kerogen, thus causing a volume expansion.
This expansion may not necessarily be very large in
the case of oil generation. Even if there was no overall
15.1 Primary Migration
volume expansion, the generation of oil would con-
tribute to the overpressure build-up, since the main
The expulsion of petroleum from a source rock into
factor is the change in void ratio when solid kerogen
adjacent rocks is called primary migration while the
is altered to fluid petroleum. To illustrate this point we
migration further along carrier beds are referred to a
may make the analogy of looking at frozen ground
secondary migration.
with lenses of ice formed during the winter. Ice is a
solid that can carry the weight of the overburden, but
when it melts in the spring the ice becomes fluid and a
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway part of the porosity unless it is expelled. Overpressure
e-mail: [email protected] develops and sometimes small mud volcanoes may
374 K. Bjørlykke
form, despite the fact that there is a reduction in however, before the overburden stress (σ v) is reached,
volume from ice to water, rather than an expansion. allowing petroleum to escape through vertical
Source rocks may include thin layers of siltstones fractures (perpendicular to the direction of least
or sandstones that can serve as pathways for the migra- stress). Even if the kerogen is distributed more evenly
tion of the petroleum fluids generated from the kero- in the source rock, the generation of fluid petroleum
gen. If, however, these more permeable layers are increases the volume of the fluid phase. Layers of
absent, the permeability of the shale matrix is in source rock with 10% TOC by weight make up about
most cases low enough for the fluid pressure to build 20% of the volume. If the water content of the source
up where petroleum is generated until fracture pres- rock is 10%, maturation and fluidisation of 50% of the
sure is reached. Should the source rock consist of kerogen would increase the fluid content (porosity) by
kerogen in a fine-grained clay-rich matrix, the flow 100% if expulsion did not occur. The excess fluid must
of oil out of the source rock is resisted both by very therefore be expelled during maturation because a
high capillary pressures and the low permeability. In shale with such high porosity would compact mechan-
such cases oil can not migrate out of source rocks ically and thus squeeze the oil out.
through the matrix. Very thin open fractures allowing The primary migration is then controlled by the rate
the expulsion of petroleum will develop when the fluid of petroleum generation and this process therefore
pressure in the source rock has reached fracture pres- seems fairly unproblematic. Either the source rocks
sure. The fracture pressure is controlled by the hori- have sufficient permeability for the petroleum to
zontal stress (σ h) which is in most cases lower than the migrate out through the rock matrix or hydro-
overburden stress (σ v). fracturing creates sufficient permeability for the pri-
The kerogen is normally not distributed homo- mary expulsion. If the source rock is very lean a
geneously in the source rocks. Organic-rich mud significant fraction of the petroleum could be retained
deposited under reducing conditions tends to be finely by the source rock, but in the case of richer source
laminated due to a lack of bioturbation, and some rocks a relatively high percentage of the oil generated
laminae may consist of almost pure kerogen will be expelled. The actual percentage of petroleum
(Fig. 15.1). expelled from source rocks is not well known, though.
If these layers of kerogen matured into petroleum The petroleum remaining in the source rock can not
which was not expelled, this fluid phase would have usually be produced by drilling wells because of the
had to support the full overburden stress (σ v). The low permeabilities. However, it may contain large
fracture pressure corresponding to (σ h) is exceeded, amounts of gas which will flow.
In recent years there has been a major development
of shale gas production, particularly in the Devonian
Vertical stress
σv and Carboniferous shales of North America such as
the Barnett Shale (Mississippian). Production is
σh enhanced by horizontal drilling and artificial fractur-
ing of the shales. The remaining oil is normally diffi-
cult to produce without mining the shale.
Oil shales are mostly source rocks that have not
been buried deeply enough to become mature and
expel petroleum. If they have been uplifted and
exposed, they can be mined and the kerogen heated
σv Solid kerogen, in ovens to about 500 C to generate the petroleum.
σh / σv = 0.6–0.8 transforming into
petroleum fluids
Fig. 15.1 Schematic illustration of a source rock. The kerogen 15.2 Secondary Migration of Petroleum
often occurs as distinct laminae which are load-bearing prior to
maturation and petroleum generation. The change from solid
kerogen to fluid petroleum therefore creates an overpressure
The flow of petroleum from source rock to reservoir
which may cause fracturing, helping the primary migration of rocks is called secondary migration and must be
petroleum out of the source rock understood in terms of two-phase and in some cases
15 Petroleum Migration 375
three-phase flow. The relative permeability for oil or and displace the oil out of the traps so that oil migrates
gas is then critical. At low oil or gas saturation the up to the surface or into a shallower trap.
hydrocarbons will only occur as small droplets in the Migration of both oil and gas takes place mostly as
water, which will not flow because of the capillary a separate phase and the driving force is the buoyancy
resistance and the fact that the buoyancy effect will be of the hydrocarbon phase in water. The density of the
very weak. petroleum (ρo) varies from 0.5 to 1.0 g/cm3. The water
The migration of a separate hydrocarbon phase is phase may have densities (ρw) of 1.00–1.20 depending
limited by the capillary forces resisting flow through on the salinity. The buoyancy force (F1) is therefore
small pore throats; the relative permeability for hydro-
carbon is then very small so that little migration can F1 ¼ ðρw ρo ÞH
occur. If there is a pressure gradient in the fluid phase
the water will flow past the petroleum droplets, which where H is the height of the continuous petroleum
are held back by capillary forces. Asphalt-rich oil, on column. For a 100 m high and narrow oil column
the other hand, which may be formed from biodegra- with a density of 0.8 g/cm3, the pressure in the water
dation, sticks to the grain surfaces which are clearly phase is 1 MPa and in the oil phase 0.8 MPa. The
oil-wet and water will flow in the pores. pressure difference between the oil phase and the
On a much larger scale, a sedimentary basin will water phase for each 100 m is thus 0.2 MPa.
always have low average oil saturation. Regional fluid It is this pressure difference which overcomes the
pressure gradients in sedimentary basins will only capillary forces resisting the migration of oil and gas.
move water since at this scale it is the only continuous The capillary forces depend on the surface properties
phase. of the grains in relation to the fluids. In water-wet
Once the primary expulsion from the source rock sandstone, water is drawn in (imbibed) to replace oil
has been achieved, the oil and gas phases will flow if the oil saturation is too high, and the excess pressure
upwards driven by buoyancy, along pathways where in the petroleum phase helps to resist that. Most
the petroleum is concentrated. In this way high oil sandstones are water-wet in contact with normal oil
saturation can be attained locally, increasing the rela- and gas but the wetting angle may vary.
tive permeability. Secondary migration requires that a Assuming water-wet conditions the capillary resis-
continuous pathway with high petroleum saturation is tant force is
established; where flow is prevented by high capillary
entry pressure or low permeability, petroleum may be F2 ¼ ð2γ cos ϕÞ=R
trapped in both small- and large-scale dead ends
(micro-traps). where γ is the interfacial tension between petroleum
When oil or gas flows upwards and accumulates in and water, φ is the wetting angle and R is the radius of
traps, it is normally not accompanied by water flow. the pore throats between the pores which the petro-
The trap is in a way a hydrodynamic “dead end”, leum has to pass.
essentially because of the low permeability of the In coarse-grained sandstones with relatively large
cap rock, unless this has fractured. As an oil or gas pores the resistance from the capillary forces is lowest
lag starts to accumulate near the top of the structure, and migration will then follow such layers.
the permeability with respect to water flow is reduced The interfacial tension between gas and water may
further. If permeability to water is nevertheless not vary from 30 to 70 dynes/cm whereas for oil the range
zero, while the migration of oil may be prevented by is 5–35 dynes/cm (Schowalter 1979). The capillary
capillary forces water may in some cases slowly seep force (F2) resisting petroleum migration is therefore
through the cap rock even if the permeability is very higher for gas than oil, but the buoyancy force (F1) is
low (Teige et al. 2005). also higher because of the lower density of gas. The
Reduction in overburden pressure due to uplift and capillary resistance to flow is as we have seen an
erosion may bring large volumes of gas out of solution inverse function of the critical pore radius (R).
in the water or oil phase, to form separate gas Migration depends on the size of the pore throats
accumulations. This gas may then fill the structures along a continuous pathway for oil migration.
376 K. Bjørlykke
The smallest pore throats are therefore critical for the the viscosity (μ). In the case of a single fluid phase the
migration, not the radius of the pores themselves. permeability is a function of the size of the
The capillary forces can be tested experimentally connections between the pores along the flow path-
by measuring the pressure required to displace water way. When two fluid phases like oil and water are
and force oil into sandstone. The pressure measured present, the permeability of one fluid phase is also a
corresponds to oil columns from 0.3 to 3 m function of the relative abundance of the two phases.
(Schowalter 1997). In siltstones the displacement pres- Oil can only flow through the percentage of the fluid
sure is much higher and in mudstones and shales the phase which is filled with oil and in a water-wet rock
displacement pressure corresponds to hundreds of there is a layer of water around each grain reducing the
metres or even kilometres of oil column. cross-section of the pore throats available for oil flow.
It is difficult to control the surface properties of oils The permeability of oil in the presence of water varies
in the laboratory. Experimentally it is easier to use as a function of the oil saturation, which is the per-
mercury as the displacing fluid and mercury centage of oil in the pore space, compared to the total
capillary-pressure curves (mercury injection curves) fluid volume. Similarly the permeability with respect
are the standard way to analyse the effective pore to water flow depends on the cross-sections of the
size distribution in reservoir rocks. Since the surface water-filled parts of the pores between grains. The
tension is known the critical pore throat radius (R) can relative permeability is the permeability of a fluid
be calculated. phase in the presence of another fluid phase compared
In sandstones the displacement pressure increases to the permeability in the same rock when only one
with increasing cementation and the resultant reduc- fluid is present. If the percentage of oil in the pores is
tion in porosity and permeability. Even relatively well less than 20–30% the relative permeability of oil is so
cemented sandstones are not normally barriers to oil low that it will move very slowly or not at all com-
migration, though. Well-cemented carbonate layers pared to water. In some cases pore networks may be
and also thin clay layers and stylolites may serve filled with water, oil and gas and then we have to
effectively as seals for further migration. The main consider 3-phase flow. Shales have normally low per-
problem is usually not the migration through meability for water and even lower relative permeabil-
sandstones, but from one sandstone body to the next ity for oil. Shales are barriers to oil migration both
through shales. Faults in sandstones may have a clay because the capillary forces resist the flow and because
smear, reducing the permeability and increasing the of the low permeability.
capillary entry pressure.
During progressive burial, faults and fractures do
not tend to be open conduits for fluid flow, but during
uplift rocks are more brittle (overconsolidated) and 15.3 Migration in Sandstones
fractures and faults may be more open.
Shales and mudstones buried to more than about Migration takes place along pathways with the lowest
3 km have relatively low porosity and permeability capillary entry pressures. These are in most cases
and high capillary entry pressure. The vertical exten- sandstones or open fractures. Mostly the migration
sion of oil columns is limited by the depth at which oil follows the upper parts of sandstone beds due to the
is generated (3–5 km) and the reservoir depth. buoyancy of oil in water. Very little of the migration
Extractions of fluids from shales in near contact with occurs vertically through sandstones.
oil and condensate in reservoirs do not show evidence The nature of the transition between the permeable
of oil saturation (Olstad et al. 1997), suggesting that sandstones and the overlying shales is therefore very
oil has not displaced water in the shales. Gas important. In a sandstone which is coarsening-
molecules, and particularly methane, are very much upwards the maximum permeability is near the top
smaller than those composing oils and can probably of the sequence just below the shale (Fig. 15.2). This
diffuse through shales, though at relatively slow rates. is the case in sandstones deposited in shallow marine
Both oil and water flow follows the Darcy Law environments (shoreface, beach and delta-front
ðF ¼ rP k=μÞ. The flux (F) is a function of the successions). When they are not well-cemented,
permeability (k), the potentiometric gradient (P) and these uppermost beds possess high porosity and
Trap
Vertical carrier
Trap
Trap Seal
Reservoir rock
Seco
ndar
ym
igra
tion
La Source rock
ter
al c Primary migration
arr
ier
Petroleum migration
Fig. 15.2 Petroleum migration in a sedimentary basin. From primary migration out of the source rock and secondary migration up
to a trap. From Steven Larter (unpublished)
permeability, and the oil saturation and relative per- sandstone geometry. Abandoned channels may
meability will therefore also be high. become separate traps and synsedimentary faults (i.e.
In fining-upward units such as fluvial sandstones growth faults) may trap migrating oil and gas. On a
and turbidites the permeability and capillary entry smaller scale there may be micro-traps defined by
pressure decrease upward. Due to its buoyancy, oil sedimentary structures or small-scale faulting, and
will flow along the top of sandstone or siltstone beds faulting will increase the volume of oil and gas lost
against the overlying mudstones or shales. Here the during migration. Foresets in cross-bedded sandstones
permeability as well as the oil saturation will be lower may contain laminae and even stylolites which may be
in siltstones and poorly sorted sandstones and the flow barriers for fluid flow, particularly oil because of the
rate and oil saturation is reduced compared to cleaner capillary resistance.
sands. Sedimentary structures such as cross-bedding
may contain laminae that are barriers for oil because
they have very high capillary entry pressure. This may
produce micro-traps for the migrating oil and gas, 15.4 Migration of Oil Through Shales
increasing the loss during migration.
The direction of migration is up the maximum There are rarely continuous connecting sandstone bod-
slope of sandstone beds in regular planar sand sheets ies from the source rock to the reservoir, and petro-
which may be folded, and this may be modelled based leum must therefore migrate through shales.
on 3D seismic. In elongate sandstone bodies like flu- Fluid flow through shales may occur in different
vial channels and certain submarine bars, however, the ways:
direction of migration is also controlled by the 1. Matrix-controlled (intergranular flow)
378 K. Bjørlykke
2. Flow in fractures produced by hydro-fracturing due Perm

to overpressure (mostly microfractures) Cap. press
3. Flow in tectonically induced macrofractures
In the case of matrix-controlled flow the capillary A
entry pressure and the permeability are a function of
the pore size distribution and diagenetic alterations. In
shales the typical pore sizes may be a few 100 Å or
less. In the Gulf Coast the pore diameter may be less
than 25 Å at 3–4 km depth and less than 10 Å at
4–5 km burial (Leonard 1993). The size of the largest
pores and their connections will however determine
B
the permeability. If the largest connecting pore size is
below 50 Å this is approaching the size of the
asphaltenes in crude oil (Tissot and Welte 1978, p.
170). A sieving effect should therefore be observed
Flow of oil through a fining-upwards sandstone (A)
relative to the size of the organic molecules if the pore and a coarsening-upwards sandstone (B).
throats are below this value. Migration of oil through
low permeability shales probably only happens along In layer A which may be a fluvial sandstone
or turbidite, more oil is lost during migration that in
fractures and not through the shale matrix. Fractures in layer B which may be a shallow marine sandstone.
shales may be formed tectonically during uplift and
extension when the shales are brittle. Relatively large Fig. 15.3 Migration along sandstones is more efficient along
coarsening-upwards units than in fining-upwards sequences
fractures may occur because the horizontal stresses
trying to close the fractures are generally small. This
fracturing of the cap rock will occur in approximately
is a common cause of oil and gas leakage from
the same position (Fig. 15.3).
reservoirs. However, nearly all onshore reservoirs
We must distinguish clearly between migration of
and many offshore reservoirs have experienced some
oil along a fault plane and across it. If there is sand on
uplift from their maximum burial depth and in a rock
both sides of the fault plane it is difficult to predict if
mechanical sense are therefore overconsolidated and
the fault will be a barrier for oil migration. Sealing
will tend to be brittle. During basin subsidence, tec-
faults are often critical for the formation of traps in
tonic shear produces fractures during ductile deforma-
rotated fault blocks. Clay smears from adjacent shales
tion and these fractures are then no more permeable
can serve as a barrier at shallow depth, while at greater
than the matrix.
depth intensive quartz cementation may reduce the
Oil fields are often highly overpressured, with
permeability and the capillary entry pressure for oil.
many of them leaking petroleum at the top of the
See further discussion on rock mechanics in sedimen-
structure which is often close to the fracture pressure.
tary basins (Chap. 11).
This means that it is the horizontal stress and the
In subsiding basins, fault planes are at shallow to
tensional strength of the rocks that control the pres-
moderate burial depth and are subjected to shear defor-
sure. When the pressure is close to fracture pressure it
mation which produces clay smearing, so that the
implies that faults are no longer the conduits for fluid
permeability along the fault is normally lower than
flow because the rock matrix would then fracture and
through the rock matrix. After tectonic displacement,
let the oil through. If faults were a zone of weakness
faults may also be subject to cementation. During
the fluid pressure should have remained below fracture
tectonic uplift, however, faults may be extensional
pressure.
and much more permeable.
In the case of traps formed by rotated fault blocks
Migration due to hydrofracturing of shales. If the
the top of the structure will usually coincide with a
fluid pressure exceeds the fracture pressure the rock
fault. Even if seismic evidence indicates gas leakage
will hydrofracture. The pressure required to fracture
from the top of the reservoir, this does not necessarily
the rock can be measured in a well by using a leak-off
mean that the leakage is along the fault, because
test (LOT test, see Chap. 11). However, the LOT
creates microfractures which develop at a lower pres- than the fluid pressure gradient. Accumulation of
sure than that required to form proper fracturing. petroleum in a trap capped by shale which does not
These microfractures should theoretically develop fracture, requires that the pressure be reduced by lat-
when the pressure exceeds the sum of the least eral flow of water to maintain pressure below fracture
horizontal stress and the tensional strength of the pressure (Fig. 15.5).
rock. Since the pressure required for large-scale
fractures to form (fracture pressure) is higher than
the LOT test value, the tensional strength which 15.5 Migration Through Tectonically
allows microfractures to form is lower than when Fractured Rocks
forming proper hydrofracturing.
In the case of microfractures (LOT tests) these A clear distinction must be made between different
probably deform the rock in a different way, so that types of tectonic fracturing and hydrofracturing,
the rock can heal once the pressure is released. although these processes can occur together. It is
Fractures developed by hydrofracturing during important also to distinguish between faults formed
leakage of oil are likely to be vertical because they during subsidence and those formed during uplift.
develop parallel to the direction of maximum stress Some of the fractures observed may have been
which is normally vertical in subsiding basins with formed during soft sediment deformation just after
little external tectonic stress. As a fracture opens, the deposition or be associated with growth faulting, and
permeability along it is increased. This reduces the thus unrelated to tectonic stress.
pressure gradient along the fault plane to less than During subsidence and uplift the direction of stress
the fracture gradient. The top of the fracture may may have changed several times, producing fractures
therefore be above fracture pressure, while the lowest with different orientations. We must therefore not
part is below fracture pressure and subject to effective assume that all the fractures were open or closed at
stress trying to close it. The fractures produced by the same time. Fractures produced during late stages
hydrofracturing must therefore propagate upwards of uplift are most likely to be open, but they have very
and are of limited vertical extent. They develop first little relevance to the conditions during oil migration.
in the least permeable parts of the shale (cap rock) However, rocks that are now outcropping have usually
which may only contain water because of low capil- had a long history of deformation.
lary entry pressure. The pressure in these very small Numerous studies have been carried out on the
water-saturated pores should not be influenced by the pattern of fracturing in outcrops on land to serve as
pressure in the petroleum phase in the sandstones analogues for subsurface fracture patterns which can
(Fig. 15.4). not readily be mapped from seismic or cores. This has
The excess pressure in the petroleum phase com- then served as a basis for sophisticated fluid modelling
pared to the water pressure, is held by the capillary and its consequence for oil migration, assuming in
forces and does not influence the pressure causing some cases that the fractures are open and more per-
onset of hydrofracturing (Bjørkum et. al. 1998). How- meable than the matrix, in others that they are less
ever once the first fracturing has occurred the petro- permeable than the matrix.
leum will be the continuous phase along the fracture,
and it is the pressure in the hydrocarbon phase which
causes leakage when the horizontal stress is exceeded. 15.6 Trapping of Petroleum Below a Cap
In the laboratory, water has been shown to flow Rock
though a cap rock while oil has been retained by the
capillary pressures (Teige et al. 2005). A cap rock traps petroleum if the flow into the trap
If we consider migration in two dimensions it is exceeds the flow out of the trap. A trap may leak
clear that it is not only the source rock that will petroleum through the matrix of the seal or through
fracture. All the overlying shales that could serve as fractures produced by overpressure or tectonically. If
potential cap rocks could reach fracture pressure and there is leakage through the matrix of a shale it is
leak petroleum if there are no lateral drainage paths. because the capillary forces are not high enough to
This is because the fracture pressure gradient is steeper resist the buoyancy of the petroleum. As we have seen,
380 K. Bjørlykke
Pressure
a Migration
pathways
Shale
Leakage
tone A
Sands
Fra
Hydr
Shale
ctu
re
osta
p
tone B
Ge
res
Fault
Sands
tic pr
os
s
ure
tat
essur
ic
p
Migration pathway Shale
res
e gra
su
for petroleum
re
d.
?
b Petroleum Pressure
leakage on
Matrix flow fractures Fracture
water flux pressure
Oil
pressure
Water
Water flow
pressure
Block B
Block A
Compaction-
driven flux
Compaction- Hydrostatic
driven flux pressure
Blocks A and B are connected but separated from the rest

of the basement:
Oil leakage (Fracture flow, F1) = compaction-driven flux (Flux A + Flux B) - matrix flow (F2)
At shallow depth the matrix flow may be relatively high and compaction mostly mechanical.
Fracture pressure can then only be reached by fluid supply from deeper
sediments driven by compaction.
Fig. 15.4 (a) Fracturing will always occur at the top of the communication the lower block (A) will not reach fracture
structure. The fluid pressure will intersect the fracture gradient pressure and fracture. The higher block (B) then serves as a
at the top because the pressure gradients inside sandstones are safety valve for block A. If they represent separate pressure cells
nearly hydrostatic. (b) If two rotated fault blocks are in pressure they could both fracture and leak
Permeability
Pressure
Shale Fracturing – water flow

(cap
rock,
seal)
Pressure in the petroleum phase
equal to capillary entry pressure
Sand-
stone
Fracture
Water pressure
Hydro- pressure
static
Fig. 15.5 Fracturing in seals is likely to occur where the pressure is higher in the oil-saturated zone. Water may therefore
permeability for water is lowest. The fracture pressure may leak from the structure when petroleum is retained. Based on
therefore be controlled by the water-saturated shale even if the Bjørkum et al. (1998)
the capillary forces are a function of the wetting angle continuously build up in the structure and the fracture
and the pore size, and the buoyancy forces are a pressure will be reached, causing the structure to leak.
function of the density of the petroleum and the thick- A cap rock thus serves as a valve which slowly allows
ness of the petroleum column. A shale may then be water or petroleum through, maintaining a constant
able to hold a limited thickness of oil or gas column in pressure equal to the fracture pressure. The flow is
the reservoir (Schowalter 1979). There is nevertheless not likely to be very episodic because a high flow
considerable evidence that fine-grained shales buried rate would lower the pressure and close the fractures.
to 3–4 km have more than enough capillary force to
hold several hundred metres of oil column. In the case
of gas the density is lower, producing higher buoy- 15.7 The Rate of Oil Migration
ancy, but the interfacial tension is also higher. Gas
may also to a much larger extent be dissolved in the Attempts have been made to calculate the rate of
water phase and diffuse through the cap rock. petroleum migration from the source rock to the reser-
It may seem natural to assume that the extra pres- voir. Many assumptions then have to be made about
sure due to the buoyancy of the petroleum may be the capillary resistance and the permeability of the
critical with respect to the fracturing because the frac- rocks through which the oil and gas migrate. The oil
ture pressure then is reached earlier. However, the saturation and the cross-section of the oil-saturated
extra pressure of the petroleum column is balanced rocks would also play an important role. There are
by the capillary forces and does not change the water good reasons, though, to think that the oil and gas
pressure critical for reaching fracture pressure migration normally is not rate-limiting for the accu-
(Bjørkum et al. 1998). Also, the fracture pressure mulation of a trap.
may be reached in the cap rock some distance above Modelling of oil migration through sand suggests
the top of the petroleum column (Fig. 15.5). Once that migration may occur at very high velocities
hydrofracturing starts and petroleum becomes the con- exceeding 100 km/million years and that the oil col-
tinuous phase in the fractures, the pressure in the umn may be thin (Sylta et al. 1997). This modelling is,
petroleum phase will resist the horizontal stress and however, based on migration through sandstones with
will determine when the fractures open and close. relatively high permeability (100 mD). The oil-
For practical purposes this may not play an impor- saturated pathway may then be only a few centimetres
tant role because if there is no adequate horizontal thick. A 100 million m3 (600 million bbl) oil field
leakage of water or petroleum, the pressure will could be filled in 3 million years with oil flowing at
382 K. Bjørlykke
just 1 cm3/s, equivalent to a tiny trickle out of a garden which defines the maximum thickness of the oil and
hose. Thicker oil columns will probably build up gas column, is represented by the vertical distance
where the migration has to pass through less perme- from the valley up to the top of the mountain.
able sandstones and siltstones which require higher In an overpressured part of a sedimentary basin the
capillary pressures. highest structures within one pressure compartment
The rate of migration is limited by the petroleum are likely to leak. Deeper structures in the same com-
generation in the source rocks in the drainage area. partment may have a high chance of retaining oil
The oil and gas can not flow faster than it is generated. because they will not reach fracture pressure and
If the migration is slower than the generation, a trap leak (Fig. 15.4).
has formed somewhere along the migration pathway. The evaluation of a trap depends very much on the
It the rate of oil generation increases, both the cross- contours on this map, which are a function of the
section of oil-saturated rocks along the pathway will accuracy of the depth conversion of the seismic data.
increase and a thicker oil column produce higher oil Faults offsetting the carrier bed and the shale seal
saturation and relative permeability. These are posi- may obstruct migration but represent at the same time
tive feedbacks that will make it possible for the migra- a possible trapping mechanism.
tion rate to increase in order to adjust to the rate of Much of what has been said about oil migration
generation. Low permeability shales along the path- holds true also for gas migration. Gas molecules, par-
way will either fracture due to the high pressure or ticularly methane, are very small and may diffuse
form a cap rock. A trap which does not have pressure through shales and mudrocks in a way which is not
communication with any higher structures will easily possible for oil. The migration of gas above structures
reach fracture pressure and leak. This will not apply to is now easily seen on seismic sections, showing that
a lower trap if there is continuous permeable sandstone many gas fields do leak but are filled at a rate which
up to a higher trap; the lower trap will not then reach compensates for the loss.
fracture pressure and leak (Fig. 15.4).
Migration of petroleum is very much influenced by
the composition of the petroleum, particularly the 15.9 Loss of Petroleum During
density, viscosity and the gas/oil ratio. This depends Migration
on the primary composition of the organic matter at
the time of deposition (organic facies). Starting with If we know the volume of source rocks in the drainage
different organofacies the composition of the oil and area and how much each volume can yield, the total
gas can be modelled as a function of different PVT volume of petroleum generated can be calculated.
conditions (di Primeo and Horsefield 2006). This may Assuming a certain expulsion efficiency, the maxi-
explain changes in petroleum composition during mum amount of petroleum which could migrate into
migration, trapping and remigration. a structure can be calculated. This has been done for
different sub-basins in the North Sea (Geoff 1983).
The percentage of oil and gas which was generated
15.8 Regional Migration but is not found in the major oil fields, could have been
lost through leakage from reservoirs or have
In a basin with well-defined shales serving as cap accumulated in smaller traps that have not been dis-
rocks the sandstones represent the migration routes. covered. Some oil and gas is probably also left behind
In the North Sea basin, Cretaceous shales are along the migration route.
important cap rocks and the base of these Cretaceous The properties of the sandstones which serve as
shales can be mapped out in great detail using 3D migration routes depend on the depositional environ-
seismics. The depth contours make the traps stand ment and subsequent diagenesis. However, shallow
out like mountains or hills on a topographic map. marine sandstones are normally coarsening-upwards
Using that analogy, the migration of oil follows the and have the cleanest and most permeable sand at the
highest ridges and fills the structures down to the top of the sequence. There will then often be a sharp
lowest contour connecting it to the next mountain, contact to the overlying shale and migration will be
which is the spill point. The closure of the structure, very efficient with relatively little loss. In fining-up
sequences, commonly found in fluvial sandstones and The probability of detecting petroleum on the land
turbidites, the buoyancy will force oil and gas up into surface or on the seabed is highest in areas of active
the finer-grained part of the sandstones where much of migration. Active seeps with gas may also be detected
it may be trapped in small pores. by side-scan sonar (Hovland et al. 2002). Good
It has been claimed that organic acids generated in examples of active seeps are found in many basins in
the source rocks could dissolve feldspar and mica in California, the Gulf of Mexico, North Sea and
the reservoir sandstones. Many of these acids are Indonesia (Abrams 1996).
rather water-soluble and will diffuse into the water Passive seeps where there is at present no active
phase in the source rock, which often contains carbon- seepage are more difficult to detect including by
ate, and during migration would be neutralised before refined geochemical methods.
reaching the reservoirs. In overpressured reservoirs the migration is likely
The sandstones adjacent to or interbedded with the to be vertical due to hydrofracturing. However once
main source rock in the North Sea basin (Kimmeridge the petroleum is out of the overpressured compartment
shale) show little or no evidence of dissolution of further migration will tend to occur laterally, often
feldspar and precipitation of authigenic kaolinite. along gently dipping sandstones in a shaly sequence.
This suggests that the generation of organic acids in This is well illustrated from Haltenbanken, offshore
the source rocks has an insignificant effect on the Norway, where petroleum is mostly leaking vertically
development of secondary porosity (Bjørlykke and from overpressured Middle to Lower Jurassic
Aagaard 1992). reservoirs (Karlsen et al. 1995) and migrating up-dip
Compared to the total buffering system of, firstly, through Palaeocene and Eocene sands up to the sea-
the silicate mineral system and, secondly, the carbon- floor near the coast (Bugge et al. 1984, Thrasher et al.
ate system, the addition of relatively small amounts of 1996).
comparatively weak acids (organic acids) will not Where there is oil migration into reservoirs that had
change the pH of the porewater significantly. been buried to depths where the temperature has
exceeded 80 C the reservoir is pasteurised so there is
little biodegradation and oil can than flow to shallow
15.10 Petroleum Seepage and depth. If the reservoir has not been heated to this
Exploration temperature there is bacteria present in the reservoir
and oil may be biodegraded to several hundred metres
Oil exploration started by drilling where there already depth. (See Chaps. 1 and 21).
was evidence of seepage. This was also the case in Oil
Creek in Pennsylvania, and in California.
When the reservoir is buried more deeply the 15.11 Summary
amount of leakage is highly variable and the migration
pathway to the surface may be very complex. The idea Primary migration of petroleum from a source rock to
that the presence of petroleum should be detected the reservoir rock may occur through the rock matrix if
directly has always seemed attractive and geochemical it contains sediments with sufficiently high permeabil-
surveys to detect oil have been carried out on a large ity and low capillary entry pressure. If petroleum can
scale both on land and on the seafloor. not flow through rock matrix the generation of petro-
A leakage of petroleum may be a good sign because leum will contribute to the build-up of the pore pres-
it shows that some petroleum is present. On the other sure until fracturing occurs.
hand the fact leakage occurs indicates that large Secondary migration has an upward component and
fractions of the petroleum generated may have been is driven by the buoyancy of the petroleum phases in
lost to the surface. Many large oilfields have no visible the porewater, and is resisted by the capillary forces.
seeps directly above the structures (Thrasher et al. The rate of migration is a function of the rate of
1996). generation and primary expulsion. The rate of migra-
A study of the Southeast Asian Basin showed that tion is therefore very low and the migration pathway
leakage was concentrated over tectonic structures such may have a very small cross-section, but with a high
as faults and diapirs. degree of petroleum saturation. A high petroleum
384 K. Bjørlykke
saturation increase the relative permeability making di Primeo, R. and Horsfield, B. 2006 From petroleum-type
the migration more efficient. organofacies to hydrocarbon phase prediction. AAPG
Bulletin 90, 1031–1058.
Migration through sandstones occurs mostly along Geoff, J.C. 1983. Hydrocarbon generation and migration from
the top of sloping beds. The depositional environments Jurassic source rocks in the E. Shetland Basin and Viking
determine the sandstone geometry, and diagenesis Graben of the northern North Sea. Journal of Geological
determines the internal flow properties. Coarsening- Society 140, 445–474.
Hantchel, T. and Kauerauf, A.I. 2009. Fundamentals of Basin
upwards sequences have a higher migration efficiency and Petroleum System Modeling. Springer, New York, NY,
than fining-upwards sequences or poorly sorted sand. 476 pp.
Migration through shales requires that they fracture Hovland, M., Garder, J.V. and Judd, A.G. 2002. The signifi-
due to overpressure or tectonic deformation to over- cance of pockmarks to understanding fluid flow processes
and geohazards. Geofluids 2, 127–136.
come the capillary resistance. A cap rock must have a Karlsen, D. A., Nyland, B., Flood, B., Ohm, S. E., and Backer-
capillary entry pressure which is higher than the pres- Owe, K. 1995. Petroleum geochemistry of the Haltenbanken,
sure in the petroleum phase. Norwegian continental shelf. In: Cubitt, J.M. (ed.), The
During progressive burial to about 4 km depth Geochemistry of Reservoirs. Geological Society Special
Publication 86, pp. 203–256.
(120 C) the shales often become almost impermeable Leonard, R.C. 1993. Distribution of subsurface pressure in the
and fracture pressure is likely to be reached unless the Norwegian Central Graben and applications for exploration.
fluids can be drained laterally. Fracturing and leakage In: Parker, J.R. (ed.), Petroleum Geology of Northwest
of petroleum will occur at the top of the highest struc- Europe: Proceedings of the 4th Conference. The Geological
Society, pp. 1295–1303.
ture in the pressure compartment where the fracture Lindgren, H. 1987. Molecular sieveing and primary migration in
pressure is lowest (least overburden). On a global basis the Upper Jurassic and Cambrian claystones source rock. In:
oil reservoirs are mostly at temperatures less than Brooks, J. and Glennie, K. (eds.), Petroleum Geology of
120 C. This is because at that depth the permeability Northwest Europe. Graham and Trotman, London,
pp. 357–364.
of shales serving as cap rocks (seals) may become Nadeau, P.H., Bjørkum, P.A. and Walderhaug, O. 2005. Petro-
very low. Without lateral drainage, fracturing and leum system analysis: Impact of shale diagenesis on reser-
leakage may occur. Quartz cementation will in most voir fluid pressure, hydrocarbon migration, and
cases cause reduced porosity at depths where the tem- biodegradation risks. In: Doré, A.G. and Vining, B.A.
(eds.), Petroleum Geology: North-West Europe and Global
perature is higher than 120 C. Perspectives – Proceedings of the 6th Petroleum Geology
During subsidence, fault zones are not likely to Conference. The Geological Society, pp. 1267–1274.
provide open conduits for oil migration unless fracture Olstad, R., Bjørlykke, K. and Karlsen, D.K. 1997. Pore water
pressure is reached. flow and petroleum migration in the Smørbukk Field area,
offshore Norway. In: Møller-Pedersen, P. and Koestler, A.G.
(eds.), Hydrocarbon Seals – Importance for Exploration and
Production. Norwegian Petroleum Society, Special Publica-
tion 7, 201–216.
Further Reading Schowalter, T.T. 1979. Mechanism of secondary hydrocarbon
migration and entrapment. AAPG Bulletin 63, 723–760.
Sylta, Ø., Pedersen, J.I. and Hamborg, M. 1998. On the vertical
Abrams, M.A. 1996. Distribution of subsurface hydrocarbon and lateral distribution of hydrocarbon migration velocities
seepage in near-surface marine sediments. In: Schumacher, during secondary migration. In: Parnell, J. (ed.), Dating and
D. and Abrams, M.A. (eds.), Hydrocarbon Migration and its Duration of Fluid Flow and Fluid-Rock Interaction. Geologi-
Near-Surface Expression. AAPG Memoir 66, pp. 1–14. cal Society Special Publication 144, pp. 221–232.
Bjørkum, P.A., Walderhaug, O. and Nadeau, P. 1998. Physical Teige, G.M.G., Hermanrud, C., Thomas, W.L.H., Wilson, O.B.
constraints on hydrocarbon leakage and trapping revisited. and Nordgård Bolås, H.M. 2005. Capillary resistance and
Petroleum Geoscience 4, 237–239. trapping of hydrocarbons; A laboratory experiment. Petro-
Bjørlykke, K. and Aagaard, P. 1992. Clay Minerals in North Sea leum Geoscience 11, 15–129.
Sandstones. In: Houseknecht, D.W. and Pittman, E.D. (eds.), Thrasher, J., Fleet, A.J., Hay, S.J., Hovland, M. and
Origin, Diagenesis and Petrophysics of Clay Minerals in Düppenbecker, S. 1996. Understanding geology as the key
Sandstones. SEPM Special Publication 47, pp. 65–80. to using seepage in exploration: The spectrum of seepage
Bugge, T., Knarud, R. and Mørk, A. 1984. Bed rock geology on styles. In: Schumacher, D. and Abrams, M.A. (eds.), Hydro-
the mid-Norwegian continental shelf. In: Spencer, A.M. carbon Migration and Its Near-Surface Expression. AAPG
et al. (eds.), Petroleum Geology of the North European Memoir 66, 223–241.
Margin: Norwegian Petroleum Society. Graham and Thyberg, B., Jahren, J., Winje, T., Bjørlykke, K. and Faleide, J.I.
Trotman, London, pp. 549–555. February 2009. From mud to shale: Rock stiffening by
micro-quartz cementation. First Break 27, 27–33.
Chapter 16
Well Logging: Principles, Applications and Uncertainties
Nazmul Haque Mondol
16.1 Introduction into the borehole on the end of an insulated electrical

cable after the drilling tool is pulled up (Fig. 16.1). The
Well logging is a means of recording the physical, cable itself is used as the depth measuring device, so
acoustic and electrical properties of the rocks that properties measured by the tools can be related to
penetrated by a well. It is carried out by service particular depths in the borehole.
companies, which work under contract for the oil Well logs are usually recorded while the logging
companies. Logging has the advantage that it measures device is being winched upward through the well. The
in situ rock properties which cannot be measured in a measurements from the instruments housed in the log-
laboratory from either core samples or cuttings. ging tool are recorded digitally at intervals of between 3
Logging started with simple electric logs measuring and 15 cm and the data is processed near the well on
the electrical conductivity of rocks, but it is now an land, or on the platform in the case of offshore. Record-
advanced and sophisticated method used routinely in ing the well log involves a number of steps, beginning
different phases of hydrocarbon exploration, field with sensing and pre-processing the measurement in the
development and monitoring. It gives us a continuous logging tool itself, transmission of this information to
downhole record that provides a detailed subsurface the surface over several kilometers of wireline, further
picture of both gradual and abrupt changes in physical processing in the logging truck computer, data storage
properties from one bed to the next. By contrast, only on disc or magnetic tape, and finally display of the data
selected parts of the reservoir rocks are cored, and on film or paper. A measured log shows many
samples of cuttings from the rest of the well give no variations from top to bottom (log display in
more than a general idea of the lithology. Fig. 16.1), and each wiggle has significance. Most
None of the logs actually measure the parameters logs are dependent on direct contact with the rock via
that are of most interest to us, such as how much oil or the walls of the well, and have to be run after successive
gas is in the subsurface, or how much is being pro- intervals of the drilling, before each stage of the steel
duced. Such important knowledge can only be derived, casing is installed in the well.
from the measured properties such as gamma radiation,
density, velocity and resistivity, using a number of
assumptions which, if true, will give reasonable 16.2 Well Logs: The Necessity
estimates of hydrocarbons. To perform a logging oper-
ation on wireline, the measuring instrument is lowered Geological sampling during drilling (cuttings)
provides a very imprecise record of the formations
encountered. Entire formation samples can be brought
N.H. Mondol (*) to the surface by mechanical coring, but this is both
Department of Geosciences, University of Oslo, Oslo, Norway slow and very expensive. In the narrowest sense, well
Norwegian Geotechnical Institute (NGI), Oslo, Norway logging supplements the analysis of cores, side-wall
e-mail: [email protected] samples and cuttings. Logs are used for a variety of
386 N.H. Mondol
Fig. 16.1 Example of a logging tool (left, Courtesy of borehole with logging tools at the end of the cable (middle).
Schlumberger). The wireline logging operation showing logging Example of a recorded gamma ray log display (right)
truck, logging cable strung into the rig, then lowered into the
purposes depending on the nature of the data recorded. Table 16.1 How is logging viewed by others? (Adapted from
Ellis and Singer 2008)
Correlation from well to well is one of the oldest and
probably the most common use of logs. Correlation is The Geologist:
What depths are the formation tops?
usually based on the shapes of the recorded curves Is the environment suitable for accumulation of hydrocarbon?
versus depth. It allows the subsurface geologist to Is there evidence of hydrocarbons in this well?
map formation depths, lateral distributions and What type of hydrocarbon?
thicknesses, and then to identify conditions that could Are hydrocarbons present in commercial quantities?
The Geophysicist:
trap hydrocarbons. Identification of the lithology of the
Are the tops where you predicted?
rock sequence is another important use. After acquiring Are the potential zones porous as you assumed from seismic data?
experience in an area it is possible for a log analyst to What does a synthetic seismic section show?
make an educated guess as to lithology by looking at The Drilling Engineer:
the log from a new well. A set of logs run in a well will What is the hole volume for cementing?
Where can you get a good packer seat for testing?
usually mean different things to different professionals Where is the best place to set a Whipstock?
(Table 16.1). Currently, over fifty different types of The Reservoir Engineer:
these logging tools exist in order to meet various infor- How thick is the pay zone?
mation needs and functions. Some of them are passive How homogeneous is the formation?
measurement devices, others exert some influence on What is the volume of hydrocarbon per cubic metre?
Will the well pay-out?
the formation being traversed.
The Production Engineer:
Pickett (1963) indicated some of the applications Where should the production well be placed?
for well logs in petroleum engineering (Table 16.2). What kind of production rate can be expected?
The applications fall into three categories: identifi- Will there be any water production?
Is the potential pay zone hydraulically isolated?
cation, estimation and production. Identification
16 Well Logging: Principles, Applications and Uncertainties 387
Table 16.2 Uses of well logging in petroleum engineering pressure in the formation pore space to prevent
(adapted from Pickett 1963) disasters such as blowouts. The drilling mud helps to
Logging applications for petroleum engineering transport the rock fragments (cuttings) from the drill
Identification: bit up to the surface (Fig. 16.2), where the cuttings are
Rock type analysed for indications of hydrocarbon. Another
Identification of geological environments
Location of fluid contacts (e.g. gas/oil, gas/water and oil/water
important function of drilling fluids is rock
contacts) stabilisation. Special additives ensure that the drilling
Fracture detection fluid is not absorbed by the rock formation in the well
Estimation: and that the pores of the rock formation are not
Estimate of hydrocarbon in place clogged.
Estimate of recoverable hydrocarbon
Reservoir pressure Because the pressure in the drilling mud must
Porosity/pore-size distribution exceed the formation pore pressure, the mud begins
Production: to enter permeable zones in the formation but is nor-
Water flood feasibility mally rapidly stopped by the build-up of a mudcake
Reservoir quality mapping
which lines the borehole wall (Fig. 16.2). The solid
Interzone fluid communication probability
Reservoir fluid movement monitoring particles are concentrated there while the fluid
penetrates the formation, first creating a flushed zone
where nearly all the primary pore fluids are replaced
concerns subsurface mapping or correlation. Estima- by the fluids from the drilling mud. This part of the
tion is the more quantitative aspect of well logging, in drilling mud is called the mud filtrate. Beyond this,
which physical parameters such as water saturation or there will be a zone where the primary pore fluids are
pressure are needed with some precision. The final partially replaced by drilling fluids, called the transi-
category consists of well logging measurements tion zone. The virgin formation fluids occupy the
which are used to monitor changes in a reservoir uninvaded/undisturbed zone further into the formation
during its production phase. (Fig. 16.2). The depth of invasion is related to the
permeability of the rock. Drilling mud extends furthest
into porous sandstones but flushing and invasion will
16.3 Drilling Muds and Borehole be rather limited in low permeability formations such
Environments as shales and tight sandstones. As the well is drilled
deeper, further invasion occurs slowly through the
During drilling a water-based, or sometimes oil-based, mudcake, either dynamically, while mud is being
slurry containing clays and other natural materials, circulated, or statically when the mud is stationary.
called drilling mud (drilling fluid), is pumped down In addition, the movement of the drill-string can dis-
the drill-string. Drilling muds are added to the lodge some of the mudcake, allowing renewed
wellbore to facilitate the drilling process by invasion.
suspending cuttings, controlling pressure, stabilising The replacement of oil by the water-based mud
exposed rock, providing buoyancy, and cooling and filtrate is by pressure-driven displacement. In water-
lubricating the drilling bit. Nowadays, drilling deeper, bearing formations the mud filtrate replaces all of the
longer and more challenging wells is made possible formation water close to the borehole (Fig. 16.3a) but
due to more efficient and effective drilling fluids. High this decreases with depth of invasion. In oil-bearing
density materials (e.g. barite, hematite) are added to formations the mud filtrate replaces all the formation
the drilling mud to increase its density and thereby its water and most of the oil close to the borehole wall,
pressure on the walls of the well. The drilling fluid again decreasing with distance into the formation
pumped down the borehole must ensure that the (Fig 16.3b). Oil-based mud filtrates replace the fluids
hydrostatic pressure in the wellbore exceeds the fluid in the invaded zone by pressure-driven displacement
388 N.H. Mondol
Fig. 16.2 The schematic diagram illustrates an idealised version zones (a). Overpressured mud indicates that it is invading porous
of what happens when fluids from the borehole invade the porous and permeable sandstones with the formation of a mudcake (a and
and permeable sandstones and low permeability shales. The mud b). During drilling, mud is pumped down the drill-string, forcing
circulation causes borehole washout in the low permeability shale the cuttings up to the surface with the return flow (c)
alone. In water-bearing formations the oil-based mud interpretation. To get good readings of the true subsur-
filtrate does not replace all the formation water even face properties of the rock, the tool has to either (1)
close to the borehole wall (Fig. 16.3c), while in oil- measure accurately through the borehole mud, mud
bearing formations the oil-based mud filtrate only cake, flushed zone and transition zone, or (2) make
replaces the oil in the formation, leaving the formation readings closer to the tool (i.e. in the flushed zone) that
water in place (Fig. 16.3d). can be reliably corrected to represent the values in the
It is very important to know the type of drilling mud uninvaded zone. Wireline companies provide correc-
used, as this will determine the way in which it tion graphs for their various tools. However, the accu-
influences what is recorded on the logs. Salinity racy of the correction diminishes as the diameter of the
variations in mud may alter rock properties. Oil-based borehole, the thickness of mud cake and the depth of
mud causes its own complications for log invasion, increase.
Fig. 16.3 Schematic indication of the distribution of pore fluids in the flushed, transition and uninvaded zones for water-based
muds (top), and oil-based muds (bottom), in water- and oil-bearing formations
16.4 Measuring Techniques MWD and LWD are types of well logging where
the measurement tools are incorporated into the drill-
Well logs measure in open hole, cased hole and paral- string and provide real-time information to help with
lel with drilling (Fig. 16.4). The conventional wireline steering the drill. These technologies were originally
logs that are run in a hole which has just been drilled, developed to partially or completely replace the open-
and before it is cased, are called open-hole logs. Open- and cased-hole logging. MWD and LWD offer the
hole logs require that the drill-string must be removed same measurements as wireline logging but in poor
from the well before the logging tools can be lowered quality, lower resolution and less coverage. Information
into the hole. Open hole logs are used to determine may be continuously transmitted to the surface by mud
most petrophysical properties of the rocks. Logs run pulses, a common method of data transmission where a
after the well is cased are called cased-hole logs. Most valve is operated to restrict the flow of the drilling mud
cased-hole logs are used to assess the integrity of the according to the digital information to be transmitted.
well completion and fluid flow into the well though This creates pressure fluctuations representing the
some cased-hole logs can also be used to determine information. The pressure fluctuations propagate within
petrophysical properties. Most logs can now be the drilling fluid towards the surface where they are
recorded while drilling (i.e. do not need the drill-string received from pressure sensors. At the surface, the
be removed) and so measurements are available con- received pressure signals are processed by computers
tinuously as drilling proceeds. These logs are called to reconstruct the information. Alternatively, the data is
MWD (Measurements While Drilling) and LWD (Log- recorded downhole and retrieved later when the instru-
ging While Drilling). ment is brought back to the surface.
390 N.H. Mondol
Fig. 16.4 Open hole (left), cased hole (middle) and LWD/ well is cased are called cased-hole logs. LWD and MWD are the
MWD (right) logging techniques. Logs run in an open hole general terms use to describe the logging techniques for
which has just been drilled but not cased. Logs run after the gathering downhole data while drilling a well
Table 16.3 lists the main differences between LWD Table 16.3 Main differences between conventional wireline
and conventional wireline logging. LWD tools are logging and LWD
large instrumented drill collars, part of the bottom Wireline LWD
hole assembly. The sensors are arranged so that they • Small, light and delicate • Big, heavy and tough
• Since the 1930s • Since the 1970s
can make relatively unimpeded measurements of the • High data speed • Slow telemetry
formation. MWD is a type of LWD where tools are • Easy communication • Limited control
encompassed in a single module in the steering tool of • Good borehole contact • Minimum borehole contact
the drill-string at the end of the drilling apparatus • Powered through cable • Batteries and mud turbine
• Takes more time • Saves time
(bottom hole assembly). Providing wellbore position, • Problem at high deviation • Can log in any direction
drillbit information and directional data, as well as • Susceptible to hole • More capable in tough
real-time drilling information, MWD uses gyroscopes, conditions conditions
magnetometers and accelerometers to determine bore- reservoir. Note that the greatest variety of logging
hole inclination and azimuth during the actual drilling. tools is run in the area that is expected to be of greatest
Some information is logged on the rig, such as a interest, i.e. the reservoir.
mud log which may record up to five or ten properties Most tools have a shallow depth of investigation,
of the drilling fluid, or a drilling log which records the resulting in the measurement of the formation in the
rate of penetration and other functions of the drilling flushed or transition zones (Fig. 16.5). Most radiation
process. The stratigraphic log or sample description tools (e.g. Gamma Ray, Neutron and Density logs)
log, records the site geologists’ identification of the have an investigation depth of less than 0.5 m. Electri-
rock samples retrieved from the drilling mud, together cal tools (e.g. Microlog, Laterolog, Induction log)
with qualitative or interpretive data concerning evi- come in various versions with a wide range of investi-
dence of the fluid content of the rock, and thus is one gation depths, from the micro-tool, which measures
of the primary sources of rock and fluid descriptions only the mud cake (a few centimetres), to the deep
for the well. penetration tools (induction log, up to 5 m). The
depth of investigation often depends upon the den-
sity/porosity of the formation. Because of the wide
16.5 Logging Tools variety of subsurface geological formations, many dif-
ferent logging tools are needed to give the best possible
Logging tools have been developed over the years to combination of measurements for the rock type
measure radioactivity (e.g. Gamma Ray, Neutron and anticipated. The following are the principles, most
Density logs), electrical properties (e.g. SP, Induction common applications and uncertainties of the most
and Resistivity logs), acoustic properties (Sonic log), important types of log:
nuclear magnetic resonance (NMR log), pressure, and
many other properties of the rocks and their contained 16.5.1 Temperature Log
fluids. Other logs, which measure properties of the
wellbore itself, are Caliper, Temperature, Image and 16.5.1.1 Generalities and Basic Principles
Dipmeter logs. Despite the availability of this rather The temperature log is a tool for measuring the
large number of devices, each providing complemen- borehole temperature. Readings from a number of
tary information, the final answers derived are mainly thermometers attached to different tool combinations
(1) the location of hydrocarbon-bearing formations, and run at different times are analysed (Fig. 16.6) to
(2) an estimate of their producibility, and (3) an assess- give the temperature at the bottom of the borehole
ment of the quantity of hydrocarbon in place in the (bottom hole temperature, BHT). Temperature in the
Fig. 16.5 Depth of investigation of common logging tools

392 N.H. Mondol
Fig. 16.6 Plot of measured

and estimated bottom hole
temperatures (BHT) and
geothermal/temperature
gradients with depth for wells
is a sedimentary basin
subsurface increases with depth. The rate at which it formula. This is often referred to as formation temper-
does so is called the geothermal gradient (gG), or ature (Tf).
simply geotherm. The first step involved in determin-
ing temperature at a particular depth is to determine T f ¼ Tms þ ½gG ð D=100Þ (16.2)
the geotherm of the region.
Low thermal conductivity rocks, such as shale,
act as a thermal insulator and have a large temper- 16.5.1.2 Uses of Temperature Log
ature gradient across them, while high thermal con- Temperature logs have many applications, with the
ductivity rocks, such as salt, permit the conduction most common being to identify zones producing or
of heat efficiently and have a small temperature accumulating fluid, to evaluate a cement or hydraulic
gradient across them. Geothermal gradient is com- fracture treatment, and to locate lost circulation zones
monly expressed in degrees celsius per 100 m and casing leaks. Geothermal gradients in sedimentary
( C/100 m). If the geothermal gradient of an area basins are extremely important for the calculations
is not known, then it can be determined by the necessary for predicting kerogen maturation, detection
following formula: of fluid movement, subsurface pressures, and for the
general modelling of basin subsidence. Since temper-
gG ¼ ð BHT Tms=TDÞ 100 (16.1) ature takes time to dissipate, a temperature log tends to
reflect the behaviour of a well over a longer time
where BHT is the bottom hole temperature, TD is the period than other measurements.
total depth and Tms is the mean surface temperature.
Typical geotherms for reservoirs are about 20–35 C/ 16.5.1.3 Uncertainties of Temperature Log
km, although significantly higher values (up to 85 C/ Temperature measurements made during drilling
km) can be found in tectonically active areas, and (MWD and LWD) consistently underestimate the for-
lower ones (0.05 C/km) in stable continental mation temperature because drilling mud is being
platforms. Once the geothermal gradient has been circulated. Temperature measurements made on
established, it is possible to determine the tempera- wireline logs sometime after the drilling fluid circula-
ture for a particular depth (D) using the following tion has stopped also underestimate the formation
temperature, but less so than MWD and LWD as the represent the caliper value minus the drill bit diameter
formation is now in the process of reheating the and thus record cavities where the well has caved in,
borehole. Drilling mud takes several days or even and also indicate the competence of the rocks cut
weeks before its temperature approaches that of during drilling (Fig. 16.7). The scale is generally
the formation water, and there is therefore considerable given in inches, which is standard for measuring bit
uncertainty involved in such temperature calculations. sizes. Table 16.4 describes the main influences on
caliper values. A hole with the same diameter as the
bit-size is called on gauge.
16.5.2 Caliper Log
16.5.2.2 Uses of Caliper Log
16.5.2.1 Generalities and Basic Principles The common uses of the caliper log are as follows:
The caliper log measures borehole diameter and shape. • Indication of borehole quality for the assessment of
Nowadays, caliper logs also record the three- the quality of other logs whose data quality is
dimensional shape of the borehole walls. It has two, degraded by boreholes that are out of gauge.
four, or more extendable arms (Fig. 16.7) which can • Contributory information for lithological assess-
move in and out as the tool is drawn up the borehole. ment, and as indicator of good permeability and
These movements are converted into electrical signals porosity zones (reservoir rocks) due to develop-
by a potentiometer. Normally caliper log readings ment of mudcake.
Fig. 16.7 Typical caliper log responses to various lithologies (right). Mutli-arm and multi-sensor caliper tool (upper-left).
Ellipticity of borehole can help to identify direction of principal stresses (lower-left). (Modified from Rider 2004)
394 N.H. Mondol
Table 16.4 Factors influencing caliper responses

Hole diameter Cause Possible lithologies
On gauge Well-consolidated formations Massive sandstones
Non-permeable formations Calcareous shales
Igneous rocks
Metamorphic rocks
Larger than Formations soluble in drilling mud Salt formation drilled with freshwater
bit size Weak formations and caving Unconsolidated sands, gravels, brittle shales
Smaller than Formations that swell and flow into borehole Swelling shales
bit size Development of mudcake with porous and Porous, permeable sandstones
permeable formations
• Estimation of mudcake thickness, hmc ¼ (dbit – dh)/

2, where h stands for the hole, in inches.
• Measurement of borehole volume, Vh ¼ (dh2/
2) + 1.2%, in litres per metre.
• Measurement of required cement volume, Vcement
¼ 0.5 (dh2 – d2casing) + 1%, in litres per metre.
• Selection of consolidated formations for wireline
pressure tests, recovery of fluid samples, and for
packer seating for well testing purposes.
16.5.2.3 Uncertainties of Caliper Log

Known challenges with caliper logging include bore-
hole spiralling. The position of the drill bit may pre-
cess as it drills, leading to spiraling shapes in the
wellbore wall, as if the hole had been drilled by a
screw. If the arms of the caliper log follow the grooves
of the spiral, it will report too high an average diame-
ter. Moving in and out of the grooves, the caliper will
give erratic or periodically varying readings. In most
cases, the borehole’s circumference will not be a per-
fect circle and therefore a caliper tool with several
Fig. 16.8 The SP tool arrangement and typical log responses
arms is required to obtain a true understanding of the for permeable sandstone and impermeable shales
size and shape of the borehole. The borehole can
change to an oval shape after drilling, which can
cause the caliper log to overestimate the size of the difference in the concentrations of electrolytes in the
borehole. liquid phases. There are three requirements for the
existence of an SP current: (1) a conductive borehole
fluid (i.e. a water-based mud), (2) a porous and perme-
16.5.3 Self-Potential/Spontaneous able bed sandwiched between low porosity and imper-
Potential (SP) Log meable formations, and (3) a difference in salinity
between the borehole, in most cases the mud filtrate,
16.5.3.1 Generalities and Basic Principles and the formation fluid. This log has no absolute
The self-potential/spontaneous potential log (SP) was scale – it is the relative changes in the SP log that are
the first wireline logging tool used in hydrocarbon important. SP is measured in mV (millivolts). A rela-
exploration. It is very simple, requiring only an elec- tive scale of 10 mV/small division is usually used.
trode in the borehole and a reference electrode at the Value range is typically approximately 50 mV
surface (Fig. 16.8). A current is created by the about the 0 mV for shale baseline, while sandstones
produce more negative (to the left) (Fig. 16.8) or

positive values based on salinity variations between
drilling mud and formation water.
SP measures the naturally occurring electrical poten-
tial (voltage) produced by the interaction of formation
water, conductive drilling fluid and shale. There are
three sources of the currents; two are electrochemical
potential (liquid junction potential and membrane
potential), and one electrokinetic potential. The liquid
junction potential and the membrane potential are the
main components of SP deflection. The Na+ and Cl
ions have different nobilities at the junction of the
invaded and virgin zones. The ion movement across
this boundary generates a current flow and hence a
potential. The greater the difference between the salin-
ity of the solutions (mud and formation water), the
greater the potential. In low porosity, low permeability
formations, the mudcake builds slowly and the electro-
kinetic potential becomes predominant. Electrokinetic
potential is generated by the flow of mud filtrate
through a porous permeable bed. It depends upon the
resistivity of the mud filtrate and will only become
important if there are high differential pressures across
the formation. The effects are normally negligible in
permeable formations because the mudcake builds
quickly and halts any further invasion.
16.5.3.2 Uses of SP Log

The SP log has several applications: (1) indication of
the shaliness of a formation, (2) correlation of strati-
graphic sequences, (3) detection of permeable beds,
(4) determination of formation water resistivity (Rw),
and (5) detection of hydrocarbons by the suppression Fig. 16.9 Example of the shale baseline and the SSP defined
on an SP log. The shale base line is the maximum positive
of the SP curve. The two most common applications
deflection (in this example) and occurs against shale. The SSP
are described below: is a maximum negative deflection and occurs against clean,
porous and permeable water-bearing sandstone
Shale Volume Estimation
The volume of shale is used in the evaluation of shaly
sand reservoirs and as a mapping parameter for both SSP (static spontaneous potential) is the SP log
sandstone and carbonate facies analysis. The follow- reading in a thick clean sand zone and SPshale is the
ing relationships are used to estimate shale volume shale baseline. This assumes a linear mixing relation-
(Vsh) from the SP log (Fig. 16.9): ship between the SP log and shale volume and has no
theoretical basis; it probably overestimates the shale
V sh ¼ ð1 PSP=SSPÞ (16.3) volume.
V sh ¼ ½ðPSP SSPÞ=ðSPshale SSPÞ (16.4) Identification of Facies and Depositional

Environments
where PSP (pseudostatic spontaneous potential) is the The SP log can be used to follow facies changes. SP
log reading in a thick homogeneous shaly sand zone, deflections often respond to changes in depositional
396 N.H. Mondol
mostly replaced by the GR log, which has a higher

resolution and is more reliable.
16.5.3.3 Uncertainties of SP Log

The SP log has several limitations such as (1) borehole
mud must be conductive: when using an oil-based mud
readings cannot be obtained from SP logs, (2) a
sequence of permeable and impermeable zones must
exist, (3) little or no deflection occurs if resistivity of
mud/mud filtrate is close/equal to resistivity of forma-
tion water, and (4) SP response will not fully develop
in front of thin beds. The log has a low vertical resolu-
tion and is rarely useful in offshore environments.
Although it can be used for correlation, it is best not
to rely solely upon it.
16.5.4 Gamma Ray and Spectral Gamma

Ray Logs
16.5.4.1 Generalities and Basic Principles

The gamma ray (GR) log measures the total natural
gamma radiation whereas the spectral gamma (SGR)
ray log measures the natural gamma radiation
emanating from a formation split into contributions
from each of the major radio-isotopic sources
(Fig. 16.11). Gamma ray log is combinable with all
other logging tools, and is almost always used as part
of every logging combination run because of its ability
to match the depths of data from each run. An advan-
tage of the gamma ray log over some other types of
well logs is that it works through the steel and cement
walls of cased boreholes and is insensitive to drilling
fluids. The unit of gamma ray log is API (American
Petroleum Institute).
The gamma radiation originates from potassium
(K40) and the isotopes of uranium-radium and thorium
Fig. 16.10 Facies identification using the SP log. A typical
fining-upwards, channel sandstone giving a bell-shaped SP curve series. Potassium is by far the most abundant of the
three elements but its contribution to the overall radio-
activity in relation to its weight is small. Each of the
environment. Characteristic SP shapes are produced in three sources emits gamma rays spontaneously. They
channels, bars, and other depositional sequences emit photons with no mass and no charge but great
where sorting, grain size or cementation changes energy. One of the characteristics of gamma rays is
with depth (Fig. 16.10). These shapes are called that when they pass through any material their energy
bells, cylinders or funnels. A typical fining-upwards, is progressively absorbed. The effect is known as
channel sandstone gives a bell-shaped SP curve. The Compton scattering, and is due to the collision
SP log is sometimes useful for stratigraphic correla- between gamma rays and electrons that produces a
tion, but is rarely used alone. Nowadays it has been degradation of energy.
Fig. 16.11 Some typical responses of gamma ray and spectral gamma ray logs against different lithologies. The spectral gamma
ray responses show individual contributions of thorium, potassium and uranium to the overall radioactivity
16.5.4.2 Uses of Gamma and Spectral gamma tool to distinguish between shales and non-shales.
Gamma Ray Logs Care must be taken not to overgeneralise these rules. For
The gamma ray is a simple but very useful log. High example a clean sandstone may contain feldspars (arkosic
vertical resolution makes it particularly well suited for sandstones), micas (micaceous sandstones) or both
depth matching and fine-scale correlation. The other (greywackes), or glauconite, or heavy minerals, any of
most common uses of gamma and spectral gamma ray which will give the sandstone higher gamma ray values
logs are (1) lithology discrimination, (2) shale volume than would be expected from a pure quartzitic sandstone.
estimation, and (3) analysis of facies and depositional Black shales (e.g. hydrocarbon source rocks) in
environment. particular, with their substantial organic content, pro-
duce marked responses on the gamma ray log because
Lithology Discrimination they normally have a higher uranium content than
A prime application of the gamma ray log is for discrimi- other shales. The Kimmeridge (Upper Jurassic) shale
nation of different lithologies (Fig. 16.12). While it can- in the North Sea contains 2–10 ppm uranium and is
not uniquely define any lithology, the information it often referred to as the “hot shale”. Normally shales
provides is invaluable when combined with information contain <1 ppm uranium, but 10–12 ppm thorium,
from other logs. Shales, organic-rich shales and volcanic which represents ~50% of the total radioactivity.
ash show the highest gamma ray values, while halite, Limestones have very low concentrations of U, Th
anhydrite, coal, clean sandstones, limestone and dolomite and K, and give very low gamma-ray responses. In
have the lowest values. This difference in radioactivity evaporite sequences, however, gamma logs are very
between shales and sandstones/carbonates allows the sensitive indicators of potassium salts.
398 N.H. Mondol
Fig. 16.12 Gamma ray log responses for common lithologies. (Modified from Rider 2004)
Shale Volume Estimation (Vsh) volume is often used to help discriminate between
In most reservoirs the lithologies are quite simple, reservoir and non-reservoir rocks. Shale volume is
being cycles of sandstones and shales or carbonates calculated in the following way: First the gamma ray
and shales. Once the main lithologies have been index (IGR) is calculated from the gamma ray log data
identified, the gamma ray log values can be used to using the following relationship
calculate the shaliness or shale volume (Vsh) of the
rock. This is important, as a threshold value of shale I GR ¼ GRlog GRmin =ðGRmax GRmin Þ (16.5)
Fig. 16.13 The chart for corrected Vsh estimation
where GRlog is the gamma ray reading at the depth of Facies and Depositional Environment Analysis
interest, GRmin is the minimum gamma ray reading Gamma ray logs help to identify thin beds and so are
(usually the mean minimum through a clean sandstone widely used for lithological correlation and depth
or carbonate formation), GRmax is the maximum matching between different logging runs. As men-
gamma ray reading (usually the mean maximum tioned earlier, the gamma ray log is often used to
through a shale or clay formation). Many measure the shaliness of a formation. In reality the
petrophysicists assume that Vsh ¼ IGR though there shaliness often does not change suddenly, but gradu-
is no scientific basis for assuming a linear relation of ally with depth. Such gradual changes are indicative of
Vsh and IGR. However, to be correct the value of IGR the litho-facies and the depositional environment of
should be entered into the chart shown as Fig. 16.13, the rock, and are associated with changes in grain size
from which the corresponding value of Vsh may be and sorting that are controlled by facies and deposi-
read. tional environments. Figure 16.14 analyses the shape
It should be noted that the calculation of shale of gamma ray log responses for various depositional
volume is a ‘black art’ as much depends upon the environments. All possible combinations of these
experience of the geoscientist in defining the GRmin shapes may be encountered.
(sand line) and the GRmax (shale line) values The cylinder shapes represent uniform deposition
(Fig. 16.13), noting that the sand line and/or shale and are interpreted as aeolian dune, tidal sand, fluvial
line may be at one gamma ray value in one part of and turbidite channel and proximal deepsea fan
the well and at another gamma ray value at another deposits. The bell shapes represent the fining-upward
level. Once the shale volume has been calculated, a sequences and are interpreted as tidal sand, alluvial
threshold shale volume may be defined which will sand, fluvial channel, point bar, lacustrine, delta, tur-
divide the well into a number of reservoir and non- bidity channel and proximal deepsea fan deposits. The
reservoir zones. funnel shapes represent coarsening-upward sequences
400 N.H. Mondol
Fig. 16.14 The gamma ray

log responses and their
application to interpret
depositional facies and
depositional environments
that are interpretated as barrier bar, beach sand and with potassium. The radiation from KCl-based muds
crevasse splay, distributary mouth bar and distal contributes to the gamma ray and increases signifi-
deepsea fan deposits. cantly the total gamma ray count (Fig. 16.15).
For standard GR logs the value measured is calcu-
16.5.4.3 Uncertainties of Gamma and lated from thorium in ppm, uranium in ppm and potas-
Spectral Gamma Ray Logs sium in percent. Due to the mass of uranium
The gamma ray log usually runs centered in the bore- concentration in the calculation, anomalous concen-
hole. If the borehole suffers from caving, the gamma trations of uranium can cause clean sand reservoirs to
ray log can be badly affected and underestimated appear shaly. The spectral gamma ray log is used to
(Fig. 16.15). The measured underestimation should provide an individual reading for each element,
be corrected from the caliper log. The mud density enabling concentration anomalies to be identified and
affects the gamma ray as higher density muds (e.g. interpreted. Sometimes non-shales also have elevated
barite) attenuate the gamma ray and will give an levels of gamma radiation. Sandstone can contain
anomalously low gamma ray reading. KCl-based uranium, potassium feldspar, clay filling or rock
muds have a natural gamma radioactivity associated fragments that cause it to have higher-than-usual
Fig. 16.15 Effect of caving

and KCl drilling mud on the
gamma ray log readings
gamma readings. Coal and dolomite may contain rays are absorbed by rock, and particularly by the water
absorbed uranium. Evaporite deposits may include in the formation. This is due to collisions with atomic
potassium minerals such as carnallite. When this is nuclei, and the absorption of the neutron radiation is
the case, spectral gamma ray logging can be done to primarily a function of hydrogen atom concentrations
identify these anomalies. (hydrogen index). Since most of the hydrogen in rocks
is present as water, neutron logs provide an expression
of the water content and thereby the porosity of a rock.
16.5.5 Porosity Logs The common density tools are GNT (gamma-ray neu-
tron tool), SNT (sidewall neutron tool), CNL (compen-
16.5.5.1 Generalities and Basic Principles sated neutron log) and DNL (dual neutron log). Neutron
Evaluating porosity is an important petrophysical task logs are displayed as neutron porosity and common
in formation evaluation. The porosity of a formation units are p.u. (porosity unit), decimal or %.
can be estimated either from neutron, density, sonic or In shales and sandstones with high clay content the
NMR logging. None of these logs measure porosity neutron logs record higher porosity because hydrogen
directly. The density, neutron and NMR logs are is also present in the clay minerals which are part of
nuclear measurements whereas the sonic log uses the solid phase. This is called the shale effect and is
acoustic measurements. When using a single porosity most pronounced in shales with high content of smec-
log, the true porosity is calculated from interpolation tite and kaolinite compared to those with mostly illite
between the values for the matrix mineral and the pore and chlorite. Limestones however give very reliable
fluid. A combination of porosity logs gives more accu- porosity values because carbonate minerals contain
rate estimates of porosity and valuable indications for little hydrogen. Calculation of porosity based on neu-
lithology. The basic principles of four porosity logs tron logs (neutron porosity) results in too low porosity
are given below: values when the pores are filled with gas because they
contain less hydrogen per volume compared to water
Neutron Log and oil. This is called the gas effect. Gas is less dense
A neutron log is obtained using a neutron source, which and has fewer hydrogen atoms per unit volume than
sends radiation into the rocks (Fig. 16.16). The neutron water and oil, and therefore has a lower hydrogen
402 N.H. Mondol
Fig. 16.16 The neutron tool and log response against mixtures of shales and sandstones. The gamma ray log responses for the
sequence are also presented
index. Neutron logs can therefore be used to detect gas A formation with a high bulk density has a high
and distinguish it from oil. number of electrons. It attenuates the gamma rays
significantly, and hence a low gamma ray count rate
is recorded at the sensors. A formation with a low bulk
Density Log
density has a low number of electrons. It attenuates the
The density tools are induced radiation tools. They
gamma rays less than a high density formation, and
bombard the formation with gamma radiation and mea-
hence a higher gamma ray count rate is recorded at the
sure how much radiation returns to the sensors. The tool
sensors.
consists of a radioactive source (usually caesium-137 or
cobalt-60) that emits gamma rays of medium energy (in
the range 0.2–2 MeV). The tool has short- and long- Sonic Log
spaced detectors (Fig. 16.17). The common density Sonic logs measure how sound travels through rocks,
tools are FDC (formation density compensated), LDT and in particular provide information about porosity.
(litho-density tool), CDS (compensated density) and With this method a probe sends out acoustic pulses
PDS (photoelectric density). Returning gamma rays which travel through the rock surrounding the well to
are measured at two different energy levels: (1) higher the other end of the logging tool, and the velocity of
energy gamma rays (Compton scattering) determine sound in the rock is recorded (Fig. 16.18). This also
bulk density and therefore porosity (e.g. FDC, CDS indicates whether a liquid or gas phase occupies the
tools) and (2) low energy gamma rays (due to photo- pore spaces. The velocity is also dependent on how the
electric effect) are used to determine formation lithol- rock is compacted, i.e. the pore distribution and the
ogy (e.g. LDT and PDS). The low energy gamma rays nature of the cements. The common sonic tools are
show little dependence on porosity and fluid and so are BHC (borehole compensated sonic), LSS (long spaced
very useful for lithology identification. The density log sonic), array sonic/full waveform sonic and DSI
unit is g/cm3 or kg/m3. The unit of the photoelectric log (dipole shear imager). The velocity measured by the
is b/e (barns per electron, equal to 10 28 m2). sonic log is however not a direct function of the
Fig. 16.17 The formation density tool and log response against shales and sandstones and their mixtures. The gamma ray log
responses for the sequence are also presented
Fig. 16.18 The sonic tool and log responses against a sandstone-shale sequence. The gamma ray log responses for the sequence are
also presented
404 N.H. Mondol
porosity. The log record is usually presented as the Nuclear Magnetic Resonance (NMR)
Δt which is the inverse of velocity (i.e. slowness). This NMR logging exploits the large magnetic moment of
is called interval transit time, and is presented on the hydrogen, an element which is abundant in rocks in the
logs on a scale of 40–140 μs/ft (μs ¼ 10 6 s) or μs/m. form of water. The NMR signal amplitude is propor-
100 μs/ft corresponds to 10,000 ft/s, or 3,048 m/s. The tional to the quantity of hydrogen nuclei present in a
interval transit time (t) is the reciprocal of the sonic formation and can be calibrated to give a value for
transit velocity (v). porosity that is free from lithology effects (Fig. 16.19).
Since the velocity of sound in water, which here A petrophysicist can also analyse the rate of decay of
means porewater, is considerably lower than it is in the NMR signal amplitude to obtain information on
minerals and rocks, the measured velocity will be the permeability of the formation. Using NMR logs
more or less inversely proportional to the rock porosity. one can distinguish between free water in the pore
In sandstones a small amount of cement (i.e. quartz space and H2O and OH groups in minerals. Also
cement) may produce a grain framework with high bound water on mineral surfaces has a different
stiffness and velocity despite it retaining a relatively NMR signature (T2 distribution). The T2 distribution
high porosity. In mudstones and shales the porosity and has several petrophysical applications:
velocity vary greatly as a function of the clay mineral- • T2 distribution mimics pore-size distribution in
ogy and the presence of carbonate or quartz cement. water-saturated rock.
Fig. 16.19 A comparison of NMR total porosity (MPHI) and distributions (NMR response), Gamma Ray, Caliper and SP
Neutron (Nϕ) and Density (Dϕ) porosities. Also shows T2 logs. (Adapted from Coates et al. 1999)
• The area under the distribution curve equals NMR

porosity.
• Permeability is estimated from the logarithmic-
mean T2 and NMR porosity.
• Empirically derived cutoffs separate the T2 distri-
bution into areas equal to free-fluid porosity and
irreducible water porosity.
16.5.5.2 Uses of Porosity Logs

The main use of porosity logs is to provide porosity
information. In combination with sonic and density logs,
the important applications are compaction trend and uplift
estimation, overpressure detection, seismic data calibra-
tion and synthetic seismogram generation. Furthermore,
the combination of density and neutron logs give the best
ways of identifying lithologies and detecting gas-bearing
zones. Numerous uses of porosity logs are given below:
Porosity Estimation Fig. 16.20 Correction chart for obtaining porosity values for
When using a single porosity log, lithology must be lithologies other than limestone (re-drawn, courtesy of
specified through the choice of a matrix value, to avoid Schlumberger)
ambiguity in calculating porosity. Combinations of poro-
sity logs may unravel complex matrix (e.g. quartz, calcite, Table 16.5 Grain (matrix) densities of some common rock-
dolomite) and fluid (e.g. brine, oil, gas) compositions, forming minerals
providing a more accurate porosity determination. Grain density Grain density
3 3
Mineral (g/cm ) Mineral (g/cm )
Porosity from Neutron Log Quartz 2.65 Halitea 2.16
The neutron tool is sensitive to the amount of hydrogen Calcite 2.71 Gypsuma 2.30
ions in the formation and to a less extent also other Dolomite 2.87 Anhydritea 2.96
Biotite 2.90 Carnalitea 1.61
elements. It is assumed that the contribution to the
Chlorite 2.80 Sylvitea 1.99
measurement by elements other than hydrogen is negli-
Illite 2.66 Polyhalitea 2.78
gible, and that the contribution to the measurement from
Kaolinite 2.594 Glauconite 2.30
hydrogen comes entirely from the fluids fully occupying
Muscovite 2.83 Kainite 2.13
the pore space. However, in real rocks, elements other a
Evaporites
than hydrogen that exist in the rock matrix do contribute
to the measurement, and hydrogen is also present in the
matrix itself (e.g. bound water in shales). The problem is the solid phase (ρmatrix) and is used to compute density
partially overcome by calibrating the tool against lime- porosity (ϕD) using the following relation:
stone. Pure limestone saturated with freshwater is used
for calibration because it contains no elements which
ϕD ¼ ρmatrix ρlog = ρmatrix ρfluid (16.6)
contribute significantly to the neutron measurements
other than hydrogen. Therefore, the porosities that are
read by the tool are accurate in limestones containing The value of the matrix density taken depends upon
freshwater. The porosities that are read by the tool in the lithology of the target zone. The most common
other lithologies or with other fluids need to be corrected reservoir rocks are sandstone (ρmatrix ¼ 2.65 g/cm3),
by a chart given in Fig. 16.20. limestone (ρmatrix ¼ 2.71 g/cm3) and dolomite
(ρmatrix ¼ 2.87 g/cm3). Clay minerals have varied
Porosity from Density Log grain densities (Table 16.5). The input of fluid density
The bulk density log (ρlog) measures the combined (ρfluid) is usually that of formation brine (1.025 g/cm3).
effects of the fluid density (ρfluid) and the density of The porosity may also be in error if the fluid density
406 N.H. Mondol
Fig. 16.21 Calibration of the density porosity (in %) calculated from formation density log (in g/cm3) against core porosities (in %)
measured in the laboratory
is misjudged. If oil or gas is present in the formation, where ϕN and ϕD are neutron and density porosities. It
porosities can be overestimated by the density has been suggested that the square-root equation is
log. Remember that the tool measures the invaded preferable as a means of suppressing the effects of
zone, so the relevant fluid is the mud filtrate in any residual gas in the flushed zone.
most circumstances. If available, the fluid densities
should be corrected to borehole temperature conditions. Porosity from Sonic Logs
The most accurate porosity determinations are The velocity of elastic waves through a given lithology
obtained from laboratory measurements on cores and is a function of porosity. Wyllie proposed a simple
so core data is used to provide accurate matrix mixing equation to describe this behaviour and called
densities for particular intervals. If there is a database it the time average equation. It can be written in terms
of core porosities for a given well, it is often advanta- of velocity
geous to plot the core porosity against the density log
derived porosity (Fig. 16.21). Δtlog Δtmatrix
ϕs ¼ (16.8)
Δtf Δtmatrix
Porosity from Neutron and Density Combination
The combination of density and neutron logs is also where Δtlog is the transit time in the formation of
used to determine porosity that is largely free of lithol- interest, Δtf is that through 100% of the pore fluid
ogy effects. Both porosity logs record apparent and Δtmatrix is that through 100% of the rock matrix
porosities that are only true when the zone lithologies (solid mineral grains), ϕs is the sonic porosity. A list of
match with predicted lithologies in porosity input values to these equations for common lithologies
calculations. By averaging the apparent neutron and and fluids is given in Table 16.6.
density porosities of a zone, effects of lithologies can
Lithology Discrimination
cancel out. The true porosity therefore is estimated
Lithology identification using the neutron-density log
either by taking an average of the two log readings
combination is a hugely important technique, and
or by applying the equation:
together with the litho-density log forms the best
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi downhole lithology identification technique available.
ϕ2N þ ϕ2D The measured neutron porosity in shales is consider-
ϕ¼ (16.7)
2 ably higher than the measured neutron porosity in
Table 16.6 Values for Dt and P-wave velocities for use in density, and its presence will give rise to
Wyllie’s equation underestimations in porosity. On the neutron-density
Material Δt (μs/ft) V (ft/s) V (m/s) combination, gas stands out very distinctly, giving a
Compact sandstone 55.6–51.3 18,000–19,500 5,490–5,950 large negative separation (Fig. 16.23).
Limestone 47.6–43.5 21,000–23,000 6,400–7,010
Dolomite 43.5–38.5 23,000–26,000 7,010–7,920 Detection of Overpressure
Anhydrite 50.0 20,000 6,096 Sonic logs can be used to detect overpressured zones
Halite 66.7 15,000 4,572
in a well. An increase in pore pressures causes a clear
Shale 170–60 5,880–16,660 1,790–5,805
drop in sonic velocity (i.e. an increase in sonic travel
Bituminous coal 140–100 7,140–10,000 2,180–3,050
time). Fluid overpressure works against any compac-
Lignite 180–140 5,560–7,140 1,690–2,180
tion trend caused by the overburden pressure. Hence, it
Casing 57.1 17,500 5,334
Water: 200,000 ppm, 180.5 5,540 1,690
is likely that overpressured zones will retain a greater
15 psi porosity than normally pressured zones. If a normal
Water: 150,000 ppm, 186.0 5,380 1,640 compaction or no compaction is observed in a shale
15 psi over some depth interval, and below it the bulk density
Water: 100,000 ppm, 192.3 5,200 1,580 begins to decrease (or the derived porosity begins to
15 psi increase) without change in lithology, it is likely that
Oil 238 4,200 1,280
one has entered a zone of overpressured fluids. In this
Methane, 15 psi 626 1,600 490
zone the overpressured fluids keep the porosity open,
stopping any compaction trend and reversing it
(Fig. 16.24).
carbonate or sandstone. This unrealistically high
porosity is a good indicator of shale, and can become Uplift Estimation
diagnostic when combined with the gamma ray log. As sediments become compacted, the velocity of elas-
One can see from the neutron-density crossplot that tic waves through them increases. If one plots the
there will be a separation of the density and the neu- interval transit time on a logarithmic scale against
tron logs for sandstone and dolomite, but no separation depth on a linear scale, a straight line relationship
for limestone (Fig. 16.22). The sandstone separation is emerges. This is called compaction trend. Compaction
called negative and the dolomite separation, which is trends are constructed for single lithologies, compar-
in the other direction and slightly larger, is called ing the same stratigraphic interval at different depths.
positive. If shale is present in the formation, the neu- It is possible to estimate the amount of erosion at
tron log reads much higher porosities and gives a large unconformities or the amount of uplift from these
positive separation. If the shale volume decreases due trends. This is because compaction is generally
to the intermixture of sandstone, the large positive accompanied by diagenetic changes which do not
separation decreases, crosses over and becomes even- alter after uplift. Hence the compaction of a sediment
tually the small negative separation associated with represents its deepest burial. Figure 16.25 compares
pure sandstone. Thus, a sequence of clearly defined the compaction trend for the same lithology in the
sand and shale formations shows switching between same stratigraphic interval in one well with that in
positive and negative separations in a neutron-density another well. The data from the well represented by
crossplot. the circles shows the interval to have been uplifted by
800 m relative to the other well because it has lower
Detection of Gas interval transit times (means more compact) but
The rules governing the relationship between neutron occurs at a shallower depth.
log porosity and the true porosity in clean formations
are valid when either water or oil fills the pores (the Seismic Data Calibration and
two fluids have essentially the same hydrogen index, Synthetic Seismogram Generation
Rider 2004). However, hydrocarbon gas has a much A sonic log in a well located on a seismic line or in a
lower hydrocarbon index resulting from its low 3D survey enables the log data to be used to calibrate
408 N.H. Mondol
Fig. 16.22 The density and neutron log responses for shale, separation occurs in shale compared to dolomite. Negative
sandstone, limestone, dolomite sequence on a compatible scale. separation against sandstone may indicate gas. (Modified from
No separation observed in limestone but a large positive Rider 2004)
and check the seismic data. In general, the sonic log seismogram with the real seismic trace measured at
resolution is about 60 cm compared to seismic the well location to improve the picking of seismic
10–50 m. Therefore, the sonic log must be averaged horizons, and to improve the accuracy and resolution
when comparing it with seismic measurement. How- of formations of interest. It should be remembered that
ever, the much higher resolution of the sonic log may the observed seismic trace is primarily a record of the
enable the log information to resolve indications of ability of interfaces between formations to reflect elas-
beds that are just beyond the resolution of the seismic tic waves. This ability is called the reflection coeffi-
technique. Note that the sonic log gives a one-way cient R. The reflection coefficient depends upon the
travel time, and the seismic technique gives a two- properties of the rock either side of the interface, and
way travel time. in particular on its acoustic impedance. The acoustic
A synthetic seismogram is a seismic trace that has impedance is the product of the seismic velocity and
been constructed from various parameters obtainable the density of the rock. Thus, if we can derive the
from log information. It represents the seismic trace density and seismic velocity of a set of formations
that should be observed with the seismic method at the from logs, we can produce a synthetic seismogram
well location. It is useful to compare such a synthetic (Fig. 16.26).
Fig. 16.23 The gas effect in the neutron-density crossplot shows large negative separation in reservoir zone. On the other hand
large positive separations occur in shales (top and bottom of the reservoir). (Modified from Rider 2004)
Fig. 16.24 Recognition of shale compaction and overpressure in shale sequences using the formation density and sonic logs.
(Adapted from Mondol et al. 2008)
410 N.H. Mondol
16.5.5.3 Uncertainties of Porosity Logs

Enlarged and rough boreholes, formation fracture,
gas in the borehole and formation, or improper
centralisation can produce signal attenuation in poros-
ity log measurements resulting in higher porosity
values. In rough holes or in heavy drilling muds, the
density data might be invalid. Porosity calculated from
density and sonic depends on the choice of matrix
density and matrix transit time, which varies with
lithology. Porosity calculations for unconsolidated
formations may yield porosity values higher than the
actual values when using the Wyllie equation (Wyllie
et al. 1956). Porosity calculated from density depends
on the choice of fluid density, which varies with fluid
type and salinity. Porosity calculated in gas-bearing
zones will be higher than the actual values because the
traveltime in gas is higher than in water. Presence of
gas (light HC) in the pore space causes density-derived
porosity to be more than the actual porosity. If the
actual lithology is sandstone, the measured neutron
porosity is less than the actual porosity. If the actual
lithology is dolomite, the log porosity is greater than
Fig. 16.25 Estimation of uplift and erosion from sonic log the actual porosity. If gas is present in the formation,
derived shale compaction trends porosities can be overestimated (Fig. 16.27).
Fig. 16.26 Synthetic seismic trace constructed from impedance and reflectivity convolved with wavelet. (Adapted from Mondol
et al. 2010)
Fig. 16.27 Effect of gas on measured neutron porosity (NPHI) and bulk density logs. (Re-drawn, courtesy of Schlumberger)
16.5.6 Photoelectric Factor/Litho-density 1 σe

Pe (16.9)
Log K Z
16.5.6.1 Generalities and Basic Principles where Pe is the photoelectric absorption index (barns/
The photoelectric factor (PEF) or litho-density log is a electron), σe is the photoelectric cross-section (barns),
new form of the density log with added features. The Z is the atomic number (number of electrons) and K is
tool is physically very similar to the density tool but it a coefficient dependent upon the energy at which the
has enhanced detectors, and the distance between the photoelectric absorption is observed (no units). A log
long spacing and the short spacing detectors has been scale running from 0 to 10 barns/electron is most often
decreased. This decrease has increased the vertical used (Fig. 16.28). The photoelectric log values of
resolution of the tool and improved its overall common minerals and fluids are given in Table 16.7.
counting accuracy. The photoelectric tool has a
caesium-137 source emitting gamma rays at 16.5.6.2 Uses of Photoelectric/Litho-density
0.662 MeV. The more efficient detectors in the litho- Log
density tool have the ability to recognise and count The PEF/litho-density log is one of the most useful logs
separately gamma rays which have high for lithology discrimination. This is simply because the
(0.25–0.662 MeV) and low (0.04–0.0 MeV) energies. tool is sensitive only to the mean atomic number of the
The high energy gamma rays are those that are formation, and at the same time is insensitive to changes
undergoing Compton scattering. The low energy in porosity and fluid saturation in the rock. Hence, the
gamma rays are those that are undergoing photoelec- absolute Pe value may often be used to indicate directly
tric absorption. The probability that a gamma ray is the presence of a given lithology, which may then be
adsorbed by the process of photoelectric absorption checked against the other tool readings for consistency.
depends upon the characteristic photoelectric cross- Uses of PEF to discriminate lithologies are shown below:
section of the process (σe). Characteristic photoelec-
tric cross-sections are measured in barns, where 1 barn Identification of Reservoir Rocks
¼ 10 24 cm2. A specific photoelectric absorption The addition of the photoelectric log to the gamma ray
index Pe is defined with the relationship and neutron-density logs provides both additional
412 N.H. Mondol
Fig. 16.28 Measurements of the photoelectric (litho-density) log for common lithologies. (Modified from Rider 2004)
validation of simple lithology picks and the resolution If there is a mixture of two mineralogies, for exam-
of ambiguities in interpretation of “complex” (multiple a sandy limestone, a crossplot technique or a
mineral) lithologies. The generalised expectations of simple mixing rule can be applied to calculate the
log patterns for shales and endmember reservoir relative proportions of the two mineralogies. In the
lithologies of limestone, dolomite and sandstone are crossplot technique, the Pe value is plotted on the
shown in Fig. 16.29. In the examination of the x-axis against bulk density on the y-axis. The point
neutron-density log overlay, dolomites and siliceous lies between lines that indicate the relative proportions
rocks (either sandstones or cherts) can be recognised of three lithologies (sandstone, dolomite and
by the curve separations. However, the close overlay limestone).
of the two could be caused either by a limestone or a
cherty dolomite (or a cherty dolomitic limestone!). Evaporite Identification
The inclusion of the photoelectric index can be used The geological interpretation of log overlays is easily
to choose between these alternatives. Similarly, a extended to sedimentary lithologies other than the com-
dolomite reading on the photoelectric index curve mon reservoir lithologies of sandstone, limestone and
could also be caused by a cherty or sandy limestone. dolomite. The common evaporite minerals of gypsum,
The simultaneous consideration of the neutron-density anhydrite and halite each have highly distinctive PEF
log overlay resolves the more likely of these two properties as shown in Fig. 16.29. Halite and anhy-
interpretations. drite have markedly low and high bulk densities,
Table 16.7 Common photoelectric log values (adapted from borehole wall, one can see that most types of drilling
Rider 2004) mud will not have a great effect on the PEF measure-
Pe ρb ments. The exception is barite which has a huge PEF
Quartz 1.81 2.65 value and will swamp all other log responses if the tool
K-feldspar 2.86 2.62 sees barite drilling mud. Therefore, in practice the PEF
Calcite 5.08 2.71 log is not used in holes drilled with barite mud.
Dolomite 3.14 2.87
Shale 3.42 2.65
Shaly sand 2.70 2.41
16.5.7 Resistivity and Conductivity Logs
Muscovite 2.40 3.29
Biotite 6.30 3.34
Glauconite 5.32 3.95
16.5.7.1 Generalities and Basic Principles
Halite 4.65 2.07 Resistivity is a fundamental material property that
Anhydrite 5.10 2.98 represents how strongly a material opposes the flow
Gypsum 4.00 2.35 of electric current (Fig. 16.30). Most rock-forming
Coal Anthracite 0.16 1.75 minerals are essentially insulators, while their
Coal Bituminous 0.18 1.47 enclosed fluids are conductors. Almost all conduction
Coal Lignite 0.20 1.19 takes place through the liquid phase, and the resistance
Barite 266.8 4.50 therefore depends primarily on the pore fluid and its
Pyrite 16.97 5.00 salt content. Hydrocarbon fluids are an exception,
Hematite 21.48 5.24 because they are almost infinitely resistive. When a
Magnetite 22.24 5.18 formation is porous and contains salty water, the over-
Pure water 0.358 1.00 all resistivity will be low. When the formation
Salt water (NaCl 120,000 ppm) 0.81 1.19 contains hydrocarbon, or contains very low porosity,
Oil 0.13 0.97 its resistivity will be high. Resistivity is also a function
Methane 0.095 0.25
of the amount of porewater relative to rock volume
(hence porosity) and the distribution of pores in the
respectively, while the very high neutron porosity of rock (permeability).
gypsum is caused by hydrogen in its water of The resistivity log must run in holes containing
crystallisation. electrically conductive mud. Each of the electric log
measurements depends on the degree to which drilling
mud invades the formation, because this will influence
Coal Identification
the electrical properties surrounding the borehole
In combination with density and neutron logs, PEF can
(Fig. 16.31). The resistivity or conductivity
be useful to identify coal (Fig. 16.29). Clastic
measurements are, of course, measuring the same
successions containing coals were commonly devel-
property of the rock, and can be interconverted by
oped in deltaic environments with shales, siltstones
Resistivity ¼ 1/Conductivity. The resistivity units
and sandstones, as well as occasional ironstones (typi-
are ohm-m2/m or simply ohm-m (Ω m). The resistivity
cally siderite). The clay mineralogy of the finer-
of a formation depends on the resistivity of the forma-
grained rocks is quite variable and can show elevated
tion water, the amount of water present, and the struc-
contents of kaolinite, particularly in palaeosols. The
ture and geometry of the pores, and can be expressed
PEF response of coals varies according to their rank
by
(0.20 for lignite, 0.17 for bitumin and 0.16 for
anthracite).
rA
R¼ (16.10)
L
16.5.6.3 Uncertainties of Photoelectric/
Litho-density Log where R is the resistivity (ohm-m), r is the resistance
Most drilling fluids have very low PEF values. When (ohms), A is the cross-sectional area (m2) and L is the
this information is combined with the fact that the litho- length (m). Most formations have resistivities in the
density tool is pad-mounted and pressed against the range 0.2 to 1,000 ohm-m. Resistivities higher than
414 N.H. Mondol
Fig. 16.29 A comparison of litho-density (photoelectric), density and neutron log responses to identify reservoir rocks (sandstone,
limestone, dolomite), evaporate (halite and anhydrite) and coal
Fig. 16.30 A general trend of electrical resistivity of reservoir rocks and fluids
Fig. 16.31 The borehole environment showing zones of electrode disposition in several body-mounted, focused
invasion of drilling fluids and measured resistivity in dif- laterologs (e.g. LLD, LLS, LLM) and induction log (IL) tools
ferent zones (left, courtesy of Schlumberger). Schematic (right, courtesy of Schlumberger)
1,000 ohm-m are uncommon in most permeable current can pass through a formation because it
formations but are observed in impermeable, low contains water with enough dissolved ions to be con-
porosity formations such as evaporites. A few low ductive. With a few rare exceptions, such as clay
porosity hydrocarbon-bearing formations with almost minerals, metallic sulphides and graphite, rock matrix
no formation water can have resistivities as high as is a good electrical conductor. The resistivity and
20,000 ohm-m. Because resistivities cannot be read conductivity tools have many varieties that are com-
accurately over the entire measurement range when monly used to investigate resistivity in different zones
displayed on a linear scale, all resistivity logs are around the borehole (Table 16.8).
presented on logarithmic grids (Fig. 16.32). Laterologs (LLS, LLM and LLD) contain three
electrodes on the tool – a central electrode and two
Resistivity and Conductivity Tools guard electrodes to focus current (Fig. 16.31). Cur-
To measure electrical rock properties, there are two rent in the guards is adjusted to maintain the same
main types of logging tool available: one measures potential as the centre electrode. The lack of poten-
resistivity (resistivity tools) and the other measures tial difference between electrodes means that current
conductivity (induction tools). Formation resistivities flows outwards horizontally. In an induction tool
are measured either by passing a known current (IL), the vertical component of the magnetic field
through the formation and measuring the electrical from the transmitting coil induces ground loop
potential (electrode or galvanic devices) or by induc- currents. The current loops in the conductive forma-
ing a current distribution in the formation and measur- tion produce an alternating magnetic field detected
ing its magnitude (induction devices). An electric by the receiver coil (Fig. 16.31).
416 N.H. Mondol
Fig. 16.32 Example of dual

(ILM and ILD) induction logs
through a hydrocarbon zone.
The drilling mud is
freshwater-based (Rmf > Rw)
where muds invade a
hydrocarbon-bearing
formation (e.g. Sw <60%).
There is high resistivity in the
flushed zone (Rxo) due to
freshwater as well as residual
hydrocarbon, high resistivity
in the invaded zone (Ri), and a
very high resistivity in the
uninvaded hydrocarbon zone
(Rt)
Table 16.8 Types of resistivity and conductivity logging tools

