G E O L O G I C A L SOCIETY S P EC I A L P U B L I C A T I O N NO.

36

Diagenesis of
Sedimentary Sequences
EDITED BY
J. D. M A R S H A L L
Department of Geological Sciences,
University of Liverpool,
Liverpool L69 3BX, UK

1987
Published for
the Geological Society by
Blackwell Scientific Publications
OXFORD LONDON EDINBURGH

BOSTON PALO ALTO MELBOURNE

 Geological Society Special Publications
Series Editor K. COE

Published for DISTRIBUTORS

The Geological Society by
BlackweU Scientific Publications USA and Canada
Osney Mead, Oxford OX2 0EL Blackwell Scientific Publications Inc.
(Orders: Tel. 0865-240201) PO Box 50009, Palo Alto
8 John Street, London WC1 2ES California 94303
23 Ainslie Place, Edinburgh EH3 6AJ (Orders: Tel. (415) 965-4081)
52 Beacon Place, Boston
Massachusetts 02 i08, USA Australia
667 Lytton Avenue, Palo Alto Blackwell Scientific Publications (Australia) Pty
California 94301, USA Ltd.
107 Barry Street, Carlton, Victoria 3053 107 Barry Street, Carlton, Victoria 3053
Australia (Orders: Tel. (03) 347-0300)

First published 1987 British Library Cataloguing in Publication Data

Diagenesis of sedimentary sequences.--
9 1987 The Geological Society. Authorization to (Geological Society special publication;
photocopy items for internal or personal use, or the ISSN 0305-8719; no. 36).
internal or personal use of specific clients, is granted 1. Diagenesis
by the Geological Society for libraries and other I. Marshall, J. D. (James D.) II. Series
users registered with the Copyright Clearance Center 552'.5 QE471
(CCC) Transactional Reporting Service, providing
that a base fee of $03.00 per copy is paid directly to ISBN 0-632-01939-5
CCC, 27 Congress Street, Salem, MA 01970, USA.
0305-8719/87/$03.00 Library of Congress
Cataloging-in-Publication Data
Printed and bound in Great Britain by Diagenesis of sedimentary sequences/edited by
William Clowes Limited, Beccles and London James D. Marshall.
p. cm. - - (Geological Society special publication;
no. 36)
Papers of a meeting held in Liverpool, Sept. 30-
Oct. 1, 1986 under the auspices of the British
Sedimentological Research Group (BSRG).
Includes bibliographies.
ISBN 0-632-01939-5
1. Diagenesis--Congresses. 2. Sedimentation and
deposition--Congresses. I. Marshall, J. D. (James
D.) II. Geological Society of London. III. British
Sedimentological Research Group. IV. Series.
QE471.D53 1988
552'.5--dc19
 Contents

Preface
MARSHALL, J. D. Diagenesis and sedimentary sequences--introduction vii

Diagenetic processes
GOLDSMITH, I. R. & KING, P. Hydrodynamic modelling of cementation patterns in modern
reefs 1
MACHEL, H. G. Some aspects of diagenetic sulphate-hydrocarbon redox reactions 15

PALMER, S. N. & BARTON, M. E. Porosity reduction, microfabric and resultant lithification in

UK uncemented sands 29

RAISWELL, R. Non-steady state microbiological diagenesis and the origin of concretions and
nodular limestones 41

WARREN, E. A. The application of a solution-mineral equilibrium model to the diagenesis of

Carboniferous sandstones, Bothamsall oilfield, East Midlands, England 55

Early diagenesis
ASTIN, T. R. Petrology (including fluorescence microscopy) of cherts from the Portlandian of
Wiltshire, UK--evidence of an episode of meteoric water circulation 73

CARSON, G. A. Silicification fabrics from the Cenomanian and basal Turonian of Devon,
England: isotopic results 87

KANTOROWlCZ, J. D., BRYANT, I. D. & DAWANS, J. M. Controls on the geometry and

distribution of carbonate cements in Jurassic sandstones: Bridport Sands, southern England
and Viking Group, Troll Field, Norway 103

PUEYO, MUR, J. J. & INGLI~SURPINELL, M. Magnesite formation in recent playa lakes, Los
Monegros, Spain 119

TABERNER, C. & SANa'ISTEBAN,C. Mixed-water dolomitization in a transgressive beach-ridge

system, Eocene Catalan Basin, NE Spain 123

SMITH, R. D.A. Early diagenetic phosphate cements in a turbidite basin 141

ZHAO XUN & FAIRCHILD, I. J. Mixing zone dolomitization of Devonian carbonates, Guangxi,
South China 157

Regional studies and burial diagenesis

BATH, A. H., MILODOWSKI, A. E. & SPIRO, B. Diagenesis of carbonate cements in Permo-
Triassic sandstones in the Wessex and East Yorkshire-Lincolnshire Basins, UK: a stable
isotope study 173

BOLES, J. R. Six million year diagenetic history, North Coles Levee, San Joaquin Basin,
California 191

EMERY, D. Trace-element source and mobility during limestone burial diagenesis--an example
from the Middle Jurassic of eastern England 201
iii
LAND,L. S. & FISHER,R. S. Wilcox sandstone diagenesis, Texas Gulf Coast: a regional isotopic
comparison with the Frio Formation 219

HARRIS, D. C. & MEYERS, W. J. Regional dolomitization of subtidal shelf carbonates:

Burlington and Keokuk Formations (Mississippian), Iowa and Illinois 237

HUDSON, J. D. & ANDREWS,J. E. The diagenesis of the Great Estuarine Group, Middle
Jurassic, Inner Hebrides, Scotland 259

LONGSTAFFE,F. J. & AYALON,A. Oxygen-isotope studies of clastic diagenesis in the Lower

Cretaceous Viking Formation, Alberta: implications for the role of meteoric water 277

PARNELL, J. Secondary porosity in hydrocarbon-bearing transgressive sandstones on an

unstable Lower Palaeozoic continental shell Welsh Borderland 297

SAIGAL,G. C. & BJORLYKKE,K. Carbonate cements in clastic reservoir rocks from offshore
Norway--relationships between isotopic composition, textural development and burial depth 313

STRONG,G. E. & MILODOWSKI,A. E. Aspects of the diagenesis of the Sherwood Sandstones of

the Wessex Basin and their influence on reservoir characteristics 325

DE WET, C. B. Deposition and diagenesis in an extensional basin: the Corallian Formation

(Jurassic) near Oxford, England 339

INDEX 355

iv
 Diagenesis and Sedimentary Sequences--Introduction
The meeting on 'The Diagenesis of Sedimentary Sequences' was held in Liverpool on 30 September
and 1 October 1986 under the auspices of the British Sedimentological Research Group (BSRG), a
Specialist Group of the Geological Society.
The aim of the meeting was to bring together an international group of research workers on
carbonate and clastic sediments, to discuss the major controls on sediment diagenesis from deposition
to deep-burial. This book contains a number of the papers presented at the meeting, together with one
that was offered but which the authors were unable to present. I believe that these papers are a good
reflection of the proceedings of a lively and diverse meeting, attended by around 150 representatives
of universities, governmental organizations and industry from nine different countries.
I leave readers to delve into the papers for themselves but would like to take this opportunity to
reiterate some of the points that I raised at the meeting concerning diagenesis and the way in which it
is approached.
Diagenesis is an integral part of the history of the fill of a sedimentary basin and needs to be treated
as such. In order to understand the post-depositional history of a sedimentary rock or, in an industrial
context, to understand the evolution of its reservoir properties, we must use information from more
general geological studies. A knowledge of depositional setting, facies architecture and burial history
are all invaluable in the deduction of a well-constrained diagenetic history. Similarly, quantitative
diagenetic investigations may contribute information, particularly about pore-fluid evolution and
temperature changes, which needs to be taken into account by those concerned with broader
syntheses of basin history and hydrocarbon prospectivity.
Many published works on diagenesis and indeed most of those in this book, are concerned with the
post-depositional history of a single lithological unit, or even a particular phase within a unit. As the
author of a paper on the diagenetic adventures of just three ammonites (Marshall 1981) I can scarcely
be too critical! Such studies concentrate attention on the interesting but perhaps atypical features of
the sediment. They are undoubtedly useful in determining the local controls on diagenesis: indeed
they often reveal just how complex the interaction of processes (cementation, neomorphism,
compaction and dissolution) can be. However, if we only work on the small scale, it becomes difficult
to determine what is typical of a sequence as a whole and indeed what is regionally rather than just
locally significant.
It is extremely difficult to be objective in sampling for any geological study and this is particularly
true for diagenetic investigations. Material is often only available from a restricted area of the basin
(determined by outcrop pattern or on a structure that has been drilled for hydrocarbons) and our
attention is automatically drawn to the 'different'. Even with careful sampling we can only ever hope
to look at a minute proportion of what is there; a colleague once estimated that he was being asked to
assess the reservoir potential of a North Sea reservoir unit on something like 10-15% of the rock in the
area! We should always be aware of this! Having said that, it is interesting, and I think particularly
welcome, to see that in a number of the papers in this book authors have been able to pool sufficient
data to treat diagenetic evolution in a regional context.
In organizing the meeting one of the main aims was to enable specialists from different areas to talk
to each other. The ways in which rocks of one type influence the diagenetic evolution of other
sediments within the same sedimentary sequence are far from clear. We might, for example, expect
chemical and isotopic buffering to preclude extremes of acidity, alkalinity or even geochemical values
as fluids pass from one rock type to another. All too often however, in diagenetic studies, authors
invoke a source, for ions, acids, fluids or whatever, from outside the rocks that they are studying. Too
often, in the past, they have not considered the feasibility of reactions or the problems of mass balance
or transfer.
Mudrocks are a fine example of an oft-quoted source or sink that we have yet to fully understand:
they are, after all, fine-grained, and full of reactive organic and inorganic chemical species which can,
through compaction, be expelled into adjacent sandstones and limestones. Several recent studies have
shown that mudrocks have complex diagenetic histories and have indicated that many reaction
products may stay very close to where they started. It is a shame that for a number of reasons more of
the 'mudrock' papers presented at the meeting are not included in this volume.
Conduits also tend to be poorly understood by workers in our field. Faults, for example, are
commonly invoked as carriers for fluids, which travel up (and occasionally down) the sedimentary
pile, yet the same fractures are known to seal hydrocarbons in place. Clearly, then, we need a more
integrated approach to constrain our grander conclusions. There is certainly scope for detailed,
 J. D. Marshall

quantitative petrographic studies of mudrocks (now possible with modern back-scatter electron
microscopy) and a need for us to talk to structural geologists. Organic geochemists and thermal
modellers also have a lot to offer the diagenetic investigator.
In planning the meeting and the layout of this book, to be consistent with the ideas expressed
above, I have tried to avoid grouping papers on purely lithological grounds. In the book therefore the
papers are, perhaps somewhat arbitrarily, arranged into three sections. The first contains a collection
of papers on 'Diagenetic Processes', the second on 'Early Diagenesis' and the third on 'Regional
Studies and Burial Diagenesis'; each contains papers dealing with both clastic and carbonate rocks. I
hope that readers will have a look at papers that lie outside their direct field of specialization.
Finally I would like to express my thanks to all the people who have helped in the organization of
the meeting and the preparation of the book. Generous financial support for the meeting was
provided by grants from the Geological Society and the Royal Society, and donations from BP
(London), BP Research, Britoil, Esso and Shell; the funds to enable colour printing were provided by
the authors' employers, the authors themselves and by additional donations from Esso and BP. The
success of the meeting was due in large part to the Liverpool postgraduates and technical staff who
took the whole thing over and ran it very smoothly ! (Thanks go to Steve, John, Greg, Greg, Jim, Paul
and Hilary.) My gratitude, for all their work, goes to the contributors, both to the meeting and the
book, as it does to the forty or so reviewers who have put so much effort into manuscripts and
maintaining scientific standards. Nick Parsons of Blackwells helped ensure rapid publication and
Hilary Davies prepared the index. I am especially grateful to my wife, Lesley, who has shown great
patience and given enormous support, practical and moral, throughout the whole enterprise.
JIM MARSHALL
Liverpool, Easter, 1987

Reference
MARSHALL,J. D. 1981. Zoned calcites in Jurassic ammonite chambers; isotopes, trace elements and
neomorphic origin. Sedimentology, 28, 867-87.

vi
 Hydrodynamic modelling of cementation patterns in modern
reefs

Ian R. Goldsmith & Peter King

S U M M A RY : Cementation patterns in modern reefs are inhomogeneous, even on the thin

section scale. The concept of the microenvironment has been developed in order to explain
this irregularity, this microenvironment being determined by both chemistry and
permeability. Identification of the rate limiting step in carbonate cement growth has led to
the suggestion that cementation patterns are controlled principally by the microenvironmen-
tal permeability. In order to test this hypothesis, modelling of the permeability
characteristics of biogenic frameworks has been attempted. Representative matrix
geometries were created mathematically and Poiseuille flow through these has produced
cementation patterns similar to those observed in thin section. Visual comparison of the
features of the real system and model results indicates that the conceptual model is realistic
and permeability may indeed be the major control on cementation.
In the future, modelling of the matrix geometry will be controlled using the measured pore
throat radius distribution of the rock. At that stage the model will be applicable not only to
the prediction of cementation patterns but to many other problems involving fluid flow
through porous media.