Flushed zone (Rxo) Invaded zone (Ri) Uninvaded zone (Rt)
Microlog (ML) Short Normal (SN) Long Normal (LN)
Microlaterolog (MLL) Medium Induction Log (ILM) Deep Induction Log (ILD)
Proximity log (PL) Shallow Laterolog (LLS) Deep Laterolog (LLD)
Micro Spherically Focused Log (MSFL) Spherically Focused Log (SFL)
16.5.7.2 Uses of Resistivity and Conductivity It also helps to determine hydrocarbon saturation.
Logs Once we know the resistivity of the drilling mud, the
Resistivity/conductivity measurements have immense porosity of the rock can be calculated from the resis-
value for detection and quantitative evaluation of tivity of that part of the rock which has been invaded
hydrocarbon productive zones and correlation of geo- by the mud. In addition, resistivity log is also useful to
logical strata. High resistivity values may indicate a identify source rocks.
hydrocarbon-bearing formation. Oil and gas have far
higher resistivity than water, so resistivity logs can be Detection of Hydrocarbon
used to locate the contacts between oil/water (OWC), By far the most important use of resistivity logs
gas/water (GWC) and gas/oil (GOC) in the reservoirs. is the determination of hydrocarbon-bearing versus
water-bearing zones. Because the rock’s matrix or Table 16.9 Sources of data for calculation of water saturation
grains are non-conductive and any hydrocarbons in the by Archie’s equation
pores are also non-conductive, the ability of the rock to Parameter Source
transmit a current is almost entirely a function of for- Rt Deep resistivity tool
mation brine in the pores. As the hydrocarbon saturation Rw SP log
of the pores increases (causing the water saturation to Calculated from water zone
Measured on RFT (repeated formation tester)
decrease), the formation’s resistivity increases. As the sample
salinity of the water in the pores decreases (as Rw ϕ Neutron, density, sonic, NMR
increases), the rock’s resistivity also increases. F Guessed
When sufficient quantities of hydrocarbons are m Measured in laboratory, Guessed
present, the deep resistivity (ILD) will show extremely n Measured in laboratory, Guessed
high resistivity because of the high saturation in
hydrocarbons (Fig. 16.32). When hydrocarbons are
present, the borehole environment becomes three equation. Table 16.9 summarises the sources of the
phase and much more complex compared to a water parameters that go into the Archie’s equation to calcu-
zone. The freshwater mud will replace the late the Sw.
hydrocarbons immediately around the borehole,
essentially replacing them through the flushed and Formation resistivity factor (F) The bulk resistivity
invaded zones, while the original hydrocarbon satura- of a rock (R0) saturated with a formation fluid of
tion is only found in the uninvaded (virgin) zone. resistivity Rw is directly proportional to the resistivity
of the fluid and can be expressed by the following
equations
Determination of Water Saturation (Sw)
Archie’s law laid the foundation for modern well log R0 ¼ FRw (16.13)
interpretation as it relates borehole electrical resistiv-
ity measurements to hydrocarbon saturations. It is a or
purely empirical law attempting to describe flow in
clean, consolidated sandstones, with varying inter- F ¼ R0 =Rw (16.14)
granular porosity. Archie’s law relates the in-situ elec-
trical resistivity of a rock to its porosity and brine where F is the formation resistivity factor and
saturation and is expressed by the following equation: describes the effect of the presence of the rock matrix.
Note that the formation factor (F) has no units because
m
Rt ¼ aϕ Sw n R w (16.11) of the ratio of two resistivities. It can be seen that
F ¼ 1.00 for a rock with 100% porosity, i.e. no

1=n
a Rw matrix, just 100% fluid. If we take 100% fluid and
Sw (16.12)
Rt ϕ m slowly add grains of rock, the porosity decreases.
However, the insulating grains of rock have negligible
where Sw is the water saturation, a is the tortuosity conductivity (infinite resistivity) compared to the
factor, m is the cementation exponent (usually in the conducting fluid. Hence, R0 will increase, which
range 1.8–2.0 for sandstones), n is the saturation expo- implies that F is always greater than unity in a porous
nent (usually close to 2), Rw is the resistivity of forma- medium. In real rocks F varies between 20 and 500. It
tion water, ϕ is the porosity and Rt is the true is a function of the porosity and permeability of the
formation resistivity as derived from a deep resistivity rock and is an expression of rock properties indepen-
log (e.g. ILD). A geoscientist, by knowing (or deter- dent of the conductivity of the porewater. For
mining) several parameters (F, a, m, and n described in sediments with a high primary porosity the formation
the following paragraphs), and by determining from factor will be an expression of the diagenetic alteration
logs the porosity (ϕ), formation water resistivity (Rw) of the rock and the relationship can be expressed as
and true formation resistivity (Rt), can determine the
formation’s water saturation (Sw) using the Archie F ¼ a=ϕm (16.15)
418 N.H. Mondol
where ϕ is the porosity and m is the cementation Porosity in a water-bearing formation can be related
exponent which varies from 1 for porous rock to 3 to shallow resistivity (Rxo) by the following equation:
for very well cemented rock (average value is 2.0). a is
a constant (tortuosity factor) which for carbonate a Rmf 1=m

rocks is about 1.0 depending on the permeability (or ϕ¼ (16.16)

Rxo
tortuosity).
Cementation exponent (m) The cementation expo- where ϕ is the formation porosity, Rmf is the resistivity
nent expresses how much the pore network increases of mud filtrate at formation temperature, Rxo is the
the resistivity. The cementation exponent has been flushed-zone resistivity, a is the tortuosity factor and
observed near 1.3 for unconsolidated sands, and is m is the cementation exponent. In hydrocarbon-
believed to increase with cementation. Common bearing zones, the shallow resistivity (Rxo) is affected
values for this cementation exponent for consolidated by the unflushed residual hydrocarbons left by the
sandstones are 1.8 < m < 2.0. In carbonates, the invading mud filtrate. These residual hydrocarbons
cementation exponent shows higher variance due to result in a value for shallow resistivity (Rxo) that is
strong diagenetic affinity and complex pore structures. too high because hydrocarbons have a higher resistiv-
Values between 1.7 and 4.1 have been observed. The ity than formation water. Therefore, the calculated
cementation exponent is usually assumed not to be resistivity porosity in hydrocarbon-bearing zones is
dependent on temperature. too low. To correct for residual hydrocarbons in the
Tortuosity factor (a) The tortuosity factor is an flushed zone, water saturation of the flushed zone (Sxo)
important parameter of formation resistivity factor must be known. Then, a formation’s shallow resistiv-
calculations in the Archie formula, which is used to ity (Rxo) can be related to porosity by the following
predict water saturation. It is equal to the square root (Sxo < 1.0):
of tortuosity, is a function of the average angle of
electrical movement with respect to the bulk fluid 1=m
a Rmf
flow and cementation exponent (m) and is related to ϕ¼ 2 (16.17)
the flow area difference between pore throat and pore sxo Rxo
body.
Saturation exponent (n) The saturation exponent
expresses the dependency on the presence of non- Source Rock Investigations
conductive fluid (hydrocarbons) in the pore-space, A source rock can significantly influence the resistiv-
and is related to the wettability of the rock. Water- ity log, depending on the maturity of the organic content:
wet rocks will, for low water saturation values, main- it has little effect when immature, but causes a large
tain a continuous film along the pore walls making the effect when it is mature. A typical shale with matrix
rock conductive. Oil-wet rocks will have discontinu- dominated by clay minerals and silts has a certain
ous droplets of water within the pore space, making water-filled porosity, whereas typically 5–15% of
the rock less conductive. The saturation exponent usu- the matrix of a source shale is comprised of organic
ally is fixed to values close to 2. matter. If the source rock is immature the pore space is
filled with water, but if the source rock is mature, the
Resistivity-Derived Porosity pores contain both water and hydrocarbons and so give
The minerals that make up the grains in the matrix of high resistivity values compared to the normal shale or
the rock, and the hydrocarbons in the pores of the rock, immature source shale. The high resistivity anomaly is
are non-conductive. Therefore, the ability of rock to thus related to the degree of maturity of source rocks
transmit an electrical current is almost entirely the (Passey et al. 1990). A high resistivity in a shale
result of the water in the pore space. Thus, resistivity interval can also be interpreted as carbonate-rich
measurements can be used to determine porosity. Nor- zones, or as highly compacted shales with very low
mally, measurements of a formation’s resistivity close porosity. If the hydrocarbon effect is to be highlighted,
to the borehole (flushed zone, Rxo, or invaded zone, Ri) the resistivity log is crossplotted with either sonic or
are used to determine porosity. When a porous, per- density logs (Fig. 16.33).
meable, water-bearing formation is invaded by drilling It is even possible to calculate the amount of TOC
fluid, formation water is displaced by mud filtrate. in a source rock from the resistivity and sonic logs.
Fig. 16.33 Sonic/resistivity overlay showing ΔlogR separation compared to measured TOC plotted against depth. The
in the organic-rich interval (zone of high gamma ray value). The measured TOC values are from sidewall cores
estimated TOC (using Passey et al. 1990 ΔlogR method)
The sonic log is plotted on a normalised scale with the 16.5.7.3 Uncertainties of Resistivity
resistivity log. When the normalised scales are correct, and Conductivity Logs
the sonic and resistivity logs track one another, regard- Temperature effects The resistivity of formation
less of compaction and compositional changes, but fluids and drilling muds varies greatly with tempera-
separate when a source rock is present (Passey et al. ture. The temperature in a borehole can be found
1990). The degree of separation is said to be related to directly from temperature logs up the entire borehole,
both degree of maturity and magnitude of TOC%, so or more traditionally, from the geothermal gradient. In
that if the level of maturity is known, the TOC% can the former case the temperature is given directly at a
be calculated using the following empirical relational: given depth. However, this is not used to calculate the
mud resistivity.
TOC% ¼ ðΔlogRÞ 10ð2:297 0:1688LOMÞ
(16.18) Effects of drilling mud The resistivity is strongly
influenced by the invasion of drilling mud into the
where TOC% is the total organic carbon in %, LOM is flushed and invaded zones. At least three resistivity
the level of maturity and ΔlogR is the curve separation measurements, each sensitive to a different distance
in resistivity unit. This method seems to be useful away from the borehole, are needed to measure a resis-
qualitatively, but quantitatively is cumbersome and tivity transitional profile. These three measured values
its validity is doubtful. represent the resistivity of flushed (Rxo), invaded (Ri)
420 N.H. Mondol
and uninvaded zones (Rt) (Fig. 16.32). Note that the (Fig. 16.34). A software program is used to correlate
degree of underestimation or overestimation in Rt and deviations on the logs and calculate the amount and
Rw leads to significant error in saturation calculation. direction of bedding dip. As a result structural dip is
Estimation of m and n In attempting to reduce errors determined. The common dipmeter tools are FMS (For-
in Sw calculation we must ensure that the m and n mation MicroScanner Sonde), SHDT (Stratigraphic
parameters in Archie’s equation are measured using High-resolution Dipmeter Tool), HDT (High Resolution
independent methods. Laboratory determined m and n Dipmeter Tool) and OBDT (Oil-based Dipmeter Tool).
values are the best ones to take, though early in a As the dipmeter is brought up the hole, the electrodes on
reservoir’s life these are not available, and so guesses each arm are in contact with the rock layers. If the rock
are used instead. Note that 20% underestimations and layer is dipping, different arms will contact the layers at
overestimations of m and n lead to an underestimation different depths. The sequence of contacts between
and overestimation of the water saturation by 0.11 and individual arms and each layer is used to compute the
0.12 respectively. These are huge errors when dip of the layer. If the layer is horizontal, all arms of the
progressed through to the STOOIP (stock tank oil dipmeter contact the layer at the same level.
originally in place) calculation. There are two common ways to present dipmeter
data; stick and tadpole plots (Fig. 16.34). A stick plot
16.5.8 Dipmeter Logs uses lines (sticks) to show the dip measurements.
Depth is recorded on the vertical axis. The angle on
16.5.8.1 Generalities and Basic Principles the stick is the dip measurement. In a tadpole plot, dip
The dipmeter is essentially a multi-arm is plotted on the horizontal axis with zero dip on the
microresistivity logging tool that provides data used left. Depth is in the vertical axis. Conventional
to compute formation dip. Three to eight spring- dipmeter tadpole plots show the four common dip
loaded arms record separate microresistivity tracks, motifs, that is uniform (green) pattern, upward-
while within the sonde, a magnetic compass records decreasing (red) pattern, upward-increasing (blue) pat-
the orientation of the tool as it is drawn up the hole tern and random (bag o’nails) pattern. Each motif can
3 4 DIP ANGLE
0 10 20 30 40°
1
2
PAD – Uniform (green)
TRAJECTORIES pattern
Upward-decreasing (red)
pattern
PLANES
OF BED
BOUNDARIES Upward-increasing (blue)
pattern
APPARENT
DIP
PLANE OF Random (bag o’ nails)
ELECTRODES
pattern
q
PAD
PAD No 1
No 2
STICK PLOT TADPOLE PLOT
Fig. 16.34 Sketch of a four-arm dipmeter tool (HDT) with the trajectories the pads take when the tool is pulled uphole (modified
from Hepp and Dumestre 1975). Also shown are two common ways (stick plot and tadpole plot) to present the dipmeter data
Fig. 16.35 The dip vs depth, and dip direction vs depth, plots The well does not penetrate the main fault, although several
from a well show an example of flattening drag towards a main minor faults are encountered. (Modified from Hesthemmer
fault. Bedding is dragged from westerly dip at shallower and Fossen 1998)
reservoir (Ness FM) to easterly dip at the base (Cook FM).
be produced by several quite different geological phe- fracture detection, directional data to provide TVD,
nomena. The position of the tadpole head shows the drift surveys and bottom hole location. Because the
amount of dip and the tail shows the direction of dip. dipmeter measures the spatial orientation of physical
properties in rocks, this method is also used to measure
16.5.8.2 Uses of Dipmeter Logs tectonic deformation of rocks, and not least the pri-
Dipmeter data reflect dip and azimuth of bedding. mary orientation of sediment particles (fabric). This is
Variations of these parameters down the boreholes useful in beach sediments as the long axis of sand
help improve our understanding of the structural geol- grains tends to be parallel with the shoreline, while
ogy in faulted reservoirs (Fig. 16.35). The dipmeter in fluvial sandstones the sand grains will be oriented
data is used to evaluate magnitude and direction of parallel with the transport direction. Dipmeter logs can
structural and stratigraphic dips, improving the evalu- therefore be very useful for reconstructing sedimen-
ation of thinly laminated sand/shale sequences, tary environments, transport direction, etc.
422 N.H. Mondol
16.5.8.3 Uncertainties of Dipmeter Logs An applied voltage causes an alternating current to

There are many uncertainties involved in interpreting flow from each electrode into the formation and then
dipmeter logs. The quality of the dipmeter data varies to be received at a return electrode on the upper part of
with the hole conditions, temperature and depth, type the tool. The microelectrodes respond to current den-
of mud used, rock type and tools. Generally FMS data sity, which is related to localised formation resistivity.
are of better quality than SHDT data due to the The tool, therefore, has a high-resolution capability in
increased button density and borehole coverage. Simi- measuring variations from button to button.
larly, SHDT gives better results than HDT due to Acoustic borehole-imaging devices are known as
improvements of the tool and the algorithms used in “borehole televiewers”. They provide 100% coverage
processing. Data from the OBDT are generally of of the borehole wall. Modern acoustic borehole imag-
poorer quality than any of the resistivity tools (related ing tools contain a magnetometer to provide azimuthal
to the use of oil-based mud), and results must accord- information. The borehole televiewer operates with
ingly be treated with care. Also it is not always possi- pulsed acoustic energy so that it can image the bore-
ble to find dipmeter patterns that are compatible with hole wall in the presence of opaque drilling muds.
sedimentological models of cross-bedded strata. Short bursts of acoustic energy are emitted by a rotat-
ing transducer in pulse-echo mode. These travel
through the drilling mud and undergo partial reflection
16.5.9 Image Logs at the borehole wall. Reflected pulses are received by
the transducer. The amplitudes of the reflected pulses
16.5.9.1 Generalities and Basic Principles form the basis of the acoustic image of the borehole
Borehole imaging has been one of the most rapidly wall. Data are usually presented as depth plots of
advancing technologies in wireline well logging. The enhanced images of amplitude and borehole radius
term “borehole imaging” refers to those logging and (Fig. 16.36). To some extent, the acoustic and electri-
data-processing methods that are used to produce cal images are complementary because the ultrasonic
centimetre-scale images of the borehole wall and the measurements are influenced more by rock properties,
rocks that make it up. The context is, therefore, that of whereas the electrical measurements respond primar-
open hole, but some of the tools are closely related to ily to fluid properties.
their cased-hole equivalents.
Borehole imaging can be performed by optical 16.5.9.2 Uses of Image Logs
imaging, electrical imaging and acoustic imaging. A Borehole image logs provide high-resolution direc-
wide range of imaging tools is available, among them tional data sets and are powerful tools in subsurface
CBIL (Circumferential Borehole Imaging Log), UBI reservoir characterisation. They span the scale gap
(Ultasonic Borehole Imaging) and CAST (Circumfer- between core and seismic observations and although
ential Acoustic Scanning Tool), with LWD imaging they do not replace cores, they provide key sedimen-
tools becoming increasingly important. Optical imag- tological and sub-seismic structural information,
ing tools were the first borehole imaging devices. allowing quantification of subsurface fracture
Today they furnish a true high-resolution colour networks and providing inputs to geomechanical and
image of the wellbore. Video cameras may record petrophysical studies. The applications of image logs
the lamination and bedding and because they are ori- range from detailed reservoir description through res-
ented, palaeocurrent direction may be inferred. This ervoir performance to enhanced hydrocarbon recov-
method may also be very useful for recording fractures ery. Specific applications are fracture identification,
and large vuggy pores. analysis of small-scale sedimentological features,
Electrical (microresistivity) imaging devices were evaluation of net pay in thinly bedded formations,
developed as an advancement on dipmeter technology. and the identification of breakouts (irregularities in
Traditionally, they have required a conductive bore- the borehole wall that are aligned with the minimum
hole fluid, but this requirement has since been obviated horizontal stress and appear where stresses around the
by oil-based mud imaging tools. In the microresistivity wellbore exceed the compressive strength of the rock).
imaging tool, the pads and flaps contain an array of It is important to detect open fractures and also partly
button electrodes at constant potential (Fig. 16.36). healed fractures which cannot be closed by increased
Fig. 16.36 (a) Measurement principle of electrical imaging Breakouts are indicated by the low acoustic amplitude of the
(microresistivity imaging) devices illustrated by Schlumberger’s reflected signal, shown here as darker areas. The breakouts are
Formation MicroImager (FMI) and (b) Example of breakout rotated because of a drilling-induced slippage of localised faults.
detection using an acoustic imaging borehole televiewer. (Courtesy of Schlumberger)
stress. Electrical images are similar to those obtained fraction of an inch. In theory, any feature that is as
by dipmeter logs. Modern image logs provide a good large as the buttons will be resolved. If it is smaller, it
overview of the lamination and also fractures. This might still be detected.
may be compared with sedimentological logging of a Acoustic imaging amplitudes are governed by sev-
core (Fig. 16.37). eral factors. The first is the shape of the borehole wall
itself: irregularities cause the reflected energy to scat-
16.5.9.3 Uncertainties of Image Logs ter so that a weaker reflected signal is received by the
The principal drawback of optical imaging is that it transducer. Examples of these irregularities are
requires a transparent fluid in the borehole. Unless fractures, vugs and breakouts. Borehole televiewers
transparent fluid can be injected ahead of the lens, work best where the borehole walls are smooth and
the method fails. This requirement has limited the the contrast in acoustic impedance is high. The scat-
application of downhole cameras. The electrical imag- tering or absorption of acoustic energy by particles in
ing tool does not provide an absolute measurement of the drilling mud can affect image resolution. This
formation resistivity but rather a record of changes in problem is more serious in heavily weighted muds,
resistivity. The resolution of electrical microimaging which are the most opaque acoustically and cause loss
tools is governed by the size of the buttons, usually a of image resolution.
424 N.H. Mondol
hydrocarbon. There are many physical parameters that

can be recorded, depending upon requirements.
Well logging is the most abundant data acquisition
technique available to help evaluate formation and
determinate fluid saturation. Modern tools may record
up to ten different types of measurement simulta-
neously during a single run. The most common
measurements are radioactive, electric and acoustic
logging. Rock properties are sensitive to changes in
lithology and fluid content. Changes in rock properties
are used to mark different formations. Well logs may
be used both as a visual qualitative tool to identify the
main rock types and also as a basis for sedimentologi-
cal and stratigraphic interpretation.
In addition to providing information about the poros-
ity and permeability of rock types, well logs can be used
directly for interpreting depositional environments.
Gamma logs and self-potential logs record characteris-
tic coarsening-upward and fining-upward sequences
very efficiently. We can therefore often recognise flu-
vial channels (fining-upward) and shallow marine
deposits (coarsening-upward). Turbidities give a char-
Fig. 16.37 Recognition of sedimentary and structural features acteristic alternation between coarse and fine material.
in microresistivity images (left). These Formation MicroImager
(FMI) images have been used to generate the dip information In delta sequences coal or lignite beds give characteris-
(middle). The combination of FMI images and dip data clearly tically strong deflections on the resistivity log. It is
differentiates the aeolian and interdune sands (right). (Courtesy important, however, to bear in mind that the response
of Schlumberger) of the logs is not due to sedimentological parameters
such as grain size and sorting, but can be indirectly
correlated with these parameters.
16.6 Summary Physical measurements are related to rock
properties such as lithology, porosity (fracture, vugs,
Well logs are the results of several geophysical primary and secondary), permeability, shaliness, and
measurements recorded in a well bore. Logging the type (oil, gas or condensate) and saturation of
provides continuous recording of a physical parameter hydrocarbons. More quantitative calculations of
of the formation as a function of depth. The primary porosity, shaliness and hydrocarbon saturation can
objectives of the well logging are (1) identification of also be made based on log data. Crossplots which are
source, reservoir and cap rocks, (2) determination of calibrated for different types of lithologies are then
depth, thickness, temperature and pressure of the reser- required. Correlation between wells can map units
voir, (3) to distinguish between oil, gas and water zones across a reservoir. Sonic and density logs provide a
in the reservoir, (4) estimation of hydrocarbon in place, link between seismic and petrophysics.
and (5) estimation of recoverable hydrocarbon. Well In conclusion, when logging is properly applied, it
logging is broadly classified into open-hole logging, can help to answer many questions from a wide spec-
cased-hole logging, MWD and LWD. In open-hole trum of special interest groups on topics ranging from
logging, the tool is lowered to the desired depth and basic geology to petroleum economics. A single tool
data is acquired as the tool is pulled up. Logs are cannot perform a complete diagnosis of lithology and
recorded to measure different physical parameters of a fluids. Usually a group of log measurements is made
well to ascertain the capacity of the well to flow in the same hole to improve interpretation. Of equal
importance, however, is the fact that logging by itself Mondol, N.H. 2010. Seismic exploration. In: K. Bjørlykke (ed.),
cannot provide answers to all questions. Coring, core Petroleum Geoscience: From Sedimentary Environments to
Rock Physics. Springer-Verlag, Berlin. doi:10.1007/978-3-
analysis, and formation testing are also integral parts 642-02332-3_17.
of any formation evaluation effort. Mondol, N.H., Fawad, M., Jahren, J. and Bjørlykke, K. 2008.
Synthetic mudstone compaction trends and their use in pore
pressure prediction. First Break 26(12), 43–51.
Passey, Q.R., Creaney, S., Kulla, J.B., Moretti, F.J. and Stroud,
J.D. 1990. A practical model for organic richness from
Further Reading porosity and resistivity logs. AAPG Bulletin 74(12),
1777–1794.
Asquith, G. and Krygowski, D. 2004. Basic Well Log Analysis. Pickett, G.R. 1963. Acoustic character logs and their applica-
Methods in Exploration Series 16, AAPG Tulsa, OK, 244 pp. tion. Journal of Petroleum Technology 15(6), 659–667.
Coates, G.R., Xiao, L. and Prammer, M.G. 1999. NMR Logging Plint, A.G., Norris, B. and Donaldson, W. 1990. Revised
Principles and Application. Halliburton Energy Services definitions for the Upper Cretaceous Bad Heart Formation
Publication H02308, Houston, TX, 234 pp. and associated units in the foothills and plains of Alberta and
Ellis, D.V. and Singer, J.M. 2008. Well Logging for Earth British Columbia. Bulletin of Canadian Petroleum Geology
Scientists (2nd ed.). Springer, Dordrecht, 685 pp. 38(1), 78–88.
Hepp, V. and Dumestre, A.C. 1975. Cluster—A method for Raymer, L.L., Hunt, E.R. and Gardner, J.S. 1980, July. An
selecting the most probable dip results from dipmeter improved sonic transit time-to-porosity transform. In
surveys. SPE Paper 5543, Society of Petroleum Engineers, SPWLA twenty-first annual logging symposium, paper P.
Dallas, TX. Rider, M.H. 2004. The Geological Interpretation of Well Logs.
Hesthemmer, J. and Fossen, H. 1998. The use of dipmeter data Blackie, Glasgow, 280 pp.
to constrain the structural geology of the Gullfaks Field, Wyllie, M.R.J., Gregory, A.R. and Gardner, L.W. 1956. Elastic
northern North Sea. Marine and Petroleum Geology 15(6), wave velocities in heterogeneous and porous media. Geo-
549–573. physics 21(1), 41–70.
Chapter 17
Seismic Exploration
Nazmul Haque Mondol
17.1 Introduction generating, recording and analysing sound waves that

travel through the Earth (such waves are also called
Of all the geophysical exploration methods, seismic seismic waves). The density and velocity changes
surveying is unequivocally the most important, primar- between rocks reflect the waves back to the surface,
ily because it is capable of detecting large-scale to and how quickly and strongly the waves are reflected
small-scale subsurface features. Simply stated, seismic back indicates what lies below. Seismic pulses for
methods involve estimation of the shapes and physical exploration surveys are generated in one of three
properties of Earth’s subsurface layers from the returns ways, employing an air-gun, vibrator or dynamite. An
of sound waves that are propagated through the Earth. air-gun source is used for marine acquisition whereas
Early wildcatters found oil by drilling natural oil seeps vibrator and dynamite are the common sources for land
and large folds (anticlines) in exposed rocks. These seismic surveys. The strength of pulses associated with
easy oil prospects were all quickly discovered and different seismic surveys varies, depending on site-
drilled, and geologists then turned to seismic surveys specific factors such as rock types, how deep the survey
to find less obvious oil and gas traps. Seismic technol- needs to image and the required source.
ogy had been used since the early 1900s to measure
water depths and detect icebergs, and by 1924, seismic
data were first used in the discovery of a Texan oil field 17.2 Basic Principles
(Milligan 2004). Several introductory and advanced
textbooks (e.g., Telford et al. 1990, Sheriff and Geldart The first step in seismic exploration is to acquire data,
1995, Yilmaz 2001) describe the principles of acquisi- which in most cases is carried out from the surface. To
tion, processing and interpretation of seismic data. This understand the seismic data, a review of the physical
chapter reviews the fundamental concepts employed in principles that govern the movement of seismic waves
seismic exploration. through layered media is necessary. A seismic source
In general, two types of seismic method (reflection at any point on the Earth generates four types of
and refraction) are common, with reflection seismic the seismic waves: compressional (P-wave), shear (S-
most widely used technique in hydrocarbon explora- wave), Rayleigh (ground roll) and Love, that travel
tion. This technique provides an image of the subsur- through the layers (Fig. 17.2). Each layer will have a
face in two or three dimensions (2D or 3D) (Fig. 17.1). specific density and velocity. Rayleigh and Love
The subsurface seismic images are produced by waves are surface waves and propagate approximately
parallel to the Earth’s surface. Although surface waves
penetrate to significant depth in the Earth, these types
of waves do not propagate directly through the Earth’s
N.H. Mondol (*)
interior and have limited significance in oil and gas
Department of Geosciences, University of Oslo; Norwegian
Geotechnical Institute (NGI), Oslo, Norway exploration. On the other hand, P- and S-waves are
e-mail: [email protected] often called body waves because they propagate
428 N.H. Mondol
Time (s)
Fig. 17.1 Subsurface imaging by 2D (left) and 3D (right) seismic reflection data
X Directi Directi
on of
propa
X on of
propa
Y gation gation
Y
Z Z
P wav S wav
e e
Partic
le mo Partic
tion le mo
tion
X Directi X Directi
on of on of
propa propa
gation Y gation
Y
Z Z
Rayle Love
igh w wave
ave
Partic Partic
le mo le mo
tion tion
Fig. 17.2 Propagation of body (P and S) and surface (Rayleigh propagation and perpendicular (in a vertical plane) so that the
and Love) waves as a function of particle motions. P-waves motion is generally elliptical – either prograde or retrograde.
shake the ground in the direction they are propagating, while Love waves shake the ground perpendicular to the direction of
S-waves shake perpendicularly to the direction of propagation. propagation and generally parallel to the Earth’s surface.
Rayleigh waves shake the ground both in the direction of (Source: Braile 2000)
outward in all directions from the source and travel compared to the line of travel), and the medium must
through the interior of the Earth and have great signif- move away from the material right next to it to cause
icance in seismic exploration. P-waves move faster the shear and transmit the wave. This takes more time,
than S-waves. The P-wave is a longitudinal wave, which is why the S-wave moves more slowly than the
the force applied in the direction that the P-wave is P-wave in seismic events. S-waves do not travel
travelling. The ground must move in that direction. through fluids as fluid has no shearing capacity.
The ground or Earth is incompressible, so the energy is The basic principle of seismic survey is to initiate a
transferred pretty quickly. In the S-wave, the medium seismic pulse from a seismic source at or near the
is displaced in a transverse way (up and down – Earth’s surface and record the amplitudes and travel
17 Seismic Exploration 429
Normal
incidence
Incident Reflected
P-wave S-wave
Reflected
P-wave ρ1V1
–
θ1 θ3 ρ 2V2
Medium 1 V1
Medium 2 V2
Refracted Fig. 17.4 A seismic section showing a large impedance con-
θ2 P-wave trast within the target zone (marked by red circle)
Refracted product of P-wave (Vp) or S-wave (Vs) velocities and

S-wave
density (ρ). The reflection coefficient (R) of a normally
incident P-wave on a boundary is given by:
Fig. 17.3 A schematic diagram of reflected and refracted ρ 2 V2 ρ 1 V1

waves generated from an incident P-wave. The angle between R¼ (17.2)
the normal to the interface of two media and an incident P-wave ρ 2 V2 þ ρ 1 V1
is the angle of incidence (θ1), and is equal to the angle of
reflection (θ3) in isotropic media. The angle of refraction (θ2) where ρ1 and ρ2 are the densities of the upper and lower
depends on the velocity of the wave in that medium layers, V1 and V2 are their respective P-wave velocities,
and ρ1 V1 and ρ2 V2 are the P-impedances of the upper
times of waves returning to the surface after being and lower layers respectively. Therefore, anything that
reflected or refracted from the interface(s) of one or causes a large contrast in impedance in the target zone
more layers. When a seismic source emits a pulse that can cause a strong reflection (Fig. 17.4). The possible
propagates through the sedimentary layers, the sound candidates include changes in lithology, porosity, pore
waves travel between the layers with different velocities fluid, degree of saturation and diagenesis. We see that
and will be refracted according to Snell’s law: the greater the difference in density and velocity of two
sin θ1 = sin θ2 ¼ V1 =V2 (17.1) layers, the greater the amount of energy which will be
reflected. Sandstone will often have significantly differ-
where V1 and V2 are the velocities of the first and ent acoustic impedance from shale, and a considerable
second media, sinθ1 and sinθ2 are the sines of the amount of sound energy will be reflected from the
incidence and refracted angles, and θ3 is the reflected boundary between a sandstone bed and a shale bed.
angle (Fig. 17.3). Snell’s law describes the changes in Limestones will tend to have both high velocities and
the direction of a wavefront as it travels in media of high densities. The result will be even greater contrast
different velocities. If the seismic wave is incident at in acoustic impedance between limestones and, for
an angle, both reflected and refracted P- and S-waves example, shales. However, this contrast will always
will be generated at an interface between two media. depend on the porosity of the limestone, though even
However, at a fluid-solid interface like the seafloor, S- rather porous limestones have relatively high velocities
waves will not exist in the fluid part. because they are usually well cemented.
A line or grid of geophones or hydrophones called
seismic receivers records the reflected and refracted
seismic signals. Reflections from the layer interfaces 17.3 Seismic Sources and Reservoirs
in the subsurface are then measured at receivers (time
measurements). If the two layers have different It is essential at this point to understand the principle
velocities, they will as a rule also have different of seismology. Seismology is based on the transmis-
densities, and part of the acoustic energy will not be sion of sound waves by the rocks of the crust. Strong
refracted, but reflected. How much of the energy is earthquakes create pressure waves (natural sources of
reflected depends on the difference in the impedance seismic waves) that are transmitted through the entire
[P-impedance (Zp) or S-impedance (Zs)], which are the Earth and detected by seismographs (receivers) on the
430 N.H. Mondol
water surrounding the gun (Fig. 17.6b). A vibrator is

s P an
wave dS an adjustable mechanical source that delivers vibra-
or S
es o P wa
N ve tory seismic energy into the ground (Fig. 17.6a). A
av
ly w Shadow s
Pa
se
on S
zone vibrator source sends a controlled-frequency sweep
ves No
nd
en
into the ground. The recorded data are then convolved
S
wa
P waves s with the original sweep to produce a usable signal.
ve
P wa
Outer Dynamite – a combination of explosive and detonator,

core Inner core
is used as a seismic source. The detonator helps to
Earthquake
ignite the explosives. When dynamite ignites, a shock
wave propagates with a speed of 3,000–10,000 m/s. It
provides an impulsive energy that can be converted
into ground motion. It is customary to drill a hole to
load dynamite and fill it with heavy mud before shoot-
ing. Dynamite can generate usable signal strengths and
surface waves
a bandwidth that covers a wide spectrum of seismic
energy. It includes a variety of energy sources based
minute mark P S on varying explosive output parameters to meet geo-
logical and climatic conditions.
17.3.2 Seismic Receivers

Fig. 17.5 A cross-section of the Earth with earthquake wave
paths defined and their shadow zones highlighted (top). A typi- Hydrophones and geophones serve as receivers for
cal seismogram (bottom) shows the fastest P-waves. The next set seismic signals. The hydrophone is a device designed
of seismic waves is the S-waves. The surface waves (Love and for use in detecting seismic energy in the form of
Rayleigh waves) travel a little slower than S-waves (which, in pressure changes in water during marine seismic
turn, are slower than P-waves) and have lower frequency
acquisition (Fig. 17.7a). It measures pressure
variations with the aid of piezoelectric material,
other side (Fig. 17.5). Seismic exploration, however,
which generates a voltage upon deformation. The
as employed by the petroleum geologist, makes use of
two piezoelectric elements in one hydrophone are
artificially generated pulses (seismic sources). The
connected and polarised so that voltages due to pres-
principle is simple; an impulse source sends acoustic
sure waves (returning signal) add and voltages due to
energy into the Earth. This energy propagates in many
one-directional acceleration will cancel. In this way
directions and is reflected and refracted when it
the influence of movements due to currents, wave
encounters boundaries between two layers. Sensors
action and so on will be minimised. Hydrophones are
(seismic receivers) placed on the surface measure the
combined to form streamers that are towed by seismic
reflected or refracted acoustic energy. These artificial
vessels or deployed in a borehole. A typical length of a
sources are much weaker than the natural seismic
streamer is about 4–6 km where a single receiver
source (earthquake) but are more focused towards
section is typically 75 m long and contains 96
areas of specific stratigraphic interest.
hydrophones which are grouped in arrays of a pre-
defined length, mostly 12.5 or 25 m.
The geophone is a device used in surface seismic
17.3.1 Seismic Sources acquisition, both onshore and on the seabed offshore,
that detects ground velocity produced by seismic
Different seismic sources are usually used in land and waves and transforms the motion into electrical
marine acquisitions. In marine environments seismic impulses (Fig. 17.7b). Geophones, unlike
energy is normally generated using arrays of air-guns, hydrophones, detect motion rather than pressure. Con-
whereas in land seismic one often uses explosives or ventional seismic surveys on land use one geophone or
vibrators. An air-gun is a device that releases highly a group of geophones per receiver location to detect
compressed air (at typically 2,000–5,000 psi) into the motion in the vertical direction. The three-component
a c
Fig. 17.6 (a) Seismic acquisition in a desert where vibrator is combined with detonators (bottom) form dynamite which is a
the most common source of seismic energy. (b) An air-gun common source of seismic energy in land acquisition where
before deployment in the water. It releases compressed air into vibratory trucks can not be used due to rugged topography
the water during marine seismic survey. (c) Explosives (top)
a b
Fig. 17.7 (a) Hydrophone, (b) geophone and (c) multi-component geophone. (Courtesy ION Geophysical)
432 N.H. Mondol
305 1219 2438 3658 4877 6096 610 1524 2743 3962 5182 305 1219 2438 3658 4877 6096 610 1524 2743 3962 5182
900
1000
1100
1200
1300
Fig. 17.8 Recorded seismic traces. Each trace consists of one because of distance of receivers from source. In conventional
recording corresponding to a single source-receiver pair. Seis- acquisition, strings of geophones hard-wired together average
mic energy recorded at the receivers arrives at different times the individual sensor measurements and deliver one output trace
(3C) geophone is used for direct measurements of seismic data. P-waves incident on an interface at other
shear waves at the seafloor. An essential feature of a than normal incidence angle can produce reflected and
seafloor seismic acquisition is the four-component transmitted S-waves. S-waves travel through the Earth
(4C) detector unit, which includes a hydrophone and at about half the speed of P-waves and respond differ-
a three-component (3C) geophone (Fig. 17.7c). The ently to fluid-filled rocks, and so can provide different
hydrophone and the vertical geophone measure pres- additional information about lithology and fluid con-
sure waves, whereas the two extra horizontal tent of hydrocarbon-bearing reservoirs.
geophones measure particle velocity associated with The recorded seismic trace is a convolution (*) of
shear wave energy. the source signal and the reflectivity sequence of the
Earth plus noise (Fig. 17.9). A seismic trace can sim-
ply be expressed by the Eq. (17.3) where multiples are
17.4 Seismic Acquisition not considered. Transmission losses and geometric
spreading are not included and the frequency-
The seismic survey is an essential part of the whole dependent absorptions are also ignored in the
cycle of petroleum exploration and production. Seis- equation.
mic surveys are carried out on land and in transition
zone, shallow marine and marine environments in S ¼ W R þ Noises (17.3)
different ways. The basic principle is an impulse
source such as dynamite, air-gun or vibrator that where S is the recorded seismic trace, R is the
sends acoustic energy into the Earth. This energy reflectivity and W is the wavelet. A wavelet is a kind
propagates in many directions. Downward travelling of mathematical function used to divide a given func-
energy reflects and refracts when it encounters tion into different frequency components and study
boundaries between two layers with different acoustic each component with a resolution that matches its
properties (Fig. 17.3). Sensors or geophones placed on scale. Accurate wavelet estimation is absolutely criti-
the surface measure the reflected acoustic energy, cal to the success of any seismic inversion. The
converting it into an electrical signal that is displayed inferred shape of the seismic wavelet may strongly
as a seismic trace (Fig. 17.8). The typical recorded influence the seismic inversion results and therefore
seismic frequencies are in the range of 5–100 Hz. P- subsequent assessments of the reservoir quality. Atten-
waves are the waves generally studied in conventional uation (amplitude loss) of seismic waves is an
Wavelet
Impedance Reflectivity Seismic trace
Fig. 17.9 Seismic trace is a result of convolution of a wavelet and the reflectivity series plus noises
important phenomenon and caused by three major common midpoint gather (CMP-gather). The number
factors: (a) Geometric spreading: progressive diminu- of traces summed or stacked is called a fold. For
tion of amplitude (proportional to the inverse of prop- instance, in 24-fold data, every stacked trace
agation distance) caused by the increase in wavefront represents the average of 24 traces. The nominal fold
area, (b) Intrinsic attenuation: energy losses due to (F) is the maximum number of traces in a CMP-
internal friction and (c) Transmission losses: reduction gather. For a standard marine seismic acquisition
in wave amplitude due to reflection at interfaces. folds (F) is given by the formula:
NΔg
F¼ (17.4)
17.4.1 Terminology Used in Seismic 2Δs
Acquisition
where N is the number of channels, Δg is the group
interval and Δs is the source interval. The rate of
A Trace is a seismic time measurement corresponding
repetition of complete wavelengths of seismic waves
to one source-receiver pair, and offset is the distance
is referred to as frequency (f) and is measured in cycles
between source and receiver for a given trace. In
per second or hertz.
practice, traces from one source are simultaneously
recorded at several receivers. Then, sources and
receivers are moved along the survey line and another
set of recordings is made. When a seismic wave travels 17.4.2 Marine Acquisition
from a source to a reflector and then back to receiver,
the elapsed time is called the two-way traveltime. The Marine seismic acquisition is generally accomplished
common depth point (CDP) is the halfway point of the using large ships with one or multiple air-gun arrays
path only where the Earth is horizontally layered; it is for sources (Fig. 17.10). Air-guns are deployed behind
situated vertically below the common midpoint the seismic vessel and generate a seismic signal by
(CMP). A gather is a family of traces (e.g. shot-point forcing highly pressurised air into the water at a given
gather is the family of all traces corresponding to the interval. Receivers are towed behind the ship in one or
same source firing). Sorting of traces by collecting several long streamer(s) that are several kilometres
traces that have the same midpoint (CMP) is called a long. Marine receivers are composed of piezoelectric
434 N.H. Mondol
Fig. 17.10 A 2D marine seismic acquisition
hydrophones, which respond to changes in water pres-

sure. In marine acquisition, seismic vessels sail along
predetermined patterns of parallel circuits
(Fig. 17.11). The length of straight segments is calcu-
lated from fold plots, and must include additional
length – run in and run out – to allow the cable to
straighten after each turn. Marine seismic acquisition
is faster than conventional land seismic because it does
not require jug hustlers to lay and pick up geophones
(Rygg et al. 1992).
Bin
The advance to 3D seismic acquisition and imaging
of the subsurface, introduced in the 1980s, was per-
haps the most important step in seismic exploration
(Beckett et al. 1995). The 3D seismic images began to
resolve the detailed subsurface structural and strati-
graphic conditions that were missing or not
discernable from 2D seismic data. With 3D seismic
acquisition potential reservoirs are imaged in three
dimensions, which allows seismic interpreters to
view the data in cross-sections along 360 of azimuth,
in depth slices parallel to the ground surface, and along
planes that cut arbitrarily through the data volume.
Information such as faulting and fracturing, bedding Fig. 17.11 Marine acquisition geometry showing seismic
vessels looping in oblong circuits. (Source: Ashton et al. 1994)
plane direction, the presence of pore fluids, complex
geological structure, and detailed stratigraphy are now
commonly interpreted from 3D seismic data sets. acquired data can be processed in a more consistent
In 2D marine data acquisition a single streamer is manner and be further manipulated using modern
deployed, whereas in 3D acquisition multiple visualisation tools. In 2D data acquisition the data
streamers are towed behind the boat (Fig. 17.12). 3D collection occurs along a line of receivers. The
Fig. 17.12 Showing a towed streamer for 3D seismic acquisition. (Courtesy ION Geophysical)
resultant image represents only a section below the In reality, midpoint distributions within a cell are not
line. Unfortunately, this method does not always pro- necessarily uniform since cable shape varies from shot
duce a clear subsurface image. 2D data can often be to shot and line to line. Such side drift of the cables is
distorted with diffractions and events produced from called feathering.
offline geologic structures, making accurate interpre- Another advanced technique of marine seismic is
tation difficult. Because seismic waves travel along ocean bottom seismic (OBS) acquisition. It gives the
expanding wavefronts they have a surface area. A possibility of direct measurement of S-wave data in
truly representative image of the subsurface is only addition to P-wave data by using ocean bottom cables
obtained when the entire wavefield is sampled. that have three component geophones (3C) and a
A 3D seismic survey where the vessel is towing hydrophone in addition (thus 4C in total). The 4C
many streamers (up to 20) at the same time with cable can be up to 6 km long with 240 stations (i.e.
multiple arrays of airguns is more capable of accu- 960 channels since 4C). A typical OBS layout involves
rately imaging reflected waves because it utilises mul- 4 or more cables (Fig. 17.13). The optimal choice of
tiple points of observations (Fig. 17.12). In the case of acquisition geometry for a 4C survey hinges on both
3D survey we have seen that seismic data are sampled geophysical and financial considerations. Most
from a range of different angles (azimuth) and source- designs can be classified as either patch or swath. In
receiver distances (offsets). After seismic processing swath designs, the source lines are parallel to receiver
the data can then be represented as 3D volume images lines, while in patch designs, source lines are perpen-
of the subsurface. In 2D processing, traces are col- dicular to receiver lines. Patch design produces seis-
lected into CMP gathers, while in 3D, traces are col- mic data of a relatively wide range of azimuths,
lected into common-cell gathers (binning). To perform whereas the swath design produces data of a limited
3D binning, a grid is first superimposed on the survey or narrow range. Moreover, the swath design offers a
area. This grid consists of cells with dimensions of half more uniform sampling of offsets with better near-
the receiver group spacing in the inline direction, offset coverage. The use of 4C OBS recording has
equivalent to the CMP spacing in 2D processing, and several advantages over conventional towed streamer
the line spacing in the crossline direction (Fig. 17.11). technology, which includes:
436 N.H. Mondol
Fig. 17.13 An Ocean Bottom Seismic (OBS) marine acquisition technique. (Courtesy ION Geophysical)
(1) Dual-sensor summation (3C geophone + hydro- The variations in ground elevation in land acquisition
phone) for the suppression of receiver-side causes sound waves to reach the recording geophones
multiples. with different traveltime. The Earth’s near-surface
(2) Utilising P–S wave conversions for enhanced layer may also vary greatly in composition, from soft
imaging (Fig. 17.14). alluvial sediments to hard rocks. This means that the
(3) Attenuation of free surface multiples when com- velocity of sound waves transmitted through this sur-
bined with towed streamer recording. face layer may be highly variable. Static corrections –
A comparison of the migrated P–S stack versus the a bulk time shift applied to a seismic trace – are
P–P stack is shown in Fig. 17.15. The P–S stack is typically used in seismic processing to compensate
produced from OBS converted wave data whereas the for these differences in elevations of sources and
P–P stack is produced from 3D towed streamer P- receivers and near-surface velocity variations
wave data. From this comparison it is clear that OBS (Ongkiehong and Askin 1988).
data can be used to successfully image through a gas
chimney.
17.4.4 Vertical Seismic Profile (VSP)
17.4.3 Land Acquisition VSP is another technique of seismic acquisition, used

for correlation with conventional seismic data (land or
A complication in land acquisition is that, unlike marine seismic). The defining characteristic of a VSP
marine data, a seismic line is rarely shot in a straight (of which there are many types) is that either the
line because of the presence of natural and man-made energy source, or the receivers (or sometimes both)
obstructions such as lakes, buildings and roads are in a borehole. In the most common type of VSP,
(Fig. 17.16). The shot points and the receivers may hydrophones, or more often geophones, in the bore-
be arranged in many ways. Many groups of geophones hole record reflected seismic energy originating from a
are commonly used on a line with shot points at the seismic source at the surface (Fig. 17.17).
end or in the middle of the receiver array. The shot The VSPs vary in the well configuration, the num-
points are gradually moved along a line of geophones. ber and location of sources and geophones, and how
Fig. 17.14 A comparison of seismic data acquired by the towed streamer (top) and OBS (bottom) techniques. The OBS survey
significantly improves the subsurface image. (Source: Thompson et al. 2007)
they are deployed. VSP uses the reflected energy of areas with significant changes or with virtually
contained in the recorded trace at each receiver posi- unchanged hydrocarbon-indicating attributes helps to
tion as well as the first direct path from source to determine new drilling sites in an already existing
receiver. VSPs include the zero-offset VSP, offset production field. For this method it is critical that the
VSP, walkaway VSP, walk-above VSP, salt-proximity observed seismic changes can be related to the fluid
VSP, shear-wave VSP, and drill-noise or seismic- flow. Differences in data acquisition, survey orienta-
while-drilling VSP. tion, processing and data quality can introduce signifi-
cant noise in a 4D analysis. Hence, such differences
must be corrected for as best as possible. Further
17.4.5 4D Seismic details of 4D seismic are discussed in Chap. 19. The
known applications of 4D seismic can be summarised
The acquisition of 4D or time-lapse seismic has as:
opened new horizons for monitoring reservoir (1) Monitoring the spatial extent of steam injection
properties such as fluids, temperature, saturation and used for thermal recovery.
pressure changes during the productive life of a field (2) Monitoring the spatial extent of the injected water
(Aronsen et al. 2004). 4D seismic is based on the front used for secondary recovery.
analysis of repeated 3D seismic data. The differences (3) Imaging bypassed oil or gas.
in seismic attributes over time are caused by changes (4) Determining the flow properties of sealing or
in pore fluid and pore pressure associated with the leaking faults.
drainage of a reservoir under production. Detection (5) Detecting changes in oil-water contact.
438 N.H. Mondol
2.25 Conventional 3D data S N
2.50
2.75
Time (sec.) 3.00
3.25
3.50
3 km
3.75 1/9-1 1/9-3R
X-component
4.50
5.00
Time (sec.)
5.50
6.00
6.50
7.00 1/9-1 1/9-3R
Fig. 17.15 Comparison of P–P stack of conventional 3D streamer data (top) and P–S stack of OBS data (bottom). Note how the OBS data
produces a much better deeper image in the presence of gas versus 3D streamer data. (Source: Granli et al. 1999)
Fig. 17.16 Land seismic acquisition
17.4.6 Permanent Seismic Monitoring compared to conventional OBS 4D survey. Such a

method is important in monitoring a reservoir injec-
Permanent seismic monitoring is becoming an impor- tion process employed to enhance recovery from a
tant tool in the reservoir management toolkit. It is a 4C producing reservoir.
fiber-optic advanced seismic acquisition technology
that is installed permanently on the seabed over a
producing field (Fig. 17.18). It reduces acquisition 17.5 Seismic Processing
time and cost. Permanent seismic monitoring helps to
improve data quality by employing more accurate Seismic technology has achieved amazing feats in
survey orientation and acquisition geometry (receiver exploration and production activities in the past few
locations) within the repeated 3D seismic surveys decades. What we record in the acquisition stage is
then be processed by employing advanced methods

within signal processing and wave-theory to get better
images of the subsurface. The prime objective in the
processing stage is to enhance the signal and suppress
the coherent and noncoherent noises and multiples
(Fig. 17.19). Coherent noise is unwanted seismic
energy that shows consistent phase from one seismic
trace to another. This may consist of waves that
travel through the air at very low velocities such as
airwaves or air blast, and ground roll that travels
through the top of the surface layer, also known as the
Layer 1 weathering layer. Multiples are internal reflections in a
layer, which occur when exceptionally large reflection
Layer 2 coefficients are present. In marine seismic the water-
bottom multiples normally dominate. Noncoherent
energy is typically nonseismic-generated noise, such
Layer 3 as noise from wind, moving vehicles, overhead power
line or high-voltage pickup, gas flares and water injec-
tion plants.
It has been stated earlier that seismic processing
is the alteration of seismic data to suppress noise,
Fig. 17.17 Acquisition of VSP. The downhole geophones
record important structural and stratigraphic data generated by enhance signal and migrate seismic events to the
a surface energy source appropriate location in space. Seismic processing
facilitates better interpretation because subsurface
structures and reflection geometries are more appar-
ent. The typical sampling rate of seismic acquisition
is 2 ms. Digital recording of the incoming wavefield
at densely spaced receiver positions ensures that the
recorded signal and noise are properly sampled and
are therefore unaliased. Aliasing is the ambiguity
that arises because of insufficient sampling. It
occurs when the signal is sampled less than twice the
cycle. The highest frequency defined by a sampling
interval is termed the Nyquist frequency and is equal
to the inverse of 2Δt, where Δt is the sampling
interval. Frequencies higher than the Nyquist fre-
quency will be folded back. In the noise-free case,
aliasing can be avoided by a finer spatial sampling
that is at least twice the Nyquist frequency of the
waveform.
It is important to note that no standard processing
sequence exists which can be routinely applied to
Fig. 17.18 Permanent installation of 4C cables on the sea all types of raw seismic data. The actual sequence
bottom over a producing field. It improves data quality by will be determined by (a) the purpose of the investiga-
ensuring more accurate receiver locations within the repeated tion, (b) extensive testing on selected parts of the
3D surveys over a period of time
dataset and (c) a trade-off between quality and cost.
The 2D seismic processing steps typically include
called raw seismic data which contains real signals static corrections, deconvolution, velocity analysis,
together with noise and multiples. This raw data must normal and dip moveout, stacking and migration.
440 N.H. Mondol
Facility noise Vehicle
energy
erating
Reverb
s
a ve Ha
ew rm
rfac on
ic
Su en
er
gy
t
s
r bla
Ai
Before DGF
Fig. 17.19 Raw seismic data with coherent and noncoherent noise (top). Noise attenuation image after autocorrelation,
deconvolution and trace muting (bottom). (Courtesy Western Geco)
The following routines are generally applied to raw 2D In the processing stage, bad measurements are
seismic data in different processing stages: edited, datuming applied and corrections of wave-
• True Amplitude Recovery (TAR) energy decay introduced. The true amplitude recovery
• Autocorrelation is applied to increase the amplitude at large travel
• CMP sorting times. The autocorrelation and deconvolution are
• Deconvolution done to compress the wavelet and to attenuate
• Trace Muting multiples. Deconvolution – a technique that can com-
• Velocity picking or velocity analysis press the source signature and eliminate multiples – is
• NMO correction applied after sorting the data into CMP gathers. The
• DMO correction trace muting is applied to get rid of unwanted energy.
• Filtering (F-K and Bandpass filtering) Contributions from the direct waves and possible head
• Stacking and waves are removed by trace muting. NMO correction
• Migration and F-K filtering are usually applied to attenuate
Fig. 17.20 (a) A CMP-gather before the F-K filtering: the respectively. (b) The same CMP gather after F-K filtering. The
primaries dipping up and the multiples dipping down in a F-K filtering accepted only primary energy (within polygon) and
time-distance display. The F-K domain (top, right) shows filtered out multiples energy (bottom, right)
energy distributions of both primary and multiples energy,
multiples. Linear coherent noises are also removed by are interpolated for each CDP. Normal moveout is
employing F-K filtering (Fig. 17.20). applied according to the following formula:
NMO correction is applied from a space-variant
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
velocity field assuming a horizontal reflector and hyper- u" 2 #
u X
bolic normal moveout algorithm. The NMO is the dif- T ð XÞ ¼ t T 2 ð0Þ þ (17.5)
ference between the travel time for a certain offset (X) V
and the vertical (zero-offset) traveltime T(0). Velocities
442 N.H. Mondol
a b
Fig. 17.21 NMO corrected CDP gathers show NMO stretch (a) and stretch muting (b) at the far offsets. Muting to remove NMO
stretch may destroy far offsets information
where T(X) is the two-way travel time for a seismic Horizontal reflector
event, X is the actual source-receiver offset distance, V

is the NMO or stacking velocity for this reflection event
and T(0) is the two-way travel time for zero offset. A
sample-by-sample velocity is built at each of the
locations where time-velocity pairs are defined. For
any point before the first velocity location or beyond
the last location, the first or last velocity function is Dipping reflector
used. Once the correct velocity function has been
interpolated, the exact moveout at each sample is
computed based on the actual source-to-receiver offset
and velocity at that time sample. NMO stretch is a
fundamental and long-standing problem in seismic
processing. After normal moveout correction the early Shotpoint
events are stretched at the far offsets (Fig. 17.21a). If we Receiver
stack this unmuted gather, the early events suffer a
severe loss of high-frequency energy, and thus resolu- Fig. 17.22 Effect of reflector dip on the reflection point. When
the reflector is flat (top) the CMP is a common reflection point.
tion. This can appreciably reduce the interpretability of
When the reflector dips (bottom) there is no CMP. A dipping
the section. There have been many attempts to solve the reflector may require changes in survey parameters, because
NMO stretch problem. The most universal is front-end reflections may involve more distant sources and receivers
or stretch muting, where samples at the beginning of a than reflection from a flat layer
trace that have suffered severe NMO stretch are zeroed
out (Fig. 17.21b). Stretch muting may leave very little In the case of dipping beds, there is no common
fold at early times, reducing the noise suppression depth point shared by multiple sources and receivers,
provided by stacking. so dip-moveout (DMO) processing becomes necessary
to reduce smearing or inappropriate mixing of data is assumed that drastic lateral velocity changes do not
(Fig. 17.22). DMO is not routinely employed; how- appear. Then we can interpret high semblance values
ever, it can be useful to solve the problem of to be usually related to primaries and low semblance
conflicting dip. values to be related to multiples (Fig. 17.23). If the
In practice, a velocity analysis has to be carried out, velocity does not increase with depth in some areas it
where the main purpose is to determine the velocity is much more complicated to distinguish between
distribution as a function of time which will give the primaries and multiples. If these basic assumptions
most accurate NMO correction. The velocity analysis of velocity analysis are violated both the velocity
is used to compute corrections in traveltime that will analysis and the stacked section can be distorted.
be applied to all the traces belonging to a CMP-gather. Such velocity anomalies can be caused by:
The most common type of velocity analysis is to • Diffractions (edges, faults)
repeat the procedure of correcting and stacking CMP • Shallow gas pockets
data for many different velocities and within a discrete • Multiples
time window. By measuring the average absolute • Complicated reflection sequence (interference)
value of the data or more precisely the semblance • Dip
within time-windows of different test velocities, and • Reflection from the side (out-of-plane).
plotting these results in a time versus velocity histo- Stacking is an important step in seismic processing.
gram, it is possible to interpret the velocity informa- Stacking represents summation of NMO-corrected
tion. This type of plot is denoted a velocity spectrum. traces in a CMP family. The collection of stacked
In general, the interpreted velocity function is picked traces forms a seismic section which gives an image
so that it goes through areas with highest values. The (slice) of the subsurface (Fig. 17.24b). The stacking
basic assumptions in the CMP method velocity analy- process has two major advantages: (a) it increases the
sis is that the geological model corresponds to a slowly signal-to-noise (S/N) ratio and (b) it amplifies primary
varying velocity as a function of depth. In addition, it energy relative to multiple energy. This second point
Fig. 17.23 An example of velocity analysis. A velocity spec- function (right). By picking maximum values from different
trum or the semblance histogram (left). Plot of velocity analysed panels a time-velocity function can be constructed
CMP data (middle) and mini-stacks based on picked velocity
444 N.H. Mondol
0 0
500 500
1000 1000
Depth (m)
Depth (m)
1500 1500
2000 2000
2500 2500
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 0 1000 2000 3000 4000 5000 6000 7000 8000 9000
Offset (m) Offset (m)
Fig. 17.24 A velocity model (left) and a stacked seismic section (right). (Source: Versteeg 1994)
depends on a good velocity analysis. In the case of an zero, and subsequently the data is recorded at the sur-
accurate velocity model (Fig. 17.24a), stacking is the face. However, the exploding reflector model does not
most efficient multiple removal method. account for reflections for which the corresponding
Migration is the process that reverses wave propa- down-going ray path differs from the up-going ray
gation effects to get clear images of the subsurface. path. In order to produce correct travel times with this
Migration processing is extremely important in geo- concept, all the velocities in the subsurface need to be
logically complex areas. The term migration came halved with respect to their actual values. The
about because, compared to stack sections, the echoes advantages of seismic migration are:
“migrate” to their true subsurface position. Migration • Diffractions are collapsed to points
is used for several reasons; the most important one is • Deeping reflections move up dip and become
to move reflectors from seismic “apparent” position to steeper
their geological “true” position. Another reason for • Triplications (bow ties) associated with synforms
doing migration is to collapse and focus diffractions. are unwrapped
Since we cannot measure the direction from where the • Crossing reflectors avoided.
signal has been reflected we assume zero offset. It is Seismic migrations are of four types: Pre-stack time
not correct to assume zero offset where the underlying and Pre-stack depth migration and Post-stack time or
Earth layer is not plane. Before doing the migration a Post-stack depth migration (Fig. 17.25). In time migra-
good velocity model is needed as an input. tion the images are displayed in two-way travel times,
The migration procedure consists of two well- and wave field extrapolation is done in a time-stepping
defined steps: downward extrapolation and imaging. way (Fig. 17.26a). In depth migration the wave-
Downward extrapolation is a simulation process stepping is done with respect to depth, and the images
(based on the one-way wave-equation), where the can be represented in a true vertical depth (Fig. 17.26b).
receivers are moved from the surface down to an arbi- Depth migration can handle strong lateral velocity
trary depth. For each downward extrapolated depth a variations. Time migration is most common in practice.
new seismic section can be formed (imaging). From
each of these sections only the part close to t ¼ 0 is
retained. A migrated section is now formed by combin-
ing these strips. The exploding reflector model or ER- 17.6 Seismic Resolution
model (Lowenthal et al. 1977) has also been used for
migration of seismic data. According to this model, a Seismic resolution is the ability to distinguish separate
zero offset seismic section can be produced, where all features; the minimum distance between 2 features so
the seismic reflectors “explode” simultaneously at time that the two can be defined separately rather than as
a b
5
Time [s]
3 km 3 km
6
Fig. 17.25 Comparison of time domain images from (a) Pre-stack time migration and (b) Post-stack time migration
one. The limit of seismic resolution usually makes us receiver, the Fresnel zone will be circular and with a
wonder, how thin a bed can we see? Normally we radius of Rf is
think of resolution in the vertical sense, but there is
rffiffiffiffiffi
λZ
also a limit to the horizontal width of an object that we Rf ¼ (17.6)
2
can interpret from seismic data.
where λ is the dominant wavelength and Z is the depth
down to the target surface. Horizontal resolution
17.6.1 Horizontal Resolution depends on the frequency and velocity of seismic
waves. If we introduce the centre frequency fc of the
The horizontal dimension of seismic resolution is pulse (i.e. representing the most energetic part), we
described by the Fresnel zone. The Fresnel zone is a have λ V=fc , with V being the wave velocity. Hence,
frequency and range dependent area of a reflector from we can rewrite the formula for the Fresnel zone as
which most of the energy of a reflection is returned and sffiffiffiffiffiffi
arrival times differ by less than half a period from the VZ
Rf ¼ (17.7)
first break (Fig. 17.27). Waves with such arrival times 2fc
will interfere constructively and so be detected as a
single arrival. Subsurface features smaller than the
Fresnel zone usually cannot be detected using seismic
waves. At spacing greater than one-quarter of the 17.6.2 Vertical Resolution
wavelength, the event begins to be resolvable as two
separate events. Migration can improve lateral resolu- Vertical resolution is the ability to separate two
tion by reducing the size of the Fresnel zone. For a features that are close together. A seismic wave can
plane reflecting interface and coincident source and be considered as a propagating energy pulse. If such a
446 N.H. Mondol
Fig. 17.26 A comparison of Pre-stack time migrated (top) and Post-stack depth migrated (bottom) images
S
The typical recorded seismic frequencies are in the
range of 5–100 Hz. High frequency and short
Z+
λ /4
Z wavelengths provide better vertical and lateral reso-

lution. One could argue that we could simply increase
the power of our source so that high frequencies
A A′
could travel farther without being attenuated. How-
Fig. 17.27 A Fresnel zone in 3D seismic is circular and has ever, there is a practical limitation in generating high
diameter A–A´ where S is the source position, Z is the depth frequencies that can penetrate large depths. The Earth
down to the target and λ is the wavelength. The size of the acts as a natural filter removing the higher
Fresnel zone helps to determine the minimum size feature that
can be seen in a seismic section frequencies more readily than the lower frequencies
(absorption effect). This means the deeper the source
of reflections, the lower the frequencies we can
wave is being reflected from the top and the bottom of receive from those depths and therefore the lower
a bed, the result will depend on the interaction of resolution we appear to have from great depths
closely spaced pulses. In order for two nearby reflec- (Fig. 17.28). The vertical resolution decreases with
tive interfaces to be distinguished well, they have to be the distance travelled (hence depth) by the ray
about λ/4 in thickness which is called the tuning thick- because attenuation preferentially robs the signal of
ness. This is also the thickness where interpretation the higher frequency components. Deconvolution can
criteria change. For smaller thickness, the limit of improve vertical resolution by producing a broad
visibility is reached and positional uncertainties are bandwidth with high frequencies and a relatively
introduced. compressed wavelet.
0–10 Hz 10–20 Hz 20–30 Hz 30–40 Hz 40–50 Hz

0
1.0
Time, sec
2.0
3.0
4.0
Fig. 17.28 Filtered seismic data showing frequency content depth of penetration of seismic energy decreases. Lower
variation with depth. Each panel has been filtered to allow a frequencies (left) penetrate deeper. Higher frequencies (right) do
different band of frequencies. As the bandpass rises, the maximum not penetrate to deeper levels. (Source: Ashton et al. 1994)
As an example, if we introduce the centre fre- will be based on an integrated use of seismic inlines,
quency fc of the energy pulse (disturbance) we obtain crosslines, time slices and horizon attributes (Dalley
the following simple relationship between the domi- et al. 1989, Hesthammer et al. 2001). The seismic
nant wavelength (λ), the wave velocity (V) and the sections or images represent slices through the geolog-
centre frequency (fc): ical model, which can be input to advanced
workstations where the actual interpretation can take
V place. Seismic data can be used in many ways such as
λffi (17.8)
fc regional mapping, prospect mapping, reservoir delin-
eation, seismic modelling, direct hydrocarbon detec-
The typical values for the dominant wavelength are tion and the monitoring of producing reservoirs. Based
then (a) λ ¼ 40 m at shallow depth (upper 300–500 m on the seismic interpretation one will decide if an area
depth), where V ¼ 2; 000 m=s and f ¼ 50 Hz, (b) λ ¼ is a possible prospect for hydrocarbon (oil or gas). If
100 m at intermediate depths (about 3,500 m), where the answer is positive, an exploration well will be
V ¼ 3; 500 m=s and f ¼ 35 Hz and (c) λ ¼ 250 m at drilled. The ultimate goal will be the drilling of pro-
depths of about 5,000 m), where V ¼ 5; 000 m=s and duction wells if the target area proves to be a commer-
f ¼ 20 Hz. For smaller thicknesses than λ/4 we rely on cial reservoir. Seismic data contain a mixture of signal
the amplitude to judge the bed thickness. For and noise. It is therefore crucial to understand the
thicknesses larger than λ/4 we can use the waveform. signature of the noise, whether it is systematic or
random, dipping or flat-lying, planar or non-planar. It
is also necessary to investigate the origin of the noise.
17.7 Seismic Interpretation The challenge of seismic interpretation is then to fully
utilise all the information contained in the seismic
Seismic data are studied by geoscientists to interpret data. Systematic noise can be related to acquisition
the composition, fluid content, extent and geometry of procedures, processing artefacts, water-layer
rocks in the subsurface. Interpretation of seismic data multiples, faults, complex stratigraphy and shallow
448 N.H. Mondol
gas. Random noise includes natural noise (e.g. wind continuity of the events on a seismic section. The
and wave motion), incoherent seismic interface and Instantaneous phase is on a scale of +180 to
imperfect static corrections. Without a sound under- 180 . The temporal rate of change of the instan-
standing of these factors as well as knowledge of the taneous phase is the instantaneous frequency.
limitation of seismic resolution, there is a danger of
misinterpreting noise as real features.
In terms of the parameters that are analysed and the
17.7.1 Seismic Facies Analysis
interpretation that may be drawn from the analyses, a
fourfold hierarchy of seismic interpretation can be
A seismic profile provides information about the
achieved. These are seismic facies analysis, seismic
properties of sedimentary rocks. Seismic facies analy-
structural analysis, seismic attribute analysis and seis-
sis is the description and geological interpretation of
mic sequence analysis. The most important parameters
seismic reflectors representing sequence boundaries. It
used for interpretation of seismic data are:
includes the analysis of such parameters as the config-
Reflection amplitudes: The strength of the reflections.
uration, continuity, amplitude, phase, frequency and
As we discussed above, the proportion of the
interval velocity. These variables give an indication of
energy reflected at the boundary between two
the lithology and sedimentary environment of the
beds is a function of the difference in the acoustic
facies (Selley 1998) (Fig. 17.29). Because seismic
impedances. If we have an alternating series of
reflections mainly represent time lines, i.e. sedimen-
different beds, the distance between the bed
tary beds which were deposited contemporaneously, it
boundaries in relation to the wavelength of the
is also possible to a certain extent to interpret the
transmitted seismic signals will play a major part.
depositional environment. Large-scale sedimentary
Reflector spacing: The distance between the reflectors
features that may be recognised by the configuration
will indicate the thickness of the bed, but there will
of seismic reflectors include prograding deltas, car-
be a lower limit to the thickness that can be
bonate shelf margins and submarine fans (Sherif
detected, which will depend on the wavelength.
1976).
Interval velocity: The interval velocity of a sequence
can provide information about lithology and poros-
ity but this will depend on the stacking velocity and
will not be very accurate. 17.7.2 Structural Analysis
Reflector continuity: The continuity of reflectors will
be a function of how continuous the sediment beds Interpreters can draw horizons and faults on in-lines,
are, information which is essential for cross-lines and arbitrary lines, as well as slices.
reconstructing the environment. Horizons can be automatically tracked on vertical
Reflector configuration: If we take the compaction seismic displays and horizontal slice displays.
effect into account, the shape of the reflecting Improved tracking algorithms for horizon interpreta-
beds gives us a picture of the sedimentation surface tion, combined with user-interpreted faults and fault
as it was during deposition. The slope of the polygons, can produce seismic-based interpretation
reflectors, for example, represents the slope of maps (Fig. 17.30). Faults on seismic sections are typi-
prograding beds in a delta sequence with later dif- cally expressed by a loss of reflection amplitude. It
ferential compaction and tilting superimposed. Ero- must be remembered that the principle we use for
sion boundaries with unconformities will in the calculating the depth to a reflecting boundary assumes
same way show the palaeo-topography during that the layering is not too far from horizontal. High-
erosion. angle faults do not reflect the sound wave back to the
Instantaneous phase: A seismic trace can be consid- receivers to give a signal and faults can therefore not
ered an analytical signal where the real part is the be seen directly on seismic sections. It is the
recorded seismic signal itself. Mathematically we truncations and offset of good reflectors which usually
can compute the complex seismic trace (imaginary allows us to identify faults. Due to special “edge
parts of the signal) and the instantaneous attributes. effects” that we observe near a fault, the reflector
The Instantaneous phase is a measure of the terminations which should define the fault are not
Shale Target reservoir Massive sands
Fig. 17.29 A time slice of a special seismic attribute (S- (green) juxtaposed against the organic-rich shales (purple) are
impedance co-rendered with Kmax curvature) that is constructed the most productive reservoirs in this horizon. (Courtesy ION
for lithology determination. Fractured, interbedded sandstones Geophysical)
located in quite their true position on the seismic

profile, making it difficult sometimes to define them
accurately. Migration of the seismic data goes some
way to remedying the problem. Folded beds will only
be realistically depicted if the folds are sufficiently
gentle that the beds have a low angle of dip.
17.7.3 Seismic Attribute Analysis
A seismic attribute is a quantitative measure of a

seismic characteristic of interest. Seismic attribute
analysis is concerned with the study of amplitude,
polarity, continuity and wave shape. One of the goals
of seismic attributes is to somehow capture maximum
information by quantifying the amplitude and morpho-
logical features seen in the seismic data through a suite
of deterministic calculations performed on a com-
puter. The extraction of seismic attributes, such as
amplitude envelope, dominant frequency, apparent
polarity and instantaneous phase can produce remark-
Fig. 17.30 A seismic section showing abundant faulting. able 3D images of subsurface rock formations
Horizons can be interpreted by variable amplitude and pick
trough-pick trace shape (top). A plan view of a horizon (bottom)
(Fig. 17.31). Such analysis may give an indication of
which can be useful to identify fault patterns from the missing the thickness and nature of the upper and lower
picks and discontinuities in colour. (Source: James 2003) contacts of a sand body (Selley 1998). Comparison
450 N.H. Mondol
a 10000
No
rth
5000
5000
10000
Time (ms)
1100
1200
0.0 1.0 1 km
10000
b No
rth
5000
5000
10000
Channel/levee deposits
Time (ms)
1100 Fig. 17.32 Interpreted 3D seismic volume showing a channel

1200 fan complex with a shelf edge. (Source: James 2009)
geometric, kinematic, dynamic or statistical features

derived from seismic data. They include reflector
0.0 1.0 1 km amplitude, reflector time, reflector dip and azimuth,
complex amplitude and frequency, generalised Hilbert
Fig. 17.31 A comparison between average absolute amplitude attributes, illumination, edge detection/coherence,
(a) and energy (b) in a horizon slice at the same stratigraphic AVO and spectral decomposition. General attributes
level. Notice that the channel/levee deposits can be recognised, are based on either the physical or morphological
mapped and detected more efficiently from the energy volume
character of the data tied to lithology or geology and
than from the amplitude volume. (After Gao 2003)
are therefore generally applicable from basin to basin
around the world. In contrast, specific attributes have a
of observed seismic waves with synthetic traces less well-defined basis in physics or geology. While a
computed from a geological model give some insight given specific attribute may be well correlated to a
into the depositional environment of the sand, and geological feature or to reservoir productivity within a
hence help to predict its geometry and internal reser- given basin, these correlations do not in general carry
voir characteristics (Fig. 17.32). over to a different basin.
There are now more than 100 distinct seismic Over the past decades, we have witnessed attribute
attributes calculated from seismic data that can be developments tracking the breakthroughs in reflector
applied to the interpretation of geological structures, acquisition and mapping, fault identification,
stratigraphy, and rock/pore fluid properties. Taner bright-spot identification, frequency loss, thin-bed
et al. (1994) divide attributes into two general tuning, seismic stratigraphy and geomorphology
categories: geometrical and physical. The objective (Fig. 17.33). More recently, interpreters have used
of geometrical attributes is to enhance the visibility crossplotting to identify clusters of attributes
of the geometrical characteristics of seismic data; they associated with either stratigraphic or hydrocarbon
include dip, azimuth, and continuity. Physical anomalies. Today, very powerful computer
attributes have to do with the physical parameters of workstations capable of integrating large volumes of
the subsurface and so relate to lithology. These include diverse data and calculating numerous seismic
amplitude, phase and frequency. Liner et al. (2004) attributes are a routine tool used by seismic
classified attributes into general and specific interpreters seeking geological and reservoir engineer-
categories. The general attributes are measures of ing information from seismic data.
a b
Probable
100 ms
gas
Gas
GOC
OWC OWC
Oil Oil
500 m 500 m
Fig. 17.33 (a) A strong soft amplitude anomaly (oil column) at column) within the reservoir with a sharp, flat base. (After
top reservoir corresponds to a clear flat event within the reser- Blom and Bacon 2009)
voir. (b) This section shows an acoustically soft body (gas
Fig. 17.34 Geological interpretation of a seismic section
17.7.4 Seismic Sequence Analysis depth will cause further subsidence, and sedimentation
will increase the load, resulting in further subsidence
Apart from structural and facies analysis, seismic to attain isostatic equilibrium. Local tectonic subsi-
signatures can be used to interpret the way the basin dence may produce a relative change in coastal onlap
has been filled in. This is a result of an interaction in a seismic profile. Regressions are defined as the
between the rate of subsidence, rate of deposition and boundary between land and sea being displaced out
the energy of the depositional environment into the basin. They may be caused by a fall in sea
(Fig. 17.34). level which will shift the coastline to further out on the
On seismic profiles we can see onlaps onto the land, shelf or to the edge of the continental slope, or
measure the height range between the lowest and progradation of a coastline or a delta front. Here
uppermost onlaps, calculate the difference in seismic unloading of some of the water load and erosion of
time, and convert this into approximate thickness. sediments led to isostatic uplift of the area landward of
However, we must remember that the thickness of the coastline, so that the measured regression is greater
the sediments deposited is due not only to a rise in than the real lowering of the sea level. Here, too, local
eustatic sea level, but also to local subsidence of this tectonic movements will play a part. Transgressions
part of the basin. The weight due to increased water (geological events during which sea level rises relative
452 N.H. Mondol
to the land and the shoreline moves toward higher 17.8 Integration and Visualisation of
ground, resulting in flooding) and regressions are not Seismic Data
always directly related to sea level changes. When a
delta builds out into the sea, there is a local regression In order to establish a reservoir model, different
on the delta even if the sea level has not fallen. If geophysical measurements such as seismic, well
sedimentation is sufficiently rapid, we can have a logs and rock physics need to be integrated. Since
local regression on a delta even with a rising sea these measurements are carried out employing
level. Along a coastline we can in fact have very different frequency regimes, there is a problem
transgressions in some areas and regressions in others, of scaling. Rock physics models of laboratory
at the same time, depending on the rate of sedimenta- measurements use ultrasound frequencies (kHz to
tion or erosion with respect to sea level change. A MHz) whereas sonic tools (well logging) operate
regression caused by a fall in sea level is called a with frequencies in the kHz range. Seismic data fall
forced regression. Changes in sea level can be due to: within the range of hertz (Hz). Assuming the velocity
• Local tectonic movements, for example uplift of a of a sedimentary layer is 3,000 m/s, a seismic
horst or subsidence of a graben structure. frequency of 30 Hz will give 100 m wavelength
• Plate-tectonic movements which can be of great whereas sonic log frequency of 30 kHz will give
extent, but are not global. 10 cm wavelength and the ultrasound frequency of
• Sea level changes. These are global and are called 300 kHz will give 1 cm wavelength. Therefore, the
eustatic sea level changes. vertical resolutions of seismic, sonic and ultrasonic
During the Quaternary, cyclic changes in sea level frequencies are 25 m, 2.5 cm and 2.5 mm, respec-
of up to 120 m accompanied the growth and decay of tively (Fig. 17.35). The major problem is then how to
continental ice sheets. These changes were rapid and best incorporate all geophysical data into a reservoir
of large magnitude and can be traced throughout much model. Rock physics investigations define how pore-
of the world. In the areas which had supported ice scale variations in properties like mineralogy, fluid
sheets, such as Scandinavia, the melting of the ice content and grain geometry affect the acoustic
led to isostatic uplift due to unloading, and this response of a core sample. As the scale of investi-
exceeded the rise in sea level so that there was a gations increases from laboratory to well log to seis-
regression. When seismic stratigraphy was established mic, low frequency measurements (seismic or well
(Vail et al. 1977), it was assumed that most of the logs) average over different rock types and sediments
variations in sea level that could be interpreted from in addition to local pore-scale variations. As a result,
seismic records were attributable to eustatic changes spatial heterogeneity and preferential sampling at
and thus could be employed for global correlation. well log or seismic scales can cause a shift in the
However accumulation of ice on the continents is the rock physics relationship away from that determined
only known process capable of producing large and in the lab. Therefore, critical understanding and
rapid global sea level changes. Such ice ages are strategies are necessary to incorporate seismic, well
known from the Quaternary, the Carboniferous- logs and rock physics to find links between qualita-
Permian, late Ordovician and the end of the tive geological parameters and quantitative geophys-
Precambrian. ical measurements. Integration strategies and
Nevertheless, transgressions and regressions can be utilisation of different geophysical data are discussed
correlated over greater or lesser distances, depending thoroughly in Chap. 18.
on the type of tectonic displacement. It is often diffi- As the use of 3D seismic increasingly becomes an
cult to distinguish between eustatic sea level changes integral part of hydrocarbon exploration, visualisation
and ones caused by more local or regional conditions techniques also continue to evolve as software and
and we therefore prefer to use the term relative sea hardware improves. Visualisation of target datasets
level change. On a delta, one delta lobe may prograde of various sizes and formats in a powerful graphical
and produce a regressive sequence while a transgres- environment is the answer to efficiently understanding
sion occurs with carbonate sedimentation on an adja- data quality problems, anomalies and trends. Such an
cent lobe. environment responds dynamically to manipulation
2D seismic
100
Resolution, m 3D seismic
10
VSP
Crosswell
1.0 Sonic seismic
imaging
Sonic
logging
0.1
Ultrasonic
0
0.1 1.0 10 100 1000
Range, m
Fig. 17.35 Plot of range versus resolution of various geophysical techniques. (Courtesy: Schlumberger Oilfield Glossary)
Fig. 17.36 Shows a 3D volume with fault (red), horizons (amplitude maps) and a producing well path through a channel. Once a
channel is identified, its shape can be extracted by visualising the 3D volume. (Courtesy: Halliburton)
when analysing many datasets on one common can- model (Fig. 17.36). 3D visualisation technique is also
vas, or it can allow travel through time in a 4D used to estimate thickness, continuity and lateral
visualisation process comparing one survey to another. extend of a reservoir and its relationship with horizons
Advance visualisation techniques allow us to interac- and faults to see the finest reservoir details.
tively explore regional datasets, integrate live well Visualising combinations of complex seismic
information into a reservoir model and understand attributes like amplitude, continuity and AVO (Ampli-
reservoir dynamics in association with a 4D predictive tude Versus Offset) provide a far more accurate
454 N.H. Mondol
picture of hydrocarbon potential, and in less time, than James, H. 2003. Has volume interpretation of structure been
viewing those attributes discretely. By integration and cracked at last? First Break 21, 38–40.
James, H. 2009. Visualizing 3D features in 3D seismic data.
3D visualisation of massive amounts of data one can First Break 27, 57–62.
add greater value to decision making to differentiate Liner, C., Li, C.-F., Gersztenkorn, A. and Smythe, J. 2004.
between economic and sub-economic wells in a mar- SPICE: A new general seismic attribute: 72nd Annual Inter-
ginal area. The ability to understand data, gain useful national Meeting, SEG, Expanded Abstracts, 433–436.
Loewenthal, D., Lu, L., Roberson, R. and Sherwood, J. 1976.
insight, and communicate these with others is greatly The wave equation applied to migration. Geophysical
enhanced by the use of interactive 3D visualisation. Prospecting 24, 380–399.
However, the ability to adjust the parameters and the Milligan, M. 2004. What are seismic surveys and how much
display interactively is crucial to exploring the data shaking do they create? Survey Notes, Utah Geological
Survey 36(3), 10–11.
and finding the combinations that highlight specific Ongkiehong, L. and Askin, H.J. 1988. Towards the universal
features and relationships. seismic acquisition technique. First Break 6(2), 46–63.
Rygg, E., Riste, P., Nottvedt, A., Rod, K. and Kristoffersen, Y.
1992. The Snowstreamer - a new device for acquisition of
seismic data on land. In: Vorren, T.O., Bergsager, E., Dahl-
Further Reading Stamnes, O.A., Holter, E., Johanses, B., Lie, E. and Lund, T.
B. (eds.), Arctic Geology and Petroleum Potential. NPF
Special Publication 2, 703–709.
Aronsen, H.A., Osdal, B., Dahl, T., Eiken, O., Goto, R., Selley, R.C. 1998. Elements of Petroleum Geology. Academic
Khazanehdari, J., Pickering, S. and Smith, P. 2004. Time Press, San Diego, CA, 470 pp.
will tell: New insights from time-lapse seismic data. Oilfield Sheriff, R.E. 1976. Inferring stratigraphy from seismic data.
Review 16(2), 6–15. AAPG Bulletin 60, 528–542.
Ashton, C.P., Bacon, B., Mann, A., Moldoveanu, N., Déplanté, Sheriff, R.E. and Geldart, L.P. 1995. Exploration Seismology.
C., Dickilreson, T.S. and Redekop, G. 1994. 3D seismic Cambridge University Press, New York, 592 pp.
survey design. Oilfield Review 6, 19–32. Taner, M.T., Schuelke, J.S., O’Doherty, R. and Baysal, E. 1994.
Beckett, C., Brooks, T., Parker, G., Bjoroy, R., Pajot, D., Taylor, Seismic attributes revisited: 64th Annual International
P., Deitz, D., Flaten, T., Jaarvik, L.J., Jack, I., Nunn, K., Meeting, SEG, Expanded Abstracts, 1104–1106.
Strudley, A. and Walker, R. 1995. Reducing 3D seismic Telford, W.M., Geldart, L.P. and Sheriff, R.E. 1990. Applied
turnaround. Oilfield Review 7(1), 23–37. Geophysics. Cambridge University Press, New York, 770 pp.
Blob, F. and Bacon, M. 2009. Application of direct hydrocarbon Thompson, M., Arntsen, B. and Amundsen, L. 2007. Full azi-
indicators for exploration in a Permian-Triassic play, off- muth imaging through consistent application of ocean bot-
shore the Netherlands. First Break 27, 37–44. tom seismic. 77th Annual International Meeting, SEG,
Braile, L.W. 2000. Seismic Waves and the Slinky: A Guide for Expanded Abstracts, 936–940.
Teachers. http://www.gg.uwyo.edu/geol2005a/lectures/ Tucker, P.M. and Yorston, H.J. 1973. Pitfalls in seismic inter-
seismics/WaveDemo.pdf pretation. Society of Exploration Geophysicists, Monograph
Dalley, R.M., Gevers, E.C.A., Stampfli, G.M., Davis, D.J., Series, No. 2, 50 pp.
Gastaldi, C.N., Ruijtenberg, P.A. and Vermeet, G.J.O. Vail, P.R., Mitchum, R.M. and Thomson, S. 1977. Seismic
1989. Dip and azimuth displays for 3D seismic interpreta- stratigraphy and global changes of sea level, Part 3: Relative
tion. First Break 7, 86–95. changes of sea level from coastal onlap. In: Payton, C.E.
Gao, D. 2003. Volume texture extraction for 3D seismic visuali- (ed.), Seismic Stratigraphy – Applications to Hydrocarbon
zation and interpretation. Geophysics 68, 1294–1302. Exploration. AAPG Memor 26, 63–97.
Granli, J.R., Arnsten, B., Anders, S. and Hilda, E. 1999. Imaging Versteeg, R. 1994. The Marmousi experience: Velocity model
through gas-filled sediments using marine shear-wave data. determination on a synthetic complex dataset. The Leading
Geophysics 64, 668–667. Edge 13, 927–936.
Hesthammer, J., Landrø, M. and Fossen, F. 2001. Use and abuse Yilmaz, O. 2001. Seismic Data Analysis: Processing, Inversion
of seismic data in reservoir characterisation. Marine and and Interpretation of Seismic Data. SEG Press, Tulsa,
Petroleum Geology 18, 635–655. 2027 pp.
Chapter 18
Explorational Rock Physics: The Link between Geological

Processes and Geophysical Observables
Per Avseth
The field of rock physics represents the link between 18.1 Quantitative Seismic Interpretation
qualitative geological parameters and quantitative Using Rock Physics
geophysical measurements. Increasingly over the last
decade, rock physics has become an integral part of The main goal of conventional, qualitative seismic
quantitative seismic interpretation and stands out as a interpretation is to recognise and map geological ele-
key technology in petroleum geophysics. Ultimately, ments and/or stratigraphic patterns from seismic reflec-
the application of rock physics tools can reduce explo- tion data. Often hydrocarbon prospects have been
ration risk and improve reservoir forecasting in the defined and drilled entirely on the basis of this qualita-
petroleum industry. tive information. Today, however, quantitative seismic
This chapter covers basic rock physics principles and interpretation techniques have become common oil
practical recipes that can be applied in the field. The industry tools for prospect evaluation and reservoir
importance and benefit of linking rock physics to geo- characterisation. The most important of these techni-
logical processes, including depositional and compac- ques include post-stack amplitude analysis (bright-spot
tional trends, is demonstrated. It is further documented and dim-spot analysis), offset-dependent amplitude
that lithology substitution can be of equal importance to analysis (AVO analysis), acoustic and elastic imped-
fluid substitution during seismic reservoir prediction. It ance inversion, and forward seismic modelling. These
is essential in exploration and appraisal to be able to techniques seek to extract additional information about
extrapolate away from existing wells, taking into the subsurface rocks and their pore fluids from the
account how the depositional environment changes, reflection amplitudes and, if used properly, they open
together with burial depth trends. In this way rock up new doors for the seismic interpreter. Seismic ampli-
physics can better constrain the geophysical inversion tudes primarily represent contrasts in elastic properties
and classification problem in underexplored marginal between individual layers and contain information
fields, surrounding satellite areas, or in new frontiers. about lithology, porosity, pore fluid type and saturation,
Finally, practical examples and case studies are pre- as well as pore pressure – information that cannot be
sented to demonstrate a best-practice workflow and gained from conventional seismic interpretation. Seis-
associated limitations and pitfalls. Rock physics models mic amplitude maps are increasingly important in pros-
are combined with well log and pre-stack seismic data, pect evaluation and reservoir delineation. As shown in
sedimentological information, inputs from basin mod- Fig. 18.1, the amplitude patterns often provide a good
elling and statistical techniques, to predict reservoir insight into depositional patterns. Seismic amplitude
geology and fluids from seismic amplitudes. maps can be very useful in the delineation of subtle
traps that are not easily revealed from conventional (i.e.
Per Avseth (*) stratigraphic and structural) seismic interpretation.
Tullow Oil Norge, Oslo, Norway However, to make sure we understand the meaning
NTNU, Trondheim, Norway of the seismic amplitudes, a quantitative link is needed
e-mail: [email protected] between the geological parameters and the rock
456 P. Avseth
Well #1 Well #2 Well #3

2.1
2.2
2.3
2.0 2.5 3.0 3.5 2.0 2.5 3.0 3.5 2.0 2.5 3.0 3.5
Vp (km/s) Vp (km/s) Vp (km/s)
Fig. 18.1 Seismic amplitude map from the Glitne Field, with velocities and the log patterns are observed as we go from the
sonic well log data for three wells penetrating a submarine fan feeder-channel to the more distal part of the lobe system.
system at different locations. Drastic changes in the seismic (Adapted from Avseth et al. 2005)
physics properties. The seismic reflections are physi- observed at a given well location to rock types
cally explained by contrasts in elastic properties, and assumed to be present nearby. Hence, during quantita-
rock physics models allow us to link the elastic proper- tive seismic interpretation of a reservoir we can do
ties to geological parameters. Hence, the application sensitivity analysis not only of fluid types, but also of
of rock physics models can guide and improve on the the reservoir quality.
qualitative interpretation (e.g. Mavko et al. 2009, The way in which geological trends in an area can
Avseth et al. 2005). Moreover, if we understand the be used to constrain rock physics models is also inves-
link between geological parameters and rock physics tigated. Geological trends can be split into two cate-
properties, we can avoid certain ambiguities in seismic gories: Depositional and Compactional. If we can
interpretation, particularly fluid/lithology, sand/shale predict the expected change in seismic response as a
and porosity/saturation. During fluid substitution, it is function of depositional environment or burial depth,
very common to assume that the rock type and poros- this will increase our ability to predict hydrocarbons,
ity are constant, neglecting the possibility that lithol- especially in areas with little or no well log informa-
ogy can change from the brine zone to the tion. Understanding the geological constraints in an
hydrocarbon zone. The link between rock physics area of exploration reduces the range of expected
and various geological parameters, including cement variability in rock properties and hence reduces the
volume, clay volume and degree of sorting, allow us to uncertainties in seismic reservoir prediction.
perform lithology substitution from rock types Figure 18.2 depicts this problem, where the only well
18 Explorational Rock Physics: The Link between Geological Processes. . . 457
How do seismic
properties change
locally within a
sedimentary basin?
What are the important

rock physics changes
as we go from one
Shelf edge basin to another?
Unconsolidated
sediments
Deep-water environment
?
Consolidated Depositional trend
rocks ?
Diagenetic trend
Fig. 18.2 Rock physics properties change with depositional environment and burial depth. These geological trends must be taken
into account during hydrocarbon prediction from seismic data
log control we have is in the shallow interval on the With the above assumptions, we only need two
shelf edge. Before extending the exploration into more elastic constants to specify the stress-strain relation
deeply buried zones, or to more distal deepwater completely, i.e. λ and μ, which we often refer to as
environments, it is important to understand the rock the Lame’s parameters. The first Lame’s parameter,
physics trends in the area. λ, can be related to the bulk modulus, K, and the
shear modulus, μ, via λ ¼ K 2μ/3. The bulk modu-
lus, K, is defined as the ratio of the hydrostatic stress,
18.2 Rock Physics Fundamentals σ0, to the volumetric strain: σ 0 ¼ Kεαα. The bulk
modulus is the reciprocal of the compressibility,
Rock physics provides a link between geologic reser- 1/K, which is used to describe the volumetric com-
voir parameters (e.g. porosity, clay content, sorting, pliance of a material. Hence, the bulk modulus
lithology, saturation) and seismic properties (e.g. expresses a material’s resistance to changing its vol-
acoustic impedance, the Vp/Vs ratio, bulk density and ume V. In a static experiment, the deformations are
the elastic moduli). Assuming an isotropic, linear elas- measured explicitly, as the relative change in volume
tic material, the stress and strain are related by the ΔV/V caused by a relative change in pressure ΔP,
famous Hooke’s law (Timoshenko and Goodier 1934; giving the static bulk modulus: K ¼ ΔP/(ΔV/V). The
Mavko et al. 2009): shear modulus is defined as the ratio of the shear
stress to the shear strain, σ ij ¼2μεij, and it expresses
σ ij ¼ λδij εαα þ 2μεij ð18:1Þ the material’s resistance to shear deformation. The
bulk and shear moduli, which have the same units as
where
stress (force/area, expressed in SI units as Pascal) can
εij ¼ elements of the strain tensor be related to seismic compressional (VP) and shear
σ ij ¼ elements of the stress tensor (VS) velocities, respectively, according to the follow-
εαα ¼ volumetric strain (sum over repeated index) ing equations:
δij ¼0 if i 6¼ j and δij ¼1 if i ¼ j.
458 P. Avseth
Rock physics models also help infer (diagnose)

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
K þ 4μ=3
VP ¼ ð18:2Þ rock texture of sandstones or shales if we know poros-
ρ
ity and the elastic-wave velocity. Such diagnostics are
rffiffiffi based on an assumption that, e.g. if velocity-porosity
μ
VS ¼ ð18:3Þ data fall on a theoretical cemented-rock trend, the rock
ρ
is cemented. This seemingly circular logic helps better
understand rock properties beyond elasticity. For
where ρ is the bulk density of the rock. The elastic example, if rock is cemented, one may expect higher
constants are then referred to as dynamic moduli, strength than in uncemented rock of the same porosity
which can differ from the static moduli due to differ- and mineralogy.
ence in strain amplitude. Seismic reflectivity at a rock A number of workers (e.g. Vernik and Nur 1992,
interface is directly related to the contrast in the Dvorkin and Nur 1996, Anselmetti and Eberli 1997)
dynamic elastic parameters. have recognised that the slope of (or impedance-
Commonly, rock physics analysis is done using porosity) trends in sandstones is highly variable and
crossplots (Fig. 18.3), where observed elastic para- depends largely on the geological processes that con-
meters, either from core, well log or seismic measure- trol porosity. Relatively steep velocity-porosity trends
ments, are plotted versus a given geological for sandstones are representative of porosity variations
parameter. Here, data can be compared with rock controlled by diagenesis, i.e. porosity reduction due to
physics models to better understand how the geologi- pressure solution, compaction and cementation.
cal parameters are controlling the elastic parameters. Hence, we often see steep velocity-porosity trends
Rock physics models can also be used to extrapolate when examining data spanning a great range of depths
beyond the available data range to examine certain or ages. The classical empirical trends of Wyllie et al.
“what if” scenarios, such as plausible fluid or lithology (1956), Raymer et al. (1980), Han (1986) and Raiga-
variations. In this way, rock physics can be used to Clemenceau et al. (1988), all show versions of the
forecast seismic response to assumed reservoir and steep, diagenetically-controlled velocity-porosity
overburden properties and conditions. trend. On the other hand, porosity change resulting
from variations in sorting and clay content tend to
yield much flatter velocity-porosity trends, meaning
that porosity controlled by sedimentation is generally
6000
expected to yield flatter trends, which we sometimes
refer to as depositional trends. Data sets from narrow
5000 depth ranges, or individual reservoirs often (though
not always) show this behaviour.
P-Velocity (m/s)
Processess that There is a wide range of different models that can

4000 give sediment strength:
compaction, stress, be used in rock physics analysis (Mavko et al. 2009,
diagenesis
Dræge et al. 2006). Every model has certain advan-
Voigt Avg.
3000 tages and limitations. We follow Box and Draper
Newly Deposited (1987) in believing that: “All models are wrong,
Clean Sand
2000
but some are useful”. There are in general three
Reuss bound different classes of models: theoretical, empirical,
Suspensions and heuristic. In this chapter we show applications
1000
0 20 40 60 80 100 of a hybrid approach where we combine theoretical
suspensions
sand-clay mixt. Porosity (%)
contact theory (granular media) or pore-shape con-
sand
clay-free sandstone
strained models with heuristic bounds to predict sed-
clay-bearing sandstone
imentary microstructure and geologic trends from
Fig. 18.3 P-wave velocity vs. porosity for a variety of water-
elastic properties.
saturated sediments, compared with the Voigt-Reuss bounds. In particular, we have found that diagenetic trends,
Data are from Yin (1992), Han (1986) and Hamilton (1956). which connect newly-deposited sediment on the Reuss
(Adapted from Avseth et al. 2005) elastic bound with the mineral point at zero porosity,
Diagenetic Trend Models

6000
5500
5000
4500
Diagenetic
Vp (m/s)
4000 Trend
3500
3000
Contact Modified
2500 Hashin-Shtrikman
Cement
Suspension Model
2000 Line
1500
1000
0 0.1 0.2 0.3 0.4 0.5
Porosity
Fig. 18.5 Appending Dvorkin-Nur’s contact cement model at

the high porosity end of the modified Hashin-Shtrikman bound
Fig. 18.4 Hashin-Shtrikman and modified Hashin-Shtrikman
improves the agreement with sands (Adapted from Avseth et al.
bounds for bulk modulus in a quartz-water system (Adapted
2010)
from Avseth et al. 2010)
can often be described accurately using the upper

Hashin-Shtrikman bound, see Fig. 18.4. In fact, we Diagenetic vs. Sorting Trends
6000
sometimes refer to it as a modified upper Hashin- Mineral Point
5500
Shtrikman bound because we use it to describe a
5000
mixture of the newly deposited sediment at critical
4500 Clean Sand
porosity with additional mineral, instead of describing Diagenetic
Vp (m/s)
Trend
a mixture of mineral and pore fluid. A slight improve- 4000
ment over the modified upper Hashin-Shtrikman 3500
bound as a diagenetic trend for sands can be obtained 3000
by steepening the high porosity end. An effective way 2500

Sorting
to do this is to use Dvorkin’s model (Dvorkin and Nur 2000
Suspension Trend
1996) for cementing of grain contacts. The contact 1500 Line
cement model captures the rapid increase in elastic 1000

0 0.1 0.2 0.3 0.4 0.5
stiffness of a sand, without much change in porosity Porosity
as the first bits of cement are added (Fig. 18.5).
The depositional or sorting trends can be described Fig. 18.6 Generalised clastic model. Sediments are deposited
with a series of modified Hashin-Shtrikman lower along the suspension line. Clean, well sorted sands will have
initial (critical) porosity of ~0.4. Poorly sorted sediments will
bounds (Fig. 18.6). By combining the contact cement have a smaller critical porosity. Burial, compaction and diagen-
model with such sorting trends, we can create lines of esis move data off the suspension line. Sediments of constant
constant depth, but variable texture, sorting and/or clay shaliness or sorting and variable age (or degree of diagenesis)
content. For more details and equations behind these fall along the (black) cementing trends. Sediments of constant
age, but variable shaliness or sorting will fall long the (grey)
models, see Avseth et al. (2005). Below, we demon- sorting trends. (Adapted from Avseth et al. 2010)
strate how we can use these models to quantify the rock
texture of sandstones based on North Sea well data.
In order to interpret the observed seismic contrast, tions. So-called inclusion-based models where solids
we also need to know the rock properties of shales. are filled with pores of different shapes and concentra-
These require more complex rock physics models to tions (e.g. Differential Effective Medium model, Self-
capture crack-like pore shapes, intrinsic microporos- Consistent Approach; see Mavko et al. 2009) have
ity, transverse isotropy and diagenetic mineral transi- proven useful for shales and low-porosity sandstones
460 P. Avseth
where pores are poorly connected (e.g. Hornby et al. C11 C33
ε¼ ð18:6Þ
1994, Johansen et al. 2002, Avseth et al. 2014), and 2C33
where contact theory and granular-based models pre-
sented above are not representative or valid. For shales C66 C44
γ¼ ð18:7Þ
and heterogeneous sandstones it may be important to 2C44
take into account the anisotropy effect on seismic
ðC13 þ C44 Þ2 ðC33 C44 Þ2
velocities (e.g. Golikov et al. 2013). The sedimento- δ¼ ð18:8Þ
logical layering of interbedded sands and shales, and 2C33 ðC33 C44 Þ
the internal texture of shale, will yield transverse isot- The constant ε can be seen to describe the fractional
ropy, where seismic waves propagating in the vertical difference of the P-wave velocities in the vertical and
direction will travel at a slower velocity than seismic horizontal directions:
waves propagating in the horizontal direction. In an
isotropic medium we are only dealing with two inde- V P ð90∘ Þ V P ð0∘ Þ
ε¼ ð18:9Þ
pendent elastic constants as given in the isotropic V P ð 0∘ Þ
version of Hooke’s law (Eq. 18.1). Assuming an aniso-
tropic, linear, elastic solid, Hooke’s law can be and is usually referred to as “P-wave anisotropy”.
expressed in compact form as follows: In the same manner, the constant γ can be seen to
describe the fractional difference of SH-wave veloci-
σ ij ¼ cijkl εkl ð18:4Þ ties (horizontally polarised shear wave velocity)
between vertical and horizontal directions, which is
where the stiffness tensor is a fourth rank tensor and equivalent to the difference between the vertical and
has 81 components. However, not all the components horizontal polarisations of the horizontally propagat-
are independent. Symmetry of stresses and strains ing S-waves:
implies that the number of independent constants
V SH ð90∘ Þ V SV ð90∘ Þ
reduces to 36. The existence of a unique strain energy γ¼
V SV ð90∘ Þ
potential further reduces it to 21. And for a transverse
isotropic medium, where the vertical axis lies along V SH ð90∘ Þ V SH ð0∘ Þ
¼ ð18:10Þ
the axis of symmetry, the number of independent V SH ð0∘ Þ
constants are reduced to 5. With simplified Voigt
The physical meaning of δ is not as clear as ε and γ,
annotation (see Mavko et al. 2009) the elastic stiffness
but δ is the most important parameter for normal move
tensor can then be expressed as:
out velocity and reflection amplitude.
Later in this chapter, we will show how we can
C11 ðC11 2C66 Þ C13 0 0 0

ðC11 2C66 Þ C11 C13 0 0 0 utilise rock physics models to calibrate seismic ampli-
tude data, and to help us predict lithology and fluids

C13 C13 C33 0 0 0
C ¼ ,
0 0 0 C44 0 0
prior to drilling, and we will also demonstrate how we
0 0 0 0 C44 0

0 0 0 0 0 C66 can take into account shale anisotropy in our analysis.
1 Rock physics theory and models for fluid effects are
where C66 ¼ ðC11 C12 Þ ð18:5Þ
2 included in a separate section later in this chapter
(Sect. 18.6).
This 6 6 matrix is symmetric and the five indepen-
dent components are annotated C11, C13, C33, C44, and
C66. These are not easily related to geological para- 18.3 Rock Physics Models for
meters as the shear and bulk moduli, but Thomsen Microstructure Interpretation
(1986), assuming weak anisotropy, expressed three
anisotropic parameters, ε, γ and δ as a function of the Rock physics models can either be used to interpret
five elastic components, where: observed sonic and seismic velocities in terms of
reservoir parameters, or they can be used to extrapo- important to include these geological factors in the
late beyond an observed range to predict certain “what rock physics analysis.
if” scenarios in terms of fluid or lithology substitution. Using the diagnostic models described in
Rock physics models can also be used to estimate Sect. 18.2, we can infer the microstructure from veloc-
expected seismic properties from observed reservoir ity-porosity data (e.g. Dvorkin and Nur 1996, Avseth
properties. Pore fluid and stress sensitivity in reservoir et al. 2000, 2010). With good local validation of the
sandstones are highly affected by reservoir heteroge- models, we can even quantify the degree of sorting
neity and sandstone microstructure, and it is therefore and cement volume from these diagnostic crossplots.
a 4500 10
4000
Qz 9
8
3500 Constant cement
trends 7
3000
Cement volume %
Vs (m/s)
6
Increasing cement volume
2500 5
Dvorkin-Nur
2000 4
contact cement
3
1500
2
1000
1
500 Shale
0
0 -1
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Porosity
b 4500 1
4000 0.9
0.8
3500
Sorting = Por/Por_C 0.7

Vs (m/s)
3000
0.6
2500
0.5
Sorting
2000
0.4
1500
0.3
1000
0.2
500 0.1
Shale
0 0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45
Porosity
Fig. 18.7 Shear wave velocity log data versus total porosity values, and for practical reasons are given the value –1 in
and superimposed diagnostic rock physics models. Using the cement volume and 0 in sorting). (Taken from Avseth et al.
models we can quantify the cement volume (a) and degree of 2009)
sorting (b) (Green data points are shale data with high GR
462 P. Avseth
Colour=Sorting
1200 1200
1.1 –1000
1400 1400 Walderhaug et al. (2000)

1.0
0.9
1600 1600 –2000
Depth (m below sea floor)

0.8
1800 1800
0.7
Depth (m)
–3000
2000 2000 0.6
0.5
2200 2200
0.4 –4000
2400 2400 0.3
0.2 –5000
2600 2600 0 10 20 30 40
0.1 Quartz cement (%)
2800 2800
0 0.5 1 0 5 10
Shale volume Cement volume %
Fig. 18.8 Estimated cement volume and sorting versus depth observations made by Walderhaug et al. (2000). Sorting, how-
for a North Sea well. Note the onset of cement at around 2,000 m ever, shows a more erratic pattern with no depth trend (colour in
depth (corresponding to c.70 C) and the increasing cement second column; see also Avseth et al. 2009)
volume with depth (second column). This is in agreement with
Figure 18.7 shows an example from Avseth et al. superimposed colour. For the relatively clean Heimdal
(2009). Here, the rock physics diagnostics is done in sandstones starting at around 2 km depth, we observe a
the Vs versus porosity domain, in order to avoid sig- clear depth trend in the cement volume. This matches
nificant pore fluid effects. The cement volume is esti- observations made by Walderhaug et al. (2000)
mated by interpolating between the constant cement (Fig. 18.8, right subplot). The sorting shows a more
volume trends. For a given constant cement trend, erratic pattern, lacking any consistent depth trend, as
sorting is defined to vary between 1 and 0. Sorting we would expect since sorting is associated with depo-
equals 1 for data points that fall right on the contact sitional trends.
cement model. Deteriorating sorting implies increas- It is essential to verify with thin section observa-
ing volume of pore-filling (non-cementing) material tions the presence of initial cementation predicted
which will cause lower porosities. The data points will from the rock physics relations. Figure 18.9 shows a
then plot further away from the contact cement model, thin-section from the relatively clean Heimdal sands,
towards zero porosity. At the zero porosity end mem- but at first glance the sandstone looks unconsolidated
ber, sorting is defined to be zero. It is important to note with grains loosely arranged and moderately well
that this definition of sorting is not the same as the sorted. A closer investigation, however, reveals the
qualitative sorting parameter defined in the field of presence of initial quartz overgrowth covering original
sedimentology, but they will be highly correlated, grain surfaces, indicated by dust rims (see arrows,
with values approaching 0 when sands are poorly sorted Fig. 18.9). This observation confirms what we see in
and values approaching 1 when sands are well sorted. the rock physics crossplots of the well log data. It is
Having estimated cement volume and sorting, we interesting that the well log data with 10s of cm reso-
can plot these as logs and compare with other petro- lution reflect what we observe at the micro-scale. A
physical logs. Figure 18.8 shows the resulting estima- comparison between cement volume estimated from
tion of cement volume and sorting. The middle subplot rock physics models and the cement volume from the
shows cement volume as magnitude, with sorting as thin-section point count analysis (only available in the
Qz-cement
Fig. 18.9 Thin sections from Heimdal Formation sands. The grains we observe dust rims representing the original grain
left image shows a loosely packed, poorly consolidated sand. surfaces that have been covered by quartz cement (arrows).
Analysis of a zoomed-in image (right) confirms the presence of Feldspar overgrowth and calcite cement also occur, though
quartz overgrowths and contact cement. On detrital quartz quartz cement is dominating. (From Avseth et al. 2009)
2050 2050
Qz cement from thin section
2100 2100
Total cement from thin section
Depth (m)
2150 2150
2200 2200
50 100 150 0 5 10
GR Cement volume %
Fig. 18.10 The estimated cement volume based on rock phys- estimated cement volume is slightly larger than the point-
ics models (black line in right-hand subplot) compared with counted quartz cement, but for most of the depth-range it
point-count cement volume in the upper 100 m of the Heimdal matches nicely with the point-counted total cement volume.
reservoir sands in the well analysed in Figs. 18.7 and 18.8. The (From Avseth et al. 2009)
upper part of the Heimdal sands in the well analysed in 18.4 Rock Physics and Depositional
Figs. 18.7 and 18.8) is shown in Fig. 18.10. The Trends
volume of quartz cement from the thin-section analy-
sis is somewhat lower than the volume estimated from There exist several rock physics models for deposi-
the rock physics models. However, if we compare the tional trends in siliciclastic environments (e.g. Marion
rock physics estimates with the total cement volume, 1990, Dvorkin and Gutierrez 2002, Xu and White
the match is very good. While most of the cement in 1995). Marion (1990) introduced a topological model
the Heimdal Fm is quartz in this well, other types are for sand-shale mixtures to predict the interdependence
also present, like feldspar overgrowths and carbonate between velocity, porosity and clay content. When
cement (see also Lehocki and Avseth 2010). clay content is less than the sand porosity, clay
464 P. Avseth
Sand, shaley Sandy shale

sand shale
3500
50 MPa
c =Æs
0.6
Sh
3000 ale
Porosity
ys
P-Velocity (m/s)
0.4 and
s
Sa
nd
ys
0.2
hal
es
0 50 100 2500
Clay content by volume Sand
Shale
2.0
P-Velocity (km/s)
1.8
2000
0.1 0.2 0.3
1.6 Porosity
0 50 100
Clay content by volume
Fig. 18.11 Left: The Yin-Marion topological model of porosity unconsolidated sands and shales at constant effective pressure of
and P-wave velocity versus clay content for shaly sands and 50 MPa. A clear V-shape trend is observed with increasing clay
sandy shales (from Marion 1990). Right: Laboratory experi- content, where velocity reaches a maximum and porosity a
ments (Yin 1992) showing P-wave velocity versus porosity for minimum when the clay content equals the sand porosity
particles are assumed to be located within the pore increasing clay content, and velocity decreases. It also
space of the load-bearing sand. The clay will stiffen has to be mentioned that clay particles can be depos-
the pore-filling material, without affecting the frame ited as lamina between sand grains or intervals of
properties of the sand. Therefore, increasing clay con- sands, and these will yield a completely different
tent will increase the stiffness and velocity of the sand- elastic response than pore-filling clays (e.g. Sams
shale mixture as the elastic moduli of the pore-filling and Andrea 2001).
material (fluid and clay) increase. Once the clay con- Until recently, shales have often been regarded by
tent exceeds the sand porosity, the addition of more geophysicists as a single type of lithology, with little
clay will cause the sand grains to become separated, as attention given during seismic data analysis to the
we go from grain-supported to clay-supported sedi- wide variation in their mineralogy, texture and poros-
ments (i.e. shales). ity. This is partly because the rock properties of clay
Marion assumed that, similar to fluids, the pore- minerals are difficult to measure in the laboratory, but
filling clay would not significantly affect the shear also because the acquisition of detailed log data and
modulus of the rock. This assumption was supported core samples in shale sequences has been given little
by laboratory measurements on unconsolidated sand- priority in the oil industry. Geologists, however, have
shale mixtures (Yin 1992). The impact on the velocity- documented the complexity of shales and there is a
porosity relationship of increasing clay content in a vast amount of published literature on the geochemis-
sand/shale mixture is depicted in Fig. 18.11. From the try and sedimentology of shales (e.g. Bjørlykke 1998,
measured data we can see that when clay content MacQuaker et al. 2007, Peltonen et al. 2008). With
increases, porosity decreases and velocity increases increased focus on cross-disciplinary integration, geo-
up to a given point called the critical clay content. physicists are starting to incorporate this geological
This point represents the transition from shaly sands to knowledge into the modelling and analysis of geo-
sandy shales. After this point, porosity increases with physical data (e.g. Dræge et al. 2006, Brevik et al.
2007, Mondol et al. 2007, Marcussen et al. 2009). As Avseth et al. (2001a) defined seismic lithofacies as
with sands, we can distinguish between depositional a seismic-scale sedimentary unit which is charac-
and diagenetic trends in shales. Depositional trends terised by its lithology (sand, silt and clay), bedding
will affect clay mineralogy, but more particularly, configuration (massive, interbedded or chaotic),
the silt content will have great impact on the seismic petrography (grain size, clay location and cementa-
properties. Avseth et al. (2005) used simple isotropic tion) and seismic properties (P-wave velocity, S-
lower Reuss bound to model vertical velocities of silty wave velocity, and density). Using the concept that
shales. More rigorous anisotropic modelling of shales seismic lithofacies represent seismic-scale sedimen-
has been performed by Hornby et al. (1994), Ruud tary units, we can improve our lateral facies predic-
et al. (2003), Johansen et al. (2002) and Dræge et al. tion, as we can link facies observed in vertical well
(2006), among others. logs to seismic attribute maps in accordance with
One of the fundamental aspects of this chapter is to Walther’s law of facies (e.g. Middleton 1973).
establish a link between rock physics and sedimentol- In the Glitne turbidite system depicted in Fig. 18.1,
ogy. More specifically, we want to relate lithofacies to Avseth et al. (2001a) identified various seismic litho-
rock physics properties. This will improve our ability facies in turbidite systems (I ¼ conglomerates/grav-
to use seismic amplitude information to interpret els, II ¼ massive sandstones, III ¼ interbedded
depositional systems, as facies have a major control sands-shales, IV ¼ silty shales, V ¼ shales), and
on depositional geometries and porosity distributions. these represent a more or less gradual transition from
Furthermore, facies occur in predictable patterns in clean sandstone to pure shale. Three subfacies of
terms of their lateral and vertical distribution, and Facies II are defined based on seismically important
can also be linked to sedimentary processes. Hence, petrographic variations within the thick-bedded sand
facies represent an important parameter in seismic facies. These subfacies were determined from core,
exploration and reservoir characterisation. thin-section and SEM analyses, in combination with
rock physics diagnostics, and include cemented clean
Fig. 18.12 Identifying seismic lithofacies from well log data; example from Palaeocene interval in the Glitne Field, North Sea.
(From Avseth et al. 2001a)
466 P. Avseth
8
2.8
7.5 IIa
IIb
Acoustic impedance
2.6
7 IIc
III
IV 2.4
6.5 V
Vp/Vs
2.2
6
IIa
5.5 2 IIb
IIc
5 III
1.8 IV
V
4.5 1.6
50 60 70 80 90 100 110 120 50 60 70 80 90 100 110 120
Gamma ray (API) Gamma ray (API)
Fig. 18.13 Acoustic impedance (in km/s∗g/cm3) and Vp/Vs bedded sands have relatively high impedance values. In the
ratio versus gamma ray for different seismic lithofacies in the Vp/Vs versus gamma ray plot, we observe a more linear trend
Glitne turbiditic field depicted in Fig. 18.1. Note the overturned as we go from clean sands to shaly facies. (From Avseth et al.
V-shape in the left hand subplot, where clean sands and shales 2001a)
have similar acoustic impedance values, whereas shaly/inter-
sands (Facies IIa), uncemented or friable clean sands Cemented sands (IIa) and laminated sands (IIc), as
(Facies IIb) and plane-laminated sands (Facies IIc). well as interbedded sand-shales have relatively high
Our seismic lithofacies can be linked to depositional impedances. The sand-shale ambiguity is not observed
sub-environments and sedimentary processes within a in the Vp/Vs versus gamma ray plot. Here we see a
deepwater clastic system (e.g. Walker 1978, Reading more linear trend where Vp/Vs increases with increas-
and Richards 1994). ing gamma ray values (i.e. clay content) as we go from
Avseth et al. (2001a) selected a type-well for iden- clean sands (Facies IIa and IIb) to shales (Facies IV
tification of seismic lithofacies from well log data and V). The overturned V-shape we observe in acous-
(Fig. 18.12). Primarily, the gamma ray log was used tic impedance can be explained physically: for grain-
to determine the different facies, as it is a good clay supported sediments, increasing clay content tends to
indicator in the quartz-rich sediments of the North Sea. reduce porosity (i.e. increase density) and therefore
Density and sonic logs were also used to ensure that stiffen the rock. However, for clay-supported sedi-
each facies occurs as significant clusters in terms of ments, porosity will increase with increasing clay con-
rock physics properties. Rock physics analysis can tent due to the intrinsic porosity of clay, and the rock
furthermore be used diagnostically to determine litho- framework will weaken. Hence, velocity will reach a
facies when direct core and thin-section data are not peak when clay content is approximately 40%, cf. the
available. This was essential in order to confirm the Yin-Marion model shown in Fig. 18.11. Higher Vp/Vs
presence of cement in Facies IIa. The cementation in ratios are expected in shales than sands, since the shear
Facies IIa is volumetrically not very significant, but in strength in shales tends to be relatively low compared
terms of elastic properties it has an important impact. to sands, due to the platy shapes of clay particles.
The seismic velocities and impedances are relatively
high because of the stiffening effect of initial cemen-
tation. 18.5 Rock Physics and Compactional
Figure 18.13 shows the different seismic lithofacies Trends
plotted as acoustic impedance versus gamma ray (left),
and Vp/Vs versus gamma ray (right). For acoustic Rock physics depth trends are important in seismic
impedance, we observe an overturned V-shape, and exploration and borehole drilling for several reasons.
an ambiguity exists between Facies IIb and IV/V. Commonly, overpressured zones can be detected from
Deposition
Porosity
Pure
North Sea Paleocene
compaction
Mechanical
shale Clean
sand
Volcanic tuff
Smectite-rich shale
Unc. sand
C. 700C ~ c.2 km
Cem. sst Illite-rich shale

compaction
Chemical
Qz
Qz
Smectite + K+ Illite + SiO2 + H2O
Depth of burial
Fig. 18.14 Schematic illustration of sand and shale compac- bound water, causing both a porosity reduction and mineralogy
tion. At c.70 C it is common to observe a change from mechan- change with depth. For quartz-rich sands, initial cementation
ical compaction to predominantly chemical compaction in tends to start at the same depth level. One possible external
siliciclastic systems. In deep-marine depositional systems, source of cement is in fact derived from the smectite to illite
smectite-rich shales will experience illitisation and release of transition in embedding shales. (From Avseth et al. 2008)
seismic velocity data, indicated by negative velocity hence the more rapid mechanical porosity reduction in
anomalies. Knowing that effective pressure and pore shales than sands. During burial to ~2 km depth, both
pressure increase linearly with depth, rock physics sands and shales are exposed mainly to mechanical
depth trends can be used to quantify overpressure. It compaction. The marine shales in the North Sea Ter-
is extremely important to locate such zones since they tiary are very rich in smectite, which gives them very
can cause hazardous blowouts during drilling. The low permeability. In thick smectite-rich shale masses,
depth trends for sands and shales can also be used to it is therefore normal to observe undercompaction and
study the expected seismic signatures of sand/shale associated overpressure even at burial depths of just
interfaces as a function of depth. several hundred metres. At about 70 C, however,
In this section, we use existing empirical porosity/ chemical alteration of smectite will commence and
depth trends for sands and shales as input to rock we expect a mineral transformation to illite. This is a
physics models of Vp, Vs and density. We can for typical mineral transformation seen in marine shales
example use Hertz-Mindlin theory (Mindlin 1949; all over the world (Bjørlykke 1998). Bound water in
Chap. 19) to calculate the velocity/depth trends for the smectite layers is released when the temperature
unconsolidated sands and shales, whereas Dvorkin- reaches this critical temperature, resulting in a poros-
Nur’s contact-cement model (Dvorkin and Nur 1996) ity decrease. Moreover, the presence of potassium
can be used for cemented sands. The depth trends cations (for instance in feldspar or mica) causes quartz
allow us to discriminate between pore fluids and lithol- to be produced as well. This quartz can precipitate as
ogies at different depths. microcrystalline quartz within the shale matrix (Thy-
Figure 18.14 shows a schematic representation of berg et al. 2009), or if connectivity allows, the quartz
shale and sand compaction curves, and a sequence of may precipitate as cement in adjacent sandstones (Pel-
interbedded turbidite sands and marine shales, typical tonen et al. 2008).
for the North Sea deep-marine environment of Ter- Even though there can be a link between the quartz
tiary age. The depositional porosity in shales is nor- cementation of sandstones and illitisation of smectite-
mally much higher (60–80%) than in sands (c. 40%), rich shales, it is important to note that the source of
but we expect a shallow crossover with depth due to quartz cement in sandstones can result from a variety
the mechanical collapse of the shales. The platy clay of geological processes (Worden and Morad 2000).
fabric in the shales is more prone to compaction than Figure 18.14 also includes a volcanic tuff layer that
the assemblage of spherically shaped grains in sands, is typically encountered in the Tertiary of the North
468 P. Avseth
Sea (i.e. within the Balder Formation). Smectite is 70–80 C. In the North Sea, this corresponds to a burial
often generated from the alteration of volcanic tuff, depth of about 2 km which is around the target depth
but tuff also includes amorphous silica that can pre- of the prolific Palaeocene and Eocene reservoir sands
cipitate as quartz cement even at temperatures below that represent major prospects for the oil industry.
70 C. Amorphous silica is thus able to dissolve at Geophysicists should be focused on these geological
lower temperatures than crystalline quartz. factors during seismic data analysis of reservoir sands,
There are also potential internal sources of quartz because dramatic changes in the seismic signatures
cement in the sandstones, either from authigenic or may reflect not pore fluid changes, but diagenetic
detrital clays in the sandstone matrix, or from the alterations in the cap-rock shales and/or reservoir
quartz grains themselves. Moreover, it has been con- sandstones.
firmed that clay minerals can both inhibit the precipi- Figure 18.15 shows well log data from a North Sea
tation, and catalyse the dissolution, of quartz in well that penetrates siliciclastic sediments and rocks of
sandstones. The presence of oil has also been shown Tertiary age, superimposed with rock physics depth
to inhibit quartz cementation in oil-wetting sandstones trends for shales and sandstones. We observe a nice
(e.g. Giles et al. 2000). There is considerable debate match between the calculated velocity depth trends for
among geologists over whether effective pressure at different lithologies and the well log data. The sand-
quartz grain contacts in sandstones can cause quartz stone rock physics depth trends are modelled by com-
dissolution and re-precipitation around the grain con- bining Hertz-Mindlin contact theory (see Chap. 19) for
tacts. More likely, time and temperature control the unconsolidated sands with the Dvorkin-Nur contact
quartz cementation (e.g. Bjørlykke and Egeberg 1993). cement model for cemented sandstones (Dvorkin and
Regardless of the complexity of the clay and quartz Nur 1996, Avseth et al. 2003, 2005, 2008). The input
diagenesis, there is empirical evidence that both the porosity/depth trends are calibrated with local com-
smectite to illite transformation in shales and the paction trends according to empirical relations (e.g.
quartz cementation of sandstones are geochemical Mondol et al. 2007, Ramm and Bjørlykke 1994). The
processes that tend to initiate concurrently at around light blue model trend curves in Fig. 18.15 show how
Seafloor
500
1000
Depth (m)
1500
2000 Gas
Oil
2500
3000
0 0.5 1 2000 3000 4000 1000 2000 3000 2 2.5
Vsh Vp (m/s) Vs (m/s) Density (g/cm3)
Fig. 18.15 Rock physics depth trends for shales (dark blue) yellow (2,099–2,151 m), and an oil zone in red (2,151–2,168 m).
and sandstones (cyan) juxtaposed on North Sea well log data The remaining interval of the Heimdal Formation is brine-filled.
penetrating a Tertiary sequence of siliciclastic sediments and The Heimdal Formation is embedded in the Lista Formation
rocks (same well data as in Fig. 18.8). A gas zone is indicated in shale. (Adapted from Avseth et al. 2008)
Delta Epsilon
1200
1400
Smectite
rich 1600
1800 c.70°C
Depth (m)
2000
Illite
rich 2200
2400
2600
2800
2000 2500 3000 500 1000 1500 2.2 2.4 2.6 0 0.1 0.2
Vp (m/s) Vs (m/s) 3 Anisotropy
Density (g/cm )
Smectite + K Illite + Si + H2O
Fig. 18.16 Modelled rock physics depth trends of shales, showing the effect of illitisation of marine smectite-rich shales. (From
Avseth et al. 2008)
the velocities for sands increase drastically as we go Consistent Approximation (SCA) are used to approxi-
from the unconsolidated regime with only mechanical mate the elongated pores and grains in shales (Hornby
compaction to the cemented regime with predomi- et al. 1994). In the present discussion, pores in chemi-
nantly chemical compaction. The onset of quartz cally compacted shales are considered to be isolated,
cement happens at about 70 C corresponding to while the pores in the mechanical compaction regime
about 2 km burial depth, in accordance with the obser- are connected (cf. Dræge et al. 2006). We define a
vations made in Fig. 18.8, and as documented from transition zone, where the properties change from the
thin-section analysis (Fig. 18.9). mechanical to chemical regime. In Fig. 18.16, the
We apply the Shale Compaction Model (Dræge initial shale (<1,500 m) is considered to be smectite-
et al. 2006, Ruud et al. 2003) to estimate the aniso- rich, while the deeper (>2,200 m) illite-rich shale is
tropic effective properties in mechanically compacted somewhat stiffer. There are two counteracting effects
shales. The first seismically important mineral reac- on anisotropy. The initial alignment of grains leads to
tion in shales is commonly the smectite to illite reac- pores becoming more aligned and hence increasing
tion. This reaction has several implications for the anisotropy. But decreasing porosity leads to decreas-
shale; the soft smectite is replaced by stiffer illite, ing anisotropy, culminating in the anisotropic proper-
which might be distributed differently in the rock, ties of the solid material at zero porosity. The pores
the reaction produces water, the volume of solids is introduce higher anisotropy than the solid, since pores
decreased (i.e. illite has a denser mineral structure than commonly are weaker orthogonal to the longest axis,
smectite), quartz is generated as a by-product, and while the solids are less dependent on direction of
porosity is reduced by chemical compaction. Where wave propagation. In addition to Vp, Vs and density,
the shales lie within the chemical compaction regime, we estimate the Thomsen parameters of anisotropy, δ
a new set of rock physics models is applied to estimate and ε, which can be significant during interpretation of
the seismic properties. The anisotropic version of a angle-dependent seismic reflectivity (AVO analysis),
Differential Effective Medium (DEM) model and Self see Sect. 18.8.
470 P. Avseth
18.6 Seismic Fluid Sensitivity and 1 1 ϕ

¼ þ ð18:13Þ
Gassmann Theory for Fluid K sat K mineral K ϕ þ Kfluid Kmineral
K fluid þK mineral
Substitution
or approximately as:
Seismic fluid sensitivity is determined by a combina-
tion of porosity and pore-space stiffness. A softer rock 1 1 ϕ
þ ð18:14Þ
will have a larger sensitivity to fluids than a stiffer K sat K mineral K ϕ þ K fluid
rock at the same porosity. The most common theory
for fluid substitution is the so-called Gassmann theory where Kfluid is the pore fluid bulk modulus. Comparing
(Gassmann 1951, Mavko et al. 2009, Avseth et al. Eqs. (18.11) and (18.14), we see that changing the pore
2005), which describes the fluid sensitivity of porous, fluid will modify the effective pore space stiffness
isotropic rocks at seismic frequency (i.e. when no (Avseth et al. 2005). From Eq. (18.14) we see also
grain-scale viscous flow effects will stiffen the rock the well-known result that a stiff rock, with large pore
frame). The Gassmann theory is very important in all space stiffness Kϕ, will have a small sensitivity to
rock physics modelling, therefore both the formulation fluids, and a soft rock, with small Kϕ, will have a
and workflow of the theory are included in this over- large sensitivity to fluids.
view chapter. Equations (18.11) and (18.13) together are equiva-
For the fluid substitution problem there are two lent to Gassmann’s (1951) relations. We can algebrai-
fluid effects that must be considered: the change in cally eliminate Kϕ from Eqs. (18.11) and (18.13) and
rock bulk density, and the change in rock compress- write Gassmann’s relations in one of the more famil-
ibility. The compressibility of a dry rock (reciprocal of iar, but less intuitive, forms:
the rock bulk modulus) can be expressed quite gener-
ally as the sum of the mineral compressibility and an K sat K dry
¼
extra compressibility due to the pore space: K mineral K sat K mineral K dry
K fluid
1 1 ϕ þ ð18:15Þ
¼ þ ð18:11Þ ϕ K mineral K fluid
K dry K mineral K ϕ
and the companion result for the shear modulus
where ϕ is the porosity, Kdry is the dry rock bulk
modulus, Kmineral is the mineral bulk modulus, and μsat ¼ μdry ð18:16Þ
Kϕ is the pore space stiffness defined by:
Gassmann’s equations (18.15) and (18.16) predict that
1 1 ∂vpore for an isotropic rock, the rock bulk modulus will
¼ ð18:12Þ
K ϕ vpore ∂σ change if the fluid changes, but the rock shear modulus
will not.
Here, vpore is the pore volume, and σ is the increment These dry and saturated moduli, in turn, ffi are related
of hydrostatic confining stress from the passing wave.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
to P-wave velocity V P = ð K þ ð 4=3 Þμ Þ=ρ and S-wave
Poorly consolidated rocks, rocks with microcracks,
pffiffiffiffiffiffiffiffi
velocity V S = μ=ρ, where ρ is the bulk density given
and rocks at low effective pressure, are generally soft by
and compressible and have a small Kϕ. Stiff rocks that
are well cemented, lacking microcracks, or at high ρ ¼ ϕρfluid þ ð1 ϕÞρmineral ð18:17Þ
effective pressure have a large Kϕ. In terms of the
popular, but idealised, ellipsoidal crack models, low In the equations above, ϕ is normally interpreted as the
aspect ratio cracks have small Kϕ and rounder large total porosity, though in shaly sands the better choice
aspect ratio pores have large Kϕ. is to use effective porosity during fluid substitution
Similarly, the compressibility of a saturated rock (Dvorkin et al. 2007). Another uncertainty stems
can be expressed as: from Gassmann’s assumption that the rock is mono-
mineralic. Clay-rich sandstones actually violate the
monomineralic assumption, but an effective mineral brine phases. The operator h i refers to a volume
modulus can be approximated via the Hill’s average or average and allows for more compact expressions,
Hashin-Shtrikman modelling (see Mavko et al. 2009). where Kfluid(x, y, z) and ρfluid(x, y, z) are the spatially
Another issue is whether the clay should be considered varying pore fluid modulus and density.
only as part of the mineral frame, or should the bound Substituting Eq. (18.18) into Gassmann’s relation is
water that constitutes part of the clay mineral be con- the procedure most widely used today to model fluid
sidered part of the fluid when we do Gassmann fluid effects on seismic velocity and impedance for low
substitution? If the latter, then the functional Gass- frequency field applications. The fluid properties are
mann porosity is actually larger than the total porosity, obtained through the Batzle and Wang (1992) equa-
but the pore fluid should be considered a muddy sus- tions, and these normally comprise brine salinity, gas
pension containing clay particles. gravity, oil reference density, gas/oil ratio (GOR),
Gassmann’s relations were originally derived to temperature and pore pressure.
describe the change in rock modulus from one pure A problem with mixed fluid phases is that velocities
saturation to another – from dry to fully brine- depend not only on saturations but also on the spatial
saturated, from fully brine-saturated to fully oil- distributions of the phases within the rock (Fig. 18.17).
saturated, etc. Domenico (1976) suggested that Equation (18.18) is applicable only if the gas, oil, and
mixed gas-oil-brine saturations can also be modelled brine phases are mixed uniformly at a very small scale,
with Gassmann’s relations, if the mixture of phases is so that the different wave-induced increments of pore
replaced with an effective fluid with bulk modulus pressure in each phase have time to diffuse and equili-
K fluid and density ρfluid given by brate during a seismic period (Mavko et al. 2009).
Equation (18.18) is the Reuss (1929) average or “iso-
1 Sgas Soil Sbr stress” average, and it yields an appropriate equivalent
¼ þ þ fluid when all pore phases have the same wave-
K fluid K gas K oil K br
induced pore pressure. A simple dimensional analysis
1
¼ ð18:18Þ suggests that during a seismic period pore pressures
K fluid ðx; y; zÞ can equilibrate over spatial scales smaller than
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ρfluid ¼ Sgas ρgas þ Soil ρoil þ Sbr ρbr Lc κK fluid =f η, where f is the seismic frequency,
κ is the permeability, and η and Kfluid are the viscosity
¼ ρfluid ðx; y; zÞ ð18:19Þ and bulk modulus of the most viscous fluid phase. We
refer to this state of fine-scale, uniformly mixed fluids
where Sgas,oil,br,Kgas,oil,br, and ρgas,oil,br are the satura-
as “uniform saturation.”
tions, bulk moduli, and densities of the gas, oil, and
In contrast, saturations that are heterogeneous over
scales larger than ~Lc will have wave-induced pore
3.0 pressure gradients that cannot equilibrate during the
Patchy (upper bound) seismic period, and Eq. (18.18) will fail. We refer to
VP
this state as “patchy saturation”. Patchy saturation, or
Velocity (km/s)
“fingering” of pore fluids, can easily be caused by

spatial variations in wettability, permeability or shali-
ness. The rock modulus with patchy saturation can be
Uniform (lower bound) approximated by Gassmann’s relation, with the mix-
ture of phases replaced by the “isostrain” (Voigt 1910)
average effective fluid (Mavko et al. 2009):
VS
1.5 K fluid ¼ Sgas K gas þ Soil K oil þ Sbr K br : ð18:20Þ

0 100
Water saturation (%) This implies that patchy saturation will cause
higher velocities and impedances than when the
Fig. 18.17 Uniform versus patchy saturation, mixing gas and same fluids are mixed at a fine scale. The fact that
water in a porous rock according to Gassmann theory the fluids cannot move freely and equilibrate during a
472 P. Avseth
Vp/Vs
Vp/Vs
Shale Shale
Initial
Decreasing Brine sand cementation
Oil sand
Sbr Brine sst
Qz mineral
Oil sand Decreasing net-to-gross
AI AI
Fig. 18.18 Rock physics templates (RPTs) can be made from fluid sensitivity of sandstones and the Vp/Vs ratio of cemented
the rock physics models in Fig. 18.6 combined with Gassmann brine sandstones can be similar to the Vp/Vs ratio of unconsoli-
theory, to define regions where expected facies and fluids will dated sands filled with oil. The effect of net-to-gross will nor-
plot. In particular, the Vp/Vs ratio is known to be a great fluid mally move data in the opposite direction in a AI-Vp/Vs
discriminator in siliciclastic environments. Homogeneous, crossplot. Hence, oil sands with relatively low net-to-gross can
unconsolidated sands filled with oil are normally well separated have higher Vp/Vs ratio than homogeneous brine sands with
from ditto brine-filled sands in an RPT of Vp/Vs versus acoustic initial cement, c.f. schematic illustration in lower crossplot.
impedance (AI), c.f. schematic illustration in the upper (Adapted from Avseth et al. 2009)
crossplot. However, the effect of initial cement will reduce the
seismic wave period results in a stiffer effective rock under investigation, using well log data to verify the
than is the case with uniform saturation where the validity of the selected RPT(s). (2) Use the selected
fluids have time to equilibrate. and verified RPT(s) to interpret elastic inversion
Equation (18.20) appears to be an upper bound, and results. RPT interpretation of well log data may also
data seldom fall on it, except at very low gas satura- be an important stand-alone exercise both for the
tions. interpretation and quality control of well log data,
and in order to assess seismic detectability of different
fluid and lithology scenarios.
18.7 Rock Physics Templates The RPTs are site (basin) specific and are con-
strained by local geological factors, including lithol-
We can combine the depositional and diagenetic trend ogy, mineralogy, burial depth, diagenesis, pressure
models presented above with Gassmann fluid substitu- and temperature. All these factors must be considered
tion, and make charts or templates of rock physics when generating RPTs for a given basin. In particular,
models for predicting lithology and presence of hydro- it is essential to include only the expected lithologies
carbons. We refer to these locally constrained charts as for the area under investigation when generating the
Rock Physics Templates (RPTs), a methodology first rock physics templates. The water depth and the burial
presented by Ødegaard and Avseth (2004). Further- depth determine the effective pressure, pore pressure
more, we expand on the rock physics diagnostics pre- and lithostatic pressure. The pore pressure is important
sented earlier as we create RPTs of seismic parameters, for the calculation of fluid properties, and for deter-
in our case acoustic impedance (AI ¼ VP ρ) versus mining the effective stress on the grain contacts of the
Vp/Vs ratio (Fig. 18.18). This will allow us to perform rock frame carrying the overburden.
rock physics analysis not only of well log data, but also In modelling RPTs we also need to know the acous-
of seismic data (e.g. elastic inversion results). tic properties of mud-filtrate, formation water and
The motivation behind RPTs is to generate an atlas hydrocarbons in the area of investigation. Required
or collection of relevant rock physics models for dif- input parameters include temperature, pressure, brine
ferent basins and areas. Then, the ideal interpretation salinity, gas gravity, oil reference density and GOR. In
workflow becomes a fairly simple two-step procedure: areas where hydrocarbons have yet to be encountered,
(1) Select the appropriate RPT for the area and depth gas gravity can be assumed (normally 0.6–0.8).
However, oil reference density is more uncertain. area of very low Vp/Vs, where we would expect hydro-
Also, the seismic response of oil can be difficult to carbon-saturated sandstones to plot. By contrast, res-
distinguish from that of brine. One should, however, ervoir heterogeneity and decreasing net-to-gross
expect oil to show values similar to low gas saturation associated with interbedded sands and shales tend to
in AI versus Vp/Vs ratio crossplots. Regarding satura- move data points in the direction of the shale cluster.
tion distribution, we have assumed uniform distribu- The cement effect is a microstructural effect, whereas
tion in the template modelling (Ødegaard and Avseth the net-to-gross is related to the scale of shale/sand
2004), which gives the famous effect where residual interbedding and is frequency dependent. In the case
amounts of gas will produce almost the same seismic where the interbedded shale is relatively soft com-
properties as commercial amounts of gas. However, a pared to the sand, the net-to-gross effect will counter-
patchy distribution of gas would have shown a more act the effect of cement on effective rock stiffness,
linear change in seismic properties with increasing gas hence these two effects will have opposite trends in the
saturation (cf. Fig. 18.17). AI-Vp/Vs crossplot. The schematic, but qualitatively
Pore fluid sensitivity in reservoir sandstones is correct illustration in Fig. 18.18 demonstrates why it is
highly affected by reservoir heterogeneity and sand- important to include these geological factors when
stone microstructure, and it is therefore important to analysing the rock physics properties and seismic
include these geological factors in the rock physics fluid sensitivity in reservoir sandstones.
analysis. As already indicated, initial cement reduces Figure 18.19 shows an RPT including data from
the pressure and fluid sensitivity of the sandstones. two neighbouring wells penetrating Palaeocene sands
Figure 18.18 shows schematically the outline of a in the North Sea (Avseth et al. 2009). One well pene-
rock physics template, where calibrated rock physics trated a thick, turbiditic gas sand with a thin oil-leg,
models have been selected that fit local data observa- whereas the adjacent well penetrated a turbidite sand
tions (well log data or seismic inversion data) of vari- filled with oil. Furthermore, one of the wells is the
ous lithologies and pore fluids. The presence of same one as was used in the rock physics diagnostics
diagenetic quartz cement will move brine-saturated in Figs. 18.7 and 18.8, and in the depth trend model-
sandstone data in an AI versus Vp/Vs crossplot to an ling above (Fig. 18.15). It turns out that the sandstone
quality changes from one well to the other, and this
drastically distorts the fluid sensitivity to hydrocar-
3 bons. The gas-saturated Heimdal sands in well 1
Contact cement model show a small increase in acoustic impedance, while
the oil-saturated sands in well 2 show a significant
I drop in acoustic impedance. This drastic change in
2.5
sandstone quality over a short distance will also yield
Vp/Vs
a corresponding change in seismic signatures, see next

Sha section below.
2 V le m
o del
II 18.8 Seismic Reservoir Characterisation
III Brine sst model (2% cement) Using AVO
1.5 IV Gas sand model
4 6 8 10 12 About 30 years ago, William Ostrander, published a
AI (km/s*g/cm3) ground-breaking paper on offset-dependent reflectiv-
ity (Ostrander 1984). He showed that gas-saturated
Fig. 18.19 RPT of Vp/Vs versus acoustic impedance (AI) for sands capped by shales would cause an amplitude
target zone (Palaeocene) of two North Sea wells. Cluster I is the
variation with offset (AVO effect) in pre-stack seismic
cap rock shale in both wells, II comprises the brine sandstones in
both wells, III and IV are reservoir sandstones in well 1 filled data (Fig. 18.20). Shortly after, AVO technology
with oil and gas, respectively. V is the upper oil zone in well 2, became a commercial tool for the oil industry, quickly
with Vp/Vs higher than the brine sandstones, and AI lower than gaining in popularity as it was now possible to
the gas sandstones in well 1. This is counter-intuitive, and must
explain seismic amplitudes in terms of rock properties.
be explained by difference in sandstone quality, c.f. Fig. 18.18.
(Adapted from Avseth et al. 2009) The technique proved successful for hydrocarbon
474 P. Avseth
CDP gather
Stack section of interest CDP GATHER
CDP Offset (or angle
location of incidence)
Target horizon
Target horizon
Time Time
AVO response at target

horizon
Geological interpretation
0.15
0.1
Reflectivity
0.05
Shale 0
–0.05
–0.1
Sandstone –0.15
with gas –0.2
–0.25
0 5 10 15 20 25 30 35 40
angle of incidence
Fig. 18.20 Schematic illustration of the principles in AVO analysis. (From Avseth et al. 2005)
prediction in many areas of the world, yet in some approximation is given by Aki and Richards (1980),
cases it failed. The technique suffered from ambigu- assuming weak layer contrasts:
ities caused by lithological effects, tuning effects and
overburden effects. It turned out that even seismic 1 Δρ 1 ΔV p
Rð θ 1 Þ 1 4p2 V 2s þ
processing and acquisition effects could cause false 2 ρ 2 cos 2 θ V p
AVO anomalies. Application of AVO analysis was ΔV S
therefore reduced. However, in many of these cases 4p2 V 2S ; ð18:21Þ
VS
it was incorrect use of the technique that was respon-
sible for the failure, not the technique itself. In the last where:
decade we have observed a revival of the AVO tech-
nique. This is due to the improvement of 3D seismic sin θ1
p¼ θ ¼ ðθ1 þ θ2 Þ=2 θ1
technology, better pre-processing routines, more fre- V P1
quent shear-wave logging that can better constrain
AVO modelling, improved understanding of rock Δρ ¼ ρ2 ρ1 ρ ¼ ðρ2 þ ρ1 Þ=2
physics properties, greater data capacity, more focus ΔV P ¼ V P2 V P1 V P ¼ ðV P2 þ V P1 Þ=2
on cross-disciplinary aspects of AVO and, last but not
least, more awareness among the users of the potential ΔV S ¼ V S2 V S1 V S ¼ ðV S2 þ V S1 Þ=2
pitfalls.
The technique provides the seismic interpreter with In the formulae above, p is the ray parameter, θ1 is
information about pore fluids and lithologies, which the angle of incidence, and θ2 is the transmission
complements the conventional interpretation of seis- angle. VP1 and VP2 are the P-wave velocities above
mic facies, stratigraphy and geomorphology. The and below a given interface, respectively. Similarly,
reflection coefficient for plane elastic waves as a func- VS1 and VS2 are the S-wave velocities, while ρ1 and
tion of reflection angle at a single interface, also ρ2 are densities above and below this interface.
referred to as the AVO response, is described by the The approximation given by Aki and Richards can
complicated Zoeppritz equations (Zoeppritz 1919). be further approximated (Shuey 1985):
For analysis of P-wave reflections, a well-known
RðθÞ Rð0Þ þ G sin 2 θ contrast in Vp/Vs ratio (i.e. the AVO gradient),
whereas ε is affecting larger offsets. Often during
þ F tan 2 θ sin 2 θ ;

ð18:22Þ AVO analysis, anisotropy is ignored, in particular in
where shallow sediments where poorly compacted shales are
assumed to be isotropic. As we showed in Fig. 18.16,

the Tertiary shales in the North Sea have some anisot-
1 ΔV P Δρ
Rð0Þ ¼ þ ; ropy, which was also demonstrated by Golikov et al.
2 VP ρ
(2013), and we will later show a sensitivity study of
1 ΔV P V 2 Δρ

ΔV S
the effect of anisotropy on the AVO signatures of a
G¼ 2 2S þ2 North Sea Tertiary oil and gas field. Golikov et al.
2 VP VP ρ VS
(2013) also demonstrated the effect of anisotropy on
Δρ 1 2V S 2

4V S 2 ΔV S elastic parameters and associated AVO signatures
¼ Rð 0 Þ þ ; due to interbedding of sands and shales within a
ρ 2 VP2 VP2 VS
reservoir.
and The most common and practical way to do AVO
analysis of seismic data is to make crossplots of the
1 ΔV P zero-offset reflectivity (R(0)) versus the AVO gradient
F¼ :
2 VP (G), assuming isotropic reservoir and cap-rock. These
attributes are estimated from pre-stack seismic gathers
This form can be interpreted in terms of different using simple least-square regressions, according to
angular ranges, where R(0) is the normal incidence Eq. (18.23). Often it is assumed that calibrated near
reflection coefficient, G describes the variation at stack seismic data, where the near offset traces have
intermediate offsets and is often referred to as the been stacked together, is representative of the zero
AVO gradient, whereas F dominates the far offsets, offset reflectivity. Furthermore, it can be assumed
near critical angle. Normally, the range of angles that the difference between the far stack seismic
available for AVO analysis is less than 30 –40 . (where the far offset traces have been stacked
Therefore, we only need to consider the two first together) and the near stack seismic is a scaled version
terms, valid for angles less than 30 (Shuey 1985): of the AVO gradient. Hence, AVO crossplots can be
made directly from near and far stack seismic sections.
RðθÞ Rð0Þ þ G sin 2 θ ð18:23Þ Brine-saturated sands interbedded with shales,
Transverse isotropy in the cap-rock shale can affect situated within a limited depth range and at a particular
the AVO gradient significantly. During feasibility locality, normally follow a well defined “background
studies of AVO, it is important to investigate this trend” in AVO crossplots. A common and recom-
effect. The P-wave reflectivity in case of transverse mended approach in qualitative AVO crossplot analy-
isotropy in the cap-rock can be expressed as the sum of sis is to recognise the “background” trend and then
the isotropic reflectivity and the anisotropic reflectiv- look for data points that deviate from this trend. Devia-
ity (Kim et al. 1993): tions from the background trend in an AVO crossplot
may be indicative of hydrocarbons.
RPP ðθÞ ¼ RIPP ðθÞ þ RAPP ðθÞ ð18:24Þ Rutherford and Williams (1989) suggested a classi-
fication scheme of AVO responses for different types
where the isotropic term is the same as Eq. (18.21) or of gas sands (Fig. 18.21). They defined three AVO
its approximations (Eqs. 18.22 or 18.23), and the classes based on where the top of the gas sands will
anisotropic term is given by: locate in an R(0) versus G crossplot. The crossplot is
split up into four quadrants. In a crossplot with R(0)
RAPP ðθÞ ¼ 0:5δ1 sin 2 ðθÞ 0:5ε1 along x-axis and G along y-axis, the 1st quadrant is
2 2
where R(0) and G are both positive values (upper right
sin ðθÞ tan ðθÞ: ð18:25Þ quadrant). The 2nd is where R(0) is negative and G is
positive (upper left quadrant). The 3rd is where both
As we see from this expression, one of the Thomsen
parameters, δ, is affecting the same offset range as the R(0) and G are negative (lower left quadrant). Finally,
476 P. Avseth
the 4th quadrant is where R(0) is positive and G is with weak intercept, but strong negative gradient.
negative (lower right quadrant). The AVO classes These can be hard to see on the seismic, because
must not be confused with the quadrant numbers. they often yield dim-spots on stacked sections. Class
Class I plots in the 4th quadrant with positive R(0) III is the AVO category that is normally associated
and negative gradients. These represent hard events with bright spots. These plot in the 3rd quadrant in
with relatively high impedance and low Vp/Vs ratio R(0)-G crossplots, and are associated with soft sands
compared to the cap-rock. Class II represents sands saturated with hydrocarbons.
Ross and Kinman (1995) separated between a class
IIp and class II anomaly. Class IIp has a weak, but
G positive intercept and a negative gradient, causing a
Class IV
polarity change with offset. This class will disappear
on full stack sections. Class II has a weak, but negative
intercept and negative gradient, hence no polarity
R(0) change. This class may be observed as a negative
amplitude on a full offset stack.
Class III Castagna and Swan (1997) extended the classifica-
Class I
tion scheme of Rutherford and Williams to include a
Class II Class IIp
4th class, plotting in the 2nd quadrant. These are
relatively rare, but occur when soft sands with gas
are capped by relatively stiff cap-rock shales charac-
terized by Vp/Vs ratios slightly higher than in the sands
Fig. 18.21 Rutherford and Williams AVO classes, originally
defined for gas sands (class I, II and III), along with the added (i.e. very compacted or silty shales).
classes IV (Castagna and Swan 1997) and IIp (Ross and Kinman Figure 18.22 shows a seismic line with two wells
1995). Figure is adapted from Castagna and Swan (1997) indicated as black lines (same wells that are analysed
Well 1 Well 2
5
1800 Near
TWT
2000 0
2200
-5
200 400 600 800 1000 1200
CDP
5
1800 Far
TWT
2000 0
2200
-5
200 400 600 800 1000 1200
CDP
5
1800 Far-Near
TWT
2000 0
2200
-5
200 400 600 800 1000 1200
CDP
Fig. 18.22 Seismic sections intersecting two wells (same wells highlights the gradient anomaly of the gas and oil discovery of
that are analysed in Fig. 18.19), including near stack, far stack, Well 1. The red ellipse highlights an adjacent oil discovery of
and the estimated far-near stack. The blue square indicates the Well 2. (Adapted from Avseth et al. 2009)
window where a background trend is defined. The yellow ellipse
10 the background Vp/Vs ratio, χ, and the slope of the

Vp-Vs relationship of the modelled shale trend, m:
5 m
κ¼ ð18:27Þ
χ
Far-Near
Commonly, m is selected to be the slope of the

0
‘Mudrock Line’ (Castagna et al. 1985), where
m ¼ 1.16. Similarly, the background Vp/Vs ratio (χ)
Oil is often set equal to 2, giving a κ value of 0.58.
-5
However, with modelled shale trends, we can estimate
Gas more realistic depth-dependent values of κ and get a
-10 better control on the expected background trend to be
-10 -5 0 5 10 used in the fluid factor (Avseth et al. 2008). Both gas-
Near and oil-filled sands will cause negative fluid factor
anomalies relative to the background brine trend.
Fig. 18.23 Intercept (Near) versus gradient (i.e. Far-Near) for
the seismic stack section in Fig. 18.22. Only data from the However, chemical compaction has a significant
selected polygons are included. The blue points represent the impact on the absolute value of the fluid factor,
background trend right above the target, the yellow points rep- which is directly related to the fact that the fluid
resent data from the gas (and oil) discovery in Well 1 in
sensitivity decreases with depth and increasing rock
Fig. 18.22, whereas the red data points represent data from the
oil discovery in Well 2 in Fig. 18.22. (Adapted from Avseth stiffness (Fig. 18.24).
et al. 2009) An improved understanding of rock physics depth
trends in sands and shales helps us to better understand
how AVO signatures change with depth. By conduct-
ing AVO depth trend modelling, we can verify the
in Fig. 18.19). One well penetrated a thick, turbiditic
feasibility of AVO analysis as a function of burial
gas sand with a thin oil-leg, whereas the adjacent well
depth and diagenesis, and thereby extrapolate to
penetrated a turbidite sand filled with oil. However, it
other burial depths. This can be a useful task in explo-
turned out that the AVO anomaly of the oil sand is
ration or appraisal, where we want to extend our
stronger than for the gas discovery (Fig. 18.23), and
search to slightly deeper or shallower prospects adja-
the oil sands show a negative bright near stack
cent to existing discoveries. This is demonstrated in
response and a class III AVO anomaly, whereas the
Fig. 18.25 where we have done AVO two-layer mod-
gas sands show a dim (i.e. close to zero) near stack
eling of reservoir sandstone capped by shale at three
response and a class II AVO anomaly. These observa-
different depths for the case shown in Fig. 18.15. (1) at
tions are counter-intuitive if the reservoir sands are
1,600 m where only mechanical compaction is occur-
similar. Avseth et al. (2009) demonstrated the impor-
ing; (2) at 2,000 m in a transition zone where initial
tance of rock texture, in particular cement volume, as
chemical compaction has taken place; and (3) at
well as net-to-gross, and how these geological factors
2,400 m where the sandstones are moderately to well
affected the seismic signatures of hydrocarbons within
cemented. In this reflectivity modelling, we include
these sands.
both isotropic (whole lines) and anisotropic (dashed
The deviation from the background trend in an
lines) AVO modelling. In the latter case we assume the
AVO crossplot can be quantified in terms of the fluid
cap rock shale to be transversely isotropic, with the
factor (Smith and Gidlow 1987, Fatti et al. 1994) as
Thomsen parameters estimated from the shale model-
follows:
ling shown in Fig. 18.16. We perform the modelling
8 for water-, oil- and gas-filled reservoir. We conclude
ΔF ¼ Rð 0 Þ κ ½ Rð 0 Þ G ð18:26Þ that the effect of anisotropy is minimal in this case,
5
and can be ignored in the further AVO analysis. The
where κ is a constant at a given depth, depending on effect of burial, on the other hand, is dramatic. Within
478 P. Avseth
Brine Brine Brine 0.1

Oil Oil Oil
1400Gas Gas Gas 0.08
Shale 0.06
Shale
1600
0.04
1800 0.02
Depth (m)
0
2000
–0.02
2200
–0.04
–0.06
2400
–0.08
2600
–0.1
2000 4000 6000 8000 1000012000 1.5 2 2.5 3 –0.5 0 0.5
Acoustic Impedance (km/s*kg/m3) Vp/Vs Fluid Factor
Fig. 18.24 Seismic depth trends and expected fluid factor. than the modelled window, beyond 2,600 m. The gas and oil
Note that both oil and gas sands show large negative fluid factor zone in the well log data also show a relatively strong fluid
values down to about 2,300–2,400 m burial depth, whereas brine factor anomaly, as expected from the depth trends. (Adapted
sands show weak or positive fluid factor values for all depth from Avseth et al. 2008)
ranges. Gas sands seem to give negative fluid factor even deeper
a few hundred metres, the AVO signatures change can be derived from near- and far-stack amplitudes
significantly, due to compaction. The oil response using the formula:
can be quite similar to either a slightly deeper gas
ΔF ¼ Near η ðFar Near Þ ð18:28Þ
response or a slightly shallower water response. In
general, we go from a class IV to III AVO regime Here, η is not the same as κ in the earlier formula-
for hydrocarbons at the shallowest level, to a class II to tion of the fluid factor, since Far-Near is not exactly
III in the middle level, and finally to a class I to II at the same as the AVO gradient. However, the two-term
the deeper level, following the AVO classification AVO makes a linear relationship between the Far-
system of Rutherford and Williams (1989), see Near and the AVO gradient, and therefore η and κ
Fig. 18.21. This shows how important it is to honour can also be easily correlated. Assuming the far stack
burial history and compaction during AVO analysis. to be around 30 and the near stack to be at 0 (nor-
We can now estimate the fluid factor attribute for mally the far stack will be representative for slightly
the seismic line shown in Fig. 18.22. The fluid factor lower angles than 30 , whereas near stack will be
representative for angles slightly higher than 0 ) we

obtain the following approximate relationship
between the gradient and the Far-Near:
Far Near ¼ Rð30Þ Rð 0 Þ

2
¼ G sin ð30Þ ¼ G 0:25 ð18:29Þ
Solving for the background trend, that is when

ΔF ¼ 0, we obtain:
20κ
Near ¼ ðFar Near Þ ð18:30Þ
8 5κ
and hence:
20κ
η¼ ð18:31Þ
8 5κ
The slope of the background trend in the Near versus
Far-Near stack is 1/η, and this shows that we can in fact
estimate the expected background trend for uncalibrated
(but offset balanced) range limited stacks using the
modelled shale trends as illustrated above (Fig. 18.16).
This can be a useful task to verify the correct amplitude
balancing between near- and far-stack data during AVO
crossplot analysis.
Based on the shale model trends, we estimate the
value of η to be of the order 2.5. This means the
background trend of the shale in the AVO crossplot
is relatively flat, equaling 0.4, i.e., Far-Near ¼
0.4Near. We observe indeed that the shale back-
ground trend is relatively flat (blue cloud in
Fig. 18.23). Using this background trend, the resulting
fluid factor attribute in Fig. 18.26 shows that both the
gas and the oil discoveries stand out as strong fluid
factor anomalies, whereas the background data is rela-
Fig. 18.25 P-to-P-wave AVO reflectivity modelling at three
different depth levels, with reference to Figs. 18.15 and 18.16: tively weak.
(1) At 1,600 m burial depth we obtain AVO class IV for brine We also estimate the difference between intercept
sands and class III for hydrocarbons. Note that the anisotropy and gradient, which has been shown to be close to
effect (dashed lines) is actually changing the sign of the gradient
ΔVs/Vs (Fatti et al. 1994, Avseth et al. 2005). This
for brine sands from slightly positive to slightly negative. (2) At
2,000 m burial depth, we observe a weaker fluid sensitivity, and attribute should not be significantly affected by pore
the brine sands have almost zero near-stack response, but there fluids, but very sensitive to lithology and cementation.
are significant negative gradients for all fluid scenarios. (3) At Figure 18.26 shows the fluid factor crossplot and line
2,400 m burial depth, the brine sandstones show a class I AVO
superimposed with the ΔVs/Vs attribute for the seismic
signature, whereas oil- and gas-filled sandstones show a more
class II AVO response. For all cases, the anisotropy starts line intersecting both well 1 and well 2. As expected,
becoming significant beyond about 30 angle of incidence, and both the gas and the oil discoveries show strong fluid
should not impact our 2-term AVO analysis for this case factors, whereas only well 1 (gas discovery) shows a
strong change in ΔVs/Vs at the top of the reservoir.
This tallies with the observations from the well log
480 P. Avseth
Fluid factor DVs/Vs

10 10
10 20
5 5
5 10
Far-Near
Far-Near
0 0 0 0
-5 -10
-5 -5
Fluid trend -10 Cement -20

-10 -10
-10 -5 0 5 10 -10 -5 0 5 10
Near Near
Well 1 Well 2
1900
Fluid factor 5
1950
TWT (ms)
2000 0
2050 -5
2100
200 400 600 800 1000 1200
CDP
1900
DVs/Vs (cement?) 20
1950
TWT
2000 0
2050
-20
2100
200 400 600 800 1000 1200
CDP
Fig. 18.26 AVO crossplots with fluid factor and ditto with show strong fluid factor values at both well locations, whereas
ΔVs/Vs as coloured attribute, for the seismic section intersecting the ΔVs/Vs attribute shows a much stronger contrast at well 1
the two wells in Fig. 18.22. The fluid factor attribute represents than at well 2. Avseth et al. (2009) documented that the upper-
the deviation from the local background shale trend. The ΔVs/Vs most part of Heimdal sandstones in well 2 are less cemented
attribute reflects rock stiffness, and is expected to be an appro- than the Heimdal sandstones in well 1 (cf. RPT observations in
priate attribute to detect initial quartz cement in the Heimdal Fig. 18.19), and this can also explain the different AVO
Formation sandstones. The resulting AVO attributes (lower) responses at the two well locations
data, and also with the rock physics estimation of local change in rock texture. With two seismic para-
relatively low cement volume at the top of the Heim- meters (R(0) and G) it is possible to discriminate the
dal sands in well 2 (see Avseth et al. 2009). It further fluid effects from the lithological effects.
explains why we have such drastically different AVO A problem when interpreting AVO crossplots is
signatures in the two wells, while the counter-intuitive that a given point in the crossplot does not correspond
AVO anomalies for oil and gas are explained by the to a unique combination of rock properties. Many
combinations of rock properties will yield the same

Clay R(0) and G. Moreover, due to natural variability in
trend geological and fluid parameters, one given geological
scenario may span a relatively large possible outcome
area in the AVO crossplot, not just a discrete point.
Hence, a hydrocarbon-like AVO response might occa-
Fluid sionally result from a brine-associated reflection, and
trend hydrocarbon-saturated sands might not always pro-
Cemented duce an anomalous AVO response; this problem was
trend seen in the RPT crossplots in Fig. 18.18. One way to
resolve this uncertainty is to create probability cross-
plots of various categories of lithology and pore fluid
scenarios. These can be based on statistical analysis of
well log data and/or rock physics models (Avseth
2000, Avseth et al. 2001a, b; Mukerji et al. 2001).
Fig. 18.27 AVO contour plots or probability density functions Each category is plotted as a “contour map”, almost
(pdfs) for main facies and fluids in the Glitne field. Only the iso- like topographical maps (Fig. 18.27). Here, the
probability contours of 50% and larger are included for each “mountain tops” represent the most likely location of
group. The R(0) and G pdfs nicely separate the three facies
groups, but there are significant overlaps. (Adapted from Avseth a given class. It is very important to be aware that the
et al. 2001a) contours of different facies and fluids are overlapping
each other. This implies that an observed set of R(0)
and G can represent more than one category. This is
Well # 1 CDP 1 CDP 1

Well # 2 Well # 3 CDP 1
off 0 323 726 1210 1694 2177
off 0 323 726 1210 1694 2177 off 0 258 581 903 1226 1613 2000
1900 1900 1900
2000 2000 2000

2000 2000 2000
Top facies IIb
Top facies III
Top_heimdal
Top_heimdal 2100 2100 2100

2100 Top_heimdal 2100 2100
Top facies IIa
2200 2200 2200 2200 2200 2200
2300 2300 2300

2300 2300 2300
Reflectivity Reflectivity Reflectivity
0.100
0.100 Well # 1 Well # 2 0.100 Well # 3
0.050
0.050 PR
0.050
PR 0
–0.050 0
0
PR
–0.100
–0.050 –0.050
–0.150
–0.100 –0.200 –0.100

Offset 0 500 1000 1500 2000 2500 (m) Offset 0 500 1000 1500 2000 (m) Offset 0 500 1000 1500 2000 2500 (m)
angle 0 7 14 21 28 34 (deg) angle 0 8 15 22 29 (deg) angle 0 7 14 20 26 32 (deg)
Fig. 18.28 Real CDP gathers (upper left), synthetic CDP gath- feeder-channel (cemented sandstone filled with brine), to the
ers (upper right), and AVO curves (lower) for three wells lobe-channel (unconsolidated sand filled with oil), and to the
penetrating the Glitne turbidite system (Fig. 18.1) at different lobe margin (interbedded sand-shale and shaly sandstone).
locations. Note the drastic change in AVO signature from the (Adapted from Avseth et al. 2001a)
482 P. Avseth
one reason why AVO analysis, used alone, can give analysis of CDP gathers (see Fig. 18.28) at the three
misleading results. In addition, these crossplots are well locations depicted in Fig. 18.1. The dramatic
often affected by noise in the seismic data. Neverthe- variability in the lateral facies distribution, going
less, as we see in Fig. 18.27, we often observe that we from a relatively proximal feeder-channel environ-
need both intercept (R(0)) and gradient (G) in order to ment to a relatively distal lobe and lobe margin envi-
separate facies and fluid types from seismic data. Or ronment, has great impact on the seismic signatures in
put another way, we often need to do AVO to discrim- this turbidite system.
inate lithology from fluids on seismic data. The next step is facies and pore fluid prediction
Avseth et al. (2001a) applied the contour plots in from 3D seismic data. 3D AVO inversion is performed
Fig. 18.27 to conduct an AVO analysis and predict on the turbidite system using Hampson-Russell’s
seismic lithofacies and pore fluids from the pre-stack AVO software. Again, we focus only on the horizon
seismic amplitudes in the Glitne Field which are representing the top of the system (Top Heimdal).
depicted in the amplitude map in Fig. 18.1. As demon- Figure 18.29 shows the three dimensional topography
strated in Fig. 18.1, the depositional trends in the (in two-way traveltime) of this seismic horizon, where
Glitne turbidite system do affect the seismic ampli- the geometries of the feeder-channel and the lobe
tudes. This is also confirmed by deterministic AVO
Interbedded
sand-shale
Oil sands
Turbidite system outline
Feeder-channel
Brine sands
Shale
Lobe-structure
Fig. 18.31 Lithofacies prediction beneath Top Heimdal seis-
100 ms
mic horizon, in the Glitne Field. Note the prediction of most
1 km
likely oil sands in the top lobe structure, as well as in the upper
part of the feeder channel. The latter is not realistic, knowing
that a well is penetrating brine-filled feeder-channel sands right
Fig. 18.29 Three-dimensional seismic topography of Top outside the amplitude map. However, as we see in the AVO
Heimdal horizon (two-way travel time). The depositional geom- contour plots in Fig. 18.27, we expect some ambiguity and
etry of a feeder-channel and fan lobe is outlined (compare to overlap between brine sands and oil sands. (Adapted from
Fig. 18.1). (Adapted from Avseth et al. 2001a) Avseth et al. 2001a)
Relatively high R(0)

Relatively large negative
(blue)
gradient G (yellow)
Relatively low R(0)

Relatively small
(yellow)
gradient (blue)
Fig. 18.30 Intercept (R(0)) and AVO gradient (G) estimated along Top Heimdal horizon in the Glitne Field. (Adapted from Avseth
et al. 2001a)
structure are outlined. The inversion gives us R(0) and the submarine fan, the most likely facies is predicted
G over the whole area across this horizon slice. Fig- to be dominated by shale. The sands are mainly pre-
ure 18.30 shows the intercept R(0) and the gradient G. dicted in the channel and lobe areas while oil is pre-
These plots allow us to predict the most likely seismic dicted in the structurally highest areas of the sand
lithofacies underlying this horizon. This is done by deposits. The depositional patterns predicted from
combining the R(0) and G inverted from the seismic the AVO analysis agree well with the expected facies
with the R(0)–G contour plots, also referred to as associations in the Glitne submarine fan. This case
probability density functions (pdfs), which are derived example demonstrates that AVO and quantitative seis-
from well log data. Before we can do this, however, mic interpretation can be particularly useful in geolog-
the inverted parameters must be calibrated to the well ical settings where facies prediction is not easily
log values (see Avseth et al. 2005 for further details). obtained from conventional seismic interpretation.
We apply the Mahalanobis distance method to cal- Finally, a case example from the Barents Sea will
culate the most likely facies group and pore fluid from be shown, where rock physics properties and asso-
the seismic data (Avseth et al. 2005). The results are ciated AVO signatures may change rapidly and in a
shown in Fig. 18.31. We predict oil-saturated sands in complex manner due to extensive tectonic uplift epi-
the lobe area where the lobe is structurally highest. sodes. Avseth et al. (2003) introduced a methodology
The rest of the lobe area is most likely water-saturated to constrain AVO classification by rock physics depth
according to the prediction. Furthermore, we predict trends. However, they only considered unconsolidated
oil-filled sands in the upper feeder-channel. Outside sediments in continuously subsiding basins. Taking
Fig. 18.32 Well log data (including Vp, density and gamma ical compaction domain, i.e. transition from sediments to rocks.
ray) from well 7201/2-1 (Myrsildre), which was a well with oil The corresponding change in a porosity-velocity crossplot is
shows in the Kolmule Fm. Note the velocity and density depth shown in the upper right corner, showing the elastic stiffening
trends. However, there is a big jump in velocity as we go from effect of chemical compaction (CC) of rocks compared to
Torsk Fm to Kolmule Fm; this corresponds to a change in mechanically compacted (MC) sediments. (Adapted from
diagenesis going from mechanical compaction domain to chem- Lehocki et al. 2012)
484 P. Avseth
into account diagenesis and tectonic events, we can mechanical to chemical compaction domain, and is
predict compaction trends and associated seismic in agreement with observations made by Storvoll
velocities in areas with more complex burial history et al. (2005) and Thyberg et al. (2009).
involving both mechanical and chemical compaction, Rock physics models are compared and calibrated
as well as uplift episodes and corresponding erosion. to the observed velocities, for both domains. The
The resulting rock physics trends help us to better Hertz-Mindlin contact theory is used for the loose
constrain AVO inversion and classification under sediments above the onset of cement, whereas the
such circumstances. Dvorkin-Nur contact cement model was utilised to
Lehocki et al. (2012) proposed a methodology for model the velocities of sandstones in the chemical
statistical classification of fluids and facies in areas compaction domain. The rock physics models are
with tectonic uplift based on rock physics depth furthermore adjusted in depth until we obtain an
trends, and they demonstrated the methodology on optimal match between observed sonic well log data
seismic AVO data from the Loppa High area in the and modelled velocities. A maximum burial depth of
Barents Sea, where there has been significant tectonic c. 2,500 m is estimated, implying that the net uplift is
uplift. The main objective of that study was to perform around 1,300 m. We therefore expect the reservoirs to
seismic reservoir prediction and hunt for Upper Creta- have been buried at temperatures high enough to initi-
ceous Kolmule Formation sands with good reservoir ate chemical compaction.
quality in a structural setting near the Loppa High. AVO data, including intercept and gradient, were
Furthermore, the goal was to do both pre- and post- extracted from a window around the seismic horizon
drill predictions of a prospect. A neighbouring well representing the Top Kolmule sst (Fig. 18.33). Based
(Fig. 18.32) was used for local calibration. This well is on the rock physics burial trends, we created AVO pdfs
located down-flank of the structural high of interest, for different lithofacies and fluids representative for the
and Kolmule Formation sands are not encountered in target burial depth (Fig. 18.34). This was done both
it. Note, however, the sudden jump in velocity as we pre- and post-drill of the prospect, with slightly differ-
go from Torsk Formation shales to Kolmule Forma- ent calibrations. The AVO pdfs were then used to
tion shales. The gamma ray log and density log show classify the calibrated AVO data. The resulting maps
only small changes at this interface. Regional work of lithofacies and fluids are shown in Fig. 18.35. The
has indicated that this is the transition from the prospect was drilled and the presence of gas in the
Fig. 18.33 AVO data from the Loppa High near the Skalle prospect. Intercept (left) and gradient (right). Note the negative
intercept and gradients around the prospect well location. (Adapted from Lehocki et al. 2012)
Fig. 18.34 Calibrated intercept and gradient with juxtaposed done prior to the prospect well, is shown to the left; the calibra-
isoprobability contours (P>0.5) of different facies and fluids, tion done after the prospect well was drilled is shown to the
including cemented sandstones (c.sand), cemented shaly sand- right. Note the slight rotation of the data and the location of the
stones (c.shl.snd), loose sands (sand), and shales (shl); with isoprobability ellipses. (Adapted from Lehocki et al. 2012)
brine (B), oil (O) or gas (G) in the sandy facies. The calibration
Fig. 18.35 AVO classification results; Lithofacies (CSS fication based on well 7120/1-2 is giving quite good results. We
Cemented shaly sandstones, CS Cemented sandstone, Snd predict somewhat less shales, and some more of the loose sands,
Loose sand, Shl Shale) to the left and pore fluids to the right. with information from the new well. The well 7120/1-2 had oil
Pre-drill of 7120/2-3 in the upper two subplots, and post-drill of shows, whereas well 7120/2-3 encountered a gas-filled hetero-
this well in the lower two subplots. Note that the fluid distribu- lithic Kolmule sandstone, as predicted both pre- and post-drill.
tions are very similar, indicating that the calibration and classi- (Adapted from Lehocki et al. 2012)
486 P. Avseth
Kolmule sandstone was confirmed for the prospect. Further Reading

The workflow was repeated post-drill to obtain an
even better calibration of the rock physics depth trends, Aki, K. and Richards, P.G. 1980. Quantitative Seismology:
and the results are also shown in Fig. 18.35; as we see Theory and Methods. San Francisco: W.H. Freeman, 948 pp.
Anselmetti, F. and Eberli, G.P. 1997. Sonic velocity in carbon-
the differences are quite small, demonstrating the ate sediments and rocks. In: Palaz, I. and Marfurt, K.J. (eds.),
benefit of the AVO information calibrated by a nearby Carbonate Seismology. Society of Exploration Geophysi-
well, prior to drilling this prospect. cists, Houston, TX, pp. 53–73.
Avseth, P. 2000. Combining rock physics and sedimentology for
seismic reservoir characterization of North Sea turbidite
systems. Ph.D. thesis, Stanford University, 200 pp.
Avseth, P., Dræge, A., van Wijngaarden, A.-J., Johansen, T.A.
18.9 Conclusions and Jørstad, A. 2008. Shale rock physics and implications for
AVO analysis: A North Sea demonstration. The Leading
In this chapter we have demonstrated how we can Edge 27, 788–797.
apply rock physics tools to relate geological para- Avseth, P., Dvorkin, J., Mavko, G. and Rykkje, J. 2000. Rock
physics diagnostic of North Sea sands: Link between micro-
meters and trends to seismic properties, and thereby structure and seismic properties. Geophysical Research Let-
predict rock and fluid characteristics from seismic ters 27, 2761–2764.
amplitude data. The focus of this chapter has been to Avseth, P., Flesche, H. and van Wijngaarden, A.-J. 2003. AVO
present a workflow for quantitative seismic interpreta- classification of lithology and pore fluids constrained by rock
physics depth trends. The Leading Edge 22(10), 1004–1011.
tion during petroleum exploration. Rock physics ulti- Avseth, P., Johansen, T.A., Bakhorji, A. and Mustafa, H.M.
mately serves as a toolbox for lithology and fluid 2014. Rock physics modeling guided by depositional and
substitution, where we can understand local well log burial history in low-to-intermediate-porosity sandstones.
observations and extrapolate to certain expected “what Geophysics 79, D115–D122.
Avseth, P., Jørstad, A., van Wijngaarden, A.-J. and Mavko, G.
if” scenarios, either in terms of depositional or com- 2009. Rock physics estimation of cement volume, sorting
pactional trends, or in terms of changes in pore fluids. and net-to-gross in North Sea sandstones. The Leading Edge
The models can also be applied to predict or classify 28, 98–108.
reservoir parameters from sonic and seismic data. We Avseth, P., Mukerji, T., Jørstad, T., Mavko, G. and Veggeland, T.
2001a. Seismic reservoir mapping from 3-D AVO in a North
have demonstrated our cross-disciplinary approach on Sea turbidite system. Geophysics 66, 1157–1176.
well log and seismic data from selected areas on the Avseth, P., Mukerji, T. and Mavko, G. 2005. Quantitative Seis-
Norwegian shelf where reservoir sands are prone to mic Interpretation – Applying Rock Physics Tools to Reduce
change drastically both because of facies variability, Interpretation Risk. Cambridge University Press, Cambridge,
376 pp.
diagenetic alterations and tectonic impact. We have Avseth, P., Mukerji, T., Mavko, G. and Dvorkin, J. 2010. Rock
also demonstrated the importance of understanding physics diagnostics of depositional texture, diagenetic
the rock physics and seismic properties of shales alterations and reservoir heterogeneity in high porosity sili-
during seismic reservoir prediction using AVO ciclastic sediments and rocks - A review of selected models
and suggested workflows. Geophysics 75, A31–A47.
analysis. Avseth, P., Mukerji, T., Mavko, G. and Tyssekvam, J.A. 2001b.
Rock physics and AVO analysis for lithofacies and pore fluid
Acknowledgements Thanks to Prof. Gary Mavko and Prof. prediction in a North Sea oil field. The Leading Edge 20, 429.
Tapan Mukerji at Stanford University for collaboration over a Batzle, M. and Wang, Z. 1992. Seismic properties of pore fluids.
long period and contributions to the work presented here. Geophysics 57, 1396–1408.
Thanks to Arild Jørstad and Hans Oddvar Augedal at Lundin- Bjørlykke, K. 1998. Clay mineral diagenesis in sedimentary
Norway for geological input to the Palaeocene sands studied in basins – A key to the prediction of rock properties; Examples
this chapter; also thanks to Aart-Jan van Wijngaarden, Erik from the North Sea Basin. Clay Minerals 33, 15–34.
Ødegaard, Torbjørn Fristad, and Anders Dræge at Statoil for Bjørlykke, K. and Egeberg, P.K. 1993. Quartz cementation in
fruitful discussions and input to the work included in this chap- sedimentary basins. AAPG Bulletin 77, 1538–1548.
ter. I also acknowledge Prof. Ran Bachrach at Tel Aviv Univer- Box, G.E.P. and Draper, N.R. 1987. Empirical Model Building
sity, and Prof. Tor Arne Johannesen at University of Bergen, for and Response Surfaces. New York, NY: John Wiley and
valuable discussions. Finally, I would like to thank Ivan Lehocki Sons, 669 pp.
at Lehocki Geospace and Håvard Buran at Lundin-Norway for Brevik, I., Ahmadi, G.R., Hatteland, T. and Rojas, M.A. 2007.
contributions to the AVO case study in the Barents Sea, and to Documentation and quantification of velocity anisotropy in
Tor Veggeland at Tullow Oil Norge for valuable input on the shales using wireline log measurements. The Leading Edge
Glitne case study presented here. 26(3), 272–277.
Castagna, J.P., Batzle, M.L. and Eastwood, R.O. 1985. Relation- Extended Abstract C007). In 74th EAGE Conference and
ships between compressional-wave and shear-wave veloci- Exhibition, Copenhagen.
ties in clastic silicate rocks. Geophysics 50, 571–581. MacQuaker, J.H.S., Taylor, K.G. and Gawthorpe, R.L. 2007.
Castagna, J.P. and Swan, H.W. 1997. Principles of AVO cross- High-resolution facies analysis of mudstones: Implications
plotting. The Leading Edge 16, 337–342. for paleoenvironment and sequence stratigraphic interpreta-
Domenico, S.N. 1976. Effect of brine-gas mixture on velocity in tions of offshore ancient mud-dominated successions. Jour-
an unconsolidated sand reservoir. Geophysics 41, 882–894. nal of Sedimentary Research 77, 324–339.
Dræge, A. 2009. Constrained rock physics modeling. The Lead- Marcussen, Ø., Thyberg, B.I., Peltonen, C., Jahren, J., Bjør-
ing Edge 28, 76–80. lykke, K. and Faleide, J.I. 2009. Physical properties of
Dræge, A., Jakobsen, M. and Johansen, T.A. 2006. Rock physics Cenozoic mudstones from the northern North Sea: Impact
modeling of shale diagenesis. Petroleum Geoscience 12(1), of clay mineralogy on compaction trends. AAPG Bulletin
49–57. 93, 127–150.
Dvorkin, J. and Gutierrez, M. 2002. Grain sorting, porosity and Marion, D. 1990. Acoustical, mechanical and transport proper-
elasticity. Petrophysics 43, 185–196. ties of sediments and granular materials. Ph.D. thesis, Stan-
Dvorkin, J., Mavko, G. and Gurevich, B. 2007. Fluid substitu- ford University.
tion in shaley sediment using effective porosity. Geophysics Mavko, G., Mukerji, T. and Dvorkin, J. 2009. The Rock Physics
72, O1–O8. Handbook. Cambridge: Cambridge University Press, 511 pp.
Dvorkin, J. and Nur, A. 1996. Elasticity of high-porosity sand- Middleton, G.V. 1973. Johannes Walther’s law of the correla-
stones: Theory for two North Sea datasets. Geophysics 61, tion of facies. Geological Society of America Bulletin 84,
1363–1370. 979–988.
Fatti, J.L., Smith, G.C., Vail, P.J., Strauss, P.J. and Levitt, P.R. Mindlin, R.D. 1949. Compliance of elastic bodies in contact.
1994. Detection of gas in sandstone reservoirs using AVO Journal of Applied Mechanics 16, 259–268.
analysis: A 3-D seismic case history using the Geostack Mondol, N.H., Bjørlykke, K., Jahren, J. and Høeg, K. 2007.
technique. Geophysics 59, 1362–1376. Experimental mechanical compaction of clay mineral aggre-
Gassmann, F. 1951. Uber die elastizitat poroser medien. Vier gates – Changes in physical properties of mudstones during
der Natur Gesellschaft 96, 1–23. burial. Marine and Petroleum Geology 24(5), 289–311.
Giles, M.R., Indrelid, S.L., Beynon, G.V. and Amthor, J. 2000. Mukerji, T., Jørstad, A., Avseth, P., Mavko, G. and Granli, J.R.
The origin of large-scale quartz cementation: Evidence 2001. Mapping lithofacies and pore-fluid probabilities in a
from large data sets and coupled heat-fluid mass transport North Sea reservoir: Seismic inversions and statistical rock
modeling. In: Worden, R.H. and Morad, S (eds.), Quartz physics. Geophysics 66(4), 988–1001.
Cementation in Sandstones. International Association of Ødegaard, E. and Avseth, P. 2004. Well log and seismic data
Sedimentologists Special Publication 29, Blackwell, Oxford, analysis using rock physics templates. First Break 22, 37–43.
pp. 21–38. Ostrander, W.J. 1984. Plane-wave reflection coefficients for gas
Golikov, P., Avseth, P., Stovas, A. and Bachrach, R. 2013. Rock sands at non-normal angles of incidence. Geophysics 49,
physics interpretation of heterogeneous and anisotropic tur- 1637–1648.
bidite reservoirs. Geophysical Prospecting 61, 448–457. Peltonen, C., Marcussen, Ø., Bjørlykke, K. and Jahren, J. 2008.
Hamilton, E.L. 1956. Low sound velocities in high porosity Clay mineral diagenesis and quartz cementation in mud-
sediments. Journal of the Acoustical Society of America stones: The effects of smectite to illite transformation on
28, 16–19. rock properties. Marine and Petroleum Geology 26(6),
Han, D. 1986. Effects of porosity and clay content on acoustic 887–898.
properties of sandstones and unconsolidated sediments. Raiga-Clemenceau, J., Martin, J.P. and Nicoletis, S. 1988. The
Unpublished Ph.D. dissertation, Stanford University. concept of acoustic formation factor for more accurate
Hashin, Z. and Shtrikman, S. 1963. A variational approach to the porosity determination from sonic transit time data. Log
elastic behavior of multiphase materials. Journal of the Analyst 219, 54–60.
Mechanics and Physics of Solids 11, 127–140. Ramm, M. and Bjørlykke, K. 1994. Porosity/depth trends in
Hornby, B.E., Schwartz, L.M. and Hudson, J.A. 1994. Aniso- reservoir sandstones: assessing the quantitative effects of
tropic effective-medium modeling of the elastic properties of varying pore-pressure, temperature history and mineralogy,
shales. Geophysics 59, 1570–1583. Norwegian Shelf data. Clay Minerals 29, 475–490.
Johansen, T.A., Jakobsen, M. and Ruud, B.O. 2002. Estimation Raymer, L.L., Hunt, E.R. and Gardner, J.S. 1980 (July). An
of the internal structure and anisotropy of shales from bore- improved sonic transit time-to-porosity transform. Paper
hole data. Journal of Seismic Exploration 11, 363–381. presented in 21st Annual Logging Symposium. Society of
Kim, K.Y., Wrolstad, K.H. and Aminzadeh, F. 1993. Effects of Professional Well Log Analysis (SPWLA), Lafayette,
transverse isotropy on P-wave AVO for gas sands. Geophys- Louisiana.
ics 58, 883–888. Reading, H.G. and Richards, M. 1994. Turbidite systems in
Lehocki, I.L. and Avseth, P. 2010. Rock physics quantification deep-water basin margins classified by grain-size and feeder
of diagenetic cement (EAGE Extended Abstract P244). In system. AAPG Bulletin 78, 792–822.
72nd EAGE Conference and Exhibition, Barcelona. Reuss, A. 1929. Berechnung der fliessgrenzen von mischkris-
Lehocki, I.L., Avseth, P., Buran, H. and Jørstad, A. 2012. tallen. Zeitschrift fur Angewandte Mathematik und
Statistical AVO classification in areas with diagenesis and Mechanik 9, 49–58.
tectonic uplift - case examples from the Barents Sea (EAGE Ross, C.P. and Kinman, D.L. 1995. Non-bright spot AVO: Two
examples. Geophysics 60, 1398–1408.
488 P. Avseth
Rutherford, S.R. and Williams, R.H. 1989. Amplitude-versus- Walderhaug, O., Lander, R.H., Bjørkum, P.A., Oelkers, E.H.,
offset variations in gas sands. Geophysics 54, 680–688. Bjørlykke, K. and Nadeau, P.H. 2000. Modelling quartz
Ruud, B.O., Jakobsen, M. and Johansen, T.A. 2003. Seismic cementation and porosity in reservoir sandstones: examples
properties of shales during compaction (SEG Extended from the Norwegian continental shelf. In: Worden, R.H. and
Abstract). In SEG Annual Meeting, Dallas, TX. 4 pp. Morad, S. (eds.), Quartz Cementation in Sandstones. Inter-
Sams, M. and Andrea, M. 2001. The effect of clay distribution national Association of Sedimentologists Special Publica-
on the elastic properties of sandstones. Geophysical Pro- tion 29, Blackwell, Oxford, pp. 39–50.
specting 49, 128–150. Walker, R. 1978. Deep-water sandstone facies and ancient sub-
Shuey, R.T. 1985. A simplification of the Zoeppritz equations. marine fans: Models for exploration for stratigraphic traps.
Geophysics 50(4), 609–614. AAPG Bulletin 62, 932–966.
Smith, G.C. and Gidlow, P.M. 1987. Weighted stacking for rock Worden, R.H. and Morad, S. 2000. Quartz cementation in oil
property estimation and detection of gas. Geophysical Pro- field sandstones: A review of the key controversies. In:
specting 35, 993–1014. Worden, R.H. and Morad, S. (eds.), Quartz Cementation in
Storvoll, V., Bjørlykke, K. and Mondol, N.H. 2005. Velocity- Sandstones. International Association of Sedimentologists
depth trends in Mesozoic and Cenozoic sediments from the Special Publication 29, Blackwell, Oxford, pp. 1–20.
Norwegian Shelf. AAPG Bulletin 89(3), 359–381. Wyllie, M.J., Gregory, A.R. and Gardner, L.W. 1956. Elastic
Thomsen, L. 1986. Weak elastic anisotropy. Geophysics 51, wave velocities in heterogeneous and porous media. Geo-
1954–1966. physics 21, 41–70.
Thyberg, B., Jahren, J., Winje, T., Bjørlykke, K. and Faleide, J.I. Xu, S. and White, R.E. 1995. A new velocity model for clay-
2009. From mud to shale: rock stiffening by micro-quartz sand mixtures. Geophysical Prospecting 43, 91–118.
cementation. First Break 27, 53–59. Yin, H. 1992. Acoustic velocity and attenuation of rocks, isot-
Timoshenko, S.P. and Goodier, J.N. 1934. Theory of Elasticity. ropy intrinsic anisotropy, and stress induced anisotropy.
New York: McGraw-Hill, 416 pp. Ph.D. thesis, Stanford University.
Vernik, L. and Nur, A. 1992. Petrophysical classification of Zoeppritz, K. 1919. Erdbebenwellen VIIIB, Ueber Reflexion
siliciclastics for lithology and porosity prediction from seis- and Durchgang seismischer Wellen durch Unstetigkeitsflae-
mic velocities. AAPG Bulletin 76, 1295–1309. chen. Goettinger Nachrichten I, 66–84.
Voigt, W. 1910. Lehrbuch der Kristallphysik. Leipzig: Teubner.
Chapter 19
4D Seismic
Martin Landrø
19.1 Introduction The business advantage of using 4D seismic on a

given field is closely correlated with the complexity of
The term 4D seismic reflects that calendar time the field. Figure 19.1 shows a 3D perspective view of
represents the fourth dimension. A more precise term is the top reservoir (top Brent) interface for the Gullfaks
repeated seismic, because that is actually what is done: a Field. Since the oil is trapped below such a complex 3D
seismic survey over a given area (oil/gas field) is repeated surface, it is not surprising that oil pockets can remain
in order to monitor production changes. Time-lapse seis- untouched even after 10–15 years of production. Since
mic is another term used for this. For some reason, the 4D seismic can be used to identify such pockets, it is
term 4D seismic is most common, and we will therefore easy to understand the commercial value of such a tool.
use it here. It is important to note that if we repeat 2D However, if the reservoir geometry is simpler, the num-
surveys, it is still denoted 4D seismic according to this ber of untapped hydrocarbon pockets will be less and
definition. Recent examples of such surveys are repeated the business benefit correspondingly lower.
2D lines acquired over the Troll gas province. The first 4D seismic surveys were probably acquired
Currently there are three major areas where 4D in America in the early 1980s. It was soon realised that
seismic is applied. Firstly, to monitor changes in a heavy oil fields were excellent candidates for 4D seis-
producing hydrocarbon reservoir. This is now an mic. Heavy oil is highly viscous, so thermal methods
established procedure being used worldwide. So far, such as combustion or steam injection were used to
4D seismic has almost exclusively been used for clastic increase its mobility. Combustion describes a burning
reservoirs and only rarely for carbonate reservoirs. This process, which normally is maintained by injection of
is because carbonate reservoirs (apart from those in oxygen or air. As the reservoir is heated, the seismic P-
porous chalk) are stiffer, and the effect on the seismic wave velocity decreases, and this results in amplitude
parameters of substituting oil with water is far less changes between baseline and monitor seismic surveys.
pronounced. Secondly, 4D seismic is being used to The best known example is maybe the one
monitor underground storage of CO2. Presently, there published by Greaves and Fulp in 1987, for which
is a global initiative to decrease the atmospheric CO2 they received the award for the best paper in Geophys-
content, and one way to achieve this goal is to pump ics. They showed that such thermal recovery methods
huge amounts of CO2 into saline aquifers. A third could be monitored by repeating conventional 3D land
application of 4D seismic (with other geophysical seismic surveys (Fig. 19.2). These early examples of
methods) is the monitoring of geohazards (landslides, seismic monitoring of thermal recovery methods did
volcanoes etc.), however this will not be covered here. not immediately lead to a boom in the 4D industry,
mainly because they were performed on small and
very shallow onshore fields. The major breakthrough
for commercial 4D seismic surveys in the North Sea
M. Landrø (*)
NTNU Trondheim, Trondheim, Norway was the Gullfaks 4D study launched by Statoil in 1995.
e-mail: [email protected] Together with WesternGeco a pilot study was done
490 M. Landrø
750
700
650
600
550
500
1852
2100
N
2348
Intra-Ness Fm.
0 1 km Amundsen Fm.
2596
2692 Fault
341 375 400 425 450 475 500 525 550 575 600 625 650 670
Fig. 19.1 3D seismic image of the top reservoir interface (top Brent Group) at the Gullfaks Field. Notice the fault pattern and the
complexity of the reservoir geometry
W104 W101
S N
a 0.3
Holt
0.4
Palo
Pinto
b 0.3
Time (second)
0.4
c 0.3
0.4
20 40 60 80
CDP
Fig. 19.2 Shows clear 4D amplitude brightening (top Holt horizon) as the reservoir is heated. Also note the extended area (marked
by white arrows) for this amplitude increase. (Greaves and Fulp, Geophysics 52, 1987)
19 4D Seismic 491
Reflections from 492 mRKB

Pre-drilling
100 150 200 250 300
0.4
0.5
0.6
4 months after re-entry
150 200 250 300
0.4
0.5
0.6
6 months
150 200 250 300 350
0.4
0.5
0.6
Fig. 19.3 Seismic monitoring of the underground flow caused section (6 months after the drilling event) the areal extent of this
by drilling a deep Jurassic well in 1989 in the North Sea. The event has increased further. Figure provided by Dag O. Larsen,
new seismic event marked by the red arrow on the middle from his SEG-presentation in 1990
section is interpreted as a gas-filled sand body. On the lower
over the major northern part of the field, and the initial 0,25
interpretation performed shortly after this monitor sur-
Relative change in Vp
0,20
vey demonstrated a promising potential.
0,15
However, the first use of 4D seismic in the North
Sea was probably the monitoring of an underground 0,10
flow (Larsen and Lie 1990). In January 1989, when 0,05
drilling a deep Jurassic well (2/4-14), Saga Petroleum
0,00
had to shut the well by activating the BOP. The rig was
(0,05)
moved off location, and a relief well was spudded. 0,0 0,2 0,4 0,6 0,8 1,0
When the rig was reconnected 3 months later, the Water saturation
pressure had dropped, indicating a subsurface fluid
Fig. 19.4 Relative change in P-wave velocity versus water
transfer. The flow and the temperature measurements
saturation, estimated from a calibrated Gassmann model. Zero
indicated a leak in the casing at 1,334 mSS. A strong water saturation corresponds to 100% oil
amplitude increase at the top of a sand layer could be
observed on the time-lapse seismic data close to this – changes in layer thickness (compaction or
well (Fig. 19.3) – showing the fluid was gas. stretching).
The areal extent of this 4D anomaly increased from A critical part of all 4D studies is to link key
one survey to the next. After 4 months this amplitude reservoir parameters like pore pressure and fluid satu-
increase was also observed on shallower interfaces, ration to the seismic parameters. Rock physics
probably connected to gas migrating upwards to provides this link. Both theoretical rock physics
shallower sand layers. models and laboratory experiments are used as impor-
tant input to time-lapse seismic analysis. A standard
19.2 Rock Physics and 4D Seismic way of relating for instance P-wave seismic velocity to
changes in fluid saturation is to use the Gassmann
When a hydrocarbon field is produced, there are several model. Figure 19.4 shows one example, where a
reservoir parameters that might change, most crucially: calibrated Gassmann model has been used to deter-
– fluid saturation changes mine how the P-wave velocity changes with water
– pore pressure changes saturation (assuming that the reservoir fluid is a mix-
– temperature changes ture of oil and water).
492 M. Landrø
In addition, various contact models have been pro-
Relative P-wave velocity

1.1
posed to estimate the effective modulus of a rock.
1
Mavko, Mukerji and Dvorkin present some of these
0.9
models in their rock physics handbook (1998). The
0.8
Hertz-Mindlin model (Mindlin 1949) can be used to
0.7
describe the properties of pre-compacted granular
0.6
rocks. The effective bulk modulus of dry random
0.5
identical sphere packing is given by: 0 2 4 6 8 10 12
Effective pressure (MPa)
( )13
C2p ð1 φ2 ÞG2 P Fig. 19.5 Typical changes in P-wave velocity versus effective
Keff ¼ (19.1)
18π 2 ð1 σÞ2 pressure using the Hertz-Mindlin model. In this case the in situ
effective pressure (prior to production) is 6 MPa, and we see that
a decrease in effective pressure leads to a decrease in P-wave
where Cp is the number of contact points per grain, φ is velocity. The black curve represents the Hertz-Mindlin model
porosity, G is the shear modulus of the solid grains, σ (exponent = 1/6, as in Eq. (19.6)), the red curve is a modified
is the Poisson ratio of the solid grains and P is the version of the Hertz-Mindlin model (exponent = 1/10) that better
fits the ultrasonic core measurements
effective pressure (that is P ¼ Peff ). The shear modu-
lus is given as:
model (i.e. introducing fluids in the pore system of the
)13 rock), will ensure that the P-wave velocity approaches
φÞ2 G2 P
(
3C2p ð1 5 4σ
Geff ¼ 2
(19.2) the fluid velocity for zero effective stress, and not zero
2π 2 ð1 σÞ 5ð2 σÞ as in the Hertz-Mindlin model (dry rock assumption).
If we assume that the in situ (base survey) effective
This leads to: pressure is P0 we see from Eq. (19.3) that the relative
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P-wave velocity versus effective pressure is given as:
Keff þ 43 Geff
VP ¼ (19.3) VP