Introduction The precipitation and growth of carbonate cements

A fundamental understanding of the controls on
Cementation patterns in reef limestones are cementation must be based on the identification
extremely inhomogeneous on a range of scales,
of the rate limiting step in the process. Crystal
down to and including thin-section level. This growth in the hydrodynamic regime present in
feature can be clearly observed in the photo- reef sediments is almost certainly transport and
micrographs shown in Figs 1 and 2.
not surface controlled (Berner 1980). Hence the
Cores taken through reefs of the SE Florida rate limiting step in the growth of calcium
carbonate platform show an irregular variation carbonate cements is the supply of solutes (Ca ++
of cementation with depth, with cements com-
and HCO3-) to the crystal surface. It is therefore
prising from zero to 15~ of the total rock volume. apparent that the variation in degree of cementa-
In the past this inhomogeneity has been tion may be governed simply by the fluid flow
explained using the concept of the microenviron-
rates, these in turn being determined by local
ment (Schroeder 1972). It has been generally
permeabilities.
considered that the major factor controlling this The suggestion is, therefore, that the microen-
microenvironment is the seawater chemistry; vironments are not chemically controlled (in
spatial variation in the chemical nature of the space) but are permeability controlled. One
pore waters therefore produces the observed could, of course, argue that the microenviron-
variation in the cement products. ments are then chemically controlled in time.
Mass balance calculations show that many
thousands of volumes of porewater need to pass
through a pore in order to fill it with cement
Modelling of permeability characteristics
(Bathurst 1975, Berner 1980); therefore the
diagenetic systems must be physically 'open' on In order to analyse the above hypothesis,
all levels, so that effective chemical exchange is computer programs developed at the BP Re-
possible (Pingitore 1982). However, this appears search Centre have been used to model the
to contradict the concept of a spatially variable permeability characteristics of carbonate frame-
chemical microenvironment which would re- stones. The modelling to date has concentrated
quire at least a partly closed diagenetic system. exclusively on coral framework as this is the

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences,

Geological Society Special Publication No. 36, pp. 1-13.
2 I . R . Goldsmith & P. King

FIG. 1. The inhomogeneous nature of cementation in a Holocene coral framestone from a reef of the SE
Florida Shelf. Cements are acicular aragonite and peloidal high magnesium calcite. Photomicrograph,
crossed polars.

Fit3. 2. The inhomogeneous nature of cementation in a coral framestone from the Belize barrier reef. Note
that some pores are devoid of cement, some have a thin isopachous coating and others are full of aragonite
fibres. Photomicrograph, crossed polars.
 Hydrodynamic modelling in modern reefs 3
major component of the Florida Holocene reefs. Theoretical basis of the model
The results described below represent volumes of
cement only; no attempt has been made to The rock matrix
explain the distribution of different morpholo-
gies and mineralogies. Most importantly, the Network codes describe a regular square grid of
models attempt to show that the distribution of capillary tubes (Fig. 3) which represents the rock
the cements is also determined by the per- matrix (rock in grey, porosity in white). In all of
meability characteristics of the rock matrix. the examples in this paper the grid is 10 by 10
A number of assumptions have been made in square, though theoretically any size can be
order to be able to describe the real system in computed.
simple terms:
(a) The real system is in a laminar flow regime, Fluidflow
i.e. flow through the pores of the system is non- There is considered to be a fixed pressure
turbulent. This assumption can be verified by gradient (P), across the grid; this drives the pore
calculating the dimensionless number--the fluid from bottom to top. Laminar flow is
Reynolds number--as follows: controlled by Poiseuille's law;
NRe _ v.r.p (1) Q = n. r 4 . p. Ap (3)
~t 8.~.l
where v is the fluid velocity in a pore of radius r, p
where Q is the fluid flux through the capillary, r
is the fluid density and ~t is the fluid viscosity.
is the capillary radius, p is the pore fluid density,
This is a standard fluid mechanical term which is
Ap is the pressure difference across the capillary,
used to determine the nature of the flow through
is the viscosity of the fluid and I is the length of
any system. The flow becomes turbulent at
Reynolds numbers between 1200 and 2000. the capillary.
The model then calculates the fluid velocities
Whilst this range is strictly only valid for flow
in each segment of the grid in the following
through a uniform, smooth cylindrical tube, flow
manner. At each node in the grid, fluid must be
in rougher tubes is still laminar up to about
conserved, so the sum of the velocities at each
NRe = 1.
node must equal zero. A standard numerical
Calculation of the Reynolds number using
iterative technique is then used to calculate the
geologically reasonable values for the flow of
velocities in each segment. An initial guess at the
pore fluids ( 1 - 1 0 m d a y - 1 ) , through pores of
pressure difference between each node is made,
radius 1-1000~tm, and standard data for the
and a velocity calculated from that using
viscosity and density of water, reveals a range for
equation (3). The iterative technique is pursued
NReof 10-1-10 -6. This clearly indicates that the
to the point when the sum of all the velocities at
real system is in a laminar flow regime.
every node in the grid is zero. Note that the inlet
(b) Solute transport is considered to be by bulk and outlet boundaries (the top and bottom of the
flow only; there is no significant ionic diffusion. diagrams), are at constant pressure, and that the
Another dimensionless number, the Peclet num- edges of the cell are connected together as though
ber (Lerman 1979) can be used to confirm this the diagrams were wrapped around a cylinder.
assumption:
Q.w
Np - (2) Cementation process
D
where Q is the fluid flux (related to fluid velocity, We are assuming that the rate controlling step in
v), w is a particular length scale and D is the the growth of the cements is the rate of supply of
diffusion coefficient for the solutes of interest, in solutes to the growing surfaces. As we have
this case Ca ++ and HCO3-. For Peclet numbers shown that the main transport of solutes is by
greater than 1, bulk flow is the most important. bulk flow, then the rate of growth is directly
Using values of the fluid velocity of 1- proportional to the flow velocity.
10 m day -1, a length scale (w), of 1 mm to 1 m The cementation process is described as a
and standard data for the diffusion coefficients decrease in the radius of the capillary tubes. This
(both approximately 10 -5 cm 2 s -1) the calculat- decrease is calculated to be directly proportional
ed Peclet number has a range of 10-105 for the to the flow velocity:
real system. Therefore, diffusive mass transport
is negligible when compared to bulk mass
dr_~=/~. v; (4)
dt
transport, so no diffusion term is included and
transport of solutes is considered to be by bulk where ri is the radius of the ith capillary, vi is the
flow only. fluid velocity in the /th capillary and /~ is a
4 I . R . Goldsmith & P. King

proportionality constant. Note that this constant (i) 2x, 2y These are two parameters which
acts as a time-scale and relates the model results control the capillary radius distri-
to geological reality. In the later examples this bution (in the x and y directions
constant is called a cementation factor. respectively) according to the fol-
In the reef system, as cementation progresses, lowing condition:
the pore radii will decrease. From equation (3), The probability of there being a
the fluid flux or velocity will also decrease and so capillary tube of radius r is 89 for;
will the rate of precipitation (from equation (4)).
1 -2<r< 1 +2 (5)
This negative feedback is built into the model by
calculating the overall result via a number of and zero otherwise (Fig. 6).
stages or time steps (usually 10). After each step A value for each parameter is
new flow velocities are calculated from the input at the start of the program.
capillary radii as modified in the previous step. This can be any number between 0
In the figures below, the cement precipitated is and 1.
shown in black.
In the 10 by 10 matrix used throughout this study,
there are 100 capillary segments in both the x-
Matrix geometry and y-axis directions. The use of a probability
The simple example in Fig. 3 is perhaps distribution, such as that above, allows a non-
representative of a supermature sandstone or a uniform capillary radii distribution to be created,
carbonate grainstone, but it is obviously not a simply by the input of a single value for the x and
good description of a coral framework as shown y directions. As the value of 2 is increased, the
in Figs 4 and 5. variety in the capillary radii also increases.
There are several parameters which may be If 2 = 0, from equation (5), r varies from 1 - 0,
used to alter the matrix geometry so that it is to 1 + 0, i.e. all the radii are equal and of unit
more representative of carbonate framestones. size. Similarly, if2 = 0.5, r varies from 0.5 to 1.5
These parameters are input by the user at the units. Increasing 2 to 0.9 produces a wider
start of the computer program and are listed at distribution in the radii, from 0.1 to 1.9 units.
the edges of the figures (Figs 3 and 7 to 15). The The probability distribution (Fig 6), shows that,
parameters are described below: for any value of 2, there is an equal probability

FIC. 3. The regular square grid of capillary tubes which forms the basis of the model matrix. In these examples
the grid is 10 x 10 square, though any size can be computed. The rock is in grey and the porosity in white.
 Hydrodynamic modelling in modern reefs 5

FIGS 4 and 5. Typical appearance of Holocene coral frameworks in thin section. This is the type of carbonate
framestone whose permeability characteristics are being modelled here. The sections are parallel and
perpendicular to the corallites. Photomicrographs, plane polarized light.

(89 that any radius within the range of 1 - 2, to respectively) which are non-conducting. This
1 + 2, will occur in the grid. therefore introduces a degree of m a t r i x
connectivity.
(ii) Px, Py These two parameters represent
the fraction of the capillary seg- T h e parameters can have any value from 0 to 1.
ments (in the x and y directions Hence, Px = 0.5 means that half of the capillary
6 I. R. Goldsmith & P. King
assuming the system to be in a laminar flow
regime. In reality, eddies and edge effects are
bound to deflect the flow of pore waters through
the rock matrix. However, introduction of
irregularities into the matrix in the following
examples does indeed cause precipitation of
1 cements on the x-axis surfaces.
1-;k 1 I*X This example serves as a standard with which
Capillary Radius r to compare the effect of changing the matrix
FIG. 6. The capillary radius distribution function. parameters. The uncemented 'gap' at the base of
all the figures is a function of the graphics
program and should be ignored.
segments in the x direction are non-conducting.
The visible effect on the matrix is that half of the
x-axis segments are sealed. Var&tion in Px, ey
The program calculates the initial and final Figures 8 and 9 show the effect of introducing a
permeabilities (K init. and K final) of the grid proportion of pores which are non-conducting.
from a simplified version of Poiseuille's law; Sealing half of the pores in the x direction has
very little effect on the cementation pattern as it
K.ap does not alter the flow path or the initial
Q- (6)
~t.L permeability (K init. for Figs 7 and 8 is
1.99 x 10-6).
where Q = fluid flux out of the top of the grid, Ap However, it is interesting to note that sealing
is the pressure difference across the grid, I-t is the half the pores in the y direction has a marked
fluid viscosity and L is the length of the grid. effect (Fig. 9). In this case the matrix tortuosity is
These permeabilities are of course only relative. greatly increased and the flow is forced along
The initial porosity can also be chosen by the capillaries in the x direction. An appreciable
user; the program will then calculate the final degree of inhomogeneity has been created. Note
porosity (~o init. and ~0 final). the concomitant reduction in initial per-
Variation of these input parameters leads to a meability. (K init. is 8 x 10-7 in Fig. 9 compared
wide range of results, some of which are shown in with 1.99 x 10-6 in Fig. 7.)
Figs 7 to 15.
Variation in 2x, )ty
Results In the examples discussed above (Figs 7 to 9), 2x
and 2, are both zero, so the capillary radius
Presentation of the results of this modelling is in distribution function (Fig. 6) reduces to a single
two parts: value and all the pore radii are equal. Varying 2
introduces a distribution of pore radii as shown
(i) the effects of the matrix geometry on the in Fig. 10, where )-x = 0, 2y = 0.5. Note that for
cementation patterns; the analogous situation, 2x = 0.5, ,,],y= 0 there
(ii) the effects of varying fl, the cementation would be very little effect on the cementation
rate constant. pattern (for the same reason as Px has no effect)
Each example below is the result of an individual as this has no effect on the flow paths. In Fig. 10
experiment created using the parameters listed at there is therefore a distribution of y-axis capillary
the edges of the relevant figure. radii, with all x-axis radii equal and all pores
conducting (Px, Py = 0).
The Poiseuille flow dependence on the capil-
Matrix geometry lary radius raised to the fourth power is evident
here (Q oz r% equation (3)). A substantial degree
Figure 7 shows the cementation pattern created of variation in the cement thickness has been
using the regular square grid as in Fig. 3. The created, with the larger pores initially receiving a
precipitated cement (shown in black) occurs in a greater volume of cement.
uniform pattern along the y-axis surfaces only.
This is because there is no pressure difference
across the capillaries in the x direction, and so Representation of coral framework
flow is in the y direction only. A combined variation of 2x, Px and Py, creates a
This result is in itself geologically unreason- matrix geometry that is a reasonable representa-
able, though mathematically valid, as we are tion of coral frameworks (Figs 11-15).
 Hydrodynamic modelling in modern reefs 7

FIG. 7. Regular square grid. Cementation (shown in black) occurs along the y-axes only because flow is in
that direction only. This example serves as a standard with which to compare the effects of changing the
matrix dscription.

FIG. 8. In this example, half of the pores in the x direction are non-conducting. The effect on the cementation
pattern is minimal and flow is still only in the y direction.
8 I . R . Goldsmith & P. King

FIG. 9. Sealing half of the pores in the y direction has a more pronounced effect on the cementation pattern.
The matrix tortuosity is increased and flow is forced along capillaries in the x direction. Note the concomitant
reduction in initial permeability (cf. Fig. 7).

FIG. 10. 2y = 0.5 creates an irregular distribution of y-axis capillary radii. As the rate of cementation is
directly proportional to the flow velocity, which is itself dependent on the fourth power of the radius, larger
pores receive substantially more cement.
 Hydrodynamic modelling in modern reefs 9

FIG. 11. Results more representative of the reality (cf. Fig. 2) can be produced by increasing the matrix
connectivity as in this example. Both dead end and non-effective porosity are apparent here, and there is a
considerable variation in the degree of cementation.

FIG. 12. Using the same matrix description as in Fig. 11, the porosity has been increased. Due to the r 4
dependence, the cementation factor has to be reduced by an order of magnitude.
 IO I.R. Goldsmith & P. King

FIG. 13. It is interesting to observe the effects o f progressively increasing ft. In this e x a m p l e fl = 500 (cf. Fig.
12, fl = 100).