P
16
ρ ¼ (19.6)
VP 0 P0
sffiffiffiffiffiffiffiffi
Geff
VS ¼ (19.4) Figure 19.5 shows this relation for an in situ effec-
ρ
tive pressure of 6 MPa. When such curves are com-
pared to ultrasonic core measurements, the slope of the
where Vp and Vs are P- and S-wave velocities, respec- measured curve is generally smaller than this simple
tively, and ρ is the sandstone density. Inserting theoretical curve. The cause for this might be multi-
Eqs. (19.1) and (19.2) into Eqs. (19.3) and (19.4) and fold: Firstly, the Hertz-Mindlin model assumes the
computing the Vp =Vs ratio, yields (assuming σ ¼ 0): sediment grains are perfect, identical spheres, which
is never found in real samples. Secondly, the ultra-
VP pffiffiffi sonic measurements might suffer from scaling issues,
¼ 2: (19.5)
VS core damage and so on. Thirdly, cementation effects
are not included in the Hertz-Mindlin model. It is
This means that according to the simplest granular therefore important to note that there are major
model (Hertz-Mindlin), the Vp =Vs ratio should be con- uncertainties regarding the actual dependency
stant as a function of confining pressure if we assume between seismic velocity and pore pressure changes.
the rock is dry. The Hertz-Mindlin model assumes the
sand grains are spherical and that there is a certain area
of grain-to-grain contacts. A major shortcoming of the 19.3 Some 4D Analysis Techniques
model is that at the limit of unconsolidated sands, both
the P- and S-wave velocities will have the same The analysis of 4D seismic data can be divided into
behaviour with respect to pressure changes, as shown two main categories, one based on the detection of
in Eq. (19.5). Combining with the Gassmann (1951) amplitude changes, the other on detecting travel-time
19 4D Seismic 493
Two complementary 4D analysis techniques In addition to these two observations, a rock phys-
1985 1995 ics feasibility study was done, and the main results are
Amplitude Pull-up summarised in Fig. 19.8. It is important to note that the
changes
Top reservoir
effects key parameter for 4D seismic is not the relative change
in the seismic parameters, but the expected change in
- Quantitative approach
reflectivity between the base and monitor surveys. If
we assume that the top reservoir interface separates the
- More noisy
OWC cap rock (Layer 1) and the reservoir (Layer 2), it is
Time shift is
- Qualitative interpretation proportional possible to estimate the relative change in the zero
- Discriminate between to change in offset reflection coefficient (Landrø et al. 1999) caused
pay thickness
drained and undrained
segments by production changes in the reservoir layer. Denoting
the acoustic impedances (velocity times density) in
Fig. 19.6 Shows the two 4D analysis techniques. Notice there Layers 1 and 2 by λ1 and λ2 , respectively, and using
are no amplitude changes at top reservoir between 1985 and
1995, but huge amplitude changes at the oil/water contact
B and M to denote base and monitor surveys, we find
(OWC). Also note the change in travel time (marked by the that the change in reflectivity is given as (Landrø et al.
yellow arrow) for the seismic event below the oil/water contact 1999):
changes, see Fig. 19.6. Practical experience has shown ΔR λM λB2 ΔAIðproductionÞ
that the amplitude method is most robust, and there- ¼ 2B ¼ : (19.7)
R λ2 λB1 ΔAIðoriginalÞ
fore this has been the most frequently employed
method. However, as the accuracy of 4D seismic has
improved, the use of accurate measurements of small Here we have assumed that the change in acoustic
timeshifts is increasingly the method of choice. There impedance in Layer 2 is small compared to the actual
are several examples where the timeshift between two impedance. Equation (19.7) means that the change in
seismic traces can be determined with an accuracy of a reflectivity is equal to the change in acoustic imped-
fraction of a millisecond. A very attractive feature of ance in Layer 2 divided by the original (pre-
4D timeshift measurement is that it is proportional to production) acoustic impedance contrast between
the change in pay thickness, and this method provides Layers 1 and 2. For Gullfaks, the expected change in
a direct quantitative result. The two techniques are acoustic impedance is 8% (assuming 60% saturation
complementary in that amplitude measurement is a change, Fig. 19.8) while the original acoustic imped-
local feature (measuring changes close to an inter- ance contrast for top reservoir was approximately
face), while the timeshift method measures average 18%. The estimated relative change in reflectivity is
changes over a layer, or even a sequence of layers. therefore 44% (8/18). This is the basic background for
In addition to the direct methods mentioned above, the success of using 4D seismic at Gullfaks, as shown
4D seismic interpretation is aided by seismic in Fig. 19.9. The effect of replacing oil with water is
modelling of various production scenarios, often com- evident on this figure, and is further strengthened by
bined with reservoir fluid flow simulation and 1D the amplitude difference map taken at the original oil-
scenario modelling based on well logs. water contact (Fig. 19.10). Several Gullfaks wells
have been drilled based on 4D interpretation, and a
rough estimate of the extra income generated by using
19.4 The Gullfaks 4D Seismic Study time-lapse seismic data is more than 1 billion USD.
One of the first commercially successful 4D examples

from the North Sea was the Gullfaks study (Landrø
et al. 1999). There are two simple explanations for this
19.5 Repeatability Issues
success: the complexity of the reservoir geometry
The quality of a 4D seismic dataset is dependent on
(Fig. 19.1) means that there will be numerous pockets
several issues, such as reservoir complexity and the
of undrained oil, and there is a strong correlation
complexity of the overburden. However, the most
between the presence of oil and amplitude brightening
important issue that we are able to influence is the
as shown in Fig. 19.7.
494 M. Landrø
Fig. 19.7 Amplitude map for the top reservoir event at amplitudes (red colours) and presence of oil. The size of one
Gullfaks. The purple solid line represents the original oil- square is 1,250 by 1,250 m
water contact. Notice the strong correlation between high
20
Vp Density Zs
16 Vs Zp Vp/Vs
12
Relative change (%)
–4
–8
–12
Pressure (–4 MPa) Pressure (4 MPa) Saturation (100%) Saturation (60%)
Fig. 19.8 Expected changes (based on rock physics) of key seismic parameters for pore pressure changes (two left columns) and for
two fluid saturation scenario changes (two right columns)
19 4D Seismic 495
1985
1996
Fig. 19.9 Seismic data from 1985 (upper right) and 1996 (lower right) and the corresponding 4D interpretation to the left (green
representing oil, and blue water)
Fig. 19.10 Seismic difference section (top) and amplitude difference map at the original oil/water contact (OWC) (bottom). Red
colour indicates areas that have been water-flushed, and white areas may represent bypassed oil
repeatability of the seismic data, i.e. how accurately • Tidal effects

we can repeat the seismic measurements. This acqui- • Noise from other vessels or other activity in the
sition repeatability is dependent on a number of area (rig noise)
factors such as: • Varying source signal
• Varying source and receiver positions (x, y and z • Changes in the acquisition system (new vessel,
co-ordinates) other cables, sources etc.)
• Changing weather conditions during acquisition • Variation in shot-generated noise (from previous
• Varying seawater temperature shot).
496 M. Landrø
0 Shot points
–500
–1000
Depth, m
–1500 Well path
–2000
–2500 Receiver
4.94
4.92
4.9
× 105 6.705 6.704 6.703
4.88 6.707 6.706
6.708 × 106
X-Coordinate, m Y-Coordinate, m
6
× 10
6.7075 N
6.707
Fig. 19.11 Two VSP-traces measured at exactly the same
position in the well, but with a slight difference in the source 6.7065
location (5 m in the horizontal direction). The distance between y-position (m)
two timelines is 50 ms. The difference trace is shown to the right
6.706
Disregarding the weather conditions (the only way 6.7055

to “control” weather is to wait), most of the items
listed are influenced by acquisition planning and per- 6.705
formance. A common way to quantify repeatability is

6.7045
to use the normalised RMS (root-mean-square)-level,
that is: 6.704
4.88 4.885 4.89 4.895 4.9 4.905 4.91 4.915 4.92
x-position (m) × 105
RMSðmonitor baseÞ 100
NRMS ¼ 2 ; (19.8)
RMSðmonitorÞ þ RMSðbaseÞ 90
80
where the RMS-levels of the monitor and base traces 70
are measured within a given time window. Normally,
60
RMS error (%)
NRMS is measured in a time window where no pro-

50
duction changes are expected. Figure 19.11 shows two
seismic traces from a VSP (Vertical Seismic Profile) 40
experiment where the receiver is fixed (in the well at 30
approximately 2 km depth), and the source co- 20

ordinates are changed by 5 m in the horizontal direc- 10
tion. We notice that the normalised rms-error (NRMS)
0
in this case is low, only 8%. 0 10 20 30 40 50 60 70 80 90 100
Shot separation distance (m)
In 1995 Norsk Hydro acquired a 3D VSP dataset
over the Oseberg Field in the North Sea. This dataset Fig. 19.12 Top: The 10,000 shot locations (map view) for the
consists of 10,000 shots acquired in a circular shooting 3D VSP experiment over the Oseberg Field. Bottom: NRMS as a
function of the source separation distance between approxi-
pattern and recorded by a 5-level receiver string in the mately 70,000 shot pairs for the Oseberg 3D VSP dataset
well. By comparing shot-pairs with different source
positions (and the same receiver), it is possible to in Fig. 19.12, where approximately 70,000 shot pairs
estimate the NRMS-level as a function of the horizon- with varying horizontal mis-positioning are shown.
tal distance between the shot locations. This is shown The main message from this figure is clear. It is
19 4D Seismic 497
important to repeat the horizontal positions both for measure. It is also important to note that the NRMS-
sources and receivers as accurately as possible; even a level is frequency dependent, so the frequency band
misalignment of 20–30 m might significantly increase used in the data analysis should be given.
the NRMS-level. During the last two decades the focus on source
Such a plot can serve as a variogram, since it shows and receiver positioning accuracy has lead to a sig-
the spread for each separation distance. Detailed stud- nificant improvement in the repeatability of 4D seis-
ies have shown that the NRMS-level increases signifi- mic data. Some of this is also attributed to better
cantly in areas where the geology along the straight processing of time-lapse seismic data. This trend is
line between the source and receiver is complex. From sketched in Fig. 19.13. Today the global average
Fig. 19.12 we see that the NRMS-value for a shot NRMS-level is around 20–30%. It is expected that
separation distance of 40 m might vary between 20 this trend will continue, though not at the same rate
and 80% and a significant portion of this spread is because the non-repeatable factors that need to be
attributed to variation in geology. This means that it tackled to get beyond 20–30% are more difficult. In
is not straightforward to compare NRMS-levels particular, rough weather conditions will represent a
between various fields, since the geological setting major hurdle.
might be very different. Still, NRMS-levels are fre-
quently used, since they provide a simple, quantitative
19.6 Fixed Receiver Cables
100 For marine seismic data, there are several ways to

Positioning Source stability
control the source and receiver positions.
Shot noise WesternGeco developed their Q-marine system,
Weather related noise where the streamers can also be steered in the horizon-
....
NRMS (%)
tal direction (x and y). This means that it is possible to

repeat the receiver positions by steering the streamers
into predefined positions. The devices are shown
attached to the streamers in Fig. 19.14.
Another way to control the receiver positions is to
0 bury the receiver cables on the seabed. One of the first
1995 2005 2015 commercial offshore surveys of this kind was
Time
launched at the Valhall Field in the southern part of
Fig. 19.13 Schematic view showing improvement in seismic the North Sea in 2003. Since then, more than 11
repeatability (NRMS) versus calendar time surveys have been acquired over this field. The buried
Fig. 19.14 Photographs of the birds (controlling devices) the cable with reasonable accuracy into a given position. (Cour-
attached to a marine seismic cable at deployment (left) and in tesy of WesternGeco)
water operation (right). These devices make it possible to steer
498 M. Landrø
cables cover 70% of the entire field and the 4D seismic 19.7 Geomechanical effects
data quality is excellent. The fact that more than 10
surveys have been recorded into exactly the same Geomechanics has traditionally been an important dis-
receiver positions, opens for alternative and new cipline in both the exploration and production of
ways (exploiting the multiplicity that more than hydrocarbons. However, the importance for geomech-
2 datasets offers) of analysing time-lapse seismic anics of time-lapse seismic monitoring was not fully
data. Other advantages offered by permanent receiver realised before the first results from the chalk fields in
systems are: the southern North Sea (Ekofisk and Valhall). Figure
– cheaper to increase the shot time interval in order to 19.15 shows map views of estimated travel-time shifts
reduce the effect of shot generated noise between base and monitor surveys at Ekofisk (Guilbot
– continuous monitoring of background noise and Smith 2002), where the seafloor subsidence had
– passive seismic been known for many years. The physical cause for
– possible to design a dedicated monitoring survey the severe compaction of the chalk reservoir is two-
close to a problem well at short notice. fold. Firstly, depletion of the field leads to pore-pressure
Despite all these advantages that permanent decrease, and hence the reservoir rock compacts
systems offer, there has not been a marked increase mechanically due to lack of pressure support. Secondly,
in such surveys so far, probably because of the rela- the chemical reaction between the chalk matrix and the
tively high upfront costs and the difficulty in water replacing the oil leads to a weakening of the rock
quantifying in advance the extra value it would bring. framework, allowing further compaction.
N N
2/4-X-27
2/4-X-09
- 0
- 2
- 4
Time (ms TWT)
- 6
Time (ms TWT)
- –10
- 8
- –8
- 10 - –6
- 12 - –4
- 14 - –2
1250 m 1250 m - 0
- 16
- 18
- 20
Fig. 19.15 4D timeshifts for top reservoir interface (left) and for the Ekofisk Formation (right). The black area is caused by the gas
chimney problem at Ekofisk, leading to lack of high quality seismic data in this area. (From Guilbot and Smith 2002)
19 4D Seismic 499
When the reservoir rock compacts, the over- and a compaction of several metres is not sufficient to
underburden are stretched. This stretch is relatively counteract the cumulative overburden effect. Similar
small (of the order of 0.1%), However, this produces effects have also been observed for sandstone
a small velocity decrease that is observable as reservoirs, such as the Elgin and Franklin fields. Nor-
timeshifts on time-lapse seismic data. Typical mally, sandstone reservoirs compact less than chalk
observations of seafloor subsidence show that the sub- reservoirs and the corresponding 4D timeshifts are
sidence is less than the measured compaction for the therefore less, but observable. Most of the North Sea
reservoir. The vertical movement of the seafloor is sandstone reservoirs show negligible compaction, and
typically less than that of the top reservoir (Settari therefore such effects are normally neglected in these
2002), though this is strongly dependent on reservoir 4D studies. However, as the accuracy of 4D seismic
geometry and the stiffness of the rocks above and improves, this effect is expected to be observed at
below. several of these fields as well. For instance the
The most common way to interpret 4D seismic data anticipated compaction of the Troll East field is
from a compacting reservoir is to use geomechanical around 0.5-1 m after 20 years of production, and this
modelling as a complementary tool. One such example will probably be detectable by 4D seismic.
(Hatchell and Bourne 2005) is shown in Fig. 19.16. A major challenge when the thickness of subsur-
Note that the negative timeshifts (corresponding to a face layers is changed during production, is to distin-
slowdown caused by overburden stretching) continue guish between thickness changes and velocity
into the reservoir section. This is due to the fact that changes. One way to resolve this ambiguity is to use
the reservoir section is fairly thin in this case and even geomechanical modelling as a constraining tool.
TWTT [ms]
1500
Top porosity
2000
2500
3000
1500
2000
2500
3000
Speed-up Slow-down
–10 10
Time-lapse timeshift [ms]
Fig. 19.16 Comparison of measured timeshifts from 4D seis- colours) continues into the reservoir. (From Hatchell and
mic data (top) and geomechanical modelling. The black solid Bourne 2005)
line represents top reservoir. Notice that the slowdown (blue
500 M. Landrø
Another way is to combine near and far offset travel From simple geometrical considerations we see that
time analysis (Landrø and Stammeijer 2004) to esti- the relation between the thickness change and the
mate velocity and thickness changes simultaneously. corresponding porosity change is:
Hatchell et al. and Røste et al. suggested in 2005 using
a factor (R) relating the relative velocity change (dv/v) dz dϕ
¼ (19.11)
to the relative thickness change (dz/z; displacement) z 1 ϕ
dv=v In the isotropic case (assuming that the rock is

R¼ (19.9)
dz=z stretched in all three directions):
Hatchell et al. found that this factor varies between dz dϕ

¼ (19.12)
1 and 5, and is normally less for the reservoir rock than z 3ð1 ϕÞ
for the overburden rocks. Laboratory measurements
(Holt et al. 2008 and Bathija et al. 2009) show that Inserting Eqs. (19.10) and (19.11) into Eq. (19.9)
the R-factor strongly depends on the stress state of the we obtain an explicit expression for the dilation factor
rock and the stress path, and that it typically decreases R:
with increasing axial stress. For sandstones and clays it
is common to establish empirical relationships aþb
R¼1 (19.13)
between seismic P-wave velocity (v) and porosity v
(ϕ). These relations are often of the simple linear
form: which is valid only for the uniaxial case and assuming
that velocity is only porosity dependent.
v ¼ aϕ þ b (19.10) Using Eq. (19.12) instead of (19.11) gives a similar
equation for the isotropic case. The relation between
where a and b are empirical parameters. If a rock is the measured time shift and the velocity and thickness
stretched (or compressed), a corresponding change in changes is given as (Landrø and Stammeijer 2004):
porosity will occur. It is important to emphasise that
Eq. (19.10) is simplified and does not include changes dT dv dz dz
¼ þ ¼ ð 1 þ RÞ
in grain contact stiffness or changes in crack density. T v z z

Assuming that the lateral extent of a reservoir is large 1 þ R dv
¼ (19.14)
compared to its thickness, it is reasonable to assume a R v
uniaxial strain condition as sketched in Fig. 19.17.
From Eq. (19.14) it is evident that we can estimate
either the thickness change or the velocity change if
the R-factor is known. Inserting Eq. (19.13) into
Uniaxial stretching of a rock
x x (19.14) yields
h dz v dT dT
x ¼ ¼S (19.15)
z 2v a b T T
h+dx x+dx
where S is the strain-timestrain (by timestrain we
simply mean the relative stretch of traveltime within
h(x + x – h) (h + dx)x + (x – h)h the layer) ratio representing how much the relative
1= 2=
x2 x(x + dx)
stretch (or compaction) is for a given measured rela-
tive time shift. If the velocity-depth trend is known and
d dx the empirical parameters a and b are known we can
=
1– x estimate S(z). As a simple example we use
v ¼ v0 þ kz, where v0 ¼ 1500 m/s and k ¼ 0.5. Fur-
Fig. 19.17 Cartoon showing the effect of increased porosity thermore, if we use the empirical relationship obtained
due to stretching of a rock sample by Han et al. (1986) for shaly sandstones we find that
19 4D Seismic 501
Both the overburden and reservoir rocks are

exposed to anisotropic stress changes, leading to
anisotropic velocity changes. How to couple
geomechanical modelling to time-lapse seismic data
is still a research topic, focusing on a variety of
methods. As the accuracy and repeatability level of
time-lapse seismic data improves, it is realistic to
assume that this research area will become more
important in the future.
Fig. 19.18 Sensitivity ratio (S) between relative stretching (strain)

and measured relative time shift (dT/T) assuming a ¼ -7000 m/s
19.8 Discrimination Between Pore
and b ¼ 3000 m/s in equation 19.15, and that v ¼ 1500 +0.5z Pressure and Saturation Effects
Although the main focus in most 4D seismic studies is

a ¼ -7000 m/s and b ¼ 3000 m/s, roughly. Using these to study fluid flow and detect bypassed oil pockets, the
values in equation 19.15 gives a sensitivity that challenge of discriminating between pore pressure
increases slightly with increasing depth, as shown in changes and fluid saturation changes occurs fre-
Fig. 19.18. However, it is important to stress that the quently. From rock physics we know that both effects
above assumption of uniaxial stretching of the over- influence the 4D seismic data. However, as can be
burden rocks might be too simple, especially for seen from Fig. 19.8, the fluid pressure effects are not
reservoirs of finite lateral extent. In such cases, the linked to the seismic parameters in the same way as
vertical and horizontal strains will have opposite sign, the fluid saturation effects. This provides an opportu-
as shown by Tempone et al. (2009; Fig. 19.19) and nity to discriminate them, since we have different rock
Tempone (2011). This means that the net volumetric physics relations for the two cases. Some of the early
strain is typically less than the vertical and horizontal attempts to perform this discrimination between pres-
strains, as shown in Fig. 19.20. If the basement layer sure and saturation were presented by Tura and
(150 m below the reservoir) is rigid, the Lumley (1999) and Landrø (1999). In Landrø’s
geomechanical modelling shows that the volumetric method (2001) the rock physics relations (based on
strains near to the surface are negative while those Gassmann and ultrasonic core measurements) are
closer to the reservoir are positive. Furthermore, we combined with simple AVO (amplitude versus offset)
observe from Fig. 19.20 that the volumetric strain equations to obtain direct expressions for saturation
below the reservoir is positive and significant. changes and pressure changes. Necessary input to this
According to Geertsma’s model (Fjær et al. 2008) algorithm is near and far offset amplitude changes
the volumetric strain will be close to zero for the rocks estimated from the base and monitor 3D seismic
surrounding the reservoir. This means that the porosity cubes. This method was tested on a compartment
change in the overburden is zero, and thus for deple- from the Cook Formation at the Gullfaks Field. Figure
tion the vertical stress decrease will be accompanied 19.21 shows a seismic profile (west–east) through this
by a horizontal stress increase (equal to half the verti- compartment.
cal stress change). For an inflated reservoir, the oppo- A significant amplitude change is observed for the
site situation will apply. Therefore, it seems very top Cook interface (red solid line in the figure), both
likely that the mono-causal porosity model discussed below and above the oil/water contact. The fact that
above does not capture the physics correctly. Stress this amplitude change extends beyond the oil/water
changes are probably as important as porosity contact is a strong indication that it cannot be solely
changes, and in many cases stress might be the domi- related to fluid saturation changes, and a reasonable
nant effect. Geertsma’s model is a simplification, and candidate is therefore pore pressure changes. Indeed, it
especially heterogeneities and reservoir geometry will was confirmed that the pore pressure had increased by
change the overburden stress path. 50–60 bars in this segment, meaning that the reservoir
pressure is approaching fracture pressure. We also
502 M. Landrø
Fig. 19.19 Vertical strain (top) caused by a compacting reservoir Corresponding results for the horizontal strain are shown in the
assuming a rigid basement below the reservoir (left) and using bottom plot. (Source: Pamela Tempone’s PhD thesis, NTNU, 2011
Geertsma’s solution (right). The reservoir extends from 750 to and SPE paper SPE-121081)
850 m in depth and has a lateral extension of 1000 m.
Fig. 19.20 Vertical strain caused by a compacting reservoir assuming a rigid basement (left), and using Geertsma’s solution (right).
(Source: Pamela Tempone’s PhD thesis, NTNU, 2011 and SPE paper SPE-121081)
19 4D Seismic 503
C-3
Top Cook
1985
OWC
1996
Fig. 19.21 Seismic section through the Cook Formation at change extends beyond the OWC-level, a strong indication that
Gullfaks. The cyan solid line indicates the original oil/water the downflank amplitude change is caused by pore pressure
contact (marked OWC). Notice the significant amplitude change changes
at top Cook between 1985 and 1996, and that the amplitude
Fig. 19.22 Estimated fluid saturation changes (left) and pore original oil/water contact. Notice that the estimated pressure
pressure changes (right) based on 4D AVO analysis for the top changes extend beyond this blue line to the west, and terminate
Cook interface at Gullfaks. The blue solid line represents the at a fault to the east. Yellow colours indicate significant changes
observe from this figure that the base reservoir event From this figure we see that in the western part of
(blue solid line) is shifted slightly downwards, by the segment most of the estimated fluid saturation
2–3 ms. This slowdown is interpreted as a velocity changes occur close to the oil/water contact. However,
drop caused by the pore pressure increase in the some scattered anomalies can be observed beyond the
segment. oil/water contact in the northern part of the segment.
Figure 19.22 shows an attempt to discriminate These anomalies are probably caused by inaccuracies
between fluid saturation changes and pressure changes in the algorithm or by limited repeatability of the time-
for this compartment using the method described in lapse AVO data. The estimated pressure changes are
Landrø (2001). In 1996, 27% of the estimated recov- more consistent with the fault pattern in the region,
erable reserves in this segment had been produced, so and it is likely that the pressure increase is confined
we know that some fluid saturation changes should be between faults in almost all directions. Later time-
observable on the 4D seismic data. lapse surveys show that the eastern fault in the figure
was “opened” some years later.
504 M. Landrø
19.9 Other Geophysical Monitoring rocks and water-filled rocks. Field tests have shown
Methods that such data are indeed repeatable, so there should
definitely be a potential for using repeated EM surveys
Parts of this section are taken from “Future challenges to monitor a producing reservoir. So far, frequencies
and unexplored methods for 4D seismics”, by Landrø, as low as 0.25 Hz (and even lower) are being used, and
Recorder (2005). then the spatial resolution will be limited. However, as
Up to now, 4D seismic has proved to be the most a complementary tool to conventional 4D seismic, 4D
effective way to monitor a producing reservoir. How- EM might be very useful. In many 4D projects it is
ever, as discussed above, there are several severe hard to quantify the amount of saturation changes
limitations associated with time lapse seismic. One taking place within the reservoir, and time-lapse EM
of them is that seismic reflection data is sensitive to studies might be used to constrain such quantitative
acoustic impedance (velocity times density). Although estimates of the saturation changes. Furthermore,
4D time shift can reveal changes in average velocity unlike conventional 4D seismic which is both pressure
between two interfaces, an independent measurement and saturation sensitive, the EM technique is not very
of density changes would be a useful complement to sensitive to pressure changes, so it may be a nice tool
conventional 4D seismic. The idea of actually measur- for separating between saturation and pressure
ing the mass change in a reservoir caused by hydro- changes.
carbon production has been around for quite some In another paper, Johansen et al. (2005) show that
time. The limiting factor for gravimetric reservoir the EM response over the Troll West gas province is
monitoring has been the repeatability (or accuracy) significantly above the background noise level, see
of the gravimeters. However, Sasagawa et al. (2003) Fig. 19.23. If we use the deviation from a smooth
demonstrated that improved accuracy can be achieved response as a measure of the repeatability level, a
by using 3 coupled gravimeters placed on the seabed. relative amplitude variation of approximately 0.1–0.2
This technical success led to a full field programme at (measured in normalised EM amplitude units) is
the Troll Field, North Sea. Another successful field observed. Compared to the maximum signal observed
example was time-lapse gravity monitoring of an aqui- at the crest of the field (2.75) this corresponds to a
fer storage recovery project in Leyden, Colorado repeatability level of 4–7%, which is very good com-
(Davis et al. 2005), essentially mapping water influx. pared to conventional time-lapse seismic. This means
Obviously, this technique is best suited for reservoirs it is realistic to assume that time-lapse EM data can
where significant mass changes are likely to occur, provide very accurate low-resolution (in x-y plane)
such as water replacing gas. Shallow reservoirs are constraints on the saturation changes observed on 4D
better suited than deep. The size of the reservoir is a seismic data. Recently, Mittet et al. (2005) showed
crucial parameter, and a given minimum size is that depth migration of low frequency EM data may
required in order to obtain observable effects. In be used to enhance the vertical resolution. In a field
quite another field, gravimetric measurements might data example, the vertical resolution achieved by this
help in monitoring volcanic activity by distinguishing type of migration is maybe of the order of a few
between fluid movements and tectonic activity within hundred metres. The EM sensitivity is determined by
an active volcano. the reservoir thickness times the resistivity,
Around 2000 the hydrocarbon industry started to underlining the fact that both reservoir thickness and
use electromagnetic methods for exploration. Statoil reservoir resistivity are crucial parameters for 4D
performed a large-scale research project that showed CSEM. This means that the resolution issue with the
that it is possible to discriminate between active EM method is steadily improving, and such
hydrocarbon-filled and water-filled reservoirs from improvements will of course increase the value of
controlled source electromagnetic (CSEM) repeated CSEM data as a complement to conventional
measurements. Since this breakthrough of CSEM 4D seismic.
surveys some years ago (Ellingsrud et al. 2002), this By using the so-called interferometric synthetic
technique has mainly been used as an exploration tool, aperture radar principle obtained from orbiting
in order to discriminate between hydrocarbon-filled satellites, an impressive accuracy can be attained in
19 4D Seismic 505
3.5
3.0
Normalized magnitudes
2.5
2.0
1.5
1.0
0.5
0.0
10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0
Offset (m)
3.5 0
Water 200
3.0 1 5 10 15 20 24
Normalised magnitude
400
Depth below sealevel (m)

2.5
at 6.5 km offset
600
2.0 800
Sediments 1000
1.5
1200
1.0
1400
0.5
TWGP
1600
0.0 1800
0 5000 10000 15000 20000 25000
Horizontal distance (m)
Fig. 19.23 Modelled (open circles) and measured (red circles) offset of 6.5 km along a profile, where the background shows
relative electrical field strength as a function of offset (top). The the reservoir model. Notice that the relative signal increases by a
values are normalised to the response outside the reservoir (blue factor of 3 above the thickest part of the reservoir. (Figure taken
circels). Bottom figure shows relative values for a constant from Johansen et al. 2005)
measuring the distance to a specific location on the more sophisticated way. If you need to adjust your
Earth’s surface. By measuring the phase differences geomechanical model to obtain correspondence
for a signal received from the same location for differ- between observed surface subsidence and reservoir
ent calendar times, it is possible to measure the rela- compaction, then this adjusted geomechanical model
tive changes with even higher accuracy. By exploiting can be used to distinguish between overburden rocks
a sequence of satellite images, height changes can be with high and low stiffness for instance, which again
monitored versus time. The satellites used for this can be translated into macro-variations of the overbur-
purpose orbit at approximately 800 km. den velocities. The deeper the reservoir, the lower the
Several examples of monitoring movements of the frequency (less vertical resolution) of the surface
surface above a producing hydrocarbon reservoir have imprint of the reservoir changes, thus this method
been reported, though there are of course limitations to can not be used directly to identify small pockets of
the detail of information such images can provide for undrained hydrocarbons. However, it can be used as a
reservoir management. The obvious link to the reser- complementary tool for time-lapse seismic since it can
voir is to use geomechanical modelling to tie the provide valuable information on the low frequency
movements at the surface to subsurface movements. spatial signal (slowly varying in both horizontal and
For seismic purposes, such a geomechanical approach vertical directions) of reservoir compaction.
can be used to obtain improved velocity models in a
506 M. Landrø
19.10 CO2 Monitoring One way to improve our understanding of how the
CO2 flows in a porous rock is to perform small-scale
This text is taken from the paper “Quantitative Seis- flooding experiments on long core samples. Such
mic Monitoring Methods” by Landrø (2008), in experiments involve injecting various fluids in the
ERCIM News 74, pp. 16–17: end of a 30–40 cm long core (Fig. 19.24). An exam-
Interest in CO2 injection, both for storage and as a ple of such a flooding experiment and corresponding
tertiary recovery method for increased hydrocarbon X-ray images for various flooding patterns is shown
production, has grown significantly over the last in Fig. 19.24. By measuring acoustic velocities as the
decade. Statoil has stored approximately 10 million flooding experiment is conducted, these experiments
tons of CO2 in the Utsira Formation at the Sleipner can be used to establish a link between pore-scale
Field, and several similar projects are now being CO2-injection and time-lapse seismic on the field
launched worldwide. At NTNU our focus has been to scale.
develop geophysical methods to monitor the CO2- Figure 19.25 shows an example of how CO2
injection process, and particularly to try to quantify influenced the seismic data over time. By combining
the volume injected directly from geophysical data. 4D travel time and amplitude changes, we have
Out
Slice
7 ml 22 ml 27 ml
37 ml 52 ml 72 ml
In
Bulk density g/cm3
Fig. 19.24 Long core (left) showing the location for the X-ray injection process. (From Marsala and Landrø, EAGE extended
cross-section (red arrow). Water injection is 50 g/l. To the right: abstracts 2005)
X-ray density maps of a core slice: 6 time steps during the
1030
1040
1050
1060
1070
1080
1090
1100
1030
1040
1050
1060
1070
1080
1090
1100
1030
1040
1050
1060
1070
1080
1090
1100
1030
1040
1050
1060
1070
1080
1090
1100
900
950
S
1000 a
1050 n
d
1100
1150
1994 1999 2001 2002
Fig. 19.25 Time-lapse seismic data showing monitoring of the CO2 injection at Sleipner. The strong amplitude increase (shown in
blue) is interpreted as a thin CO2 layer. (Printed with permission from Statoil)
19 4D Seismic 507
developed methods to estimate the thickness of CO2 waves is that the incidence angle of a seismic wave is
layers, which makes it possible to estimate volumes. larger than the critical angle θC :
Another key parameter is CO2 saturation, which can
be estimated using rock physics measurements and v1
sin θC ¼ (19.16)
models. Although the precision in both 4D seismic v2
methods and rock physics is increasing, there is no
doubt that precise estimates are hard to achieve, and where v1 and v2 denote the velocity above and below
therefore we need to improve existing methods and the top reservoir (or the target) interface. It is evident
learn how to combine several methods in order to from this equation that there has to be a velocity
decrease the uncertainties associated with these mon- increase in order to create head waves for incidence
itoring methods. angles greater than the critical angle. This is also one
of the major limitations for this method: it does not
work for interfaces where the velocity contrast is weak
or decreasing. Another weakness of the method is that
19.11 Time-lapse Refraction Methods
it relies on dominantly horizontal wave propagation,
and therefore it is most robust for detecting velocity
Refraction methods are among the oldest methods in
changes that are laterally extensive. Small, localised
seismic exploration. The term refraction is not unique
4D anomalies are better resolved by using conven-
or precise. According to Sheriff (2002) it has two
tional 4D reflection methods. However, in a situation
meanings, first to change direction (following Snell’s
where the velocity changes caused by production are
law) and second to involve head waves. In this chapter
small (less than a few per cent), the time lapse refrac-
I will use a relaxed interpretation which means that the
tion method might be attractive. A simple synthetic
term includes head waves, diving waves and
example based on a six-layer model is shown in
reflections close to critical angle. Maybe long offset
Fig. 19.26, where we observe that the major difference
time-lapse seismic is more correct. Within petroleum
signal occurs for offsets close to the critical offset. In
geophysics, one of the first proposals to exploit this
this case, the velocities of the cap rock and the reser-
type of data for 4D purposes was presented by Landrø
voir layer are 2000 and 2500 m/s, respectively. For the
et al. 2004. It was suggested that 4D refraction analy-
monitor survey, we have assumed that the reservoir
sis could be of potential benefit for carbonate
velocity increases to 2600 m/s. The differences
reservoirs. The basic condition for generation of head
observed in Fig. 19.26 are caused by a combination
offset (m) offset (m)

0 2000 4000 6000 2000 4000 6000
0 0
1 1
2 2
t (s)
t (s)
3 3
4 4
Fig. 19.26 Synthetic 4D refraction example: Modelled base close to the critical offset around 5–6 km. The vertical axis is
survey (left) for a 6-layer model, and the corresponding differ- from 0 to 5 seconds, and the horizontal offset axis is from 0 to
ence section, assuming that the velocity of the reservoir layer is 8 km. (Figure from Landrø et al. 2004)
changed from 2500 m/s to 2600 m/s. The largest changes occur
508 M. Landrø
Fig. 19.27 RMS (Root Mean Square, after fk-filtering) analy- offsets when the velocity of the reservoir layer is slightly
sis of the synthetic data example demonstrating that the ampli- decreased (2% and 4% reduction, respectively)
tude increase close to critical offset is moving towards shorter
of travel time and amplitude changes. We notice that within the reservoir layer, as discussed by Zadeh et al.
the differences at zero offset are negligible, as 2011.
expected. In the second example from Valhall we observe no
A reliable and simple way to detect the change in changes between survey 1 and 6, followed by a signif-
critical angle is to use a windowed RMS (Root Mean icant offset shift between survey number 6 and
Square) technique, as shown in Fig. 19.27. A clear 8 (Fig. 19.29). This example clearly demonstrates the
shift in the RMS-curve is observed assuming small advantage of acquiring several frequent surveys: it is
velocity changes (2 and 4%). An fk-filter has been possible to monitor detailed reservoir changes that
applied to the data in order to enhance the refracted might occur over a half year interval. When a perma-
signal in Fig. 19.27. Is it likely that we can observe nent array is installed at a field, each survey is rela-
similar effects for field data? One concern is that long tively cheap since a single source vessel is sufficient to
offset seismic data is less repeatable and that noise acquire a repeat survey.
created by multiple reflections and refractions in the Another application of time lapse refraction is to
water layer (see Landrø 2007) will distort a clean and detect changes in shallow sediments. One such exam-
precise picking of such offset shifts. ple is sketched in Fig. 19.30, where an underground
An example using the LoFS (Life of Field Seismic) blow-out (from well 2/4-14, the same as shown in
acquired at the Valhall Field demonstrates that it is Fig. 19.3), causes gas to be trapped in a thin sand
possible (Fig. 19.28). We clearly see an offset shift layer. In this case, the most robust 4D refraction
similar to that for the synthetic case. We observe attribute is the time shift. The time shifts attributed
changes between LoFS survey number 1 and 6 and to the velocity reduction are observed from conven-
that this change continues when compared to LoFS tional site survey data acquired before and after the
survey number 8. The interpretation of these shifts is, blow-out, as shown in Fig. 19.31. For comparison, a
however, not as simple as for the synthetic case, since corresponding gather outside the gas-filled area is
the Valhall field is compacting. The compaction at the shown, clearly demonstrating no time shifts in this
reservoir level causes stretching of the overburden case. For site survey data, the amplitude repeatability
shale, and some of the offset change is attributed to is not of the same quality as for normal seismic data.
this effect and some to saturation and pressure changes There are several potential causes for this lower
19 4D Seismic 509
Fig. 19.28 Field data example using LoFS data from Valhall. We observe a clear shift towards shorter offsets for the interpreted
amplitude increase close to critical offset (4000–4500 m). (Figure from Zadeh et al. 2011)
Fig. 19.29 Example 2 using Valhall LoFS data: Here we observe no changes between LoFS1 and LoFS6, followed by a significant
change of approximately 300 m for the LoFS8 survey. (From Zadeh et al. 2011)
repeatability level: smaller source arrays, more Figure 19.32 shows a systematic estimation of time
variability in streamer sensitivity and shallower shifts using offsets ranging from 1000 m to 1150 m
towing depth (causing more noise in rough weather) (Zadeh and Landrø 2011). Despite some scatter in the
are some of them. time shift estimates, we observe a clear increase close
510 M. Landrø
Fig. 19.30 Schematic plot showing how 4D refraction analysis can be used to detect shallow gas leakage
Fig. 19.31 Observed time shift for the refracted wave close to (M¼Monitor) traces after. Typical time shifts within the gas
the blow-out well (left) and similar display far away from the anomaly are of the order of 3-4 ms. (Zadeh and Landrø 2011)
well (right). Black traces (B¼Base) are before blow-out and red
Fig. 19.32 Estimated refraction time shifts for three offsets close to the blow-out well (14): A positive time shift of up to 4 ms is
observed. (From Zadeh and Landrø 2011)
to the blow-out well (14). The maximum time shift is A third example of time lapse refraction is shown in
4 ms. The thickness of the gas-filled sand layer is Fig. 19.33, where time shifts are observed close to a
approximately 10 m. region where steam is injected into a heavy oil
19 4D Seismic 511
Fig. 19.33 Time lapse refraction example from Peace River heavy oil field in Alberta. Significant velocity reduction caused by
steam injection into row 8. (Figure from Hansteen et al. 2010)
reservoir in Alberta, Canada (Hansteen et al. 2010). presented so far illustrate that the technique has a
When steam is injected, the stiffness of the rock is potential. One weakness is that it is hard to map the
reduced, leading to a velocity decrease. This velocity time lapse anomalies precisely in depth. Fortunately,
decrease leads to a corresponding increase in time full wave form inversion techniques have developed
shift, as observed in Fig. 19.33. significantly over the last decade (Virieux and Operto
Hilbich (2010) uses time-lapse refraction seismic 2009, Sirgue et al. 2009). A few examples of this
tomography to study seasonal changes in the ground technique have also been presented for 4D
ice in the Alps. He concludes that the method is applications (Routh et al. 2009). Therefore, there is
capable of detecting temporal changes in alpine hope that time lapse refraction methods might develop
permafrost and can identify ground ice degradation. into a more precise and accurate technique. Despite
A major strength of the method is according to Hilbich this lack of more advanced analysis methods, it is
the high vertical resolution potential and the ability to important to underline that if 4D refraction is used as
discriminate phase changes between frozen and a tool for detecting shallow gas leakage, the important
unfrozen material over time. A significant limitation issue is to detect and locate approximately where the
is the low penetration depth due to strong velocity leakage occurs. Detailed and accurate mapping of the
contrasts between the active layer and the permafrost 4D anomaly is not crucial. Should more accurate
table. mapping be required, this may be done across the
Presently, time lapse refraction is not well located area by conventional reflection survey.
established as a 4D technique. The few examples
512 M. Landrø
19.12 Future Aspects Ivar Brevik, Paul Hatchell, Peter Wills, Jan Stammeijer, Thomas
Røste, Rodney Calvert, Alexey Stovas, Lyubov Skopintseva,
Andreas Evensen, Amir Ghaderi, Dag O. Larsen, Kenneth
The most important issue for further improvement of Duffaut, Jens Olav Paulsen, Jerome Guilbot, Olav Barkved,
the 4D seismic method, is to improve the repeatability. Jan Kommedal, Per Gunnar Folstad, Helene Veire, Erling
Advances in both seismic acquisition and 4D Fjær, Pamela Tempone, Rune Martin Holt, Ola Eiken, Torkjell
Stenvold, Fredrik Hansteen and Alberto Marsala are all thanked
processing will contribute to this process. As sketched
for co-operation and discussions.
in Fig. 19.13, it is expected that this will be less
pronounced than in the past decade. However, it is
still a crucial issue, and even minor improvements
might mean a lot for the value of a 4D seismic study. Further Reading
Further improvements in repeatability will probably
involve issues like source stability, source positioning, Amundsen, L. and Landrø, M. December 2007. 4D seismic –
shot time interval, better handling of various noise Status and future challenges, Part II. GeoExpro, 54–58.
sources. Maybe in the future we will see vessels Barkved, O. 2004. Continuous seismic monitoring. 74th Annual
towing a superdense grid of sensors, in order to obtain International Meeting, The Society of Exploration
Geophysicists, 2537–2540.
perfect repositioning of the receiver positions. Barkved, O., Buer, K., Halleland, K.B., Kjelstadli, R., Kleppan,
Another direction to improve 4D studies could be T. and Kristiansen, T. 2003. 4D seismic response of primary
to constrain the time-lapse seismic information by production and waste injection at the Valhall field. 65th
other types of information, such as geomechanical Meeting. European Association of Geoscientists and
Engineers, Extended Abstracts, A22.
modelling, time-lapse EM, gravimetric data or inno- Barkved, O.I. and Kristiansen, T. 2005. Seismic time-lapse
vative rock physics measurements. effects and stress changes: Examples from a compacting
Our understanding of the relation between changes reservoir. The Leading Edge 24, 1244.
in the subsurface stress field and the seismic Bathija, A.P., Batzle, M.L. and Prasad, M. 2009. An experimen-
tal study of the dilation factor. Geophysics 74, E181–E191.
parameters is still limited, and research within this Calvert, R. 2005. Insights and methods for 4D reservoir moni-
specific area will be crucial to advance the 4D. New toring and characterization, EAGE/SEG Distinguished
analysis methods like long offset 4D, might be used as Instructor Short Course 8.
a complementary technique or as an alternative Christensen, N.I. and Wang, H.F. 1985. The influence of pore
pressure and confining pressure on dynamic elastic
method where conventional 4D analysis has limited properties of Berea sandstone. Geophysics 50, 207–213.
success. However, long offset 4D is limited to Davis, K., Li, Y., Batzle, M. and Reynolds, B. 2005. Time-lapse
reservoirs where the velocity increases from the cap gravity monitoring of an aquifer storage recovery project in
rock to the reservoir rock. Leyden, Colorado. 75th Annual International Meeting. The
Society of Exploration Geophysicists, GM1.4.
The link between reservoir simulation (fluid flow Eiken, O., Zumberge, M., Stenvold, T., Sasagawa, G. and
simulation) and time-lapse seismic will continue to be Nooen, S. 2004. Gravimetric monitoring of gas production
developed. As computer resources increase, the feasi- from the Troll Field. 74th Annual International Meeting. The
bility of a joint inversion exploiting both reservoir Society of Exploration Geophysicists, 2243–2246.
Ellingsrud, S., Eidesmo, T., Johansen, S., Sinha, M.C.,
simulation and time-lapse seismic data in the same MacGregor, L.M. and Constable, S. 2002. Remote sensing
subsurface model will increase. Despite extra comput- of hydrocarbon layers by seabed logging (SBL): Results
ing power, it is reasonable to expect that the non- from a cruise offshore Angola. The Leading Edge 21(10),
uniqueness problem (several scenarios will fit the 972–982.
Fjær, E., Holt, R.M., Horsrud, P., Raaen, A.M. and Risnes, R.
same datasets) will require that the number of Earth 2008. Petroleum Related Rock Mechanics, 2nd edition. pp.
models is constrained by models predicting the distri- 402-409. Elsevier, ISBN 978-0-444-50260-5.
bution of rock physical properties based on sedimen- Foldstad, P.G., Andorsen, K., Kjørsvik, I. and Landrø, M. 2005.
tology, diagenesis and structural geology. Simulation of 4D seismic signal with noise – Illustrated by
WAG injection on the Ula Field. 75th Annual International
Meeting. The Society of Exploration Geophysicists, TL2.7.
Acknowledgments Statoil is acknowledged for permission to
Furre, A-K., Munkvold, F.R. and Nordby, L.H. 2003. Improving
use and present their data. The Research Council of Norway is
reservoir understanding using time-lapse seismic at the
acknowledged for financial support to the ROSE (Rock Seismic)
Heidrun Field, 65th Meeting. European Association of
project at NTNU. Lasse Amundsen, Lars Pedersen, Lars
Geoscientists and Engineers, A20.
Kristian Strønen, Per Digranes, Eilert Hilde, Odd Arve Solheim,
19 4D Seismic 513
Gassmann, F. 1951. Über die elastizität poroser medien. methods – Possibilities and limitations. First Break 19,
Vierteljahrsschrift der Naturforchenden Gesellschaft in 671–677.
Zürich 96, 1–23. Landrø, M., Digranes, P. and Strønen, L.K. 2002. Pressure
Gosselin, O., Aanonsen, S.I., Aavatsmark, I., Cominelli, A., effects on seismic data – Possibilities and limitations.
Gonard, R., Kolasinski, M., Ferdinandi, F., Kovacic, L. and Paper presented at the NPF biennual conference,
Neylon, K. 2003. History matching using time-lapse seismic Kristiansand, Norway, 11–13 March, 2002.
(HUTS), SPE paper 84464. Landrø, M., Digranes, P. and Strønen, L.K. 2005. Pressure
Guilbot, J. and Smith, B. 2002. 4-D constrained depth conver- depletion measured by time-lapse VSP. The Leading Edge
sion for reservoir compaction estimation: Application to 24, 1226–1232 (December issue).
Ekofisk Field. The Leading Edge 21(3), 302–308. Landrø, M. and Duffaut, K. 2004. Vp–Vs ratio versus effective
Han, D., Nur, A. and Morgan, D. 1986. Effects of porosity and pressure and rock consolidation – A comparison between
clay content on wave velocities in sandstones. Geophysics, rock models and time-lapse AVO studies. 74th Annual Inter-
51, 2093–2107. national Meeting. The Society of Exploration Geophysicists,
Hansteen, F., Wills, P.B., Hornman, K. and Jin, L. 2010. Time- 1519–1522.
lapse refraction seismic monitoring. 80th Annual Interna- Landrø, M., Nguyen, A.K. and Mehdizadeh, H. 2004. Time
tional Meeting. Society of Exploration Geophysicis, 4170- lapse refraction seismic – A tool for monitoring carbonate
4174. fields? 74th Annual International Meeting. The Society of
Hatchell, P.J. and Bourne, S.J. 2005. Rocks under strain: Strain- Exploration Geophysicists, 2295–2298.
induced time-lapse time shifts are observed for depleting Landrø, M., Nguyen, A.K. and Zadeh, M.H. 2004. Time lapse
reservoirs. The Leading Edge 24, 1222–1225. refraction seismic – a tool for monitoring carbonate fields?
Hatchell, P.J., Kawar, R.S. and Savitski, A.A. 2005. Integrating 74th Annual International Meeting. The Society of Explora-
4D seismic, geomechanics and reservoir simulation in the tion Geophysicists, 2295–2298.
Valhall oil field. 67th Meeting. European Association of Landrø, M. and Skopintseva, L. 2008. Potential improvements
Geoscientists and Engineers, C012. in reservoir monitoring using permanent seismic receiver
Hilbich, C. 2010. Time-lapse refraction seismic tomography for arrays. The Leading Edge 27, 1638–1645.
the detection of ground ice degradation. The Cryosphere 4, Landrø, M., Solheim, O.A., Hilde, E., Ekren, B.O. and Strønen,
243–259. L.K. 1999. The Gullfaks 4D seismic study. Petroleum Geo-
Holt, R.M., Fjær, E., Nes, O.-M. and Stenebråten, J.F. 2008. science 5, 213–226.
Strain sensitivity of wave velocities in sediments and sedi- Landrø, M. and Stammeijer, J. 2004. Quantitative estimation of
mentary rocks. ARMA (American Rock Mechanics Associ- compaction and velocity changes using 4D impedance and
ation; conference paper) 08-291, 1-8. traveltime changes. Geophysics 69(4), 949–957.
Johansen, S.E., Amundsen, H.E.F., Røsten, T., Ellingsrud, S., Landrø, M., Veire, H.H., Duffaut, K. and Najjar, N. 2003.
Eidesmo, T. and Bhuyian, A.H. 2005. Subsurface Discrimination between pressure and fluid saturation
hydrocarbons detected by electromagnetic sounding. First changes from marine multicomponent time-lapse seismic
Break 23, 31–36. data. Geophysics 68, 1592–1599.
Kommedal, J.H., Barkved, O.I. and Howe, D.J. 2004. Initial Larsen, D. and Lie, A. 1990. Monitoring an underground flow
experience operating a permanent 4D seabed array for reser- by shallow seismic data: A case study. SEG Expanded
voir monitoring at Valhall. 74th Annual International Abstracts 9, 201 pp.
Meeting. The Society of Exploration Geophysicists, MacLeod, M., Hanson, R.A., Bell, C.R. and McHugo, S. 1999.
2239–2242. The Alba Field ocean bottom cable survey: Impact on devel-
Koster, K., Gabriels, P., Hartung, M., Verbeek, J., Deinum, G. opment. The Leading Edge 18, 1306–1312.
and Staples, R. 2000. Time-lapse seismic surveys in the Mavko, G., Mukerji, T. and Dvorkin, J. 1998. The Rock Physics
North Sea and their business impact. The Leading Edge 19 Handbook. Cambridge University Press, Cambridge, ISBN
(03), 286–293. 0521 62068 6, 147–161.
Kragh, E. and Christie, P. 2002. Seismic repeatability, Mehdi Zadeh, H., Landrø, M., Mythen, B.A., Vedanti, N. and
normalized rms, and predictability. The Leading Edge 21, Srivastava, R. 2005. Time lapse seismic analysis using long
640–647. offset PS data. 75th Annual International Meeting. The Soci-
Landrø, M. 1999. Repeatability issues of 3-D VSP data. Geo- ety of Exploration Geophysicists, TL3.7.
physics 64, 1673–1679. Mindlin, R.D. 1949. Compliance of elastic bodies in contact.
Landrø, M. 2001. Discrimination between pressure and fluid Journal of Applied Mechanics 16, 259–268.
saturation changes from time lapse seismic data. Geophysics Nes, O.M., Holt, R.M. and Fjær, E. 2000. The reliability of core
66(3), 836–844. data as input to seismic reservoir monitoring studies. SPE
Landrø, M. 2002. Uncertainties in quantitative time-lapse seis- 65180.
mic analysis. Geophysical Prospecting 50, 527–538. Osdal, B., Husby, O., Aronsen, H.A., Chen, N. and Alsos, T.
Landrø, M. 2007. Attenuation of seismic water-column noise – 2006. Mapping the fluid front and pressure buildup using 4D
tested on seismic data from the Grane field. Geophysics 72, data on Norne Field. The Leading Edge 25, 1134–1141.
V87–V95. Parr, R., Marsh, J. and Griffin, T. 2000. Interpretation and
Landrø, M. 2008. The effect of noise generated by previous integration of 4-D results into reservoir management,
shots on seismic reflection data. Geophysics 73, Q9–Q17. Schiehallion Field, UKCS. 70th Annual International
Landrø, M., Digranes, P. and Strønen, L.K. 2001. Mapping Meeting. The Society of Exploration Geophysicists,
reservoir pressure and saturation changes using seismic 1464–1467.
514 M. Landrø
Rogno, H., Duffaut, K., Furre, A.K., Eide, A.L. and Kvamme, L. Tempone, P., Landrø, M., Fjær, E. and Inoue, N. 2009. Effects
1999. Integration, quantification and dynamic updating – on time-lapse seismic of a hard rock layer beneath a
Experiences from the Statfjord 4-D case. 69th Annual Inter- compacting reservoir (SPE-121081).
national Meeting. The Society of Exploration Geophysicists, Tura, A. and Lumley, D.E. 1999. Estimating pressure and satu-
2051–2054. ration changes from time-lapse AVO data. 69th Annual
Røste, T., Stovas, A. and Landrø, M. 2005. Estimation of layer International Meeting, The Society of Exploration
thickness and velocity changes using 4D prestack seismic Geophysicists, Expanded Abstracts, 1655–1658.
data. 67th Meeting. European Association of Geoscientists Tøndel, R. and Eiken, O. 2005. Pressure depletion observations
and Engineers, C010. from time-lapse seismic data at the Troll Field. 67th EAGE
Routh, P., Palacharla, G., Chikichew, I. and Lazaratos, S. 2012. Meeting, Abstract, C037.
Full wavefield inversion of time-lapse data for improved Vasco, D.W., Datta-Gupta, A., Behrens, R., Condon, P. and
imaging and reservoir characterization. SEG Expanded Rickett, J. 2004. Seismic imaging of reservoir flow
Abstracts, 1–6. properties: Time-lapse amplitude changes. Geophysics 69,
Sasagawa, G.S., Crawford, W., Eiken, O., Nooner, S., Stenvold, 1425–1442.
T. and Zumberge, M. 2003. A new sea-floor gravimeter. Virieux, J and Operto, S., 2009. An overview of full-waveform
Geophysics 68, 544–553. inversion in exploration geophysics. Geophysics 74, WCC1-
Settari, A. 2002. Reservoir compaction. SPE paper 76805, 62–69. WCC26.
Sheriff, R.E. 2002. Encyclopedic dictionary of applied geophys- Watts, G.F.T., Jizba, D., Gawith, D.E. and Gutteridge, P. 1996.
ics. Society of Exploration Geophysicists. ISBN 0-931830- Reservoir monitoring of the Magnus Field through 4D
47-8. time-lapse seismic analysis. Petroleum Geoscience 2,
Sirgue, L., Barkved, O.I., Van Gestel, J.P., Askim, O.J. and 361–372.
Kommedal, J. H. 2009. 3D waveform inversion on Valhall Zadeh, M.H. and Landrø, M. 2011. Monitoring a shallow sub-
wide-azimuth OBC. EAGE expanded abstracts, U038. surface gas flow by time lapse refraction analysis. Geophys-
Stovas, A. and Landrø, M. 2004. Optimal use of PP and PS time- ics 76, O35–O43.
lapse stacks for fluid pressure discrimination. Geophysical Zadeh, M.H., Landrø, M. and Barkved, O.I. 2011. Long-offset
Prospecting, European Association of Geoscientists and time-lapse seismic: Tested on the Valhall LoFS data. Geo-
Engineers 52, 301–312. physics 76, O1–O13.
Tempone, P. 2011. Effects of reservoir compaction on seismic Zimmer, M., Prasad, M. and Baggeroer, A.B. 2002. Pressure and
and gravity monitoring. PhD thesis, NTNU, 2011:6, ISBN porosity influences on Vp–Vs ratio in unconsolidated sands.
978-82-471-2519-9. The Leading Edge 21(2), 178–183.
Chapter 20
Interpretation of Marine CSEM and Marine MT Data for

Hydrocarbon Prospecting
Ståle Emil Johansen and Pål T. Gabrielsen
20.1 Introduction Controlled source electromagnetic (CSEM) sound-

ing uses EM energy transmitted by a horizontal elec-
Remote sensing techniques record variations in tric dipole (HED) source to detect contrasts in
petrophysical parameters such as acoustic or electric subsurface resistivity. Although EM techniques have
properties. Seismic sounding is by far the most common been used for many years, detection of subsurface
of such tools. Seismic techniques can detect strata and hydrocarbons by CSEM sounding is still a new disci-
structure in the subsurface that are potential hydrocar- pline within HC prospecting.
bon (HC) traps. By interpreting the seismic a detailed A full scale CSEM sounding research test was first
geological model can be constructed, but seismic data done offshore Angola in 2000 by Statoil, and showed
has limitations in direct prediction of pore fluid compo- that the CSEM technique could detect deeply buried
sition. Given detection of a structural geometry that hydrocarbons. The first commercial test was done in
may contain HC within porous sedimentary rocks, the 2002 by EMGS on the Ormen Lange Field offshore
main remaining uncertainty is normally whether the Norway. For some time the interpretation of CSEM
pore space is filled with saline water or HC. For this data was hampered by the lack of statistically signifi-
reason only 10–30% of exploration wells penetrate cant calibration data demonstrating that deeply buried
commercial oil or gas reserves in many areas. HC accumulations were detectable by the method.
Controlled source electromagnetic (CSEM) data However, development of new equipment opened the
has been used for several years for general mapping way for improved acquisition, processing and inter-
of electrical properties of the subsurface. In the late pretation of CSEM data, and in 2003 the Troll CSEM
1990s Statoil scientists Svein Ellingsrud, Terje data was collected by EMGS. This data was the first
Eidesmo and Ståle Johansen developed the use of irrefutable evidence for direct detection of a deeply
the CSEM method for remote identification of buried hydrocarbon accumulation by subsea CSEM
hydrocarbons in a marine setting. The early tests of sounding. The technique has later shown very good
the method were positive and in co-operation with results when applied correctly in the prospecting work
NGI, Statoil established EMGS that further developed flow. These results have opened a new frontier in
and commercialised the CSEM technique for hydro- hydrocarbon exploration.
carbon exploration. In recent years different types of CSEM acquisition
have evolved including horizontal dipole source with
seabed nodes, horizontal dipole source with cable
S.E. Johansen (*) towed receivers, and vertical dipole source with seabed
NTNU, Trondheim, Norway
nodes. Here we will describe the seabed node receiver
method using a horizontal towed dipole source.
P.T. Gabrielsen
The marine magnetotelluric method (MMT) is
EMGS (ElectroMagnetic GeoServices ASA), Trondheim,
Norway another offshore geophysical method measuring sub-
e-mail: [email protected] surface resistivity. This method is using the Earth’s
516 S.E. Johansen and P.T. Gabrielsen
Air 1010 Ω m
Water 0.33 Ωm
Not to scale
Source Rx Rx Rx Rx
Overburden 1–2 Ωm
HC Reservoir 30–500 Ωm
Half Space 1–2 Ωm
10–8 4
Electric magnitudes (V/A m2)
10–9 3
Normalised magnitudes
Dominated by
10–10 energy transmitted
2
via subsurface
10–11 1
Dominated by energy Dominated by energy
Subsurfa
transmitted via water and ce Resist transmitted via
or (HC R 0
10–12 sediment/water interface Halfs eservoir) air/water interface
pace
(no H
C Re
10–13 servo
ir)
–1
10–14 –2
0 approximately 10000 m
Source - Receiver Distance
Fig. 20.1 (Top) Schematic sketch of air-water-sediment geom- (Bottom) Electric magnitudes measured at a single receiver as
etry and receiver (Rx) layout on seabed during towing of a function of source-receiver distance. Red curve shows the
electromagnetic source. Black arrows denote refracted trans- expected response from a model including a high-resistivity hydro-
mission of electromagnetic signals via the air/water interface. carbon reservoir. (Lower curves) Blue curve is the significantly
Blue arrows denote direct transmission of electromagnetic weaker response from a model without a hydrocarbon reservoir.
signals through water and by refraction along the seabed. Red (Upper curves) To better visualise theoretical differences in
arrows denote guided transmission of electromagnetic signals amplitudes between the two cases the reservoir curve (red)
via a buried high-resistivity layer (hydrocarbon reservoir). was divided (normalised) by the no-reservoir curve (blue)
natural varying electromagnetic field as source. On land mainly controlled by variations in porosity, permeabil-
the method (MT) was introduced in the 1950s by the ity and pore connectivity geometries in addition to
geophysicists Cagniard and Thikonov. MT can improve pore fluid properties and temperature. It is the high
the interpretation of deep structures as it has very low resistivity of hydrocarbon filled reservoir rocks
frequency content, and it can image conductive (30–500 Ωm) compared with reservoirs filled with
sediments below thick resistors such as salt or basalt. saline formation water (0.5–2 Ωm) that makes EM
For this reason the MMT application has created inter- sounding a good tool for detection of subsurface HC.
est also in the hydrocarbon exploration industry. In marine CSEM sounding a horizontal electrical
dipole is towed close to the seabed emitting a low
frequency (0.1–20 Hz) signal which is recorded by
20.2 CSEM Basic Concepts stationary seabed receivers (Fig. 20.1). Seabed
receivers record the EM responses as a combination
20.2.1 Background of energy pathways including signal transmission
directly through seawater, reflection and refraction
Resistivity variations in rocks are generally controlled via the seawater/air interface, refraction and reflection
by the interplay between highly resistive minerals and along the seabed, and reflection and refraction via
pore fluids including low resistive saline water and/or possible high resistivity subsurface layers.
highly resistive hydrocarbons. Sedimentary rocks Low frequency EM signals decay exponentially
have a wide range of resistivities (0.2–1,000 Ωm) with distance. The distance required to attenuate an
20 Interpretation of Marine CSEM and Marine MT Data for Hydrocarbon Prospecting 517
EM signal by the factor e 1 (appr. 0.37) is defined as Forward modelling will give a good indication
p
the skin depth (503 ρf [m]). It is about 551 m in whether an EM survey will add new information or
seawater (0.3 Ωm) and 1423 m in 2 Ωm sediment and not. When data are acquired, adequate processing is
108 m in air (1010 Ωm) for a 0.25 Hz signal. EM critical for good data quality. To get the best result, the
signals are rapidly attenuated in seawater and seafloor EM data must also be interpreted as early as possible
sediments saturated with saline water, and these signal in the processing work flow. Inversion is an integrated
pathways will dominate at near source-to-receiver part of the interpretation work flow, and this is the best
offsets (~3 km). In high resistivity and relatively thin way to depth convert the EM data. When data are
(20–200 m) subsurface media, such as hydrocarbon inverted it is also possible to integrate EM data with
filled reservoirs (30–500 Ωm), the energy is guided other geophysical and geological data.
along the layers and attenuated less depending on the
property of the layer. Guided EM energy is constantly
leaking back to the seafloor and is recorded by the EM 20.2.3 Sensitivity Study and Survey Design
receivers. Energy is also refracted along the air/water
interface. This energy is commonly termed the air- Prior to acquiring CSEM data a sensitivity study is
wave and dominates at far offsets depending on performed. In this way we can find out if the defined
water depth and the subsurface resistivity. The target can be measured by the CSEM technique. A
refracted energy from high resistivity subsurface typical target is a hydrocarbon prospect. All available
layers will dominate over directly transmitted energy knowledge is gathered to build a representative Earth
when the source-receiver distance is comparable to model. Resistivity both for the prospect and the back-
twice the depth to this layer. The detection of this ground is included in the Earth model.
guided and refracted energy is the basis of CSEM as A prospect’s sensitivity to the CSEM method can
used when exploring for thin HC layers. be analysed by forward modelling. In the modelling
the expected CSEM response from the prospect is
calculated. A prospect’s sensitivity will depend on
20.2.2 CSEM Work Flow several factors such as: complexity of the background
geology, prospect size, prospect geometry, prospect
Typical steps in a CSEM work flow are shown in thickness, resistivity contrast between the prospect
Fig. 20.2, and in the following we focus on the steps and background rocks, and burial depth for the pros-
most important for interpretation of CSEM data. The pect. Based on the results from the sensitivity study the
sensitivity study is a vital step before time and money actual survey design is decided.
is invested in a survey. Here it can be evaluated if Sensitivity to a target can be defined by several
targets are within reach of the selected EM method. methods such as normalised magnitude (NMVO)
response or phase difference response (PDVO).
These attributes are created by dividing the target
response by the background response (Fig. 20.1).
The weakness of these measures is that they do not
consider the whole field (i.e. they look at magnitude
and phase separately) and they do not take the
expected data uncertainty and noise into account. Fig-
ure 20.3 shows a better way to consider sensitivity
which takes phase, magnitude, data uncertainty and
background noise into account. The figure shows a 1D
modelled magnitude and phase response at 1.0 Hz
from a background model (blue curve) and a model
Fig. 20.2 CSEM work flow from sensitivity study to interpre-
including a reservoir (red curve). The green curve is
tation and data integration. Seismic data and existing geological the absolute field difference (magnitude and phase) of
knowledge is necessary to fully utilise the potential of the these two responses and represents the scattered field
CSEM data caused by the resistive reservoir. The purple curve is
Fig. 20.3 Illustration on how sensitivity to a hydrocarbon and is the absolute difference between the background and
reservoir can be calculated. Both magnitude and phase vs. target response (both magnitude and phase included). The data
source receiver offset responses are shown for a target (red is sensitive to the target only if the scattered field response is
curve) and background (blue curve). The difference response larger than the expected data uncertainty and noise (purple
(green curve) represents the scattered field from the target only curve). See text for details
the assumed data uncertainty and noise when acquir- maximum sensitivity in this example is at approximate
ing the data. This is taken from an experience data 5.0 km source-receiver offset, and is the peak
base. For short to intermediate offsets the acquisition sensitivity.
measurement uncertainty will dominate the curve The left diagram in Fig. 20.4 shows a generalised
(scalable noise), and at long offset the background workflow for the sensitivity study where all prior
noise will dominate (additive noise). The latter is knowledge is used to build an Earth resistivity model
seen from 6.0 km offset where the noise curve flattens. for forward modelling. The results of the modelling
From these curves we can define the sensitivity by will decide whether the project is seen as not feasible
dividing the scattered field response (green curve) by (project end) or feasible. In the latter case the work
the expected data uncertainty and noise (purple curve). flow will continue with survey design and acquisition.
Hence a sensitivity <1 will have the scattered field Results from a sensitivity study are shown to the right
response lower than the data uncertainty and noise (no in Fig. 20.4. On the vertical axis is the peak sensitivity
sensitivity). This can be seen from offsets less than as shown in Fig. 20.3. We have divided the peak
2.0 km and offsets larger than 9.5 km. A sensitivity >1 sensitivity into three different zones: low, medium
is in the grey shaded area and means that the scatter and high sensitivity (red, yellow and green zones).
field response from the target is higher than the Low sensitivity is less than 1, and means that the target
expected data uncertainty and noise. This is where is not likely to be detected. Medium sensitivity is
CSEM is sensitive to the target. Note that even though between 1 and 3, and means that the target could be
the magnitude difference between the target and back- detected. High sensitivity is larger than 3, and means
ground at 5.5 km source-receiver offset is zero, the that the target is very likely to be detected. Along the
grey sensitivity area is large. This is because the phase horizontal axis the transverse resistance is plotted.
difference between the two scenarios at this offset is Transverse resistance is the product of the net pay
also large. Thus using the field response (magnitude thickness of the target and the resistivity of the target.
and phase) gives a better sensitivity measure than For example, a transverse resistance of 2,000 Ωm2
using the phase and magnitude separately. The could be a target with 50 m net pay and target
Fig. 20.4 CSEM sensitivity study procedure (left) and target target A has an identical target size and geometry as B, but a
sensitivity diagram (right). A, B and C show sensitivity for burial depth of 2,000 m below mudline. The curve describing
three different targets as a function of transverse resistance target C has the same target burial depth as B, but with a larger
(target net pay thickness and resistivity contrast). The curve target area. Red area is regarded as low sensitivity (<1), yellow
describing target B is the sensitivity to a given target with a area is medium sensitivity (1–3) and green area is high CSEM
burial depth of 1,500 m below mudline. The curve describing sensitivity
resistivity contrast of 40 Ωm, or a target net pay thick- reservoir resistivity) target A is not likely to be
ness of 40 m and a resistivity contrast of 50 Ωm. These detected, target B could be detected and target C
two scenarios will give approximately equal CSEM is very likely to be detected.
response if the burial depths are comparable. The In addition to forward modelling studies providing
transverse resistance is one of the main parameters CSEM sensitivity to a target, a sensitivity study can
that control the CSEM sensitivity. also include inversion of the synthetic data. An
To demonstrate the effect of burial depth, target extended sensitivity study is typically done when the
size and target form, three examples are included in initial sensitivity results are in the moderate zone
Fig. 20.4. Peak sensitivity for the three targets (A, (yellow). Based on a detailed sensitivity study, acqui-
B and C) with increasing transverse resistance is sition parameters such as frequency bandwidth and
plotted in the diagram. The curve describing target receiver-source configuration can be selected. It is
B is the sensitivity to a given target with a burial also worth noting that a sensitivity study is not, of
depth of 1,500 m below mudline. The curve course, the absolute truth. The selected parameters
describing target A has an identical target size and can still be uncertain. Therefore as a part of a complete
geometry as B, but a burial depth of 2,000 m below sensitivity study, the study itself should be evaluated,
mudline. The curve describing target C has the and the results risked before they are used in the
same target burial depth as B, but with a larger evaluation.
target area. If we follow the curve of target B we
can see that this target is not likely to be detected
by CSEM if the transverse resistance is less than 20.2.4 Data Acquisition and Processing
900 Ωm2 (e.g. 18 m net pay and 50 Ωm reservoir
resistivity), but is very likely to be detected if the Both 2D and 3D CSEM acquisition can be performed.
transverse resistance is larger than 2,500 Ωm2 (e.g. In early 2000, when the CSEM method was
50 m net pay and 50 Ωm reservoir resistivity). The established in HC exploration, only 2D data was
difference between the curves A, B and C shows acquired. However, since 2008 3D acquisition has
how the sensitivity will vary as a function of target been the preferred way to acquire CSEM data. It
area, shape and burial depth. For a given transverse provides better data coverage, and gives an improved
resistance value the sensitivity will increase with for image of the subsurface. Today 2D acquisition is typi-
example decreasing burial depth and/or increasing cally performed to tie CSEM survey lines to nearby
target area. For example at a transverse resistance wells and discoveries, or as long regional lines for
of 1,500 Ωm2 (e.g. 30 m net pay and 50 Ωm structural imaging.
Fig. 20.5 (a) 3D CSEM acquisition with a receiver grid layout and a dipole source. (b) Deployment of the source. (c) Deployment
of the receivers
When acquiring 3D marine CSEM data, receivers antennas will therefore be in random directions for
are placed on the seabed in a grid covering the area of each receiver.
interest. The source is towed above the receiver grid in When all the receivers are placed on the seabed the
a predefined pattern (towlines) while all receivers on CSEM source is deployed (Fig. 20.5b) and towed
the seabed are recording (Fig. 20.5a). A 2D CSEM above the receivers. The source is an electric dipole
acquisition has receivers placed only along one emitting an electromagnetic field. The horizontal elec-
receiver line and the source is towed directly above tric dipole antenna consists of two electrodes
these receivers. Hence, there are no receivers off the separated from each other, and with electrical contact
actual source towline. to the seawater. The source transmits a continuous and
The receivers consist of electric antennas and mag- designed signal with a low fundamental frequency.
netic coils typically measuring in two horizontal per- The frequency range spans from 0.1 to 20 Hz, with
pendicular directions (Fig. 20.5c). They are placed on the main energy output at the lower part of the spec-
the seabed simply by letting them sink freely down to trum. The peak current can vary from 1,000 to 8,000 a,
pre-planned positions. The orientation of the receiver depending on the actual design of the transmitter.
Fig. 20.6 Types of antenna configuration and EM fields. When inline and broadside position respectively. Receiver Rx003 is
the source- and receiver antennas are in line with each other the located at a position acquiring azimuthal data. This receiver will
configuration is called inline. When the antennas are parallel measure both inline and broadside data. A 2D survey with
and perpendicular to each other the antenna configuration is receivers only along the source direction (Rx001) will only
called broadside. Receivers Rx001 and Rx002 are placed in an record inline data
During a 3D acquisition the source antenna and resistivity of the Earth. This is one of the reasons
receiver antenna will form varying configurations. why 3D data acquisition is superior to 2D acquisition.
When the source- and receiver antennas are in line After data acquisition the receivers are released from
with each other the configuration is called inline and the seabed and data can be downloaded for processing
the data collected are inline data. When the antennas (Fig. 20.7). The recorded time series on each receiver
are parallel and perpendicular to each other the are calibrated and transformed to the frequency domain
antenna configuration is called broadside and the col- using Fourier transform. The pre-conditioning step
lected data are broadside data. In Fig. 20.6 the two involves dipole antenna scaling and stacking of dual
antenna configurations (inline and broadside) are channels on the receivers. The antennas of the receivers
shown. The position of receivers Rx001 and Rx002, are oriented randomly because receivers are sinking
with respect to the source position, are placed at an freely down to the seabed. When using the receiver
inline and broadside position respectively. Most of the data for quality control and attribute analysis, inline
antenna configuration within a 3D CSEM survey will and broadside components are calculated for all
be antenna configuration with an azimuth component. receivers. This is done through a processing step
Such data is called azimuthal data. Receivers Rx002 where receivers are mathematically rotated, and
and Rx003 are located in a position acquiring azi- receiver antennas become either inline or broadside
muthal data, and will measure both inline and broad- oriented. An example of rotated receiver data at
side data. 0.25 Hz frequency is given in Fig. 20.8. The receiver
The Earth is anisotropic and has different resistivity data is plotted as magnitude vs. source-receiver offset
properties in the vertical and in the horizontal direc- (MVO) and phase vs. source-receiver offset (PVO).
tion. The inline antenna configuration is mainly sensi- These data are used as input for attribute analysis and
tive to the vertical resistivity, while the broadside for inversion algorithms to produce Earth resistivity
configuration is mainly sensitive to the horizontal models.
Fig. 20.7 Simplified CSEM processing workflow from receiver (demodulation). Pre-conditioning involves receiver channel
data download to 3D resistivity cube. The recorded time series on summation and scaling of the source dipole moment. Before
each receiver are downloaded and calibrated. Each receiver is attribute analysis and inversion the receiver data is mathemati-
transformed from the time domain to the frequency domain cally rotated into inline and broadside components
Fig. 20.8 Magnitude vs. offset (MVO) to the left and phase vs. and describe offsets when the source is approaching the
offset (PVO) to the right for one receiver at frequency 0.25 Hz. receiver. Positive offsets are called out-tow and describe offset
Along the x-axis is the source-receiver offset where offset 0.0 is when the source is moving away from the receiver. Also com-
the location of the receiver. Negative offsets are called in-tow pare with Fig. 20.1
20.2.5 Inversion measure. In general terms an inversion problem can

be solved by finding the model m which explains the
When we do electromagnetic acquisition (CSEM and measured data d using an operator G. The operator G
MT) the goal is to find the resistivity distribution in the describes the explicit relationship between the model
subsurface. However, what we measure is the electric (m) and the observed data (d):
and magnetic fields recorded at the seabed. This is a
typical inverse problem as we seek a parameter (the m ¼ G 1 ðd Þ
resistivity of the Earth) which we cannot directly
For the CSEM and MT methods the model m is the d ¼ GðmÞ

Earth resistivity distribution that we are seeking and d
is the electric and magnetic field measured by the The final inverted resistivity model is obtained by
receivers. However, we have several problems with finding a model, m, which gives the minimum differ-
this equation. First of all we do not know the inverse ence between the synthetic data, d, and the measured
operator G 1. Second the operator G, which is forward data using e.g. least square methods. These inversion
modelling based on Maxwell equations, is nonlinear. results will depend on the start model used, the
Third the solution to our problem is ill posed. To regularisation terms, a priori information applied to
understand what ill posed means lets define a well constrain the results, and the number of dimensions
posed problem. A well posed problem means that a used in the forward modelling algorithm, G (1D, 2D or
solution exists. There is a unique solution (same data 3D). Due to the non-uniqueness of the solution,
measurement can only be explained by one model), performing inversion is just as much an interpretation
and the solution is stable (a small change in the model task as a data processing step.
creates a small change in the data measurement). If Figure 20.9 shows a schematic diagram of how this
one or more of these terms are broken we have an ill type of inversion works. Inversion of electromagnetic
posed problem. For CSEM and MT data our largest data starts by making a qualified guess of an initial
challenge is the non-uniqueness. This means that Earth resistivity model. This is the start model, m0.
several resistivity models can create the same data The start model can be similar to the background
measurements. We can also be challenged with an model used in the sensitivity study, or an updated
unstable solution. Because of the complexity of our model using new information from real data analysis
problem, we must do approximations and find the and lower dimension inversion results. Forward
solution in an iterative way. Instead of finding the modelling is performed on the start model using
Earth resistivity model by using the inverse operator receiver- and source geometry which is identical to
G 1 (which is unknown), we perform forward the real data acquisition configuration. The forward
modelling to simulate synthetic data based on a modelled data, d, are compared with the measured
given model: data. The difference between these two data sets
Fig. 20.9 Flowchart for inversion. See text for details

Fig. 20.10 Unconstrained (upper) and constrained (lower) 3D on a very good understanding of the background resistivity.
CSEM inversion results are overlain depth migrated seismic This knowledge is used in the start model for the constrained
data and well data (blue line). The example is from the Western inversion. The constraints favour resistivity model updates
Troll oil province in the North Sea, offshore Norway. The within the reservoir level defined by seismic horizons. Data
improved resolution of the reservoir is clearly seen in the courtesy EMGS
constrained inversion result. Note that this example is based
gives the data cost which expresses how well the start be regarded as restrictions to the inversion solution in
model explains the measured data (misfit). order to comply with these data. They will therefore
Because the inversion solution is non-unique it limit the solution space and force the inversion
may be necessary to introduce regularisation. In the towards a particular solution. Examples of constraints
regularisation process additional information is added can be that the inversion is only allowed to update the
to solve the ill-posed problem. In our case regulariza- model within a certain layer defined by seismic
tion terms could for example be to seek a smooth horizons (Fig. 20.10), or that the updates within a
horizontal layered resistivity model. The data- and layer should be within a resistivity interval taken
regularisation cost adds up to the total cost function, Φ: from well log data. Constrained inversion will usually
give an Earth model with improved resolution. How-
Φ ¼ Data cost þ Regularisation cost ever, it is important to note that the restrictions put on
the model update can also lead to a wrong result. If the
The goal of the inversion is to update the resistivity assumptions used in the constraints are wrong, the
model in such a way that the total cost function is solution will also be wrong. Constrained inversion is
minimised. One loop within the flowchart involves therefore usually performed in areas where one has
forward modeling of a model and calculation of the good prior knowledge, and it is recommended that an
cost function. If the cost function (or misfit) is unac- unconstrained inversion is always performed for
ceptable the inversion performs new updates of the comparison.
resistivity model (miþ1) and starts a new loop. A The final output from an unconstrained or
complete loop is called one iteration within the inver- constrained inversion is an Earth resistivity model.
sion. The inversion iterates until the total misfit goes This model is only one of several possible solutions
below a certain value, or is stopped by the user when it which fit the measured data, and the final results
converges towards one value with respect to iteration should be carefully evaluated. First, the entire data
number. misfit, or how well the synthetic data from the model
An inversion process that only uses the electromag- describes the measured data, should be inspected. Has
netic data itself and no other input is called an uncon- the total misfit from the inversion converged with
strained inversion. A further step in the inversion increasing iterations? Are there areas where the misfit
process is to apply constraints to the model based on is too large? Second, the final resistivity model needs
other geological or geophysical data. Constraints can to be evaluated with respect to the general geological
model. Do the resistivity variations follow the subsurface. It will mainly provide information about
expected trends? How well do resistivity anomalies lateral resistivity changes, but can also give rough
fit with the anticipated results? If the results are not as estimates of burial depth to a resistor. A deeply buried
expected, it does not necessarily mean that the inver- resistor will show up at larger source-receiver offsets
sion results are wrong. It can just as well be that the and lower frequencies than a shallow resistor. As a
existing geomodel is wrong and needs to be redefined. rule of thumb the burial depth to a resistor can be
It is within this integrated interpretation process approximated to half the source-receiver offset at the
between electromagnetic-, seismic- and well-data onset of the anomaly. This approximation assumes a
that valuable additional information can be extracted. 1D Earth. Interpretation of receiver data attributes
(such as NMVO and PDVO) is important as it directly
evaluates the recorded data. The Troll case is included
20.2.6 Interpretation and Integration as an example of receiver interpretation. In the other
of CSEM Results cases the interpretation is based on inversion results.
Inverted resistivity models can be directly
Interpretation of CSEM data involves explaining the integrated with depth converted seismic and well log
resistivity changes that we observe from real CSEM data. When interpreting CSEM inversion results it is
data, and relating these to lithology and fluid changes useful to look at vertical depth sections as well as
in the subsurface. The interpretation needs to be integrated resistivity maps. Depth windows for resistivity maps
with other geophysical and geological data to build a can be selected by constant depth intervals, or to
geological model that honours all the available data. follow seismic depth converted horizons. Such maps
When interpreting CSEM data it is important to give the lateral outline of resistive anomalies which
keep in mind that the ability to detect hydrocarbon can be correlated with seismically defined structures.
reservoirs is not only dependent on the resistivity It is also important to quality control and interpret
contrast between the reservoir and the surroundings, the CSEM data at different processing levels from
but also on burial depth, target size and net pay thick- attributes in the receiver domain to full 3D inversion.
ness of the reservoir. Furthermore the acquisition Figure 20.11 shows an example where a resistor is
geometry will also be of importance. A dense receiver consistent at different processing steps. A significant
grid will be able to detect smaller targets than a coarse normalised magnitude anomaly (purple shadow) is
receiver grid. The presence of a resistor could indicate clearly visible early in the processing (receiver level,
a hydrocarbon reservoir, but could also be explained upper left picture). In this image the source-receiver
by other resistivity anomalies. It could for example be offset is plotted along the vertical axis. The burial
caused by low porosity rocks (e.g. cemented or tightly depth to the resistor can be approximated to 1.5 km
compacted sandstones), highly organic-rich shale which is half the source-receiver offset of the anomaly
(source rocks), carbonates, volcanic rocks, salt or onset. The same anomaly also stands out in the 2.5D
other high resistivity strata or structures in the subsur- inversion result along the line (middle left picture).
face. Further, integration of seismic data, well logs and The burial depth and lateral extent agree with the
geological understanding will help to reduce the pos- initial interpretation of the NMVO attribute. The
sible explanations of an observed resistivity anomaly. lower left picture shows the 3D inversion result
Interpretation of CSEM data can be done at a very along the same acquisition line and the areal extent
early stage in the processing flow by analyzing relative of the anomaly is shown in map view in the upper right
differences between receivers. This can for example picture in Fig. 20.11. The anomaly also conforms to a
be done on normalised magnitude vs. offset (NMVO) faulted structure when overlain seismic data (lower
displays, or phase difference vs. offset plots (PDVO). right picture). In this case the analyses done on the
An increase in NMVO indicates a resistivity increase CSEM data at different processing levels are consis-
in the subsurface relative to the reference receiver tent, and indicate a subsurface resistor with significant
used for the normalisation (Fig. 20.1). Similarly a contrast to the surroundings. However, the detailed
decrease in PDVO (Fig. 20.3) will indicate a resistivity nature of the resistor can only be understood by
decrease in the subsurface. Such interpretations will using all available geophysical and geological knowl-
give a first indication of resistivity changes in the edge in the area.
Fig. 20.11 Real data example from the Norwegian Sea through 2.5D inversion (middle left) and 3D inversion (lower
showing a resistive anomaly which is consistently identified left). The map outline of the anomaly also correlates at depth
from normalised magnitude vs. offset attribute (Upper left) with a faulted structure (right). Data courtesy EMGS
20.2.6.1 Electrical Anisotropy 20.2.6.2 Apparent Anisotropy

Sediments and layered strata can be strongly electri- As explained above, 3D CSEM acquisition provides
cally anisotropic. This is a combination of intrinsic both inline- and broadside data. The inline data is
anisotropy (e.g. anisotropy within the shale) and sensitive to the vertical resistivity (Rv) in the Earth,
large-scale anisotropy (e.g. anisotropy due to thin while the broadside data is sensitive to the horizontal
bed layering). If anisotropy is not taken into account resistivity (Rh) in the subsurface. Azimuthal data
it can easily lead to false inversion- and interpretation measures both inline and broadside data, and is there-
results. Most forward modelling and inversion fore sensitive to both horizontal and vertical
algorithms assume a TIV media (Transverse Isotropic resistivity.
with a Vertical axis of symmetry). This means that In addition it is only the in-line data that is sensitive
electrical properties of subsurface rocks are described to thin horizontal resistive layers. The broadside data
by a horizontal and a vertical resistivity component. is not able to detect thin horizontal resistive layers, but
The anisotropy is defined as the vertical divided by the measures background resistivity trends. This is
horizontal resistivity component. Typically, vertical because horizontal currents will flow around a thin
resistivity will often be higher than horizontal resistiv- resistor while vertical currents will not. The situation
ity due to the horizontal layering of beds and strata. is opposite for thin vertical resistors. Here vertical
Today most resistivity well log tools only measure the currents can flow around the resistor while horizontal
horizontal resistivity. In such cases it is only possible current will not. In this case the vertical resistor will be
to directly calibrate the horizontal resistivity model to imaged better in the horizontal resistivity model than
the well logs. in the vertical resistivity model.
When tri-axial resistivity logging tools measuring The difference in sensitivity can be illustrated by
both directions become more common, calibration using a synthetic example (Fig. 20.12). The upper
also of the vertical resistivity model derived from panels show the Earth model’s vertical (left column)
CSEM data will be possible. Better understanding of and horizontal resistivity (right column). The thin
the electrical anisotropy will improve our ability to resistor (red target) is a hydrocarbon reservoir with
interpret CSEM data. equally high vertical and horizontal resistivity.
Fig. 20.12 Unconstrained 3D inversion of synthetic CSEM (b). Due to difference in sensitivity only the inverted Rv model
data. (a) The resistivity model used for forward modelling, images the thin target, the inverted horizontal resistivity model
vertical resistivity model, Rv, (left) and horizontal resistivity images the background resistivity. See text for a more detailed
model, Rh, (right). The inverted resistivity result is shown in explanation
Hence, the anisotropy in the thin hydrocarbon layer is layer is not measured, only the background horizontal
equal to one. Based on this Earth resistivity model, resistivity is detected. Apparent anisotropy can be
synthetic 3D CSEM data are simulated, and uncon- used as an interpretation attribute for better separating
strained 3D inversion is performed on the synthetic thin resistors, possibly associated with hydrocarbon
data. The inversion result for both the vertical and reservoirs, from thicker resistors associated with back-
horizontal resistivity models is shown in the lower ground variations. Both thin and thick resistors will
panel in Fig. 20.12. The results confirm that the show up as resistive anomalies in the vertical resistiv-
CSEM method is not sensitive to the horizontal resis- ity model, but only the former will also have an appar-
tivity within the thin layer. It is only the inverted ent anisotropy anomaly.
vertical resistivity model that reconstructs a resistive
anomaly at reservoir level. The inverted horizontal
resistivity model does not reconstruct the thin reser- 20.3 Marine Magnetotelluric: Basic
voir, but provides an image of the background Concepts
resistivity.
For these reasons, thin resistors will show high 20.3.1 Background
electrical anisotropy (Rv/Rh). We call this apparent
anisotropy as it is not the “true” anisotropy in the Marine magnetotelluric (MMT) is an offshore geo-
thin layer, but rather an effect of the difference in physical method measuring the Earth’s resistivity
sensitivity between inline- and broadside data. There- using the Earth’s natural varying electromagnetic
for, for thin resistors the horizontal resistivity of the field as source. Magnetotelluric (MT) on land has
existed for many years, but it is only in the last decade receiver spacing of 1–5 km. Often a MMT survey is
it has been widely used offshore. While CSEM is used done in combination with CSEM acquisition. How-
primarily for detection of relatively shallow and thin ever, due to the low frequency signal it is important
horizontal resistors such as hydrocarbon reservoirs, that enough listening time, undisturbed by the con-
MMT can only image larger-scale structures. It can trolled source, is allowed when planning a survey.
improve the interpretation of deep structures, and due Typically it is necessary to listen for several days in
to the very low frequency content the energy can order to get enough samples to describ the low fre-
penetrate several tens of kilometres into the subsur- quency signal (Fig. 20.13). Required listening time
face. More importantly it can image conductive strata will also vary with latitude. Close to the equator the
below thick resistors such as salt or basalt. Because of signal is weaker, and longer listening time is needed to
this, MMT application has also created interest in the maximise the number of periods describing the low
hydrocarbon exploration industry. Presence of salt or frequency signal.
basalt in a basin often complicates seismic imaging, MMT surveys in areas with large water depth will
and MMT’s ability to see through such layers has only obtain the lower frequency signals. This filtering
shown to be very valuable. of the high frequencies has implications for the imag-
The naturally varying electromagnetic field is set ing capabilities and will result in reduced resolution in
up by two external sources. The low frequency content the shallowest part of the section.
(<1 Hz) is caused by solar winds (stream of protons
and electrons) interacting with the Earth’s magneto-
sphere, while the high frequency content (>1 Hz) is 20.3.3 Basic Interpretation of MMT Data
worldwide lightning, mostly around the equator,
which charges the Earth’s ionosphere. These complex The Earth resistivity can be described by impedance
processes create a plane electromagnetic field tensors. These tensors can be found by analysing the
travelling vertically downward through the atmo- relationship between the recorded E and H fields (elec-
sphere. Most of the energy is reflected back again tric and magnetic). The processed receiver data results
when it hits the Earth surface, but part of the signal in apparent resistivity and impedance phase as a func-
penetrates into the subsurface. This field induces tion of the signal period (T). In Fig. 20.14 the TE- and
currents in the Earth called telluric currents which TM modes are plotted as a function of period. The
again set up a secondary magnetic field. The strength transverse electric mode (TE) is the electrical
of the secondary field is dependent on the electrical polarised mode where the electric field is in the geo-
properties in the Earth. Measuring this signal will logical strike direction. The transverse magnetic mode
therefore provide information about subsurface resis- (TM) is the magnetic polarised mode where the mag-
tivity. In marine magnetotelluric it is the low fre- netic field is in the geological strike direction.
quency signal set up by the interaction between the The apparent resistivity plot can be directly
solar wind and the magnetosphere that will dominate. interpreted as it gives the resistivity average of the
The conductive seawater will filter out all frequencies subsurface. The time period can roughly be transferred
above approximately 1 Hz. The signal strength also to minimum approximate burial depth by using the
pffiffiffiffiffi
varies with latitude. Areas close to the poles have skin depth ðδ 500 ρT Þ. For example a signal period
stronger source signals than areas close to the equator. (T) of 10 s with apparent resistivity (ρ) of 1.5 Ωm in
Fig. 20.14 (upper) corresponds to a minimum burial
depth of 2.0 km. The two different modes start to
20.3.2 Acquisition of MMT Data separate at a signal period of 100 s (Fig. 20.14,
lower), indicating that there is a deeply buried three
An MMT survey is acquired by placing receivers on dimensional structure at minimum 7 km depth.
the seabed. The MMT signal is extremely weak, so Displaying the apparent resistivity or impedance
high sensitivity receivers, recording both the electric phase for a complete receiver line with the signal
and magnetic fields, are needed. Receivers can be period along the y-axis, and the different receiver
placed as a single line, or in a receiver grid with typical responses along the x-axis can therefore provide a
Fig. 20.13 Receiver recording as a function of time (days recorded during day time. The grey square indicates a time
along x-axis) and frequency (y-axis). Both the controlled source window which can be used for MMT processing. This window
signal (Christmas trees) and the time-varying MMT signal are is not influenced by the controlled source acquisition
visible in the plot. Note that the strongest MMT signal is
Fig. 20.14 Apparent resistivity (upper) and impedance phase transverse magnetic mode (TM) is the magnetic polarised mode
(lower) processed from a recorded electromagnetic signal. Blue where the magnetic field is in the geological strike direction.
circles are the TE mode and red crosses are the TM mode. The The separation of the two modes with increasing period
transverse electric mode (TE) is the electrical polarised mode (decreasing frequency) suggests that there is a 3D structure at
where the electric field is in the geological strike direction. The large burial depth
Fig. 20.15 Apparent resistivity display for the TE mode where in resistivity. This could for example be interpreted to represent
lateral distance is along the x-axis and signal period (~depth) is the basement of a sedimentary basin. Stand-alone interpretation
along the y-axis. Such displays can quickly give an idea of the of MMT data can be difficult, and integration with other geo-
dominant structures in the subsurface. Here the transition from logical or geophysical data is important
blue to yellow colours indicates the depth to a significant change
good indication of the main subsurface structures 1,400 m below sea level. Water-bearing Sognefjord
(Fig. 20.15). Fm. sands and overburden sediments show resistivities
The apparent resistivity and impedance phase data in the 0.5–2 Ωm range. The TWGP was very well
is used as input to inversion. The inversion transforms suited for CSEM surveying, due to the high reservoir
the recorded data to a depth converted resistivity resistivities, well defined field edges, the low and
model (see inversion section for details). MMT inver- relatively constant resistivities in the geological layers
sion algorithms only invert for the horizontal resistiv- above the reservoir (overburden), constant water
ity model as the natural source is dominantly a plane depth, smooth seafloor, and the moderate distribution
horizontal field. The polarisation of the field is also of the HC filled reservoir.
why the MMT method will see through horizontal
resistors such as basaltic layers. MMT data is used in
combination with CSEM data to map structures
20.4.2 CSEM Forward Modelling
beneath basaltic layers. MMT data can also be effec-
tive in mapping subsurface salt structures.
When interpreting CSEM data at receiver level the
CSEM response over the HC accumulation is com-
pared with the CSEM response in a reference area
20.4 Troll Case Study immediately outside the accumulation. It is also criti-
cal to understand CSEM responses from high resistiv-
20.4.1 Troll West Gas Province ity bodies, other than the HC reservoir itself, which
can potentially generate significant responses.
In this case study we present CSEM data across the Three dimensional modelling shows how the
Troll West Gas Province (TWGP), offshore Norway. CSEM response is affected by the geometry of the
The case is included as an example of pre-inversion reservoir and the effect of varying receiver layout
interpretation of marine CSEM data at receiver level. geometries. With the ultra-low frequencies (0.25 Hz)
The Troll Field complex is the largest gas discovery used in CSEM sounding, it is the resistivity variations
on the Norwegian Continental Shelf, located in in the illuminated subsurface that dominate the
300–360 m of water in the northeastern part of the response measured at the seafloor. All measured and
North Sea. The field is subdivided into three separate modelled CSEM responses are normalised to a refer-
compartments. Troll East is by far the largest and ence case with no subsurface HC accumulation.
contains two thirds of the HC reserves. For the In this early case study, modelling had two main
CSEM survey the much smaller TWGP was selected goals; firstly, to establish the optimal survey location
(Fig. 20.16). The reservoir interval is Jurassic and receiver geometry; and secondly, to quantify the
sandstones (Sognefjord Fm.) with up to 160 m gas expected CSEM response from the subsurface HC
column, and a thin oil leg at the base. HC-filled accumulation relative to that of a reference area out-
sands show high average resistivities around side the accumulation.
200–500 Ωm, and occur at a burial depth of about
Fig. 20.16 Geological section across Troll West Gas Province Thin line is towline for CSEM source and thick line indicates
(TWGP) together with resistivity data from exploration well 31/ approximate position of CSEM sea floor receivers
2-1. Outline of field and survey layout is shown on inset map.
The geological models and results from the shown in Fig. 20.18. The selected ON and OFF
modelling are shown in Fig. 20.17. Average resistivity receivers record near identical magnitudes out to
(Ωm) values used in the modelling are based on avail- 3.5 km, but between 3.5 and 9.0 km offset the ON
able well information. A TWGP reservoir resistivity of receiver shows systematically higher magnitudes
250 Ωm is calculated from trimmed mean resistivity compared to the OFF receiver. The corresponding
values from well 31/2-1 (Fig. 20.16). There is a strong normalised MVO predicted from forward modelling
increase in simulated CSEM response above TWGP is also shown in the figure and there is good agreement
along the selected survey line. The modelled response between the forward simulations and the measured
also increases with increasing reservoir width, and for MVO values. Normalised MVO data for the OFF
5 km width the simulated response is more than three receiver are close to 1.0 which is in accord with
times stronger than the response from a non-reservoir subsurface sediments comprising only water-wet
case. lithologies along this line segment.
Normalised MVO data for the ON-receiver are
close to 1.0 out to ca 3.5 km but increase gradually
to levels above 3.0 at 7.5 km offset before gradually
20.4.3 Results and Interpretation
decreasing with higher offset. Magnitudes measured
by the ON-receiver are reasonably similar to the
The survey was acquired in 2003 and designed with 24
modelled response for TWGP, and in accordance
EM receivers across TWGP as illustrated in
with the position of this line segment over the main
Figs. 20.16 and 20.18. Data quality had reliable infor-
gas accumulation (Fig. 20.18).
mation to an offset of c.8 km. Normalised values of
In order to display CSEM sounding systematics
electric magnitudes vs. offset measured by two repre-
along the entire survey line, median normalised
sentative receivers, one above (ON) the HC accumu-
magnitudes were calculated at 6.5 0.5 km offset
lation and one outside (OFF) the HC accumulation, are
On-Reservoir Off-Reservoir
5.0 0
4.5 0.33 Ω m 0.33 Ωm

0.33 Ω m 200
Receiver
4.0
Model E
400
Depth below Sealevel (m)

3.5
Model A 5 Ωm
Normalized magnitudes
600
3.0
1.4 Ωm Model B
Model F
2.5 800 1 Ωm
1.4 Ω m 2 Ωm
2.0 1000
Model C
1.5
Model D 1200 1 Ωm
1.0
1400
0.5 250 Ω m 2 Ωm
250 Ω m 2 Ωm
1600
0.0 1.5 Ω m 1.5 Ω m
1.5 Ω m
–0.5 1800
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Source - Receiver Distance (m)
Fig. 20.17 Modelled CSEM responses from simplified Troll modelling (dashed lines) are included. Models A, C and D are
geological models. Columns to the right show plane-layer geo- responses from varying reservoir widths (5 km, 2.5 km and
logical models representing ON and OFF reservoir situations. 1.25 km respectively). In model B the width increases from
The differences in overburden resistivity between ON and OFF 3.5 km (left) to 5 km (right), representing an oblique crossing
reservoir are due to lateral lithological changes. The cross-section of Troll reservoir. Models E and F are responses from plane-layer
is a modification of the east-west cross-section in Fig. 20.16. modelling
Results from both 3D modelling (solid lines) and plane-layer
for all receivers. This intermediate offset interval by the survey. Due to extensive seismic surveying and
records close to maximum difference in measured drilling in this part of the North Sea, the geology of the
magnitudes ON and OFF TWGP. Median normalised area is very well known.
magnitudes at 6.5 0.5 km offset for all receivers are
illustrated in Fig. 20.18, with data plotted at 3.25 km
offset halfway between source and receivers. Median 20.5 Gulf of Mexico Case Study
normalised magnitudes are around 1.0 south west of
TWGP (OFF), increase gradually up to 2.7 over the CSEM data and seismic data measure different Earth
apex of the HC accumulation (ON), and decrease properties. The CSEM method is sensitive to resistiv-
gradually down to 1.2 over the northwestern fringes ity variations, while the seismic method is sensitive to
of the field. The gradual increase is caused by the fact velocity and density variations. Both methods are
that both source and receiver must be above the target therefore indirectly sensitive to HC saturation in a
to record maximum returned energy from the HC reservoir. The seismic method can normally see
accumulation. For TWGP this is the case only for a deeper into the subsurface than the CSEM method. It
narrow zone above the middle part of the HC accumu- is always important to control that the measured
lation. We also see that the western slope of the anom- targets are well inside the detectability range for the
aly is steeper than the eastern. A likely explanation for method used. Especially for the CSEM method sensi-
this is the asymmetric shape of the accumulation with tivity studies should always be performed to check if
a fault termination at the western extension and a more the actual prospect is within reach of the method.
gradual eastward thinning as shown in Fig. 20.18. Seismic velocity will have a sharp decrease when
There are no observations of high resistive strata or introducing even small amounts of gas in a water-filled
structures that we know of, other than the TWGP, sandstone reservoir. Since oil has comparable
which can explain the high resistivity values measured velocities to saline water the velocity effect is much
3.5
3.0
Normalised magnitudes 2.5
2.0
1.5
1.0
0.5
0.0
10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0
Offset (m)
3.5 0
Water 1 5 10 20 24 200
15
3.0
Depth below Sea level (m)

400
Normalised magnitude
2.5
600
at 6.5 km offset
2.0 800
Sediments 1000
1.5
1200
1.0
1400
TWGP
0.5
1600
0.0 1800
0 5000 10000 15000 20000 25000
Horizontal Distance (m)
Fig. 20.18 (Top) Modelled and measured CSEM receiver data response normalised by the modelled OFF reservoir response.
across Troll. Red curve represents a typical response from a (Bottom) Median normalised magnitudes at 6.5 0.5 km
receiver measurement ON the reservoir (Rx 24) divided by a source-receiver distances along the survey line across Troll.
(normalised) reference receiver OFF the reservoir (Rx 2). The Normalised magnitudes are posted at common midpoints
source was towed SW-NE and shown responses are on the SW in- 3.25 km from receivers. For better visualisation median
towing side of both receivers. Blue curve is the reference receiver magnitudes are overlain on the simplified geological model
normalised by the modelled plane-layer response representing the along the survey line. The correlation between the location of
OFF reservoir situation. White curve is the modelled ON reservoir the CSEM anomaly and the Troll reservoir is excellent
smaller if oil, instead of gas, is present in the reservoir. indications of increased hydrocarbon saturation the
If a HC-saturated reservoir is to be detected by the chance to find a reservoir with hydrocarbons is signifi-
CSEM method the hydrocarbon saturation has to be cant. If only the seismic data shows indication of
high. It must reach 40–60% before the resistivity will hydrocarbon the chance to find a high saturation reser-
significantly increase (Fig. 20.19) and the reservoir voir is reduced. The evaluation will depend on the
can be detected by CSEM sounding. Oil and gas will expected fluid response in the seismic data. If only
have similar effects and both will increase the resistiv- the CSEM data shows indications of high resistivity,
ity in the reservoir. As a result of this it is challenging with no hydrocarbon indications in the seismic data,
for the CSEM method to distinguish between different lithology and other non-HC interpretations should also
HC phases in the subsurface. be carefully evaluated.
The relationships and properties mentioned above This Pemex case study demonstrates how seismic-
have implications for interpretation of a given HC and CSEM data can be used as a combined DHI tool
prospect. If both seismic data and CSEM data show for better HC prediction. The prospect is located in
104
Seawater/gas in 50% porous sandstone
1.6
Resistivity (Ωm, log scale)

103
P-wave velocity (km/s)

’s
hie
1.2
Vel
c
Ar
102
ocit
y 0.8
101
Fizz gas
100 0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Gas saturation
1.0 0.5 0.0
Water saturation
Fig. 20.19 P-wave velocity and resistivity as a function of hydrocarbon saturation). Note that the resistivity curve is loga-
hydrocarbon saturation (modified from Constable 2010). Acous- rithmic. Commercial hydrocarbon reservoirs typically have
tic p-wave velocity is sensitive to small volumes of gas in the water saturation lower than 0.4. The curves show that it is
porewater, but not very sensitive to further increasing gas possible to differentiate low saturated hydrocarbon reservoirs
saturation. Introducing oil instead of water in the pore space from high saturated hydrocarbon reservoirs by combining seis-
only has a moderate impact on p-wave velocity. Resistivity mic and CSEM data
shows a large increase for low water saturation (high
deep water in the Gulf of Mexico. The trap is an have low hydrocarbon saturation as there is no resis-
anticline structure with a well-known regional reser- tive anomaly associated with the seismic amplitude
voir interval approximately 700 m below the mud line. anomaly.
The prospect is covered by both 3D seismic and 3D This combined seismic and CSEM interpretation
CSEM data. was confirmed by a well showing high gas saturation
A seismic amplitude anomaly defines the maxi- and good reservoir sand at the top of the structure
mum extent of the prospect, and a seismic flat spot down to the flat spot. The same sand was also
defines a possible internal gas/fluid contact identified in an appraisal well down-dip on the struc-
(Fig. 20.20). One possible model explaining these ture, but here the reservoir sandstone had low hydro-
seismic observations is that the flat spot represents carbon saturation as predicted in the interpretation.
the gas/oil contact, and that the seismic amplitude
anomaly below the flat spot is caused by presence of
oil in the reservoir. 20.6 Barents Sea Case Study
From the 3D inverted CSEM data we see that a
resistive anomaly is consistent with the outline of the This case study shows how the combination of 2D
seismic flat spot (Fig. 20.21). There is no CSEM seismic data and 3D CSEM data can be interpreted
anomaly associated with the seismic anomaly below and used in exploration. 3D CSEM data covers the
the flat spot. The combined DHI observations are Johan Castberg field (Skrugard and Havis discoveries)
interpreted to be due to variation in hydrocarbon satu- on the Polheim sub-platform in the Barents Sea
ration within the prospect. The crest of the structure (Fig. 20.22). The survey also includes the exploration
was interpreted to have high saturation gas as there is a well 7219/9-1 to the southwest of the field. The
strong DHI indicator in both the seismic- and CSEM Skrugard and Havis discoveries made in 2012 were
data. The lower part of the structure was interpreted to based on observations on 3D seismic data (double flat
Fig. 20.20 Seismic depth slice showing outline of seismic fluid contact. One possible model explaining these seismic
anomaly (red dotted line) and flat spot (white dotted line). A observations is that the flat spot represents the gas/oil contact,
seismic amplitude anomaly defines the maximum extent of the and that the seismic amplitude anomaly below the flat spot is
prospect, and a seismic flat spot defines a possible internal gas caused by presence of oil in the reservoir. Data courtesy Pemex
Fig. 20.21 Inverted resistivity data overlain seismic depth (pink colours). The combined DHI observations are interpreted
slice from Fig. 20.20. From the 3D inverted CSEM data we to be due to variation in hydrocarbon saturation within the
see that a resistive anomaly is consistent with the outline of the prospect. Included are the positions of the gas discovery well
seismic flat spot (blue-yellow-red colours). There is no CSEM and a low saturation gas appraisal well. Data courtesy Pemex
anomaly associated with the seismic anomaly below the flat spot
spot), but CSEM data also played an active role in on the seabed with a spacing of 3 km. The source was
evaluation of the prospects. The case example shows towed over each receiver line in a south-north orienta-
how CSEM data can detect the hydrocarbons in place tion. The coarse receiver grid was used to effectively
and distinguish between the dry well (7219/9-1) and scan a large area and is typically used in frontier areas
the discovery wells. It is therefore an excellent case to where limited information is available. Here the acqui-
be used as a reference for further work in similar play sition geometry was designed to cover a given area,
models in the area. and was not based on prospect outlines. Such a coarse
The CSEM data was acquired in 2008 as a 3D receiver spacing will put some limitations on the target
survey where receivers were placed in a coarse grid size that the CSEM data are able to detect. Ideally the
Fig. 20.22 Average CSEM maps from unconstrained 3D Black squares are receiver positions. Solid lines A and B show
inversion calculated in a window 500 m above to 500 m below line location for seismic lines in Figs. 20.23 and 20.24. The blue
the Base Cretaceous unconformity. (a) Horizontal resistivity, square in the small overview map shows survey location in the
Rh, (b) Vertical resistivity, Rv and (c) Apparent anisotropy. Barents Sea. Data courtesy EMGS
prospects that can be detected should be within size for generate new leads and prospects not yet identified by
a commercial discovery in the region. In the Barents the seismic. Especially stratigraphic traps are often
Sea such CSEM grids are actively used by oil difficult to identify on seismic data.
companies for licencing round exploration, and also In 1988 well 7219/9-1 was drilled at the crest of a
in some cases for drilling decisions. fault block with main targets in Jurassic sandstones
Identifying prospective areas based on CSEM data (Stø, Nordmela and Tubåen formations). A total thick-
alone is challenging. Also 2D seismic data will have ness of 350 m of sandstones was encountered with
its limitations due to coarse coverage and possible porosities ranging from 16 to 18%. In addition, a
imaging problems. However, the combination of the thinner section of sand in the Lower Cretaceous sec-
two data types can be a powerful tool in exploration. tion was found (Knurr formation). The well was dry
Seismic data is used for geological understanding, with oil shows. Some years later Statoil made the two
structural mapping, play development, and to map new discoveries, Havis and Skrugard, in a similar play
prospect outlines. Resistivity trends and anomalies model further to the northeast (Fig. 20.22).
observed in the CSEM data can be better understood When interpreting 3D CSEM data it is useful to
when integrated with seismic data. In this way, lack of extract resistivity maps centred on target horizons in
3D seismic information can to some extent be order to get a lateral overview of resistivity variations
compensated for by the 3D resistivity information and anomalies. As explained in Chap. 2, a horizontal
obtained from CSEM data. The EM data can also, in resistivity (Rh) map will provide information about the
some cases, carry structural and geometric informa- regional trends, while a vertical resistivity (Rv) map
tion that can guide the interpretation between the seis- will show regional trends as well as thin horizontal
mic 2D lines. Once a prospective area has been resistors (potential hydrocarbon reservoirs). A map of
identified on seismic, 3D CSEM data can be used to the anisotropy attribute (Rv/Rh) can also be created.
upgrade or downgrade prospects. CSEM data can also Such a map will show the regional anisotropy, but will
in addition highlight areas where thin horizontal attribute and the position of the discovery well. The
resistors are present. resistivity anomaly is located at the crest of the fault
In Fig. 20.22 resistivity maps (Rh and Rv), and block and correlates with the Skrugard HC discovery.
anisotropy map created from unconstrained inversion Here, CSEM data has the potential to increase the
of 3D CSEM data, are shown. All maps are calculated seismically determined chance of success as the
with a 1,000 m wide window centred on the Base CSEM anomaly correlates with the location of the
Cretaceous unconformity. The horizontal resistivity prospect. The rotated fault block drilled by well
map shows the regional resistivity trend in the data. 7219/9-1 (Fig. 20.24) does not show a resistive anom-
This is also the case for the vertical resistivity map, but aly at the crest of the fault block. This well was drilled
in addition two distinct anomalies appear. One is at the on a similar play model as the Skrugard and Havis
location of the Havis discovery, and the other (Lead 1) discoveries, but was a dry well. This combination of
is southeast of Well 7219/9-1. Only a very weak subsurface responses from the discoveries and the dry
anomaly is associated with the Skrugard discovery well can be used as calibration to de-risk similar
well 7220/8-1. prospects in the area where both seismic data and
Figure 20.22c shows the apparent anisotropy. In CSEM data are available.
this map three main anomalies are clearly observed. Figures 20.22 and 20.24 show a resistivity anomaly
Havis and Lead 1 are already identified in the vertical east of well 7219/9-1 (lead 1). This anomaly is different
resistivity map. In addition the Skrugard discovery is from the Skrugard and Havis anomalies as it is not
now identified as a strong anomaly. The Skrugard and located at the crest of the fault block, but in a down-
Havis discoveries are located in rotated fault-blocks in dip position. One possible interpretation of this anom-
Early to Middle Jurassic sandstones. Figure 20.23 aly is to couple it to a stratigraphic trap within the
shows a line crossing the Skrugard discovery well. syn-rift deposits. It could also be explained by general
Overlying the seismic data is the apparent anisotropy lithology variations or possibly by the presence of
Fig. 20.23 2D seismic line (from MCG) and CSEM data (from well locations (right). The discovery sits at the crest of the
EMGS) across the Skrugard discovery (Line A in Fig. 20.23). rotated fault block where the red apparent anisotropy anomaly
Overlain is the apparent anisotropy from the CSEM data and the is seen
Fig. 20.24 2D seismic line (from MCG) and CSEM data (from data and the well locations (right). The anomaly can be
EMGS) across the dry well 7219/9-1 drilled in 1988 (Line B in associated with a different play than the Skrugard and Havis
Fig. 20.23). Overlain is the vertical resistivity from the CSEM discoveries. See text for explanation
organic-rich and resistive source rocks. More data is Eocene times the break up was initiated, and a period
needed for a more detailed interpretation of this CSEM with thermal uplift, massive volcanism, extension and
anomaly. subsequent seafloor spreading was initiated.
West of Shetland the volcanic complex consists of
flow-basalts, volcanoclastics and intrusives. During
20.7 West of Shetland Case deposition the complex advanced towards the east.
Today it pinches out east of the Rosebank hydrocarbon
20.7.1 Background discovery West of Shetland (Fig. 20.25). A typical
succession consists of volcanoclastics at the base, a
Due to the large velocity contrast between sediments mix of volcanoclastics and flow basalts in the middle,
and many basaltic rocks most of the acoustic energy is and thick flow-basalts at the top. At the eastern edge
reflected and scattered back from basaltic layers. the succession confluences with the Flett NW
Therefore, seismic imaging of rocks beneath basaltic prograding delta creating the Rosebank play concept.
strata is difficult. In the Rosebank prospect the Flett delta is dominated
Basaltic rocks have high resistivities while water- by fluvial to marginal marine deposits.
wet sedimentary rocks have low resistivities. In resis- Along the eastern edge, the thinnest part of the
tive rocks the electromagnetic energy will be less volcanic complex is to some extent known from seis-
attenuated compared to conductive rocks and the mic surveys and drilling. However, the gross distribu-
energy will penetrate deeper into the subsurface. This tion and nature of the inferred sedimentary basins
case example shows how the combination of CSEM, beneath the volcanic complexes is not known. The
MMT and seismic data can improve interpretation of reason for this is the mentioned problem the seismic
basaltic layers as well as deeper structures beneath the technique has seeing through the basaltic complex and
basalts. into the sedimentary succession below. Of particular
In Paleocene times before continental break up interest is to map the thickness of the volcanic com-
between Greenland and Norway/Great Britain, the plex, and also to identify a possible transition from the
present day continental shelfs were situated next to inferred conductive sedimentary basin below the
each other Fig. 20.25. In Late Paleocene to Early basalt to a deeper resistive basement.
Fig. 20.25 (Upper) Pre break-up (Late Paleocene to Early hydrocarbon discovery. The complex consists of volcaniclastics
Eocene) reconstruction of Greenland and Norway/Great Britain. and flow basalts. At the eastern edge the succession confluences
Present day continental shelves were situated next to each other. with the Flett delta creating the Rosebank play concept. Loca-
(Lower) The volcanic complex pinches out east of the Rosebank tion of profile is shown in the upper right figure
CSEM sounding is here used to map the electrical basaltic layers are thin along the eastern edge, seismic
properties in the upper part of the succession, and data can to some extent see inside and through the
MMT is used to map the deeper part. The electromag- complex. But this is not the case when the thickness
netic data is acquired along a profile from the increases in a westerly direction (Figs. 20.25 and
Rosebank Field in the UK sector to the Brugdan pros- 20.27)
pect in the Faroe sector (Fig. 20.26) Both the individually inverted results and the joint
CSEM/MMT inversion results are shown in
Fig. 20.27. The Inverted CSEM result images the
basaltic complex quite well, but is hampered by atten-
20.7.2 Results
uation in the inferred low resistive sedimentary rocks
beneath the basalts. The transition from the overlying
Seismic data identifies the top of the basaltic complex
conductive sediments to the resistive volcanics, as
quite well, but has problems identifying the base of
well as the gradual thinning of the volcanic unit
the basalt as well as deeper structures. Where the
Fig. 20.26 Location of the combined CSEM/MMT Rosebank—Brugdan survey profile west of Shetland
towards the east, is very well imaged. The transition to the west, and more gradual shallowing towards the
from resistive to conductive strata can be gradual. This east. East of the ridge the data show significant deep-
in combination with gradual thinning of the resistive ening into another sedimentary basin. This conductive
strata and general low resolution of the CSEM data structure correlates with the position of the West
makes it difficult to define the lateral edge precisely. Shetland Basin.
The inverted MMT data has low resolution, but is able This case study is a very good example of how
to define deep and large-scale structures. It is not able seismic-, CSEM- and MMT data can be combined to
to define the shallower volcanic complex in any detail. reveal structures in an area were the individual data
In the joint inversion result the total gross resistiv- set alone cannot resolve the problem.
ity distribution is calculated. Two deep basin-shaped
conductive structures separated by a resistive struc- 20.8 Implications for HC Play-
tural high are present along the EM section and Prospect Evaluation
(Fig. 20.28). The low resistivity values below the
basaltic complex are comparable to the values above EM methods’ general ability to map electrical
the complex. This strongly indicates that sedimentary properties in the subsurface has proven very good.
strata are present below the basalts. The strata are However, when using CSEM and MMT data for HC
shaped as a low relief half graben bounded to the exploration it should be kept in mind that they are low
east by a resistive ridge. Further, based on the shape resolution methods. Accurate depth estimation to
of the resistive bodies, the inferred basin could be resistive anomalies can also be a challenge. In addi-
bounded by larger-scale faulting (stepwise deepening) tion, for the CSEM method, depth penetration is
Fig. 20.27 2D seismic data (a), MMT inversion (b) and CSEM base of the basaltic complex, and to some extent also deeper
inversion (c). Seismic data identifies the top of the basaltic structures. Vertical black lines (in Fig. b and c) show well
complex. It has problems identifying base of the basalts and positions and black solid circles are top and base of the basalts
cannot image strata below the basaltic layers. MMT data identified in the wells. MMT and CSEM data courtesy EMGS
identifies large-scale structures. CSEM data identifies top and
Fig. 20.28 In this joint CSEM- and MMT inversion both the deeper structures and the basaltic complex are identified. See text for
discussion. Data courtesy EMGS
significantly smaller compared to the seismic reflec- HC is found both in traditional sandstone reservoirs
tion method. and in reservoirs with volcanic influence. Although the
For analysis of basins, petroleum systems, plays HC may have migrated from the Shetland Basin east
and prospects the individual HC companies have of the basement high, these are all positive indications
their own proprietary HC assessment methodology. that the basin below the basalts could have a working
The work flows are based on their experience, knowl- petroleum system.
edge, data access, and method performance statistics.
Today, seismic data and well data are the dominating
data type. Here, we discuss the potential that electro- 20.8.2 General Prospect Identification
magnetic data has for adding new and valuable infor-
mation to the HC evaluation process. Offshore EM To map out lateral variations in the electrical
data has been available for some years, but it takes properties, 3D CSEM data is needed. Collecting both
time to get enough experience to include a new data inline and broadside data enable measuring vertical
type in the evaluation process. and horizontal resistivity. When several EM
anomalies are identified, from the resistivity measure-
ment alone, it is difficult to know what is causing the
20.8.1 Basin Scale Analysis various anomalies, and additional information is
needed. In combination with seismic data, CSEM
A sedimentary basin can have one or several working data can be used for prospect identification. Seismic
petroleum systems, and a system can contain one or is the superior tool for identification of structural traps,
several petroleum plays. In a working petroleum sys- but to identify stratigraphic traps is a much more
tem, source rock, reservoir rock and overburden rock difficult task, and CSEM data can here give crucial
must be present. In addition trap formation, hydrocar- additional information. In such a work flow CSEM is
bon migration and accumulation must occur. Finally, used as a general prospect mapping tool and not as a
the timing of events must also be correct. DHI tool. The mapped anomalies can have several
In the case study from West of Shetland, seismic interpretations other than subsurface hydrocarbons.
imaging of the potential sedimentary strata is ham- These can be stratigraphic and lithological variations,
pered by the basalts covering the area. The inverted structural topography, intrusives, extrusives, salt,
EM data shows a conductive unit below the resistive carbonates or evaporites. When CSEM data is used
basaltic complex. This unit is interpreted as a sedimen- in this way for prospect and lead identification within a
tary basin, and EM data can in this way improve the play, in the end, the prospect must also be mapped on
HC evaluation at basin scale. Gross depth, shape, size 3D seismic data to become a drillable prospect.
and electrical properties of the basin can be interpreted
from the EM results.
If source rocks are present in the region, the 20.8.3 DHI Analysis
interpreted EM data can give a first indication of
potential for maturation and large-scale HC migration Since CSEM was introduced to HC exploration it has
patterns. If information about heat flow and tempera- mainly been used as a direct hydrocarbon indicator.
ture gradients exist, theoretical oil and gas windows The method can then potentially influence the prospect
can be suggested for the interpreted basin. ranking process. When it is used in this way it should
The approximate thickness and shape of the resis- always be used in combination with all available geo-
tive basaltic complex is also mapped. The base of the physical and geological data. In the cases described
basalts is deepest in the centre of the basin and above, except for the West of Shetland case, the
shallows gradually towards the basin margins. In this method has been used in this way.
case, if the basaltic layers work as vertical fluid One way to use the results from a CSEM DHI
barriers, potential hydrocarbons could migrate analysis in the risking process, is for modification of
towards the margins of the basin. The Rosebank dis- the original geological chance of success for the
covery is situated in such a position (Fig. 20.28). Here, prospect.
0.9
0.8
Posterior (modified) probability

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Prior probability
Fig. 20.29 Modification of geological risk by introducing CSEM data in prospect evaluation. By using performance tracking the
prediction strength for the CSEM data can be quantified and used in the evaluation process. See text for explanation
To do this, information about prediction strength divided into two categories: Right prediction or
for the new data is needed. In this way the value of the Wrong prediction. Now, in a case were the geological
new information can be quantified. When evaluating chance of success for a prospect is decided, it can
prediction strength, the strong and weak sides of the be modified by including CSEM data. If the Right/
CSEM technique must be taken into consideration. On Wrong ratio from the performance data base is
the strong side is the ability to measure general resis- high the impact on the geological chance of success
tivity variations, areal distribution of the resistivity will also be high. Figure 20.29 can be used to illustrate
anomaly, and in some case also the thickness and the effect adding new information from CSEM
volume of the measured objects. The ability to map data will have on a prospect’s geological chance of
these parameters is strongly dependent on acquisition success.
geometry and complexity of the subsurface. The If it is decided that the geological chance of success
method’s weak sides are for example vertical resolu- for your prospect is 20%, it will plot at 0.2 on the
tion, depth conversion and strong attenuation with horizontal axis. Your performance data base for the
depth. CSEM data have told you, that for the setting your
Another important step in evaluating the prediction prospect is in, the prediction strength is very good with
strength of CSEM data is to perform a thorough sensi- a Right/Wrong ratio as high as 0.8. The 0.8-line in
tivity study including detailed forward modelling. Fur- Fig. 20.29 is green line number three up from the 0.5
ther, statistical performance data for similar CSEM black line. Then, if the acquired CSEM data are posi-
targets can be included and used in a risk modification tive, the modified chance of success in this example
procedure. would increase from 0.2 to 0.5. If the CSEM data are
However, the lack of significant performance data negative the modified probability should be decreased
is a challenge when using a relatively new data type to approximately 0.06.
like CSEM data. In the ideal case all relevant subsur- CSEM data can of course also be used without
face scenarios that can explain the CSEM observations having a performance data base available, but as expe-
should be included, also those not involving rience with the method builds up, this is one way to
hydrocarbons. include the new knowledge in the prospect risking
However, the performance data base can be process. In many cases, for many companies, a statis-
simplified, and the measured CSEM cases can be tical performance database is not available. In such
cases it is particularly important to integrate the Gabrielsen, P. and Fanavoll, S. 2012. Application of CSEM data
CSEM information with all other geophysical and in exploration—A case study from the Nordland Ridge,
Norwegian Sea. SEG International Geophysical Conference
geological data available. and Oil and Gas Exhibition, Istanbul. Extended abstract.
Gabrielsen, P.T., Brevik, I., Mittet, R. and Løseth, L.O. 2009.
Acknowledgements We thank Rune Mittet for valuable Investigating the exploration potential for 3D CSEM using a
comments on the manuscript, and we thank Pemex, MCG and calibration survey over the Troll Field. First Break 27(6),
EMGS for permission to use the data examples. 67–75.
Gabrielsen, P.T., Abrahamson, P., Panzner, M., Fanavoll, S. and
Ellingsrud, S. 2013. Exploring frontier areas using 2D seis-
mic and 3D CSEM data, as exemplified by multi-client data
over the Skrugard and Havis discoveries in the Barents Sea.
First Break 31(1), 65–73.
Further Reading Hesthammer, J., Fanavoll, S., Stefatos, A., Danielsen, J.E. and
Boulaenko, M. 2010. CSEM performance in the light of well
Alcocer, J.A.E., Garcı́a, M.V., Soto, H.S., Baltar, D., Paramo, results. The Leading Edge 29, 34–41.
V.R., Gabrielsen, P., et al. 2013. Reducing uncertainty by Johansen, S.E., Amundsen, H.E.F., Røsten, T., Ellingsrud, S.,
integrating 3D CSEM in the Mexican deep-water explora- Eidesmo, T. and Bhuyian, A.H. 2005. Subsurface
tion workflow. First Break 31(4), 75–79. hydrocarbons detected by electromagnetic sounding. First
Bhuyian, A.H., Landrø, M. and Johansen, S.E. 2012. Three- Break 23, 31–36.
dimensional modeling and time-laps sensitivity for subsur- Johansen, S.E., Amundsen, H.E.F. and Wicklund, Tor A. 2007.
face CO2 storage. Geophysics 77(5). Interpretation of SeaBed Logging (SBL) data. Leading Edge,
Bhuyian, A.H., Thrane, B.P., Landrø, M. and Johansen, S.E. January Volume.
2010. Controlled source electromagnetic three-dimensional Johansen, S.E., Brauti, K., Fanavoll, S., Amundsen, H.E.F.,
grid-modelling based on a complex resistivity structure of Wicklund, T.A., Danielsen, J., et al. 2008. How EM survey
the seafloor: effects of acquisition parameters and geometry analysis validates current technology, processing and inter-
of multi-layered resistors. Geophysical Prospecting 58, pretation methodology. First Break 26, 83–88.
505–533. Key, K.W. 2003. Application of broadband marine
Bhuiyan, A.H., Wicklund, T.W. and Johansen, S.E. 2006. High- magnetotelluric exploration to a 3D salt structure and a
resistivity anomalies at Modgunn Arch in the Norwegian fast-spreading ridge. Doctor of Philosophy thesis. University
Sea. First Break 24, 39–44. of California, San Diego.
Buland, A., Løseth, L.O., Becht, A., Roudot, M. and Røsten, T. Kong, F.N., Westerdahl, H., Ellingsrud, S., Eidesmo, T. and
2011. The value of CSEM data in exploration. First Break Johansen, S.E. 2002. Seabed logging: A possible direct
29(4), 69–76. hydrocarbon indicator for deepsea prospects using EM
Chave, A.D. and Cox, C.S. 1982. Controlled electromagnetic energy. Oil & Gas Journal, May 13.
sources for measuring electrical conductivity beneath the Lu, X. and Xia, C. 2007. Understanding anisotropy in marine
oceans 1. Forward problem and model study. Journal of CSEM data. 77th SEG Annual Meeting, San Antonio.
Geophysical Research 87(B7), 5327–5338. Løseth, L.O. 2007. Modelling of controlled source electromag-
Constable, S. 2010. Ten years of marine CSEM for hydrocarbon netic data. Doctor of Philosophy thesis. Department of Phys-
exploration. Geophysics 75(5), 75A67–75A81. ics at the Norwegian University of Science and Technology.
Constable, S. and Chester, J. 2006. Weiss Mapping thin resistors Mittet, R. and Gabrielsen, P.T. 2013. Decomposition in upgoing
and hydrocarbons with marine EM methods: Insights from and downgoing fields and inversion of marine CSEM data.
1D modeling. Geophysics 71, G43–G51. Geophysics 78, E1–E17.
Cox, C.S., Constable, S.C., Chave, A.D. and Webb, S.C. 1986. Mittet, R. and Morten, J.P. 2012. Detection and imaging sensi-
Controlled source electromagnetic sounding of the oceanic tivity of the marine CSEM method. Geophysics 77,
lithosphere. Nature 320, 52–54. E411–E425.
Eidesmo, T., Ellingsrud, S., MacGregor, L.M., Constable, S., Mittet, R. and Morten, J.P. 2013. The marine controlled-source
Sinha, M.C., Johansen, S.E., et al. 2002. Sea Bed Logging electromagnetic method in shallow water. Geophysics 78,
(SBL), a new method for remote and direct identification of E67–E77.
hydrocarbon filled layers in deepwater areas. First Break 20, Morten, J.P. and Bjørke, A.K. 2010 Fast-track marine CSEM
144–152. processing and 3D inversion. 72nd EAGE Conference &
Ellingsrud, S., Sinha, M.C., Constable, S., MacGregor, L.M., Exhibition incorporating SPE EUROPEC 2010, Barcelona.
Eidesmo, T. and Johansen, S.E. 2002. Remote sensing of Extended abstract C004.
hydrocarbon layers by Sea Bed Logging (SBL): Results Morten, J.P., Roth, F., Karlsen, S.A., Timko, D., Pacurar, C.,
from a cruise offshore Angola. The Leading Edge 21, Olsen, P.A., et al. 2012. Field appraisal and accurate
972–982. resource estimation from 3D quantitative interpretation of
Fanavoll, S., Ellingsrud, S., Gabrielsen, P., Tharimela, R. and seismic and CSEM data. The Leading Edge 31, 447–456.
Ridyard, D. 2012. Exploration with the use of EM data in the Weiss, C. and Constable, S. 2006. Mapping thin resistors and
Barents Sea: The potential and the challenges. First Break hydrocarbons with marine EM methods, Part II—Modeling
30(4), 89–93. and analysis in 3D. Geophysics 71, G321–G332.
Chapter 21
Production Geology
Knut Bjørlykke
Production Geology has become an important field lifetime of the reservoir, providing a fourth dimension
and in mature sedimentary basins more geologists (time) to make it a 4D survey, see Chap. 19. It is then
and geophysicists are involved in production than often possible to follow the change in the oil/water
exploration. Today more of the global reserves of contact, gas/oil contact or gas/water contact by the
petroleum are added by increasing the recovery in density and velocity contrast which is a function of
existing fields than by discovering new fields. petroleum saturation. This may enable poorly drained
The development of horizontal drilling has made it parts of the reservoir to be identified and the financial
possible to produce oil much more efficiently than feasibility of drilling a new production well to drain
from vertical wells. It is now possible to drill wells this part to be assessed.
up to 8–9 km along a very complex path draining Securing optimal production from oil and gas fields
several small reservoir compartments. Horizontal dril- is an important challenge both from an economical
ling has also opened up for production from low per- and environmental perspective. The presence of free
meability reservoirs and from shales. See Chap. 23 on gas provides good pressure support for the oil produc-
unconventional oil and gas. tion. The oil must normally be produced before the
During high oil prices much more can be invested gas, otherwise production of the gas will reduce the
in methods to increase oil and gas recovery. reservoir pressure markedly and distort the oil/gas
The production of oil and gas requires detailed contact, making it difficult to recover the remaining
mapping of the reservoir with respect both to the oil. There is also gas dissolved in the oil and water
distribution of petroleum and the flow properties of which will bubble up from the oil phase and the water
the reservoir rock. phase when the pressure is reduced during production.
After the discovery of a new field it may therefore (Fig. 21.1).
be necessary to drill some delineation wells to acquire
more information about the extent of the reservoir and
the distribution of reservoir properties. Production 21.1 Capillary Forces
wells and injection wells must be planned carefully
to secure optimal recovery. Each new production well The pores in the reservoir rocks may be filled with oil
will provide a large database which needs to be and water and in some cases there may be also gas so
interpreted; this will include information about pres- that we have three phases. This complicates and
sure barriers in the reservoir. In many cases 3D seis- retards fluid flow in the reservoir during production.
mic surveys will be repeated during the producing The contact between the solid phase, mostly minerals
and the fluid phases is then very important. Minerals
have different preferences to be wetted by fluids such
as water, oil and gas. This is measured by the wetting
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway angle which depends on temperature, pressure and the
e-mail: [email protected] composition of the fluids.
546 K. Bjørlykke
Gas
GOC
Oil with dissolved gas

Vr = Oil-saturated rock volume
OWC
Water-saturated
Water with dissolved gas (methane and carbon dioxide)
Fig. 21.1 Simple petroleum trap with a gas/oil contact (GOC) and an oil/water contact (OWC)
If the wetting angle is more than 90 the oil/water

contact in the pipe will be below the general OWC,
which means the oil is drawn downwards because it is
preferentially wetting the inside of the pipe (Fig. 21.2); R
with lower wetting angles the OWC is raised.
The pores in a reservoir rock are usually filled with F1 Oil
some water in addition to the oil or gas. If we sub-
merge a very thin pipe (e.g. of glass) in oil and water,
the water will rise above the oil/water contact. This θ
H = Capillary rise
capillary rise is a function of the radius of the pipe bore
OWC
(R) and the wetting angle (θ). F2
The capillary forces (F1) must be in equilibrium
Water
with the gravitational forces (F2) which is the differ-
ence in density between oil ðρo Þ and water ðρw Þ over
the capillary rise H (Fig. 21.2). We then obtain Fig. 21.2 Capillary rise of water in a very thin pipe of
H ¼ 2γ cos θ=Rgðρw ρo Þ and we see that the capil- water-wet material. The capillary forces acting upwards are
lary rise is a function of the interfacial tension (γ) and F1 ¼ 2Rπγcosθ. Here R is the radius of the pore throat, γ the
the wetting angle (θ). This is assuming that the rocks surface tension and θ is the wetting angle. The contact between
fluid surface and the grain surface in the pore throat is thus 2Rπ
are water-wet ðθ < 90 Þ. and this must be multiplied with the surface tension (γ) and the
γ is the interfacial tension between gas and water, wetting angle (θ) between the fluid and the grain surface. The
varying within the range 30–70 dynes/cm for gas. gravitational forces acting downwards are F2 ¼ πR2gH(ρw ρo).
For oil, γ is 5–35 dynes/cm (Schowalter 1979). In a Here (ρw ρo) is the difference between the density of water and
the density of oil. At equilibrium the forces acting downwards
sediment, R represents the radius of the intergranular must be equal to the forces acting upwards so that F1¼ F2 and
pore throats (Fig. 21.2), θ is the wetting angle between we can calculate capillary rise (H) by H ¼ 2γcosθ/Rg(ρw ρo)
petroleum or water against a mineral surface
(Fig. 21.3). The wetting angle depends on the surface oil-wet than siliceous sandstones and have a higher
properties of different minerals, as well as on the wetting angle.
composition of the petroleum and water, and on the Coarse-grained sandstones with little cement, and
temperature. The capillary resistance is thus also a fractured rocks, have the lowest capillary entry
function of the lithology. In sandstones the mineral pressures (Fig. 21.4).
surfaces may be dominated by quartz, mica and clay We see that the capillary rise H is larger if the pores
minerals and the system is water-wet. Biodegraded oil are small and the wetting angle is low.
may nevertheless wet these mineral surfaces and pro- In fine-grained sediment the OWC is drawn
duce an oil-wet system. Carbonates tend to be more upwards higher than in coarse-grained sediments.
21 Production Geology 547
Mineral surface Mineral surface

θ Oil
Water θ
Oil droplet
droplet
Water-wet Oil-wet
Fig. 21.3 Wetting angle for oil in water against a mineral surface. If the wetting angle is higher than 90 the system is oil-wet. At
lower wetting angles it is water-wet
Poorly sorted
sandstone
Capillary pressure
Medium
sorted
sandst.
Well
Fractured reservoir sorted
sandst.
0 20 40 60 80 100% Water saturation
100 80 60 40 20 0% Oil saturation
10 20 30
b MPa
1 km Cap rock
Px
Px = Capillary pressure
Gas
GOC
2 km Sandst.
Oil
OWC
Water
3 km
Water (ρw = 1.00–1.05 g/cm3)
Fig. 21.4 (a) Capillary pressure as a function of sediment com- density. The pressure difference between the petroleum phase
position controlling the size of the pore throats. Fractured and the water in the pores is the buoyancy directed upwards.
reservoirs have the largest pore throats and the lowest capillary This is equal to the capillary pressure directed downwards. As a
pressures. (b) The pressure in the hydrocarbon phase increases result petroleum will be able to fill smaller pores (that have higher
upwards from the oil/water contact (OWC) due to the lower capillary entry pressures) with increasing height above the OWC
During production the capillary resistance from fine- small pores, water is drawn (imbibed) upwards due
grained sand will be greater, due to the small pore to capillary forces. The difference in OWC which is
throats. In fine-grained sandstones or siltstones with recorded on the resistivity logs may be as much as a
548 K. Bjørlykke
few metres. Since the pores in the reservoir rock usu- Capillary pressure
ally also contain some water, the percentage of oil or Pc
gas in the pores is called the oil saturation or the gas
saturation.
The petroleum saturation depends on the composi-
tion of the petroleum phase and its surface properties,
on the mineral surfaces and on the pressure difference
between the petroleum and the water phase. This pres-
sure difference increases with increasing height above Drainage
the OWC due to the lower density of the hydrocarbon
column (Fig. 21.4b). As a result, petroleum will enter
smaller pores higher up in the reservoir, and the degree
Imbibition
of petroleum saturation there will be higher if every-
thing else remains constant.
Most sandstone reservoirs are water-wet and this Swc Sw
100%
means that the minerals are lined with a thin layer of Water saturation
water. The smallest pores, i.e. between clay minerals
or along the contacts between quartz grains, will then Fig. 21.5 Drainage and imibibition of oil and water as a func-
be filled with water. The percentage of water in the tion of water saturation and capillary pressure (Pc)
pores (water saturation) may be from 80 to 90% in
well-sorted sandstones to 50–60% in sandstones with Clay
micro-
higher clay content. Also clay minerals formed during porosity Water
diagenesis (kaolinite, chlorite and illite) will have Sand

grain
rather low water saturation if the system is water-wet
Oil Water
because oil can not enter into the smaller pores due to Oil flow
capillary forces. The gas saturation will normally be Pore

throat,
capillary
higher. In reservoirs with biodegraded heavy oil, how- resistance
ever, the wetting angle is normally more oil-wet. Oil
In carbonate reservoirs the wetting angle is nor-

mally higher than in sandstone reservoirs but it may
vary as a function of the composition of both the oil
and the porewater. Fig. 21.6 Schematic figure showing how oil may be by-passed
In the Ekofisk Field it has been shown that injection as water flows around the oil. Oil must overcome a greater
of seawater has changed the wetting angle towards a capillary resistance to flow through the pore throats
more water-wet system, thus contributing to a higher
recovery. This is probably due to the effect of the high to a value Swc. This is called the connate water satu-
sulphate and Mg++ concentrations in the seawater ration. It is not possible to expel more water (reduce
which change the surface charge of the carbonate the water saturation) by oil beyond this value due to
minerals. the high capillary pressures Pc. When water flows
When the wetting phase (i.e. water) increases, the back into the rock (imbibition), the capillary pressure
contact angle (wetting angle) will increase and reduce curve follows a different path, reaching a point where
the capillary resistance. This is called imbibition and a the capillary pressure is zero. The water can therefore
core with high oil saturation may take up (imbibe) not displace all the oil and we can not get back to
water. When the wetting phase decreases, this is called 100% water saturation. This is because the oil phase is
drainage. The contact angle forms a hysteresis in no longer continuous and there are only droplets of oil
water-wet reservoirs as a function of displacing oil in water. The droplets of oil will meet strong capillary
and water (Fig. 21.5). resistance when trying to pass through pore throats and
Starting with 100% water saturation, the water is water can then flow through the reservoir without
displaced by oil so that the water saturation is reduced moving the remaining oil (Fig. 21.6).
21.2 Determination of Petroleum We see that the calculation of oil in place is based
Reserves on factors that are impossible to determine accurately.
Each value, such as porosity or oil saturation, may be
When an exploration well has detected oil an estimate represented by a probability distribution and the cal-
must be made of the reserves that can be produced. culation of oil is then based on statistical analysis. If
In many cases additional wells (appraisal wells) are the estimated average porosity in a reservoir is for
needed before the full extent of the field and its finan- example 18% we are interested in the probability of
cial viability for production can be determined. the average porosity being significantly lower or
The total initial volume of oil in the reservoir is higher (e.g. 15 or 21%).
called the volume of oil in place (Vp). The fraction of the oil in place that can be produced
The percentage of that oil which can be produced is is called the recovery and this can vary greatly, from
called recovery or the recovery factor. 20–30% to 60–70%. The number of wells and their
The volume of rock above the oil/water contact position are critical for field production, and the recov-
(OWC) and below the cap rock, Vr, is determined by ery can also be stimulated by injection of gas or a
3D seismic and data from the wells. This requires that chemical. While 20–30 years ago the maximum recov-
the (OWC) is correctly defined, but it may not be ery was considered to be 40–50%, it has since
visible on the seismic records, nor necessarily be at increased due to improved technology. Recovery in
the same depth throughout the reservoir. highly porous and permeable reservoirs may approach
Good reservoir intervals are often called pay zones, 70%. Towards the end of the production life of a field
whereas the tight zones are called non-pay zones. the wells may produce much more water than oil, but
Intervals (layers) where the porosity is low (<10%) the water is then re-injected.
and the permeability below a few milli Darcy (mD)
may not be capable of producing significant amounts
of oil even if they lie above the oil/water contact. In 21.3 Reservoir Energy
fractured rocks (e.g. fractured limestones) oil can be
produced at very much lower porosities. The ratio A reservoir may be overpressured, giving it the poten-
between the volume of reservoir rocks (N) that possess tial to flow to the surface. If the wells are not managed
sufficient reservoir quality to produce significant properly a blow-out may occur. Even with reservoirs
amounts of oil (pay), and the total volume (G), is at hydrostatic pressure in the water phase, the
called net to gross (N/G). We then need to estimate pressures in the oil and gas will be higher due to
the average porosity (φa). their buoyancy relative to water and they will flow
The volume of oil or gas in place (Vp ) is then: towards the surface. When oil is rising up a well the
pressure is falling and gas may then bubble out of
Vp ¼ Vr N=G φa Sat solution in the oil phase, as well as in the water
phase. This will reduce the density and further
Calculation of the volume above the OWC also increase the buoyancy effect so that a very high flow
depends on correct depth conversion of seismic travel rate results. The formation of free gas is an important
times. If the velocity of the overlying rocks changes part of the mechanism that causes blow-outs in wells.
across the field, this will influence the calculated
volume. The oil/water contact may not always be
horizontal across a structure and there may be sub-
21.4 Relative Permeability
compartments within the field with different OWCs.
The average porosity φa is difficult to determine. We
Permeability is a term expressing the resistance to fluid
may have a very wide range of porosities in the same
flow. It is a function of the rock properties and is
reservoir, from good reservoir quality to rather poor.
independent of the viscosity of the fluid phase. How-
Sandstones with less than 10–12% porosity usually
ever, when there are more than two fluid phases present
have such low permeability that they are not considered
in the pores, the flow and the permeability are different
to be producing parts (zones) of the reservoir.
compared with when there is only one fluid phase.
550 K. Bjørlykke
Oil Water 100

100
80
Relative permeability
80
for gas and oil

Relative permeability
Gas
for oil and water
60
60 Production
of oil and Oil
water 40
40
of only water
20
Production
Production
20 of only oil
0
0 20 40 60 80 100% Oil saturation
0 Water
0 20 40 60 80 100% saturation 100 80 60 40 20 0 Gas saturation
100 80 60 40 20 0% Oil saturation
Fig. 21.8 Relative permeability of oil and gas. We see that if
the oil saturation is less than about 35% only gas will flow. Even
Fig. 21.7 Diagram showing the relative permeability with a
with high oil saturation (80%) there will still be some production
mixture of water and oil. Similar relations exist between oil and
of gas
gas, and between gas and water
Production
Injection well
In pores containing two fluid phases, the permeabil- well Aquifer support
ity of one of the fluid phases is expressed as the from meteoric
Pressure support water
relative permeability (kr). This is the permeability from expanding gas
with two fluid phases divided by the permeability
with only one fluid phase. In the case of oil and water:
kr ¼ koþw =kw
Spill
point
The relative permeability is a function of the per-
centage of each fluid phase in the pore space and also
of the wetting properties (wetting angle) of the phases. Pressure
support due
The permeability with both oil and water present to compaction
(ko+w) is lower than the permeability with only water
(kw) (Fig. 21.7). Fig. 21.9 Different types of flow of water into the reservoir
providing pressure support. This may be from the volume of
Two phase flow will always reduce the total flow
sandstones communicating with the reservoir, from adjacent
rate (Figs. 21.7 and 21.8) and this is critical for flow shales and siltstones, from meteoric water and from water
near the well and in the pipes. injection
We see that if the oil saturation is less than about
35% only gas will flow. Even with high oil saturation Here p is the pressure, V is the volume and T is
(80%) there will still be some production of gas. In temperature (K). n is the molecular weight of the gas
some cases we may have three phases together (water, and R is the universal gas constant.
oil and gas) and this is very difficult to model. When oil and gas are produced, the pressure within
the reservoir is reduced and water then flows into the
reservoir rock (Fig. 21.9). The rate of flow from the
21.4.1 Behaviour of Reservoir Fluids
surrounding rocks controls the rate of pressure drop
during production (Fig. 21.9). If the reservoir is
The reservoir fluids obey the phase rules. For an ideal
connected to another, very large and extensive, reser-
gas we have:
voir there will be very significant pressure support
from the surrounding rocks, which are often referred
p V ¼ nRT
to as the formation. In a relatively isolated sand body
surrounded by low permeability shales the pressure
Pressuring at increased effective stress or chemical effects of the

Pressure constant Supercritical fluid water flooding.
temperature
The equation of state is:
Liquid Liquid
Critical V P=T ¼ C
Gas
Bubble line point
Heating at If the temperature is constant, the pressure is a
constant
Gas and pressure
function of the change in volume, thus
Dew line
liquid
Gas C ¼ ΔV=VΔp
Temperature
ΔV ¼ CVΔp
Fig. 21.10 Simple phase diagram showing how gas will form
in the reservoir as pressure is reduced during production. When Fluid flow during production can be calculated
gas is produced, lower temperatures and pressures at the surface
by applying the conservation of mass. The volume of
may cause liquids to form (condensate). If the pressure is
reduced in the reservoir gas, methane in particular will bubble hydrocarbons produced must be replaced either by
out of solution from both the oil and water phases. This has a fluids flowing into the reservoir from outside or by
feedback effect which helps to maintain reservoir pressure. At an expansion of the fluid phases with corresponding
the critical point gas and liquid have the same properties. This is
reduction in pressure.
a temperature and pressure where liquids can be transformed to
gas without any change in volume. For water this is 22.1 MP When we produce petroleum from a reservoir the
(221 bar) and 374 C. For hydrocarbons it depends very much on volume of produced petroleum can at least partly be
their composition replaced by water flowing into the reservoir from
surrounding rocks. The drop in volume ðΔV Þ is then

will drop fairly quickly during production. The rate of a function of the volume of produced petroleum Vpro
pressure reduction is thus a function of the volume of minus the volume of water flowing into the reservoir
sandstones connected to the oil-saturated reservoir. If which is referred to as the water drive ðVwd Þ.
the sandstone volume is large or if it is communicating
with siltstones and sandy shale with significant perme- ΔV ¼ Vprod Vwd
ability, there will be more pressure support to the
reservoir during production. Changes in the reservoir The drop in pressure due to produced oil, gas and
pressure may also cause changes in the petroleum water is a function of:
phase (gas/oil ratio, formation of condensate etc.). (1) The compressibility of the fluids present. Oil with
At the critical point gas and liquid have the same a low content of dissolved gas has relatively low
properties (Fig. 21.10). This is the temperature and the compressibility and the pressure will be reduced
pressure where liquids can be transformed to gas with- rather rapidly. If free gas is present as a separate
out any change in volume. For water this is 22.1 MP phase above the oil, it will be highly compressible
(221 bar) and 374 C. For hydrocarbons it depends and will expand as the oil is produced and help to
very much on their composition. maintain the higher pressures in the oil phase. Oil
A reservoir may be thought of as a large container; saturated with gas will form a separate gas phase
during production we can therefore calculate reservoir when the pressure is reduced and thus maintain the
pressures using material balance equations. If there is pressure.
little compaction of the reservoir during production, (2) The supply of fluids from outside the reservoir.
the volume available for the fluids (the porosity) will This may be from meteoric water connecting to
be almost constant. The main change is due to elastic the surface where there is a continual supply of
deformation due to loading or unloading. In some water, or by the injection of water or gas.
cases as in the Ekofisk Field, where the reservoir (3) Supply of water from around the reservoir. Pres-
rock is mechanically weak, there is significant com- sure reduction in the reservoir during production
paction of the reservoir during production due to creates a potential for water to flow from the
552 K. Bjørlykke
surrounding rocks. Reservoir sandstones that are the oil will bubble out as a separate phase and add to
surrounded by low permeability shales will have a the gas drive.
very low supply of water (water drive) and the
pressure will therefore drop faster during produc- 21.7 Compaction Drive
tion. This is particularly true when reservoir
sandstones are offset laterally by faults. In most If the fluid pressure is reduced, the increase in effec-
cases the permeability of shales is so low as to be tive stress may cause mechanical compaction, reduc-
almost irrelevant in the time span during which ing the pore volume. While this will also contribute to
petroleum is produced from a field. However, if the maintenance of high pressure, it may damage the
the reservoir is part of a thick sandstone sequence, reservoir by grain crushing and closing fractures (if
the pressure support provided from a larger sand present). In the Ekofisk Field, however, the compac-
volume may be very significant. If we only con- tion of the reservoir provides a drive that enhances
sider the water phase, the pressure drop in a sand- recovery of the oil.
stone reservoir is a function of the volume of
sandstone in communication with the reservoir.
21.8 Water Injection
Water is injected into the reservoir to replace the

21.5 Meteoric Water Drive produced petroleum and thus maintain high pressure
in the reservoir. It will also help to maintain the pres-
Reservoirs connected to an aquifer that is part of the
sure gradients towards the production wells. Injected
meteoric water flow system will experience very little
water will follow the most permeable layers at the base
drop in pressure during production because the pro-
of sandstones because it is denser than oil. Injection
duced petroleum is quickly replaced by meteoric water
of water to displace oil or gas is called “immiscible
from the aquifer.
replacement” because oil is not soluble in water.
While the supply of meteoric water may be rela-
Water will displace oil found in pockets (depressions)
tively fast, compaction-driven water flow is many
against low permeability shales (Fig. 21.13b).
orders of magnitude lower and can be ignored.
The vertical distribution of permeability and capil-
In a closed system the response to production and
lary resistance within a bed is important for the flow of
the associated reduction in reservoir pressure is a
oil and water and the total recovery (Figs. 21.11 and
function of the compressibility of the water (Cw), the
21.12). We see that in coarsening-upwards sequences
compressibility of the oil (Co) and the compressibility
the highest permeability and lowest capillary resis-
of the gas (Cg). There will also be changes in the
tance are at the top, where oil or gas flow will
reservoir volume (porosity) because reduced fluid
pressure will result in higher effective stresses, but Water injection Oil production
this response will normally be rather small unless the
Permeability
reduction is sufficiently large to cause mechanical
compaction with grain crushing. Shale
When the pressure is reduced there will be an
expansion of both the water and the petroleum in the Water-sat.
Sandstone
reservoir, but this is relatively small. Oil-sat.
21.6 Gas Expansion Drive

Capillary
Gas has a high compressibility and if free gas is Sandstone coarsening upwards resistance
present in the reservoir this gas phase will expand as i.e., shallow marine sandstones
the pressure is reduced. The pressure drop relative to Fig. 21.11 Production from a coarsening-upwards unit, e.g. a
the volume of oil produced will therefore be relatively marine shoreface sequence. The injected water will sink in to the
small. In addition, gas that was originally in solution in less permeable layers and displace oil also from there
Water injection Oil and water a Oil – microtrap

production Displacement of oil by gas
Permeability Gas
Shale
Oil-saturated
Sandstone Gas
Water-sat. Oil – microtrap
Water prod.
b
Gas
Displacement
Sandstone fining upwards Capillary Oil of oil by water
resistance Water
i.e., fluvial channel or turbidites
Fig. 21.12 Production from a fining-upwards sequence, e.g.

fluvial channel facies. The injected water will sink to the lower- Fig. 21.13 (a) Displacement of oil by gas in small traps during
most permeable layers and then may break through to the production. (b) Displacement of oil by water in a reservoir with
production well, leaving most of the oil in the less permeable mostly gas
beds behind. Addition of polymers to increase the viscosity of
the water may result in a better sweep and produce more oil
Tracers may be employed to see if water from the
injection well has reached the production well.
preferentially occur due to the buoyancy relative to
water. In fining-upwards sequences the injected water
will tend to follow the basal permeable part, causing
21.10 Other Methods for Enhanced
early water breakthrough to the production well.
Recovery
When there is breakthrough of water from the injec-

21.9 Gas Injection tion well to the production well, the flow of water may
be reduced by adding polymers that increase the vis-
When gas is injected in to the reservoir it is partly cosity of the water. The pressure gradient will then
soluble (miscible) in both the oil phase and the water increase in the oil phase and it may be possible to
phase. The most common gases used are methane, displace more oil towards the production well.
nitrogen and more recently also carbon dioxide. Gas When the oil saturation is low, particularly towards
injection contributes to the maintenance of the pres- the end of production, movement of the remaining oil
sure in the reservoir. Gas is less dense than oil and is drops is prevented by the capillary forces. By adding
efficient in terms of displacing oil from small micro- surfactants (soap) the oil drops may be broken down
traps where pockets of oil are found above water. Gas into an emulsion which can flow with the water to the
will therefore sweep the upper parts of sandstone beds production well.
in a reservoir and displace the oil downwards, so that it Many enhanced recovery methods are more effec-
can flow out of the small traps (Fig. 21.13a). tive in onshore oil fields where the distance between
The injected gas may be methane that is produced the wells is much less than offshore. During high oil
along with the oil. These gases are also to a large prices it may be economical to spend more on
extent dissolved in the oil phase during the injection. chemicals to stimulate production.
The use of CO2 is now encouraged for injection There are three main enhanced recovery methods:
in reservoirs because this represents a storage of I. Thermal processes: These are processes which
carbon dioxide, reducing the contributions to global increase the temperature of the oil and thus lower
warming. its viscosity. The most common are:
Carbon dioxide has been applied, for example, in (a) Steam injection. This is the main recovery
the Sleipner Field in the North Sea, but not in the method for heavy oil and tar sand.
reservoir. (b) In situ combustion. Supplying oxygen to par-
tially burn heavy oil or coal to increase the
554 K. Bjørlykke
temperature and reduce the viscosity of oil, and tar sand. For it to be possible to produce oil, the
to produce gas from coal. pressure difference between the oil in the reservoir
II. Chemical processes: These involve injection of rock and the well must exceed the resistance to flow
chemicals into the reservoir. The most important exerted by capillary forces (in water-wet reservoirs)
types are: and viscosity forces (friction).
(a) Injection of caustic (alkaline) chemicals which Biodegraded oil is normally found at relatively
reduce surface tension (surfactants). shallow depths, and the temperature of the oil is low
(b) Polymers which are injected into water to and the viscosity consequently high. By injecting
increase its viscosity to prevent water steam the temperature can be raised and the viscosity
breakthrough. lowered considerably. There are examples of the vis-
III. Injection of various types of gas to increase the cosity of oil being reduced from 300 to 10 cP after
miscibility of hydrocarbon phases. These may steam injection, and biodegraded oil is often too vis-
include carbon dioxide, neutral gases (e.g. nitro- cous to flow at all without steam injection. Steam
gen) or hydrocarbon gases. injection is particularly important in California,
I. Thermal Processes: These are only used with where 300,000 barrels/day are produced using this
reservoirs with highly viscous oil, where an increase method. Steam injection must be repeated at regular
in temperature will greatly reduce the viscosity. This intervals because the viscosity increases again as the
will normally be heavy, asphaltic oil which has been reservoir cools off. Sandstones which are to be
subjected to bacterial breakdown and evaporation. Oil subjected to steam injection must also have high oil
reservoirs of this type are normally fairly shallow (less saturation, because otherwise water and steam can
than 1,500 m). Deeper reservoirs will have a high flow past the oil. Steam is recovered from the pro-
temperature anyway, so there is less to be gained by duced oil, and in some fields one out of every three
artificially heating the oil. barrels produced is used for steam production.
In situ combustion can be induced by pumping air II. Chemical Processes: When oil flows out of a
down into the reservoir, and the combustion can be reservoir rock into a producing well, it must overcome
regulated by addition of oxygen. The hydrocarbons the capillary forces involved in two-phase flow. By
which do not burn will then become hotter and less reducing the surface tension between oil and the water
viscous. Burning also produces gases, which may which is injected into the wells, the capillary forces can
increase production. Special wells can be used for be reduced. The addition of surfactants, a sort of soap,
injecting air, and combustion spreads from these to will reduce the surface tension so that the wettability
the production wells, forcing a zone of hot water, gas changes. The wettability can be measured by means of
and light oil ahead of it to the production wells. Com- the contact angle, i.e. the angle which the liquid phases
bustion may also increase the porosity of the reservoir make with a mineral surface, for example (Fig. 21.3).
rock, making it possible to produce later on from the If an oil reservoir is oil-wet, it means that oil is
wells in which combustion has taken place. more strongly bound to the mineral surface than water.
Steam injection is responsible for 90% of all pro- By reducing the surface tension in the water phase, the
duction through secondary recovery methods apart wetting angle is altered and the system may become
from water injection. One of the major reasons for water-wet, and the oil will become more mobile so
the low recovery percentage from oil reservoirs is that the relative permeability to oil is increased. A
that the oil is too viscous. This is particularly true of mixture of different chemicals (slug) is injected into
oil which has been subjected to bacterial degradation oil reservoirs to reduce the surface tension. It includes
(biodegradation), leaving the viscous asphaltic part of micro-emulsions, soluble hydrocarbons, micelles and
the oil behind as heavy oil (low gravity 16–17 API). frequently also alcohol and various salts.
The Cretaceous Athabasca tar sand in Alberta, One serious drawback of this method is that
Canada, is the most famous example, and contains chemicals can easily be adsorbed onto the reservoir
huge oil reserves. When oil reservoirs with such rock. Clay minerals in particular have high ion
heavy biodegraded oil are exposed at the surface, the exchange and adsorption capacities. The specific sur-
oil is almost solid at normal temperatures, and is called face (m2/g) of rocks is also of significance.
Consequently chemicals are often pumped down in expelling mixture, and gas and water are injected
advance (preflush) to reduce adsorption of the subsequently.
chemicals which are intended to reduce surface ten- CO2 has a critical temperature of about 31 C, and is
sion. This treatment ought also to take into account therefore gaseous at all reservoir pressures. Carbon
which clay minerals occur in the rock. dioxide is very soluble in oil, which increases in vol-
Polymers. In oil reservoirs with an undesirably high ume when it is saturated with respect to CO2. This
relative water permeability, large amounts of oil will creates a sort of gas drive.
remain in the reservoir rock. Water will flow where Increasing the CO2 content also reduces the viscos-
there is least resistance to its movement, e.g. along ity and density of oil, making it very mobile. The
cracks and permeable sand beds, and injection of method is used for oil reservoirs with oil saturation
water will then have little effect. This is particularly of 25–55%. Neutral gases such as nitrogen may also be
relevant with oil reservoirs with low oil saturation injected into reservoirs, but far greater pressure is then
(high water saturation). If the viscosity is increased required to make the gas soluble in oil.
so that the pressure gradient in the water phase
increases, the gradient in the oil phase will also
increase. Polymers reduce the mobility of the water, 21.11 Changes During Production
causing a piston-like replacement mechanism. With
increased viscosity, the polymer-bearing water will With production from sandstone reservoirs, the physi-
exert greater shear forces on the oil phase, so that oil cal properties of sandstone must be taken into account,
drops are more easily carried along. Water containing e.g. distribution of porosity and permeability. The
polymers is no longer a Newtonian liquid, but a actual production, however, will lead to changes in
pseudoplastic liquid in which the apparent viscosity the reservoir rock which may have an adverse effect
is reduced by the rate of deformation. on the reservoir properties, i.e. cause reservoir dam-
Polymers will also have a tendency to be adsorbed age. This may happen in two ways:
onto the surface of minerals so that their effect decreases 1. Chemical reactions between minerals and liquids
with increasing distance from the injection well. which are used in drilling or injection of water.
Addition of Alkaline (Caustic) Chemicals. Strong 2. Mechanical damage through relatively loose clay
alkaline solutions, for example NaOH, will also minerals or other small grains being carried by the
reduce the surface tension of aqueous phases. Injection fluid flowing towards the well, and obstructing the
of a mixture of these alkaline solutions may therefore pore throats.
increase both the water-wetness of the reservoir As oil flows into a well, the flux (cm3/cm2) is
and oil production. The most common chemicals inversely proportional to the distance to the well. If
include sodium and potassium hydroxide, sodium we can improve the permeability of the reservoir
orthosilicate, sodium carbonate and sodium phos- nearest the well, we may be able to step up the pro-
phate. The method is used largely in sandstone duction rate considerably. For this reason chemicals
reservoirs. Small amounts of gypsum or anhydrite are sometimes injected into the reservoir to improve
will cause Ca(OH)2 to precipitation and neutralise the permeability. We then need to understand some of
caustic soda (NaOH). the chemical properties of minerals in order to be able
The object of gas injection is to produce a fluid to predict the reactions which will result from treat-
phase which will dissolve the reservoir oil. The flow of ment with acids and other chemicals.
such a fluid through reservoirs with low oil saturation As far as chemical reactions are concerned, clay
may dissolve the oil and carry it along to the produc- minerals are important because they have a large spe-
tion well. Hydrocarbon gases which are in the liquid cific surface and ion exchange capacity. The specific
phase at reservoir pressure (Liquid Petroleum Gas – surface varies with the size and shape of the mineral
LPG) will be fully miscible with oil. This requires that grains. Kaolinite will typically have a specific surface
the reservoir temperature be below the critical temper- of 5–30 m2/g, chlorite 10–50 m2/g, illite and smectite
ature for the gas. At higher temperatures the gas is in (montmorillonite) >100 m2/g. The high specific sur-
gaseous form regardless of pressure, and is not fully face of illite and smectite is due to the fact that they
miscible with oil. Propane is often used in the oil- often occur as very thin sheets or fibres.
556 K. Bjørlykke
Kaolinite has very low solubility in water, and strong for example of quartz and felspar, which may block the
acid (HF) is required to dissolve it. Kaolinite is therefore pore throats. This effect can be reduced by lowering the
chemically stable and its ion exchange capacity is lower production rate. In the laboratory this type of mechani-
than that of other clay minerals. Chlorite, on the other cal formation damage can be tested in an experimental
hand, is soluble in acids such as HCL. If acid is used to flow rig. If the measured permeability increases tempo-
dissolve minerals like chlorides in order to increase the rarily when the flow direction is reversed, formation
permeability around the well, it is important to avoid iron damage has been caused by “a moving clay and silt
being precipitated as iron hydroxide (Fe(OH)3). This fraction”. Sandstones with a high percentage of second-
could reduce the porosity very severely because it ary porosity will have relatively large pores with small
forms pore-filling cement. To avoid this, one can add pore throats. The ratio between the pore diameter and
chemicals which form complexes with iron (e.g. citrate). the pore throat is often referred to as the aspect ratio.
Illite and smectite are not very soluble in HCL, and Pores with high aspect ratios will be particularly vul-
hydrofluoric acid (HF) must be added to dissolve these nerable to mechanical formation damage. Pore geome-
minerals. They are sensitive to variations in salinity try is also important with respect to two-phase flow (oil
and in particular to the K+ content. K+ and other alkali and water) because of the capillary forces that have to
ions will be held in the interlayer position in smectite be overcome in the pore throats. This is even more the
so that water is expelled and the mineral contracts in case with condensate reservoirs, where we may have
volume. On the other hand, injection of fresh or three phases (oil, gas, water).
slightly saline water into a reservoir may lead to The planning of production from a reservoir
uptake of water and expansion of smectite and mixed involves a detailed production strategy and optimal
layer minerals, which may greatly reduce the porosity. positioning of production and injection wells. The
Sodium smectites can swell to 5–10 times their origi- trend production will take with time is simulated on
nal volume. Swelling can be reduced by injecting extremely powerful computers which are fed with a
dilute acid (HCl) or water with a high salinity (e.g. very large number of parameters relating to the reser-
KCl). Water is also often injected to maintain reservoir voir. Simulations depend very heavily on the geometry
pressure. Offshore drilling platforms in particular of the reservoir and internal communication, e.g.
depend on using seawater. However, seawater may between sandstone bodies, the exact position of the
have adverse effects on the reservoir. The high sealing fault, etc. The internal properties of the reser-
sulphate concentration in seawater may cause precipi- voir can be measured on cores from the wells, but a
tation of baryte (BaSO4) which is the least soluble of three-dimensional representation of the distribution of
the common sulphate minerals. Seawater also contains porosity and permeability and the pore geometry must
sulphate-reducing bacteria which will start to reduce be constructed, using models for the detailed deposi-
SO24 to H2S in the reservoir. This is most undesirable tional environment and diagenetic alteration.
since H2S is a very poisonous gas which also has a It may now be possible to follow fluid flow within
highly corrosive effect on steel during production. the reservoir compartments during production using
To avoid unforeseen chemical damage to the reser- 4D time-lapse seismic data (Lynn et al. 2014).
voir rock, it is necessary to carry out very thorough
mineralogical analyses so as to be able to predict the
effect of various types of chemical stimulation. 21.12 Carbonate Reservoirs and
Physical damage to the reservoir, as previously men- Fractured Reservoirs
tioned, is due to mineral grains being loosened and
carried by oil and water, eventually clogging the flow Many of the same principles apply to carbonate
paths. If the reservoir is water-wet, fine-grained reservoirs as apply to sandstones. There is a tendency
minerals will tend to remain suspended in the aqueous for carbonate reservoirs to be less water-wet than sand-
phase, and will not come into the oil phase very easily stone reservoirs, but this depends on the composition of
because of surface tension. When considerable the oil and on the temperature and pressure. Pores often
quantities of water are produced together with oil, have a high aspect ratio, particularly in limestones
therefore, the aqueous phase is particularly likely to where we often have intragranular porosity inside
carry clay minerals and other small mineral particles, fossils or vuggy mouldic porosity after selective disso-
lution of organic fossils. The reservoir properties of
bioclastic and reef limestones are largely a result of the 4D seismic may be successful in some fields for
palaeo environment, which determined the distribution detecting changes in the gas/oil contact (GOC) and in
of fauna. some instances also the oil/water contact (OWC).
Fractures are generally much more important in Repeated seismic surveys at 1–3 years intervals may
carbonate than in sandstone reservoirs. Stylolites are detect an upward movement of the OWC as the reser-
sometimes found in sandstone reservoirs that have voir is depleted. The difference in impedance may,
been buried to at least 3.4–4 km. In limestones however, not be as pronounced at the primary OWC
stylolites develop at much shallower depths, often because there may be 30–40% oil saturation in the
forming a thin clay surface that is quite impermeable, drained intervals, It may then be possible to detect
and which may divide a reservoir up into compart- parts of the reservoir which are not drained and
ments or even serve as a cap rock. which are large enough to justify an additional well.
In fractured reservoirs there is a high percentage of
porosity due to fractures rather than more uniformly
distributed porosity. Fractures are most extensively 21.14 Well-to-Well Tracers
developed in brittle rocks like well-cemented
limestones and chert. They may also form in metamor- Tracers can be added to the water in the injection well
phic and igneous rocks. Fractures are often invaded by and their arrival in the production well can provide
drilling mud, causing lost circulation, and this may important information about the reservoir. If the tracer
damage the reservoir near the well. The flow of petro- is recorded in the production well relatively soon after
leum through a fractured reservoir during production the injection there must be a high permeability con-
is very complex, and it is difficult to map out the nection in the form of an open fault or fracture or a
fracture system of the subsurface in sufficient detail. high permeability sand layer. The time of the arrival of
As a result of reduced pressure during production, the tracers makes it possible to calculate the flow
fractures may start to close due to increased net stress. velocity of the water phase. The tracers must remain
By injecting water at very high pressures (higher than dissolved in the water and should not be adsorbed on
the geostatic pressure) fractures may be widened the minerals. Since they become part of the produced
(hydrofracturing), a method also sometimes used in water they should be environmentally acceptable.
other reservoirs to create or widen fractures close to Many of the tracers are radioactive but only with
wells and improve the permeability of the critical area β radiation, such as tritiated water (HTO), 36 Cl and
around the well. The introduction of some coarse- 22
Naþ . Most of them have a relatively short half life,
grained material like sand and also other types of
except 36 Cl .
granular materials may help to wedge the fractures
It is also possible to use non-radiocative tracers, but
and prevent them from closing.
extremely low levels must be detectable (in the ppb
Onshore oil fields have much closer well spacing
range) because of the strong dilution. Gas tracers may
than offshore fields because of the lower drilling costs.
also be used.
This means that many of the enhanced recovery
The concentration of the tracer as a function of time
methods can be used much more effectively. At high
during production may provide very important infor-
oil prices mature oil fields can have a very long tail
mation about the rock properties and the flow in the
production and be economical even with very high
reservoirs. Different types of tracers may to different
water production (>90% water).
degrees interact and be adsorbed onto the reservoir
rocks.
21.13 Monitoring Production
During production, comparison of pressure changes 21.15 Reservoir Models and Field
in the production and injection wells will provide Analogues
information about fluid communication within the res-
ervoir. These data are fed into vary large and complex When planning the production of a reservoir, a rela-
reservoir simulation models. tively detailed reservoir model is required.
558 K. Bjørlykke
Seismic data has limited resolution and in offshore Reservoir modelling has become a special field where
reservoirs the spacing between wells may be relatively the flow during production is simulated mathemati-
large, usually several 100 m or a few kilometres. It is cally based on assumptions about the distribution of
difficult to construct a 3D model when the producing rock types with different porosity and permeability.
units are offset by faults that are below seismic resolu- See Chap. 22.
tion so that they can not be detected. Outcrops of similar Drilling and well completion.
reservoir rocks may be used to provide quantitative data A team of specialists is therefore required to pro-
on the geometry of sand bodies and the distribution of duce a reservoir efficiently.
minor faults. to be used in the reservoir model. This depends very much on reservoir modelling
It is however critical that the field analogue was which is a complex process were the fluid flow in the
deposited in a similar sedimentary environment and reservoir is simulated.
plate tectonic setting. Many factors such as sediment
supply, climate, water depth, tidal range, etc. have to
be similar. Further Reading
All reservoir analogues that are exposed have been
buried to a certain depth and then uplifted to the
Archer, J.S. and Wall, C.G. 1994. Principles of Petroleum Engi-
surface. It is difficult to reconstruct the burial history neering. Principles and Practice. Graham and Trotman,
in terms of temperature and effective stress so that the London, 362 pp.
diagenetic effects can be compared with reservoirs at Burchette, T.P. 2012. Carbonate rocks and petroleum reservoirs:
a geological perspective from the industry. In: Garland, J.,
their maximum burial depth.
Neilson, J.E., Laubach, S.E. and Whidden, K.J. (eds.),
The structures observed in the field analogue must Advances in Carbonate Exploration and Reservoir Analysis.
be analysed so that the deformation which occurred Geological Society Special Publication 370, pp. 17-37.
during uplift can be distinguished from that produced Dandekar, A.Y. 2006. Petroleum Reservoir Rocks and Fluid
Flow Properties. CRC Publications, Boca Raton, FL, 460 pp.
during subsidence.
Dickey, P.A. 1979. Petroleum Development Geology. Petro-
Offshore production from a separate platform is leum Publishing Co., Tulsa, OK, 398 pp.
expensive and requires a relatively high production to Dikkers, A.J. 1985. Geology in Petroleum Production.
offset the production costs. Still many oil fields can be Developments in Petroleum Science 20, Elsevier Science,
239 pp.
produced with a tail production stimulated by different
Hyne, N.J. 2012. Nontechnical Guide to Petroleum Geology,
methods. Onshore oil and gas fields on the other hand Drilling and Production, 3rd. edition. Pennwell Corporation,
can be economical at rather low production rates. Tulsa, Oklahoma, pp. 698.
Loudon, R.C. 1996. Principles of Petroleum Development Geol-
ogy. PTR Prentice Hall. New Jersey. 267pp.
21.16 Conclusion Lynn, G.J., Ellis, A.C., Brain, J., Parker, R., Lortzer, G-J. and
Michelet, S. 2014. Gannet F field - Unexpeted fluid flow
between reservoirs identfied from time-lapse seismic data.
Production geology encompasses many different First Break 32(4), 53-59.
specialities in geology, geophysics and engineering Robinson, A., Griffiths, P., Price, S., Hegre, J. and Muggeridge,
subjects: A. 2009. The Future of Geological Modelling in Hydrocar-
bon Developments. Geological Society Special Publication
Reservoir sedimentology/diagenesis. 309, 232 pp.
Reservoir geophysics. Tiab, D. and Donaldson, E.C. 2004. Petrophysics. Elsevier,
Reservoir modelling (modelling of fluid flow). Amsterdam, 889 pp.
Chapter 22
Introduction to Reservoir Modelling
Jan C. Rivenæs, Petter Sørhaug and Ragnar Knarud
22.1 Introduction 22.1.1 The Static Reservoir Model
This chapter gives the reader an overview of the basic The static reservoir model is usually referred to as the
elements in reservoir modelling. geological model (often abbreviated “geomodel”), and
Reservoir modelling is the process of transferring is a digital numerical model describing the initial state
the available subsurface data and knowledge into a of the reservoir before any production of hydrocarbons
digital (computerised) numerical representation of has taken place, i.e. representing the spatial distribution
the subsurface, and is now considered an essential of rock types with different reservoir properties. Hence,
part of understanding and developing oil and gas a major task for the static model is calculating the initial
resources. Reservoir models are used for calculation in-place volumes of hydrocarbons, which is a quantifi-
of initial volumes in place, optimisation of drainage cation of what is “down there”, without consideration of
strategy for a hydrocarbon field, calculation of recovery method or economical viability. These volu-
recoverable volumes, generation of production metric calculations can be performed for an entire res-
profiles, evaluation/generation of depletion plans, ervoir or individually for any layers or fault segments
and as a planning tool for detailed placement of that have been defined during model construction.
production and injection wells. Due to the multidis-
ciplinary input and the increased options for
advanced visualisation, digital 3D geological models
are important as a collaborative working tool for the 22.1.2 The Dynamic Reservoir Model
integration of different subsurface disciplines
(Figs. 22.1 and 22.2). The dynamic reservoir model is typically based on the
Digital reservoir models may serve different framework and spatial distribution of reservoir
purposes, and we typically distinguish between static properties from the static reservoir model. Equations
reservoir models and dynamic reservoir models. for fluid flow in porous media are applied in order to
simulate the distribution/movement of reservoir fluids
as a function of production and/or injection. The
dynamic reservoir model also takes into account
changes related to varying reservoir pressure and
hydrocarbon saturation as a function of production/
injection over time. Fluid properties, such as density,
viscosity, phase equilibrium and fluid mobility, interact
with changes in pressures and fluid saturations.
To summarise, the static model will estimate the
J.C. Rivenæs (*) P. Sørhaug R. Knarud
Statoil Norway natural in-place volumes in the subsurface, while the
e-mail: [email protected]; [email protected]; [email protected] dynamic model will predict what is recoverable. What
560 J.C. Rivenæs et al.
Fig. 22.1 3D modelling tools are excellent for co-visualisation model grid with properties, and (D) dynamic grid with properties.
of multiple data types. Here is an example from a Norwegian The visualization is improved by contemporary use of visual
offshore field, displaying (A) seismic data, (B) wells, (C) static techniques like various fence slices and partial transparency
Fig. 22.2 The Gullfaks field (offshore Norway) is heavily faulted. The reservoir model shown here displays all faults included in
the model work
is recoverable will be dependent on various factors, 22.1.3 Main Model Elements

including the production costs, and there are many
large accumulations of hydrocarbons that are simply Both the static model and the dynamic model may be
too expensive to recover given current expectation of split into two major elements:
future oil and gas prices. • The reservoir container (bulk volume) limiting the
spatial distribution of the reservoir. The reservoir
22 Introduction to Reservoir Modelling 561
interval is defined by stratigraphic surfaces/layering model? What is the purpose of the model? What kind
and faults, represented by a three-dimensional grid. of questions is the model going to provide answers to?
• The property model, defining how petrophysical Business objectives may vary substantially depending
reservoir properties and other relevant properties on where an individual field is in its life cycle. Here
are distributed in three dimensions within the reser- are some examples of modelling objectives and the
voir container/grid. It usually also includes a consequences to the scope of the work:
numerical description of the spatial distribution of • Calculating oil or gas in-place volumes. This objec-
the reservoir fluids, e.g. oil and/or gas saturation tive is often the most important in the exploration
(through water saturation modelling). and early field development phase. Generally, it
To provide high quality input to a model, it is implies the use of more simplistic models with
important to stress that a reservoir model represents a less focus on the detailed architecture of the reser-
synthesis of several sub-surface disciplines’ descrip- voir. Usually, several alternative models are made
tion and understanding of the reservoir. The goal is to to capture volume uncertainty (see Sect. 22.5).
establish a common model that can be used for deter- • Calculating recoverable volumes and establishing
mining and optimising the commercial value of a production profiles. This objective becomes a pri-
hydrocarbon accumulation. Close communication, ority in the field development and production stage.
co-operation, and planning between the involved It generally requires a more detailed and sophisti-
disciplines are key points for constructing a successful cated description of reservoir architecture and dis-
model that is useful for its intended purposes. As there tribution of reservoir properties to be able to
are many ways to create a digital model, having a clear simulate how fluids are flowing through the reser-
and well defined business objective for the modelling voir during production. A common way to calibrate
effort is another important success factor. Disciplines the models with production data is so-called history
that typically are involved with modelling are matching (Sect. 22.4.3).
petrophysics, various disciplines in geology (such as • Optimising the depletion plan for a field. This objec-
sedimentology, petrography, diagenesis, stratigraphy, tive is tightly linked to the previously described
biostratigraphy, geochemistry and structural geology), business objective and therefore has many of the
geophysics, production technology, reservoir engi- same requirements. The depletion plan describes
neering, statistics and computer sciences. Due to how to optimally drain the reservoir. This includes
the major implementation of reservoir models in not only optimising the recoverable reserves, but
the oil industry in the last two decades, being a also the economy of the field in question. Some
“geomodeller” or “reservoir modeller” has become a typical aims of depletion planning could include:
discipline by itself. – Evaluating depletion vs. gas and/or water
injection.
– Optimising number of wells.
22.2 Model Planning – Evaluating well geometries, e.g. vertical, slanted
or horizontal wells. Would it be economical or
Before starting any modelling work there are some funda- beneficial to use branched wells?
mental questions that need answers, as model work may be • Well planning. This objective may demand an even
time consuming and requires significant manpower. One higher degree of detail in the reservoir model.
should also collect all possible data and knowledge of the
reservoir to be modelled, including experience from previ-
ous models. Hence, proper model planning is important 22.2.2 Input Data and Databases
before starting the actual model work.
There are numerous data sources that can be utilised
for reservoir modelling (Fig. 22.3). Examples of
22.2.1 Business Objectives important data are given in the following sub-sections.
The process of constructing a new reservoir model 22.2.2.1 Geophysical Data

should always start by defining the business The dominant geophysical input to reservoir
objectives, e.g. why are we making a reservoir modelling is seismic data. Other geophysical data
Fig. 22.3 Reservoir models integrate many data sources. These (f), concept models (g), production data (h), and earlier reservoir
include seismic data (a), core data (b), outcrop analogues (c), models (i). These data originate from various subsurface
thin sections (d), biostratigraphy (e), well logs and correlations disciplines at different scales, abundance and quality
such as magnetic, electromagnetic or gravity analysis derived data are the most important for reservoir
are currently of less importance. Data from seismic modelling:
surveys includes: • Porosity logs derived from petrophysical well log
• Interpreted horizons and faults, usually depth measurements (usually the density log) and core
converted (Fig. 22.3a). data (Fig. 22.3f).
• Seismic inversion data that provide important • Microscopic images (thin sections and SEM) and
information on rock properties, e.g. porosity and laboratory analysis (XRD) for mineralogy and pore
fluid distribution. geometries taken from core samples (Fig. 22.3d).
• 4D seismic data provide additional crucial informa- • Horizontal permeability logs derived from
tion on reservoir behaviour during production, for petrophysical well log measurements, core data,
instance connectivity and segmentation. and well testing.
• Input on vertical permeability (often expressed as
the Kv/Kh ratio between vertical and horizontal
22.2.2.2 Petrophysical Data
permeability), from log analysis, core analysis,
Well log data are a fundamental input to reservoir
and lamina bed modelling.
modelling. It is important to acknowledge that well
• Petrophysical lithology curves, often referred to as
logs do not give petrophysical data such as porosity,
“flag” curves, such as calcite-cement flags, coal
saturations and permeabilities directly; rather all these
flags and reservoir flags. The flag logs are important
data types are the result of some kind of processing of
in petrophysical evaluation to improve the determi-
basic well logging tools. For example, porosity logs
nation of reservoir parameters. For instance, the use
will be derived from density, sonic or neutron logs,
of resistivity logs to compute hydrocarbon
while saturations are computed from resistivity logs or
saturations is only valid in intervals flagged as
NMR (see Chap. 16). The following logs and core-
“net reservoir”. The reservoir flag is an indication • Well test data and production history in general.
of which intervals (based on petrophysical criteria) These data are usually applied during calibration of
may be regarded as reservoir, or non-reservoir, reservoir model, known as history matching
intervals. (Sect. 22.4.3).
• Input to fluid and saturation models – such as satu-
ration logs, saturation height functions, fluid 22.2.2.5 Databases and Data Management
contacts, and fluid properties. All data input to reservoir modelling need to be easily
All logs are associated with a variable degree of accessible. A database is an organised collection of
uncertainty. Permeability logs are regarded as more data, and includes digital data as well as paper reports.
uncertain than porosity logs, and horizontal wells have The purpose of a database is to provide efficient
usually less accurate measurements than vertical wells retrieval for the task to be performed – in this case
(Chap. 16). reservoir modelling.
A major challenge is that data need to be thoroughly
quality checked prior to modelling as even apparently
22.2.2.3 Geological Data small errors in parameters such as well data may be
Geological data and analysis embraces a wide variety magnified when propagated into the models, which in
of input to reservoir models, both quantitative and turn can lead to erroneous models and hence wrong
qualitative. It is vital to understand the genesis and decisions with large financial consequences.
depositional signature of the subsurface interval of The subsurface database may be vast. Seismic data
interest, in order to delineate the overall architecture alone requires enormous amounts of data storage. A
of the reservoir. For example, the geological data input common problem is that different software products
could be: have their own databases, and transferring data
• Reservoir zonation and sedimentological descrip- between different databases and applications can be
tion, including identification of flow units demanding. As software tools are upgraded, data
(modelling facies, see Sect. 22.3.3.4) and barriers. import and export formats may change, and new data
Sedimentological core descriptions are a very types may be required.
important input here. Working with the database prior to the actual
• Various stratigraphic models, including sequence modelling is an important and time consuming task,
stratigraphic models and lithostratigraphic models and sometimes the database preparations and quality
(Chaps. 7 and 8). This also includes palogeographic control may take more resources than the actual
maps which may be quite important for understand- modelling. It is important to acknowledge this early in
ing trends and architectural topology within reser- the model planning, to avoid delays and errors in the
voir units in the model construction. actual modelling.
• Evaluation of compaction and diagenesis, includ-
ing impact on reservoir property distribution (see
Chaps. 4 and 21). 22.2.3 The Conceptual Model
The conceptual model can be defined as a synthesis of

22.2.2.4 Reservoir Technical Data our understanding of the reservoir targeted for
Reservoir technical data includes data that are related modelling, based on all available input data from all
to dynamic understanding of the reservoir. Typical the subsurface disciplines (see Sect. 22.2.2). In many
reservoir technical data are: cases, it will also be crucial to include additional
• Pressure data, both those observed prior to produc- information, such as well data from neighbouring
tion and the pressure development during fields, regional models and understanding, experience
production. from analogue producing fields, and/or outcrop data.
• Various fluid data and PVT data, such as Typically the conceptual model can be summarised
viscosities, fluid densities, Bo, Bg (shrinkage factors with a number of drawings and qualitative
for oil and gas), etc. descriptions prior to the modelling.
Fig. 22.4 Comparison of (a) chronostratigraphic and (b) units will significantly affect flow patterns. Hence it is critical
lithostratigraphic correlation styles [modified from van Wag- for the reservoir model to have a solid conceptual model as basis
oner et al. (1990)]. Picking the “correct” correlation of sand
22.2.3.1 Sequence Stratigraphy and 22.2.3.2 Link Between Geology and

Architectural Elements Petrophysical Properties
The conceptual model is an important “driver” for In addition to positioning architectural elements in the
the resulting digital model, and understanding the 3D model, it is important to understand and predict
link between stratigraphy and reservoir architecture how mineralogy and diagenesis will affect the
is essential. Figure 22.4 illustrates how a sequence petrophysical properties in the reservoir. The diage-
stratigraphic correlation will alter the reservoir netic processes, taking an unconsolidated sediment to
connectivity dramatically compared with a simpler a rock is thoroughly discussed in Chap. 4, and many
lithostratigraphic correlation. Using the concept aspects in this process are important considerations
sketches as guidelines for the modelling is important, when constructing a model.
and comparing the initial concept drawings with the The mineralogy, grain size, sorting and post-burial
digital result is a valuable exercise (Fig. 22.5). How- chemical processes have huge effects on porosity,
ever, it is important to acknowledge that the concept permeability, saturation and wettability. These
itself has significant uncertainties, and the modelling properties impact both on static volumes and dynamic
process (e.g. the data analysis, Sect. 22.3.3.3) can flow behaviour in the reservoir. For example, the rela-
reveal information that may force a revision of the tive permeability functions are quite important for the
conceptual model. Thus it is common that the uncer- dynamic flow (see Chap. 21), and they depend heavily
tainty of the concept model is tested by building alter- on fluid properties, pore shape, sorting, grading, grain
native models (see Sect. 22.5). coating minerals, cement types, mechanical strength
and microfractures. These, in turn, are functions of
primary mineralogy, diagenesis and the basin burial
Fig. 22.5 Conceptual model for a reservoir interval in Oseberg directly comparable, but the sketch shows that a channelised
Sør, compared with a digital model realisation. The conceptual system in South changes into a shoreline system towards North
model (left) is a hand-drawn sketch, while the computer model and gets progressively deeper. This is replicated in the digital
(right) is a screenshot from the modelling tool. Colours are not model (Image courtesy of Paul J. Valle)
history etc. (see Chaps. 4 and 11). Hence it is crucial to • As seismic resolution is limited, additional zones
understand how sediments may change properties dur- may be included from well log zonation and corre-
ing burial and uplift in order to make a good assess- lation, and the conceptual model. This refinement
ment of the petrophysical data that are applied in the will outline the geological framework.
static and dynamic models. • Based on the geological framework, one or more 3D
reservoir grids are constructed. In many cases a high
resolution geological grid (also called static grid) is
made, where the actual property modelling takes
22.3 Constructing the Static Reservoir place, while a coarser flow simulation grid is
Model constructed for the purpose of dynamic modelling,
see Sect. 22.4.
Essentially, the static reservoir model is constructed in • Property modelling is the step where each cell is
two steps. First a structural model expressed as a 3D assigned various petrophysical values (such as
modelling grid is generated, and then this grid is filled porosity, permeabilities and fluid saturations) and
with rock properties. An overview on this process is other values (such as facies and region identifiers).
outlined in Fig. 22.6. The main elements in the model • If both a high resolution static grid and a coarser
construction are: flow simulation grid are generated, the process of
• Model the seismic horizons and faults. Seismic transferring properties to the coarser grid is known
interpretation provides the most important input as reservoir property upscaling.
for building the gross skeleton of the reservoir, In the next sections, these steps will be explained in
and make the seismic framework. more detail.
Fig. 22.6 Overview of a typical model construction work flow. (step e). An optional test grid may be made for quality control
Seismic interpretation is usually done in time domain and depth- (step f). Based on this, two more 3D grids are made (steps g, h),
converted (steps a–d). Isochore maps are made and merged which are filled with properties (steps i, j)
with the seismic framework to make the geological framework
22.3.1 Seismic and Geological Framework
The seismic framework typically defines the top and

base reservoir and in some cases internal reflectors
(Fig. 22.6, steps a–d). Faults are interpreted from
discontinuities in the seismic reflectors. As the seismic
interpretation usually is done in the Two-Way Time
(TWT) domain, a depth conversion will be needed. As
the seismic resolution usually is limited to a few
horizons in the reservoir, additional zones are com-
monly added based on well correlation (supported by
the conceptual model). This provides the basis for
creating isochore (thickness) maps, which are merged
into the seismic framework (including faults) by the
modelling software. The resulting mix of interpreted
seismic surfaces, faults and calculated intermediate
horizons from well correlation and isochores makes
the geological framework, which can be considered as
the most precise model of the structural elements Fig. 22.7 While the geocellular model may contain millions of
(Fig. 22.6, steps e–h). Note that different modelling cells, it is still a major simplification as each cell can only hold
one set of parameters. Hence variability seen in cores and
software has different approaches to the details of this
outcrops will be averaged
work flow, and some software incorporates parts of
zonal modelling directly in the 3D gridding process.
22.3.2 3D Gridding for Static Modelling
The geological framework (comprising interpreted

and calculated horizons) is usually surface oriented,
i.e. it is made by a number of maps which make up a
skeleton in three dimensions. However, properties
may have values in any position between the surfaces.
In order to populate the digital model with properties,
we introduce an additional data type: the 3D grid. The
3D grid consists of numerous cells to build the reser- Fig. 22.8 The size of a typical static grid cell superimposed on
an outcrop in Ainsa, Spain (image courtesy of John Thurmond)
voir volume, analogue to a “flexible” Lego brick
model, and we often refer to this a geocellular
model. What is important to understand in the property wells logs will represent (Figs. 22.7 and 22.8; see
modelling is that each grid cell in the model is also discussion in Sect. 22.4.2).
assigned one value for each reservoir property to be The number of grid cells in a model (hence the
modelled. The typical size for a grid cell is from 20 m average grid cell size) is often a trade-off between
to a few hundred metres lateral resolution (X and Y the need for the necessary resolution to be able to
direction of grid) and 0.5 m to up to 10’s of metres represent geological information adequately and
vertical resolution (Z direction of grid); hence computational storage and execution time. Large
petrophysical values are averaged over a much larger grids (i.e. many grid cells) will be progressively
volume than primary input such as core plugs and more time consuming to work with, both from a
a b
Fig. 22.9 Going from grid (a) to discrete facies model (b) and further to a continuous property model (here porosity) (c)
geomodelling perspective but especially from a dynamic calculation of production profiles and reservoir
simulation point of view. The maximum number of grid management.
cells in a model is limited by prevailing computer
power, but as an estimate static models can handle
up to 100 million cells which is typically two orders 22.3.3 Property Modelling
of magnitude more than dynamic models can deal
with. There are a few important requirements When the grid construction is finished, the next step is
concerning the geological modelling grid: to fill in the cells with properties. The most important
• The grid should be as uniform as possible, hence properties are the petrophysical ones (porosity,
the grid cell size should not vary too much in the permeabilities, saturations), but the grid may contain
a large number of properties. We can classify the
model, for each zone. The underlying methods for
properties into two main categories:
property modelling commonly assume a uniform
grid, and the result may be distorted if the grid is Discrete or Categorical Properties These properties
non-uniform. hold a single integer value per cell. This cell value is
• The grid orientation both vertically and laterally commonly just an alias or indicator for a given feature,
should follow main directions of the heterogeneities. such as a facies name. For example, cells with integer
It is important to have a good multidisciplinary value 1 may be associated with facies “channel-sand”,
co-operation when designing the geological modelling while cells with value 2 may be associated with facies
grid, taking into account not only the input parameters “deltaplain-mudstone”, etc. The use of numbers and
but also the modelling objectives and practical use of associated facies names will vary from case to case.
the grid. A simple grid (i.e. coarser grid resolution and Other examples of discrete variables are region and
fewer grid cells) designed for a crude in-place volume fault block identifiers.
estimates will have quite different requirements rela- Continuous Properties These values represent a sca-
tive to a grid constructed for a static model that shall lar representation of a petrophysical property directly,
serve as basis for a dynamic simulation model for the such as porosity (which is usually given in a fractional
form, such as 0.25). The properties are not restricted to making reasonable decisions on the basis of
traditional petrophysical properties; e.g. they may such analysis, and is an extensive science.
include other physical properties such as geophysical Geostatistics is distinct from statistics in three main
data (acoustic impedance, P wave velocity, etc.). aspects: (1) focus on the geological origin of the data,
In order to achieve realistic results using a property (2) explicit modelling and treatment of spatial
model, a mix of categorical and continuous properties correlation between data, and (3) treatment of data at
are used. Best known is the “two-stage” approach different volume scales and levels of precision
applied for petrophysical properties. First, the main (Deutsch 2002).
flow units are modelled using a categorical property; The geomodeller will need a certain level of skills
so-called facies modelling (Sect. 22.3.3.4). Next, each within geostatistics as the modelling software draws
facies is populated independently with continuous heavily upon the geo-statistical methods and terminol-
petrophysical properties (Sect. 22.3.3.5). The principles ogy. A sufficient skill level is usually a good under-
of going from a facies model to continuous properties standing of geostatistical principles and terms; the
are illustrated in Fig. 22.9. detailed mathematics behind it is usually not required.
The next question is how shall the property model However, it is common that oil companies employ a
be populated in practice? What input data are avail- number of statistical experts who are available as
able? Generally speaking, we have the following data support when doing reservoir modelling in general,
available: and in particular the uncertainty analysis associated
• Well data, in the form of a zone log, facies log, poros- with it (Sect. 22.5).
ity, saturation and permeability logs. These are usually
treated as hard data, meaning that well observations 22.3.3.2 Blocking of Well Data
should be conditioned properly in a model. Before starting the actual property modelling, the
• Seismic data, in the form of depth-converted available well data need to be sampled into the scale
interpretations and attribute cubes. These data of the geological modelling grid. This process is most
have usually poorer resolution than the cells we commonly referred to as blocking of wells/well data
model. Seismic data are regarded as less certain but could also be called upscaling of well data. This
than well data, and the use of seismic data will involve sampling and averaging of petrophysical log
vary a lot from field to field, depending on seismic data into the scale of the geological grid.
resolution and availability. In general, seismic data Blocking of well data is not a trivial task. The well
should be treated as semi-quantitative. logs (see Chap. 16) are a result of an averaging due to
• The conceptual model (see Sect. 22.2.3) may be the logging tools, particular measurement technique,
regarded as “soft” data, meaning that is cannot be and the sampling increment is usually approximately
conditioned directly; rather the model is “tuned” to 15 cm for conventional logs. The actual resolution is
acknowledge the main features in the concept usually less, typically from 30 cm to several metres,
description. depending on the tool and measurement conditions.
This means that small-scale laminae (from mm scale
22.3.3.1 The Role of Geostatistics to tens of cm scale) commonly seen in cores are not
Before we proceed into actual property modelling captured by the logs, and shoulder bed effects (the
methods, it is important to emphasise the role of smearing or averaging when going from one lithology
geostatistics in reservoir modelling. Both the analysis to another) can be significant. When these logs are
and modelling of facies and petrophysical properties blocked, a further averaging is done. Horizontal
are heavily driven by geostatistical principles and wells have additional challenges. An example of well
methods, and the use of geostatistical methods is also blocking is shown in Fig. 22.10. Loss of detail from
increasing within structural framework modelling. the original logs may be compensated by using a finer
The field of statistics in general is concerned with grid; however using too many grid cells will be a
quantitative methods for collecting, summarising and disadvantage as it will increase the needed computa-
analysing data, as well as drawing conclusions and tion power later in the modelling process.
Fig. 22.10 Blocking well data. This example shows blocking there will be some loss of details (arrows), and the geomodeller
(upscaling) of a facies log and a porosity (ϕ) log. For the facies will have to decide if that is acceptable for the purpose of the
log, the dominant facies will be used (majority rule). For the model
porosity log, an arithmetic average will be applied. In both cases
a b
c d
Fig. 22.11 Principles in data analysis. In (a), a histogram gives In (b) the porosity is crossplotted with depth to reveal depth
the data range for porosity. Outlier data (red arrow) must be trends. In (c), the porosity vs. permeability correlation is plotted,
identified and removed if they are evaluated as erroneous data. while in (d), the spatial correlation (variogram) is displayed
22.3.3.3 Data Analysis important choice of whether to do facies modelling or