X : 0.500
x

A = 0.000
y

P = 0.500 Ntime=

10
P = 0.500
y

fl=4000.O ~m,t:O.271

K
Init

2.70E--06

2.13E--10

FIG. 14. In this example fl = 4000 and much of the effective porosity has been filled by cement.
 Hydrodynamic modelling in modern reefs II

FIG. 15. Here/3 is very large and the grid has been fully cemented. Note that as flow is laminar, dead end
porosity remains uncemented. Non-effective porosity receives no fluid flow and so is also unaltered.

~00
+++++++'~++ "t"+-t,t- + + + + + +
+
- 9O
++%++
~ so

._ +
+

E
~> 40

9-g 30

_~ 20

IO
+
A 1 1 i L L 1 ~ ! i J J
0.5 1.0 1.5 2.0 2 5 3.0 5.5 4.0 4.5 50 55 6.0 6.5 7.0
Log (pore diameter [Angstroms])

FIG. 16. The pore throat radius distribution for an unaltered Holocene coral framestone. This is a capillary
pressure curve derived by mercury injection porosimetry. In future modelling these types of data will be used
to control the capillary radius distribution of the model matrix.

In Fig. l l, Px and Py are both 0.5, thereby variation in the degree of cementation is appar-
increasing the matrix connectivity. Many of the ent. The lower left corner of Fig. 11 has received
features apparent in the real system are now no cement whatsoever as all the porosity is either
present in the model results. Both dead-end and non-effective or dead-end. Some of the pores in
non-effective porosity are present and a wide the upper left side have received small volumes
I2 I . R . Goldsmith & P. King

of cement, but the majority of the cement has Holocene coral framestone in Fig. 16. This
been 'precipitated' in the centre and right side of modification will permit modelling of the
the grid. permeability characteristics of any rock
Using this matrix description, the following given its pore throat distribution.
examples show the effects of varying the porosity (2) At present the model operates in two
and the cementation factor//. dimensions only which clearly involves some
In Fig. 12 the input parameters are the same as fundamental approximations. Expansion of
for Fig. 11, but the initial porosity figure has the capillary grids to three dimensions would
been increased. The capillary radii are therefore be advantageous. This will necessitate
increased and again the Poiseuille dependence changes in the method of visualization of the
on r 4 is evident: to produce roughly the same results, either by observation of successive
volume of cement as in Fig. 11, the cementation 2-D sections or by producing a capillary
factor 13 has to be reduced from 2000 to 100. radius distribution after cementation.
(3) Modelling on a larger scale requires a
detailed knowledge of the porosity distribu-
The cementation factor/~ tion on that scale followed by use of those
The parameters used in Fig. 12 produce the most data for the description of the model matrix.
accurate representation of coral framestone (4) Cementation interacts dynamically in both
geometry as observed in thin section. Using this space and time with other processes of reef
basis it is interesting to observe the effects of formation (Schroeder & Zankl 1974). Reef
increasing fl on the cementation patterns (Figs construction, reef destruction and sedimen-
13, 14 and 15). In Fig. 15, fl is very large and the tation alter the initial depositional per-
grid has been fully cemented, with the exception meability characteristics of the reef lime-
of non-effective and dead-end porosity. Com- stones, so affecting the continuing cementa-
parison of this result with the photomicrograph tion patterns. Some aspects of these pro-
in Fig. 2 suggests that those empty pores adjacent cesses will be included in the modelling in the
to fully cemented ones are probably non- future.
effective porosity only revealed in the two-
dimensional section. Such pores have very little Conclusions
or no fluid flow through them.
Whilst the last four examples are separate The results of this first stage model are very
results created by increasing the cementation promising. Many of the features apparent in the
parameter 13, it is possible to use the model in an real system have been reproduced using a fairly
incremental manner such that successive cement simple mathematical description of the frame-
zones are created. Figures 7 to 15 were created stone matrices.
using 10 time steps, with only the final result The creation of the two-dimensional matrix
displayed. By setting the program to display the has indicated that a significant proportion of the
results of each time step, a series of 10 cement porosity observed in thin section may in fact be
zones is produced. The thickness of each zone non-effective. This therefore explains why areas
progressively decreases according to the depen- devoid of cement are often observed in an
dance on the fourth power of the radius. otherwise well cemented sample. In addition,
much of the porosity may be poorly interconnect-
ed with the major flow conduits (dead-end
Discussion porosity), and will therefore receive substantially
less cement than the well connected pores.
The modelling to date has been on the thin The results of the cementation experiments
section level only. Our aim is to be able to model
indicate that larger pores initially receive a
the cementation patterns on a variety of larger
greater volume of cement than the smaller pores
scales. In order to do this, some of the limitations
and confirm that the rate of cementation will be
of the present network codes need to be
reduced as the pores become restricted.
modified.
The most important conclusion of this study is
(1) The use of a regular square grid restricts the that the model indicates that the variation in the
variation in capillary radii essentially to one volumes and the spatial distribution of precipi-
order of magnitude. Modification of the tated cement is controlled by the microenviron-
network codes will eventually allow control mental permeability alone. We have not yet
of the capillary radius distribution of the developed a statistical method to compare
model using the measured pore throat radius rigorously the results and the reality, but visual
distribution of the rock. An example of the comparison suggests that the conceptual model is
pore throat radius distribution is given for a successful.
 Hydrodynamic modelling in modern reefs 13
ACKNOWLEDGMENTS: We gratefully acknowledge the and Dr M. E. Tucker for supervision throughout this
permission of the British Petroleum Company to research. We are grateful to Professors R. N. Ginsburg
publish this work. One of us (IRG) would like to thank and E. A. Shinn for kindly donating the samples on
NERC and British Petroleum for financial assistance which this research has been carried out.

References
BATHURST, R. G. C. 1975. Carbonate Sediments and SCHROEDER, J. H. 1972. Fabrics and sequences of
their Diagenesis. Elsevier, Amsterdam. submarine carbonate cements. GeologischeRunds-
BERNER, R. A. 1980. Early Diagenesis: a Theoretical chau, 61, 2, 708-30.
Approach. Princeton University Press, New
Jersey. --& ZANKL, H. 1974. Dynamic reef formation: a
LERMAN, A. 1979. Geochemical Processes, Water and sedimentological concept based on studies of
Sediment, pp. 64-65. Wiley, New York. Recent Bermuda and Bahama Reefs. Proceedings
PINGITORE, N. E. 1982. The role of diffusion during of the 2nd International Coral Reef Symposium, 2,
carbonate diagenesis. Journal of Sedimentary Pe- 314-28.
trology, 52, 27-39.

IAN R. GOLDSMITH,Department of Geological Sciences, South Road, University of Durham,

Durham DH1 3LE, UK.
PETER KING, Reservoir Technology Branch, BP Research Centre, Chertsey Road, Sunbury
on Thames, Middlesex TWl6 7LN, UK.
Some aspects of diagenetic sulphate-hydrocarbon redox reactions
H. G. Machel

SUMMARY: Sulphate-hydrocarbon redox-reactions occur at two specific diagenetic

temperature/thermal maturity levels: less than about 75-85~ (0.2-0.3~0 Ro), and more than
100-140~ (> 1.5% R0), respectively. In low-temperature/maturity environments these
redox reactions take place only with the mediation of bacteria. In high-temperature/maturity
environments these reactions take place thermochemically, and certain catalysts must
interact in order to overcome the high activation energies and to sustain the reactions at
geologically significant rates.
The reaction products and by-products may be identical for both temperature/maturity
levels: altered and oxidized hydrocarbons (including bitumen), hydrogen sulphide, metal
sulphides (including Mississippi Valley Type deposits), elemental sulphur, carbonates
(mainly calcite and dolomite), and other minerals. An important by-product of these redox
reactions may be porosity resulting from the dissolution of solid sulphates and/or the
carbonate host rock. The net reaction is exothermic, and the released heat may generate a
geothermal hot-spot in some cases.

Introduction variety of dissolved sulphates and hydrocarbons.

Later experiments showed that TSR can take
The association of dissolved sulphates and place at temperatures at least as low as 175~
hydrocarbons is thermodynamically unstable in (summarized in Trudinger et al. 1985). Geologi-
diagenetic environments (for the purpose of this cal evidence, however, suggests lower minimum
paper, the term hydrocarbons includes any type temperatures, perhaps 100-140~ (e.g. Powell &
of organic matter, carbohydrates, kerogen, crude Macqueen 1984, Siebert 1985, Machel & Krouse,
oil, bitumens, dissolved and gaseous organic in prep. a). In either case, TSR is restricted to
compounds). Hence, redox reactions occur, high-temperature diagenetic and hydrothermal
either with or without the mediation of bacteria. environments and probably is the main process
For simplicity, these reactions are discussed generating large quantities of hydrogen sulphide
from the vantage point of sulphate reduction, in numerous deep subsurface sour gas provinces
which always implies concomitant hydrocarbon of the world (Orr 1977). Presumably TSR is also
oxidation. Bacterially mediated sulphate reduc- involved in the formation of sulphide ore
tion is called BSR (bacterial sulphate reduction), deposits, i.e. some Mississippi Valley Type
and abiologically mediated sulphate reduction is deposits (Powell & Macqueen 1984).
called TSR (thermochemical sulphate reduc- Despite their frequent occurrence, some
tion). TSR is also called non-biogenic (Barton aspects of sulphate-hydrocarbon redox reactions
1967), non-microbial (Orr 1974), organic (Oh- are not well understood. Evidence for reaction
moto & Rye 1979), thermal (Siebert 1985), and paths, products, and conditions is scattered in
abiological (Trudinger et al. 1985) sulphate the literature and partly contradictory, especially
reduction. regarding the associated changes in pH, carbon-
BSR was discovered by a bacteriologist who ate alkalinity, and precipitation or dissolution of
published his findings in a journal of bacterio- carbonates. A major objective of the present
logy, parasitology, infectious diseases, and paper is to clarify some of these problems.
hygiene (Beijerinck 1895). BSR occurs in a large Secondly, it is not clear under what circum-
variety of sedimentary and low-temperature (less stances sulphate-hydrocarbon redox reactions
than about 85~ diagenetic environments in- proceed without the mediation of bacteria. In
cluding ground water aquifers (e.g. Champ et al. particular, Trudinger et al. (1985) suggested that
1979), marine sediments (e.g. Berner 1980), TSR may not occur at geologically significant
reefal carbonates and layered or diapiric evapor- rates at diagenetic temperatures because of
ites (e.g. Feely & Kulp 1957, Sassen 1980), and kinetic barriers and the paucity of reactive
clastic rocks (e.g. Coleman 1985). First data on hydrocarbons. Therefore, another objective of
TSR were presented by Toland (1960) who the present paper is to discuss reaction kinetics
performed hydrothermal experiments with a and catalysts for TSR. Elemental and isotopic

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences,

Geological Society Special Publication No. 36, pp. 15-28. I5
I6 H.G. Machel
TABLE 1. Reaction scheme for bacterial and thermochemical sulphate reduction. See text for explanation

Reaction 1: paraffinic hydrocarbons 9- - ~ biodegraded hydrocarbons

Reaction 2: crude oil light crude oil + H2S + CH 4

Reaction 3 a: 4R-CH 3 + 3SO42" + 6H + 9- - ~ 4R-COOH + 4 H 2 0 + 3H2S

b: R-CH 3 + 2R--CH 2 + CH 4 + 3SO42" + 5H + 3R-COOH + HCO 3" + 3 H 2 0 + 3H2S

c: 2 C H 2 0 § SO42" 2HCO 3" + H2S

Reaction 4 a: 2H2S + 0 2 2S ~ + 2 H 2 0

b-1 : 3H2S § SO42" + 2H + 4S ~ + 4 H 2 0

b-2: H2S + SO42" + 2H + S ~ + 2 H 2 0 + SO 2

c: H2S + hydrocarbons --~ S ~ + altered hydrocarbons

d: S 2" ,~ S~

Reaction 5: 4S ~ + 1.33(-CH2- ) + 2.66 H 2 0 + 1.33 OH" 4H2S + 1.33 HCO 3"

Net Reaction: hydrocarbons + SO42" altered hydrocarbons + bitumen

HCO 3" + H2S (+CO2?) + heat

characteristics of the reaction products of BSR free of base and transition metals (the addition of
and TSR will be discussed in a forthcoming these metals is discussed separately), and meta-
paper (Machel & Krouse, in prep. b). Some of stable or transient reaction products such as
these characteristics are mentioned below where polysulphides are not included. It should also be
necessary. kept in mind that each reaction in Table 1 is a net
mass and charge balance reaction consisting of
several sub-reactions. These sub-reactions are
Important reactions omitted from Table 1 for simplicity.

Many reactions have been proposed for BSR and

Reaction 1 : biodegradation of hydrocarbons
TSR. However, no previous reaction scheme
provides a satisfactory explanation for all the Biodegradation of hydrocarbons is not a pre-
important reactants and products occurring in requisite where methane is the main carbon
diagenetic environments. Therefore, a reaction source, or where sulphate reduction proceeds
scheme consisting of previously published and of thermochemically. However, biodegradation of
modified reactions is suggested that are believed hydrocarbons is a prerequisite for BSR in
to be a better representation of sulphate-hydro- environments devoid of methane for two rea-
carbon redox reactions in diagenetic environ- sons: (1) sulphate-reducing bacteria generally
ments (Table 1). The reactions in Table 1 have cannot digest n-paraffins, which are common in
been (a) demonstrated experimentally, (b) ob- non-biodegraded hydrocarbons such as crude
served in natural environments, (c) calculated oil; and (2) sulphate-reducing bacteria depend on
thermodynamically, or (d) inferred on the basis the metabolic residues of hydrocarbon biodegra-
of elemental or isotopic evidence, as indicated in dation as nutrients e.g. various organic acids
each case for verification. The redox-steps in (experimentally demonstrated by Bailey et al.
reactions 3, 4, and 5 (Table 1) are valid for 1973, Nazina et al. 1985, Jobsen et al. 1979). This
bacterial and for abiological sulphate-hydrocar- is supported by evidence from natural environ-
bon redox reactions, as are the major reaction ments of BSR where most hydrocarbons are
paths of many redox reactions (e.g. the forma- biodegraded oils, i.e. typically heavy to medium
tion of pyrite: Morse et al. 1987). BSR and TSR gravity naphthenic oils with most n-paraffins
are therefore discussed together for each of these and isoprenoids removed (Davis & Kirkland
redox-steps. However, the reaction scheme in 1970, Philippi 1977). Accordingly, reaction 1
Table 1 is valid only for systems that are initially (Table 1) represents the sequential and partially
 Sulphate-hydrocarbon redox reactions I7

overlapping oxidation of (a) n-paraffins, (b) partly oxygenated compounds (experimentally

isoprenoids, and (c) lower-ring naphthenes and shown by Toland 1960, Davis & Yarbrough
aromatics with increasing extent of biodegrada- 1966, Kiyosu 1980, Trudinger et al. 1985 and
tion. As used in this context, the term biodegra- references therein). As in the case of the sulphate
dation designates hydrocarbon oxidation/de- species, reaction rates vary for different hydro-
composition by aerobic bacteria. Hydrocarbon carbons, and certain compounds are non-reac-
decomposition by fermenting, sulphate reduc- tive (Toland 1960, Trudinger et al. 1985).
ing, or methanogenic bacteria is not called For these reasons, and because there are
biodegradation. conflicting data in the literature regarding the
reaction products, reaction 3 has been written in
three alternative ways (Table 1). Reactive or-
Reaction 2: thermal cracking o f crude oil ganic compounds are: alkanes/paraffins in reac-
Reaction 2 (Table 1) is thermal cracking of crude tion 3a; alkanes/paraffins, alkenes/olefins, and
oil which, among other reaction products, results methane in reaction 3b; and carbohydrate in
in the formation of methane and (a few per cent) reaction 3c. Obviously, reactions 3a, 3b and 3c
H2S (e.g. Orr 1977). Reaction 2 may be involved are not the only possibilities. For example,
in TSR because the resulting H2S is a catalyst for simultaneous and intermediate reactions are
reactions 3 and 5 via reaction 4b (experimentally possible, or methane could be the only organic
shown: e.g. Toland 1960). Reaction 2 is not reactant. Excepting the last possibility, the main
necessary for BSR. reaction products of reaction 3 are organic acids
(i.e. carboxylic acids for paraffins, olefins, and
alkylated aromatic compounds as reactants),
Reaction 3: &itial S - O bond rupture and reduction
bicarbonate ions, and hydrogen sulphide (experi-
o f S 6+
mentally and theoretically shown for BSR: e.g.
Reaction 3 consists of several sub-reactions Rickard 1969, Berner et al. 1970; for TSR: e.g.
between sulphates and hydrocarbons involving Toland 1960). One important reaction product is
the initial S-O bond rupture and reduction of S6§ not included in Table 1 : some TSR experiments
to lower valence states (most S6§ is reduced to produced 'black precipitates' (e.g. Toland 1960,
$2-). Reaction rates vary for different sulphate Kiyosu 1980). The equivalent to these precipi-
and hydrocarbon species, and not all species are tates in natural environments is bitumen. Subor-
reactive. Firstly, only dissolved sulphate can be dinate reaction products, also omitted from
utilized during BSR and TSR. Low-temperature Table 1 for simplicity, are liquid and gaseous
experiments demonstrate that sulphate-reducing inorganic and organic sulphur compounds, in-
bacteria cannot directly decompose solid sul- cluding polysulphides (e.g. Toland 1960, Burnie
phates, although they can solubilize solid sul- 1979). Also, small amounts of CO2 and CO gas
phates via removal (reduction) of sulphate from evolved during BSR and TSR laboratory experi-
solution (experimentally shown by Bolze et al. ments (e.g. Toland 1960, Ward & Brock 1978,
1974 and McCready & Krouse 1980). Similarly, Kiyosu 1980). Where methane is the only organic
TSR of solid sulphates is not possible in reactant, all carbon is oxidized to carbonate
diagenetic environments: non-aqueous mixtures species and some CO2 and CO, and no bitumen
of hydrocarbons and solid sulphates remain is formed.
unreactive even at temperatures of 180-315~ There is abundant geological evidence for an
(Toland 1960). Furthermore, the reaction rate increase in carbonate alkalinity in solution, and
depends on the specific sulphate species in for precipitation of carbonates, as a result of
solution, and some species (i.e. NaSO4-) appear sulphate reduction in a variety of diagenetic
to be non-reactive even at temperatures up to environments (e.g. Berner et al. 1970, Curtis
300~ (Kiyosu 1980). Regarding the hydrocar- 1977, Berner 1980, Jeffries & Krouse 1984).
bons, a clear distinction has to be made between Therefore, reactions 3a, 3b and 3c (Table 1) have
BSR and TSR. Sulphate reducing bacteria can been written in such a way as to allow for
utilize a large variety of low-molecular weight carbonate precipitation. For example, reaction
organic compounds including many products of 3a involves a net increase in pH because the
aerobic biodegradation, but generally not n- organic acid released is very weak. This would
paraffins (e.g. Jobson et al. 1979, Peck 1984, lead to carbonate precipitation utilizing carbon-
Nazina et al. 1985). On the other hand, relatively ate (and metal) ions that are already in solution
few hydrocarbons are known to react abiologi- (these carbonates would not contain organic
cally with sulphate, but they include n-paraffins: carbon, as observed in some geological environ-
low-n-alkanes (i.e. methane, ethane), low-n- ments: Machel & Krouse, in prep. b). Alternati-
alkenes, n-octadecane, carbohydrates (i.e. su- vely, carbonates would precipitate as a result of
gar), alkylated aromatic compounds, and other, an increase in carbonate alkalinity (release of
I8 H. G. Machel

HCO3- with organic carbon) in the cases of electron acceptor to yield sulphide, but some
reactions 3b and 3c. It should be kept in mind, reduce nitrate, nitrite, and other compounds and
however, that there is considerable controversy can grown even in the absence of detectable
regarding the changes of pH and carbonate sulphate (Peck 1984, Postgate 1984). For reac-
alkalinity, and regarding carbonate precipitation tion(s) 3 to proceed thermochemically, catalysts
or dissolution resulting from sulphate reduction. and temperatures in excess of 100-140~ are
For example, experimental evidence suggests necessary (see discussion of kinetics, below).
that the pH may increase or decrease during BSR
depending on the types of hydrocarbons utilized Reaction 4: part&l oxidation o f S 2- to S ~
by the bacteria (e.g. Birnbaum & Wireman
1984). Chemical models are partly contradictory Sulphide-sulphur oxidation may take place in
and suggest carbonate dissolution or precipita- several alternative ways. Reactions 4a and 4b
tion depending on the involvement of other represent reversible reactions for abiological
phases and components, i.e. silicates, carbon- oxidation of sulphide, step 4c represents the non-
ates, sulphides, and ammonia. For example, reversible abiological oxidation of sulphide by
Berner et al. (1970) predicted an increase in hydrocarbons, and step 4d represents a variety of
carbonate alkalinity and concomitant carbonate non-reversible bacterial reactions.
precipitation in organic-rich silicate muds. The Reaction 4a: Abiological oxidation of sulphide
data of Berner et al. (1970) also suggest that with molecular oxygen is possible where HzS
carbonate precipitation occurs in chemically escapes into the atmosphere, and where H2S
'inert' and carbonate sediments. On the other comes into diffusive or hydrological contact with
hand, Gardner's (1973) model predicts carbonate dissolved oxygen (Berner 1980, Morse et al.
dissolution in 'inert' and carbonate sediments as 1987). Abiological sulphide oxidation with dis-
a result of sulphate reduction. It also predicts solved oxygen was experimentally demonstrated
carbonate precipitation along with sulphide for temperatures as low as 70~ and it is
precipitation in sediments containing iron ox- thermodynamically likely for temperatures in
ides. This is at variance with the models and excess of 25~ (Davies et al. 1970). Importantly,
observations of Anderson (1983) and Anderson elemental sulphur should be formed only in
& Garven (1987) who predict carbonate dissolu- environments of relatively low partial pressure of
tion if sulphides precipitate, unless the acid oxygen, or where oxygen is supplied relatively
released during sulphide precipitation is con- slowly to the reaction site (reaction 4a). Where
sumed by other reactions, i.e. by sulphate oxygen is abundant, hydrogen sulphide is oxi-
reduction or reactions involving silicates. Reac- dized to sulphuric acid according to
tions 3a, 3b and 3c (Table 1) conform with the
models of Berner et al. (1970), Anderson (1983) H2S + 202 *-+2H+ + SO42-.
and Anderson & Garven (1987), and with the This reaction is omitted from Table 1 because
evidence from most natural environments of elemental sulphur is not a reaction product.
BSR/TSR, in that carbonates precipitate where
base and transition metals are scarce or absent. Reaction 4b." Abiological oxidation with excess
Reaction(s) 3 represent BSR and TSR. In the dissolved sulphate at low pH is not likely at low
case of BSR, the temperature must be cool temperatures but possible during TSR (experi-
enough to permit bacterial metabolism, that is mentally and theoretically shown by Husain
less than about 45~ for most sulphate-reducers, 1967, Davies et al. 1970, Husain & Krouse 1978).
although some are known to metabolize between Water and elemental sulphur are the only
about 60 and 92~ (Peck 1984, Trudinger et al. reaction products where the hydrogen sulphide
1985, Nazina et al. 1985, Stetter et al. 1987). concentration is high (reaction 4b-l). Sulphur
Furthermore, the environment must be anaero- dioxide is produced along with elemental sulphur
bic because sulphate-reducing bacteria cannot where little hydrogen sulphide is available
grow under aerobic conditions (experimentally (reaction 4b-2) (experimentally shown by Husain
shown: e.g. Jobson et al. 1979). On the other & Krouse 1978).
hand, sulphate-reducing bacteria are a taxono-
mically and physiologically extremely diverse Reaction 4c: Above about 100~ H2S reacts
group of microorganisms consisting of many abiologically with (mainly saturated) hydrocar-
genera (Desulfo-x) that can grow over a wide bons to produce elemental sulphur and NSO-
range of temperatures (about +0-92~ salini- compounds (organic compounds containing N,
ties, and water compositions (Peck 1984), which S, O, e.g. resins), if catalysts such as silica gel
explains their common occurrence in low-tem- and/or clays are present (experimentally shown:
perature diagenetic environments. All sulphate- e.g. Rudakova & Velikovskii 1947, Ho et al.
reducing bacteria use sulphate as terminal 1974).
 Sulphate-hydrocarbon redox reactions I9

Reaction 4d: Bacterial oxidation of sulphide was with isotopically light organic carbon are found
experimentally shown and observed in natural in natural environments of BSR and TSR (e.g.
environments (Chen & Morris 1972, Pawlowski Powell & Macqueen 1984, Sassen in press).
et al. 1979, Sieburth 1979, Paull et al. 1984). Furthermore, reaction 5 could destroy light
These bacteria may be anaerobes (e.g. photosyn- saturated hydrocarbons and generate naphthenic
thetic bacteria, Chromatium) or aerobes (Thio- acid, aromatic compounds, and numerous inor-
bacillus, Beggiatoa, and other genera). However, ganic and organic sulphur compounds (polysul-
even bacteria that need free oxygen live as phides, mercaptans, thialkenes, thiopenes, and
microaerophiles or anaerobes, because a thin others: e.g. Ho et al. 1974, Burnie 1979).
and almost anaerobic aqueous boundary layer
surrounds the colonies even in turbulent water
N e t mass balance
(Jorgensen & Revsbech 1983). Oxygen reaches
the bacteria only via diffusion through this Using a reaction such as reaction 3c (Table 1) as
boundary layer. Another important factor for a net mass balance reaction is correct (e.g. Berner
bacterial sulphide-sulphur oxidation is pH. Chen et al. 1970), but it does not account for all
& Morris (1972) suggested that the reaction rate important reaction products. Reactions 4 and 5
during aerobic bacterial oxidation of sulphide is do not seem to be necessary to account for net
highest at intermediate pH (about 8). However, sulphate reduction, but these reactions are likely
the optimum pH depends on the type of bacteria to occur simply because the oxidizing bacteria or
carrying out oxidation. For example, Beggiatoa oxidizing agents are present. Once instigated,
and Thiotrix thrive best at slightly alkaline pH, reactions 3, 4 and 5 proceed simultaneously
but Thiobacillus thrives best in rather acidic because some of the reaction products (i.e.
environments (Larkin, pers. comm. 1986). hydrogen sulphide, elemental sulphur) are used
The electron acceptors during bacterial sul- up as they are generated. Therefore, reactions 3,
phide-sulphur oxidation may be different for the 4 and 5 can be summarized in a net mass balance
various genera and environments, and usually reaction (Table 1) representing the reaction
involve several enzyme-catalysed steps. There- products if all sulphur generated during reaction
fore, the simplest representation of the reaction 4 is used up in reaction 5. Hydrocarbons are
path during bacterial sulphide-sulphur oxidation expressed with words rather than formulae
is the valence change of sulphur (reaction 4d). because there are numerous possibilities to write
Regarding mass balance, however, reaction 4a and balance the net reaction. Important reaction
may be taken to represent bacterial sulphide- products are altered (non-polymerized) hydro-
sulphur oxidation in natural oxygenated environ- carbons, bitumen, HCO3-, and H2S. No bitu-
ments, and reaction 4c may be taken to represent men is formed where methane is used as the only
anaerobic bacterial sulphide-sulphur oxidation. carbon source. Some carbon may evolve as
Reaction 4b is probably not an alternative in this carbon dioxide gas. The net reaction is exother-
context, because no bacteria are known that can mic (discussed below).
couple sulphide oxidation with sulphate reduc- For simplicity, polysulphides and other com-
tion (Larkin, pets. comm. 1986). pounds with intermediate valence states of
sulphur have not been included in the reaction
Reaction 5: reduction of S ~ to S z- scheme of Table 1 although they are undoubtedly
involved (e.g. Davis & Kirkland 1970, Orr 1974,
At temperatures in excess of about 100~ Krouse 1977, Morse et al. 1987). For example,
elemental sulphur is an active oxidizing agent for elemental sulphur and excess sulphide may react
many organic compounds (represented by the abiologically to form polysulphides (experimen-
methylene functional group in Table 1), resins tally shown by Chen & Morris 1972). Polysul-
and asphaltenes generally being the most reac- phides, generated in this way, or during reactions
tive (experimentally shown with bacteria: e.g. 3 or 5, may react with bicarbonate to form
Stetter & Gaag 1983, Peck. 1984; and abiologi- elemental sulphur (Davis & Kirkland 1970), and
cally/thermochemically: Pryor 1962, Douglas & polysulphides may be used in reaction 5 instead
Mair 1965, Valitov 1974). Therefore, reaction 5 of S~ (experimentally shown: e.g. Orr 1977).
(Table 1) is valid for BSR and TSR. As in the Furthermore, various bacteria may mediate
case of reaction 3, reaction 5 is a very simplified redox-steps that are not shown in Table 1. For
representation of the involved redox-steps. Ma- example, bacteria oxidizing sulphide to sulphur
jor reaction products are HzS and bicarbonate (reaction 4) may be accompanied by other
ions (and some COz and CO). Experiments also bacteria (e.g. certain species of Thiobacillus,
produced 'viscous black tar' and 'dark viscous Thiomicrospora, and other genera: Sieburth 1979)
gum' with up to 11~ S (Douglas & Mair 1965). which further oxidize the elemental sulphur to
Similarly, sulphurized bitumens and carbonates sulphate if the environment obtains some dis-
20 H. G. Machel