The next important step in property modelling is data not should be based on the modelling objectives, the
analysis. The blocked well data (and other data if conceptual model, the need for accuracy and the avail-
relevant) are analysed using geostatistical techniques able time and resources.
in order to be able to set up the modelling process Whether facies is modelled in a simple fashion (e.g.
optimally. The analysis includes: just two or three facies) or more complex facies patterns
• Histogram plotting, which shows the data distribu- depends on the conceptual model and model purpose.
tion of a given property (Fig. 22.11a). Models that should be input to dynamic simulation
• Scatter plots of the properties versus spatial usually require facies models as the facies dictates
dimensions. This will show if the property has flow units and heterogeneity which can have significant
trends that are important to acknowledge in the impact on fluid movements in the subsurface.
modelling. One trend typically found is a porosity From a model technical view there are several
decrease with depth (Fig. 22.11b). available techniques for making a facies model. The
• Crossplotting of various properties (multivariate main techniques, with a short description of pros and
analysis). Commonly, a correlation between poros- cons, are briefly mentioned here:
ity and the logarithm of permeability is found
(Fig. 22.11c). Pixel-Based Facies Modelling In contrast to object-
• Spatial correlations are found by variogram analy- based methods, the pixel-based methods use the 3D
sis. This is a common technique in geostatistics, grid directly in the modelling. A number of different
analysing how the property will vary in a spatial pixel based methods are available.
context (Fig. 22.11d). • Truncated Gaussian fields (Deutsch 2002, Chap. 6).
The data analysis provides the fundamental input to This methodology is useful for modelling large
both facies modelling methods and modelling of con- scale facies objects that appear in a specific order,
tinuous properties. like the boundaries between the lower, middle and
upper part of a shoreface, or the boundary between
shoreface and delta plain (Fig. 22.12a). Truncated
22.3.3.4 Facies Modelling Gaussian fields are usually unsuitable for flow units
The word “facies” has several different definitions in which have distinct shapes.
geology; for reservoir modelling, a facies is an archi- • Sequential indicator simulations (Deutsch 2002,
tectural element that has a certain geometry and a Chap. 6). This methodology can output a large
properly defined distribution of its internal continuous number of random patterns and has the advantage
properties. As an example, a fluvial reservoir zone of quite fast modelling when conditioning to a large
may comprise channel facies with high porosities number of wells. However, the output from indica-
and a quite specific elongated shape, while the tor methods may have difficulties in matching the
overbank facies has poor reservoir properties with no conceptual geological model.
particular shape (only filling the space between chan- • Multi-point statistics (Deutsch 2002, Chap. 11).
nel bodies). The boundary between each facies may be This technique utilises training images, providing
quite sharp, see example in Fig. 22.9. It is important to an intuitive way for the geologist to describe the
understand that facies modelling is an optional step; is geology. This methodology has the advantage of
not required in order to get a model that can be used quite fast modelling when conditioning to a large
for volumetrics and flow. However, a facies model has number of wells, and may in some cases replace the
the advantage of introducing geologically reasonable need for using object-based methods.
shapes into the geomodel, thus honouring the concep-
tual geological understanding. Most reservoir models Object-Based Facies Modelling This is a class of
utilise facies modelling in some sense in order to facies modelling techniques that utilises a statistical
improve the petrophysical architecture, although com- marked-point process to create facies objects based on
plex facies models are demanding to construct. The geometrical input (Deutsch 2002, Chap. 7). Object
a b
c d
Fig. 22.12 Facies modelling often involves combining several techniques. In (a), a truncated Gaussian method is applied, while
different object models are applied in (b) and (c). These are merged by simple grid operations into a resulting facies realisation, in (d)
based methods place the various facies bodies into a result. An example of a merged facies model is shown
background facies, according to statistical rules and in Fig. 22.12d.
geometrical relationships. Object modelling is intui-
tive in the way that input is provided by the modeller 22.3.3.5 Petrophysical Property Modelling
on the shape and size of the objects to be modelled, but Petrophysical property modelling is the process of
may be complex to parameterise and to condition (fit) distributing reservoir properties into the geological
with all observations, such as blocked wells and seis- grid. This is performed individually for each reservoir
mic data. Sometimes the methods are slow to run zone of the geological grid. If facies have been
(taking hours on high-end computers). The most typi- modelled the properties are distributed individually
cal example of object modelling is fluvial channel for each modelled facies for each reservoir zone.
deposits which are modelled as elongate sinuous A very common approach for petrophysical prop-
objects within a “background” of the delta plain erty modelling in all modern modelling software is the
deposits (Fig. 22.12b). use of geostatistical Gaussian methods (Deutsch 2002,
Depending on the modelling tool, most or all of the Chap. 8). This implies that the properties that will be
techniques can be conditioned to (fit to) several data modelled have a distribution that is close to a Gaussian
types. Most vital are well data, as well observations (Normal) distribution, or a distribution that can be
are “hard” data that should fit naturally in the model easily transformed to a Gaussian shape. To achieve
after facies modelling is done. In addition, most tools the Gaussian distribution, a transformation sequence is
can use seismic conditioning, which will guide the needed, and establishing that sequence is an important
modelling algorithm to generate a pattern or feature part of the data analysis.
(such as “sand-probability”) extracted from the seis- The Gaussian modelling has two output modes:
mic data. 1. Prediction mode or kriging mode gives the
To improve the result from facies modelling, sev- expected average outcome. This is illustrated in
eral different techniques can be applied within one Fig. 22.13a. Observations (wells) will stand out as
unit and finally merged in order to achieve a realistic bull’s eyes, and outside the influence range (which
Fig. 22.13 Petrophysical modelling based on Gaussian methods. (a) Shows kriging mode, while three realisations from stochastic
mode are shown in (b), (c) and (d)
is given by the variogram, cf. Fig. 22.11d), the An example of a model from a North Sea field is
model values will tend towards the average. There given in Fig. 22.14. Here stochastic facies are filled
will only be one single outcome “realisation” in with Gaussian simulation of porosity.
kriging mode. An important weakness with kriging
is that the result becomes too smooth, and therefore
unsuitable for realistic flow modelling.
22.3.4 Volumetrics and Model Ranking
2. A Gaussian simulation (stochastic mode) will
add variability between well observations
Most of the available modern geomodelling tools offer
(Fig. 22.13b–d). Due to the sparseness of data
predefined panels for volumetric calculations. They
(wells), the modelling software will use random
are all based on the same classic equations for volume
generators in combination with the geostatistical
calculations of oil in place (Vp)
model to populate the cells. The result will appear
more realistic and reproduce the variability seen in
Vp ¼ Vr N=G φ Osat
scatter plots and histograms from the data analysis
(Fig. 22.11), and Gaussian simulations are usually
(see Sect. 21.2) When standardising to surface pres-
preferred for flow models.
sure and temperature, a shrinkage factor (Bo) is also
The input settings (such as the specification of
needed. A similar equation is applied for gas in place
variogram models) for kriging or Gaussian simulation
volumes.
are generally the same, and it is worth noting that the
In simple cases, no facies are modelled, and
average of an infinite number of stochastic Gaussian
properties are modelled with kriging or other interpo-
realisations would generate the same result as the
lation techniques. In such cases volume calculations
kriging model.
are straightforward. However, today most major oil
Fig. 22.14 Example from an offshore Norwegian field model, with facies realisations (left) which are input to porosity realisations
(right). Both the facies and the petrophysical modelling within each facies are based on stochastic methods
companies are utilising stochastic techniques for visualise the outcome space and to run volumetric
facies and petrophysical property modelling, as these calculations on all of these model realisations.
produce more realistic variability. This means that the Often tens to several hundreds of stochastic model
modelled distribution of properties and resulting realisations are generated. For computer capacity
volumetrics represent one possible outcome within a reasons, it is rare that all realisations are taken further
larger outcome space. Hence it will be necessary to run to flow simulation models, hence a ranking of
multiple realisations of the geomodel, to be able to realisations is required. Picking representative model
realisations may be challenging. One possible Dynamic models can also deal with cell-face
approach may be to choose the realisations that are properties. This means that the numerical properties
volumetrically closest to the average off all the model are not limited to the cell-centre, but also add
realisations. However, it will also be important to properties controlling the flow between two adjacent
make a visual inspection to ensure that the most repre- cells. This is quite often used to model the effect of
sentative model realisations are chosen, and it may thin barriers and baffles, both of stratigraphical origin
also be necessary to investigate those realisations and those created by faults.
that represent end-members in some sense, such as
volumetrics or connectivity. In a complex situation
with a large number of model realisations, so-called
22.4.1 3D Gridding for Flow Modelling
streamline simulations (which are fast and simplified
flow simulations) may be used as a guide for picking
The grid used in flow simulation is usually coarser
an appropriate set of model realisations.
than those that are used for static models. To save
The fact that many reservoir models today are
computational time, flow simulation grids usually
based on stochastic modelling techniques must be
have coarser cells in parts of the reservoir which are
carefully considered when using models for well
of less interest, such as attached water bodies
planning or other decisions. For example, a horizontal
(aquifers) which may be important for reservoir
well optimised for penetrating highly permeable layers
energy, but do not require the detailed cells required
in one model realisation, might be positioned in low
for two- or three-phase flow near oil and gas
permeability layers in a different stochastic realisation
producers. This is possible to achieve as the flow
of the same model, due to the stochastic distribution of
simulator can handle varying cell sizes, in contrast to
facies and petrophysical properties. Hence, it is of
property modelling methods which commonly require
critical importance to consider multiple realisations
quite uniform grids. A comparison of such grids is
when doing such analysis.
shown in Fig. 22.16.
There are cases where the static and dynamic
modelling is performed on the same grid, and this
22.4 Constructing the Dynamic
simplifies the modelling. In addition to the require-
Reservoir Model
ment for a relative uniform grid, this also demands an
in-depth analysis of proper scales, so-called REV
While geological (static) models are the domain of
(Representative Elementary Volume) analysis, which
geologists, geophysicists and petrophysicists, the
is discussed in Sect. 22.4.2.
dynamic models are in the hands of the reservoir
engineers. Flow simulation is a major activity in oil
companies, as such simulations can be a very valuable
tool to understand the reservoir dynamics, optimise the 22.4.2 Upscaling from Static to Dynamic
production and compute recovery factors, Chap. 20 Models
(Fig. 22.15). Flow simulations are quite computation-
ally demanding, and require powerful “number- As previously stated, there are several ways of defin-
crunchers” that are occupied continuously, as each ing the scale relationship between the geomodel and
flow simulation run may take from several hours to the dynamic simulation model. In most cases, the
many days. geomodel is constructed at a finer scale (for accuracy),
Most dynamic models inherit both the structural and a property upscaling step is necessary in order to
and property model from the static model, and tight have a simulation model with a number of grid cells
integration between the disciplines is required for an giving reasonable computational time for the dynamic
optimal result. In addition, the dynamic model needs simulations. We will not go into the details of various
to incorporate special data types only relevant for flow approaches and techniques for upscaling the
modelling, such as relative permeabilities and a num- parameters of the geological model from geogrid
ber of temperature and pressure dependent fluid scale to simulation grid scale, however a few main
properties. principles can be outlined:
Fig. 22.15 Example from a flow model, taken from a major (c) the well will produce primarily gas, while the model reveals
North Sea field. In (a), the reservoir is filled with oil (green remaining oil in the lowermost intervals, which would eventu-
colour). During production, gas from gas cap (red) will expand ally require a new dedicated well to be produced
and replace the oil, moving towards the well (b). In final stage
Fig. 22.16 Grids for flow models have other requirements than
grids for static property modelling. In this case, the dynamic grid
has varying cell sizes that are fully acceptable in flow
simulators, while the static grid shows a uniform grid that is
required by most modelling algorithms
• A proper handling of scales from one data-type to

another is one of the most challenging tasks within
reservoir modelling.
• The upscaling from a static model to a dynamic
model has historically received much attention, but
the actual scale change is relatively small compared Fig. 22.17 An overview of typical measurements and/or data
to other upscaling operations we do (Fig. 22.17). types, and the relative upscaling factors
For instance, the upscaling of core plug data to well
log resolution is a much larger leap, but is quite average that can be difficult to find, in particular
often ignored or treated simplistically. for permeabilities.
• Quite often, core analysis data for relative For simple scalar properties such as VSH (volume
permeabilities and capillary pressures are applied of shale), porosity and saturations, the upscaling
directly in the dynamic model, which implies a 109 methods are straightforward, based on weighted arith-
upscaling factor! (Fig. 22.17) metic averages. For categorical properties (such as
• While porosity and saturations are trivial to facies) the so-called majority rule is applied, e.g. the
upscale, the permeability tensor is much more chal- dominating facies will “win” (an average of facies has
lenging. Even more difficult are the saturation no physical meaning). For upscaling of permeability
dependent functions, such as relative permeabilities and saturation functions, a large number of methods
and capillary pressures. exist, and choosing the optimal approach is case
• Applying measurements from one scale at another dependent and hence challenging. A review of differ-
scale is not necessarily wrong; if the reservoir ent upscaling methods is provided by Christie (1996).
interval or facies is homogenous, or the spatial A serious weakness with the standard work flow is
scale of variation is relative large, then core plug that many upscaling steps are neglected, or carried out
samples may be sufficiently valid for large grid inappropriately. One common error is to apply core-
cells. plug values (permeability, porosity, etc) directly at
• The goal with every upscaling exercise is to repre- larger scales (Fig. 22.17). In addition, only the sam-
sent the small-scale data using an appropriate aver- pling size is considered, but the scale of heterogeneity
age; it is the method to achieve the appropriate within the rocks is actually unrelated to sampling.
Fig. 22.18 History matching loop. Modelled dynamic behaviour is compared with actual, and calibration of the model may be done
in the dynamic model directly (inner loop) or the static model (outer loop)
These issues are addressed in alternative upscaling 22.4.3 History Matching

philosophy called geologically based upscaling,
which is based on separation of scales seen in the History matching is an important work process within
sedimentological patterns. One goal here is to investi- reservoir modelling, and has long been a major part of
gate if the static geogrid and the dynamic simulation conducting reservoir simulation studies for the pur-
grid can be constructed at the same scale. The property pose of effectively managing reservoirs. It is defined
modelling will be based on a hierarchical procedure, as the adjustment of reservoir model input parameters
where the scales of the different heterogeneity levels so the model’s predictions match historical production
are identified. To identify the REV levels is essential data. This adjusting is necessary as long as the static
here as they define the smallest volume where the model (which provides input to the dynamic model) is
measurements yield a proper representation of the not capable of conditioning on production behaviour.
whole volume. The approach is used to provide a History-matched models are used to test develop-
systematic upscaling work flow from the finest scale ment or depletion plans, and thus provide the basis for
(pore level) to flow simulation grid level. For further business decisions about how the reservoir will be
reading, see e.g. Liu et al. (2002), Pickup et al. (2005) produced. The question is how should the model be
and Scaglioni et al. (2006). adjusted? There are two end-members of doing history
If the project has enough time available, the ulti- matching (Fig. 22.18):
mate check of the dynamic impact of upscaling is to 1. The inner loop. After the model has been upscaled
run simulations on the geogrid scale. Running a from the static model, the reservoir engineer adjusts
dynamic simulation on a fine-scaled geomodel might the flow simulation model in order to match model
take weeks, however this can represent a very good with production data. For example, the reservoir
consistency check. engineer may suggest that the permeability should
be multiplied with a factor ten in selected areas to Our insight of the reservoir evolves through the life
improve match. of a field. As the field moves from the initial explora-
2. The outer loop. In this case the reservoir engineer tion phase, via the field development phase, through
works together with the geomodeller to modify the the early production phase and eventually ends up in
static model that provides input to the dynamic the late stage IOR (Improved Oil Recovery) phase, the
model. amount of available static and dynamic reservoir
There is no doubt that the “inner-loop” is most information increases. Hence the development of res-
frequently applied as it is simpler and results will be ervoir understanding and modelling is a dynamic and
quicker. However, it breaks the link with the static continuous process. Uncertainty is a complex term,
model, and the models can diverge significantly. In and can be classified as follows:
many cases, the reservoir engineer can achieve a suffi-
cient history match with values that cannot be Uncertainty Due to Lack of Data Although we
defended from the static model and concept model. assume we know the geostatistical model, we lack
Because the goal of history matching is to develop a data in the reservoir. For instance, at some co-ordinate
model or models that can be used to make reliable between wells, we cannot predict if we have facies A
predictions of future performance, it is important to or facies B, as both can be possible. This is illustrated
ensure that history match changes are consistent with in Fig. 22.14, where the stochastic method predicts a
the geologic concept that has been inferred from avail- different facies and porosity at a single position in the
able geologic data. If the history match changes are reservoir.
not consistent with geologic data, it is less likely that
the predictions of the resulting model will be reliable. Uncertainty Due to Lack of Knowledge This relates
That is, if history match changes are not consistent more to the “human factor” of uncertainty. As the data
with a geologic concept, it will be difficult to extrapo- analysis is not definite, we may end up with many
late to other parts of the reservoir, or in some cases, alternatives (scenarios) regarding how to set up the
even to later times (Lun et al. 2012). model. For instance, in a fluvial reservoir, we may
Hence, working along the outer loop will more or have one case with narrow channels with a North-
less ensure that the model is updated according to the South direction, while another scenario could be to
concept model and in doing so will promote cross- have broader channels trending East-West. We usually
disciplinary consistency. However, as the outer loop is treat this by scenario modelling, i.e. building alterna-
more demanding and time consuming, a hybrid solu- tive models.
tion is also attractive. This implies that the reservoir
engineer tests alterations to model parameters in the Uncertainty Due to Imprecise Measurements All
inner-loop, but at frequent intervals includes the outer- the measurements we perform have some uncertainty,
loop in order to calibrate the static model and ensure and it is important to take this into account. For
that changes made are within acceptable ranges. instance, the density and resistivity logs are uncertain
due to uncertainties in well logging, and those
uncertainties propagate in the estimation of the
22.5 Uncertainty Handling derived petrophysical properties such as porosity and
saturations.
A key aspect of reservoir modelling is the handling of
uncertainty. With limitations in seismic resolution, Uncertainty Due to Unknown Unknowns These
number of available wells etc., there will always be a are the uncertainties we do not test for – we simply
significant uncertainty associated with the description do not know their existence. One example could
and understanding of the subsurface. The impact of be that none of the concept models we test are the
this uncertainty can to a certain degree be estimated right one. Another example is that the software
through constructing alternative scenarios and statisti- performs erroneous calculations we are not aware of,
cal realisations of the model. The aim of this is to or (usually more frequent) that the reservoir modeller
quantify the impact on static and dynamic volumes, makes mistakes that are not captured by the quality
impact on depletion planning and well placement. control.
Quantifying and communicating uncertainty from References and Further Reading

reservoir models is a big topic, and is one which is not
covered in any detail here. It is important to under- Bjørlykke, K. 2010. Petroleum Geoscience: From Sedimentary
stand and communicate to the decision makers (for Environments to Rock Physics. Springer, 508 pp.
Christie, M. 1996. Upscaling for reservoir simulation. Journal of
example of a field development or an IOR initiative) Petroleum Technology 48(11), 1004–1010. ISSN: 0149-
that the deterministic outcomes of a modelling exer- 2136
cise (such as volumes, production profiles or well Damsleth, E., Tjolsen, C.B., Omre, H. and Haldorsen, H.H.
positions) are only one valid solution in an often 1992. A two-stage stochastic model applied to a North Sea
reservoir. SPE (Society of Petroleum Engineers) Reservoir
large outcome space. If some of the input parameters Engineering 44(4), 402–408.
are skewed (e.g. having a large probability of a fluid Deutsch, C.V. 2002. Geostatistical Reservoir Modeling. Oxford
contact being deeper rather than shallower) this will University Press, 376 pp.
affect the uncertainty analysis, and the statistical mean Deutsch, C.V. and Meehan, D.N. 1996. Geostatistical
techniques improve reservoir management. Petroleum Engi-
value may deviate significantly from the deterministic neer International 999, 21–27.
base case model. Haldorsen, H.H. and Damsleth, E. 1990. Stochastic modeling.
Journal of Petroleum Technology 42, 404–412.
Liu, K., Paterson, L., Wong, P. and Qi, D. 2002. A sedimento-
logical approach to upscaling. Transport in Porous Media
22.6 Summary 46(2), 285–310.
Lun, L., Dunn, P., Stern, D., Oyerinde, A., Chorneyko, D.,
The purpose of this chapter is to introduce some of the Stewart, J., et al. 2012. A procedure for integrating geologic
concepts into history matching. SPE Annual Technical Con-
key elements and success factors for 3D reservoir
ference and Exhibition. SPE-159985.
modelling, which now is a major activity in most oil Pickup, G.E., Stephen, K.D., Ma, J., Zhang, P. and Clark, J.D.
and gas companies. Proper planning with clearly 2005. Multi-stage upscaling: Selection of suitable methods.
stated business objectives is a good starting point for Transport in Porous Media 58(555), 191–216.
Scaglioni, P., Ruvo, L. and Cozzi, M. 2006. Implicit net-to-gross
successful reservoir modelling. Constructing a high
in the petrophysical characterization of thin-layered
quality database of modelling input data can be time reservoirs. Petroleum Geoscience 12(4), 325–333.
consuming, however it will ultimately save time for Thurmond, J.B., Løseth, T.M., Rivenæs, J.C., Martinsen, O.J.,
the project. The construction of reservoir models is an Xu, X. and Aiken, C. 2005. Using outcrop data in the 21st
century–New methods and applications, with example from
integrated and multidisciplinary process, involving
the Ainsa Turbidite System, Ainsa, Spain. Ch. 6 in Deep-
static and dynamic data. Many disciplines provide water outcrops of the world atlas: Tulsa, Oklahoma: Ameri-
input to the geomodel, and the geomodel is most can Association of Petroleum Geologists Special Publication
often handed over to reservoir engineers for dynamic CD-ROM.
van Wagoner, J.C., Mitchum, R.M. Jr., Campion, K.M. and
simulations. Keeping all project members actively
Rahmanian, V.D. 1990. Siliciclastic sequence stratigraphy
involved and the ability to communicate clearly in well logs, cores and outcrops: Concepts for high-
between all the involved disciplines is a key factor resolution correlation of time and facies. Tulsa: AAPG
for achieving a good result. Through modelling alter- Methods in Exploration Series, No. 7. The American Asso-
ciation of Petroleum Geologists, 55 pp.
native conceptual scenarios, and by making multiple
Worthington, P.F. and Cosentino, L. 2005. The role of cutoffs in
stochastic realisations, the uncertainty related to integrated reservoir studies. SPE (Society of Petroleum
estimated static and dynamic volumes can be Engineers) Reservoir Evaluation and Engineering 8(4),
quantified. 276–290. SPE-84387.
Zakrevsky, K.E. 2011. Geological 3D Modelling. EAGE
Publications, 261 pp.
Acknowledgement John Thurmond and Sarah Riordan
provided many useful comments that improved the manuscript
significantly.
Chapter 23
Unconventional Hydrocarbons: Oil Shales, Heavy Oil,

Tar Sands, Shale Oil, Shale Gas and Gas Hydrates
Knut Bjørlykke
For many decades conventional oil which could be successful (in 1857 in Pennsylvania), mining for
produced at low cost was present in abundance. A heavy oil continued in many parts of the world includ-
low oil price gave no incentive to look for other ing Germany. In southern California (Ventura and Los
types of resources. It is now clear, however, that we Angeles basins) oil was mined from the early 1860s to
are gradually running out of new sedimentary basins to the 1890s because the heavy oil there would not flow
explore and that the reserves of conventional oil which to the wells (Fig. 23.1).
can be produced cheaply are limited. Tar sands are sandstone reservoirs which have been
This is the reason why many of the major oil filled with oil at shallow depth <2 km (<70–80 C) so
companies now invest in what is often called unconven- that the oil has become biodegraded. Reservoir rocks
tional hydrocarbons. There are large reserves of these which have been buried more deeply and then uplifted
and the main types are oil shales, heavy oil, tar sand, before the oil migration may be sterilised at higher
shale oil and shale gas. Production of hydrocarbons temperatures and are less likely to be biodegraded.
from these sources occurs mainly in onshore basins When the sandstone has not been buried to more
and may be challenging environmentally. than about 2 km depth the sand will remain mostly
Gas hydrates may also become an important source uncemented as loose sand because of a lack of quartz
of hydrocarbons but the development of methods to cement.
produce from such accumulations is still at an early Tar sand contains asphaltic oil rich in asphaltenes
stage. and resins. It has a high content of aromatics and
The reserves of unconventional hydrocarbons are naphthenes compared to paraffins, and a high content
very large, probably larger than those of conventional of nitrogen, sulphur and oxygen (NSO). Most of the
oil. We will therefore not run out of fossil hydrocarbons hydrocarbon molecules have more that 60 carbon
for a long time but the prices will be higher because atoms and the boiling point and viscosity are therefore
these sources are more difficult to produce, particularly very high.
if strict environmental standards should be met. The viscosity of the biodegraded oil is very low and
the oil must be heated so that the viscosity is reduced
23.1 Heavy Oil and Tar Sands before it can be produced by drilling wells. There are
transitions between reservoirs with heavy oil and nearly
Heavy oil seeping out on the surface has been known solid bitumen. Heating can be achieved by soaking the
for a long time and this was easy to exploit for use in reservoir with injected steam so that the heat of conden-
small quantities. Even after drilling for oil became sation to water helps to heat the oil. This is called cyclic
steam injection when the steam is injected in the pro-
duction well and left to soak for a few weeks before
production starts when the oil is warmer. The steam
K. Bjørlykke (*)
Department of Geosciences, University of Oslo, Oslo, Norway may also be injected in a nearby well and driven
e-mail: [email protected] towards the production well. Much of the energy is
582 K. Bjørlykke
Sun
S N
Ventilating shaft
Asphaltum
s
ray
’s
Sun
San
Mirror ds &
Sha shale
Dump Tunnel Oi les
Wat l forma
er b tion
eari
ng s
trat
a
Sulphur Mountain
Fig. 23.1 Mining for heavy oil in the Sulfur Mountain near the tunnel and the sunlight would contribute to the heating of the
Santa Paula, South California (from 1860 to 1890). Mine heavy oil so that it could flow. See the Oil Museum in Santa
tunnels could be 6–700 m long and a mirror was used to light Paula (http://www.oilmuseum.net/)
used to heat the water in the reservoir and it is important reserves of conventional petroleum. There are cur-
to reduce the inflow of water from adjacent rocks. One rently large mining operations at Athabasca
method includes freezing the ground at a distance from (Fig. 23.2). At surface temperatures, which are low
the well to avoid the flow of water towards the well. It is in northern Canada, this tar sand is rather hard
also possible to burn some of the oil in the subsurface to (Fig. 23.3) because of the high viscosity of the oil.
provide heat to heat the oil. The oil must be separated from the sand using hot
It has also been proposed to heat the oil electrically, water (Fig. 23.4). The oil (tar) is very viscous and
possibly powered by a nuclear reactor to reduce the may be denser than water (API <10).
CO2 emissions from burning oil to produce heat. Tar sand is produced by in situ mining where the
The tar sands in Alberta, Canada (Athabasca) are of sand is excavated with heavy equipment (Fig. 23.1) The
Middle Cretaceous age (Aptian, 100 million years). separation of tar from the sand requires very large
The main reservoir rock is the McMurray Forma- amounts of water. About 350,000 m3 of hot water is
tion, a sandstone representing fluvial to tidal used yearly in the treatment of the Athabasca tar sand.
environments, and it is important to find thick There are very large volumes of heavy oil and tar in
sequences of sand with few clay layers. The oil was sandstone reservoirs which cannot be produced at nor-
generated from older source rocks during the mal surface temperatures. In Alberta alone this may
Laramide folding of the Rocky Mountains to the represent 1–2 trillion barrels. Heavy oil with API grav-
west and migrated into the Athabasca sands during ity between 10–20 has viscosities exceeding 10,000
the Cretaceous when these fluival sandstone were centipoise (cP) while bitumen with API gravity below
still at a relatively shallow depth. Starting in the 10 may be very hard and solid with viscosity exceeding
Eocene (50 million years) the sand was uplifted and 1 mill cP. Heating the tar to 200–300 C the viscosity
most of the overburden eroded. The oil was then becomes very much reduced so that it can flow to a
biodegraded by bacteria. These tar sands contain 1.7 well, The heating may occur by soaking the reservoir
trillion bbl (270 109 m3) of bitumen in-place, com- with steam for several weeks before production starts.
parable in magnitude to the world’s total proven This is called CSS (Cyclic Steam Stimulation).
23 Unconventional Hydrocarbons: Oil Shales, Heavy Oil. . . 583
Fig. 23.2 Athabasca tar sand. The oil-impregnated sand is mined and the heavy oil is separated from the sand with hot water
a
Sand Sand
grain grain
Bitumen H2O
Sand
Sand grain
grain
San
Fig. 23.4 Schematic representation of tar sand. The oil may

also stick more closely to the grain surfaces, so the system is
more oil-wet
Another method mostly used in Alberta is called

SAGD (Steam Assisted Gravity Drainage). Here two
b horizontal wells are drilled into the reservoir with tar
Principles of SAGD
ProducƟon
sand. Steam is injected into the upper well to heat up the
well InjecƟon well tar so that it becomes viscous enough to be produced
from the lower well. The tar (bitumen) must generally
be heated to 200 C for the viscosity to be reduced to
Overburden 200 -400 m
about 10 cP. Water and air is then injected to displace
InjecƟon of steam
the steam further into the sandtone reservoir. Each m3
of oil produced may require from 2-10 GJ of energy.
Layer with tar sand
When cheap gas is used as energy for the heating
Oil from heated tar sand
process it may be profitable to produce oil even if there
should be a negative energy balance.
As the steam condenses to water it gives off heat in the
reservoir. A recovery between 40–60% may be expected.
Fig. 23.3 (a) Cores from the Athabasca tar sand. The oil sand is
highly viscous, almost solid, but will flow at high temperatures. The produced oil will also contain some water and
Dark intervals are loose sand mostly with more than 30% poros- clay which must be removed before the oil can be
ity, filled with biodegraded heavy oil. Grey layers are thins processed. The water is then re-injected or used for the
shales which will reduce the vertical mobility of the heated oil
generation of steam, to reduce the use of new freshwater.
during production, illustrated in (b)
584 K. Bjørlykke
It is also possible to produce oil and sand from top due to its low density but will sink when
wells and then separate the sand out as is done when condensed to water. The shales will reduce the vertical
mining the tar sand. permeability and the drainage, but the heating will
Another method is to pump air down boreholes and cause some fracturing which may allow more vertical
burn some of the oil to produce heat. flow. This method requires rather thick and homoge-
Fluvial sandstones like the lower Cretaceous neous sand. Steam can also be injected into one well
McMurray Fm of the Athabasca Tar sand in Alberta for a few weeks until the oil is heated and then pro-
represent a very complex reservoir with respect to duced from the same well. This is called cyclic steam
migration and filling of hydrocarbons. Each channel stimulation (CSS).
sand may represent a small trap which will be first Large amounts of energy and water are required to
filled with early mature oil from the source rock. As produce the steam for this method. If petroleum is
the oil becomes more mature and less dense it may burned to produce the heat, high emissions of CO2
displace the first oil to arrive and finally gas may will result not only from when the oil is burnt as fuel
migrate in and displace most of the oil into shallower but also from producing it. The steam may be mixed
structures. This is a good example of fill and spill in a with solvents (gas) and this may reduce the amount of
series of rather small compartments. heat (steam) required. Much of the water used to
Fluvial channels which are abandoned by channel produce steam can be recovered and used again.
avulsion are filled with clay and mud and they may For each 3 bbl of oil produced one barrel is burned
represent clay plugs which are barriers for fluid flow to produce the steam required for the production.
and sometimes seals in these reservoirs (Fustic et al. 50–70 kg of CO2 is typically produced and emitted
2012). When oil and gas has migrated into shallow for earch barrel of oil produced mainly due to the
sandstones like the McMurray Fm where the production of steam. This heavy oil should also be
temperatures are below 80 C, bacteria are present. mixed with gas during the refining and a supply of
The sand will be loose and indurated due to little or gas is therefore important.
no quartz cement and the temperatures will also be low Heating the oil with electricity has also been pro-
enough for biodegradation at a high rate. Most of the posed and hydroelectric or nuclear power would
biodegradation occurs near the oil/water contact and reduce the CO2 emissions.
since new oil/water contacts may develop during 173 billion bbl (27.5109 m3) of crude bitumen,
migration there may be complex zones of biodegrada- which is 10% of the total bitumen in place in Alberta,
tion. The highest viscosity oil is usually found near the has been considered to be, economically recoverable
base of the oil column where the biodegradation has using current technology at oil prices above 60–100
been most intense (Jones et al. 2008). The OWC USD/barrel. In recent years the price of gas has
contact may however have changed during the fill dropped due to shale gas. Heating the oil with steam
and spill, resulting in more than one intensively produced by gas is now more economical than using
biodegraded zone. Variations in the oil viscosity oil. In some cases there may a negative energy balance
within the reservoir play an important role in the so that the energy from gas exceeds the energy in the
planning of production. produced oil.
Only about 20% is close enough to the surface to be The environmental problems associated with oil
economically mined and the rest must be heated in production from tar sand are considerable and very
place. large amounts of water are required both for the pro-
A cubic metre of oil, mined from the tar sands, duction of steam for subsurface operations and for the
needs 2–4.5 m3 of water. processing of loose sand excavated from surface pits.
Oil may be extracted by steam-assisted gravity As a result the release of CO2 is higher by 20–25%
drainage (SAGD) (Fig. 23.3b). Two parallel horizontal than with normal petroleum production if the CO2 is
wells are drilled so that one is about 5 m below the not captured and sequestrated.
other. Steam is injected in the shallowest well and the In Venezuela there are also very large reserves of
heated, less viscous, oil is drained into the lower well heavy oil and tar sand such as the Orinoco tar sand.
where it is produced. The steam will remain near the While the average surface temperature in Northern
Alberta is only slightly above 0 C the surface temper- most of its hydrocarbons during the Caledonian fold-
ature in Venezuela is much higher (>20 C) and the oil ing in the late Silurian and early Devonian. In Sweden
therefore needs less heating to reduce its viscosity. the Upper Cambrian alum shales have not been buried
Much of the oil in Venezuela is heavy oil and it so deeply and therefore contain more oil which can be
occurs in a foreland basin in front of the Cordilleran distilled off by pyrolysis at 400–500 C.
Mountains in a similar plate tectonic position to the In the Baltic region the lowermost Ordovician shales
heavy oil in Alberta. Also the oil in the Middle East is are even less mature so that more of the kerogen remains
located in a similar foreland basin but here the and in Estonia this shale is mined for oil on a large
reservoirs have not been uplifted so close to the sur- scale. Here the reserves are very large (0.6 109 sm3
face and are therefore in most cases not so oil equivalents – o.e).
biodegraded. Oil shales are used in electric power generation and
It is possible to make an emulsion between bitumen provide 60% of Estonia’s stationary energy. It is also
and water which has low viscosity and which can flow used for oil production and refining. The waste is
in pipelines and be burned directly to produce electric 70–80 Mt. of semi-coke and the mounds exceed
power. The heavy oil and bitumen is rather rich in 100 m in height. The waste is very fine grained and
sulphur which should be removed to avoid pollution. It is alkaline with significant concentrations of sulphides
can also be mixed with gas so that it can be used to and heavy metals.
produce normal oil products. On a global basis oil shales represent a reserve
There are other important tar sand deposits in the almost as large as conventional oil. Reserves of oil
USA (Utah) and Africa (Congo and Madagascar). which can be produced from oil shale are difficult to
Tar sand has only recently been included in data on estimate but they may total close to 5 trillion barrels
world oil reserves and as a result Canada has become (800 109 sm3). This is about equivalent to the
one of the nations with the highest oil reserves. estimated world reserves of conventional oil and gas
(480 sm3 o.e.). More than 80% of the well known
reserves are located in the US but there are probably
23.2 Oil Shales other regions with oil shales which have not been
recorded and evaluated.
Oil shales are source rocks, usually mudstones and The Green River Shale in Colorado, Utah and
shales, with a high organic content (TOC), which Wyoming is a gigantic source of hydrocarbons. This
have not been buried deeply enough to become suffi- organic-rich mudstone was deposited in very large
ciently mature for most of the hydrocarbons to be lakes during the Eocene and the organic matter was
generated. Although they may contain some mostly freshwater algae (Fig. 23.5a, b).
hydrocarbons they must be heated in an oven (pyroly- In Australia there are also very large reserves of oil
sis) to 400–500 C so that most of the petroleum can be shale and they have been used for oil production
generated from the remaining kerogen. already in the latter part of the 19th Century when
Oil shales must therefore be mined near the surface more than 100,000 tonnes of shale were processed.
in quarries and then heated in large ovens so that the In areas including Queensland and South Australia
petroleum can be distilled off. there are Cenozoic shales or mudstones deposited in
Source rocks may be uplifted close to the surface freshwater associated with coal deposits but there are
after deeper burial; it is the temperature history that also older (Permian) oil shales. Strict environmental
determines how much of the kerogen is altered to oil restrictions on oil shale mining have been imposed
and gas. here, particularly because of the association with The
Some source rocks may have been buried to more Great Barrier Reef.
than 5–6 km (160–170 C) and they have then Also in other parts of the world oil shales may
generated and expelled most of the hydrocarbons, but represent large potential oil reserves but they may
some oil and particularly gas may remain. The Upper not be produced with the present technology.
Cambrian alum shale which is found in the Oslo Oil shales must however be mined and heated in
region is a good example of a rich source rock which ovens (pyrolysis) to generate the petroleum and the
has been buried to at least 200 C (5–7 km) and lost energy for the heating is taken from the burning of the
586 K. Bjørlykke
Inset area
b
Big piney
sin
G
Ba
re
en
Wyoming
riv
er
Rock springs
r
rive
en
Gre
Washakie
Evanston basin
Utah
Vernal
Colorado
Uinta
Piceance
Basin creek
Price basin Rifle
River
do
10 0 10 20 30
ra
Grand
lo
Co
Miles junction
Fig. 23.5 (a) Green River Shale. This shale has a high content River oil shale was deposited in shallow lakes and is of Tertiary
of organic carbon from algae and contains mostly Type I kero- (Eocene) age. This shale is very extensive in Utah, Wyoming
gen. It is far from mature and the shale must be mined and and Colorado (adapted from Smith 1980)
heated to 400–500 C to form oil by pyrolysis. (b) The Green
oil shales. The release of CO2 is therefore high also pure methane with low sulphur content and is referred
during production. Oil shales may only contain 5–10% to as sweet gas. Wyoming has very large reserves of
organic matter and the volume of waste will then be coal bed methane in the Powder River, Bighorn, Wind
10–20 times the oil produced. The waste consists of River and Green River basins. These are Cretaceous
coke and also smectite formed in the heating process and Tertiary coals which have been buried more
and is very difficult to store. It is also rich in heavy deeply and then uplifted. Gas may be produced from
metals, including vanadium and uranium which are these coals down to depths of 1.5–2 km (5,000 ft). The
typical of black shales. reserves of CBM are large in the USA (201012 m3)
Production of oil from oil shales requires very large and Canada also has major reserves.
amounts of water, so in dry areas the water supply may Methane is constantly being formed by bacteria at
be a limiting factor. There are therefore considerable shallow depth and it may be argued that this is at least
environmental problems linked to the exploitation of to some extent a renewable resource but the rate of
oil shales as a major source of oil. accumulation is slow compared to our consumption.
Attempts has also been made to heat the shale in the Artificial growth of algae may produce some oil and
ground so that the environmental problems related to gas and consume CO2.
water supply and storage of waste is reduced.
By establishing an underground barrier by freezing
the ground using refrigerated fluid a “freeze wall” 23.4 Shale Gas and Shale Oil
is established. The purpose is to prevent the
hydrocarbons from escaping and groundwater from Organic-rich shales which have been buried to depths
entering into the area which is heated. where most of the oil and gas has been generated and
Some source rocks may have enough permeability to expelled may nevertheless contain considerable
serve as reservoir rocks too. The Miocene Monterey amounts of gas. The gas remaining in these shales is
Formation in California is an organic-rich diatomaceous present in very small pores and may also be partly
source rock, which is also a reservoir rock. Oil can in this adsorbed on remaining organic matter or its residue
case be produced because of tectonic fracturing which (coke) and on clay minerals. The shales have been
has enhanced the permeability. Source rocks may also uplifted and may therefore have small extensional
be interbedded with thin sandstones or limestones and fractures, but they must be hydrofractured by water
then only a very short migration is required. injection to increase the permeability.
Barnett Shale is tight shale of Mississippian age in
Texas, containing at least 2.5 trillion cubic feet of gas.
23.3 Coal Bed Methane (CBM) It is referred to as a tight gas reservoir. Much of the gas
is in urban areas such as the Dallas-Forth Worth area.
Coal and kerogen rock with a high content of plant The permeability of the shale matrix is generally
material (Type III kerogen) are the main source rocks very low but there may be thin silty layers and also
for gas which will migrate to a reservoir rock or up to fractures that increase the effective permeability.
the surface. Relatively large quantities of gas (meth- Hydraulic fracturing can be carried out to further
ane) will, however, be retained in the coal because its increase the permeability, and horizontal drilling also
micro-porous structure provides a very large surface helps to produce more gas. The Woodford Shale
area for adsorbed gas. This can be exploited by drilling (Devonian) in Oklahoma can be almost 100 m thick.
into the coal. Commercial production of methane from Devonian tight gas shales include the Middle Devo-
coal is common in the US. Near the surface gas can nian Marcellus Shales in the Appalachians. These are
leak off from coal but where coal is buried a few now mostly at 1–2 km depth but have previously been
hundred metres much of the gas is still retained in buried much deeper (5–6 km or more).
the coals and can be produced by drilling. Coal has Shale gas is estimated to account for 50% of the gas
very low permeability and the flow of gas to the wells produced in North America by 2020.
depends on thin fractures (cleats) developed during Higher gas prices will also result in increased inter-
uplift. The gas produced from coal is normally very est in shale gas in other parts of the world.
588 K. Bjørlykke
The occurrence of gas in onshore shales has been and produced water may also create environmental
well known in many European countries but very little problems.
has been produced so far. Shale Oil is produced from source rocks, mostly
In the UK Lower Carboniferous shales and coal shales, which have not been buried as deeply as those
were buried deeply in front of the Variscian metamor- with mostly gas and have only reached the oil window.
phic front in the south of England and adjacent parts of Oil is then produced by fracturing in the same way as
the North Sea basin. The Carboniferous Bowland for gas. Shale oil may be found in the same formations
Shale in NW England has been regarded as a potential and sedimentary basins as shale gas but where the
target for shale gas production in Britain. Also in parts maximum burial depth (temperature) has been lower.
of Northern England and Scotland have Upper Until about 10 years ago it was not considered
Palaeozoic sediments been buried deep enough to possible to produce oils from shales. In recent years
generate significant gas. At 4–5 km burial depth gas oil production from shales has increased rapidly par-
is generated from primary oil. The shales are then so ticularly in the US, increasing their domestic produc-
hard and indurated that they can sustaina fracture tion and reducing their import.
network produced by natural fracturing during uplift The fact that it is now possible to produce oil from
and during fracking. shale has changed petroleum geology and the distribu-
In the Liassic and Kimmeridge clays (shales) there tion of reserves. Conventional oil is limited to
are potential source rocks for both thermal gas and reservoirs rocks with sufficient quality and a structural
biogenic gas. They are immature in many areas and trap. Shale oil may be found in very extensive layers of
may not be hard enough for effective fracturing. When shale which have had a maximum burial temperature
developed as carbonate-cemented shales they become corresponding to the oil window.
brittle but the fractures may then become cemented up. The late Cretaceous Eagle Ford Formation of
The total reserves of shale gas is very difficult to South Texas has since 2008 developed into a major
estimate due to uncertainty about gas prices and the oil producer. This is a 100 m thick transgressive
energy policy of each country. organic-rich black shale with some carbonate layers
There are considerable reserves of shale gas which and siltstones. This makes it a good producer in many
can be produced in Europe. Most countries have how- areas. The Eagle Ford shale was deposited on top of
ever introduced strong restrictions on the development the Woodbine sandstone and overlain by the Austin
of shale gas. In many countries such as Germany and Chalk.
France many of the potential areas for such production The Bakken Formation (late Devonian) in the
are rather densely populated and there are strong envi- Willistone basin is found subcropping in much of
ronmental concerns about pollution of groundwater North Dakota, Montana and Saskatchewan but is not
etc. In Germany, Poland and France there are shales exposed. It is black shale deposited as organic-rich
of Jurassic, Permian and Carboniferous age which mud with layers of dolomite. The average porosity is
could be produced, but it is uncertain to what extent about 5% and the permeability is very low (0.04 mD)
they will be produced. Much of these resources are in but due to new fracturing technology production rates
densely populated areas with strict environmental have increased very significantly and the reserves have
restrictions, and France has banned fracking. Also in recently been estimated to about 24 billion bbl.
Spain, Romania and Bulgaria there are large reserves
in terms of shale gas.
Shale gas represents very large reserves of 23.5 Gas Hydrates (Clathrates)
hydrocarbons which can be used directly as gas, but
can also be converted to diesel fuels for cars and Gas hydrates are crystalline solids almost like ice,
trucks. This gas can also be mixed with heavy oil consisting of gas (mostly methane) surrounded by
and tar sand to make regular petrol. Production of water. Most common is methane clathrate which is
gas from shales requires much water for fracturing, water molecules bonded by hydrogen and with gas
0 10 20 30 40 °C When gas hydrates dissolve (melt) one volume of

Sea level
gas hydrate produces 160 volumes of gas. The source of
Stability limit of
the methane is mostly biogenic, from organic rich
1 km
gas hydrates sediments, but gas hydrates may also fill the pores in
sand beds. During the glaciations gas hydrates were
2 km Sea floor
Stable gas hydrate more widespread than now and occurred also beneath
the seafloor in basins like the North Sea (Fig. 23.7). Gas
3 km
hydrates are potentially a very important source of gas.
Gas hydrates unstable Production of methane from sediments with gas
4 km
hydrates may become possible from sand beds
Fig. 23.6 Stability of gas hydrates. Gas hydrates are stable at
cemented with gas hydrates. It is however difficult to
high pressure and low temperatures and are therefore found predict how economic it would be to produce from gas
beneath the deeper part of the continental slopes. As the hydrates.
temperatures in the sedimentary basins increase with depth,
gas hydrates become unstable
23.6 Summary
trapped within its structure. It is stable at high
Unconventional oil and gas as discussed above repre-
pressures and low temperatures (Fig. 23.6). It there-
sent very large energy reserves which are much greater
fore typically occurs on the slopes and below the
than conventional oil. The environmental problems
seafloor in deep oceans. At about 2.0 km water depth
associated with producing these reserves are however
(20 MPa water pressure) the temperature must not
great, particularly for oil shale and tar sand.
exceed about 18 C and at 0.5 km below the seafloor
There is a need for new research to develop methods
(2.5 km depth – 25 MPa pressure) the temperature
which can minimise both the local environmental
must be below 20 C. The temperature near the sea-
impact and the release of CO2 into the atmosphere.
floor is usually less than 2–3 C but the geothermal
When burning hydrocarbons to produce energy the
gradients may be about 30 C/km and it is therefore
carbon dioxide can be separated out and stored in the
clear that gas hydrates can only exist a few hundred
subsurface to prevent emissions to the atmosphere, see
metres below the seafloor (Fig. 23.7).
chapter 24.
0
20 Presen
t day c
onditio
40 ns
Li
60 mi
t of
Glacial period Gas
Pressure (bar)
ga
80 condition range s and
water
hy
100
dr
at
es
120
tab
140
il
Gas hydrate
ity
160
180
200
2 4 6 8 10 12 14 16 18 20
Temperature (°C)
Fig. 23.7 During glacial periods, gas hydrates formed in basins like the North Sea due to the low surface temperatures in front of
the ice (from Fichler et al. 2005)
590 K. Bjørlykke
Further Reading Jones, D.M., Head, I.M., Gray, N.D., Adams, J.J., Rowan, A.K.,
Aitken, C.M., Bennett, B., Huang, H., Brown, A., Bowler, F.
B.J., Oldenburg, T., Erdmann, M. and Larter, S.R. 2008.
Crude-oil biodegradation via methanogenesis in subsurface
Adams, J., Larter, S., Bennett, B., Huang, H., Westrich, J. and
petroleum reservoirs. Nature 451, 176–180.
van Kruisdijk C. 2012. The dynamic interplay of oil mixing,
Judd, A. and Hovland, M. 2007. Seabed Fluid Flow. Cambridge
charge timing, and biodegradation in forming the Alberta oil
University press. 475pp.
sands: Insights from geologic modeling and biogeochem-
Keller, E.A. 2011. Environmental Geology (9th edition).
istry. In: Hein, F.J., Leckie, D., Larter, S. and Suter, J.
Prentice Hall. ISBN-10 0321643755E. 624pp.
(eds.), Heavy-oil and Oil-sand Petroleum Systems in Alberta
Riedel, M., Willoughby, E., and Chopra, S. 2001. Geophysical
and beyond: AAPG Studies in Geology 64, pp. 1–80.
Characterization of Gas Hyddrates. Geophysical
Caineng zou et al. 2013. Unconventional Petroleum Geology.
Developments 14, SEG, Oklahoma. 392pp.
Elsevier, 373pp.
Salvador, A. 2005. Energy: A Historical Perspective and 21st
Camp, W.K., Diaz, E. and Wawak, B. 2013. Electron Microscopy
Century Forecast. AAPG Studies in Geology 54, 33–121.
of Shale Hydrocarbon Reservoirs. AAPG Memoir 102, 260pp.
Selly, R.C. 2012. UK shale gas: The story so far. Marine and
Davies, R.J., Mathias, S.A., Moss, J., Hustoft, S. and Newport,
Petroleum Geology 32, 100–109, Review article.
L. 2012. Hydraulic fractures: How far can they go? Marine
Smith, J.W. 1980. Oil Shale Resources of the United States.
and Petroleum Geology 37, 1–6.
Minereral and Energy Resources, Colorado School of Mines,
Fichler, C., Henriksen, S., Rueslaatten, H. and Hovland, M.
vol. 23(6), 30 pp.
2005. North Sea Quaternary morphology from seismic and
Williams, A., Larter, S.R., Head, I. et al. 2001. Biodegradation
magnetic data: indications for gas hydrates during glacia-
of oil in uplifted basins prevented by deep burial steriliza-
tion? Petroleum Geoscience 11, 331–337.
tion. Nature 411, 1034–1037.
Hein, F.J., Leckie, D., Larter, S. and Suter, J.R. 2013, Heavy-oil
and Oil-sand Petroleum Systems in Alberta and Beyond.
AAPG Studies 64, 131pp.
Chapter 24
Carbon Capture and Storage (CCS)
Helge Hellevang
24.1 Introduction that the reservoir has a seal (cap rock), and sufficient
injectivity (porosity and permeability) to store large
Carbon Capture and Storage (CCS) involves capturing volumes of CO2. These issues are discussed below.
CO2 and then storing it so that it is not emitted to the
atmosphere. The background for CCS is a concern that
increasing atmospheric CO2 concentrations will cause 24.2 CO2 Capture
global climate changes, ocean acidification and a sea
level rise, with dramatic negative consequences for Capture refers to the separation of CO2 from natural
large populations (IPCC 2013). gas or flue (exhaust) gas where the concentration of
Although capture technologies in principle can be CO2 may be as low as 3–4%. In CCS, capture is
implemented for any natural gas or flue gas sources, followed by transport and finally underground storage.
costs put practical constrains on what kind of There is considerable research activity to improve the
operations that can implement CCS. Generally, any efficiency of present-day capture methods and to
large point source of CO2 (>1 Mt CO2/year), such as invent new and better technologies, the main motiva-
coal- or gas-fired power plants, natural gas producers tion being to reduce the cost of the capture process.
and cement factories, has sufficient CO2 production to Capture, whether it is by pre- or post-combustion,
justify installation of CCS technology. The cost may requires significant amounts of energy. For a coal-
however be rather prohibitive. fired power plant, an extra 20–25% of energy is typi-
CO2 may be captured from major point sources. If cally required for capturing and compressing the CO2.
several point sources are in the same geographical There are already several technologies available for
region, then captured CO2 should be transported to a capturing CO2. A review of the advances in these
common hub before being injected into a geological technologies is provided by Figueroa et al. (2008)
formation (Fig. 24.1). This lowers the cost of and only a brief overview will be given here. In post-
implementing the pipeline network and thereby the combustion capture, CO2 is removed from flue gases
total cost of the CCS operation. The geological stor- formed from the combustion of primarily coal. The
age units can be of various kinds, such as saline coal is burned in an air atmosphere, in contrast to oxy-
aquifers, coal seams and abandoned oil and gas combustion that utilizes an oxygen atmosphere with
fields. The requirement is that CO2 is stored as a little nitrogen. In pre-combustion capture, CO2 is
dense supercritical phase (scCO2), that typically removed from the fuel before combustion. This tech-
requires more than 700–800 m burial depth, and nology is applied to gasification plants. In separating
CO2 from the gas streams, various technologies are
being used, such as gas-phase separation, sorption to
liquids or solids, and hybrid processes such as sorp-
H. Hellevang (*)
Department of Geosciences, University of Oslo, Oslo, Norway tion/membrane systems. Each of these technologies
e-mail: [email protected] has their strengths and weaknesses, and the choice of
592 H. Hellevang
Fig. 24.1 CO2 can be captured at major point emitters, such as coal-fired power plants, and piped through a network of pipelines to
major storage units, such as the Utsira Sand
technology must be based on the processing plant, and several minor CO2 point sources (<1 Mt/year) are in
whether the technology will be implemented on a geographically limited area, the sources (flue gas or
existing plants (retrofit), or integrated into new plants. natural gas) can be piped to a common capture facility.
In addition to the size of the facilities, geographical
location plays a role. Because of the costs and risks
associated with pipelines, the distance to the storage
24.3 What Is Required to Store CO2? reservoir should be as short as possible.
24.3.1 CO2 Point Sources

24.3.2 The Storage Reservoir: Injectivity
Size matters in CCS. A large part of the cost of CCS is and CO2 Storage Capacity
related to the capture facilities and the transport net-
work to the CO2 storage site (steel pipelines and The requirements of a CO2 storage deposit are to a large
compressors). The possibility of capturing large extent the same as those for petroleum reservoirs. Injec-
amounts of CO2 at one plant lowers the cost of CCS tion requires that the pressure of the injection well is
($/ton stored CO2). This is critical in order to imple- larger than the reservoir pressure. The amount of over-
ment CCS globally at a scale required to meet the pressure required is given by the injectivity of the
IPCC 2 C target. CCS therefore requires point sources reservoir, which is determined by several factors, such
of CO2, which can be cement factories, coal-fired as the size of the reservoir unit and its hydraulic con-
power plants, or production facilities for natural gas ductivity. The size of the reservoir unit refers to the
(all with a potential of about 1–3 Mt CO2/year). If volume that can absorb the pressure build-up resulting
24 Carbon Capture and Storage (CCS) 593
from the CO2 injection and formation water The requirement of a seal or cap-rock is therefore
mobilisation. For extensive reservoir units such as the analogous to petroleum reservoirs. Chapter 1 gives
Utsira Sand, pressure is dispersed through a large vol- an overview over different structural and stratigraphic
ume and pressure build-up is minimal, and the traps, and Chap. 12 explains the detailed tectonic and
injectivity is therefore very high. The Tubåen Forma- structural settings of hydrocarbon traps.
tion at Snøhvit, on the other hand, is divided into Although CO2 storage is in many ways analogous
pressure compartments by steep faults with low perme- to hydrocarbon storage, there is one important differ-
ability and hydraulic conductivity, resulting in rapid ence: CO2 is reactive and forms carbonic acid that may
pressure build-up (Pham et al. 2011). Hydraulic con- react with, and change the properties of, the cap-rock
ductivity depends on the permeability of the reservoir (Alemu et al. 2011). The effect of this alteration on an
and the viscosity of the fluid, with high permeabilities intact caprock at ambient temperatures (for most cases
(and low viscosities) favouring water mobilisation and probably at less than 100 C) is however likely very
preventing pressure build-up. limited as the reactions are slow and diffusion of CO2
A storage deposit must in addition to injectivity into a tight shale will be very slow as well (Gaus et al.
have sufficient capacity. If pressure build-up is not 2005; Lu et al. 2009). If CO2 is allowed to migrate into
considered, CO2 storage capacity for saline aquifers fractures or faults that penetrate the seal, then the
can be estimated by the following relationship (see picture is more uncertain, and it is not well understood
CO2 Storage Atlas Norwegian Sea, 2012): if and when reactions will be self-sealing, and when
reactions will enhance flow.
mCO2 ¼ V b ϕ N=G ρCO2 χ (24.1)
where Vb, ϕ, and N/G are the bulk volume, porosity 24.4 The Fluid Properties of CO2
and Net/Gross ratio of the storage reservoir, ρCO2 is the
density of CO2, and χ is the storage efficiency factor. CO2 may occur as a solid, a liquid or a vapour,
We see that injectivity and storage capacity are linked determined by the temperature and pressure
through the reservoir volume and porosity (affecting (Fig. 24.2). Properties such as density and viscosity
permeability; see Chap. 10). The density of CO2 varies vary greatly between these phases. At 304.25 K and
from about 2 kg/m3 at the Earth’s surface (a bit heavier 7.38 MPa, CO2 is at the so-called critical point (Crp),
than air) to more than 500 kg/m3 at conditions deep and above this point CO2 is termed supercritical. At
underground where CO2 is supercritical. The CO2 temperatures above the Crp (but pressure below), CO2
properties are described in more detail in section is a critical gas, whereas if only pressure is above the
24.4. From equation (24.1), we see that the storage Crp, CO2 will be a critical liquid with high density
capacity is proportional to the CO2 density. The final similar to natural gas. The viscosity of scCO2 is about
variable is the storage efficiency factor. This is defined 0.05–0.06 cp at 100 bar and 37 C (Zabaloy et al.
as the fraction of the total reservoir pore space that can 2005) (corresponding to the Utsira storage reservoir),
be filled with scCO2, with values typically being a few slightly lower than for water under the same con-
percent or lower. The most important factor ditions. The storage capacity is a function of CO2
controlling the storage efficiency factor is reservoir density and therefore also of burial depth. At the
heterogeneity. Spatial differences in permeability Earth’s surface, CO2 takes up approximately 22.4 l/
lead to flow of CO2 along preferential fluid pathways, mol gas (about 0.51 l/g). At these conditions, only a
and CO2 is prevented by the capillary entry pressure couple of grams of CO2 could have been stored per
from migrating into fine-grained units. litre pore space. The situation is however very differ-
ent at the higher temperatures and pressures found
underground. The density of CO2 increases with pres-
24.3.3 The Seal (Cap Rock) sure and decreases with temperature in a quite com-
plex way (Fig. 24.3). At low temperatures, below the
As for oil and natural gas, the density of CO2 stored Crp, pressure increases leads to a phase transition from
underground is lower than formation water, so CO2 gas to liquid and a large increase in the CO2 density.
will migrate buoyantly towards the surface. Above the Crp, no such phase transitions occur and
594 H. Hellevang
Fig. 24.4 Mutual solubilities of CO2 and H2O. (a) The amount
Fig. 24.2 CO2 phase stability diagram (s ¼ solid, l ¼ liquid, of CO2 dissolved in the water-rich phase increases with pressure
g ¼ gas). Phase boundaries are given by the solid curves, but decreases with temperature; (b) the amount of H2O
whereas the triple point (T and P where s + l + g are all stable) dissolved in the CO2-rich phase increases with temperature,
and critical point (Crp) are indicated by solid circles whereas pressure first has a positive effect before the solubility
decreases at increasing pressures. Calculations done by R. Miri
(UiO) using the SAFT-EOS approach
reservoirs, which most likely will lie at depths greater

than 1 km, CO2 will have densities between about 200
and 700 g/cm3 compared to >1,000 g/cm3 for brine,
giving a significant buoyancy, i.e. upward force.
To avoid corrosion of pipelines and formation of
hydrates (ice-like particles of water and CO2), CO2
should be injected as a dry phase with as little water as
possible in the stream. However, as soon as CO2 is
injected into a reservoir, CO2 mixes and dissolves in
water and some water dissolves back into the CO2
phase and reacts to form carbonic acid. The mutual
solubilities of CO2 and water depend on temperature
and pressure and on the salinity of the formation water.
Fig. 24.3 CO2 densities as a function of temperature and pres- Although the mutual solubilities are quite limited
sure (from 0 to 70 Mpa and 273 to 453 K). Superimposed are (Fig. 24.4), they have profound effects on the storage
CO2 density curves estimated by three different geothermal
operation both during the injection period (typically
gradients; one being a ‘normal’ North Sea gradient (35 /km)
and two other curves being 10 /km from the normal gradient. lasting for a few decades) and during the subsequent
Density data from the NIST database long-term storage.
During injection, water that is close to the well
pressure increases lead to a steady increase in density. will dissolve into the dry CO2 and the salinity of the
If CO2 is stored underground at normal pressure brine will increase. This may result in the formation
(hydrostatic; see Chap. 10), the geothermal gradient of evaporites, clogging the pore space and reducing
determines the phase properties. Superimposed on the the injectivity. The potential that one pore volume of
density plot in Fig. 24.3 are three geothermal brine has to change the porosity depends on the initial
gradients, one being quite average for the North Sea salinity. If the salinity is 1 mol NaCl/L pore space, a
(35 C/km) and the two others being 10 C/km. We complete dry-out would clog approximately 3% of
see that there are large differences in density, espe- the pore space (assuming a halite molar weight and
cially at shallow depth. It is also clear that CO2 storage density of 58.44 g/mol and 2.17 g/cm3 respectively).
This appears to be quite limited. However, as pores

dry out, additional brine may be brought in by capil-
lary suction and osmotic pressure differences, lead-
ing to further salt precipitation. Dissolution of CO2
into water affects the storage operation on both short
and long time-scales. First, dissolved CO2 reacts with
water and forms carbonic acid, and the pH may drop
to 3 or even lower depending on the buffer capacity
and CO2 pressure. This results in dissolution of the
minerals present in the reservoir and formation of
new secondary phases. In hundreds to thousands of
years, these mineral reactions may form carbonates,
immobilising CO2 for long time spans (see next
section).
A second effect of dissolved CO2 is that it Fig. 24.5 Solubility of CO2 in NaCl-dominated brines with salt
increases the density of the water. The solubility of contents from 0.0 to 5.8 moles per kg water (kgw). Note the
CO2 at 100 bars and 37 C in a solution of normal “salting out effect”, how the solubility decreases with increasing
salt content. Calculations done by R. Miri using the SAFT-EOS
seawater salinity (~0.5 mol NaCl per kg water (kgw)) approach
is approximately 1.1 mol/kgw. At full CO2 satura-
tion, density is about 1,034 kg/m3, which is 1.6%
higher than the solution without CO2 (about
1,018 kg/m3). It has been suggested that the density
contrast between CO2-saturated formation water and
normal brine may result in convection cells, making
the heavier CO2-charged solutions sink downwards,
and bringing ‘fresh’ brine upwards which increases
the dissolution rate of the scCO2 (Elenius and Gasda
2013). Viewed in the light of earlier studies on the
possibility of density currents in stratified and verti-
cally heterogeneous sandstones (Bjørlykke et al.
1988), conclusions on the extent of such mixing
still have to be taken with caution. Moreover, the
solubility of CO2 in water is strongly a function of
salt content. Higher salinities results in less dissolved
CO2 (Fig. 24.5), and less density contrast between
fully saturated CO2-brine and brine with no CO2
(Fig. 24.6). The potential for density currents for Fig. 24.6 Density difference (ΔDensity) defined as the differ-
ence between the density of fully CO2-saturated brines and
high-salinity brines may therefore be even lower. brines without dissolved CO2. The difference is getting lower
and lower as the salinity increases, mainly because of the
salting-out effect leading to decreasing CO2 solubility with
24.5 CO2 Trapping Mechanisms increasing salinity. Calculations done by R. Miri using the
SAFT-EOS approach
Four mechanisms are commonly considered responsi-

ble for keeping CO2 underground after injection: (1) 24.5.1 Structural Trapping
structural trapping; (2) capillary or residual trapping;
(3) solubility trapping; and (4) mineral trapping At short time-scales (during the injection period), most
(Benson and Cole 2008). The risk of leakages and CO2 will be a separate buoyant phase. The vertical
the time dependency of risk are related to these four rock permeability and the density contrast between the
mechanisms (see Fig. 6 in Benson and Cole 2008).
596 H. Hellevang
formation water and scCO2 (see Fig. 24.3) will deter- where x is mole fraction, ϕ and γ are the fugacity and
mine the upward flux of CO2. Chapter 12 and Sect. 1.8 activity coefficients of CO2, and KH is the Henry’s law
provide an overview of structural elements that can coefficient. The molecular CO2 reacts with water to
prevent vertical migration of CO2 or hydrocarbons. form carbonic acid, which dissociates in two steps to
form bicarbonate and carbonate:
24.5.2 Capillary (Residual) Trapping CO2 þ H2 O ¼ H2 CO3

H 2 CO3 ¼ HCO3 þ H þ ; (24.3)
As CO2 flows upwards and laterally below structural
or stratigraphic elements, water must be displaced. HCO3 ¼ CO23 þH þ
Such displacement of a wetting fluid by a non-wetting

fluid is termed drainage. In siliciclastic reservoirs, where the equal sign (‘¼’) denotes equilibrium. Both
mineral grains will be water-wet, and a thin film of the Henry’s law constant and the equilibrium
water will therefore remain on the mineral grains after constants for the carbon speciation are temperature
a CO2 sweep. If water later flows through the area and pressure dependent. As a result of the CO2 disso-
rich in free CO2, some bubbles of CO2 may not be lution, pH drops to 3 or even lower at the CO2
able to overcome the capillary entry pressure of the pressures encountered during CO2 storage.
smaller pore throats, and will be trapped. The latter Because new interfaces between the undissolved
process is termed imbibition. In modelling drainage CO2 and formation water are constantly made through
and imbibition, the relation between the water and the mobilisation of the free-phase CO2, or through
CO2 saturations and the relative permeabilities must density-driven mixing of CO2 and formation water,
be known. The van Genuchten model is one of several more and more CO2 dissolves with time. Dissolution
ways of relating relative permeabilities to CO2 and is therefore favoured in sloping aquifers where the
H2O saturations (van Genuchten 1980). The model is CO2/water interface is large and in heterogenous
empirical and requires experimental data. Values for reservoirs where the plume spreads through a larger
irreducible saturations, which are the minimum satu- volume (Fig. 24.7). If vertical mixing through density-
ration values CO2 and H2O can have, and the shape of driven plume formation can occur (see last part of
the saturation/permeability curves, depend on factors Sect. 24.4), then this will promote further dissolution.
such as sediment sorting and mineral grain size.
24.5.4 Mineral Trapping

24.5.3 Solubility Trapping (CO2 Carbonatisation)
Injected CO2 will first occur primarily as a free phase, The increased acidity of the formation water resulting
but will start to dissolve in the formation water. The from CO2 dissolution increases the solubility of the
rate of dissolution depends on factors such as the CO2/ primary minerals (i.e. minerals in the reservoir prior to
water interface surface area, and diffusion of dissolved CO2 injection), leading to dissolution of these phases
CO2 away from the interface. Solubility trapping will and precipitation of a new stable mineral assemblage.
therefore be most efficient in heterogeneous reservoirs The rate of dissolution varies from mineral to mineral
where the CO2 plume spreads over larger volumes of and is strongly a function of the reservoir temperature.
the reservoir providing more interfacial surface area From the primary phases, the clay minerals, feldspars
between CO2 and formation water. The amount of and carbonates play important roles during CO2 stor-
dissolved molecular CO2 depends on factors such as age. Primary carbonates such as calcite dissolve fast
the CO2 partial pressure, temperature and salinity, and (within days) and transform some of the carbonic acid
can be generally expressed by: to alkalinity (HCO3 ), buffering some of the pH drop
induced by the CO2:
PCO2 ϕ
xCO2 ¼ ; (24.2)
KH γ
CaCO3 þ H 2 CO3 ¼ Ca2þ þ 2HCO3 : (24.4)
Fig. 24.7 Injection of CO2 in the Johansen Formation #2 Harmonic mean averaged model, Kz ¼ 0.082 mD. Including
(modified from Sundal et al. 2014). Plume distributions after the flow baffles leads to significantly larger lateral spread of the
20 years. (a, b) Case #1 includes 15 thin flow baffles. (c, d) Case plume
The pH commonly increases from about 3 to 5 in this (Hellevang and Aagaard 2013, Hellevang et al. 2013).
reaction (Pham et al. 2011). The rates of feldspar and Feldspars release Ca2+ and Na+, supporting calcite,
clay mineral dissolution are many orders of magnitude ankerite and dawsonite (NaAl(OH)2CO3) formation.
slower than for carbonate dissolution, especially at In many quartz-rich reservoirs that are poor in Me2+,
low temperatures. These reactions however play both simulations (e.g. Johnson et al. 2004) and natural
important roles in the immobilisation of CO2 in the observations (e.g. Gao et al. 2009) have suggested that
long term, i.e. mineral trapping of CO2. Clay minerals, dawsonite may be the dominant secondary carbonate
such as chlorites, contain divalent metal cations (Me2+) that forms and traps CO2.
that can react with dissolved CO2 and form secondary Estimates of the mineral-trapping potential of a
carbonates (the inverse of reaction 24.4). The most reservoir require prediction of the following: (1) the
common secondary Me2+-carbonates predicted to amount (volume) of formation water being affected by
form during CO2 storage are ankerites the CO2 injection, which depends on the drainage
(CaFexMg1 x(CO3)2, with x 1), siderite (FeCO3), pathway; (2) the representative average primary and
and in some cases disordered dolomite and magnesite. secondary mineralogy; and (3) the dissolution rates of
The formation of the latter two in numerical simulations the primary minerals and growth rates of secondary
may however be an artifact of the kinetic models being phases. All these points require a good understanding
used overpredicting rates of Mg-carbonate formation of the depositional environment and the diagenetic
598 H. Hellevang
Fig. 24.8 CO2 carbonatisation potential for (a) Utsira Sand; sediments) depends strongly on the initial sediment composi-
and (b) Gulf Coast sediments (modified from Hellevang and tion, and that the range in predictions using different kinetic
Aagaard 2013). The figure shows that the total amount of models is very large. ‘Alpha’ here denotes fraction of the min-
carbonates that can form (~10 times more in Gulf Coast eral surface area being reactive (see Hellevang et al. 2013)
overprint. In the kinetic models, reactive surface areas 1,030–1,100 kg/m3) is used to map the extent of the
are commonly regarded as a highly uncertain parame- CO2 plume. This method has been very successful in
ter. Knowledge on how much of the surfaces of reac- monitoring the CO2 plume migration for the Sleipner
tive minerals (e.g. feldspars and chlorites) are in injection (Eiken et al. 2011). Microseismicity caused
contact with the formation water (provides the specific by fracture or fault activation may also provide valu-
surface area as m2/L porewater) is therefore essential. able information on the effect of pressure build-up and
Figure 24.8 shows predictions of the long-term fluid migration in the sub-surface. The status of CO2
trapping potential for two reservoirs, the Utsira Sand, storage operations is obtained indirectly by monitoring
and a representative of a feldspar-rich Gulf Coast the reservoir pressure. If the pressure approaches the
sediment (Frio Formation, Lower Tertiary). The figure estimated cap-rock fracture pressure, preventative
shows how reducing the reactive surface area by one measures have to be taken. Such measures can be
order of magnitude (parameter α) has a large impact reduction of the injection rate or water production.
on the predictions, and that the three different kinetic However, in some cases, such as for the Tubåen For-
models used also lead to a large range in predicted mation at Snøhvit, the storage capacity is very limited
carbonatisation. Because of the uncertainties in the and alternative storage reservoirs have to be found. In
kinetic models, there is at present a focus to improve most cases, a combination of seismic monitoring and
both the kinetic models and the parameters used in the measurements of the well pressures is employed.
models.
24.7 Examples of CO2 Storage Projects

24.6 How Can We Know that CO2 Is
Safely Stored? There are today (October 2013) 12 CCS projects that
are operational, but a large number of others are in the
To ensure that the CO2 is safely stored underground, planning and executive phases (Global CCS Institute).
various monitoring technologies are in use, with seis- One of the projects that has received most attention is
mic and well-pressure monitoring being two important at Sleipner (North Sea), being the first industrial-scale
methods. Seismic monitoring is built on comparing dedicated CCS project started in 1996. Some projects
seismic signals in time and space, so called time- such as In Salah, Algeria, were running until recently,
lapse or 4D seismics (see Chap. 19). By doing so, the but that project ended after illustrating how several
difference in the density of supercritical CO2 (approx- monitoring technologies can be used to successfully
imately 400–800 kg/m3) and water (approximately track the fate of stored CO2. Below is a short overview
of the Sleipner, Snøhvit (Barents Sea), In Salah, and the seismic images is the presence of a chimney-like
Weyburn-Midale (Saskatchewan, Canada) CCS structure in the storage reservoir just above the injector.
projects. This structure possibly formed shortly after injection
commenced, by liquefaction of the sediments breaking
up the thin intra-reservoir shale layers (Hermanrud et al.
24.7.1 Sleipner CCS Project 2009). Although CO2 has migrated through what was
thought to be a tight shale layer in the cap-rock and into
Natural gas produced at Sleipner contains about 9.5% an overlaying sand wedge, there are no indications of
CO2. In order to ship this natural gas in a liquid form further leakage to shallower depths.
(LNG) to customers it has to be cooled to 162 C,
which may cause formation of CO2 ice aggregates. To
avoid this, most of the CO2 is captured during produc- 24.7.2 Snøhvit CCS Project
tion, i.e. removed from the natural gas stream. Before
1996, CO2 captured from the Sleipner field was Gas produced from the Lower to Middle Jurassic Stø
released directly to the atmosphere. A few years Formation at Snøhvit contains 5–8% CO2. This gas is
before, in 1991, Norway had established a tax on transported by pipelines to the processing plant at
CO2 emissions (in 1999, about 18 US$/ton CO2 on Melkøya, where CO2 is captured and transported back
emissions of ~1 Mt CO2/year). This represented a underground. CO2 storage started in 2008 with injection
considerable financial levy on the operators of into the early Jurassic Tubåen Formation, which lies
Sleipner. In 1996, Statoil and partners started to inject just below the gas-producing Stø Formation In contrast
CO2 captured at Sleipner into the Utsira Formation, to the Utsira Sand, the stratigraphic sequences in the
and the injection has continued successfully at a rate of Barents Sea region have experienced a complex history
approximately 1 Mt CO2/year. of burial and later tectonic uplift. The present-day depth
The CCS operation at Sleipner was feasible of the Tubåen is about 2,400 m at the injection site, and
because of the properties of the Utsira storage reser- temperature and pressure before injection were 95 C
voir located near the Sleipner field. The Utsira Sand (Hansen et al. 2013) and 240 bars (assuming hydro-
sediments are basin-restricted marine lowstand static). The depth and temperature at maximum burial
deposits of Miocene-Pliocene age (Zweigel et al. may however have been more than 3,200 m and 130 C.
2004), only weakly compacted and with high porosity After a short period of injection, the operation experi-
(30–40%). The sand is very extensive (see Fig. 24.1) enced problems with rapid pressure build-up,
and is in many ways an optimal storage reservoir. At interpreted to be caused by reactions clogging the
800 m below seafloor, temperatures and pressures are pore space and reducing the permeability around the
about 37 C and 100 bar, which is above the critical well (Hansen et al. 2013). In the Tubåen Formation,
point of CO2. As overburden, the Utsira Formation has these reactions could be: (1) salt formation (scale)
a 250 m clay-rich sequence of the Nordland group caused by the drying-out of highly saline brine around
which provides the seal, with a further several hundred the well; or (2) CO2 hydrate (clathrate) formation. As
metres of coarser-grained Quaternary sediments above CO2 entered the reservoir at low temperature (4 C) and
(Zweigel et al. 2004). The mudstones deform in a pressures were higher than 290 bars (Hansen et al.
ductile manner and stress changes are therefore not 2013), the latter is more likely (Kvamme and Tanaka
expected to lead to fractures that could provide leak- 1995, Duan and Sun 2006), although the highly saline
age pathways for the CO2. brine depresses the freezing point of CO2 hydrate (Duan
The injected CO2 has been monitored by 3D and Sun 2006). Salt (halite) formation caused by injec-
seismics surveys carried out before, and several times tion of dry CO2 has been predicted numerically (Muller
after, injection started in 1996. The seismic images et al. 2009) and in laboratory experiments (Bacci et al.
show that CO2 has spread out laterally below nine 2013), and although not yet observed in situ, could also
distinct stratigraphic layers. A bulk of the CO2 is still very well be the reason for the pressure build-up. After
trapped under these layers and this has reduced the the initial pressure build-up was solved by chemical
amount of free CO2 that has reached the cap rock treatment, the operation continued. Unfortunately, the
(Hermanrud et al. 2009). A much discussed feature of tight and compartmentalised Tubåen Formation turned
600 H. Hellevang
out to have only a limited storage capacity, and pressure capacity of the Weyburn-Midale reservoirs, including
continued to increase until injection was abandoned in both the EOR phase and later CO2 storage, is
2011. Injection was then moved to the shallower Stø estimated to be between 50 and 60 Mt. The
Formation, below the gas reserves. As CO2 migrates Weyburn-Midale CO2 monitoring and storage project,
updip and will mix with the natural gas, there is an supported by the International Energy Agency Green-
expectance that additional storage sites will be needed house Gas R&D program (IEA GHG), has invested in
in the future to continue the CO2 storage operation. an extensive geophysical and geochemical monitoring
programme, including passive seismic and soil gas and
fluid geochemistry monitoring (Whittaker et al. 2011).
24.7.3 In Salah CCS Project
The In Salah CCS project in central Algeria is an 24.8 What Is the Future of CCS?
example of an onshore CCS operation. During the
lifetime of the project, between 2004 and 2011, 3.8 The main obstacles to implementing CCS on a large
Mt CO2 was stored in a Carboniferous sandstone unit scale today are high costs and, for many regions, a lack
1.9 km below the surface. The project demonstrated a of a market for CO2. In North America, CO2 has been
large number of monitoring technologies, such as the used for enhanced oil recovery for decades, and this
use of Interferometric Synthetic Aperture Radar illustrates that the costs of capture and transport can be
(InSar) to monitor surface heave, 4D seismics to more than compensated if a market exists to sell the
track the underground CO2 migration, geophones to CO2. A second challenge is that the cost of releasing
detect microseismicity, well-head sampling of geo- CO2 to the atmosphere is still too low to make it
chemical tracers, etc. (Ringrose et al. 2013 and financially viable for most emitters to store CO2.
references therein). The combination of all these Below is a short review of present-day activities to
technologies enabled the tracking both of possible utilise and store CO2 in the USA, and a suggested
hydrofractures and potential migration of the CO2 solution to lower the cost and allow large-scale storage
outside the Krechba hydrocarbon lease, and remedial by gathering emissions from many point sources into
action (operation and monitoring responses). The In one shared pipeline network and storage site.
Salah CCS project has therefore provided highly valu-
able information on the whole chain, from planning to
operation and abandonment. 24.8.1 CCUS/EOR
CCUS (and CCUS/EOR), is short for Carbon Capture

24.7.4 Weyburn-Midale CCS-EOR Project Use (or Utilization) and Storage, and Enhanced Oil
Recovery (EOR). By using captured CO2 in industrial
The Weyburn-Midale project is at present the world’s processes such as EOR, CCS may become profitable.
largest CCS operation, and is an example of a CCS There are at present nine CCUS/EOR projects in the
Enhanced Oil Recovery (EOR) project. The CCS operational phase and several in the planning and exec-
operation started in 2000, and currently nearly 3 Mt utive stage, most of them in the USA (Global CCS
CO2 are annually injected in multiple wells and used Institute, Oct. 2013). Of these, the largest-scale is the
for enhanced oil recovery (Whittaker et al. 2011). At Weyburn-Midale CCS-EOR project described above.
the planned end of the operation, in 2035, the amount Onshore CCUS/EOR has been a great success. The
of stored CO2 is expected to be more than 40 Mt. The same is not yet true for offshore CCUS/EOR. There are
injected CO2 is a byproduct of coal gasification at the several reasons for this: (1) seawater is readily available
Great Plains Synfuels Plant in North Dakota, USA. offshore and can be used instead of the less available
The target reservoirs are of lower Carboniferous CO2; (2) the costs of installing CO2 capture facilities
age, at a present-day depth of 1,300–1,500 m. The may be considerable; (3) injected CO2 can lead to
rocks were deposited in a peritidal sequence, with corrosion of already existing infrastructure (wells etc.)
carbonates deposited at highstands and evaporitic and investment in new infrastructure (in addition to the
dolomites formed at lowstands. The total storage capture facilities) may be required; (4) stored CO2 may
be produced back along with the oil. Whether CCS/ 24.9 Summary
EOR will be implemented offshore on a larger scale in
the future depends on the cost-benefit picture and the Carbon capture and storage is one out of several
availability of CO2. There are limits to the degree that strategies to mitigate rapidly increasing atmospheric
water can enhance oil recovery, and CO2 is known to CO2 levels. CCS involves capture of CO2 from natural
further increase the amount of oil that can be extracted gas or flue gas, and transport and storage in geological
and thus extend the lifetime of oil fields. Moreover, new formations. There are several technologies available for
technologies using CO2 for EOR are being developed. capture, and the choice of technology depends on the
One recent discovery is that CO2 foams (rather than gas processing plant and the cost of operation. Because of
or supercritical) can significantly improve the recovery the relatively high costs of capture, there are large
of medium to heavy oils and that foams are twice as research programmes underway to improve the
effective compared to water or gas/supercritical CO2 energy-efficincy of current methods and to come up
(Rassenfoss 2012). with new and more cost-effective capture technologies.
The captured CO2 can be stored underground in geo-
logical formations, preferentially as a dense supercriti-
24.8.2 Large Scale Transport and Storage cal phase, but this also requires large amounts of
energy. The main requirements of a storage formation
Most of the discussion so far has been on capture and are that it has sufficient injectivity and storage capacity
storage from single point sources. This limits CCS and a tight cap rock. Moreover, as water is mobilised by
projects to a few Mt/a. The total emissions from point the injected CO2 and the pressure waves caused by the
sources in larger regions in Europe, Asia, and America injection, one has to ensure that freshwater resources
may however reach hundreds of Mt/a. By connecting are not affected by the storage operation. In the future,
smaller (<20–40 Mt) CO2 suppliers through hubs, and to meet the IPCC target of globally storing Gt of CO2
transporting large amounts (hundreds of Mt/a) to per year rather than the few Mt/a stored today, storage
reservoirs such as the Utsira Sand with very large stor- operations will have to be scaled up. One way to do this
age capacity, the cost of stored CO2 per ton may be is to collect CO2 from several large point sources and
reduced significantly. The cost of pipelines and infra- store it in sandstones, such as the Utsira Sand, that
structure is shared among several stakeholders, and possess a very high storage capacity (Fig. 24.1).
larger dimensions of pipelines allow a greater volume
CO2 to be transported per ton steel used (volume/cross-
section ratio of pipelines increases). The amount of
pressure a pipeline can hold is a function of the diameter
of the pipeline, with smaller pipes withstanding higher References
pressures. The upper limit of pipeline dimensions must
therefore be weighed against the pressures required to Alemu, B.L., Aagaard, P., Muntz, I.A. and Skurtveit, E. 2011.
transport the CO2 as a dense phase. Caprock interaction with CO2: A laboratory study of reac-
tivity of shale with supercritical CO2 and brine. Applied
To allow such large-scale CCS in Europe requires
Geochemistry 26(12), 1975–1989.
regulations across national boundaries, and there are Bacci, G., Durucan, S. and Korre, A. 2013. Experimental and
several complicated issues that must be sorted out. numerical study of the effects of halite scaling on injectivity
One is the ownership and risk responsibility of the and seal performance during CO2 injection in saline aquifers.
Energy Procedia 37, 3275–3282.
transport and storage of a mixture of CO2 from several
Benson, S.M. and Cole, D.R. 2008. CO2 sequestration in deep
sources. Who will be held responsible in case of a sedimentary formations. Elements 4, 325–331.
pipeline rupture? Who owns the stored CO2 and will Bjørlykke, K., Mo, A. and Palm, E. 1988. Modelling of thermal
be responsible for a monitoring programme? It is convection in sedimentary basins and its relevance to diage-
netic reactions. Marine and Petroleum Geology 5, 338–351.
natural that in Europe the European Union and
Duan, Z. and Sun, R. 2006. A model to predict phase equilib-
associated nations such as Norway, should facilitate rium of CH4 and CO2 clathrate hydrate in aqueous electro-
such large scale CCS. lyte solutions. American Mineralogist 91, 1346–1354.
602 H. Hellevang
Eiken, O., Ringrose, P., Hermanrud, C., Nazarian, B., Torp, T. sequestration partitioning. In: Bains, S.J. and Worden, R.H.
A. and Høier, L. 2011. Lessons learned from 14 years of (eds.), Geological Storage of Carbon Dioxide. Geological
CCS operations: Sleipner, In Salah and Snøhvit. Energy Society Special Publication 233, pp. 107–128.
Procedia 4, 5541–5548. Kvamme, B. and Tanaka, H. 1995. Thermodynamic stability of
Elenius, M.T. and Gasda, S.E. 2013. Convective mixing in hydrates for ethane, ethylene, and carbon dioxide. Journal of
formations with horizontal barriers. Advances in Water Physical Chemistry 99, 7114–7119.
Resources 62, 499–510. Lu, J., Wilkinson, M., Haszeldine, R.S. and Fallick, A.E. 2009.
Figueroa, J.D., Fout, T., Plasynski, S., McIlvried, H. and Long-term performance of a mudrock seal in natural CO2
Srivastava, R.D. 2008. Advances in CO2 capture technol- storage. Geology 37, 35–38.
ogy—The U.S. Department of Energy’s carbon sequestra- Muller, N., Qi, R., Mackie, E., Pruess, K. and Blunt, M.J. 2009.
tion program. International Journal of Greenhouse Gas CO2 injection impairment due to halite precipitation. Energy
Control 2, 9–20. Procedia 1, 3507–3514.
Gao, Y., Liu, L. and Hu, W. 2009. Petrology and isotopic geo- Pham, T.H.V., Maast, T.E., Hellevang, H. and Aagaard, P. 2011.
chemistry of dawsonite-bearing sandstones in Hailaer basin, Numerical modeling including hysteresis properties for CO2
northeastern China. Applied Geochemistry 24, 1724–1738. storage in Tubåen Formation, Snøhvit field, Barents Sea.
Gaus, I., Azaroual, M. and Czernichowski-Lauriol, I. 2005. Energy Procedia 4, 3746–3753.
Reactive transport modelling of the impact of CO2 injection Rassenfoss, S. 2012. Increased oil production with something
on the clayey cap rock at Sleipner (North Sea). Chemical old, something new. Journal of Petroleum Technology, Oct.
Geology 217, 319–337. issue, 36–39.
Global CCS Institute. http://www.globalccsinstitute.com. Ringrose, P.S., Mathieson, A.S., Wright, I.W., Selama, F.,
Hansen, O., Gilding, D., Nazarian, B., Osdal, B., Ringrose, P., Hansen, O., Bissell, R., et al. 2013. The In Salah CO2 storage
Kristoffersen, J.-B., et al. 2013. Snøhvit: The history of project: lessons learned and knowledge transfer. Energy
injection and storing 1 Mt CO2 in the fluvial Tubåen Fm. Procedia 37, 6226–6236.
Energy Procedia 37, 3565–3573. Sundal, A., Hellevang, H., Miri, R. and Aagaard, P. 2014. On
Hellevang, H. and Aagaard, P. 2013. Can the long-term poten- layer specific CO2 plume distributions and variability in min-
tial for carbonatization and safe long-term CO2 storage in eralization potential. Submitted to the Fourth EAGE CO2
sedimentary formations be predicted? Applied Geochemis- Geological Storage Workshop, Stavanger, April 23–25, 2014.
try 39, 108–118. van Genuchten, M.Th. 1980. A closed-form equation for
Hellevang, H., Pham, V.H. and Aagaard, P. 2013. Kinetic predicting the hydraulic conductivity of unsaturated soils.
modeling of CO2-water-rock interactions. International Jour- Soil Science Society of America Journal 44, 892–898.
nal of Greenhouse Gas Control 15, 3–15. Whittaker, S., Rostron, B., Hawkes, C., Gardner, C., White, D.,
Hermanrud, C., Andresen, T., Eiken, O., Hansen, H., Janbu, A., Johnson, J., et al. 2011. A decade of CO2 injection into
Lippard, J., et al. 2009. Storage of CO2 in saline aquifers— depleting oil fields: monitoring and research activities of
lessons learned from 10 years of injection into the Utsira the IEA GHG Weyburn-Midale CO2 monitoring and storage
Formation in the Sleipner area. Energy Procedia 1, 1997–2004. project. Energy Procedia 4, 6069–6076.
IPCC. 2013. Climate change 2013: The physical science basis. In: Zabaloy, M.S., Vasquez, V.R. and Macedo, E.A. 2005. Viscos-
Stocker, T.F., Quin, D., Plattner, G.-K., Tignor, M., Allen, ity of pure supercritical fluids. Journal of Supercritical Fluids
S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V. and Midgley, 36, 106–117.
P.M. (eds.), Contribution of Working Group I to the Fifth Zweigel, P., Arts, R., Lothe, A.E. and Lindeberg, E.G. 2004.
Assessment Report of the Intergovernmental Panel on Climate Reservoir geology of the Utsira Formation at the first
Change. Cambridge: Cambridge University Press, 1535 pp. industrial-scale underground CO2 storage site (Sleipner
Johnson, J.W., Nitao, J.J. and Knauss, K.G. 2004. Reactive area, North Sea). In: Baines, S.J. and Worden, R.H. (eds.),
transport modeling of CO2 storage in saline aquifers to Geological Storage of Carbon Dioxide. Geological Society
elucidate fundamental processes, trapping mechanisms and Special Publication 233, pp. 165–180.
Chapter 25
Geology of the Norwegian Continental Shelf
Jan Inge Faleide, Knut Bjørlykke and Roy H. Gabrielsen
25.1 Introduction 25.1.1 Regional Geological Setting
In the preceding chapters we have included only a few The Norwegian continental shelf comprises three main
regional examples and case studies because of space provinces (Figs. 25.1 and 25.2):
limitations. The present chapter will, however, provide • North Sea
some examples. The North Sea and other parts of the • Mid-Norwegian continental margin
Norwegian continental shelf contain several different • Western Barents Sea
petroleum provinces which can illustrate some of the Prior to continental break-up and the onset of sea-
general principles of petroleum geology and geophys- floor spreading in the Norwegian-Greenland Sea
ics. The geological evolution of these sedimentary (Fig. 25.3) these provinces were part of a much larger
basins provides a necessary background to understand epicontinental sea lying between the continental
the distribution of source rocks and the timing of petro- masses of Fennoscandia, Svalbard and Greenland.
leum migration. The structural history of rifted basins, Thus there are many similarities in the stratigraphy
passive margins and also uplifted basins such as the (Fig. 25.4) and geological evolution of the various
Barents Sea is critical to the trapping of oil and gas. provinces but also important differences – in particular
These basins are very well documented by seismic and during Cretaceous-Cenozoic times.
well data. The Norwegian Petroleum Directorate also The sedimentary basins at the conjugate continental
provide information about exploration and production margins off Norway and Greenland and the adjacent
on their web pages (http://www.npd.no). shallow seas, the North Sea and the western Barents
The Norwegian continental shelf is unique with Sea, developed as a result of a series of post-Caledonian
respect to the amount of information available about rift episodes until early Cenozoic time, when complete
petroleum exploration and production and also geo- continental separation took place. The Norwegian Mar-
logical and geophysical information and maps (http:// gin comprises the mainly rifted volcanic margin off-
www.npd.no/en/Topics/Geology/). You will there find shore mid-Norway (62–70 N) and the mainly sheared
updated information about exploration and production margin along the western Barents Sea and Svalbard
and also references to publications and reports. This is (70–82 N) (Fig. 25.2). Physiographically, the Norwe-
a supplement to this book. Furthermore, updated gian margin consists of a continental shelf and slope that
lithostratigraphic tables from the Norwegain Offshore vary considerably in width and morphology (Fig. 25.1).
are available from NPD (http://www.npd.no/en/
Topics/Geology/Lithostratigraphy/).
25.1.2 Petroleum Exploration
J.I. Faleide (*) K. Bjørlykke R.H. Gabrielsen
When gas was found onshore in the Groningen Field in
Department of Geosciences University of Oslo, Oslo, Norway
e-mail: [email protected]; [email protected]; the Netherlands in 1958 it soon generated interest in
[email protected] the North Sea itself. However, it was difficult to say
604 J.I. Faleide et al.
northeast England and Denmark were assumed to con-

EB
tinue beneath at least parts of the North Sea, and Car-
°
80
boniferous coal seams and Permian sediments were
already well known in northeastern England. There
were also a few indications that younger sedimentary
rocks were present beyond the Norwegian coast. The
moraines on Jæren contained small amounts of proba-
75° bly Cretaceous material, there were erratic blocks of
Jurassic age on Froan and pieces of coal from Tun in
Barents
Verran, off Trøndelag. Much farther north, the fairly
Sea thin Jurassic and Cretaceous sequences on Andøya had
long been known. It was difficult, though, to know how
Norwegian-
Greenland much reliance could be placed on these scattered
Sea 70°
observations; not least, one had no idea of how thick
such Mesozoic sediments would be.
lt was difficult to know if the equivalents of the well
VP
known source rock, the Upper Jurassic Kimmeridge
Clay extended into the Norwegian Shelf and if it was
65° buried deeply enough for it to be mature.
A confirmation of the presence of thick sediment
sequences on the shelf was obtained in seismic refraction
studies in the Barents Sea and the central North Sea in
the late 1950s, indicating accumulations in the order of
60° several kilometres. Seismic refraction studies supported
by potential field data in the Skagerrak, performed in
1963–1964, confirmed the presence of more than 3 km
of sediments also in this area. The early seismic data that
North
Sea was shot in the North Sea had very poor quality and it
55° was not until drilling on the Norwegian shelf
commenced in 1966 that it was possible to have an
informed opinion on the potential for oil and gas.
–5° 0° 5° 10° 15° 20° The Norwegian continental shelf, extending from
the baseline to the limit approved by the UN Commis-
Fig. 25.1 Regional setting (bathymetry/topography) of the sion on the Limits of the Continental Shelf, amounts to
Norwegian Continental Shelf and adjacent areas. EB ¼ Eurasia 2.2 million km2. About half of this acreage has bedrock
Basin, VP ¼ Vøring Plateau
in which petroleum may be found, and half of that has
been opened for petroleum exploration. The first
what lay off the Norwegian coast beneath the Quater- licences were awarded in 1965 and the first commercial
nary sediments. Moraine material which had been discovery in the Norwegian part of the North Sea (block
recovered from the seabed during marine geological 2/4) was made in 1969. The motivation for this drilling
investigations in the 1950s contained clasts that almost was to test a possible Permian (Rotliegendes sandstone)
without exception were basement rocks. reservoir. Instead, however, an unexpected discovery of
The pre-Quaternary geology of the Norwegian Con- hydrocarbons was made in a domal structure in Upper
tinental Shelf was virtually unknown prior to the early Cretaceous Chalk. Continued exploration on this play
1960s, as there had been no seismic measurements and type resulted in a number of discoveries in the Ekofisk
few other indications of what was beneath the cover of area of the central North Sea.
Quaternary, partly glacial, sediments. Before any seis- Large fields in the North Sea, such as Sleipner,
mic or borehole data were obtained in the North Sea, Statfjord and Gullfaks, were discovered during the
one had to extrapolate from the geology of the first 10 years after the Ekofisk Field was proven, and
surrounding land areas. The Mesozoic rocks of all four are still in production. Most of the other large
25 Geology of the Norwegian Continental Shelf 605
°
Eurasia
80
Basin
Greenland
°
Barents 78
Sea
°
76
Svalbard
74°
72°
Norwegian- 70°
Greenland
Sea
68°
JMMC
66°
64°
Finland 62°
Norway
60°
Sweden
58°
56°
North Denmark
Late Cretaceous - Sea
Palaeocene 54°
Late Jurassic - UK
Early Cretaceous
Late Palaeozoic
–5° 0° 5° 10° 15° 20°
Fig. 25.2 Main structural elements of the Norwegian Conti- Faleide et al. 2008). For more details see close-up maps in
nental Shelf and adjacent areas related to different rift phases Figs. 25.5, 25.10 and 25.14. JMMC ¼ Jan Mayen micro-
affecting the NE Atlantic region (modified/updated from continent
fields on the shelf were found between 1979 and 1984 are more complex, and even today important aspects
(see Sect. 25.2.3 for more details on the North Sea of the petroleum geology remains enigmatic – partly
exploration history). due to geological complexities.
Following the exploration success in Norwegian Large parts of the Norwegian continental shelf
waters south of 62 N, the areas to the north were are now well explored. Better knowledge of the
opened for exploration in 1979. Results and experi- geology and advances in technology lead to a higher
ence gained from the North Sea were utilised when discovery rate, but the finds have mostly been small.
areas further north were opened, and areas with a Consequently, the growth in resources has been rel-
similar geological setting to the northern North Sea atively low for the last 20 years. The largest discov-
were tested first (Haltenbanken offshore Mid-Norway ery made in this period was Ormen Lange in the
and Tromsøflaket in the southwestern Barents Sea). Norwegian Sea in 1997. Recent discoveries in the
The exploration histories of the areas north of 62 N North Sea (Edvard Grieg and Johan Sverdrup) and
a Present b Chron 13 (c. 33 Ma) c Breakup (c. 55 Ma)
°
˚
80
80
80
Eure
ka
(not n Oroge
resto n
red) y
75° 75° 75°
70° 70° 70°
MC
JM
C
M
JM
65° 65° 65°
60° 60° 60°
55° 55° 55°
–5° 0° 5° 10° 15° 20° –5° 0° 5° 10° 15° 20° –5° 0° 5° 10° 15° 20°
Fig. 25.3 Plate tectonic reconstructions of the NE Atlantic. (c) Reconstruction to time of breakup (c.55 Ma) based on
(a) Present (same as Fig. 25.2), (b) Reconstruction to chron 13 unpublished work of Faleide et al.
(c.33 Ma) using the rotation pole of Gaina et al. 2009,
Barents Sea (Johan Castberg and Gohta/Alta) have 25.2 North Sea
added considerable oil resources which will contrib-
ute to substantial production in years to come. 25.2.1 Structure
Technological advances have led to development
of discoveries in areas of great water depth and far The North Sea is an example of an intracratonic
from shore. The technological advances have also basin; that is to say a basin which lies on continental
made a higher proportion of the resources profitable crust. The prerequisite for forming major sedimen-
to recover. On average, fields on the Norwegian shelf tary basins on continental crust is that the crust (and
have a recovery factor of 46% for oil. This is high mantle lithosphere) is thinned, resulting in subsi-
compared with oil provinces in other parts of the dence to maintain isostatic equilibrium. The North
world. Continuous research and development of tech- Sea has been subjected to periods of stretching/thin-
nology are required to raise this further. ning and subsidence during late Carboniferous,
Significant volumes remain to be found and pro- Permian-Early Triassic and Late Jurassic times.
duced on the Norwegian continental shelf. About Each rift phase was followed by a thermal cooling
40–50% of the estimated oil and gas reserves have so stage, characterised by regional subsidence in the
far been produced. The reserves will depend on future basin areas.
oil and gas prices and on environmental restrictions. It The Northern North Sea province is dominated by
may still be possible to make large discoveries in less the Viking Graben, which continues into the Sogn
explored areas, such as in deep waters in the Norwegian Graben towards the north (Fig. 25.5). These grabens
Sea, in the Barents Sea and in areas that are not yet are flanked by the East Shetland Basin and the Tampen
open. Oil production reached its peak in 2000–2001, but Spur to the west, and the Horda Platform to the east
gas production is still increasing. For more facts about (Figs. 25.6 and 25.7). These are Jurassic-Cretaceous
the petroleum activity on the Norwegian shelf see http:// features, and the main crustal thinning took place in
www.npd.no/en/Publications/Resource-Reports/2014/. the late Middle to Late Jurassic, followed by thermal
Southern Norwegian- Southern Tampen Trøndelag Halten Vøring East Tromsø Nordkapp B., SE Barents
Platform Svalbard
North Sea Danish Basin Viking Graben Spur Terrace Basin Greenland Basin Bjarmeland P. Basin
Quat
Pleist
Neogene
Plio
Mio
Tertiary
Oligo
Palaeogene
Eoc
Palaeo
Maast
Camp Utsira high Kviting
U Sant Chalk Group
Coni
Tur
Cretaceous
Cen
Alb
Apt Helvetiafjellet
L Barr
Haut ? ? Klippfisk
Valang
Ryaz
Volg Mandal Spekk Hekkingen Janusfjellet Hekkingen
Brae Draupne Rogn ?
Tau
Ula
U Kimm
Oxf Heather
klø rde -
Call Sandnes Hugin Melke
ft
Va
Jurassic
Bath Garn
Brent
M Bajoc Stø Condensed Stø Nordmela eq?

Bryne Sleipner sequence
Aalen Ile
?
Drake Tofte
Toarc
Kl ill
er
Ne
int
Plienc Cook Tilje ? Condensed
L Sinem Burton Kap
sequence
Nordmela
Åre
Hett Stewart
d
jor
Tubåen
atf
Rhæt ?
St
U Nor
Triassic
Carn
Ladin ? Botneheia
M Anis
L Scyth
ek ik-
Tatar ?
cre ldv
U Kazan
Fo
Permian
Kung
?
L Artin
Sakm
Assel ? ?
?
Gzel
Steph
U Kasim ? ? ? ?
Carboniferous
?
WF M Moscov ?
Bash
Nam
Serpuk ?
L Vise ?
Din
Tour ? ?
Marginal evaporite deposits, sabkha

Continental clastics, unspecified Coastal-, delta- and flood plain deposits Shallow marine carbonates Spiculite
Continental clastics, mainly sands Marginal marine and shallow marine deposits, mainly sands Volcanic deposits Coal
Continental clastics, mainly shales and silts Shallow marine deposits, mainly shales Carbonaceous shales Coarse clastics
Halite Deeper marine deposits, mainly shales Clastic mixing in carbonates, Source rocks
sands in shales
Fig. 25.4 Lithostratigraphic summary for the Norwegian Continental Shelf and adjacent areas (modified from Brekke et al. 2001)
subsidence and sediment loading in the Cretaceous. The Norwegian Central North Sea Province
However, the Viking Graben and its margins are encompasses the northwestern part of the Central Gra-
underlain by an older major rift basin of assumed ben (Figs. 25.5, 25.8 and 25.9), which is mainly a
Permian-Early Triassic age. The axis of this rift sys- Jurassic-Cretaceous structure, and its northeastern
tem is thought to lie beneath the present Horda Plat- margin. The strata of the Central Graben area were
form. It is bounded by the East Shetland Platform in affected by halokinesis already in the Triassic, and
the west and the Øygarden Fault Zone in the east. major structuring accordingly occurred already in
Structures within this area are characterised by large pre-Jurassic times. However, salt movements have
rotated fault blocks with sedimentary basins in asym- locally continued into the Tertiary. Jurassic rifting
metric half-grabens associated with lithospheric and generation of large, rotated fault blocks resulted
extension and crustal thinning (Fig. 25.6). The area in extreme erosion in places. The complex structural
was presumably also strongly affected by post- pattern established during this stage was further com-
orogenic (post-Caledonian) extension in Middle to plicated by Cretaceous inversion. The Norwegian-
Late Devonian times. Danish Basin, east of the Central Graben, also contains
1 gB rB SG
M M
TS
2
ESB
60
VG
ESP HP
SB OG
58 UH
ÅG
WG
SH 4
SkG
5
FB
NDB
56
CG STZ
6
RFH HG
MNSH
3
54
Late Cretaceous -
Palaeocene
Late Jurassic -
Early Cretaceous
Late Palaeozoic
–5 0 5 10
Fig. 25.5 Main structural elements in the North Sea and adja- MNSH ¼ Mid North Sea High, MrB ¼ Marulk Basin, NDB ¼
cent areas (close-up of Fig. 25.2 – modified/updated from Norwegian-Danish Basin, OG ¼ Oslo Graben, RFH ¼
Faleide et al. 2008). Location of interpreted regional profiles Ringkøbing-Fyn High, SB ¼ Stord Basin, SG ¼ Sogn Graben,
(black), crustal transects (red) and seismic examples (blue) SH ¼ Sele High, SkG ¼ Skagerrak Graben, STZ ¼ Sorgenfrei-
shown in Figs. 25.4–25.9. CG ¼ Central Graben, ESB ¼ East Tornquist Zone, TS ¼ Tampen Spur, UH ¼ Utsira High, VG ¼
Shetland Basin, ESP ¼ East Shetland Platform, HG ¼ Horn Viking Graben, WG ¼ Witchground Graben, ÅG ¼ Åsta
Graben, HP ¼ Horda Platform, MgB ¼ Magnus Basin, Graben
many salt structures but has not been affected by The offshore part of the Oslo Rift in the Skagerrak
significant rifting (Figs. 25.8 and 25.9). is characterised by NE-SW striking half-grabens
The geometric shape of the sedimentary basin is (Fig. 25.9). Cross-sections across the Skagerrak show
influenced by the structure in the underlying rocks tilted half-grabens filled with down-faulted Upper
(basement) and by the thickness of the continental Carboniferous-Lower Permian and Lower Palaeozoic
crust. Most of the North Sea is underlain by a strata that are unconformably overlain by Triassic
Caledonian basement. Long-lived zones of weakness sedimentary rocks. Although a thick succession of
inherited from the Caledonian Orogeny played a role Lower Palaeozoic strata has been preserved in the
in the later evolution of the North Sea basin. The Skagerrak Graben, a substantial part of this pre-rift
southeastern North Sea and Skagerrak are underlain unit has been eroded in the central part of the rift.
by a Precambrian basement covered by a Lower From strata preserved in the onshore Oslo Graben we
Palaeozoic sedimentary succession. In the south the know that the Lower Palaeozoic succession consists
North Sea basin is bounded by the Hercynian mainly of Cambrian-Ordovician and Silurian platform
(Variscan) mountain range which runs E-W through series and Upper Silurian-lowermost Devonian clastic
Germany, northern France and southwest England. sedimentary rocks, with the latter being deposited in
The contraction occurred during the Carbonifererous- the Caledonian foreland basin. In the western part of
Permian. Uplift in this area resulted in vast amounts of the Skagerrak, the late Palaeozoic structures of the
sediment being deposited in the area to the north and Oslo Rift link up with the Sorgenfrei-Tornquist Zone
this initiated the formation of the North Sea basin. (also termed the Fennoscandian Border Zone in
Magnus E. Shetland Tampen Viking Horda Øygarden

NW Basin Basin Spur Graben Platform Fault Zone SE
0
s twt
10
1
15
0 50 100 150 200 250 km
Magnus E. Shetland Tampen Viking Horda Øygarden

NW Basin Basin Spur Graben Platform Fault Zone SE
0
10
km
20
30
1
40
0 km 50
Neogene Upper Cretaceous Jurassic Crystalline basement
Triassic-Upper
Palaeogene Lower Cretaceous Palaeozoic High velocity lower crustal body
Fig. 25.6 Interpreted regional deep seismic line and crustal transect across the northern North Sea (modified from Christiansson
et al. 2000). See Fig. 25.5 for location
Denmark). The Sorgenfrei-Tornquist Zone strikes in a 25.2.2.1 Devonian

NW-SE direction from the western Skagerrak across The Caledonian Orogeny led to uplift and formation of
northern Jutland and the Kattegat into Scania a major mountain chain along western Scandinavia
(Fig. 25.2). If the effects of post-Permian structuring and Scotland, East Greenland and a southern branch
along the Sorgenfrei-Tornquist Zone are restored, late into Poland. Thick red continental sediments, which in
Palaeozoic rift structures similar to those seen in the Britain are known as Old Red Sandstone, were depos-
Oslo Rift are observed. The major NW-SE trending ited in Devonian time in response to the extensional
faults are associated with dextral strike-slip collapse of the Caledonides. On the Norwegian main-
movements during Late Carboniferous-Early Permian land we have Devonian basins in western Norway
times. (Hornelen, Håsteinen, Kvamshesten, Solund) which
are filled with thick conglomeratic sediments. There
are also Devonian deposits further north, for example
25.2.2 Stratigraphy/Evolution on the island of Smøla. The Devonian sandstones of
western Norway are metamorphosed and can not be
A stratigraphic summary for the North Sea is given in reservoir rocks, but they are somewhat more porous in
Fig. 25.4. southern England.
NW East Shetland Basin Tampen Spur Viking Graben Horda Platform SE

0
2
s twt
2
5
25 km
Fig. 25.7 Regional seismic line across the northern North Sea (courtesy Fugro and TGS). See Fig. 25.5 for location and Fig. 25.6
for stratigraphic information
Although Devonian sediments in the northern 25.2.2.2 Carboniferous

North Sea have been reached in only a few wells, Following the markedly dry climate which prevailed
there are reasons to believe that Devonian sediments through Devonian time in the North Sea region, the
are present regionally in the deeper parts of the pre- Carboniferous period gradually became more humid.
Triassic half-grabens beneath the Horda Platform, Northwest Europe moved northwards from the arid
Viking Graben and East Shetland Basin. The pres- belt of the southern hemisphere into the humid equa-
ence of Upper Palaeozoic rocks, of both Devonian torial belt. This provided the basis for all the coal
and Lower Permian age (Rotliegendes), has also deposits. There was also a marked transgression over
been confirmed by drilling on the East Shetland large areas. The strike-slip movements along the
Platform. Seismic data reveal large sedimentary Greenland/Fennoscandia plate boundary ceased at
basins beneath the platform, thought to contain the transition from Devonian to Carboniferous, and
Upper Palaeozoic (Devonian-Carboniferous) rocks. ever since this has been an area of diverging plate
In this context it is important to note that oil is movement and rift formation until final continental
produced from Devonian sandstones in the Embla break-up and onset of seafloor spreading in earliest
Field of the Central Graben. Eocene time. There was rifting in the Midland Valley
The Caledonian plate movement changed from of Scotland, continuing along the Forth Approahes
subduction to Late Devonian lateral (strike-slip) into the Witch Ground Graben which was a volcanic
movement between Greenland and Fennoscandia, centre in the North Sea. Black mud deposited in these
including along the Great Glen Fault. In Scotland graben structures forms source rocks for oil.
this was marked by active volcanism, especially in There was no rifting in England at that time and
the Midland Valley Rift. the Carboniferous Limestone was deposited as a shelf
carbonate during the Early Carboniferous. North of the
SW Mid North Sea High Central Graben Norwegian-Danish Basin NE

0
2
s twt
3
5
25 km
Fig. 25.8 Interpreted regional deep seismic line and crustal transect across the northern North Sea (modified from Christiansson et
al. 2000). See Fig. 25.5 for location
carbonate platform and in much of the North Sea, the Netherlands, and it is these which provided the
deeper water shales predominated, with some carbon- source of the gas in this region. Fluvial sandstones
ate. Sandstones are encountered higher up in the between the coal seams can be reservoirs both for oil
Lower Carboniferous and these can be important res- and gas. Black marine shales were also deposited in
ervoir rocks in the southern North Sea (Fell Sand- the mid-Carboniferous in the southern part of the
stone). At the Lower/Upper Carboniferous boundary North Sea. In the Oslo Region there are thin
is the Yoredale Formation, consisting of cyclic sandstones and carbonate sediments of the Asker
deposits of marine carbonates and shales, and fluvial Group yielding Upper Carboniferous fossils,
sandstones. These have been interpreted as the result demonstrating a connection with the North Sea basin.
of eustatic changes in sea level due to glaciations in
the southern hemisphere, but local tectonics and even
25.2.2.3 Permian
sedimentation processes can also produce cyclicity.
During the early Permian, uplift of the Hercynian
The Hercynian (Variscan) mountain range formed
(Variscan) mountain range continued and sedimentary
along the subduction zone through Germany and
basins developed in front of it in the southern North
northern France close to south England. It was uplifted
Sea and within subsided areas of the range itself. At
as a marked topographic feature, and at its foot (the
the same time, northwest Europe was pushed farther
Variscan foredeep) major sedimentary units were
northwards from the equator, into the dry belt of the
deposited, derived from erosion in the mountains.
northern hemisphere. The high mountain range to the
Thick beds of coal developed from the swamp areas
south also contributed to severe aridity in the North
on the deltas. In Britain these sandy delta deposits are
Sea basin and most of northwest Europe. This region
known as the Millstone Grit. Contemporary coal
was then located in the middle of a large continent in a
deposits are found in the southern North Sea and in
similar position to the dry areas to the north of the types of sediment are well-exposed in southwest
present Himalayas. England near Exmouth but also in northeast England,
The latest Carboniferous-Permian extension in the near Durham and Newcastle. In the North Sea basin
North Sea region (Fig. 25.2) was linked to the the Upper Rotliegend contains an extensive, mostly
Variscan fold belt by strike-slip movements on a series aeolian, sandstone called the Auk Formation.
of NW-SE trending zones of weakness (among others The climate behind the mountain range was dry,
the Sorgenfrei-Tornquist Zone). The rifting was and a marine evaporite basin eventually developed,
associated with widespread and massive igneous possibly with a narrow passage through the Viking
activity. This is well known from the onshore part of Graben to the open ocean in the north to a seaway
the Oslo Rift (Oslo Graben) but similar rocks, both between Norway and Greenland. There was, however,
extrusives and intrusives, are present in the subsurface an important connection eastwards through Poland.
offshore (e.g. Skagerrak Graben, Sorgenfrei-Tornquist The deposition of the Zechstein salt commenced in
Zone, Central Graben; Fig. 25.9). the southern Permian basin and spread across most of
Sedimentation in the North Sea region was the North Sea basin during Late Permian time but is
dominated by two E-W aligned basins separated by absent in the northern North Sea. Sabkha deposits
the Mid-North Sea and Ringkøbing–Fyn highs. The accumulated in areas marginal to the evaporite basins.
basins were infilled with Rotliegend continental
sediments, most rapidly in the southern basin which
25.2.2.4 Triassic
was nearest the mountains and subsided fastest.
Rifting continued into earliest Triassic time and the
Alluvial fans and aeolian dune sand accumulated
Triassic to Middle Jurassic succession reflects a pat-
most of the sediment supply from the south. These
tern of repeated outbuilding of clastic wedges from the
NW Skagerrak Graben SE
0
Cenozoic Palaeozoic
Cretaceous Salt (Zechstein) 2
s twt
Jurassic Basement 4
4 10 km
Triassic
6
SW Sorgenfrei-Tornquist Zone Skagerrak-Kattegat Platform NE

0
2 ?
s twt
4 10 km
6
SW Central Graben Ringkøbing-Fyn High Norwegian-Danish Basin NE

0
2
s twt
4
5 10 km
6
W Central Graben Ringkøbing-Fyn High Horn Graben Norwegian-Danish Basin Sorgenfrei-Tornquist E

Zone
0
2 ?
s twt
?
4 ?
6 ?
6
0 100 200 300 400 500 570 km
Fig. 25.9 Regional profiles across the SE North Sea, Skagerrak and Kattegat (modified from Heeremans and Faleide 2004). See
Fig. 25.5 for location
Norwegian and East Shetland hinterlands within a The Triassic fluvial sandstones may have variable
generally evolving post-rift basin. There was still con- sorting and reservoir quality depending upon facies
siderable sediment supply from the Variscan and distance to basement sources. In the fluvial chan-
mountains to the south and there is also evidence of nel facies there may be well developed clay coatings
an uplift of Scandinavia. A broadly similar geometry on quartz grains, retarding quartz cementation. The
of the megasequences in both continental Triassic and clay coatings have probably formed in the vadose
marine Jurassic successions was related to subsidence- zone where surface water has filtered down to the
rate variations. Differential subsidence across faults groundwater table.
throughout Triassic time has also been reported. The
Øygarden Fault Zone, forming the eastern margin of 25.2.2.5 Jurassic
the Permo-Triassic basin, was active throughout most The transition from Triassic to Jurassic approximately
of the time interval. The sedimentation rate was in coincides with a change from continental to shallow
most cases high enough to keep up with the subsi- marine depositional environments. The climate also
dence, resulting in a rather flat landscape with gently gradually became more humid in the Jurassic as north-
flowing rivers. If the sediment supply had been less, west Europe was pushed northward out of the arid belt
these rift basins would have been marine. Along the at about 30 N. A transgression in the Early Jurassic
rift structure along the middle of the North Sea the (Lias) led to the accumulation of black shales over
thickness of the Triassic sediments may exceed 5 km. large parts of NW Europe and they are exposed in
The underlying Permian salt started to form diapirs Yorkshire in northeast England and Dorset in southern
and Triassic sediment was eroded, or not deposited, at England. This is because a transgression over an
the top of the structures. uneven land surface results in a shallow sea with
The climate in northwest Europe during the Trias- poor vertical circulation in the water. The Upper Lias
sic was still arid, and continental red beds continued to (Toarcian) in particular contains good source rocks for
be predominant. In England these are referred to as the oil and gas in southern parts of the North Sea. In
New Red Sandstone because they are similar to the Britain and much of the North Sea the Lias deposits
Devonian continental sediments (Old Red Sandstone). consist of black shales with thin carbonate and sand
The equivalent facies in Germany is the Bunter beds. Iron-rich layers with siderite (FeCO3) and
sandstone of Lower Triassic age. In south England chamosite (Fe-chlorite) are common. Beds rich in
east of Exmouth there are also good exposures of these two minerals were the basis for iron mines
Triassic deposits with sandstones and conglomerates which were especially important during the industrial
(Sherwood Sandstone Group) and mudstones (Mercia revolution in England, Germany and France.
Mudstone Group). The Sherwood Sandstone is a very In the northern part of the North Sea fluvial and
important groundwater reservoir in large areas of partly marine sandstones of Lower Jurassic age
England. In Southern Norway north of Hamar, the (Lunde and Statfjord formations) are important reser-
Brumundal sandstone is an outlier of a more extensive voir rocks in the Viking Graben. The Statfjord Forma-
cover of Permian red bed facies. tion is succeeded by the Dunlin Group, which is a dark
In the Upper Triassic we find carbonates marine shale but normally without enough organic
(Muschelkalk) and salt deposits (Keuper salt) in the content to become a significant source rock. Then
southern part of the North Sea. In the central and follows the Brent Group sandstone, a prograding
northern areas, however, continental clastic sedimen- delta sequence which forms the main reservoir rock
tation continued right up to the end of the Triassic in the northern North Sea. This sandstone was depos-
(Rhaetian). Sabkha environments fringed the evapo- ited in a delta that drained the central part of the North
rite basins and caliche, that is carbonate precipitation Sea towards the marine embayment to the north,
in soil profiles, is typical. Towards the end of the between the Shetland and Horda platforms. It was
Triassic the climate became less arid with more nor- sourced from an uplifted area in the south associated
mal fluvial sedimentation and gradually also marine with Middle Jurassic (Bajocian-Bathonian) volcanic
sedimentation when the Statfjord Formation was activity. The volcanic centre was located south of the
deposited. Viking Graben and east of Scotland (Moray Firth). On
the other side of the uplifted area, the sediments were
transported southwards, exemplified in good coastal side is called the Draupne Formation. The rift topog-
sections in Yorkshire. raphy produced numerous, locally overdeepened,
The lower part of the Brent Group consists of basins with poor bottom water circulation. Only a
upward-coarsening beds of mica-rich sandstones that relatively small proportion of the organic production
represent prograding delta deposits, especially distrib- was oxidised while the sedimentation rate was fairly
utary mouth bars. This is the Etive and Rannoch high. The organic-rich shales of the Upper Jurassic are
formations. The middle part is the Ness Formation, thus the prime source rock in the North Sea, and
comprised mostly of fluvial delta facies with channels, provided the main petroleum source in both the
crevasse sand deposits, lagoonal deposits and coal Statfjord and Ekofisk areas. The thickness of the
beds. These delta top facies were deposited during a Upper Jurassic sediments along the rift axis may
northward progradation of the Brent Delta. At the top reach 3,000 m. The deposition of organic-rich shales
of the Brent Group, the Tabert Formation is a gener- continued into the Early Cretaceous in some of the
ally well-sorted sandstone formed by the reworking of basins. The edges of the rotated blocks suffered ero-
deltaic deposits during a transgression. The sediment sion and a few were not buried until the Late Creta-
supply from the south was no longer able to keep up ceous. The majority of the faults die out before the
with the basin subsidence, resulting in a gradual Cretaceous but a few continue up into younger beds.
drowning of the Brent Delta. Each of these delta facies The rifting resulted in rotated fault blocks containing
has its characteristic reservoir properties, which are a sandstone reservoirs of Lower and Middle Jurassic age
function both of primary sorting and mineralogy and (sandstones of the Brent Group and Statfjord Forma-
subsequent diagenetic modification. tion). Small fan deltas developed along the rift and
Nearly all samples from these deltaic reservoir also deepwater sandstones including debris flows.
rocks show evidence of meteoric water flushing and Some of these are good reservoir rocks occuring
have well developed authigenic pore-filling kaolinite within the Upper Jurassic source rocks (see Chap. 12).
and partly dissolved feldspar (see Chap. 4). At burial Upper Jurassic sandstones deposited on submarine
depth >4 km (130 C) kaolinite have reacted with K- slopes are much less leached and have less kaolinite
feldspar and is replaced by illite. than the Brent sandstones. They may however contain
There is normally little clay coating on quartz in the grain-coating micro-quartz which retards quartz
Brent sandstones and quartz cementation may be cementation, thus preserving porosity. This is linked
extensive so that the porosity is reduced to 10-15% to a siliceous sponge (Rhaxella), see Chap. 4.
at 4 km depth. In much of the North Sea basin Kimmeridge shale
In the Late Jurassic, volcanism was very much equivalents are immature but in the Upper Jurassic rift
reduced and the areas within and surrounding the rift structures this source rock is buried to much greater
systems subsided in response to lower geothermal depth corresponding to the oil window, and even gas
gradients. At the same time, normal faulting along window.
the Viking Graben led to the rotation of basement
blocks and their overlying sediments. The shoulders 25.2.2.6 Cretaceous
of the tilted fault blocks were exposed to erosion, The last phase of rifting in the North Sea in the Late
removing Lower-Middle Jurassic and locally even Jurassic was followed by a major transgression, but the
Upper Triassic strata. The Late Jurassic (Oxfordian) uplifted rift stuctures remained dry and were islands
transgression covered the Viking Graben with a thick for most of the Early Cretaceous. There is a major
drape of clayey sediments of the Heather Formation, unconformity between the Cretaceous and the Jurassic
while the coarser clastics (sand) were deposited as except in the deep parts of the rifts where there may
turbidites and in deltas along the basin margins. have been continuous sedimentation (Figs. 25.6 and
Some of the deltas appear to have been controlled by 25.7). The Base Cretaceous Unconformity is very
the same structures that have determined the location well marked on most seismic sections from the North
of the fjords in western Norway. Sea. Fault activity diminished during the Cretaceous,
The uppermost Jurassic Kimmeridge Clay Forma- and the Cretaceous subsidence was due primarily to
tion is transgressive and often forms a several hundred crustal cooling after the Jurassic rifting. The transition
metres thick rich source rock which on the Norwegian from syn- to post-rift configuration was strongly
diachronous, suggesting that the thermal state of the highs. In the Viking Graben the carbonate content
system was not homogeneous at the onset of the post- diminishes northwards and we do not have pure lime-
rift stage. stone (Chalk) facies like that in the southern and
Three stages can be identified in the post-rift Cre- central part of the North Sea. Instead, shales predomi-
taceous development of the northern North Sea: (1) nate, though often with a significant carbonate content.
The incipient post-rift stage (Ryazanian–latest Albian) At the close of the Cretaceous and beginning of the
was characterised by different degrees of subsidence. Cenozoic, compressive movements were felt from the
The major structural features inherited from the syn- Alpine Orogeny to the south. Part of this movement
rift basin (e.g. crests of rotated fault blocks, relay was accommodated along diagonal fault zones, such
ramps and sub-platforms) had a strong influence on as the Sorgenfrei-Tornquist Zone in Scania and the
the basin configuration and hence the sediment distri- Kattegat. The Polish Basin and parts of the North
bution. (2) In the middle stage (Cenomanian–late Sea area were uplifted (inversion) and eroded.
Turonian) the internal basin relief became gradually
drowned by sediments. This is typical for basins where 25.2.2.7 Cenozoic
sediment supply outpaces or balances subsidence, as The early Cenozoic rifting, break-up and onset of
was the case in the northern North Sea. Thus, the seafloor spreading in the NE Atlantic gave rise to
influence of the syn-rift basin topography became differential vertical movements also affecting the
subordinate to the subsidence pattern determined by North Sea area. The sedimentary architecture and
the crustal thinning profile, which in turn relied on breaks are related to tectonic uplift of surrounding
thermal contraction and isostatic/elastic response to clastic source areas, thus the offshore sedimentary
sediment loading. (3) The mature post-rift stage record provides the best age constraints on the Ceno-
(early Coniacian–early Palaeocene) was characterised zoic exhumation of the adjacent onshore areas.
by the evolution into a wide, saucer-shaped basin Major depocentres sourced from the uplifted Shet-
where the syn-rift features were finally erased. Since land Platform and areas along the incipient plate
thermal equilibrium was reached at this stage, subsi- boundary in the NE Atlantic formed during Late
dence ceased, and the pattern of basin filling became, Palaeocene-Early Eocene times. A local source area
to a larger degree, dependent on extra-basinal pro- also existed in western Norway. Tectonic subsidence
cesses (see Chap. 12). accelerated in Palaeocene time throughout the basin,
The Lower Cretaceous shales are black, but only with uplifted areas to the east and west sourcing
locally do they form good source rocks. Conditions prograding wedges, which resulted in large
subsequently became more oxidising as bottom circu- depocentres close to the basin margins. Subsidence
lation improved. In the Early Cretaceous grey to red- rates outpaced sedimentation rates along the basin
dish oxidised shales were also deposited, which axis, and water depths in excess of 600 m are
become increasingly calcareous upwards. The Lower indicated. Uplift along the Sorgenfrei-Tornquist Zone
Cretaceous shales (Cromer Knoll Formation) are shal- caused erosion into top Chalk along the southeastern
low to deep marine mudstones with little sand. flank of the North Sea basin.
In the Late Cretaceous the sea attained its trans- Prominent ash layers of earliest Eocene age are
gressive maximum and clastic sedimentation almost found throughout the entire North Sea and also further
ceased across large areas of northwest Europe. Parts of north. Onshore in Denmark the volcanoclastic
Scandinavia were probably also covered by the Creta- sediments are known as “moler”. The extensive volca-
ceous sea. Sedimentation was dominated by plank- nism was related to the opening of the North Atlantic
tonic carbonate algae (coccolithoporids) which and both the Eocene and Oligocene mudstones are
formed a lime mud, the main component of Chalk, dominated by smectitic clays formed from the volca-
though it also includes some foraminifera and nic ash. These smectitic mudstones are characterised
bryozoa. The main development of the Chalk was in by seismic velocities that are low, even compared to
the Campanian and Maastrichtian, but sedimentation the overlying Neogene sediments. Ash layers rich in
continued up into the Danian of the Palaeocene, for the volcanic glass (hyaloclastites or tuff) may, however,
most part through resedimentation of earlier Chalk develop into hard layers due to diagenesis. They will
deposits which had been uplifted along the central then be characterised by high seismic velocities,
particularly after burial to more than 2 km depth and valleys to the Oslo region and deepened the Oslofjord.
quartz cementation, giving rise to strong seismic The ice flow continued along the south coast of
marker horizons (top Balder Formation). Norway and eroded a trough in the Tertiary and Meso-
In Eocene times progradation from the East Shet- zoic sediments. This is up to 700 m deep and continues
land Platform was dominant and major depocentres up along western Norway. Thick Pleistocene sedimen-
developed in the Viking Graben area, with deep water tary fans were deposited at the slope in front of the
along the basin axis. The Palaeocene and Eocene bathymetric trough.
submarine fans were built up with turbidite sands Periods of ice loading resulted in a tectonic tilting
carried out into the central part of the North Sea with of much of the shelf, particularly in the inner part. This
the Utsira High limiting their eastward extent. Parts of has in some cases influenced oil migration and resulted
Fennoscandia were probably covered by sea during in tilted OWC (e.g. in the Troll Field). Ice loading
Middle-Late Eocene times. has also caused compaction so that Quaternary
At the Eocene-Oligocene transition, southern moraines and glaciomarine sediments have become
Norway became uplifted. This uplift, in combination overconsolidated.
with prograding units from both the east and west, At most places on the shelf the Holocene (the last
gave rise to a shallow threshold in the northern North 10,000 years) is represented by only a thin layer of
Sea, separating deeper waters to the south and north. silty sediment, chiefly reworked from Quaternary or
The uplift and shallowing continued into Miocene older sediments exposed on topographic highs or
time when a widespread hiatus formed in the northern along the margins of the North Sea. Very little “new”
North Sea, as revealed by biostratigraphic data. Mio- sediment has been supplied from land during the
cene outbuilding from the north into the southeastern Holocene. This is because the fjords act as very effi-
North Sea was massive (Fig. 25.8). Coastal cient sediment traps, collecting the sediment from
progradation of the Utsira Formation in the northern the rivers. Because fjords are deep but have shallow
North Sea reflects Late Miocene-Early Pliocene uplift thresholds, not much clastic sediment reaches the
and erosion of mainland Norway. This relatively thick shelf. On the seafloor there are in many areas fre-
sandstone is a good aquifer and is used for injection of quent depressions which are typically a few tens of
CO2 in the Sleipner Field. It was still relatively warm metres across and several metres deep. These are
in the Early Pliocene, compared to the Late Pliocene called pockmarks and have formed by fluid, generally
when mountain glaciation started to develop. gas, seeping from deeper layers to emerge at the
The Late Pliocene-Early Pleistocene basin configu- seafloor.
ration was dominated by the progradation of thick The Cenozoic sedimentation was relatively rapid
clastic wedges in response to uplift and glacial erosion and the clayey sediments had little time to compact
of eastern source areas (Figs. 25.7 and 25.8). Consid- sufficiently to reduce the water content. Some beds
erable uplift of the eastern basin flank is documented by therefore display plastic folding and diapir structures
the strong angular relationship and tilting of the com- due to the under-compacted clays, especially in the
plete Cenozoic succession below the mid-Pleistocene Eocene. Polygonal faults are also common in these
angular unconformity. The Plio-Pleistocene sediments mudstones. They form a network which are from sev-
are partly glacial and partly marine representing eral hundred metres to 1 km across.
reworked glacial sediments and are typically poorly
sorted. Such sediments compact readily because of the
poor sorting of glacial and glaciomarine sediments and 25.2.3 Exploration History and Petroleum
periods of glacial advances may have contributed to the Provinces/Systems
compaction.
The ice sheets advanced repeatedly out onto the Exploration has been taking place in the North Sea for
shelf, but often deposited much of their debris load over 40 years and most of the large Norwegian fields
relatively near to land. Only during relatively short are situated here. This is a mature part of the continen-
periods did the glaciers cover most of the North Sea tal shelf and most of the plays are confirmed. The
basin. When major ice sheets developed on the Baltic Norwegian part of the North Sea can be divided into
Shield much of the ice flowed southwards along the two petroleum provinces characterised by different
petroleum systems/plays: (1) the Central Graben area, one of the few other occurrences where Chalk forms
and (2) the northern North Sea (Fig. 25.5). a reservoir rock, but the fields there are fairly small by
The most important source rocks in the North Sea comparison.
are shales in the Upper Jurassic. Thick beds of shale Chalk is comprised of microscopic
rich in organic matter were deposited over most of (0.001–0.005 mm) skeletons of planktonic algae
the North Sea area in Late Jurassic time. In the Nor- which floated in the surface waters. After they died
wegian sector, these belong to the Draupne Formation they sank to the seafloor to accumulate as a calcareous
in the Viking Group or the Mandal Formation in the ooze. First when the scanning electron microscope
Tyne Group. Middle Jurassic coal is another important (SEM) was developed was it possible to study these
source rock, mainly for gas. In the southern part of the algae. The coccolithoporid skeleton consists of calcite,
Norwegian sector, this coal is found in the Bryne which makes the sediment more stable during diagen-
Formation in the Vestland Group. In the northern esis. We do not find as much solution and
part of the North Sea, these coal seams are found in reprecipitation as would have been the case if the
the Brent Group. Source rocks may possibly be found fossils had been aragonitic. There was uplift and
in Carboniferous or older strata in the southern part some erosion of the Chalk in the Danian (earliest
of the North Sea, but these have so far not been Cenozoic) with a degree of resedimentation of Chalk
confirmed. beds along the slopes of the shallower areas, i.e.
slumping. These redeposited Chalk sediments have
25.2.3.1 Central Graben proved to be the ones with the best reservoir
In the Central Graben area rifting took place during characteristics.
Permian-Early Triassic and Middle-Late Jurassic The Chalk beds were then buried by Palaeocene
times, and Zechstein salt was deposited north of the and Eocene clays which sealed the Chalk and
Mid North Sea High (Figs. 25.8 and 25.9). When this prevented the freshwater circulation which we other-
area was first drilled, the target was the Permian wise often find in carbonate rocks along continental
sandstones beneath the salt. It was by accident that shelves. These clays have high contents of expanding
oil was discovered in the Upper Cretaceous rocks clay minerals (smectite) and form a layer with very
(Chalk). Ekofisk was the first field to be discovered, low permeability, so that porewater could not be
but now there is a string of fields with Upper Creta- forced out quickly enough with respect to the subsi-
ceous reservoirs, both on the Norwegian side and in dence rate. The effective stresses and hence compac-
the Danish Sector. tion have therefore been strongly reduced, and this is
There are other sandstone reservoirs too, such as an important factor explaining why the porosity can
those of Upper Jurassic age in the Ula and Gyda fields. still be as high as 30–35%. Both mechanical compac-
This lead is found south of the Brent delta, which tion and chemical pressure solution are reduced by the
excluded the Brent Group from forming reservoir overpressure. The Ekofisk reservoir lies at over 3 km
rocks there. The area was uplifted and Middle Jurassic depth and without overpressure there would not have
sediments are mostly absent. Later, oil has been found been such good porosity. In the shallower Valhall
in Upper Palaeozoic reservoir rocks which had been reservoir the effective stresses are even lower.
the original prospecting target in the area. In the The Ekofisk structure is a dome-shaped anticlinal
Embla Field the reservoir rock is a sandstone believed which has formed in response to an underlying salt
to be of Devonian age. diapir (Upper Permian). The structure covers c.50 km2
The Ekofisk Field, found in 1969, was the first large and has a closure height of 244 m. The oil column is
oilfield in Europe. The Chalk in the Ekofisk reservoir 306 m in thickness, which means there is oil below the
is of the same type as we have onshore in Denmark structure’s lowest point (spill point). This situation
and eastern England in stratigraphic levels requires an additional diagenetic trap, i.e. low perme-
approaching the Cretaceous/Cenozoic boundary ability coupled with the capillary forces in the Chalk
(Maastrichtian and Danian). Chalk lithologies had hinder the oil from seeping out laterally from below
previously been assumed to be far too fine-grained to the structural trap. The porosity within the reservoir
be reservoir rocks, and Ekofisk was the first large rock varies from 30 to 35% to tighter layers with
oilfield of this type. The Austin Chalk in Texas is 0–20%. However, in addition to this primary porosity
there is a secondary porosity provided by fissures. the Rhaetian-Sinemurian Statfjord Formation, in the
Fissures are particularly important in connecting the Aalenian-Bathonian Brent Group, and locally in the
small pores, so that the permeability increases to circa Pliensbachian-Toarcian Cook Formation of the Dunlin
one to a few millidarcy. Without the fissuring the Group. Sealing faults are frequently important for
permeability would be much lower and it would have this trap type. Also Upper Jurassic (Bathonian-
been impossible to produce the reservoir. The fissures Kimmeridgian) sandstones provide good reservoirs,
are probably related to the horizontal tension which perhaps particularly in the platform areas, where
developed when the beds were bent upwards above the fault block rotation is moderate (e.g. the Troll Field),
salt structure, so that the cracks are mostly vertical. and in mixed structural-stratigraphic traps along fault-
The high overpressure also aids the fissuring process. block crests. Post-rift marine sands (Cretaceous and
The source rock for the hydrocarbons at Ekofisk is the Cenozoic) add to the prospectivity of the northern
underlying Kimmeridge Clay which attains its optimal North Sea. Indeed, the Cenozoic marine sands of the
maturity at this depth. Frigg type were the major target of early exploration
Gas injection and water injection have been impor- activity.
tant for maintaining reservoir pressure and preventing In the northern North Sea there are no Permian salt
the fissures from closing. The reservoir rock has also deposits, but there are large thicknesses of Permian
been deliberately fissured by hydrofracturing, to and Triassic sediments at depth (Fig. 25.6). The large
increase the permeability. The platform at Ekofisk oil fields in the northern part of the North Sea have
has sunk several metres during production, and the Brent Group sandstones, and often also Statfjord For-
subsidence has continued despite the pressure being mation sandstones, as reservoir rocks. This applies to
maintained with water injection. According to the laws Statfjord, Gullfaks, Vigdis, Visund, Snorre and
of soil- and rock-mechanics the subsidence should Oseberg. The traps consist of rotated fault blocks
cease when the effective stresses are not increasing, formed during rifting in the Late Jurassic. They stood
but here there has probably also been a chemical as islands in the sea, and much of the Upper Jurassic
compaction which is not simply a function of effective and Lower Cretaceous is absent through non-
stresses. This may be due to water substituting for the deposition or erosion. At the top of the Gullfaks struc-
oil during production. The injected sea water contains ture we only find a thin shale from the Upper Creta-
high concentrations of sulphate and also Mg++ which ceous, and the structure probably remained above sea
were not present in the porewater of the reservoir. This level until then.
led to increased compaction (water-weakening) and The source rock is the Kimmeridge Clay from the
also changed the surface properties and wettability. Upper Jurassic (Draupne Formation), and the oil has
These factors helped to increase the recovery from migrated up to the top of the fault blocks, but stratigra-
the Ekofisk Field. There are several satellite fields, phically downwards from Upper to Middle or Lower
including Cod, Albuskjell, Tor, Eldfisk, West Ekofisk Jurassic. In the Snorre Field the erosion at the top of
and Edda. the rotated fault blocks has reached right down to
sandstones in the Upper Triassic (Lunde Formation)
25.2.3.2 Northern North Sea which is the reservoir rock together with the Statfjord
The northern North Sea contains the majority of the Formation. The Statfjord Field was discovered in 1974
giant fields discovered so far in the Norwegian conti- and was the first giant oil field on the Norwegian shelf
nental shelf. The area is, however, still under intensive after Ekofisk. The field lies in blocks 33/9 and 13/12,
exploration, and the existing infrastructure makes and extends into the British Sector. The Statfjord Field
even relatively small and complex traps interesting. is large, estimated to contain about 575 mill. tons oil
The province is extensively drilled, particularly along and 85 · 109 m3 gas and NLG. Since the sand has such
the margins of the basin, and a whole series of trap high porosity and permeability, the production rate
types have been investigated. The most common trap was very high (330,000 barrels a day in 1997). Most
type is provided by rotated fault blocks along both of the oil and gas has now been produced. Statfjord is
margins of the Viking Graben (Figs. 25.6 and 25.7), located on part of the platform-like Tampen Spur
generated during Bathonian-Ryazian extension. The which is situated north of the Viking Graben and east
reservoirs are commonly found within sandstones of of the East Shetland Basin (Figs. 25.6 and 25.7). The
Tampen Spur is one of several fault blocks which eroded and redeposited as marine sediments as part
subsided and rotated along low-angle faults in Middle of the Tarbert Formation.
to Upper Jurassic time. The reservoir rocks consist of The reservoir properties are a function of these
the Statfjord Formation (Lower Jurassic) and the Brent depositional conditions, both on account of primary
Group (Middle Jurassic). sorting and because diagenetic changes are usually
The Statfjord Formation is comprised of fluvial determined by the primary mineralogical composition
sandstones which are replaced towards the top by and sorting. The sandstones of the Brent Group are
shallow marine deposits. The Dunlin Shale succeeds generally immature with a high content of feldspar
the Statfjord Formation, and has enough organic mate- and mica indicating a basement source and a relatively
rial to be a source rock even if its contribution is very short transport into the prograding Brent delta. This
modest compared to the Kimmeridge Clay (Draupne implies that all sandstone bodies where flushed inten-
Formation). During the Middle Jurassic there was sively by freshwater shortly after deposition. Nearly
considerable volcanic activity in the central part of every sample of Brent Sandstone therefore contains
the North Sea (Rattray Formation). This volcanism abundant evidence of feldspar leaching and pore-
and the high geothermal gradients caused this area to filling authigenic kaolinite. Carbonate-cemented beds
be uplifted. The Triassic and Permian sandstones were are often associated with marine environments with
then eroded and the sediments were for the most part aragonitic fossils which have dissolved and formed
transported northwards in a fluvial system which cement.
debouched in the Brent delta. The delta built out The Brent delta was transgressed by the mudstones
across almost the entire area between west Norway of the Heather Formation towards the end of the Late
and Shetland, and sediment was also supplied from Jurassic. The main source rock for the area is the
these areas to the delta. A several hundred metre thick Upper Jurassic Kimmeridge Clay. It was deposited
series called the Brent Group was deposited, over a large part of the North Sea basin and also
consisting of five formations that represent different onshore present Britain. In the northern North Sea it
facies. First the Broom Formation was deposited, is referred to as the Draupne Formation and the
mostly on the British side, and the Oseberg Formation Mandal Formation in the Norwegian Central Graben.
on the Norwegian side. These were deposited as local It may be more than 500 m thick along the rotated fault
fans related to faults along the sides of the initial rift. blocks which emerged as islands. The irregular bottom
The Rannoch Formation consists of a mica-bearing topography in the basin gave poor circulation and
sandstone which represents an upwards-coarsening stagnant conditions with the deposition of a good
sequence – from “offshore” to “shoreface” – and source rock. The crests of many of the blocks were
represents the actual progradation of the Brent delta. so heavily eroded during rotation that most of the
The mica settled out of suspension outside the fluvial Jurassic sequence is missing from them.
channels. The Etive Formation consists of sandstones Shale at less than 3.5 km depth on the shallower
deposited on the delta front above wave base and is parts of the structures is not sufficiently mature to have
therefore better sorted. The Ness Formation represents generated oil or gas in significant quantity. Where the
the fluvial delta-top facies. Here there are meandering Kimmeridge Clay is downfaulted deeply enough along
fluvial channels and crevasse splays. The clay and the listric faults, oil has migrated up into the tectoni-
mica contents are higher in the levée and overbank cally overlying but stratigraphically underlying
deposits. The Ness Formation represents the maxi- Statfjord and Brent Groups. Most of the oil comes
mum northward extent of the delta. The Tarbert For- from the area north of the Snorre Field where the
mation is much better sorted, having been deposited as source rock (Draupne Formation) is buried to
marine delta-front sediments while the delta was being 4–5 km, but there is also some from the areas south
transgressed by the sea. The sediment supply from the of Gullfaks. Such areas that have generated a lot of oil
Mid-North Sea High was reduced as this area was are called kitchen areas.
gradually transgressed and sediment supply to the In the Troll Field the reservoir rock is of Upper
delta could then no longer keep pace with the subsi- Jurassic age and contains mostly gas formed in the
dence. Part of the Ness Formation was probably deepest parts of the basin (>4.5 km).
The oil began to migrate into the reservoir during 25.3 Mid-Norwegian Shelf/Margin
early to mid-Tertiary times, when the source rock
(Kimmeridge Clay) began to be mature. The oil/ 25.3.1 Structure
water contact in the Statfjord Formation and the
Brent Group is at different levels which relate to two Along strike the mid-Norwegian margin comprises
different pressure cells. The reservoir lies at 2.5–3 km three main segments (Møre, Vøring and Lofoten-
depth and the sandstones have little quartz cement, Vesterålen), each 400–500 km long, separated by the
such that loose sand is often recovered instead of East Jan Mayen Fracture Zone and Bivrost Lineament/
solid core samples. There is also a danger of some Transfer Zone (Fig. 25.10).
sand coming up during oil production. The Møre Margin is characterised by a narrow shelf
During recent years, increasing focus has been set and a wide/gentle slope, underlain by the wide and
on the prospectivity of sands in the post-rift sequences, deep Møre Basin with its thick Cretaceous fill
and some discoveries in different settings have been (Figs. 25.11 and 25.12). The inner flank of the Møre
made (e.g. the Agat and Grane fields). The greatest Basin is steeply dipping basinward and the crystalline
exploration success during recent years is the Edvard crust thins rapidly from >25 km to <10 km. The
Grieg and Johan Sverdrup oil discoveries in southern sedimentary succession is thickest along the western
parts of the Utsira High (see Chap. 26 for more part of the basin, 15–16 km, decreasing landwards to
details). 12–13 km. The Møre Basin comprises sub-basins
Some of the reservoirs in the northern North Sea separated by intrabasinal highs formed during Late
leak oil or gas, and the gas in particular is seen in Jurassic-Early Cretaceous rifting. Most of the struc-
seismic reflection data. The leakages are probably tural relief was filled in by mid-Cretaceous time. Sill
the result of pressure having built up to the point intrusions are widespread within the Cretaceous suc-
where the cap rock has fractured, as at Gullfaks cession in central and western parts of the Møre Basin,
(block 34/10). This means that some of the oil has and lava flows cover the western part. Seaward of the
migrated up through the Cretaceous shale cap rock Faeroe-Shetland Escarpment, at the Møre Marginal
and into the Cenozoic succession. Parts of this oil and High, thickening of the crystalline crust and
gas have accumulated in Palaeocene and Eocene shallowing of the pre-Cretaceous sediments and top
sandstones that were deposited as turbidites crystalline basement occur near the continent-ocean
emanating from the Shetland Platform to the west. transition.
After deposition, the eastern side towards the The 500 km wide Vøring Margin comprises, from
Norwegian coast was uplifted, imparting a gentle southeast to northwest, the Trøndelag Platform, the
westward dip to the top of these sandstone beds. Halten and Dønna terraces, the Vøring Basin and the
They now form traps, as in the Frigg Field where Vøring Marginal High (Figs. 25.10, 25.11, 25.12 and
the reservoir is Eocene turbidite sands. Heimdal, 25.13). The Trøndelag Platform has been largely sta-
Balder and Grane fields also contain Palaeocene and ble since Jurassic time and includes deep basins filled
Eocene sandstones, which are good reservoir rocks. by Triassic and Upper Palaeozoic sediments. The
Eocene clays rich in smectite make good cap rocks. Vøring Basin can be divided into a series of sub-basins
These clays are lightly compacted, have low seismic and highs, mainly reflecting differential vertical
velocities and often form small diapirs. During the movements during the Late Jurassic–Early Cretaceous
Palaeocene a lot of sediment was supplied from the basin evolution.
west, but these beds also dip westwards and rarely The Vøring Plateau is a distinct bathymetric feature
have closure towards the east. (Fig. 25.1), and includes the Vøring Marginal High
In the northern North Sea a gas field (Peon) is found and the Vøring Escarpment. The Vøring Marginal
in the Quaternary glacial sediments only about 160 m High consists of an outer part of anomalously thick
below the seafloor. This shows that compacted glacial oceanic crust, and a landward part of stretched conti-
sediments may serve as a cap rock and be capable of nental crust, covered by thick Early Eocene basalts
trapping gas. and underplated by mafic intrusions.
Late Cretaceous-
Palaeocene
Late Jurassic-
Early Cretaceous
Lofoten
Late Palaeozoic
8 Basin 7
11
68
UR
9
VMH
RiB
VE HG NH
VB
NS UH
GR TB
66 12
VS Trøndelag
NR
Norway Platform
13
Basin Vøring
Basin DT HB
HT
64 FB
10 RsB
VH FH
MMH
Møre
FC
Basin MT
62
0 5 10 15
Fig. 25.10 Main structural elements of the mid-Norwegian HG ¼ Hel Graben, MMH ¼ Møre Marginal High, MTFC ¼
Margin and adjacent areas (close-up of Fig. 25.2 – modified/ Møre-Trøndelag Fault Complex, NH ¼ Nyk High, NR ¼
updated from Faleide et al. 2008). Location of interpreted Nordland Ridge, NS ¼ Någrind Syncline, RiB ¼ Ribban
regional profiles (black), crustal transects (red) and seismic Basin, RsB ¼ Rås Basin, TB ¼ Træna Basin, UH ¼ Utgard
examples (blue) shown in Figs. 25.11, 25.12 and 25.13. DT ¼ High, UR ¼ Utrøst Ridge, VE ¼ Vøring Escarpment, VH ¼
Dønna Terrace, FB ¼ Froan Basin, FH ¼ Frøya High, GR ¼ Vigra High, VB ¼ Vestfjorden Basin, VMH ¼ Vøring Marginal
Gjallar Ridge, HT ¼ Halten Terrace, HB ¼ Helgeland Basin, High, VS ¼ Vigrid Syncline
The Bivrost Lineament separates the Vøring and a series of basement highs along the shelf edge.
Lofoten-Vesterålen margins, marking the northern ter- Beneath the slope, break-up-related lavas mask a sedi-
mination of the Vøring Plateau and the Vøring Mar- mentary basin whose detailed mapping is hampered by
ginal High, as well as the Vøring Escarpment. The poor seismic imaging. The continental crust on the
Bivrost Transfer Zone is a major boundary in terms Lofoten-Vesterålen margin appears to have experi-
of margin physiography, structure, break-up enced only moderate pre-break-up extension,
magmatism and lithosphere stretching; break-up contrasting with the greatly extended crust in the
related magmatism is more voluminous south of it Vøring Basin farther south.
while the less magmatic Lofoten-Vesterålen margin
was more susceptible to initial post-opening
subsidence.
25.3.2 Stratigraphy/Evolution
The Lofoten-Vesterålen margin is characterised by
a narrow shelf and steep slope (Figs. 25.10, 25.11 and
A stratigraphic summary for the Mid-Norwegian
25.12). The sedimentary basins underneath the shelf
Shelf/Margin is shown in Fig. 25.4. The pre-opening,
are narrower and shallower than on the Vøring and
structural margin framework is dominated by the NE
Møre margins. Typically they form asymmetric half-
Atlantic-Arctic Late Jurassic–Early Cretaceous rift
graben structures with changes in polarity bounded by
episode responsible for the development of major
NW UR RiB LR VfB SE
0 Plio-Pleistocene
Upper Cretaceous
7 Glacial Sediments
2
Cenozoic Lower Cretaceous
s twt
4
Jurassic-Triassic-
Breakup Lava
6 Upper Palaeozoic
10 km
?
8 Intrusion
NW Vøring Marginal High VE Vøring Basin Trøndelag Platform SE

0
4
s twt
8
8 HG NH NS UH TB NR HB 10 km
10
NW Vøring Marginal High VE Vøring Basin Halten Terrace Trøndelag Platform SE

0
2
?
4 ? ?
s twt
8
9 10 km
10
NW Møre Marginal High Møre Basin SE

0
2
?
4
s twt
?
?
?
?
6
?
8
10 10 km
10
Fig. 25.11 Regional profiles across the Mid-Norwegian Margin (profiles 7–8 modified from Blystad et al. 1995, and profiles 9–10
modified from Tsikalas et al. 2005). See Fig. 25.10 for location and abbreviations
Cretaceous basins such as the Møre and Vøring basins was characterised by regional subsidence and deposi-
off mid-Norway, and the deep basins in the SW tion of large sediment volumes. The Lower-Middle
Barents Sea (Fig. 25.3). Prior to that, Late Palaeozoic Jurassic strata (mainly sandstones) reflect shallow
rift basins formed between Norway and Greenland and marine deposition prior to the onset of the next major
in the western Barents Sea along the NE-SW rift phase.
Caledonian trend. It has been suggested that the main A shift in the extensional stress field vector to NW-
Late Palaeozoic–Early Mesozoic rift episodes took SE is recorded by the prominent NE Atlantic-Arctic
place in mid-Carboniferous, Carboniferous–Permian late Middle Jurassic–earliest Cretaceous rift episode,
and Permian–Early Triassic times. Sediment packages an event associated with northward propagation of
associated with these movements are poorly resolved, Atlantic rifting. Considerable crustal extension and
mainly because of overprint by younger tectonism and thinning led to the development of major Cretaceous
burial by thick sedimentary strata. basins off mid-Norway (Møre and Vøring basins) and
On the mid-Norwegian margin, the Trøndelag Plat- East Greenland, and in the SW Barents Sea (Harstad,
form (Froan Basin) and Vestfjorden Basin record sig- Tromsø, Bjørnøya and Sørvestsnaget basins). These
nificant fault activity in Permian–Early Triassic time. basins underwent rapid differential subsidence and
Permian–Triassic extension is generally poorly dated, segmentation into sub-basins and highs.
but is best constrained onshore East Greenland where By mid-Cretaceous time, most of the structural
a major phase of normal faulting culminated in the relief within the Møre and Vøring basins had been
mid-Permian and further block faulting took place in filled in and thick Upper Cretaceous strata, mainly
the Early Triassic. The later Triassic basin evolution fine-grained clastics, were deposited in wide basins.
NW LB UR RB LR VB SE
0
Cenozoic Oceanic basement
10 Cretaceous Extrusives
Lower Mesozoic -
7+ km/s lower crustal
Upper Palaeozoic
20 body (LCB)
Continental basement
km
30
7 Lofoten rifted margin
0 100 200 km 300
Vøring Basin Trøndelag Platform

NW VMH VE GR VS RsB DT FB SE
0
10 ?
?
20
km
30 11 Vøring rifted margin
0 100 200 300 400 500 600 km
Møre Basin
NW NB MMH VH SE
0
10
20
km
12 Møre rifted margin

30
0 100 200 300 400 500 600 km
Fig. 25.12 Regional crustal transects across the Mid-Norwegian Margin (modified/updated from Faleide et al. 2008). See
Fig. 25.10 for location and abbreviations
Pulses of coarse clastic input with an East Greenland break-up lavas. On the Møre and Vøring margins, the
provenance appeared in the Vøring Basin from early Palaeocene epoch was characterised by relatively deep
Cenomanian to at least early Campanian times. water conditions. Depocentres in the western Møre
Break-up in the NE Atlantic was preceded by and Vøring basins were sourced from the uplifted rift
prominent Late Cretaceous–Palaeocene rifting. At zone in the west. The northwestern corner of southern
the onset of this rifting, the area between NW Europe Norway was also uplifted and eroded, and the
and Greenland was an epicontinental sea covering a sediments were mainly deposited in the NE North
region in which the crust had been extensively weak- Sea and SE Møre Basin.
ened by previous rift episodes. The main period of Final lithospheric break-up at the Norwegian mar-
brittle faulting occurred in Campanian time followed gin occurred near the Palaeocene–Eocene transition at
by smaller-scale activity towards break-up. The 55 Ma. It culminated in a 3–6 m.y. period of massive
Campanian rifting resulted in low-angle detachment magmatic activity during break-up and the onset of
structures that updome thick Cretaceous sequences early seafloor spreading. At the outer margin (e.g.
and sole out at medium-to-deep intracrustal levels on Møre and Vøring margins), the lavas form character-
the Vøring and Lofoten-Vesterålen margins. istic seaward-dipping reflector sequences (Fig. 25.11)
Late Cretaceous–Palaeocene rifting at the Vøring that drilling has demonstrated to be subaerially and/or
Margin covers a 150 km wide area bounded on the neritically erupted basalts. These have become diag-
east by the Fles Fault Complex and the Utgard High nostic features of volcanic margins. During the main
(Fig. 25.10). Along the outer Møre and Lofoten- igneous episode at the Palaeocene–Eocene transition,
Vesterålen margins, most of the Late Cretaceous–Pa- sills were intruded into the thick Cretaceous
laeocene deformation is masked by the lavas, but the successions throughout the NE Atlantic margin,
structures appear to continue seawards underneath the including the Vøring and Møre basins. Magma
intrusion into organic-rich sedimentary rocks led to cooling and formation of glaciers. Large Plio-
formation of large volumes of greenhouse gases that Pleistocene depocentres formed fans in front of bathy-
were vented to the atmosphere in explosive gas metric troughs scoured by ice streams eroding the
eruptions forming several thousand hydrothermal shelf. The Naust formation consist of mostly glacial
vent complexes along the Norwegian margin. sediments which offshore Mid-Norway may reach a
The mid-Norwegian margin experienced regional thickness of more than 1,000 m.
subsidence and modest sedimentation since Middle
Eocene time and developed into a passive rifted mar-
gin bordering the oceanic Norwegian-Greenland Sea. 25.3.3 Exploration History and Petroleum
Mid-Cenozoic compressional deformation (including Provinces/Systems
domes/anticlines, reverse faults and broad-scale inver-
sion) is well documented on the Vøring margin, but its The mid-Norwegian continental shelf was opened for
timing and significance are debated. The main phase exploration in 1980, when a limited number of blocks
of deformation is clearly Miocene in age but some of on Haltenbanken were announced in the fifth licencing
the structures were probably initiated earlier in Late round, and it has had fields in production since 1993.
Eocene–Oligocene times. These are situated within the major fault complexes
There is increasing evidence on the Norwegian which separate the platform area to the east from the
margin for Late Miocene outbuilding on the inner deep Cretaceous basins in the west, and along the
shelf (Molo Formation) indicating a regional, moder- margin of the Trøndelag Platform (the Halten and
ate uplift of Fennoscandia. Over the entire shelf there Dønna terraces) (Figs. 25.10 and 25.11). In recent
is a distinct unconformity, which changes on the slope years blocks have also been awarded in greater water
to a downlap surface for huge prograding wedges depths within the deep Vøring and Møre basins.
(Fig. 25.13) of sandy/silty muds sourced on the main-
land areas around the NE Atlantic and the shelf. This
horizon marks the transition to glacial sediment depo- 25.3.3.1 Haltenbanken
sition during the Northern Hemisphere glaciations The Haltenbanken area (Halten and Dønna terraces)
since about 2.6 Ma. Latest Pliocene sedimentation is has good source rocks and reservoir rocks, and also
interspersed with ice-rafted debris signifying regional structural traps (Fig. 25.11). On the eastern side of the
main fault (the Klakk Fault), water has drained
Vøring Basin Trøndelag Platform

W VS HHA RB DT NR HB E
0
3
s twt
13
8
25 km
Fig. 25.13 Regional seismic line across the Mid-Norwegian Margin (courtesy Fugro and TGS). See Fig. 25.10 for location and
abbreviations and Fig. 25.11 for stratigraphic information
eastwards through the Jurassic sandstones up to the plagioclase, and illitisation of kaolinite is then
coast so that pressure has not built up much above prevented (See Chap. 4). These compositional
hydrostatic pressure; this has been an important factor differences may reflect source areas both from the
in hindering leakage. The main problem in the area is east and from the west (Greenland).
that many of the reservoirs are very deep and oil can The Draugen Field has Upper Jurassic sandstone
have escaped due to overpressure. Typical trap types reservoirs and is fairly shallow, just 1.5 km. The
drilled so far on the mid-Norwegian continental shelf source rock in this area is not mature and the oil
are rotated fault blocks of Jurassic age in the horst- found at Draugen has actually had a relatively long
and-graben terrane of the Halten Terrace, and complex migration path from deeper-lying areas to the west.
fault blocks within the major fault zones. Here too the The subsidence and sedimentation rates were high
source rock formed during rifting in the Late Jurassic. during the Quaternary, providing a sequence about
It is called the Spekk Formation, which approximately 1 km thick (Fig. 25.13). This is considerably greater
equates with the Draupne Formation in the North Sea than in most of the North Sea. Thus neither the reser-
and the Kimmeridge Formation in England. In the voir nor source rocks have been deeply buried for a
Early Jurassic Åre Formation there are coal beds, geologically long period. This has left the source rocks
which may be the source for the gas and possibly somewhat immature for their depth, but also means
some of the oil. that the reservoir rocks have less quartz cement.
The most important reservoir rocks are of Early and
Middle Jurassic age, with alternating sandstones and 25.3.3.2 Vøring and Møre Basins
shales from the Åre, Tilje, Tofte, Ile and Garn This area was first opened for exploration at the end of
formations (Fig. 25.4). The fields in the east, for the 1990s, with great expectation because of the large
example Heidrun Field, are not so very deep, but in structures that could be seen on the regional seismic.
the Smørbukk Field the reservoir rocks lie at more Inversion structures (mainly huge, gentle anticlines
than 4 km. Also here it was block rotation during with wavelengths in the order of tens of kilometres)
rifting which created the majority of the structures. are common in the deep Cretaceous basins, and pro-
West of the Smørbukk Field there is a large fault, vide traps of considerable size. These were generated
and to the west of that there is high overpressure in the during Tertiary inversion.
Jurassic sandstones. Several wells have been drilled The geology proved to be very different from the
here that show signs of there having once been oil North Sea and Haltenbanken. There had been consid-
present, and that it has leaked out. This is presumably erable magmatic activity associated with the break-up
due to the high pressure having reached fracture pres- between Norway and Greenland and the initiation of
sure. In the Kristin Field, however, some of the oil is seafloor spreading. Tertiary lavas had flowed out
still in place. The Lavrans and Kristin fields lie at across great tracts of land and intrusive dykes and
more than 5 km, where the temperature can reach sills penetrated the Cretaceous sediments. There was
170–180 C. Here the porosity and permeability in some concern that this could have raised the tempera-
the reservoir rocks are critical, and it is the degree of ture so much that all the oil had become gas. However,
quartz cementation which essentially determines it was found that this heating had been quite local and
whether there is sufficient porosity to produce petro- had had little overall effect.
leum. The sand grains are coated with chlorite in much More significant was the extreme thickness of the
of the Tilje Formation, and in places also in the Garn Cretaceous sequence, up to 6–7 km. This meant that
Formation. This hinders quartz precipitation and thus the Upper Jurassic source rocks had already matured
helps maintain the relatively high porosities (20–25%) by the mid-Cretaceous. This was before the reservoir
despite the great burial depth. Where the chlorite is rocks, of Upper Cretaceous and Lower Tertiary age,
absent and the sand is pure quartz, as in other parts had even been deposited, and before the structures we
of the Garn Formation, the porosity is quite low now see (e.g. Ormen Lange Dome, Helland Hansen
(<10–12%). Precipitation of illite from kaolinite Arch, Gjallar Ridge) had formed. It thus appears that
and K-feldspar has in many cases reduced the the structures have only trapped some of the gas
permeabillity significantly at a depth of >3.8-4.0 km. formed at depth from the oil. Gas has been found in
Some of the sandstones contain however dominantly several prospect in the western Vøring Basin, in Upper
Cretaceous sandstones. One of these gas discoveries is characterised by a general westward migration of the
now under development (Aasta Hansteen). It has been rifting, the formation of well-defined rifts and pull-
possible to map the distribution of sands from ampli- apart basins in the southwest, and the development of a
tude analysis of the seismic data. belt of strike-slip faults in the north. Apart from epei-
The large Ormen Lange gas field was proven in rogenic movements which produced the present day
1997. It is located at water depths of 800–1,100 m in elevation differences, the Svalbard Platform and the
the eastern Møre Basin within the Storegga Slide area. eastern part of the regional basin have been largely
The gas is found in Palaeocene and uppermost Creta- stable since Late Palaeozoic times.
ceous reservoir rocks (sandstones).
25.4.2 Stratigraphy/Evolution
25.4 Barents Sea
The Barents Sea is underlain by a thick succession of
25.4.1 Structure Palaeozoic to Cenozoic strata showing both lateral and
vertical variations in thickness and facies –
The Barents Sea covers the northwestern corner of the characterised by Upper Palaeozoic mixed carbonates,
Eurasian continental shelf. It is bounded by young evaporites and clastics, overlain by Mesozoic-
passive margins to the west and north that developed Cenozoic clastic sedimentary rocks (Fig. 25.4). Direct
in response to the Cenozoic opening of the information on the nature of the crystalline crust
Norwegian-Greenland Sea and the Eurasia Basin, beneath the Barents Sea sedimentary basins is scarce,
respectively (Figs. 25.1 and 25.2). but the available, mostly indirect, evidence indicates
The western Barents Sea is underlain by large that the basement underlying much of its western part
thicknesses of Upper Palaeozoic to Cenozoic rocks was metamorphosed during the Caledonian Orogeny
constituting three distinct regions (Figs. 25.14, 25.15 and that the structural grain within the Caledonian
and 25.16). (1) The Svalbard Platform is covered by a basement may have influenced later structural devel-
relatively flat-lying succession of Upper Palaeozoic opment. The Caledonian crystalline basement has a
and Mesozoic, mainly Triassic, sediments. (2) A NE-SW grain and was folded during Silurian time.
basin province between the Svalbard Platform and
the Norwegian coast is characterised by a number of 25.4.2.1 Upper Palaeozoic
sub-basins and highs with an increasingly accentuated Within the Caledonian domain Devonian molasse
structural relief westwards. Jurassic-Cretaceous, and sediments were deposited in intermontane basins
in the west Palaeocene-Eocene, sediments are pre- undergoing extensional collapse. A Devonian tectonic
served in the basins. (3) The continental margin regime comprising both extensional and compres-
consists of three main segments: (a) a southern sheared sional events is so far only known on Svalbard, located
margin along the Senja Fracture Zone; (b) a central on the northwestern corner of the Barents Shelf. Here,
rifted complex southwest of Bjørnøya associated with thick Devonian strata are found in a north-south
volcanism and (c) a northern, initially sheared and trending graben structure and the graben fill is discor-
later rifted margin along the Hornsund Fault Zone. dantly overlain by Carboniferous strata (Fig. 25.17).
The continent-ocean transition occurs over a narrow Lower to lower Upper Carboniferous strata, mainly
zone along the line of Early Tertiary break-up and the clastics, were deposited in extensional basins ranging
margin is covered by a thick Upper Cenozoic sedimen- from wide downwarps to narrow grabens. A 300 km
tary wedge. wide rift zone, extending at least 600 km in a north-
The post-Caledonian geological history of the easterly direction (Fig. 25.14), was formed mainly
western Barents Sea is dominated by three major rift during mid-Carboniferous times. The rift zone was a
phases, Late Devonian?-Carboniferous, Middle direct continuation of the northeast Atlantic rift
Jurassic-Early Cretaceous, and Early Tertiary, each between Greenland and Norway, but a subordinate
comprising several tectonic pulses. During Late tectonic link to the Arctic rift was also established.
Palaeozoic times most of the Barents Sea was affected The overall structure of the rift zone is a fan-shaped
by crustal extension. The later extension is array of rift basins and intrabasinal highs with
78
Late Cretaceous -
Palaeocene
Late Jurassic -
Early Cretaceous
Late Palaeozoic
76
KR
Svalbard
18 Platform
HF
Z
GH
74
KF
19
C 21
FSB Bjarmeland
MHPlatform 22
SH MB
NB 14
20 72
BB
VVP
MR NH
OB
Finnmark
VH
LH
PS
Platform
P
15
SR
16 HfB
17 TB
SB
23
70
TFP
SF
Lofoten
Z
Basin
HB
68
10 15 20 25
Fig. 25.14 Main structural elements in the western Barents Sea Complex, KR ¼ Knipovich Ridge, LH ¼ Loppa High, MB ¼
and adjacent areas (close-up of Fig. 25.2 – modified/updated Maud Basin, MH ¼ Mercurius High, MR ¼ Mohns Ridge,
from Faleide et al. 2008). Location of interpreted regional NB ¼ Nordkapp Basin, NH ¼ Nordsel High, OB ¼ Ottar
profiles (black), crustal transects (red) and seismic examples Basin, PSP ¼ Polheim Sub-platform, SB ¼ Sørvestsnaget
(blue) shown in Figs. 25.15, 25.16, 25.17, 25.18 and 25.19. Basin, SFZ ¼ Senja Fracture Zone, SH ¼ Stappen High, SR ¼
BB ¼ Bjørnøya Basin, FSB ¼ Fingerdjupet Sub-basin, GH ¼ Senja Ridge, TB ¼ Tromsø Basin, TFP ¼ Troms-Finnmark
Gardarbanken High, HB ¼ Harstad Basin, HfB ¼ Hammerfest Platform, VH ¼ Veslemøy High, VVP ¼ Vestbakken Volcanic
Basin, HFZ ¼ Hornsund Fault Zone, KFC ¼ Knølegga Fault Province
orientations ranging from northeasterly in the main rift Fingerdjupet, Maud and Ottar basins have been
zone to northerly at the present western continental interpreted as rift basins formed at this time
margin. The structural style is one of interconnected (Figs. 25.15, 25.18, 25.19 and 25.20). Fault
and segmented basins characterised by half-graben movements ceased in the eastern areas towards the
geometries. end of the Carboniferous and the structural relief was
The Carboniferous rift phase resulted in the forma- gradually infilled and blanketed by a platform succes-
tion of several interconnected extensional basins filled sion of Late Carboniferous-Permian age. The lower
with syn-rift deposits and separated by fault-bounded part of this succession passes upwards from cyclical
highs. Structural trends striking northeast to north dolomites and evaporites to massive limestones. It
dominate in most of the southwestern Barents Sea includes a widespread evaporite layer of latest
(Fig. 25.14) where the Tromsø, Bjørnøya, Nordkapp, Carboniferous-earliest Permian age mapped
NW Maud Basin MH Ottar Basin Nordkapp Basin Finnmark Platform SE

0
2
s twt
6
10 km
14
8
NW Loppa High Ottar Basin NH Nordkapp Basin Finnmark Platform SE

0
2
s twt
6
15 10 km
8
SW Sørvestsnaget Basin Bjørnøya Basin Fingerdjupet Sub-basin NE

0
Neogene
2 Palaeogene
4 Upper Cretaceous
s twt
Lower Cretaceous -
6 Upper Jurassic
Middle Jurassic - Triassic
8 Upper Palaeozoic
16 10 km
10 Salt
W Lofoten Basin Sørvestsnaget Basin Senja Ridge Tromsø Basin Hammerfest Basin E
0
2
s twt
8
17 10 km
10
Fig. 25.15 Regional profiles across the western Barents Sea (modified from Faleide et al. 1993, Breivik et al. 1995). See Fig. 25.14
for location and abbreviations
regionally in the SW Barents Sea (also on the NE growth of marginal carbonate build-ups, a regional
Greenland shelf/margin). shallow-water carbonate platform was established dur-
The thickness of the salt layer in the Nordkapp ing Sakmarian-Artinskian times. Carbonate sedimen-
Basin (Fig. 25.18), which is inferred to have reached tation occurred in the entire region until late Early
4–5 km locally, implies that a substantial fault- Permian time when clastic deposition started to domi-
generated depression was in place or developed during nate the area (Fig. 25.4). The change to platform type
early Late Carboniferous times. Not all of this thick- sedimentation marked the initial development of a
ness corresponds to a fault-defined relief, however, regional sag basin which continued to subside in the
because salt was also deposited in the basin during Late Permian during the deposition of cherty
the subsequent phase of differential thermal subsi- limestones and shales.
dence. Still, a considerable fault-bounded basin must The western part of the rift system was affected by
have existed. renewed faulting, uplift and erosion in Permian-Early
From Bashkirian to Artinskian/Early Kungurian Triassic times. Normal faulting along the western
times, carbonate deposition took place on a broad margin of the Loppa High (Figs. 25.19 and 25.20) as
shelf. Carbonate build-ups were common at basin well as the uplift, tilting and erosional truncation of the
margins and intrabasinal highs controlled by underly- high itself are of sufficient magnitude to indicate a
ing older structures. Following Gzelian-Asselian/ significant Permian-Early Triassic rift phase affecting
Sakmarian deposition of basinal evaporites and the N-S structural trend. Evidence of fault movements
0 W HFZ CB BFZ E
10
20
km
30 18 Svalbard margin
0 100 200 300 km
W HFZ KFC Svalbard Platform E

0 Cenozoic
Cretaceous
10 Lower Mesozoic -
Upper Palaeozoic
Devonian
20
Continental basement
km
Oceanic basement
30 19 Hornsund sheared margin
Extrusives
0 100 200 300 km
W LB Vestbakken Volcanic Province Bjørnøya Basin Loppa High E

0
10
20
km
30 20 Vestbakken rifted margin
0 100 200 300 km
Fig. 25.16 Regional crustal transects across the western Barents Sea-Svalbard margin (modified/updated from Faleide et al. 2008).
See Fig. 25.14 for location and abbreviations
is found as far north as the Fingerdjupet Sub-basin. which is complex. There seems to be an increased
The erosional surface associated with this rift phase content of coarse clastic rocks in the younger intervals.
extends from the Loppa High to the Stappen High, and A similar increase is also observed in an easterly and a
Permian tectonic activity is known to have occurred on southerly direction towards the main sediment source
Bjørnøya and on the Sørkapp High in the southern part area. Marine conditions prevailed in Late Permian and
of Spitsbergen. Therefore the rift phase probably Early Triassic times followed by a shallowing and
affected a narrow northerly trending zone along the partial exposure of some areas. The continental regime
entire present western margin. prevailed in large areas in Middle Triassic time and the
northward and westward prograding deltaic system
25.4.2.2 Mesozoic continued to infill the regional basin.
In Early Triassic time a regional deepwater basin In the central and northern basin areas marine
covered much of the Barents Sea. During the relatively conditions existed throughout the Middle Triassic. A
short Triassic period large amounts of clastic good Middle Triassic source rock is known from
sediments were deposited. The Uralian highland to Svalbard and this is probably present in large parts of
the east was an important sediment source. Sediments the western Barents Sea. In the Late Triassic the
were also shed into the Barents Sea from the Baltic shoreline was moved back to the southern and eastern
Shield and from other local source areas. Differential borders of the SE Barents Sea basin. The Triassic
loading triggered salt diapirism in the Nordkapp Basin ended with regression and erosion. Late Triassic
(Fig. 25.18) and possibly later also in other salt basins sediments were also derived from other source areas,
in the SW Barents Sea. mainly in the northwest.
The Triassic strata are dominated by shales and The Lower-Middle Jurassic interval is dominated
sandstones, the vertical and lateral distribution of by sandstones throughout the Barents Sea. These
ISFJORDEN
Grumantbyen
W BFZ E
Festningen Colesbukta Fault
0
s twt
2
?
?
10 km
24
4
VAN MIJENFJORDEN
W Ishøgda E Palaeogene
0
Lower Cretaceous -
Upper Jurassic
Middle Jurassic -
24
s twt Triassic
2
Permian -
Carboniferous
25
Devonian
25
4 Continental basement
Fig. 25.17 Svalbard geology in seismic lines from fjords (modified from Faleide et al. unpublished)
NNW Maud Basin Bjarmeland Platform Ottar Basin Nordsel High Nordkapp Basin SSE
0
2
s twt
21
25 km
Fig. 25.18 Regional seismic line across the SW Barents Sea (Nordkapp Basin – Bjarmeland Platform – Maud Basin) (courtesy
Fugro and TGS). See Fig. 25.14 for location and Fig. 25.15 for stratigraphic information
sandstones form the main reservoir in the SW Barents The late Middle Jurassic sequence boundary marks the
Sea (Hammerfest Basin; Fig. 25.20). They probably onset of rifting in the southwestern Barents Sea,
also covered the Loppa High and Finnmark Platform whereas unconformities within the Upper Jurassic
but were partly eroded during later tectonic activity. sequence reflect interplay between continued faulting
SW Veslemøy High Loppa High Bjarmeland Platform NE

0
2
s twt
22
25 km
Fig. 25.19 Regional seismic line across the SW Barents Sea (Bjarmeland Platform – Loppa High – Veslemøy High) (courtesy
Fugro and TGS). See Fig. 25.14 for location and Fig. 25.15 for stratigraphic information
NNW Bjørnøya Basin PSP Loppa High Hammerfest Basin NE

0
Snøhvit Goliath
3
s twt
6 23
25 km
Fig. 25.20 Regional seismic line across the SW Barents Sea (Hammerfest Basin – Loppa High – Bjørnøya Basin) (courtesy Fugro
and TGS). See Fig. 25.14 for location and Fig. 25.15 for stratigraphic information
and sea-level changes. The sequence is so thin that it is The Late Jurassic-earliest Cretaceous structuring in
generally impossible to resolve these unconformities the SW Barents Sea was characterised by regional
on the seismic data. The shales and claystones contain extension accompanied by strike-slip adjustments
thin interbedded marine dolomitic limestone and rare along old structural lineaments, developing the
siltstones or sandstones toward the basin flanks, Bjørnøya, Tromsø and Harstad basins as prominent
reflecting relatively deep and quiet marine rift basins (Fig. 25.15). The evolution of these basins
environments. was closely linked to important tectonic phases/events
in the North Atlantic-Arctic region. Rifting continued completeness. In the Tromsø Basin a 1,200 m shale
in Early Cretaceous time and an important phase of succession has been drilled while seismic data indicate
Aptian faulting is documented in the SW Barents Sea. that the sequence reaches 2,000–3,000 m in rim
Several phases of Late Mesozoic and Early Cenozoic synclines in the central basin. Whereas the Tromsø
rifting gave rise to very deep basins in the SW and Sørvestsnaget basins were depocentres through
Barents Sea. A tentative Middle Jurassic horizon most of this period, areas further east were either
corresponding to the top of the shallow marine transgressed only during maximum sea-level and/or
sandstones, also marking the onset of late Mesozoic display only condensed sections.
rifting, can be followed from the seafloor down to 9 s
twt (16–17 km) over a relatively short distance
25.4.2.3 Cenozoic
(Fig. 25.15).
The Cenozoic structuring was related to the two-
The Lower Cretaceous comprises three sedimen-
stage opening of the Norwegian-Greenland Sea
tary units from Valanginian to Cenomanian. Shales
and the formation of the predominantly sheared
and claystones dominate, with thin interbeds of silt,
western Barents Sea continental margin (Fig. 25.3).
limestone and dolomite. These strata make up the
Continental break-up and onset of seafloor spreading
main basin fill in the deep SW Barents Sea basins
were preceded by a phase of rapid late Palaeocene
and they dominate the regional subcrop pattern. The
subsidence. The Palaeogene succession rests
marine environments throughout the deposition of
unconformably on the Cretaceous and this deposi-
these units were dominated by distal conditions with
tional break at the Cretaceous-Tertiary transition
periodic restricted bottom circulation.
(Maastrichtian-Danian) occurs throughout the south-
In Early Cretaceous time the northern Barents Sea
western Barents Sea.
was characterised by widespread magmatism without
The western Barents Sea-Svalbard margin devel-
any signs of faulting. Extrusives and intrusives (sills
oped from a megashear zone which linked the
and dykes) belonging to a regional Large Igneous
Norwegian-Greenland Sea and the Eurasia Basin dur-
Province (LIP) in the Arctic are documented from
ing the Eocene opening. The first-order crustal struc-
both onshore (Svalbard and Franz Josef Land) and
ture along the margin and its tectonic development is
offshore areas in the northern Barents Sea. The
mainly the result of three controlling parameters: (1)
magmatism was most extensive during Barremian?-
the pre-break-up structure, (2) the geometry of the
Aptian times but both older and younger dates have
plate boundary at opening and (3) the direction of
been reported. The present distribution in the northern
relative plate motion. The interplay between these
Barents Sea is modified by later uplift and erosion. The
parameters gave rise to striking differences in the
Early Cretaceous magmatism within the Arctic LIP
structural development of the different margin
had important palaeogeographic implications. It
segments of sheared and/or rifted nature. The
caused regional uplift and the Neocomian in the
continent-ocean boundary is clear along the sheared
north and east was characterised by southward sedi-
Senja margin. The central rifted margin segment
ment progradation. This depositional regime is diffi-
southwest of Bjørnøya was associated with
cult to document in the western Barents Sea due to
magmatism in the Vestbakken Volcanic Province
later uplift and erosion. The magmatism and regional
both during break-up at the Palaeocene-Eocene transi-
uplift was related to rifting and break-up in the
tion and later in the Oligocene. In the north, strike-slip
Amerasia (Canada) Basin and the formation of the
movements between Svalbard and Greenland gave rise
Alpha Ridge.
to compressional (transpressional) deformation within
Little or no Upper Cretaceous sediments were
the Spitsbergen Fold and Thrust Belt (Fig. 25.17).
deposited in the Barents Sea except in the SW Barents
Compressional deformation is also observed in the
Sea which continued to subside in response to faulting
Barents Sea east of Svalbard, showing that stress
in a pull-apart setting. The Wandel Sea Basin is a NE
related to transpression at the plate boundary west of
Greenland equivalent where Late Cretaceous strike-
Svalbard was transferred over large distances. Domal
slip faulting and pull-apart basin formation is well
structures observed in the eastern Barents Sea may be
documented (Figs. 25.2 and 25.3). The Upper Creta-
related to this compressional regime.
ceous succession varies in thickness and
The opening of the Greenland Sea was complex, southern Barents Sea were opened for exploration in
involving jumps in the location of the spreading axis 1980 and the first discovery was made in 1981.
and the splitting off of microcontinents. Since earliest A great variety of traps (fault and salt structures,
Oligocene time (magnetic chron 13) Greenland moved stratigraphic pinchout) and sealing mechanisms exist
with North America in a more westerly direction rela- in the Barents Sea area, and several different play
tive to Eurasia (Fig. 25.3). This gave rise to extension, models have proven hydrocarbon accumulations. The
break-up and onset of seafloor spreading also in the early discoveries were dominated by gas and gas con-
northern Greenland Sea west of Svalbard, and a deep- densate accumulations (e.g. Snøhvit Field), but several
water gateway between the North Atlantic and Arctic oil discoveries have been made recently in the SW
was finally established sometime in the Miocene. This Barents Sea (see below). Potential reservoirs are
had major implications for the palaeo-oceanography distributed across several stratigraphic levels from
and climate in the region. Devonian to Tertiary and comprise both sandstone
The late Cenozoic evolution was characterised by and carbonate lithologies. Jurassic sandstones are the
subsidence and burial of the margins by thick main proven reservoir but hydrocarbons have also
sediments in a clastic wedge derived from the uplifted been found in sandstones of Triassic and Cretaceous
Barents Sea area. Late Cenozoic uplift and erosion of age. Several proven and potential source rock units are
the Barents Sea has removed most of the Cenozoic present in the Barents Sea. The most important are the
sediments, and even older strata. The subcrop pattern Upper Jurassic and Triassic organic-rich (anoxic)
is dominated by Mesozoic units. The erosion seems to shales. Lower Cretaceous, Permian and Carboniferous
have been most extensive in the western Barents Sea, shales, as well as Lower Permian evaporites and
especially in the northwestern part including Svalbard Lower Jurassic coals, also have source potential.
where more than 3,000 m of sedimentary strata have About 25 discoveries have been made in the
been removed. In the SW Barents Sea the erosion Barents Sea, most of them in the Hammerfest Basin
estimates are in the range 1,000–1,500 m. In the east- (Fig. 25.20) where their reservoirs are in sandstones,
ern Barents Sea, little work has so far been done on mainly Jurassic, as in the Snøhvit Field. Deeper
Cenozoic uplift and erosion, however, erosion has discoveries have also been made, such as in Triassic
been calculated to be between 250 and 1,000 m. sandstones in 7122/7-1 (Goliath Field) and 7125/4-1
High seismic velocities at the top of the bedrock in (Nucula Field). Oil and gas have also been found in
the NW Barents Sea, decreasing south- and eastwards, Triassic sandstones in the Nordkapp Basin, which is
reflect the pattern of differential uplift and erosion. dominated by salt tectonics (Fig. 25.18). Gas and oil
The Neogene and Quaternary, resting unconformably have been found in carbonates of Carboniferous to
on Palaeogene and Mesozoic rocks, thickens dramati- Permian age on the Finnmark Platform. Recently,
cally in the huge sedimentary wedge at the margin. several oil discoveries have been made in the SW
The glacial sediments are dated as Late Pliocene to Barents Sea at or close to the Loppa High (Johan
Pleistocene/Holocene in age. Castberg and Gohta/Alta).
The most significant exploration problem in the
Western Barents Sea relates to the severe uplift and
erosion of the area that took place during the Ceno-
25.4.3 Exploration History and Petroleum
zoic. The quantity of sediments removed, and the
Provinces/Systems
timing of this removal, are still a matter of debate,
but it is agreed that the uplift and erosion have had
More than three decades of research and petroleum
important implications for oil and gas exploration in
exploration in the Barents Sea have revealed a deep
the Barents Sea. Residual oil columns found beneath
and complex sedimentary basin system affected by a
gas fields in the Hammerfest Basin indicate that the
variety of geological processes. A large petroleum
structures were once filled, or partially filled, with oil.
potential has been proven including multiple source
The removal of up to 2 km of sedimentary overburden
and reservoir intervals. However, there are still many
from the area has had critical consequences for these
uncertainties and problems linked to understanding
accumulations: exsolution of gas from the oil, and
the prospectivity of the Barents Sea. Areas in the
expansion of the gas due to the decrease in pressure,
resulted in expulsion of most of the oil from the traps. results in a reduction in differential stresses during
Seal breaching and spillage probably also occurred as tectonic deformation. In the parts of the Barents Sea
a result of the uplift and tilting. A further consequence where the uplift and erosion has been more than 1.5–
of these late movements was the cooling of the source 2.0 km the maximum temperature may have exceeded
rocks in the area, which effectively caused most 70–80˚C and Jurassic or Cretaceous rocks are quartz
hydrocarbon generation to cease. cemented to some degree. This also means that the
The Neogene uplift and erosion has had important temperature has pasteurised the reservoir rocks with
consequences for Barents Sea petroleum systems – respect to bacteria. Oil migrating into such reservoir
different from other basins at the Norwegian Conti- rocks will not easily be biodegraded. An example to
nental Shelf. illustrate this may be the Wisting discovery where
Much of the North Sea basin and Haltenbanken normal unbiodegraded oil is found at very shallow
offshore mid-Norway have been dominated by depth, 2–300 m below the seafloor.
continued subsidence and sedimentation. Permian During glaciations maximum ice loadings of up
and Jurassic rifting resulted in uplift and erosion but to 2 km or more could add up to 10–20 MPa to the
most of the sedimentary sequences making up the total overburden stress depending on the sea level.
basin fill are now at their maximum burial depth. Overpresssured water below the ice would reduce the
This simplifies the reconstruction of the burial history effective stress transmitted to the seafloor, but increase
also with respect to temperature and time for the pore pressure in the rocks below the seafloor. This
modelling of source rock maturation and migration. means that more gas will be dissolved in the porewater
Similar time-temperature integrals may be used to of the underlying sedimentary rocks, and in reservoirs
model diagenetic processes like quartz cementation more gas could also dissolve in the oil phase.
in reservoir rocks to predict reservoir quality. Particu- During interglacial periods with no ice load,
larly at Haltenbanken up to 1.5 km of Plio-Pleistocene reduced pressure will cause gas to bubble out of the
basin subsidence and sedimentation imply that porewater and appear as shallow gas. It is important to
reservoirs now at 4 km burial depth were at 2.5 km distinguish between shallow biogenic gas produced by
depth 1.5–3.0 million years ago. The limited time at bacterial processes and thermogenic gas formed from
elevated temperature (>100˚C) resulted in limited maturation of source rocks at greater depth. This may
quartz cement compared to sandstones with less recent be shallow gas due to upwards migration and also due
subsidence. In the Northern North Sea Quaternary to erosion of the overburden.
uplift has affected the eastern parts including the The Quaternary cover of glacial sediments may be
Troll Field and has also caused late tilting of some of hard and well compacted by the ice load and could in
the sequences thus influencing petroleum migration some cases also serve as a cap for oil and gas. The
and the fluid contacts. timing for trapping of petroleum would however have
In the western part of the Barents Sea late Neogene been very short (from the last glaciation). Because of
uplift and erosion may total 1.5–2 km. The timing of their poor sorting of clay, silt and sand glacial
maximum burial depth and temperature is then more sediments compact very effectively at moderate effec-
difficult to estimate with respect to maturation, migra- tive stresses (see Chap. 11). In the northern North Sea
tion and sometime remigration of oil and gas into the Peon field is an example where the reservoir is only
traps. Maturation of source rocks is not only a function at about 200–300 m depth in Pleistocene glacial
of the maximum burial temperature but of the time sediments which have been compacted by glacial
exposed to temperatures above 90–100˚C. loading.
Sandstones undergo diagenesis and quartz cemen- The late uplift of the Barents Sea basin and proxi-
tation both during subsidence and uplift as an expo- mal parts of the Norwegian offshore basins may have
nential function of temperature when it is above about caused leakage of petroleum reservoir. The many pro-
80˚C. Chemical compaction due to quartz cementation lific onshore basins in the world show however that not
and porosity loss will therefore continue also during all traps for oil and gas have leaked during uplift and
uplift. If the rock volume is reduced the shrinkage erosion.
Further Reading and Implications for Petroleum Exploration. Geological