solved oxygen, and should there be an appropri- number of partially reduced sulphur compounds
ate balance between oxygen, sulphide, carbon (S ~ polysulphides, sulphite, thiosulphate) by
dioxide, organic compounds, and pH (experi- reaction with dissolved sulphate (similar to
mentally shown by Iwatsuka & Mori 1960, reaction 4b, Table 1 ; experimentally shown: e.g.
Nakai & Jensen 1964, Ivanov 1968, Larkin Toland 1960). These compounds appear to be
1981). Of course, oxidation of sulphur to sulphate more reactive than SO42- or HzS, forming H2S,
by oxygen may also proceed abiologically, but carboxylic acids, CO,, NSO-compounds, and
this takes place much more slowly (Chen & other polysulphides via reactions 3 and 5. Hence,
Morris 1972). two possibilities exist for the initial sequence of
reactions during TSR, assuming that reaction 2 is
the most likely process to provide 'catalytic' H2S
in deep subsurface diagenetic environments: 2,
Geological and geochemical 4b and/or 4c, 5, 3; or 2, 4b and/or 4c, 3, 5 (note:
implications reduced sulphur other than H2S , such as NSO-
compounds from early diagenesis, may also act
The reactions listed in Table I have a number of as catalysts: Burnie 1979). Whatever the initial
geological and geochemical implications, de- sequence, reactions 3, 4 and 5 proceed simulta-
pending on the types of reactive hydrocarbons, neously once they are instigated because H2S
presence or absence of base and transition generated in reactions 3 and 5 is (at least
metals, presence of catalysts, and temperature partially) recycled in reaction 4. As in the case of
(mainly controlling bacterial metabolism and BSR, S~ accumulates as a net reaction product
kinetics of TSR). only if reaction 5 does not proceed or is retarded.
This happens if the hydrocarbons suitable for
TSR are no longer available, or if they are not
Initial order of reactions for non-gaseous hydro-
supplied fast enough (see discussion of reaction
carbons
kinetics, below).
Regarding BSR with non-gaseous hydrocarbons
as a carbon source (i.e. crude oil, but also solid Methane
and dissolved hydrocarbons), Table 1 is a In natural environments, methane is the only
simplification of complex associations of bacte- gaseous hydrocarbon abundant enough for large-
ria performing specific redox-steps (including scale sulphate reduction. Methane could be
intermediate redox-steps that are not shown for derived from methanogenic bacteria (in low
simplicity). Most bacteria mediating reactions 1 temperature/maturity environments), or from
and 4 are aerobes, whereas those mediating thermal cracking of hydrocarbons (reaction 2 : in
reactions 3 and 5 are anaerobes. Initially, the high-maturity environments). The first case was
normal sequence of reactions for BSR would be observed or inferred in natural BSR environ-
1-3. If the environment obtains some oxygen (i.e. ments on the basis of isotopic and circumstantial
at the groundwater/oil interface), reaction 4a will evidence (e.g. Barnes & Goldberg 1976, Kirk-
proceed to form S~ which will accumulate as a land & Evans 1976). On the other hand,
net reaction product only if reaction 5 is retarded thermodynamic calculations indicate that TSR
or inhibited. This may take place if the bacteria should take place between methane and dis-
necessary for reaction 5 are not present, or if the solved sulphate at temperatures lower than
environment is not conducive to growth. This, in 200~ (Barton 1967), and geological examples
turn, depends mainly on a sufficient supply of occur in several sour gas provinces in the world
suitable nutrients (including specific hydrocar- (e.g. Orr 1977). In both cases, reactions 3 to 5 can
bons that can be metabolized). If the environ- be simplified because bitumen would not be
mental conditions are appropriate, reactions 3, 4 formed: all methane would be oxidized to
and 5 proceed simultaneously at or near the HCO 3- (and/or CO2, CO).
aerobic/anaerobic interface, which may be an
oil/water contact or a diffusion zone within
Stable reaction products and by-products
anaerobic sediments overlain by aerobic water.
Should the whole environment become closed H2S evolves as a separate gas phase if the system
and totally anaerobic, reactions 3 and 5 will take does not contain, or has used up, base and
over until the nutrients and reactants are transition metals. During BSR, H2S could be
depleted. generated as long as reactants and nutrients for
The initial sequence of reactions is different the bacteria are available, and as long as the HzS
and possibly variable for TSR of non-gaseous concentration is below the toxic level for bac-
hydrocarbons. This depends mainly on the initial terial metabolism. Hence, for large quantities of
availability of 'catalytic' H2S, which may form a H2S to be formed, the system must be open and
 S u l p h a t e - h y d r o c a r b o n r e d o x reactions 2I

allow for continuous inward diffusion of sulphate If transition or base metals are present,
as well as a sink for hydrogen sulphide. In fact, disseminated or stratiform base metal or Missis-
H2S in most natural environments of BSR is sippi Valley type deposits could be formed.
bonded as metal sulphides, organic compounds, Minerals such as pyrite, galena, and sphalerite
or escapes as gas from the reaction site, and does are precipitated as a result of sulphide generation
not accumulate as reservoir gas (e.g. Orr 1977, during reactions 3 and 5. In these cases, partial
Krouse 1980). However, the possibility of bio- dissolution of the host rock should occur (at least
genic H2S migrating into a shallow trap and initially), because the precipitation of sulphides
forming a sour gas reservoir cannot be ruled out. generally are strongly acid-generating reactions
On the other hand, deep reservoirs with more (several related phenomena are discussed in
than a few per cent H,_S (in reservoir gas) are Anderson 1983, Coleman 1985, Anderson &
suspected to have undergone TSR. Using crude Garven 1987, Morse et al. 1987). In addition,
oil with about 3~ sulphur as representative of BSR/TSR themselves could generate acidity due
sulphur-rich crude, Orr (1977) calculated that no to the release of CO2, perhaps in reaction 3.
more than 2-3 vol-~ H2S (in reservoir gas) can However, the low acidity resulting from this
evolve from thermal cracking of crude oil. Crude carbon dioxide is probably negligible in most
oil may contain more than 3~ sulphur (Sassen, natural environments. On the other hand, much
pers. comm. 1986) and perhaps release more than larger quantities of CO2 evolve from thermal
3~ H2S, but deep reservoirs with 20-80~ HzS cracking of kerogen or crude oil (e.g. Krouse
are invariably due to TSR (Orr 1977). 1983). Hence, dissolution by thermogenic car-
Elemental sulphur accumulates as a net reac- bonic acid in deeply buried hydrocarbon reser-
tion product if reaction 5 is inhibited. This voirs may compound potential dissolution
happens if the system runs out of reactive caused by precipitation of sulphides.
hydrocarbons (BSR/TSR), or if reactive hydro- Subordinate reaction by-products that form in
carbons are not supplied fast enough (TSR). the vicinity of BSR and TSR as a result of
BSR and TSR form many distinctive NSO- released and/or consumed phases are cerrusite,
compounds (discussed in Ho et al. 1974, Burnie barite, fluorite, and gases such as nitrogen and
1979). Some of these compounds are contained in helium (e.g. Barton 1967, Dunsmore 1971,
the solid organic precipitates (bitumens) which Anderson 1983, Siebert 1985).
should be formed where hydrocarbons other than Naturally, one or several of these reaction
methane are utilized. These bitumens must be products and by-products may be absent. The
differentiated from those generated by biodegra- availability and mass proportions of the reac-
dation or thermal maturation on the basis of tants determine which products and by-products
elemental and isotopic criteria (which has been are formed, and in which proportions.
attempted in isolated cases: Macqueen & Powell
1983, Powell & Macqueen 1984, Sassen 1986 and
in press, Machel & Krouse, in prep. b). Temperature ranges and reaction kinetics of BSR
and TSR
Several minerals could precipitate as the result
of BSR/TSR if the respective metal ions are (a) The major reaction pathways are similar for
present or (b) transported to the reaction site, or biological and abiological sulphate-hydrocarbon
(c) if the reaction products of BSR/TSR are redox reactions. However, BSR and TSR appear
transported into an environment containing to be mutually exclusive processes in natural
metal ions. Firstly, the presence of alkali earth environments. One of the best indications for
metals will result in precipitation of carbonates this phenomenon are the natural occurrences and
(mainly calcite and dolomite), either as cement isotopic compositions of H2S, suggesting that
or as replacement of the dissolving sulphates BSR and TSR take place at two particular
(mainly gypsum and anhydrite), as a direct result temperature/thermal maturity levels: at low
of reactions 3 and 5 because of the generated levels of less than about 75-85~ (equivalent to
bicarbonate. Additionally, Davis & Kirkland about 0.2-0.3~ Ro), and at high levels in excess
(1970) suggested that the reaction of polysul- of 100-140~ (> 1.5~ Ro) (Burnie 1979, Krouse
phides with bicarbonate also leads to calcite 1980, Sassen 1985, see Fig. 1; note: these
precipitation, and Schneider & Nielson (1965) temperature-maturation correlations are only
suggested that bicarbonate may form as a result generalized approximations because of the time-
of bacterial oxidation of sulphide to sulphur dependence of thermal maturation).
(reaction 4), with subsequent precipitation of The lower temperature/maturity level coin-
this bicarbonate as calcite. Other carbonates cides with BSR, because sulphate reducing
(ankerite, siderite, witherite, strontianite) are bacteria cannot thrive at temperatures of the
also formed as a result of BSR/TSR where the upper level. Bacteria can survive up to tempera-
respective metals are available. tures at which their cytoplasm does not boil
22 H. G. Machel

T H E F R A M E W O R K OF
HYDROCARBON GENERATION AND DESTRUCTION
THERMAL MATURITY
% VITRINITE REFLECTANCE
0.2
BIODEGRADATION BIOGENIC
OF C R U D E O I L 0.3 METHANE

I j0,4
ONSET OF T H E R M A L - - - - I 0.5
GENERATION OF OIL L
ill
n"
MAIN PHASE OF 0.7
I- MEDIUM-GRAVITY OIL o.8
,, <r
GENERATION (0.70%) I MAIN PHASE OF
THERMAL GENERATION
O0 PRODUCTION LIMIT FOR-1 OF DRY GAS (0.90%)
r LIQUID HYDROCARBONS, I
ONSET OF DILUTION BY 1.5
M. CO2, H2S (1.50%)
PRODUCTION LIMIT F O R - - 1 2.0
,,r WET GAS (2.0%)
.... ONSET OF GAS
IJ.I 3.0 DILUTION BY CO2
m H2S (2.5%)
I-
PRODUCTION LIMIT FOR . . . . . 4.0 . . . . . . . . . PRODUCTION LIMIT FOR
DRY GAS (4.0%) 5.0 DRY GAS (4.0%)

RESERVOIR PRODUCTS FROM RESERVOIR PRODUCTS FROM

HYDROGEN-RICH MARINE KEROGEN LOW-HYDROGEN TERRESTRIAL KEROGEN

F]G. 1. Generalized relationships between thermal maturity and hydrocarbon generation and destruction. Key
vitrinite reflectance values useful in hydrocarbon exploration are shown in parentheses. Biogenic methane and
the gases resulting from oil biodegradation are also included because they are significant reservoir constitu-
ents. Note that H2S is formed mainly in association with marine kerogen at low maturity levels (0.2-0.3~
vitrinite reflectance), and at high maturity levels (greater than about 1.5~ vitrinite reflectance). Reproduced
from Sassen (1985).