Society Special Publication 196, pp. 1–12.
Doré, A.G., Lundin, E.R., Jensen, L.N., Birkeland, O., Eliassen,
P.E. and Fichler, C. 1999. Principal tectonic events in the
Bell, R.E., Jackson, C., Elliott, G.M., Gawthorpe, R.L., Sharp,
evolution of the northwest European Atlantic margin. In:
I.R. and Michelsen, L. 2014. Insights into the development
Fleet, A.J. and Boldy, S.A.R. (eds.), Petroleum Geology of
of major rift‐related unconformities from geologically
Northwest Europe: Proceedings of the 5th Conference. Geo-
constrained subsidence modelling: Halten Terrace, offshore
logical Society, pp. 41–61.
mid Norway. Basin Research 26, 203–224.
Ebbing, J., Lundin, E., Olesen, O. and Hansen, E.K. 2006. The
Bergh, S.G., Eig, K., Kløvjan, O.S., Henningsen, T., Olesen, O.
mid-Norwegian margin: A discussion of crustal lineaments,
and Hansen, J.A. 2007. The Lofoten Vesterålen continental
mafic intrusions, and remnants of the Caledonian root by 3D
margin: A multiphase Mesozoic-Palaeogene rifted shelf as
density modelling and structural interpretation. Journal of
shown by offshore-onshore brittle fault-fracture analysis.
the Geological Society 163, 47–60.
Norwegian Journal of Geology 87, 29–58.
Eldholm, O., Tsikalas, F. and Faleide, J.I. 2002. Continental
Blystad, P., Brekke, H., Færseth, R.B., Larsen, B.T., Skogseid, J.
margin off Norway 62–75 N: Paleogene tectono-magnetic
and Tørudbakken, B. 1995. Structural elements of the Nor-
segmentation and sedimentation. In: Jolley, D.W. and Bell,
wegian continental shelf, Part II: The Norwegian Sea
B. (eds.), North Atlantic Igneous Province: Stratigraphy,
Region. Norwegian Petroleum Directorate Bulletin 8.
Tectonic, Volcanic and Magmatic Processes. Geological
Breivik, A.J., Faleide, J.I. and Gudlaugsson, S.T. 1998.
Society Special Publication 197, pp. 38–68.
Southwestern Barents Sea margin: Late Mesozoic sedi-
Engen, Ø., Faleide, J.I. and Dyreng, T.K. 2008. Opening of the
mentary basins and crustal extension. Tectonophysics
Fram Strait gateway: A review of plate tectonic constraints.
293, 21–44.
Tectonophysics 450(1), 51–69.
Breivik, A., Faleide, J.I., Mjelde, R., Flueh, E. and Murai,
Evans, D., Graham, C., Armour, A. and Bathurst, P. (eds.). 2003.
Y. 2014. Magmatic development of the outer Vøring margin
The Millenium Atlas: Petroleum geology of the central and
from seismic data. Journal of Geophysical Research: Solid
northern North Sea. Published by the Geological Society.
Earth 119, 1–23.
Faleide, J.I., Gudlaugsson, S.T. and Jacquart, G. 1984. Evolu-
Breivik, A., Gudlaugsson, S.T. and Faleide, J.I. 1995. Ottar
tion of the western Barents Sea. Marine and Petroleum
Basin, SW Barents Sea: A major Upper Paleozoic rift basin
Geology 1, 123–150.
containing large volumes of deeply buried salt. Basin
Faleide, J.I., Kyrkjebø, R., Kjennerud, T., Gabrielsen, R.H.,
Research 7, 299–312.
Jordt, H., Fanavoll, S. and Bjerke, M.D. 2002. Tectonic
Brekke, H. 2000. The tectonic evolution of the Norwegian Sea
impact on sedimentary processes during the Cenozoic evo-
continental margin with emphasis on the Vøring and Møre
lution of the northern North Sea and surrounding areas. In:
basins. In: Nøttvedt, A. et al. (eds.), Dynamics of the Nor-
Dore et al. (eds.), Exhumation of the Circum-Atlantic Conti-
wegian Margin. Geological Society Special Publication 167,
nental Margins: Times, Mechanisms and Implications for
327–378 pp.
Petroleum Exploration. Geological Society Special Publica-
Brekke, H., Sjulstad, H.I., Magnus, C. and Williams, R.W. 2001.
tion 196, pp. 235–269.
Sedimentary environments offshore Norway – An overview.
Faleide, J.I., Solheim, A., Fiedler, A., Hjelstuen, B.O.,
In: Martinsen, O. and Dreyer, T. (eds.), Sedimentary
Andersen, E.S. and Vanneste, K. 1996. Late Cenozoic evo-
Environments Offshore Norway – Paleozoic to Recent,
lution of the western Barents Sea-Svalbard continental mar-
NPF Special Publication 10, pp. 7–37 (Norwegian Petroleum
gin. Global and Planetary Change 12, 53–74.
Society and Elsevier Science B.V.).
Faleide, J.I., Tsikalas, F., Breivik, A.J, Mjelde, R., Ritzmann,
Christiansson, P., Faleide, J.I. and Berge, A.M. 2000. Crustal
O., Engen, Ø., Wilson, J. and Eldholm, O. 2008. Structure
structure in the northern North Sea – An integrated geophys-
and evolution of the continental margin off Norway and the
ical study. In: Nøttvedt, A. (ed.), Dynamics of the Norwe-
Barents Sea. Episodes 31, 82–91.
gian Margin. Geological Society Special Publication 167,
Faleide, J.I., Vågnes, E. and Gudlaugsson, S.T. 1993. Late
pp. 15–40.
Mesozoic-Cenozoic evolution of the southwestern Barents
Clark, S.A., Glørstad-Clark, E., Faleide, J.I., Schmid, D., Hartz,
Sea in a regional rift-shear tectonic setting. Marine and Petro-
E.H. and Fjeldskaar, W. 2014. Southwest Barents Sea rift
basin evolution: Comparing results from backstripping and
Færseth, R.B. 1996. Interaction of Permo-Triassic and Jurassic
time-forward modelling. Basin Research 26, 550–566.
extensional fault-blocks during the development on the
Dimakis, P., Braathen, B.I., Faleide, J.I., Elverhoi, A. and
Northern North Sea. Journal of the Geological Society 153,
Gudlaugsson, S.T. 1998. Cenozoic erosion and the preglacial
931–944.
uplift of the Svalbard-Barents Sea region. Tectonophysics
Færseth, R., Gabrielsen, R.H. and Hurich, C.A. 1995. Influence
300, 311–327.
of basement in structuring of the North Sea Basin, offshore
Doré, A.G., Cartwright, J.A., Stocker, M.S., Turner, J.P. and
southwest Norway. Norsk Geologisk Tidsskrift 75, 105–119.
White, N.J. 2002. Exhumation of the North Atlantic margin:
Færseth, R.B. and Lien, T. 2002. Cretaceous evolution in the
Introduction and background. In: Doré, A.G., Cartwright, J.
Norwegian Sea – A period characterized by tectonic quies-
A., Stocker, M.S., Turner, J.P. and White, N.J. (eds.), Exhu-
cence. Marine and Petroleum Geology 19, 1005–1027.
mation of the North Atlantic Margin: Timing, Mechanisms
Gabrielsen, R.H. and Doré, A.G. 1995. The history of tectonic Europe. Geological Society Special Publication 223, pp.
models on the Norwegian continental shelf. In: Hanslien, S. 159–177.
(ed.), Petroleum Exploration and Exploitation in Norway - Henriksen, E., Ryseth, A.E., Larssen, G.B., Heide, T., Rønning,
Past Experiences and Future Challenges. A Celebration of 25 K., Sollid, K., et al. 2011. Tectonostratigraphy of the greater
Years. Norwegian Petroleum Society Special Publication 4, Barents Sea: Implications for petroleum systems. Geological
pp. 341–375. Society, London, Memoir 35, pp. 163–195.
Gabrielsen, R.H., Færseth, R.B., Jensen, L.N., Kalheim, J.E. and Hjelstuen, B.O., Eldholm, O. and Faleide, J.I. 2007. Recurrent
Riis, F. 1990. Structural elements of the Norwegian conti- Pleistocene mega-failures on the SW Barents Sea margin.
nental shelf, Part II: The Norwegian Sea Region. Norwegian Earth and Planetary Science Letters 258, 605–618.
Petroleum Directorate Bulletin 6. Hjelstuen, B.O., Sejrup, H.P., Haflidason, H., Nygård, A.,
Gabrielsen, R.H., Færseth, R.B., Steel, R.J., Idil, S. and Kløvjan, Ceramicola, S. and Bryn, P. 2005. Late Cenozoic glacial
O.S. 1990. Architectural styles of basin fill in the northern history and evolution of the Storegga Slide area and adjacent
Viking Graben. In: Blundell, D.J. and Gibbs, A.D. (eds.), slide flank regions, Norwegian continental margin. Marine
Tectonic Evolution of the North Sea Rifts. Clarendon, and Petroleum Geology 22, 57–69.
Oxford, pp. 158–179. Jarsve, E.M., Faleide, J.I., Gabrielsen, R.H. and Nystuen, J.P.
Gabrielsen, R.H., Kyrkjebø, R., Faleide, J.I., Fjeldskaar, W. and 2014. Mesozoic and Cenozoic basin configurations in the
Kjennerud, T. 2001. The Cretaceous post-rift basin configura- North Sea. IAS Special Publication 46, pp. 417–452.
tion of the northern North Sea. Petroleum Geoscience 7, Jarsve, E.M., Maast, T.E., Gabrielsen, R.H., Faleide, J.I.,
137–154. Nystuen, J.P. and Sassier, C. 2014. Seismic stratigraphic
Gabrielsen, R.H., Odinsen, T. and Grunnaleite, I. 1999. Struc- subdivision of the Triassic succession in the Central North
turing of the Northern Viking Graben and the Møre Basin; Sea; integrating seismic reflection and well data. Journal of
the influence of basement structural grain and the particular the Geological Society 171, 353–374.
role of the Møre-Trøndelag Fault Complex. Marine and Johansen, S.E., Ostisty, B.K., Birkeland, Ø., Fedorovsky, Y.F.,
Petroleum Geology 16, 443–465. Martirosjan, V.N., Bruun Christensen, O., Cheredeev, S.I.,
Gaina, C., Gernigon, L. and Ball, P. 2009. Palaeocene–Recent Ignatenko, A.A. and Margulis, M. 1993. Hydrocarbon poten-
plate boundaries in the NE Atlantic and the formation of the tial in the Barents Sea region: Play distribution and potential.
Jan Mayen microcontinent. Journal of the Geological Soci- In: Vorren, T.O. et al. (eds.), Arctic Geology and Petroleum
ety 166, 601–616. Potential, NPF Special Publication 2. Elsevier, New York,
Gernigon, L. and Brönner, M. 2012. Late Palaeozoic architec- NY, pp. 273–320.
ture and evolution of the southwestern Barents Sea: Insights Jordt, H., Faleide, J.I., Bjørlykke, K. and Ibrahim, M.T. 1995.
from a new generation of aeromagnetic data. Journal of the Cenozoic stratigraphy of the central and northern North Sea
Geological Society 169, 449–459. Basin: Tectonic development, sediment distribution and
Gernigon, L., Brönner, M., Roberts, D., Olesen, O., Nasuti, A. provenance areas. Marine and Petroleum Geology 12,
and Yamasaki, T. 2014. Crustal and basin evolution of the 845–879.
southwestern Barents Sea: From Caledonian orogeny to Laberg, J.S., Andreassen, K., Knies, J., Vorren, T.O. and
continental breakup. Tectonics 33, 347–373. Winsborrow, M. 2010. Late Pliocene–Pleistocene develop-
Gernigon, L., Ringenbach, J.-C., Planke, S., LeGall, B. and ment of the Barents Sea ice sheet. Geology 38, 107–110.
Jonquet-Kolstø, H. 2003. Extension, crustal structure and Lundin, E.R. and Doré, A.G. 1997. A tectonic model for the
magmatism at the Outer Vøring Basin, North Atlantic Mar- Norwegian passive margin with implications for the NE
gin (Norway). Journal of the Geological Society 160, Atlantic: Early Cretaceous to break-up. Journal of the Geo-
197–208. logical Society 154, 545–550.
Gernigon, L., Ringbach, J.C., Planke, S., Planke, S. and Le Gall, Lundin, E.R. and Doré, A.G. 2002. Mid-Cenozoic post-breakup
B. 2004. Deep structures and breakup along rifted margins: deformation in the “passive” margins bordering the
Insights from integrated studies along the outer Vøring Basin Norwegian-Greenland Sea. Marine and Petroleum Geology
(Norway). Marine and Petroleum Geology 21, 363–372. 19, 79–93.
Glennie, K.W. 2009. Petroleum Geology of the North Sea, Basic Marello, L., Ebbing, J. and Gernigon, L. 2013. Basement
Concepts and Recent Advances, Wiley, 636 pp. inhomogeneities and crustal setting in the Barents Sea from
Glørstad-Clark, E., Faleide, J.I., Lundschien, B.A. and Nystuen, a combined 3D gravity and magnetic model. Geophysical
J.P. 2010. Triassic seismic sequence stratigraphy and paleo- Journal International 93, 557–584.
geography of the western Barents Sea area. Marine and Minakov, A., Faleide, J.I., Glebovsky, V.Y. and Mjelde, R. 2012.
Petroleum Geology 27, 1448–1475. Structure and evolution of the northern Barents–Kara Sea
Gudlaugsson, S.T., Faleide, J.I., Johansen, S.E. and Breivik, A. continental margin from integrated analysis of potential fields,
1998. Late Palaeozoic structural development of the south- bathymetry and sparse seismic data. Geophysical Journal
western Barents Sea. Marine and Petroleum Geology 15, International 188, 79–102.
73–102. Mjelde, R., Faleide, J.I., Breivik, A.J. and Raum, T. 2009.
Heeremans, M. and Faleide, J.I. 2004. Permo-Carboniferous Lower crustal composition and crustal lineaments on the
rifting in the Skagerrak, Kattegat and the North Sea: Evi- Vøring Margin, NE Atlantic: A review. Tectonophysics
dence from seismic and borehole data. In: Wilson, M., 472, 183–193.
Neumann, E.-R., Davies, G., Timmerman, M.J., Heeremans, Mjelde, R., Raum, T., Breivik, A.J. and Faleide, J.I. 2008.
M. and Larsen, B.T. (eds.), Permo-Carboniferous Rifting in Crustal transect across the North Atlantic. Marine Geophys-
ical Research 29, 73–87.
Mjelde, R., Raum, T., Breivik, A., Shimamura, H., Murai, Y., Ritzmann, O. and Faleide, J.I. 2007. Caledonian basement of the
Takanami, T., Brekke, H. and Faleide, J.I. 2005. Crustal western Barents Sea. Tectonics 26, TC5014.
structure of the Vøring Margin, NE Atlantic derived from Ro, H.E., Stuevold, L.M., Faleide, J.I. and Myhre, A.M. 1990.
OBS-data: Geological implications. In: Doré, A.G. and Vin- Skagerrak Graben; the offshore continuation of the Oslo
ing, B. (eds.), Petroleum Geology: North West Europe and Graben. Tectonophysics 178, 1–10.
Global Perspectives – Proceedings of the 6th Petroleum Roberts, A.M., Lundin, E.R. and Kusznir, N.J. 1997. Subsidence
Geology Conference. Geological Society, pp. 803–813. of the Vøring Basin and the influence of the Atlantic continen-
Mosar, J., Eide, E.A., Osmundsen, P.T., Sommaruga, A. and tal margin. Journal of the Geological Society 154, 551–557.
Torsvik, T. 2002. Greenland-Norway separation: A Ryseth, A., Augustson, J.H., Charnock, M., Haugerud, O.,
geodynamic model for the North Atlantic, Norwegian Jour- Knutsen, S.-M., Midbøe, P.S., Opsal, J.G. and Sundsbø, G.
nal of Geology 82, 281–298. Cenozoic stratigraphy and evolution of the Sørvestsnaget
Neumann, E.-R., Wilson, M., Heeremans, M., Spencer, E.A., Basin, southwestern Barents Sea. Norwegian Journal of
Obst, K., Timmerman, M.J. and Kirstein, L.A. 2004. Geology 83, 107–130.
Carboniferous-Permian rifting and magmatism in Southern Skogseid, J., Pedersen, T. and Larsen, V.B. 1992. Vøring Basin:
Scandinavia, the North Sea and northern Germany: A Subsidence and tectonic evolution. In: Larsen, R.M. et al.
review. In: Wilson, M., Neumann, E.-R., Davies, G.R., (eds.), Structural and Tectonic Modelling and Its Application
Timmerman, M.J., Heeremans, M. and Larsen, B.T. (eds.), to Petroleum Geology; Proceedings, Norwegian Petroleum
Permo-Carboniferous Magmatism and Rifting in Europe. Society Special Publication 1, pp. 55–82.
Geological Society Special Publication 223, pp. 11–40. Skogseid, J., Planke, S., Faleide, J.I., Pedersen, T., Eldholm, O.
Nøttvedt, A., Gabrielsen, R.H. and Steel, R.J. 1995. and Neverdal, F. 2000. NE Atlantic continental rifting and
Tectonostratigraphy and sedimentary architecture of rift volcanic margin formation. In: Nøttvedt, A. (ed.), Dynamics
basins; with reference to the northern North Sea. Marine of the Norwegian Margin. Geological Society Special Publi-
and Petroleum Geology 12, 881–901. cation 167, pp. 295–326.
Odinsen, T., Christiansson, P., Gabrielsen, R.H., Faleide, J.I. Steel, R.J. 1993. Triassic-Jurassic megasequence stratigraphy in the
and Berge, A.M. 2000. The geometries and deep structure of northern North Sea: Rift to post-rift evolution. In: Parker, J.R.
the northern North Sea rift system. In: Nøttvedt, A. (ed.), (ed.), Petroleum Geology of Northwest Europe. Proceedings of
Dynamics of the Norwegian Margin. Geological Society the 4th Conference. Geological Society, pp. 299–315.
Special Publication 167, pp. 41–57. Tsikalas, F., Eldholm, O. and Faleide, J.I. 2005. Crustal struc-
Ohm, S.E., Karlsen, D.A. and Austin, T.J.F. 2008. Geochemi- ture of the Lofoten-Vesterålen continental margin, off
cally driven Exploration models in uplifted areas: Examples Norway. Tectonophysics 404, 151–174.
from the Norwegian Barents Sea. AAPG Bull 92, 1191–1223. Tsikalas, F., Faleide, J.I. and Eldholm, O. 2001. Lateral
Olaussen, S., Larsen, B.T. and Steel, R.J. 1994. The Upper variations in tectono-magmatic style along the Lofoten-
Carboniferous-Permian Oslo Rift; basin fill in relation to Vesterålen margin off Norway. Marine and Petroleum Geol-
tectonic development. In: Embry, A.F., Beauchamp, B. and ogy 18, 807–832.
Glass, D.J. (eds.), Pangea; Global Environments and Tsikalas, F., Faleide, J.I., Eldholm, O. and Blaich, O.A. 2012.
Resources. Canadian Society of Petroleum Geologists, The NE Atlantic conjugate margins. In: Roberts, D.G. and
Calgary, Memoir 17, pp. 175–197. Bally, A.W. (eds.), Phanerozoic Passive Margins, Cratonic
Osmundsen, P.T. and Ebbing, J. 2008. Styles of extension off- Basins and Global Tectonic Maps. Amsterdam: Elsevier.
shore mid-Norway and implications for mechanisms of doi:10.1016/B978-0-444-56357-6.00004-4.
crustal thinning at passive margins. Tectonics 27, TC6016. Tsikalas, F., Faleide, J.I., Eldholm, O. and Wilson, J. 2005. Late
Osmundsen, P.T., Sommaruga, A., Skilbrei, J.R. and Olesen, O. Mesozoic-Cenozoic structural and stratigraphic correlations
2002. Deep structure of the mid-Norway rifted margin, Nor- between the conjugate mid-Norway and NE Greenland con-
wegian Journal of Geology 82, 205–224. tinental margins. In: Doré, A.G. and Vining, B. (eds.), Petro-
Ottesen, D., Dowdeswell, J.A., Rise, L. and Bugge, T. 2013. leum Geology: North West Europe and Global Perspectives
Large-scale development of the mid-Norwegain shelf over – Proceedings of the 6th Petroleum Geology Conference.
the last three million years for hydrocarbon reservoirs in Geological Society, pp. 785–801.
glacial sediments. In: Huuse, M. et al. (eds) 2012. Worsley, D. 2008. The post-Caledonian development of Svalbard
Glaciogenic Reservoirs and Hydrocarbon Systems. Geol. and the western Barents sea. Polar Research 27, 298–317.
Soc. London. Spec. publ. 368, 53–73. Ziegler, P.A. 1982. Geological Atlas of Western and Central
Ramberg, I.B., Bryhni, I., Nøttvedt, A. and Rangnes, K. 2008. Europe. Shell International, The Hague.
The Making of a Land – Geology of Norway. Trondheim: Ziegler, P.A. 1988. Evolution of the Arctic-North Atlantic and
Norwegian Geological Society, 624 pp. the western Tethys. American Association of Petroleum
Rassmussen, E.S., Heilmann-Clausen, C., Waagstein, R. and Geologists Memoir 43, 198 pp.
Edvin, T. 2008. The Tertiary of Norden. Episodes 31, 1–7. Ziegler, P.A. 1992. North sea rift system. Tectonophysics 208,
Ren, S., Faleide, J.I., Eldholm, O., Skogseid, J. and Gradstein, F. 55–75.
2003. Late Cretaceous-Paleocene development of the NW
Vøring Basin. Marine and Petroleum Geology 20, 177–206.
Chapter 26
Exploration Strategy
Hans Chr Rønnevik
26.1 Play Concepts context it is important to understand the behaviour

reference curve for the past successes and failures
The purpose of exploration for petroleum is to make of the exploration (Fig. 26.2). The successful
oil and gas discoveries that can form the basis for breakthroughs can be related more to the performance
commercial development and production of oil and of the actions rather than the predictions.
gas fields (Fig. 26.1). The formulation of ideas or Early in the exploration phase of an area the
models for where to drill to make discoveries is concepts will be referred to as frontier play concepts
based on assessments of the generation, migration (Fig. 26.3). When the models are documented by their
and trapping of petroleum. Combinations of these successful application and significant follow-up
factors are referred to as play concepts. At the very discoveries are made, they are referred to as growth
start of the exploration of an area within a sedimentary plays. When only smaller, albeit frequent, discoveries
basin the factual basis for assessing the various neces- are made they are referred to as mature play concepts.
sary factors is sparse and oil companies are confronted In the frontier phase the possibilities, assessments and
with a wide range of possible scenarios that can lead to discovery frequency are in general less than 25%, in the
petroleum discoveries. An open approach is, however, growth phase these factors are between 25–50% and in
needed in relation to what is the most likely scenario the mature phase more than 50%. The resource growth
both in the early and later phases of exploration. The for each play concept follows an S-curve.
explicit knowledge can very easily shadow for the The perception of maturity is a subjective process and
implicit potential. Exploration is a continuous learning frequent revitalisations of mature areas are experienced.
process where the key to success is found in details The S-curve also reflects the knowledge creation
that are difficult to appreciate. process and clearly demonstrates that at any time one
The play concepts should always contain, but never should harvest using the knowledge at hand while cre-
be constrained by, available facts and knowledge. ating new models to obtain a stable reserve growth over
Awareness of the unknown will always be the main time. Todays impossibilities will be tomorrow’s general
driver for generative learning. In the evaluation of play accepted truths. The explicit knowledge reflects the past
concepts a maximum number of parameters must be and the future is always in the tacit domain.
considered, and the limitations of current theories and The reality can only be revealed through drilling
technology should be appreciated. Continuously and there will in general be significant differences
improving technology has to be applied when matur- between the pre- and post-drill assessment of reality.
ing the play types and prospects in an area. In this The discovery wells need to be followed by delinea-
tion wells to clarify subsurface uncertainties and
assess the resource potential.
All parameters of unsuccessful as well as success-
H. Chr Rønnevik (*)
Lundin, Norway ful wells need to be re-analysed after drilling. The
e-mail: [email protected] main obstacle for learning is the human self-
640 H. Chr Rønnevik
referencing that can only by compensated for by

including all facts from many different geological
disciplines. The facts available at any time and new
perceptions of reality can only be corrected by drilling.
Efficient exploration of an area requires frontier,
growth and mature exploration drilling to be carried
out in parallel. Due to the size of the discoveries the
frontier exploration concepts will be the most cost effi-
cient over time (Fig. 26.4). A linear approach where one
play type or area at a time is explored can easily abort an
exploration process prematurely due to self-fulfilling
prophecies. In general 1/3 of the drilling activity efforts
should be spent on frontier prospects, 1/3 on growth
prospects and 1/3 on mature/delineation drilling.
Successful breakthroughs become rapidly obvious.
Hence, for companies to harvest on their own
breakthroughs the potential follow-up acreage should
be secured prior to start of exploration drilling activities.
Securing acreage in clusters can also lower the reserve
Fig. 26.1 Oil dripping out of a core from the Volgian reservoir threshold needed for commercial development by
in Johan Sverdrup Field. The reservoir is marine sand with
multi-Darcy permeability enabling several discoveries to constitute a larger field.
Individual companies need to have prospects
related to various play types in their portfolio, and a
200 16
180
Reserve growth from ExplWells 14
# Wildcat Wells
160
# New Discoveries 12
# Appraisal Wells
140
10
120
No wells
MMboe
100 8
80
6
60
4
40
2
20
0 0
0.5 1 0.5 0.5 0.6 0.67 0.75 0.5 0.38
Bøyla Tellus
Luno S Apollo Caterpillar Salina Luno 2
Alvheim Volund E. Grieg Gaupe Viper/Kneler J. Sverdrup Skalle Albert Gohta
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Fig. 26.2 Lundin organic growth exploration curve. The behaviour reference curve for the reserve growth shows the break-through
nature of reserve growth. The reserves are referred to the time of the exploration wells
26 Exploration Strategy 641
Truths and facts:

1. Thoughts are instant, thinking takes time (self-referencing).
2. Facts are the experienced parts of the truth. The truth must contain, but not be constrained by facts.
3. What is the likelihood of having experienced the truth based on facts?
4. The truth is so large that everybody can see a part of it but nobody the whole.
5. We should keep on unfolding reality by interacting with it (David Bohm).
- always 1.0
Psuccess Exhausted
- usually
- general 0.75
- by and large
Mature
- more often than not
- half the time 0.5
0.5
- often Many One
- sometimes truths at truth at Growth
the same the time
- occasionally time
- been known to happen 0.25
- never
Front
- cannot be true 0.1
0.0
Activity/time
Fig. 26.3 Sustainable organic growth needs a balanced exploration. Frontier, growth and mature exploration drilling should occur
in parallel
Frontier Growth Mature Stepouts Total

to be defined as drillable commercial prospects
No of wells 15 15 10 6 46 (Fig. 26.5). The critical part of the assessment process
Expected prospect size 300 100 30 20
Probability 0.15 0.25 0.5 0.75
is to construct time and space relations between gen-
Expected reserves (mmboe) 675 375 150 90 1290
eration of traps, preservation of porosity, and the
migration and retention of petroleum.
Fig. 26.4 Frontier exploration is most cost efficient. The size of There will always be alternative ways of combining
the prospects compensate for the lower discovery rate the parameters into equally likely scenarios. Possibil-
ity analyses need to reflect this by trying to define
country needs a diversity of companies that secure the many alternate ways of linking the critical parameters.
plurality necessary to unfold the resource potential. The alternate deterministic relations should reflect an
understanding of the geological processes that can
have happened.
26.2 Probability Analysis The alternative to constructing possible scientific
relations is to treat the factors as independent. The
Assessment of the inherent uncertainties in evaluating consequence of this approach results in the paradox
exploration prospects are referred to as probability or that the more factors that are used the lower the proba-
risk analysis. The analyses are concerned with the bility of success becomes. This is however a commonly
same parameters that define the various play types, used approach and can be misleading as a decision tool,
i.e. the presence of rock sequences with organic mate- as it underestimates the value of knowledge.
rial that have been buried deep enough to generate Probability analysis should be based on the aware-
petroleum of various types, and assessment of porous ness that the available facts at any time are only the
rocks or fractures where the petroleum can migrate experienced part of the truth. New actions will always
until it is retained is porous reservoirs of sufficient size uncover new facts and knowledge. Hence an
- always P success Probability assessments:

- usually
The simultaneous presence of
reservoir, generation, migration
- general and trapping of petroleum
0.75
- by and large
● The same facts can support
alternative possible models or scenarios
- more than not
- half the time ● Stochastic modelling based on independent

0.5
2007 Avaldsnes parameters is no substitute for thinking
- often after Luno; 0,45
● Decisions reflect the input that is always

- sometimes 2004 Luno; 0,38 insufficient in relation to the truth
- occasionally ● Probability analyses are time contextual

0.25
- has happened
Avaldsnes
before Luno
- never
- impossible 0.0
Fig. 26.5 Probability analysis is a subjective tool. The scien- The successful Luno (Edvard Grieg) discovery changed the
tific relationship between the various factors related to petro- Avaldsnes (Johan Sverdrup) prospect to a high possibility
leum generation, migration and trapping has to be understood. prospect
assessment of the unknown should be included in any A conscious data acquisition strategy is necessary
possibility analysis. It should be accepted that the for securing fact-based subsurface knowledge crea-
unknown is largest in the frontier phase, but the tion. The basic principles for acquisition of data
learning process should continue and significant should be:
breakthroughs can occur after long time. The analysis • Sampling of drill cuttings
should always be based on open perceptions of reality. • Coring through reservoir sequences and hiatuses
Facts and successful ruling models will always contain • Continuous processing of all petrophysical data
the risk of shadow for the possibilities. during drilling
• Sampling of fluids
• Production testing to understand reservoir
26.3 Operation continuity
During the operations it is important to continu-
In implementing knowledge-based exploration ously relate new data to a maximum number of
strategies in frontier areas the larger prospects, prefer- parameters in relation to the models that were used
ably with stacked play possibilities, are selected first as a basis for the prospect evaluation. It is imperative
for drilling (Fig. 26.6). In addition the conditional to continuously adapt the operation to the reality
links between alternative prospects should be assessed revealed during the drilling progress in order to enable
prior to selecting the drilling sequence. The first well corrective actions to be taken early. Changes in the
on a structure will be located so it can test the reserve perceptions of reality are most efficiently done by the
potential necessary for a commercial development. In people that formulated the basic concepts before dril-
addition it should secure as much new geological ling. These people have the necessary knowledge
information as possible that will be relevant for a about the simplification process behind the volumetric
rapidly improved understanding of the whole area. models and play types used.
Gohta stacked prospects

A A'
TWT
(ms)
1000
Triassic 1
Triassic 2
1500 Amplitude anomalies

Amplitude anomalies
Pe
rm
ian
2000
7120/1-1R2
7120/1-3 Gohta
7120/1-4
A'
Cross-section
A
N
Fig. 26.6 Diversity of prospect types increases the learning speed. It is preferable to test several stacked prospective levels with the
first wells in the exploration of frontier areas
26.4 Exploration Culture

between the people involved in knowledge creation
Knowledge creation is a self-organising team effort need to be based on openness and trust.
that requires a common purpose for all involved Learning is a continuous process of action and
(Fig. 26.7). The individuals should have the freedom change. Knowledge changes in steps and the way
to improve their specialty skills to realise the full forward should not be shackled by past knowledge.
knowledge creation potential of the team. This Continuous development of better practice is a must.
requires a high degree of communication and co- The skill to act should always be combined by the
operation. The greater the diversity of special strengthened will to believe in the improved concepts
disciplines involved, the greater the knowledge gener- that form the basis for stable activities. The learning
ation potential of the team becomes. The relations
Purpose
Project
Organisation/ Initiative/
Knowledge/ Information/
Structure Exformation Chaos
Fig. 26.7 Knowledge needs a common purpose and open network. The individual should have freedom to improve their specialty
within the purpose of the team
speed of the team depends on the quality of the conscious continuous drilling actions. The exploration
relations between humans. effort started out as impossibility, and the shelf or
Generative learning requires an action-oriented, parts of it have been considered mature at frequent
bottom-up culture glued together by shared visions. intervals. Renewed drilling activities based on
The shared visions must be based on a common con- increased diversity have revitalised the discovery pro-
sciousness of how reality is perceived. The basis for cess several times. The authorities have been instru-
perceptions of reality should honour the following: mental in relation to such revitalisations. The last
• Possibility analyses are influenced by the ruling revitalisation occurred around year 2000 when the
authoritative opinions of the time authorities compensated for lack of diversity of
• New knowledge is created beyond the filter of companies by inviting new players to take part in the
preconceived ideas exploration activities. This effort has lead to several
• A concept must be based on, but not limited by, all new significant discoveries in earlier assumed mature
that is considered facts areas.
• New knowledge is frequently created by actions The first play type to be established on the NCS was
• Cost benefit analysis can hinder generative learning the early Palaeogene deep sea fan play that resulted in
• Efficient learning should attack uncertainties, not the Balder oil field and the Frigg and Odin gas fields.
certainties The first definite commercial discovery was the seren-
• New awareness relates to increased diversity of dipitous revelation that the chalk in the Ekofisk For-
geological aspects and not more of the same of mation, considered to have poor reservoir potential,
what is well known actually contained producible oil.
The initial success led to a technical national man-
agement of the resources from 1972 onwards by
26.5 Exploration History and Strategy of establishing the Norwegian Petroleum Directorate
the Norwegian Continental Shelf (NPD) and Statoil as instruments for the political
management of the nation’s offshore hydrocarbon
The initial exploration phase on the Norwegian Conti- resources. An important aspect of the NPD’s resource
nental Shelf (NCS) during 1965–71 was led by a management was the proper storage and retrieval of
diversity of major companies with different geological as well as geophysical data.
perceptions of play concepts in the shallow water Areas to be allocated for future exploration drilling
parts of the North Sea basin. The unfolding of the were from then on selected by the NPD and put up for
petroleum potential of the NCS has been a continuous competitive bidding and awards among the petroleum
learning effort that demonstrates the need for companies. Work commitments were agreed through
100
Produced and discovered resources
90 Yet to find resources (NPD mid case)
High estimate 89,3
ind
80 to f
Yet
72
70,5
70
37 North of
North Sea
60
billion boe
50
40
35 North Sea
30
20
10
0
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030
Fig. 26.8 The reserve growth of the Norwegian Continental Shelf reflects breakthroughs. The breakthroughs are related to
new play types or new areas
negotiations for each new licence in order to secure the field exploration than frontier exploration, with the
necessary funding of seismic and drilling activities natural self-fulfilling result that smaller discoveries
considered necessary from a national point of view. were made and it was assumed that the NCS (Norwe-
A mixture of established and new plays where selected gian Continental Shelf) had reached maturity. At the
in the various allocations rounds. This early period same time an increasing number of mergers and
included the Statfjord allocation, and the third, fourth, acquisitions of companies took place, reducing the
fifth, sixth, seventh and eighth concession rounds. The number of operators and drilling activities on the
fourth and subsequent rounds were designed after the NCS. The authorities stimulated new activities in the
Jurassic had been established as the third major play Barents Sea by opening up acreage for strategic dril-
type, and were based on the first total resource esti- ling in the 11th round phase 2 in 1987, without any
mate of the shelf (Fig. 26.8). The fourth round, which major success. The 15th concession round in 1996
followed the third round that prematurely was consid- opened up major inverted Tertiary structures in the
ered as a disappointment, was the major breakthrough Norwegian Sea for exploration, which resulted in the
round ever in relation to reserve growth on the NCS. discovery of the Ormen Lange and Åsta Hansteen gas
Fields like Snorre, Oseberg, Gullfaks, Brage and Troll fields. At the same time the authorities initiated the
was the major outcome. Barents Sea Project with awards in 1997 to stimulate
The opening of licence areas on Haltenbanken and increased activities. The main result of this project was
in the Barents Sea in 1980, 1981 and 1982 was based the discovery of the Goliath oil field.
on national mapping by the NPD. The chosen areas The three Norwegian companies Statoil, Norsk
where selected to cover various play types in order to Hydro and Saga Petroleum were conducting 80% of
secure continuity. The same philosophy was applied the drilling activities on the NCS in the 1990s. When
on Trænabanken in 1982. The result of the initial these companies were merged into first two and then
exploration phase resulted in gas and oil discoveries subsequently one company, due to a perception that
on Haltenbanken and gas discoveries on Tromsøflaket, the Norwegian shelf was too mature to sustain many
but no discoveries were made on Trænabanken. companies, the authorities revitalised the diversity
The oil price drop in 1985–1986 resulted in a strategy pursued in the 1970s. This diversity strategy
reduction of drilling activity and more focus on near- has been a success, both in the North Sea, Norwegian
Alvheim STOOIP Alternate oil columns

0.040
Kamelon East Kamelon Heimdal field
0.035 No extra wells
Kneler well
0.030
Boa well
Kneler
0.025 Kameleon-East well Boa
Gekko
0.020 5m
30m
17m
0.015 60m
0.010 Result
0.005
Hamsun
0.000 Kobra
N 5 km
50 120 190 260 330
GOC corresponding with observed far offset anomaly
Alternate volumes (mill bbls) based on experienced facts
Inline
Heimdal
Distribution for Kneler Oil Column
Cro
0.1
sslin
0.08
Cro
0.06
e
sslin
0.04
Result
e
0.02
0
-10 0 10 20 30 40 50 60 70
Variation in potential oil columns

AVO anomalies
Inline
Fig. 26.9 The Alvheim Field is the story of gas satellites that became a large oil field. The exploration drilling proved oil reserves
beyond two standard deviations in the pre-drill model for the prospects
Sea and Barents Sea. In all areas fourth generation seismic indications interpreted as gas AVOs. The post-
exploration has resulted in new successful play types drill results demonstrated the shortcomings of a single
that clearly demonstrate that the maturity of the shelf method. Additional oil discoveries were the spinoff of
was fiction due to lack of exploration drilling. rapid learning from these results.
Examples of new discoveries are the Alvheim, Edvard The Haugaland High (Utsira High South) was
Grieg, Ivar Aasen, Johan Sverdrup, Johan Castberg, explored for 40 years before the breakthrough discov-
Wisting and Gohta fields and discoveries. ery was made (Fig. 26.10). The first well, Esso’s 16/2-
Prior to the Alvheim oil field discovery in 2003 the 1, was drilled in 1967 and had good oil shows in the
prospects were assumed to be gas tie-in satellites to the Tor Formation and Basement. These shows were later
Heimdal field (Fig. 26.9). The exploration drilling referred to as the Ragnarrock discovery and delineated
resulted in a significant standalone oil field consisting by Statoil in 2007 by the drilling of 16/2-3 and -4
of three discoveries that became the hub for other oil wells. The delineation concluded that the chalk and
tie-in satellites. This demonstrates the importance of basement reservoirs in this area have limited commer-
defining the resource potential in a subsurface-based cial potential.
context and not in relation to what could be nice to The prolific hydrocarbon nature of the Haugaland
have from a production point of view. The drilling of High area was demonstrated by the Luno 16/1-8 dis-
the Kneler prospect within the Alvheim Field proved covery (Edvard Grieg) in 2007, the Draupne 16/1-9
an under-saturated non-biodegraded oil discovery discovery (Ivar Aasen) in 2008, the Avaldsnes 16/2-6
amidst the saturated Gekko and Kameleon and later discovery (Johan Sverdrup) in 2010 (Fig. 26.11) and
Boa oil discoveries. The discoveries were drilled on Luno 2 16/4-6 discovery in 2013.
18
2013 - Luno II Discovery
BP Amoco
Elf
16
Det Norske
Wintershall
14
Norsk Hydro
Esso
Statoil
12
Lundin
2010 - Johan Sverdrup Discovery

Number of wells
10
2008 - Ivar Aasen Discovery

8
2007 - Edvard Grieg Discovery

6
0
1967 1972 1977 1982 1987 1992 1997 2002 2007 2012
Fig. 26.10 The learning curve demonstrates 40 years of with the presence of oil in basement. The Edvard Grieg discov-
disappointments before the first discovery on the southern Utsira ery 2007 opened up for the discovery of the Johan Sverdrup
High. The result from the first well in 1967 was encouraging discovery in 2010
The Edvard Grieg and Johan Sverdrup discoveries fringing the high in the west were defined. The
on the Haugaland High contain under-saturated non- prospects and leads were assumed to be filled by
biodegraded oil originating from the same Draupne saturated light oil similar to the 16/1-7 West Cable
formation source rocks. These discoveries flank the discovery.
saturated biodegraded Ragnarrock discovery at the Following the discovery successes the strategy has
crest of the high. involved the application of continuously improved
The pre-drill possibilities for oil in the Johan geophysical and geological databases and interpreta-
Sverdrup discovery were considered remote by most tion techniques to refine the new play concepts. New
companies, but the possibility of an oil-filled reservoir 3D seismic acquisition techniques used in 2008, 2009
was backed up with seismic indications of and 2012 have given continuously better pictures of
hydrocarbons venting vertically into the overlying the subsurface (Fig. 26.12).
Utsira Formation followed by lateral migration up- The Edvard Grieg discovery might have been
dip in a westward direction. This observation linked overlooked without coring and testing, due to the
the Edvard Grieg and Johan Sverdrup prospects. Fol- feldspar-rich nature of the reservoir rocks
lowing the discovery of Edvard Grieg in 2007 the (Fig. 26.13). The increased understanding of the
Johan Sverdrup prospect became a high probability hydrocarbon habitat of the southern Utsira High
prospect which was proven in 2010. implies a reassessment of the geological history from
The initial play model employed on the Haugaland Palaeozoic to Pleistocene. The main learning concerns
High was based on the presence of a thin veneer of recent and on-going oil migration. The Johan Sverdrup
Volgian sand, older inlier basins and fractured porous structure (Fig. 26.14) obtained its current structural
basement shallower than 1940 m MSL on the high position during the last 1.5 million years, which
itself. In addition, structural and stratigraphic traps constrains the time available for migration and
J72 basinal muds W Oil

J72 basinal muds J72 basinal muds
Draupne Sandstone J71 E. Utsira Lundin Norway Licenses
J72 Deep marine muds Partner
J72 Draupne Spiculite Operator
Basement subcrop
Zechstein Faults
1940-50 m MSL
1900 m MSL
A
Basement
A'
0 10 km
Hypothesis: Saturated system with 40-50 m oil leg.
A A'
NW 16/1-14 16/1-8 16/2-5 16/2-6 SE
1000
Johan Sverdrup Discovery
Ragnarock Discovery
Edvard Grieg Chalk
Apollo Discovery Rødby/Sola
2000
Paleocene
Utsira High Paleozoic
Jurassic/Triassic
3000
Fig. 26.11 The basic assumption was sufficient. The drilling demonstrated under-saturated and saturated oil in new reservoir types
on the NCS. The oil fill of the reservoirs is late and is still ongoing
trapping of the same oil type into both fields. The Middle Jurassic inversion that stripped off the Lower
varying oil/water contact between 1922 and 1935 m Jurassic and older sequences down to basement over
MSL and deeper residual oil in the Johan Sverdrup large parts of the high. There are also marked tectonic
discovery demonstrates the effect of glacial down events in the Callovian and Kimmeridgian that have
warping and rebounds during the Pleistocene. influenced the presence of reservoir sequences.
Both the Edvard Grieg and Johan Sverdrup The main Volgian reservoirs in the Edvard Grieg
discoveries contain reservoirs that are new on the discovery are proximal continental deposits and con-
NCS. In both discoveries Volgian reservoirs are the sist of high quality eolian sand, good quality braided
main reservoirs that are overlying older reservoirs by a river and alluvial fan sands. There are also proximal
major regional unconformity. The unconformity was coarse sandy conglomerates of Triassic age, weath-
created by large-scale tectonic epeirogene basin inver- ered basement gravel and Valanginian shell fragment
sion. Renewed weathering and erosion sourced the sand. All the reservoir types have been production
various Volgian sandstone reservoirs in the two tested.
discoveries. The Volgian inversion on the Haugaland The main reservoir in the Johan Sverdrup Field is
High is superimposed on the thermal-driven early the shallow marine Volgian sand with extreme
Edvard Grieg Discovery
Original
3D BroadSeis/BroadSource
2008 2012
2D Benchmark Line
3D OBC
Edvard Grieg Discovery
3D GeoStreamer Broadseis
2009
Fig. 26.12 The vintage 3D has been replaced by broadband 3D after the discoveries. The broadband 3D seismic gives a holistic
picture of the shallow and deep geology simultaneously
16/1-8 formation evaluation Logs read 40% shale in 100% sand

Chalk
Reservoir
Gamma log from core no 1. This is what it looks like
16/1-8 Pressure (bar) The First Oil
Oil
y= 13,942x - 780,28 Water
R² = 0,9994 All
1,0583
PVT
Linear (Oil)
Linear (Water)
Oil
Water
Depth (M)
y= 9,6528x + 64,362
R² = 1
Fig. 26.13 Coring was a necessary trigger for the discoveries on the Utsira High. Indirect data types (wireline logs) need to be
calibrated. Without cores the Edvard Grieg discovery could have been overlooked
SW 16/2-15 16/2-11 16/2-11A 16/2-6 T2 16/2-13 S NE
1840
Volgian 16/2-6
1922
1920 Pre Volgian Discovery well
1934
2000 Z-scale=10
SW
16/2-6
-
NE
Fig. 26.14 The Johan Sverdrup Field was found by well 16/2-6 in 2010. The discovery well was located to give a maximum
amount of information. The delineation drilling took place in 2011, 12, 13 and 14
Wisting discovery
Johan Castberg Field

Skalle discovery
Gohta discovery
Salina discovery
Lundin Norway Licenses: Partner Operator
Fig. 26.15 The Barents Sea has experienced breakthrough in relation to oil potential. The ongoing oil filling is compensating for
the leakage during the ice ages
permeability documented by the seven production discoveries is ongoing active oil migration that
tests. Beneath the Volgian unconformity there is Mid- compensates for leakage of oil during the glaciation
dle to Upper Jurassic estuarine to fluvial sand periods. The indication of this process has been there
sequences with good reservoir qualities. Beneath the for a long time, but the possibilities were shadowed for
base Middle Jurassic unconformity there are marine to by the ruling view that focused on the leakage problem
continental Lower Jurassic sands of very good quality. and not on the active ongoing petroleum migration
Zechstein karstified limestone and dolomite are also which has been documented by sampling. The
present as reservoirs. reservoirs are Late Triassic to Middle Jurassic
The Luno 2 discovery is a Late Triassic/Jurassic except for Gotha with karstified Permian carbonate
inlier basin with reservoirs in fluvial facies sandstones. reservoirs.
This oil discovery is sourced from the Heather/ The new discoveries on the NCS related to both old
Draupne formations. and new play concepts since 2003 demonstrate that the
The Volgian inversion on the Haugaland High is exploration process should always be open in relation
reflected regionally on the NCS and formed the Mid to which parameters are necessary. The seed for new
North Sea High, Utsira High, Tampen High, Manet knowledge will always be outside the bandwidth of
High, Frøya High, Nordland Ridge, Loppa High and preconceived ideas and the truth can never be voted on
Fedinsky High. Hence analogue plays have and will be by majority or size. Diversity and not more of the same
found around these highs. This demonstrates the is needed in generative knowledge creation. The fol-
importance of having a regional geological perspec- lowing wording of Piet Hein is a good guidance for
tive in formulating efficient exploration strategies. exploration; “To know what thou knowest not is in
Renewed exploration of the Barents Sea resulted in essence omniscience”.
the discoveries of the Johan Castberg field in 2011 and Updates on the petroleum discoveries and produc-
the Gohta and the Wisting discoveries in 2013 tion are found on: www.NPD.no (Norwegian petro-
(Fig. 26.15). Common for all these light oil leum directorate)
Index
A Atlantic, 9, 10, 15, 65, 68, 87, 114–116, 154, 182, 184, 188, 193,
Accommodation space, 20, 248–250, 266, 267, 343 219, 225, 227, 262, 264, 266, 267, 281, 324, 605, 606,
Acoustic logs, 424 615, 621–624, 626, 633
Acquisition fold, 433 Attenuation, 410, 433, 436, 440, 446, 543
Acritarchs, 234–236 Authigenic, 59, 65, 96, 98, 101, 103, 110–111, 113, 123,
Aeolian, 20, 30, 32, 38, 40, 51, 52, 58, 63, 64, 77, 78, 87, 111, 125–127, 131, 133, 134, 136, 137, 139, 140, 207, 228,
115, 139, 188, 192, 209, 219, 223, 226, 399, 424, 612 250, 290, 295, 353–356, 383, 468, 614, 619
Air-gun, 255, 427, 430–433, 435 Autocorrelation, 440
Algae, 4, 6, 7, 9, 24, 27, 58, 65, 77, 104, 106, 122, 124, 151, 152, AVO. See Amplitude versus offset (AVO)
154, 155, 157–165, 168, 170, 171, 176–179, 181–187,
190, 192, 194, 199, 201, 202, 212, 222, 234, 236,
361–363, 365, 585–587, 615, 617 B
Alkanes, 368–370 Backstripping, 265, 266, 302
Alluvial fans, 59–62, 334, 648 Bahamas, 21, 77, 81, 157, 179, 181, 182, 184, 187–190, 192,
Alvheim oil field, 646 198, 202, 207, 209
Amorpheous silica, 257 Bandpass filter, 440
Amplitude map, 453, 456, 482, 494 Bandwidth, 430, 446, 519, 651
Amplitudes, 201, 259, 346, 404, 422, 423, 428, 432, 433, 440, Barchans, 63, 64
447–451, 453, 455, 456, 460, 465, 473, 476, 478, 479, Barents Sea, 14, 154, 160, 184, 483, 484, 534–538, 599,
482, 486, 489–495, 501, 503, 504, 506, 508, 509, 516, 603–606, 622, 626–634, 645, 646, 650, 651
534, 535, 626 Barrier island, 16, 17, 20, 76, 78–80
Amplitude versus offset (AVO), 257, 450, 453, 455, 469, Basaltic rocks, 101, 111, 538
473–486, 501, 503 Base Cretaceous unconformity, 536, 537, 614
Anhydrite, 15, 192, 193, 207, 223, 226, 283, 344, 397, 405, 407, Basement, 15–17, 25, 61, 94, 139, 263, 266, 268, 273, 275, 276,
412, 413, 555 293, 295, 306, 319, 321, 332, 357, 501, 502, 530, 538,
Ankerite, 152, 198, 226, 597 541, 542, 604, 608, 613, 614, 619–621, 626, 646–648
Anthracite, 413 Basin analysis, 33, 90, 121, 145, 255–271, 351
Antidunes, 52, 53 Basin modelling, 2, 145, 271, 277, 301, 340, 351, 367, 455
API gravity, 369, 370, 582 Beach rock, 198, 199
Apparent anisotropy, 526–527, 536, 537 Bedforms, 42, 50–54, 63, 74, 83, 190
Appraisal well, 22, 534, 535, 549 Bernoulli, 42
Aquifers, 125, 126, 276, 281, 287, 289, 295–297, 489, 552, 575, Binning, 435
591, 593, 596, 616 Biodegradation, 18, 135, 281, 371, 383, 554, 584
Aquitards, 287, 297 Biosparite, 203
Aragonite, 19, 22, 96, 112, 123, 139, 143, 151, 152, 155, 157, Biostratigraphy, 23, 231, 233–240, 242–244, 561, 562
159, 163, 165, 167, 171, 172, 176–179, 181–183, 185, Biozones, 232, 238–240, 242
187, 189, 193, 195–200, 202, 203, 206, 207, 209–214, Bitumen, 2, 10, 133, 135, 138, 581–585
228, 246, 353, 358, 359 Bjørnøya, 622, 626, 627, 629, 631, 632
Archie’s law, 417 Black shales, 10, 14, 15, 123, 213, 229, 250, 273, 361, 368, 397,
Arctic, 160, 222, 626, 632 587, 588, 613
Arenite, 119, 120, 130, 142 Blue-green algae, 7, 58, 154, 182, 201
Arkose, 119, 139, 266 Boa oil, 646
Aromatic, 369–371, 581 Borehole break-out, 308
Arrhenius, 132, 366 Borehole televiewers, 422, 423
Asphalt, 8, 18, 369, 370 Bottom hole temperature (BHT), 391
Athabasca, 554, 582–584 Bouma sequence, 46
DOI 10.1007/978-3-642-34132-8, # Springer-Verlag Berlin Heidelberg 2015
654 Index
Braided stream, 20, 66 Carnallite, 401

Break-up, 324, 328, 336, 603, 610, 615, 621, 623, 625, 626, 632, Cased-hole logs, 389, 390, 424
633 CCD. See Carbon compensation depth (CCD)
Brent, 126, 131, 143, 489, 490, 613, 614, 617–620 CCS. See Carbon capture and storage (CCS)
Brent delta, 614, 617, 619 CDP. See Common depth point (CDP)
Brightspot, 455, 476 Cellulose, 6, 10, 363
Brittle deformation, 324 Cementation, 17, 22, 122, 128–130, 132–135, 138–144, 146,
Bryozoa, 171–173, 212, 615 147, 168, 180, 182, 195–200, 210–214, 255, 276, 288,
Burial diagenesis, 91, 104, 137, 142, 144, 146, 200–202, 226, 295, 297, 298, 301, 302, 306, 311–313, 316, 317, 354,
352, 353 357–359, 376, 378, 384, 396, 418, 458, 462, 465–468,
479, 492, 613, 614, 625, 634
exponent, 417, 418
C Cenozoic, 154, 166, 170, 175, 237, 242, 243, 367, 585, 603,
Calcareous algae, 77, 124, 152, 157–160, 164, 168, 176, 177, 615–616, 620, 624, 626, 632–633
179, 184, 185, 187, 192, 194, 199, 212 Central Graben, 162, 282, 332, 336, 607, 608, 610, 612,
Calcisphere, 162, 163 617–619
Calcite, 24, 56, 93, 95–99, 112, 123, 124, 130, 137–139, 143, Chalk, 14, 16, 19, 56, 160–162, 194, 212, 213, 221, 310, 317,
151–153, 159–161, 165, 167, 170, 172, 175–179, 182, 323, 358, 489, 498, 499, 588, 604, 615, 617, 644, 646
183, 185, 189, 192–200202–204, 206–215, 226–229, Chamosite, 103, 123, 209, 227, 613
234, 246, 283, 290, 295, 353, 358, 359, 405, 413, 463, Channel flow, 41–42, 54, 75, 80
596, 597, 617 Chemical compaction, 129, 132, 133, 140, 145, 203, 212–214,
Caledonian, 326, 585, 603, 608–610, 622, 626 257, 296–298, 301, 302, 304, 305, 311, 312, 314–316,
Caledonides, 609 352, 354, 356–359, 467, 469, 477, 483, 484, 634
Caliche, 60, 226, 613 Chert, 56, 119, 167, 207, 220–222, 412, 557
Caliper log, 393–394, 400 Cherzy equation, 42, 61
Capillary Chlorite coating, 123, 133, 134, 141, 142, 313
entry pressure, 127, 145, 375–379, 383, 384, 546, 547, Chlorozoan association, 184
593, 596 Chronostratigraphy
forces, 11, 12, 61, 62, 92, 111, 195, 197, 287, 375, 376, 379, minerals, 231
381, 383, 545–548, 553, 554, 556, 617 sediments, 240, 242, 243
residual trapping, 595, 596 Clay minerals, 33, 36, 39, 40, 47, 49, 56, 73, 92, 94, 100–107,
resistance, 11, 14, 375, 377, 381, 384, 546–548, 552 109–112, 114, 119, 120, 127, 130, 132, 144, 186, 188,
Cap rock, 1, 12, 14–16, 18, 21, 33, 145, 223, 225, 267, 328, 375, 201, 206, 208, 217–219, 221, 256, 279, 281, 282, 290,
378, 379, 381, 382, 384, 424, 468, 473, 475–477, 493, 313–316, 351–353, 355, 356, 358, 359, 365, 401, 404,
507, 549, 557, 591, 593, 598, 599, 601, 620 405, 418, 464, 468, 471, 548, 554–556, 587, 597, 617
Carbonate Closure, 14, 15, 26, 319, 324, 342, 382, 617, 620
classification, 179, 202–204 CMP. See Common midpoint (CMP)
compaction, 351, 358–359 Coal, 2, 5, 7, 13, 24, 26, 33, 51, 59, 65, 75, 95, 234, 242, 363,
environments, 180, 183–209 365, 397, 401, 407, 413, 414, 424, 553, 562, 585, 587,
minerals, 65, 94, 97, 99, 151, 152, 185, 210, 281, 284, 588, 591, 592, 600, 604, 610, 611, 614, 617, 625, 633
359, 401 Coal bed methane (CBM), 587
platform, 21, 183, 188–193, 195, 201, 209, 611, 628 Coarsening-upwards, 21, 59, 66, 73, 75, 77, 78, 249, 250, 376,
reservoirs, 1, 4, 20–22, 33, 119, 210–214, 317, 489, 548, 378, 382, 384, 399, 424, 552
556–557 Coccolithophore, 154, 157, 159, 160, 162, 164, 176, 193, 194,
skeleton, 151, 210 201, 212
turbidites, 193 Coccoliths, 152, 160, 161, 164, 177, 193, 246, 310
Carbon capture and storage (CCS), 591–601 Coccosphere, 160
Carbonate compensation depth (CCD), 113, 114, 165, 193, Coherent noise, 439–441
264, 270 Collapse breccia, 207, 211
Carbon dioxide (CO2), 5, 7, 8, 10, 24–26, 29, 95, 98, 99, 104, Common depth point (CDP), 433, 441, 442, 482
105, 111, 113, 114, 124, 135, 137, 151–155, 158, 179, Common midpoint (CMP), 433, 435, 440–443, 533
185, 190, 192–194, 202, 213, 222, 245, 281, 283, 296, Compaction(al), 1, 10, 15–17, 20, 21, 31, 40, 45, 48, 49, 55, 56,
361, 363, 364, 367–371, 489, 506–507, 553–555, 582, 70, 73, 78, 80, 84, 121–122, 127–130, 132, 133, 135,
584, 587, 589, 591–601, 616 138–140, 142–146, 196, 199, 200, 202–204, 208,
capture, 591, 592, 599–601 210–214, 226, 227, 255, 257, 260, 265, 276, 290,
storage, 24, 489, 506, 553, 592–594, 596–600 294–299, 301, 302, 304–318, 327–329, 337, 351–359,
Carboniferous, 65, 165, 171, 172, 175, 262, 374, 452, 588, 600, 373, 407, 408, 410, 419, 448, 458, 459, 466–469, 477,
604, 606, 610–611, 617, 626–628, 633 483, 484, 491, 498–500, 505, 508, 551, 552, 563, 616,
Carboxyl acid, 363 618, 634
Index 655
driven flow, 202, 275, 276, 281, 288, 293–294, 297–299, 552 Desert, 47, 61–64, 87, 104, 106, 107, 110, 111, 122, 125, 139,
trends, 353, 354, 358, 359, 405, 407, 410, 466–469, 484 192, 218, 224, 226, 227, 262, 431
Compression, 55, 84, 268, 286, 301, 304, 306, 307, 309–310, Devonian, 63, 65, 112, 162, 167, 168, 175, 184, 326, 374, 585,
315, 337, 338, 340, 345 587, 588, 607–610, 613, 617, 626, 633
Compressional wave, 427 DHI, 533–535, 542–544
Compton scattering, 396, 402, 411 Diagenesis, 1, 10, 20, 21, 28, 31, 36, 59, 91, 98, 104, 120–122,
Concretions, 56, 123, 124, 211, 213, 221, 227, 228 136, 137, 139–142, 144–146, 152, 162, 168, 195–203,
Conductivity, 127, 267, 273, 274, 276, 286, 292, 385, 392, 413, 211–214, 217, 226, 245, 246, 267, 277, 287, 289, 290,
415, 417, 592, 593 316, 351–353, 382, 384, 429, 458, 459, 468, 472, 477,
Conductivity log, 413–420 483, 512, 548, 558, 561, 563, 564, 615, 634
Constrained inversion, 524 Diagenetic processes, 1, 20, 33, 36, 39, 90, 91, 94, 119, 121,
Continental rifting, 266–267, 324 122, 124, 133, 138, 140–143, 179, 196, 199, 282, 287,
Contraction, 55, 56, 263, 265, 286, 304, 319, 320, 323, 327, 329, 292, 301, 634
340, 342, 345, 347, 608, 615 Diagenetic trends, 458, 459, 465, 472
Controlled source electromagnetic (CSEM), 504, 515–544 Diatoms, 7, 63, 65, 106, 107, 112, 123, 124, 179, 192, 193,
Convections, 138, 273, 276, 284, 290–293, 297, 321, 595 219–222, 228, 234–236, 283, 315, 354
Conventional wireline log, 389, 390 Differential subsidence, 330, 613, 622
Convolute lamination, 54 Diffraction, 260, 435, 443, 444
Coral, 21, 84, 154, 159, 162, 168–171, 178, 184–187, 189, 190, Digital reservoir models, 559
192, 199, 210 Dipmeter logs, 20, 391, 420–423
Core description, 563 Dip-moveout (DMO), 439, 440, 442, 443
Cores, 1, 2, 19, 22, 33, 49, 82, 123, 127, 129, 136, 142, 145, 146, Dish structures, 54
182, 206, 233, 238, 246, 283, 288, 294, 330, 347–349, Distribution coefficients, 92, 95–96
362, 379, 385, 406, 419, 422, 423, 425, 452, 458, DMO. See Dip-moveout (DMO)
464–466, 492, 501, 506, 548, 556, 562, 563, 567, 569, Dolomite, 93, 152, 157, 183, 192–194, 197, 202, 205–212, 214,
576, 577, 583, 620, 649 224–226, 234, 281, 302, 397, 401, 404, 405, 407, 408,
Cretaceous, 9, 10, 14, 21, 65, 81, 97, 100, 112, 114, 124, 126, 410, 412–414, 588, 597, 600, 627, 632, 651
140, 161, 162, 164, 181, 182, 184, 188, 193, 194, 221, Dolomitisation, 168, 205–208, 214, 234
225, 226, 240, 242, 247, 262, 267, 271, 282, 296, 298, Domes, 3, 15–16, 223, 267, 274, 275, 283, 292, 624
326, 329, 354, 355, 382, 536, 537, 554, 582, 584, 587, Downward extrapolation, 444
588, 604, 607, 614–615, 617, 618, 620–626, 631–634 Draupne, 362, 646
Crinoids, 173–175, 198, 199 formation, 614, 617–619, 625, 647, 651
Critical point (Crp), 551, 593, 594, 599 Drilling muds, 22, 23, 282, 353, 387–389, 391–395, 401, 410,
Crossplot technique, 412 413, 416, 417, 419, 422, 423, 557
Crude, 2, 293, 369, 370, 378, 568, 584 Dunes, 20, 38, 43, 48, 50–53, 63, 77, 190, 226
CSEM. See Controlled source electromagnetic (CSEM) Dunham’s classification, 202, 204–205
Cuttings, 1, 3, 22, 23, 33, 145, 233, 238, 246, 314, 334, 385, 387, Dunlin, 613, 618, 619
388, 642 Dynamic reservoir model, 559–560, 575–579
Cyanobacteria, 7, 154–159, 163, 164, 177, 178, 182, 192, Dynamic simulation model, 568, 575
207, 224
E
D Early–Middle Jurassic sandstones, 537
Darcy, 19, 25, 275, 276, 286, 289, 291, 294, 297, 376, 549 Earth resistivity model, 518, 521, 523, 524, 527
flow, 285–288 East Shetland Basin, 606, 608, 610, 618
Data acquisition, 424, 434, 437, 519–523, 642 East Shetland Platform, 607, 608, 610
Dead sea, 172, 179, 194 Echinoderms, 152, 173–175, 177, 184, 187, 192, 197, 198
Debris flow, 44, 47–48, 50, 59, 60, 84, 87, 193, 212, 219, 614 Echinoids, 173–175, 177
Deconvolution, 439, 440, 446 Effective stress, 21, 48, 49, 73, 103, 121, 128, 129, 132, 134,
Dedolomitisation, 208 142, 143, 145, 146, 196, 200, 214, 257, 293–298,
Deformation mechanisms, 323–324 301–307, 309–317, 351, 352, 356–359, 363, 472, 492,
Delamination model, 326, 330 551, 552, 558, 617, 618, 634
Delta Ekofisk, 16, 267, 310, 317, 498, 604, 614, 617, 618, 644
front, 59, 68–74, 218, 258, 376, 451, 619 field, 19, 161, 162, 194, 212, 310, 358, 498, 548, 551, 552,
lobe, 70, 71, 86, 258, 452 604, 617
Deltaic, 7, 32, 58, 82, 121, 127, 237, 238, 271, 290, 413, formation, 498
614, 629 Electrical anisotropy, 526, 527
Deltaplain-mudstone, 568 Electrokinetic potential, 394, 395
Density log, 256, 391, 402–405, 407–414, 418, 424, 466, 484, Electromagnetic, 2, 353, 504, 515, 516, 520, 523, 525, 527–529,
562, 579 538, 539
Depth migration, 444, 504 Electromagnetic acquisition, 522
656 Index
Enhanced oil recovery (EOR) project, 600–601 Flute cast, 46, 53–55
Enhanced recovery, 1, 553–557 Fluvial deposits, 60, 65, 67
Ephemeral, 60, 65, 222, 225 Fluvial flow, 45, 46, 71
Erosion, 4, 21, 26, 31, 38, 41–43, 46, 49–54, 58–68, 70, 73, Fluvial reservoir zone, 571
75–77, 79, 80, 83–85, 91, 100, 104, 110–111, 113, 115, Folk’s classification, 202–204
116, 127, 137, 139, 144, 180, 182, 190, 217, 218, 226, Foraminifera, 7, 19, 100, 112, 124, 162, 164–166, 184, 185, 192,
228, 238, 248–250, 252, 259–261, 263, 266, 275, 276, 193, 198, 234, 236, 237, 240, 245, 615
283, 290, 301, 305, 306, 318, 321, 335–337, 342, 344, Foramol association, 184–185
346, 364, 375, 407, 448, 451, 452, 484, 607, 611, Fore-arc basin, 87, 120, 270, 271
614–618, 628, 629, 632–634, 648 Formation factor, 279, 417
Evaporate, 8, 61, 97, 192, 193, 207, 222, 224, 225, 371 Formation resistivity factor, 417–418
Evaporites, 50, 65, 94, 95, 98, 113, 114, 156, 161, 179, 192, 207, Formation water, 279, 281–284, 292, 303, 387, 388, 395, 396,
211, 217–229, 234, 246, 266, 267, 281–284, 294, 299, 415, 417, 418, 472, 516, 593–598
303, 344–347, 397, 401, 405, 412–413, 415, 542, 594, 4C seismic, 432, 435, 438
612, 626–628, 633 4D seismic, 2, 21, 23, 437–438, 489–512, 557, 562, 598, 600
basins, 64, 65, 114, 183, 193, 222, 223, 225, 226, 267, 612 Fractured reservoirs, 124, 146, 211–212, 221, 547, 556–557
minerals, 207, 222, 344, 412 Fractured rocks, 379, 546, 549
Exhumation, 21, 615 Fracture pressure, 21, 295–298, 303, 307, 308, 356, 374,
Experimental compaction, 127, 128, 142, 143, 145, 312–315, 378–382, 384, 501, 598, 620
353, 359 Fresnel zone, 445, 446
Exploration, 1–3, 17, 20–25, 27, 28, 31–33, 73, 116, 121, 145, Froude number, 42, 43, 52, 53
233, 237, 271, 279, 283, 288, 305, 307, 317, 319, 321,
328, 330, 342, 343, 364, 365, 367, 383, 385, 394,
427–486, 498, 504, 507, 515, 516, 528, 531, 534, 536, G
540, 542, 545, 549, 561, 603–605, 616–620, 624, 625, Gamma logs, 3, 397, 424
633–634, 639–651 Gamma ray log, 396–401, 407, 466, 484
Extension, 21, 33, 133, 211, 282, 288, 304, 306, 309, 319, 321, Gas
323, 326–332, 334, 335, 338, 343, 345, 357, 378, 502, effect, 401, 409
532, 538, 607, 612, 618, 621, 622, 626, 631, 633 hydrates, 24, 581, 588–589
Extensional basin, 324–338, 340, 626, 627 injection, 553, 555, 618
Gassmann model, 491
Gauge, 393, 394
F Gaussian simulations, 573
Facies modelling, 569, 571–572 Geo-cellular model, 567
Faecal pellets, 82, 180–181, 190 Geochemistry, 2, 47, 64, 90–116, 151–152, 361–371, 561, 600
Far stack, 475, 476, 478 Geological framework, 565–567
Fault, 2, 16–17, 54, 106, 119, 211, 221, 260, 268–270, 275, 279, Geological modelling grid, 568, 569
294–295, 306, 319, 347–349, 359, 368, 421, 434, 490, Geomechanic, 301–318, 422, 498–501, 505, 512
525, 552, 559, 593, 607 Geomechanical effects, 498–501
Fault architecture, 347–349 Geometric spreading, 433
Feathering, 435 Geophone, 2, 3, 429–432, 434–436, 439, 600
Feldspar, 19, 33, 39, 68, 100, 101, 103–107, 110, 119, 120, Geophysical monitoring, 504–505
124–127, 130, 136, 137, 139–141, 143, 144, 146, 207, Geostatic pressure, 557
217–219, 246, 290, 312, 315, 355, 383, 397, 400, 413, Geostatistics, 569, 572, 573, 579
463, 467, 596–598, 614, 619, 625, 647 Geothermal gradient (gG), 12, 89, 128–130, 137, 139, 141, 143,
Fining-upwards, 20, 66, 67, 80, 82, 85, 377, 378, 384, 396, 399, 144, 220, 257, 263, 265–267, 270, 271, 273–277, 291,
424, 553 292, 314, 321, 342, 343, 356, 363, 365–368, 392, 419,
F-K filter, 440, 441 589, 594, 614, 619
Flaser bedding, 52, 82 Gibbsite, 107, 111, 127, 217, 218, 282
Flett delta, 538, 539 Glacial, 9, 60, 63, 64, 97, 110, 115, 189, 194, 196, 218, 248, 249,
Flocculation, 44, 64, 75, 218, 314 276, 304, 314, 315, 353, 354, 589, 604, 616, 620, 624,
Flow simulations, 493, 565, 574, 575, 577, 578 633, 634, 648
Fluid Glitne field, 456, 465, 481, 482
communication, 279, 284, 348, 387, 557 Goliath field, 633, 645
factor, 477–480 Grain flow, 48
inclusions, 135–137, 207 Grain size, 31, 32, 34–40, 43, 45, 49–51, 53, 60, 63, 66, 68, 90,
pressure, 11, 279, 285, 288, 295, 299, 301–304, 307, 310, 107, 119, 127, 128, 143–145, 176, 179, 181, 202, 204,
374, 378–380, 501, 552 212, 214, 217, 218, 246, 299, 312, 315–317, 351, 352,
Fluid (hydrodynamic) potential, 195, 285–288, 303 356, 396, 399, 424, 465, 564, 596
Flushed/transition zones, 126, 387–389, 391, 406, 416, 418, Grapestone, 189
419, 432, 469, 477 Gravity flow, 44, 212, 335
Index 657
Green river formation, 65 Invertebrates, 164–179

Green river shale, 4, 585, 586 Ionic potential, 92–94
Greywacke, 120, 139, 204, 397 Island arcs, 87–89, 115, 116, 270, 271
Groningen field, 603 Isostasy, 263, 302, 329
Ground water, 281, 587, 588 Isotopes, 92, 96–100, 245, 284, 294, 366, 369, 396
Growth faults, 15, 17, 73, 74, 126, 345, 377, 379
Gullfaks field, 146, 489, 490, 501, 560
Gypsum, 15, 157, 161, 192, 193, 207, 208, 222–226, 234, 279, J
282, 283, 344, 371, 405, 412, 413, 555 Johan Sverdrup Field, 640, 646, 648, 650
Jurassic, 11, 13, 14, 21, 51, 63, 65, 99, 100, 124, 126, 129–131,
135–137, 140, 162, 164, 167, 168, 182, 188, 194, 211,
H 225, 242, 266, 267, 271, 276, 277, 282, 317, 335, 353,
Halite, 15, 192, 207, 222, 224, 225, 274, 283, 299, 344, 397, 354, 361, 383, 397, 491, 530, 536, 537, 599, 604, 606,
405, 407, 412, 413, 594, 599 607, 612–614, 617–619, 621, 622, 625, 626, 629–634,
Halokinesis, 319, 321, 336, 346, 607 645, 648, 651
Haltenbanken, 14, 100, 134–137, 294, 353, 383, 605, 624–625,
634, 645
Hardground, 199, 212, 221, 250 K
Haugaland High, 646, 647, 651 Kattegat, 609, 612, 615
Heat flow, 89, 263, 270, 273–277, 294, 329, 363, 542 KCl-based muds, 400
Heat transport, 273–277, 294 Kerogen, 10, 13, 26, 27, 213, 266, 283, 284, 298, 351, 356,
Heavy oil, 3, 18, 24, 26, 28, 135, 138, 268, 370, 371, 489, 510, 361–368, 370, 373, 374, 392, 585–587
511, 548, 553, 554, 581–589, 601 Kimmeridge Clay/shale, 13, 14, 123, 137, 232, 361, 362, 588,
Heidrun, 625 614, 618, 619
Hercynian, 608, 611 Kimmeridge (Upper Jurassic) shale, 13–14, 397
Hertz-Mindlin model, 492
History matching, 563, 578–579
Hjulstrøm’s curve, 43 L
Horda platform, 327, 606–608, 610, 613 Lacustrine, 63–65, 194, 266, 332, 399
Horizontal wells, 26, 561, 563, 569, 583 deposits, 60, 63–65
Humic kerogen, 363, 364 Lakes, 4, 8, 9, 45, 60, 61, 63–68, 70, 82, 91, 95, 111, 114, 122,
Hummocky, 83 123, 127, 136, 182, 194, 195, 206, 218, 222, 225–227,
stratification, 53, 78 245, 266, 304, 332, 352, 436, 585, 586
Hydraulic conductivity, 286, 592–593 Land seismic, 427, 430, 434, 438
Hydraulic radius, 60, 61 Lava, 63, 116, 226, 620
Hydrocarbon (HC) traps, 14, 319–349, 386, 515, 593 Leak-off test (LOT), 307, 308, 378–379
Hydrogen index, 368, 401, 407 Lignin, 8, 363
Hydrophone, 429–432, 434–436 Lime mud, 154, 159, 163, 164, 179, 190, 192, 199, 202, 615
Hydrostatic pressure, 67, 193, 202, 297, 303–304, 314, 387, 549, Liquefaction, 45, 48, 49, 54, 73, 84, 599
594, 625 Liquefied flow, 48–49
Listric faults, 17, 55, 89, 267, 321, 330, 334, 619
Lithoclasts, 180, 198
I Lithosphere, 87, 248, 321–323, 326–330, 345, 368, 621
Ice, 5, 54, 56, 63, 97, 103, 218, 222, 261, 263, 265, 298, 357, 373, Lithospheric thinning, 326
374, 452, 511, 588, 589, 594, 599, 616, 624, 634, 650 Lithostratigraphy, 231–233, 243–244
IGV. See Intergranular volume (IGV) Load cast, 54, 55
Illitisation, 136, 137, 467, 469 Lofoten, 621, 623
Image logs, 2, 423–424 Logging while drilling (LWD), 389, 390, 392, 393, 422, 424
Imaging, 308, 344–347, 422, 423, 428, 434–437, 444, 519, 528, LOT. See Leak-off test (LOT)
538, 542, 621 Love waves, 427, 428
Impedance, 256–259, 354, 410, 423, 429, 448, 449, 455, 457,
458, 466, 471–473, 476, 493, 504, 528–530, 557, 569
Incidence angle, 432, 507 M
In Salah CCS project, 600 Magmatism, 334, 621, 632
Instantaneous phase, 448, 449 Magnetotelluric (MT), 515, 523, 527, 528
Intergranular volume (IGV), 129, 130 Magnitude vs. source-receiver offset (MVO), 517, 521, 522, 531
Interval velocity, 448 Mandal formation, 617, 619
Intraclasts, 180, 181, 202 Marine magnetotelluric method (MMT), 515, 516, 527–530,
Inversion, 114, 133, 284, 292, 336, 343–344, 432, 455, 472, 473, 538, 540, 541
482–484, 511, 512, 517, 519, 521–527, 530, 536, Marine seismic, 2, 430, 431, 433–436, 439, 497
539–541, 562, 607, 615, 625, 648, 651 Mass flow, 44, 47–48, 120, 336, 337
658 Index
Maturation, 6, 13, 14, 27, 213, 267, 270, 271, 274, 283, 319, 326, 330, 335–337, 353, 355, 361, 362, 364, 367, 382,
324, 332, 340–342, 363–368, 373, 374, 392, 542, 634 383, 397, 459, 462, 465–468, 473, 475, 489, 491, 493,
Maxwell equations, 523 496–499, 504, 524, 530, 532, 553, 573, 576, 588, 589,
Measurements while drilling (MWD), 389, 390, 392, 393, 424 594, 598, 603–620, 625, 634, 644, 645, 651
Mechanical compaction, 21, 121, 122, 127–129, 135, 142, 143, North Sea field, 573, 576
145, 196, 200, 210, 211, 213, 294, 296–298, 301, 304, Norwegian continental shelf (NCS), 321, 530, 603–634, 644,
307, 312, 315–317, 351, 352, 356–358, 467, 469, 477, 645, 648, 651
483, 552 Norwegian-Danish basin, 607, 608
Mesozoic, 65, 68, 114, 116, 124, 152, 154, 166, 170, 175, 192, Norwegian petroleum directorate (NPD), 29, 603, 644, 645, 651
193, 221, 237, 243, 255, 262, 263, 281, 282, 604, 622, Norwegian Sea, 48, 56, 526, 593, 605, 645
626, 629–633 Nuclear magnetic resonance (NMR) logging, 391, 401, 404–405
Meteoric water/flow, 18, 97, 103, 121–127, 130, 138–142, 145, Nyquist frequency, 439
159, 189, 194–196, 199, 210, 212, 213, 271, 273, 276,
279, 281–284, 287–290, 294, 297–299, 304, 550–552,
614 O
Micrite, 177, 179, 180, 202, 203, 228 Ocean bottom seismic (OBS), 435–438
Micritic envelope, 178, 179, 196 Ocean water circulation, 154
Micritisation, 177–179, 181, 192 Oil
Microfossils, 1, 23, 233–234, 236, 238–240, 246, 262 emplacement, 135, 138
Micropalaeontology, 1, 233 reserves, 4, 23–24, 28, 144, 209, 554, 585, 646
Microquartz, 123, 130, 131, 133, 141–143, 355, 614 saturation, 125, 135, 145, 375–377, 381, 382, 548–550,
Mid-north sea high, 608, 617, 619, 651 553–555, 557
Midpoint gather, 433 seeps, 2, 8, 18, 221, 371, 427, 581, 617
Migration, 2, 4, 6, 10–14, 16, 22, 27, 33, 50, 52, 79, 80, 82, 139, shale, 3, 4, 24–28, 363, 364, 371, 374, 376, 581–589
208, 260, 266, 284, 319, 324, 342, 345, 347, 367, 368, window, 10, 12, 13, 588, 614
373–384, 439, 440, 444, 445, 449, 504, 542, 581, 584, Oil-based mud filtrates, 387, 389, 396, 422
587, 596, 598, 600, 603, 616, 625, 626, 634, 639, 641, Oil-wet, 135, 211, 375, 418, 468, 546–548, 554, 583
642, 647, 651 Ooids, 21, 22, 128, 129, 181–183, 202, 204, 205, 210, 211
Mineral trapping, 130, 459, 555, 596–598 Oomoulds, 183, 211
Mississippi, 66, 68, 70–71, 73, 76, 115, 218, 267, 276, 290 Opal (A and CT), 130, 143, 144, 220, 221, 257, 283
MMT. See Marine magnetotelluric method (MMT) Open-hole logs, 389, 424
More, 620–626 Organic geochemistry, 2
Mouldic porosity, 19, 22, 171, 182, 210–211, 214, 556 Organic silica, 131, 143, 219
Mudcake, 387, 388, 393–395 Ormen Lange Field, 515, 645
Mud filtrate, 282, 387, 388, 394–396, 406, 418, 472 Oslo Graben, 608, 612
Mud flow, 39, 47–48, 59, 60 Oslo Rift, 608, 609, 612
Mud pulses, 389 Ostracods, 177, 180, 184, 192, 199, 234–236
Multiples, 173, 197, 257, 432–436, 439–444, 447, 508, 560, Overpressure, 21, 49, 55, 73, 103, 128, 142, 143, 211, 257,
574, 580, 600, 633 285–287, 290, 294–299, 303, 304, 306, 308, 311, 317,
MVO. See Magnitude vs. source-receiver offset (MVO) 356–358, 365, 373, 374, 378, 379, 382–384, 388, 405,
MWD. See Measurements while drilling (MWD) 407, 409, 466, 467, 549, 592, 617, 618, 625
Overpressured zones, 407, 466
Oxygen index, 368
N
Naphtenes, 365
NCS. See Norwegian continental shelf (NCS) P
Near stack, 475–479 Packstone, 166, 204, 205
Nereites facies, 58 Palaeomagnetism, 246
Net/gross (N/G), 20, 145, 549, 593 Palaeozoic, 8, 68, 87, 111, 139, 154, 159, 162, 164–167, 172,
Net reservoir, 563 174, 175, 194, 207, 221, 242, 243, 246, 262, 318, 588,
Neutron logs, 401–402, 405, 406, 408, 413, 414, 562 608–610, 612, 617, 620–629, 647
Neutron porosity, 401, 406, 411, 413 Paraffin, 365, 368, 369, 581
NMO, 440–443 Parasequences, 247, 249–252
NMVO. See Normalized magnitude (NMVO) Passive margins, 15, 59, 267–268, 324, 340, 357, 363, 603, 626
Non-coherent noise, 439, 440 Patch, 52, 84, 210, 223, 435, 471, 473
Non-skeletal carbonate grains, 179–183 Pelagic carbonate, 140, 193–194
Normalized magnitude (NMVO), 517, 518, 525, 526, 531, 532 Pellets, 9, 43, 82, 87, 180, 181, 190, 192, 203
North Sea, 10, 11, 13–15, 23, 56, 64, 78, 81, 100, 126, 127, Peloids, 178, 181, 202
129–131, 135–137, 139, 140, 161, 162, 184, 194, 212, Permeability, 1, 17–20, 33, 103, 119, 151, 217, 245, 279,
220, 223, 232, 237, 240, 242, 255, 262, 266, 274, 276, 286–291, 294–298, 301, 353, 374–378, 387, 404, 467,
281, 282, 290, 292, 294, 296, 299, 304, 306, 317, 324, 549–556, 562, 564, 584, 591, 593, 617, 640
Index 659
Permeability logs, 563, 569 P–S stack, 436, 438

Permian, 15, 63, 64, 139, 162, 163, 165, 167, 170–172, 175, 182, PSTM. See Pre-stack time migration (PSTM)
212, 223, 262, 282, 293, 452, 585, 604, 608–613, Pull-apart basins, 338, 626, 632
617–619, 622, 627–629, 633, 634 Pure shear, 326, 328–330, 332, 336
carbonates reservoirs, 651 PVO. See Phase vs. source-receiver offset (PVO)
Persian gulf, 179, 190–193, 223–225, 267 P-wave, 255, 407, 427–430, 432, 435, 458, 460, 464, 465, 470,
Petroleum 474, 475, 489, 491, 492, 500, 534, 569
composition, 382 Pyrolysis, 4, 362, 364, 367, 585, 586
generation, 12, 276, 299, 365–367, 374, 382, 383, 642
migration, 10–14, 373–384, 603, 634, 651
seepage, 383 Q
Petrophysical property(ies), 389, 561, 564–565, 568, 569, Quartz cement, 122, 127, 129–142, 274, 276, 295, 312, 313,
572–573, 579 315, 317, 354, 355, 358, 463, 467, 468, 473, 480, 584,
modelling, 565, 568, 572–573 620, 625, 634
pH, 8, 92, 94–96, 104, 107, 108, 111, 113, 125, 135, 147, 152, Quartz overgrowth, 132–135, 138, 462, 463
153, 179, 185, 192, 193, 202, 221, 222, 226, 227, 281,
283, 383, 595–597
Phase difference response (PDVO), 517, 525 R
Phase vs. source-receiver offset (PVO), 521, 522 Radiactive logs, 2, 23
Phosphorite, 199, 228–229 Radiolaria, 7, 107, 193, 219–221, 228
Photoelectric absorption, 411 Radiometric dating, 231, 233, 240, 243–245
Photoelectric factor (PEF)/litho-density log, 411–413 Raw seismic, 439, 440
Photosynthesis, 4–6, 9, 95, 98, 112, 113, 153, 155, 158, 190, Rayleigh, 291, 427, 428, 430
193, 194, 222, 245, 283, 361 convection, 291
Phreatic zone, 197 Receiver array, 436
Phytoplankton, 4, 6, 219, 363 Red algae, 77, 152, 157–159, 162, 177, 184, 185, 201
Piezometric, 285, 297, 303, 304, 430, 433 Redox boundary, 8, 10, 103, 122, 123, 209, 227, 280, 282, 294
P-impedance, 429 Reefs, 6, 14, 16, 17, 19, 21, 33, 58, 78, 84, 158, 159, 162, 163,
Pisolites, 182 168, 170, 171, 184–192, 199–201, 204–207, 210, 214,
Planar faults, 334, 338 255, 557, 585
Plants, 1, 4, 7, 8, 12, 24, 26, 27, 32, 61, 65, 73, 87, 98, 104, 112, Reflected angle, 429
122, 124, 151, 154–164, 178, 187, 192, 194, 228, 236, Reflected wave, 429, 435
240, 245, 263, 361, 363, 365, 366, 439, 587, 591, 592, Reflection, 192, 193, 220, 256–260, 329, 336, 345, 347, 422,
599–601 427–429, 433, 439, 442–446, 448, 455, 456, 460, 474,
Plastic deformation, 324 475, 481, 504, 507, 511, 516, 620
Plate tectonics, 261, 262, 271, 306, 321, 323, 452, 558, 585, 606 coefficient, 256, 257, 411, 429, 439, 474, 475, 493
Point bar, 20, 66, 67, 82, 399 Reflectivity, 12, 13, 27, 365, 410, 432, 433, 458, 469, 475, 477,
Porewater, 95, 103, 146, 213, 275, 281, 283, 285, 289, 299, 303, 479, 493
304, 354, 373, 534, 618, 634 Refracted angle, 429
composition, 145 Refracted wave, 429, 510
flow, 145, 304 Refraction, 36, 82, 256, 427, 429, 507–511, 516, 604
Porosity Regression, 59, 70, 75, 180, 224, 228, 248, 261, 262, 451, 452,
depth, 133, 140, 200, 293, 297, 352, 358 475, 629
depth trend, 142–143, 352, 468 Relative permeability, 286, 375, 376, 382, 384, 549–552, 554,
log, 401–410, 562, 570 564, 575, 577, 596
loss, 128, 129, 132, 142–144, 146, 213, 293, 301, 351, 357, Reservoir
358, 634 modelling, 559–580
Post-mortem destruction, 175–177 quality, 22, 28, 90, 120–124, 127, 130, 135, 136, 140–146,
Potentiometric, 285–289, 291, 296, 303, 376 151, 168, 171–173, 175, 182–183, 203, 210, 269, 273,
Potentiometric gradients, 285–288, 376 274, 277, 301, 387, 432, 456, 484, 549, 613, 634
P–P stack, 436, 438 rocks, 1, 6, 10–12, 14, 16, 18–22, 25–28, 31, 33, 90, 120,
Pre-stack depth migration (PSDM), 444 124, 127, 130, 136, 138–140, 145, 146, 160, 162,
Pre-stack time migration (PSTM), 444–446 192–194, 200, 203–205, 209–214, 217, 221, 269, 271,
Production monitoring, 447, 557 277, 313, 336, 342, 345, 371, 373, 376, 383, 385, 393,
Progradation, 20, 59, 61, 66, 70, 75, 77, 85, 86, 192, 243, 247, 405, 411–412, 414, 498–501, 512, 516, 542, 545, 546,
249, 250, 258, 267–269, 334, 451, 614, 616, 619, 632 548–550, 554, 555, 558, 581, 582, 587, 609, 613, 614,
PSDM. See Pre-stack depth migration (PSDM) 617–620, 624–626, 634, 647
Pseudostatic spontaneous potential (PSP), 395 Resistivity-derived porosity, 406, 407, 418
PSP. See Pseudostatic spontaneous potential (PSP) Resistivity log, 391, 413–416, 418, 424–426, 547, 562, 579
660 Index
Reynold’s number (Re), 40 Seismic lithofacies, 465, 466, 482, 483

Rift/rifting, 11, 15, 63–65, 68, 89, 110, 119, 139, 222, 225, 226, Seismic monitoring, 438, 489, 491, 498, 598
229, 262–264, 266–268, 276, 294, 324, 326–328, 330, Seismic processing, 435, 436, 438–444
332–335, 340, 344, 603, 605–608, 610, 612–615, Seismic pulses, 257, 428, 430
618–621, 623, 625–630, 632, 634 Seismic pumping, 295–296, 304
Rift stages, 327–330, 332, 340, 615 Seismic reservoir, 455–457, 473–486
Ringkøbing-Fyn High, 608, 612 Seismic resolution, 444–448, 565, 579
Rock-Eval, 367–368 Seismic signal processing, 429
Rock physics models, 452, 456, 458–463, 469, 470, 472, 481, 484 Seismic signatures, 467, 473, 477, 482
Rotliegend, 63, 139, 604, 610, 612 Seismic source, 364, 427–432, 436
Seismic stratigraphy, 255–271, 450, 452
Seismic trace, 257, 408, 410, 432, 433, 436, 448, 493, 496
S Seismic waves, 255, 259, 427, 429, 430, 432, 433, 435, 445,
Sabkha, 157, 191, 223–226, 612, 613 450, 460, 472, 507
Saga Petroleum, 491, 645 Self-potential/spontaneous potential log (SP), 394–395, 424
Salt Semblance, 443
deposits, 15–16, 192, 207, 223, 266, 283, 344, 613 Sequence stratigraphy, 1, 2, 90, 121, 124, 247–252, 255–271,
diapers, 292 289, 564
domes, 15–16, 223, 259, 267, 275, 283, 292, 306 Sequence types, 251–252
Sampling, 32, 39, 385, 435, 439, 569, 577, 600, 642, 651 SGR. See Spectral gamma ray (SGR) log
Sand reservoirs, 395 Shadow zone, 430
Sandstone Shale
fluvial, 17, 20, 21, 53, 139, 140, 377, 383, 421, 584, 611, 613 gas, 4, 24–26, 28, 217, 356, 371, 374, 581–590
marine, 20, 134, 139, 140, 243, 382, 613 matrix, 374, 378, 467, 587
reservoir, 21, 143, 277, 290, 313, 383, 473, 477, 534, 551 Shear wave (S-wave), 255, 317, 427–430, 432, 435, 437, 460,
Sapropelic kerogen, 362–363 461, 465, 470, 474, 492
Saturation effects, 501–503 Shelf, 21–22, 39, 62, 69, 77, 81–85, 112, 162, 184, 190, 193,
Saturation exponent, 418 219, 225, 228, 236–238, 249, 251, 252, 258, 261, 267,
Seabed, 8, 10, 17, 46, 52, 77, 103, 112, 113, 116, 122, 123, 164, 269, 271, 281–282, 290, 321, 326, 335, 338, 448, 450,
165, 187, 188, 190, 199, 200, 237, 261, 270, 281, 282, 451, 486, 530, 538, 539, 603–634, 644–651
294, 383, 430, 438, 497, 504, 515, 516, 520–522, 528, Shore face, 20, 21, 77, 139, 192, 219, 249, 376, 552, 571, 619
535, 604 Shot-point gather, 443
Seal, 3, 12, 144, 287, 295–299, 343, 376, 379, 381, 382, 384, Siderite, 56, 96, 103, 123, 152, 208–209, 227, 250, 282, 302,
584, 591, 593, 599, 634 413, 597, 613
Sea urchins, 173, 174, 178 Signal-to-noise (S/N) ratio, 443
Secondary porosity, 19, 127, 145, 159, 167, 168, 199, 203, 208, Silica deposits, 217–229
210–211, 214, 290, 383, 556, 618 S-impedance, 429
Sediment Simple shear, 326, 327, 329–330, 332
texture, 34–36, 49, 58, 90 Skagerrak, 604, 608, 609, 612
transport, 31–91, 119, 121, 190, 328, 334, 336, 337 Skagerrak Graben, 608, 612
Sedimentary basin, 1–3, 8, 17, 18, 25, 27, 28, 33, 44, 68, 90, 91, Sleipner, 506, 598, 599, 604
94, 100, 116, 121, 125, 128, 132, 135, 137, 138, field, 506, 553, 616
141–143, 145, 146, 151, 208, 213, 217–219, 232, 242, Slumping, 54, 55, 73, 84, 617
243, 247, 260, 261, 263–266, 268, 270, 271, 273–277, Smectite, 22, 39, 40, 47, 56, 62, 101, 102, 106, 107, 110, 111,
279–299, 301–349, 351, 356, 357, 364–366, 368, 370, 125, 130, 134, 136, 137, 139, 144, 217–219, 274, 279,
371, 373, 375, 377, 382, 430, 438, 540, 542, 545, 581, 281, 282, 302, 314–317, 351–354, 467–469, 555, 556,
588, 589, 603, 606–608, 610, 621, 626, 633 587, 617, 620
Sedimentary environments, 1, 27, 49–50, 119, 226, 448, 558 Snell’s law, 256, 429, 507
Sedimentary facies, 1, 2, 28, 57–59, 89, 91, 124, 146, 243, Snells refraction, 256
259–260, 287, 328 Snøhvit CCS project, 599–600
Sedimentary structures, 31, 32, 43, 49, 50, 53, 54, 59, 60, 78, 83, Snorre Field, 618, 619, 645
231, 377 Solubility trapping, 595, 596
Seepage, 3, 8, 50, 371, 383 Sonic (acoustic) logs, 256, 257, 391, 401–404, 407–409, 419,
Seismic acquisition, 431–436, 438, 439, 512, 647 424, 452, 456, 466, 484, 562
Seismic attributes, 301, 437, 448–451 Sorgenfrei-Tornquist Zone, 608, 609, 612, 615
Seismic exploration, 427–454, 507 Source rock, 1, 4–6, 8–16, 25–27, 31, 33, 64, 65, 80, 84, 91, 110,
Seismic facies analysis, 448 111, 119, 120, 123, 124, 137, 139, 157, 159, 160, 162,
Seismic fluid sensitivity, 470–472 192, 200, 213, 214, 217, 221, 229, 233, 237, 245, 266,
Seismic framework, 565–567 267, 271, 274, 276, 277, 283, 298, 328, 332–334, 340,
Seismic interpretation, 260, 347, 447–452, 455, 456, 483, 342, 343, 354, 356, 361–371, 373–375, 377, 379,
493, 565–567 381–383, 397, 416, 418–419, 424, 525, 538, 542, 582,
Index 661
584, 585, 587, 588, 603, 604, 610, 613–615, 617–620, Tilje formation, 134, 625
624, 625, 629, 633, 634, 647 Time-lapse seismic, 437, 489, 491, 493, 497–499, 501,
Sparite, 179, 202, 203 504–507, 512, 556
Spectral gamma ray (SGR) log, 396, 397, 400–401 Time migration, 444, 445
Spekk formation, 354, 625 Time slice, 48, 260, 447, 449
Spiculite, 167 Time-stratigraphy, 240–243
Spill point, 14, 319, 382, 617 Time-temperature index (TTI), 277, 366, 367
Sponges, 56, 123, 124, 130, 131, 143, 154, 164, 166–167, 176, Tortuosity factor, 299, 417, 418
178, 187, 199, 212, 219, 221, 315, 354, 614 Trace fossils, 49, 56–58, 82, 187
Spontaneous potential (SP) logs, 394–396, 417 Trace muting, 440
Stacking, 249, 336, 439, 440, 442–444, 448, 521 Transgression, 20, 21, 59, 60, 70, 75, 76, 78, 79, 100, 112, 161,
Statfjord allocation, 645 184, 185, 192, 219, 228, 248–250, 252, 258, 259, 261,
Statfjord field, 129, 146, 321, 604, 618 262, 264, 265, 267, 269, 336, 451, 452, 610, 613, 614
Static correction, 436, 439, 448 Transmission loss, 432, 433
Static reservoir model, 559, 565–575 Trap, 2, 6, 11, 12, 14–18, 21, 25–27, 33, 73–75, 80, 154, 155,
Static spontaneous potential (SSP), 395 182, 184, 186, 188, 200, 223, 227, 233, 267, 270, 295,
Statoil, 489, 504, 506, 515, 536, 599, 644–646 319–349, 353, 368, 375, 377–379, 381, 382, 386, 427,
Steam injection, 371, 489, 511, 553, 554, 581 455, 515, 534, 536, 537, 542, 546, 553, 584, 588, 593,
Stokes law, 34 597, 616–618, 620, 624, 625, 633, 634, 641, 647
Stratigraphic trap, 6, 14, 16–18, 21, 80, 319, 321, 324, 328, 332, Travertine, 194–195
334–337, 340, 342, 345, 346, 536, 537, 542, 593, 647 Triassic, 136, 162, 164, 167, 170, 171, 237, 247, 262, 607, 608,
Streamer, 430, 433–438, 497, 509 610, 612–614, 617–619, 621, 622, 626, 628, 629, 634,
Stress-strain, 309–311, 457 648, 651
Strike-slip, 17, 268–270, 304, 319, 321–323, 337–340, 345, 347, Troll field, 23, 504, 530, 616, 618, 619, 634
609, 610, 612, 626, 631, 632 Troll West Gas Province (TWGP), 504, 530–532
Stromatolite, 154–157, 164, 180, 192, 207 True amplitude recovery (TAR), 440
Stromatoporoids, 159, 167–168 Truncated Gaussian fields, 571
Structural trapp(ing), 2, 6, 14–17, 25, 26, 267, 270, 319–321, TTI. See Time-temperature index (TTI)
324, 327, 328, 331, 334, 338, 340, 342, 343, 345, 542, Tubåen formation, 536, 593, 598, 599
588, 593, 595–596, 617, 624, 647 Tufa deposits, 194
Stylolites, 22, 132, 133, 142, 201, 202, 208, 211–214, 355, 356, Turbidites, 17, 20–22, 38, 39, 45, 46, 50, 58, 70, 83–85, 87, 89,
358, 359, 376, 377, 557 120, 140, 193, 212, 250, 269, 270, 290, 377, 383, 399,
Submarine canyons, 84–87 465, 467, 473, 477, 481, 482, 614, 616, 620
Submarine fans, 85–87, 337, 448, 456, 483, 616 Turbidity currents, 44–47, 64, 73, 83–85, 219
Subsidence TWGP. See Troll West Gas Province (TWGP)
basin, 147, 297, 298, 304, 328, 329, 373, 378, 392, 614 Two-way time (TWT) domain, 567
tectonic, 60, 70, 265, 332, 615 Two-way travel time, 409, 442, 444, 482
Svalbard, 161, 167, 603, 626, 629, 630, 632, 633
Swath, 435
S-wave. See Shear wave (S-wave) U
System tract, 249–251 Uninvaded/undisturbed zone, 388, 389, 416, 417, 419–420
Uplift, 2, 8, 11, 21, 26, 48, 60, 67, 76, 104, 120, 126, 133, 140,
144, 151, 210–212, 218, 219, 225, 226, 245, 248, 259,
T 261, 263, 266–269, 271, 275, 283, 286–288, 290, 294,
Tampen Spur, 606, 608, 618 297, 298, 301, 304–306, 314, 315, 318, 321, 327, 330,
Tarbert formation, 619 336, 342–344, 351, 357, 370, 371, 373–376, 378, 379,
Tar sand, 24–26, 28, 268, 371, 553, 554, 581–589 404, 407–409, 451, 452, 483, 484, 538, 558, 565, 581,
Temperature log, 391–393, 419 582, 585, 587, 588, 599, 603, 608, 609, 611, 613–617,
Tertiary, 4, 51, 65, 100, 111, 124, 140, 162, 165, 188, 199, 206, 619, 620, 623, 624, 628, 632–635
224, 226, 228, 237, 240, 246, 255, 261, 263, 271, 296, Upwelling, 7, 21, 84, 114, 154, 184, 185, 220–221, 228, 361
298, 467, 468, 475, 506, 586, 587, 598, 607, 615–620, Utsira High, 608, 616, 646, 647, 649, 651
625, 633 Utsira Sand, 592, 593, 598, 599, 601
structures, 626, 645
3D seismic, 2, 23, 48, 377, 434, 435, 437, 438, 446, 450, 452,
474, 482, 490, 501, 534, 536, 542, 549, 647, 649 V
Thermal alteration, 26, 27, 361 Vadose, 196
Thermal convection, 138, 273, 284, 290–293, 297 Vadose zone, 195–197, 226, 613
Thixotropic, 47, 103 Variscan, 608, 611–613
Thrust systems, 340–342 Velocity analysis, 439, 440, 443, 444
Tidal Velocity spectrum, 443
bundles, 82, 83 Vertical seismic profiles (VSP), 260, 436–437, 439, 496
channels, 20, 75, 80, 82, 83, 180, 190, 192 Vesterålen, 621, 623
662 Index
Vestland group, 617 Wavelength, 51, 52, 66, 77, 83, 101, 158, 190, 257, 259, 346,
Vibrator, 427, 430–432 433, 445–448, 452, 625
Viking Graben, 324, 326, 327, 330, 332, 335, 606–608, 610, Wavelet, 410, 432, 433, 440, 446
613–616, 618 Weathering, 8, 24, 25, 31, 39, 47, 49, 59, 60, 67, 91–94, 99–101,
Void ratio, 18, 284, 356, 363, 365, 373 104–116, 119–122, 124, 125, 131, 144, 146, 151, 153,
Volcanism/volcanic, 39, 65, 77, 87, 89, 97, 107, 110, 111, 113, 180, 183, 188, 194, 195, 209, 211, 217–219, 225–229,
115, 116, 120, 125, 130, 139, 188, 217–220, 225, 226, 238, 283, 315, 368, 439, 648
231, 246, 263, 264, 266, 269–271, 277, 315, 353, 354, Well logs, 2, 123, 144, 145, 232, 279, 289, 292, 317, 318, 351,
363, 397, 467, 504, 525, 538–540, 603, 610, 613–615, 385–424, 452, 455–458, 462, 465, 466, 468, 472, 473,
619, 623, 626 478, 479, 481, 483, 484, 486, 493, 524–526, 562, 565,
Volgian reservoirs, 640, 648 577, 579
Vøring, 604, 620–626 West of Shetland, 538–540, 542
VSP. See Vertical seismic profiles (VSP) Wetting angle, 375, 381, 546–548, 550, 554
Weyburn-Midale project, 598, 600
Wildcat well, 22
W Wilson cycle, 324–326
Wacke, 120, 139, 204, 397 Wyllie’s equation, 255, 256, 407, 410
Water drive, 271, 551, 552
Water injection, 284, 288, 439, 506, 550, 552–554, 594, 618
Water saturation, 127, 138, 387, 417, 418, 420, 491, 534, 548,
555, 561 Z
Water-wet, 12, 135, 138, 375, 376, 531, 546–548, 554–556, 596 Zechstein, 15, 212, 223, 266, 282, 294, 612, 617, 651
Waveform, 402, 439, 447, 511 Zero-offset, 437, 441, 442, 444, 475, 508
Wavefront, 429, 433, 435 Zooplankton, 4, 7, 9, 153, 219, 363

2015 Book PetroleumGeoscience PDF

Uploaded by

Copyright:

Available Formats

2015 Book PetroleumGeoscience PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2015 Book PetroleumGeoscience PDF

Uploaded by

Copyright:

Available Formats

Knut Bjørlykke Editor

With contributions from Per Avseth, Jan Inge Faleide,

ISBN 978-3-642-34131-1 ISBN 978-3-642-34132-8 (eBook)

# Springer-Verlag Berlin Heidelberg 2010, 2015

Cover design: deblik, Berlin

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

The second edition of Petroleum Geoscience is an updated and corrected version of

Adrian Read ([email protected]) has been of great assistance as Text Editor in

Oslo, Norway Knut Bjørlykke

1 Introduction to Petroleum Geology . . . . . . . . . . . . . . . . . . . . . . . . . . 1

15 Petroleum Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373

Per Avseth, Tullow Oil, Oslo, Norway

Introduction to Petroleum Geology

K. Bjørlykke (ed.), Petroleum Geoscience: From Sedimentary Environments to Rock Physics, 1

Deposition of source rocks

3–4 km Cap rock

Cretaceous mudstones and shales

Triassic and Permian

Upper Jurassic source rock

Other compounds, such as pentane, are even less solu- H2O

6–8 km – 200°C Burial curve

5 Gas 150° 2.0

Fig. 1.8 Kimmeridge Clay (Upper Jurassic) unconformably

south England and in many parts of the North Sea

III Deposition of 2–3 km of sediments

halotectonic movements have formed dome structures (d) Fault Traps

The name growth fault goes back to the early days of

1.8.2 Stratigraphic Traps

These are traps which are partially or wholly due to

Reservoir Geology is not a well-defined discipline.

1.10 Porosity and Permeability

Any rock with sufficiently high porosity and perme-

1.9 Other Types of Trap VR ¼ φ=ð1 φÞ

Geochemistry of carbonate minerals

More and more sophisticated methods are therefore

WELL FOR Cap rock

Organic matter - mostly algae

Oil Oil window

Coal bed methane

Eﬀect of heang (pasteurisaon) of reservoir

Sand reservoir Biodegraded oil/tar sand

Normal oil - little

unconventional sources represents a serious environ- Further Reading

K. Bjørlykke (ed.), Petroleum Geoscience: From Sedimentary Environments to Rock Physics, 31

Amount of sediment on Sedimentation balance

2.2 Grain-Size Distribution in Solid

Resistance due to Lightly-cemented sandstones can be disintegrated by

acceleration term g in Stokes’ Law. There are also

0.01 2.4 Grain-Size Distribution Parameters

Better sorting Poorer sorting

2.6 Grain Shape (or aerodynamic) principles involved. When a liquid

Laminar flow Turbulent flow

Shear stresses against air

Fig. 2.8 Different forms of transport in water. Sediment grains

a Bedforms as a function of flow velocity and grain size

Average velocity cm/s Anti-dunes

0.1 0.2 0.3 0.4 0.6 0.8 1.0 1.5 2.0 mm

Flood High tide Ebb tide