(which may be in excess of 100~ if the pressure The upper temperature/thermal maturity level
is high enough), but they become dormant as a ( > 100-140~ > 1.5~ Ro) coincides with TSR
certain temperature is exceeded. No sulphate- which is very slow or inhibited at low
reducing bacteria are known that metabolize at temperatures, even though the reactions may
temperatures in excess of about 92~ and most have a large negative free energy change of
sulphate-reducing bacteria do not metabolize at reaction. The activation energy is also very large
temperatures in excess of about 45~ (at 1 atm: and differs for various hydrocarbons (experi-
Peck 1984, Trudinger et al. 1985, Nazina et al. mentally shown: e.g. Toland 1960). For TSR of
1985, Stetter et al. 1987). Hence, BSR can take an 'average' hydrocarbon, the activation energy
place only in near-surface and shallow subsur- has been estimated to be about 50 kcal (Dhan-
face environments. In these environments, BSR noun & Fyfe 1972). Acordingly, TSR has been
proceeds almost instantaneously on a geological performed in the laboratory only at high tem-
time-scale and does not depend on any catalysts peratures (in excess of 175~ and geologically
(other than enzymes within the cells). This was significant reaction rates were measured only
demonstrated experimentally in the laboratory above about 250~ On the basis of this evidence,
(e.g. Nakai & Jensen 1964), and in the field. For and the apparent paucity of reactive hydrocar-
example, water injection for enhanced petroleum bons, Trudinger et al. (1985) concluded that TSR
recovery in several shallow Russian and Canadi- may not be possible at geologically significant
an oil fields resulted in rapid H2S production. rates in natural diagenetic environments. On the
Within less than 10 years, up to about 10~ H,S other hand, Dunsmore (1971), Powell & Mac-
was generated in previously sweet reservoirs queen (1984), Machel (1985), Machel & Krouse
because of the introduction of bacteria into (in prep. a), and Sassen (in press), as well as a
hitherto sterile reservoirs (Ashirov 1962, Krouse number of other studies (Kartsev et al. 1959,
1980). Gutsalo & Krivosheya 1966, Barton 1967, An-
 Sulphate-hydrocarbon redox reactions 23

dreev et al. 1968, Dunsmore 1971, Reznikov Hanor & Workman 1986), some of which
1971, Orr 1974, 1977, 1982, Drean 1978, Ohmoto may catalyse TSR (and perhaps BSR).
& Rye 1979, Anderson, 1983, Macqueen & (6) The oxidation of H,S is known to be
Powell 1983, Krebs & Macqueen 1984, Powell & catalysed by metal ions in water in the
Macqueen 1984), have provided cumulative sequence Ni 2+ > Co z+ > Mn 2+ > Cu 2+
geological, theoretical, and circumstantial evi- > Fe z+ > Ca 2+ = Mg 2+, and by organic
dence which is considered sufficient proof by the compounds such as phenols, aldehydes,
present author for the occurrence of TSR at aniline, urea, and vanillin (Morse et al.
diagenetic temperatures as low as at least 135- 1987).
140~ (three examples are discussed in the (7) Clay minerals (particularly montmorillon-
section on 'Selected geological examples', be- ite) and silica gel have been shown experi-
low). The discrepancy between this writer's and mentally to catalyse TSR (Rudakova &
Trudinger et al.'s (1985) assessment is based on Velikovskii 1947, Dhannoun & Fyfe 1972).
the role catalysts and kinetic factors play that are (8) H2S from thermal decomposition of hydro-
not present or involved in most laboratory carbon NSO-compounds (reaction 2) drives
experiments. There is no shortage of potential step 4 to the right (also invoked by Orr 1974,
catalysts in natural environments and some have Ohmoto & Rye 1979, Powell & Macqueen
been identified experimentally for TSR. Also, 1984).
the reaction chain in Table 1 contains several (9) Reactions 3 and 4b are driven to the right by
reaction loops that provide for 'autocatalysis', in low pH (represented by H + on the left sides;
the sense that reaction products of one step may experimentally shown by Toland 1960;
increase the reaction rate and degree of comple- references in Trudinger et al. 1985). There-
tion of another step. The main kinetic factors fore, precipitation of carbonates and par-
and catalysts can be summarized as follows (this ticularly sulphides, which are associated
list is certainly incomplete). with a decrease in pH (e.g. Anderson 1983,
Coleman 1985, Siebert 1985) and often
(1) Salts (soaps) of organic acids are known to accompany TSR, would drive reactions 3
catalyse TSR in an unknown manner and 4b to the right. Of course, simultaneous
(Kartsev et al. 1959). Also, the formation of thermal cracking of organic matter during
soaps could lower the pH and thus drive maturation also generates acidity, as do
reaction 3 to the right. other reactions not related to the redox-
(2) The formation of compounds using up scheme of Table 1 (i.e. clay mineral diagen-
reaction products of reversible reactions esis, and other processes: Giles & Marshall
would drive these steps to the right. For 1986). Acids derived from these processes
example, S~ is used up by polysulphide may be generated within the system of
formation, favouring reactions 4a and 4b. BSR/TSR, or invade from elsewhere, and
Similarly, S~ is used up in reaction 5 and by would also drive reactions 3 and 4b to the
bitumen, and by the resin and asphaltene right.
fractions in crude oil ('sulphurization': (10) Reaction 5 is driven to the right at high pH,
Kaplan et al. 1963, Orr 1974, Milner et al. represented by OH- on the left side (experi-
1977, Powell & Macqueen 1984), again mentally shown by Toland 1960).
favouring reactions 4a and 4b. Of course, (ll) Oxidation of crude oil should proceed faster
sulphurization also occurs with sulphur than oxidation of methane because crude oil
compounds of valence states other than is thermodynamically less stable than meth-
zero, but this does not favour any of the ane (Barton 1967).
reactions in Table 1. (12) The rate of reaction 3 depends on the
(3) Formation of certain compounds by one sulphate species (i.e. SO~ z-, HSO4-,
reaction increases the rate and degree of NaSO,~-) and temperature (Kiyosu 1980),
completion of another reaction. For exam- as well as on the type of hydrocarbons (e.g.
ple, formation of elemental sulphur in Toland 1960, Trudinger et al. 1985). The
reaction 4 increases the rate and completion rate of reaction 5 probably also depends on
of reaction 5 which uses up sulphur. the types of hydrocarbons utilized.
(4) Formation of complexes, including polysul-
phides, is known to catalyse redox-reactions Given the presence of at least some of the above
(Stumm & Morgan 1970, Berner 1971, catalysts and kinetic factors, and an intermediate
Morse et al. 1987). pH, the reaction scheme could be triggered
(5) Organic acids are involved in a variety of thermochemically. A minimum temperature of
processes, i.e. complexing of metals, decar- about 135-140~ (Machel 1985, Siebert 1985),
boxylation, and oxidation by metals (e.g. perhaps only about 100~ (Rye & Williams 1981,
z4 H. G. Machel
Macqueen & Powell 1983, Sassen 1986) is during TSR). Naturally, the faster the reaction
suggested for the onset of TSR even under the rate, the greater the potential geothermal
most favourable conditions (as inferred by these anomaly.
authors from fluid inclusion and thermal matura-
tion data from carbonate rocks and Mississippi Selected geological examples
Valley type deposits of Canada, the US Gulf
coast, and Australia). Once the reactions are Because of the limited temperature tolerance of
instigated, several reaction loops provide for most bacteria, BSR has a relatively shallow
'autocatalysis'. The most important loops are depth limit in natural environments. The maxi-
probably: sulphur generated in reactions 4b and mum depth for BSR is probably close to 3000 m.
4c appears as reactant in reaction 5 driving this Scarce living microflora occurs down to depths of
reaction to the right; the removal of sulphur from 3290 m (equivalent to about 85~ and a pressure
reaction 4b drives this reaction to the right; H2S of about 350 bar: Ashirov 1962, Rosanova &
generated in reaction 5 appears as reactant in Khudyakova 1974), but these deep environments
reactions 4b and 4c, again driving these reactions are anoxic, and large-scale BSR is not likely to
to the right (reaction 4a is not an alternative in occur because of the paucity of nutrients for the
this context because natural TSR environments sulphate-reducers (from aerobic biodegradation
are anoxic). Additionally, polysulphides form of hydrocarbons). Hence sulphate reduction, the
similar reaction loops. Therefore, the reaction associated aerobic biodegradation of hydrocar-
rate of this scheme may increase considerably, bons, and aerobic sulphide-sulphur oxidation,
provided the system begins with a surplus of the are generally bracketed between the 'aeration
initial reactants: reactive sulphates and hydro- zone' of exposed rocks (0-200 m depth) and the
carbons. In fact, shortage of these reactants is the bottom of the oxygenated groundwater zone
only factor that may inhibit exponential increase (generally about 600 m) (Andreev et al. 1968).
of the reaction rate. For example, if hydrocar- Accordingly, carbonates, bitumen, and elemen-
bons seep to the reaction site only slowly, this tal sulphur resulting from BSR and partial
could be the rate-limiting step. Similarly, if the reoxidation of sulphide occur generally in near-
sulphates dissolve or are supplied only very surface environments (i.e. in salt dome cap rocks
slowly, this will be the rate-limiting step. of the United States and Europe: Feely & Kulp
1957, Dessau et al. 1962, Pawlowski et al. 1979)
and along/below the oil-water interface in shal-
Heat
low oil reservoirs (i.e. in Russia: Ashirov 1962).
The net reaction (Table 1) is exothermic. The Meteoric water is believed to be the vehicle to
heat released was estimated to be about 30 or transport microbes into contact with subsurface
10 kcal tool-1 calcium sulphate by Feely & Kulp hydrocarbon pools, i.e. via faults, fractures, and
(1957) and Dhannoun & Fyfe (1972), respective- other conduits (e.g. Milner et al. 1977), and
ly. It is not clear, however, whether these authors oxygenated groundwater probably is the cause of
took into account the heat released during H~S oxidation to S~ in most shallow sulphur
reactions 4 and 5, which may also be exothermic. deposits. Where bitumen is absent, and the
This is suggested by observations that the carbon isotope ratios of precipitated carbonates
bacterial oxidation of sulphur to sulphate (in are lower than those of oil, methane is indicated
quarry dumps of rocks containing elemental as the main reactant hydrocarbon (e.g. in the
sulphur) is accompanied by a marked production limestone buttes of west Texas: Kirkland &
of heat (Pawlowski et al. 1979). Hence, 10 kcal Evans 1976). Perhaps both crude oil and meth-
mo1-1 calcium sulphate is a conservative esti- ane may be utilized in some cases.
mate of the minimum heat released during TSR. The natural occurrence of TSR is probably
This heat may be sufficient to generate a restricted to environments that have a minimum
geothermal hot-spot at the reaction site, if the temperature of about 100-140~ For mass
reactions proceed fast enough. This was invoked balance reasons, any deeply buried sour gas
by several authors to explain present or past reservoir with more than a few per cent H2S is
positive geothermal anomalies (e.g. Bush 1970, suspected to be formed by TSR. On the other
for Pine Point; Dunsmore 1971, for Keg River hand, uplifted rocks that underwent sulphate-
and Pine Point). These interpretations are hydrocarbon redox reactions may have under-
supported by calculations (Dhannoun & Fyfe gone BSR or TSR. Perhaps the most convincing
1972) indicating that heat production in a 1 km examples for these two possibilities have been
thick rock column containing 30% calcium described from western Canada.
sulphate would be of the same order of magni- Examples for TSR at deep burial are the
tude as normal geothermal heat flow, if TSR subsurface Keg River and Nisku Formations.
occurred in the order of 1 Ma (taking 10 kcal Dunsmore (1971) showed for the Middle Devon-
mo1-1 calcium sulphate as the heat released ian Keg River Formation that (a) the present
 Sulphate-hydrocarbon redox reactions 25
reservoir temperatures of about 70-100~ corre- 60~ and (b) insoluble bitumens are enriched in
sponding to depths of about 1500-2300 m, isotopically heavy sulphur relative to their
represent abnormally high temperatures (these precursor bitumen. Considering all the data,
temperatures and depths also represent the these authors interpreted TSR to have taken
minima since uplifting and erosion began in place at a temperature of about 100~ (or
the early Tertiary); (b) these abnormally high somewhat higher, should the fluid inclusion
temperatures are not due to elevated heat flow temperatures be too low; Macqueen, pers.
from below; (c) most reaction products of TSR comm. 1987). The heat anomaly in the mineral-
are present; and (d) the composition of present ized zone was not interpreted to be the result of
Keg River oil field brines and hydrocarbons, as TSR. Rather, heat was thought to be imported by
well as sulphur isotope data, are consistent with the basinal fluids carrying the metals.
TSR. On the basis of these results, Dunsmore
(1971, p. 55) suggested that, at present, 'heat may
be originating at the site of petroleum occur- Conclusions
rence'. Dunsmore also concluded that the present
geothermal anomalies may be inherited from The preceding discussion sheds some light on the
TSR that has gone to completion at greater environmental conditions and kinetics of sul-
depths. phate-hydrocarbon redox reactions. It is con-
Regarding the Upper Devonian Nisku For- cluded that both bacterial and thermochemical
mation, Machel (1985) and Machel & Krouse (in sulphate reduction are common phenomena in
prep. a) demonstrated that sour gas (up to 30~o diagenetic environments, and that these pro-
H2S), elemental sulphur at a former oil/water cesses occur at mutually exclusive temperature/
contact, and saddle dolomites with organic thermal maturity levels. Bacterially mediated
carbon and isotopically depleted oxygen, must redox reactions generally proceed much faster
have resulted from TSR. Importantly, these than thermochemical redox reactions. Given the
reservoirs are presently at depths in excess of presence of certain catalysts, however, thermo-
about 3400 m, which are their shallowest depths chemical sulphate reduction may proceed at
since oil emplacement. Isotopic, fluid inclusion, geologically significant rates and may even
and thermal maturation data indicate that (1) the generate a geothermal hot spot.
maximum burial temperature was between 135 Future work should focus on clarification of
and 170~ (2) these reservoirs have been the specific roles of various catalysts for thermo-
hydrodynamically isolated since oil emplace- chemical sulphate reduction. Also, elemental
ment at similar depths in the late Cretaceous; and isotopic criteria should be developed that
and (3) no meteoric water intruded since then. discriminate products of bacterial and thermo-
Hence, BSR is not an alternative because these chemical sulphate reduction from one another
reservoirs were sterile at the time of sulphate and from other processes (as discussed in Machel
reduction. & Krouse, in prep. b).
An example for BSR and TSR in an uplifted
rock sequence is the Devonian Pine Point
ACKNOWLEDGMENTS:H. R. Krouse introduced me to
Mississippi Valley Type district in northwestern the subject of sulphate-hydrocarbon redox reactions.
C a n a d a . Macqueen & Powell (1983) and Powell Several enthusiastic diurnal and nocturnal discussions
& Macqueen (1984) presented isotopic and with him paved the way for this paper. A. Aplin, J. M.
compositional data on reservoir bitumens, inte- Larkin, R. W. Macqueen, R. Raiswell, R. Sassen and
grated into thermal maturation, petrographic, Z. Sofer critically read earlier versions of the manu-
isotopic, and fluid inclusion data of the mineralic script, and their numerous suggestions greatly im-
precipitates. Among other details, they could proved the final version. G. M. Anderson and J. W.
demonstrate that (a) the maximum temperature Morse kindly provided unpublished manuscripts. The
permission for reproduction of Fig. 1 by R. Sassen is
of the mineralized Pine Point district was about
greatly appreciated. Last but not least, C. H. Moore
100~ (perhaps 10-30~ higher: Macqueen, was a patient and constant source of encouragement
pers. comm. 1987), which represents a thermal without which this study would not have been possible.
palaeo-anomaly because the maximum tempera- This research was supported by the Basin Research
ture of the surrounding rocks was only about Institute of Louisiana State University.

References
ANDERSON,G. M. 1983. Some geochemical aspects of versity of Missouri-Rolla, Rolla, Missouri, pp.
sulfide precipitation in carbonate rocks. In: 61-76.
KISVARSANYI,G., GRANT, S. K., PRATr, W. P. & --& GARVEN, G. 1987. Sulfate-sulfide-carbonate
KOENIG, J. W. (eds). International Conference associations in Mississippi-Valley-type lead-zinc
on Mississippi Valley Type lead-zinc deposits. Uni- deposits. Economic Geology, in press.
26 H. G. Machel
ANDREEV, P. F., BOGOMOLOV, A. I., DOBRYANSKII, DAVIS, J. B. & KIRKLAND,D. W. 1970. Native sulfur
A. F. & KARTSEV, A. A. 1968. Transformation deposition in the Castile formation, Culberson
of Petroleum in Nature. Pergamon Press, Oxford. County, Texas. Economic Geology, 65, 107-21.
ASrXIROV, K. B. 1962. Life activity of formational - - , STANLEY,J. P. & CUSTARD,H. C. 1970. Evidence
microflora as an index of geologic environment against oxidation of hydrogen sulfide by sulfate
and processes obtaining in petroliferous forma- ions to produce elemental sulfur in salt domes.
tions. In: KUZNETSOV,S. I. (ed.). Geologic Activity American Association of Petroleum Geologists Bulle-
of Microorganisms. Transactions of the Institute of tin, 54, 2444-7.
Microbiology, IX, 84-91. -- & YARBROUGH, H. F. 1966. Anaerobic oxidation
BAILEY,N. J. L., JOBSON,A. M. & ROGERS, M. A. 1973. of hydrocarbons by Desulfovibrio desulfuricans.
Bacterial degradation of crude oil: comparison of Chemical Geology, 1, 137~4.
field and experimental data. Chemical Geology, 11, DESSAU,G., JENSEN, M. L. & NAKAI, N. 1962. Geology
203-21. and isotopic studies of Sicilian sulfur deposits.
BARNES, R. O. & GOLDBERG, E. D. 1976. Methane Economic Geology, 57, 410-38.
production and consumption in anoxic marine DHANNOUN,H. Y. & FYFE, W. S. 1972. Reaction rates
sediments. Geology, 4, 297-300. of hydrocarbons with anhydrite. Progress in
BARTON, P. B. 1967. Possible role of organic matter in Experimental Petrology, 2, 69-71.
the precipitation of the Mississippi Valley ores. In: DOUGLAS, A. G. & MAIR, B. J. 1965. Sulfur: role in
BROWN, J. S. (ed.). Genesis of Stratiform Lead- genesis of petroleum. Science, 147, 499-501.
Zinc-Barite-Fluorite Deposits. Economic Geology, DREAN, T. A. 1978. Reduction of sulfate by methane,
Monograph 3, 371-8. The Economic Publishing xylene, and iron at temperatures of 175 to 350~
Company, Lancaster, PA. MS Thesis, Pennsylvania State University.
BEIJERINCK, M. W. 1895. f3ber Spirillum desulfuricans DUNSMORE, H. E. 1971. Diagenetic model for middle
als Ursache von Sulfat-reduktion. Centralblattff~r Devonian Keg River Formation, Rainbow area,
Bakteriologie, Parasitenkunde, lnfektionskrankhei- northwestern Alberta. Unpublished BSc Thesis,
ten und Hygiene, Abteilung L Originale, 1, 1-9, 49- University of Calgary.
59, 104-14. FELLY, H. W. & KULP, J. L. 1957. Origin of gulf coast
BERNER, R. A. 1971. Principles of Chemical Sedimento- salt-dome sulphur deposits. American Association
logy. McGraw-Hill, New York. of Petroleum Geologists Bulletin, 41, 1802-53.
--1980. Early Diagenesis. Princeton University GARDNER, L. R. 1973. Chemical models for sulfate
Press, New Jersey. reduction in closed anaerobic marine environ-
, SCOTT, M. R. & THOMLINSON,C. 1970. Carbon- ments. Geochimica et Cosmochimica Acta, 37, 53-
ate alkalinity in the pore waters of anoxic marine 68.
sediments. Limnology and Oceanography, 15, 544- GILES, M. R. & MARSHALL,J. D. 1986. Constraints on
9. the development of secondary porosity in the
BIRNBAUM, S. J. & WIREMAN, J. N. 1984. Bacterial subsurface: re-evaluation of processes. Marine and
sulfate reduction and pH; implications for diagen- Petroleum Geology, 3, 243-55.
esis. Chemical Geology, 43, 143-9. GUTSALO, L. K. & KRIVOSHEYA,V. A. 1966. Impor-
BOLZE, C. E., MALONE, P. G. & SMITH, M. J. 1974. tance of sulfates from subsurface waters in
Microbial mobilization of barite. Chemical Geo- prospecting for oil and gas in Dnepr-Donets
logy, 13, 141-3. trough. International Geological Reviews, 9, 1144-9.
BERNIE, S. W. 1979. A sulphur and carbon isotope HANOR, J. S. & WORKMAN,A. L. 1986. Distribution of
study of hydrocarbons from the Devonian of dissolved volatile fatty acids in some Louisiana oil
Alberta, Canada. PhD Thesis, University of field brines. Applied Geochemistry, l, 37-46.
Alberta, Edmonton. Ho, T. Y., ROGERS, M. A., DRUSHEL,n. V. & KROONS,
BUSH, P. R. 1970. Chloride brines from sabkha C. B 1974. Evolution of sulphur compounds in
sediments and their possible role in ore formation. crude oils. American Association of Petroleum
Institute of Mining and Metallurgy Transactions Geologists Bulletin, 58, 2338-48.
Section B, 79, 137-44. HUSAIN, S. A. 1967. Sulphur isotope exchange reac-
CHAMP, P. R., GULENS, J. & JACUSON, R. E. 1979. tions. PhD Dissertation, University of Alberta,
Oxidation-reduction sequences in ground water Edmonton, Alberta.
flow systems. Canadian Journal of Earth Science, --& KROUSE, H. R. 1978. Sulphur isotope effects
16, 12-23. during the reaction of sulphate with hydrogen
CHEN, K. Y. & MORRIS, J. C. 1972. Kinetics of sulphide. In: ROBINSON, B. W. (ed.). Stable
oxidation of aqueous sulfide by 02. Environmental Isotopes in the Earth Sciences. Department of
Science and Technology, 6, 529-37. Scientific and Industrial Research Bulletin, 220,
COLEMAN, M. L. 1985. Geochemistry of diagenetic 207-10.
non-silicate minerals: kinetic considerations. IVANOV, M. V. 1968. Microbiological processes in the
Philosophical Transactions of the Royal Society of genesis of native sulfur deposits. Institute for
London, A, 315, 39-56. Physical Science and Technology, Maryland, No.
CURTIS, C. D. 1977. Sedimentary geochemistry: envi- 1850.
ronments and processes dominated by involve- IWATSUKA, H. & MORI, T. 1960. Studies on the
ment of an aqueous phase. Philosophical Trans- metabolism of a sulfur-oxidizing bacterium. I.
actions of the Royal Society of London, 286, Oxidation of sulfur. Plant Cell Physiology, 1, 163-
353-72. 72.
 Sulphate-hydrocarbon redox reactions 27

JEFFRIES, M. O. & KROUSE, H. R., 1984. Isotope MCCREADY, R. G. L. & KROUSE, H. R. 1980. Sulfur
geochemistry of stratified water bodies on north- isotope fractionation by Desulfovibrio vulgaris
ern Ellesmere Island, Canadian arctic. Zentralin- during metabolism of BaSO,. Geomicrobiology
stitut far Isotopen- und Strahlenjbrschung Mittei- Journal, 2, 55-62.
lungen, Leipzig, 84, 159-69. MIGDISOV, A. A., CHERKOVSKIY,S. L. & GRINENKO,
JOBSON, A. M., COOK, F. D. & WESTLAKE, D. W. S. V. A. 1974. The effects of formation conditions on
1979. Interaction of aerobic and anaerobic bacte- the sulfur isotopes of aquatic sediments. Transla-
ria in petroleum biodegradation. Chemical Geo- tions from Geokhimiya No. 10, 1482-502.
logy, 24, 355-65. MILNER, C. W. D., ROGERS, M. A. & EVANS, M. A.
JORGENSEN, B. & REVSBECH, N. P. 1983. Colourless 1977. Petroleum transformation in reservoirs.
sulfur bacteria Beggiatoa sp.. and Thiovulum spp. Journal of Geochemical Exploration, 7, 101-53.
in 02 and H2S microgradients. Applied and MORSE, J. W., MILLERO, F. J., CORNWELL, J. C. &
Environmental Microbiology, 45, 1261-70. RICKARD, D. T. 1987. The chemistry of the
KAPLAN, I. R., EMERY, K. O. & RITTENBERG, S. C. hydrogen sulfide and iron sulfide systems in
1963. The distribution and isotopic abundance of natural waters. Earth Science Reviews, 24, 1-42.
sulphur in recent marine sediments off southern NAKAI, N. & JENSEN, M. L. 1964. The kinetic isotope
California. Geochimica et Cosmochimica Acta, 27, effect in the bacterial reduction and oxidation of
297-331. sulfur. Geochimica et Cosmochimica Acta, 28, 1893-
KARTSEV,A. A., TABASARANSKII,Z. A., SUBBOTA,M. I. 912.
& MOGILEVSKII,G. A. 1959. Geochemical Methods NAZINA, T. N., ROZANOVA, E. P. & KNZETSOV, S. I.
of Prospecting and Exploration for Petroleum and 1985. Microbial oil transformation processes
Natural Gas. University of California Press, accompanied by methane and hydrogen-sulfide
Berkeley. formation. Geomicrobiology Journal, 4, 103-30.
KIRKLAND, D. W. & EVANS, R. 1976. Origin of OHMOTO, H. & RYE, R. O. 1979. Isotopes of sulfur and
limestone buttes, gypsum plain, Culberson Coun- carbon. In: BARNES, H. L. (ed.). Geochemistry of
ty, Texas. American Association of Petroleum Hydrothermal Ore Deposits, pp. 509-65. Wiley,
Geologists Bulletin, 60, 2005-18. New York.
KIYOSU, Y. 1980. Chemical reduction and sulfur- ORR, W. L. 1974. Changes in sulfur content and
isotope effects of sulfate by organic matter under isotopic ratios of sulfur during petroleum matura-
hydrothermal conditions. Chemical Geology, 30, tion-study of Big Horn Basin Paleozoic oils.
47-56. American Association of Petroleum Geologists Bulle-
KREBS, W. & MACQUEEN, R. 1984. Sequence of tin, 58, 2295-318.
diagenetic and mineralization events, Pine Point -- 1977. Geologic and geochemical controls on the
lead-zinc property, Northwest Territories, Can- distribution of hydrogen sulfide in natural gas. In:
ada. Bulletin Canadian Petroleum Geology, 32, 434- CAMPOS, R. 8s GONI, J. (ed.). Advances in Organic
64. Geochemistry, pp. 571-97. Enadisma, Madrid,
KROUSE, H. R. 1977. Sulfur isotope studies and their Spain.
role in petroleum exploration. Journal of Geo- - - 1 9 8 2 . Rate and mechanism of non-microbial
chemical Exploration, 7, 189-211. sulfate reduction. Geological Society of America,
- - 1980. Stable isotope geochemistry of non-hydro- Annual Meeting, Abstracts with Programs, 14,
carbon constituents of natural gas. Proceedings 580.
lOth Worm Petroleum Congress, pp. 85-92, Bucha- PAULL, C. K., HECKER, B., COMMEAU, R., FREEMAN-
rest, Romania. LYNDE, R. P., NEUMANN, C., CORSO, W. P.,
1983. Stable isotope research in support of more GOLUmC, S., HOOK, J. E. & CURRAY, J. 1984.
effective utilization of gas fields in Alberta. Biological communities at the Florida Escarpment
Alberta - - Canada Energy Resource Research Fund resemble hydrothermal vent taxa. Science, 226,
Agreement U-30. 965-7.
LARKIN, J. M. 1981. Isolation of Thiothrix in pure PAWLOWSKI, S., PAWLOWSKA,K. & KUBICA, B. 1979.
culture and observation of a filamentous epiphyte Geology and genesis of the Polish sulfur deposits.
on Thiothrix. Current Microbiology, 4, 341-67. Economic Geology, 74, 475-83.
MACHEL, H. G. 1985. Facies and diagenesis of the PECK, H. D. 1984. Physiological diversity of the
upper Devonian Nisku Formation in the subsur- sulfate-reducing bacteria. In: STROHL, W. R.
face of central Alberta. Unpublished PhD Thesis, TUOVINEN, O. H. (eds). Microbial Chemoauto-
McGill University, Montreal. trophy, pp. 231-335. Ohio State University Press,
- - & KROUSE, H. R. in prep. a. Thermochemical Columbus.
sulphate reduction in deeply buried Upper Devon- PHILIPPI, G. T. 1977. On the depth, time and
ian Nisku carbonates of the Alberta subsurface. mechanism of origin of the heavy to medium-
- - & --in prep. b. Products and distinguishing gravity naphthenic crude oils. Geochimica et
criteria of bacterial and thermochemical sulphate- Cosmochimica Acta, 41, 33-52.
hydrocarbon redox-reactions in diagenetic POSTGATE, J. R. 1984. The Sulphate-reducing Bacteria.
environments. Cambridge University Press.
MACQUEEN, R. W. & POWELL, T. G. 1983. Organic POWELL, T. G. & MACQUEEN, R. W. 1984. Precipita-
geochemistry of the Pine Point lead-zinc ore field tion of sulfide ores and organic matter: sulfate
and region, Northwest Territories, Canada. Eco- reactions at Pine Point, Canada. Science, 224, 63-
nomic Geology, 78, 1-25. 6.
28 H. G. Machel

PRYOR, W. A. 1962. Mechanisms of Sulfur Reactions. SCHNEIDER, A. & NIELSON, H. 1965. Zur Genese des
McGraw-Hill, New York. elementaren Schwefels im Gips von Weenzen
REZNIKOV, A. N. 1971. The conversion of petroleum (Hils). Beitriige zur Mineralogie und Petrographie,
gases in a high-temperature environment. Geolo- 11, 705-18.
giya Nefti i Gaza, 4, 44-8. SIEBERT, R. M. 1985. The origin of hydrogen sulfide,
RICKARD, D. T. 1969. The chemistry of iron sulfide elemental sulfur, carbon dioxide, and nitrogen in
formation at low temperatures. Stockholm Contri- reservoirs. Timing of siliciclastic diagenesis: rela-
butions to Geology, 20, 67-95. tionship to hydrocarbon migration. Sixth Annual
ROSANOVA, E. P. & KHUDYAKOVA,A. I. 1974. A new Research Conference, Gulf Coast Section, Society
non-sporeforming thermophilic sulfate-reducing of Economic Paleontologists and Mineralogists
organism. Desulfovibrio thermophilus nov. sp. Mi- Foundation, pp. 30-31.
crobiology, 43, 908-12. SIEBURTH, J. McN. 1979. Sea Microbes. Oxford
RUDAKOVA, E. F. & VELIKOVSKn,A. S. 1947. Condi- University Press, New York.
tions for the formation of sulphur compounds and STEYrER, K. O. & GAAG, G. 1983. Reduction of
sulphur in crude oils. Neftyanoe Khozyaistro, 25, molecular sulphur by methanogenic bacteria.
49-54. Nature, 305, 309-11.
RYE, D. M. & WILLIAMS,N. 1981. Studies of the base STETTER, K. O., LAUERER,G., THOMM, M. & MEUNER,
metal sulfide deposits at McArthur River, North- A. 1987. Isolation of extremely thermophilic
ern Territory, Australia: III. The stable isotope sulfate reducers: evidence for a novel branch of
geochemistry of the H.Y.C. Ridge, and Cooley archaebacteria. Science, 236, 822-4.
deposits. Economic Geology, 76, 1-26. STUMM, W. • MORGAN, J. J. 1970. Aquatic Chemistry.
SASSEN, R. 1980. Biodegradation of crude oil and Wiley-Interscience, New York.
mineral deposition in a shallow Gulf Coast salt TOLAND, W. G. 1960. Oxidation of organic compounds
dome. Organic Geochemistry, 2, 153-66. with aqueous sulfate. Journal of the American
-- 1985. Basic geochemical strategies. Oil and Gas Chemical Society, 82 1911-6.
Journal 1985. TRUDINGER, P. A., CHAMBERS, L. A. & SMITH, J. W.
-- 1986. Crude oil destruction and bitumen precipi- 1985. Low-temperature sulphate reduction: biolo-
tation in deep carbonate reservoirs of the Smack- gical versus abiological. Canadian Journal of Earth
over Formation. Third Annual Meeting, Society for Science, 22, 1910-8.
Organic Petrology, Abstract and Program, pp. 24-6. VALITOV, N. B. 1974. Elemental sulphur as a factor in
Lexington, Kentucky. the generation of hydrogen sulphide in deep-lying
- - in press. Geochemical and carbon isotope studies carbonate reservoir rocks. Doklandy Akademii
of crude oil destruction, bitumen precipitation, Nauk SSSR, 219, 206--8.
and sulfate reduction in the deep Smackover WARD, D. M. & BROCK, T. D. 1978. Anaerobic
Formation. Organic Geochemistry. metabolism of hexadecane in sediments. Geomi-
crobiology Journal, 1, 1-9.

H. G. MACHEL, Basin Research Institute, Louisiana State University, Baton Rouge,

Louisiana 70803-4101, USA. Present address: Department of Geology, University of
Alberta, Edmonton, Alberta T6G 2E3, Canada.
Porosity reduction, microfabric and resultant lithification in UK
uncemented sands
S. N. Palmer & M. E. Barton

SUMMARY: Studies of the extent of diagenetic change in matrix-free, uncemented,

quartzose sands ranging in age from the Jurassic to the Recent in the UK have been carried
out as part of a geotechnical research programme. All the sands studied are thought to have
experienced only a relatively small depth of burial and the extent of diagenetic change is
consequently small. Previous studies of the in situ fabric of sands in this category have been
limited owing to the sampling difficulties created by their very friable nature. Careful
sampling, however, has succeeded in obtaining undisturbed material and has facilitated
studies of the porosity, microfabric and degree of lithification. Distinctive changes,
progressive with age, include reduction of porosity, an increase in the numbers and
complexity of grain contacts and an increasing degree of lithification. The cause of these
diagenetic changes are discussed and it is concluded that the evidence strongly favours
pressure solution of the detrital quartz grains as the dominant process.

This is a study of very clean, virtually matrix- the in situ fabric and allowing laboratory impreg-
free, uncemented, mature quartzose sands of nation (Palmer & Barton 1986). The Quaternary
ages ranging from Recent to Jurassic in the UK. sands are impregnated in situ. Eight of these
The primary motivation for the research is sands (Table 1) have a similar median grain size,
geotechnical. Surprisingly, despite other areas of similar sorting (Fig. 1) and quartz dominated
progress in soil and rock mechanics, very little mineralogy as shown in Table 2. The particle
geotechnical work has been done on the transi- shapes and roundness of the various sands show
tion from loosely compacted, unlithified sands to some variation, but all eight sands have similar
compact, indurated sandstones. Dusseault & minimum remoulded (see below) porosities. It is
Morgenstern (1979) coined the phrase 'locked important to note that all the sands have minor
sand' to denote the intermediate state where the matrix (not greater than 2.0~ in any case) and
original depositional porosity is reduced, the negligible cement content (not greater than
grains are inter-locked (or 'locked') and the sand 0.5~o). It can be considered therefore that these
has acquired a small but measurable degree of sands have sufficient similarities in the sedimen-
lithification without true cementation. The auth- tary characteristics influencing depositional and
ors have investigated the extent to which these diagenetic fabrics (Meade 1966, Berner 1971,
characteristics are present in U K sands (Barton Wolf & Chilingarian 1976, Chilingarian 1983) to
et al. 1986b, Barton et al. in prep.) The study has permit comparison between them in respect of
included observations on the porosity, micro- their porosity reduction and microfabric.
fabric and lithification and has revealed the The uncemented nature of these sands, par-
extent of diagenetic changes in these uncement- ticularly those of Mesozoic age, is distinctive
ed sands. Because of the friable nature of such since there are other sand deposits of equivalent
sediments, studies of the in situ fabric of weakly burial histories which are considerably more
lithified sands have tended to be neglected and it lithified due to the development of an authigenic
is therefore considered that the results are of mineral phase. Indeed, as noted in Table 1, the
significant sedimentological interest. petrographic features discussed here are not
necessarily present throughout the complete
stratigraphic thickness of the strata detailed. The
Sands studied Mesozoic strata, for example, contain horizons
possessing the textural features described in this
Samples of matrix-free, uncemented, quartzose paper, but also frequently contain zones of
sands have been obtained from numerous quar- cemented and/or matrix-rich (>5~o in this
ries and natural exposures. The Mesozoic and context) sands. The reason for the lack of
Tertiary sands, although readily disaggregated interstitial cement is unknown and requires
by manipulation, with care, can be block- further investigation. There is no evidence of
sampled (Barton et al. 1986a) thereby preserving secondary porosity or cement dissolution. The

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences,

Geological Society Special Publication No. 36, pp. 29-40. 29
30 S. N. Palmer & M. E. Barton
TABLE 1.

Sands Age Estimated Approx. Location

studied maximum thickness
Stratigraphic Absolute depth of of sand Site NGR
Ma (bp) burial unit*
(m) (m)

Holkham Beach,
Modern-day Holocene 0.0 0.2 >2 Wells-next-the-Sea, TF 892 455
beach sand Norfolk

Eastern Bavents,
Norwich Crag sand Pleistocene c. 1.65 20 c. 3 Southwold, TM 518 780
Suffolk
Barton sand, Late Becton Bunny,
Barton Beds (Bed K), Eocene c. 42 169 c. 5 Barton-on-Sea, SZ 425 092
Barton Formation Essex

Thanet sand, Late Linford Pit,

c. 57 295 c. 30 Thurrock, TQ 667 799
Thanet Formation Palaeocene Essex

Nine Acres Pit,

Woburn Sands Albian/Aptian c. 100 302 c. 60 Leighton Buzzard, SP 939 277
L. Cretaceous Bedfordshire

Folkestone sand, Albian West Heath Commons,

Foikestone Beds L. Cretaceous c. 100 474 c. 78 West Harting, SU 785 227
West Sussex
South Cave Pit,
Kellaways Sand Callovian,
M. Jurassic c. 139 700 c. 10 South Cave, SE 920 330
Kellaway_ Beds Humberside
Wittering Grange Pit,
Grantham sand, Aalenian,
c. 170 780 c. 7 Wittering, TL 048 101
Grantham Formation M. Jurassic Cambridgeshire

* The petrographic features detailed for each sand in this paper do not necessarily occur throughout the sand's total thickness.

(2) 5 4 3 2
Coarse Silt Very Fine Sand Fine Sand Medium Sand
100,

m
t /f ,///~
/ .//,'/
!

/,'" / . W /
0

/
/ ./," ,,A 3"
/.-" ../' ./i"
O3

0
0.032 0.063 0.125 0.250 0.500
Particle Size (ram)

FIG. 1. Particle size distribution o f the sands, illustrating their very similar grain sizes a n d sorting.
 Porosity reduction & UK uncemented sands 31
TABLE 2.
,...-., Petrographic Mineralogical
g components (%) composition (%)

Sands studied X ,~ ~ ~ O

Modern-day beach sand 0.16 0.58 99.2 0.8 0.0 47.2 38.8 97 3 0 0 T
Norwich Crag sand 0.19 0.63 97.7 1.8 0.2 43.1 38.5 93 5 1 T 1
Barton sand 0.15 0.54 98.0 1.7 0.3 35.6 39.0 96 3 1 T T
Thanet sand 0.10 0.47 97.0 1.9 0.5 35.1 40.7 93 3 2 I 1
Woburn sand 0.14 0.55 98.4 1.5 0.1 35.0 38.7 94 3 1 1 1
Folkestone sand 0.18 0.56 98.3 1.2 0.5 34.5 37.9 96 2 T 2 T
Kellaways sand 0.12 0.51 97.3 1.7 0.4 34.1 39.6 95 3 1 T 1
Grantham sand 0.14 0.57 98.6 1.1 0.3 33.6 38.6 94 2 2 1 1

* Detrital grains of diameter greater than 20/am.

t Detrital grains of diameter less than 20 Iam.
$ Authentic mineral development.

porosity and microfabric are therefore consi- remain speculative or unknown. It is considered
dered to be primary features. necessary, however, in such a study as this to
Ideally, it would be most appropriate to make make some estimate of the burial depth for each
comparisons between the sands' porosities and deposit to serve as a rough guide for comparative
microfabrics in terms of factors such as depth of purposes.
burial, maximum past temperature and other With the exception of the modern beach
conditions likely to influence their diagenesis. material, the sands are shallow water deposits
Unfortunately, the majority of these factors which, in the majority of cases, have been
preserved as the deposits of stable, shelf areas
(Whittaker 1985) and, while the Tertiary depos-
0-- Be
its are associated with basins, the samples
studied are either from the upper stratigraphic
C B levels or from the marginal areas and are
similarly unlikely to have experienced any great
depth of burial. Estimates of the maximum
burial depths are given in Table 1. These figures
are maximum thicknesses based on the lithostra-
...1:
tigraphy presented in the relevant Geological
a Society Report volumes (Mitchell et al. 1973,
Curry et al. 1978, Rawson et al. 1978, Cope et al.
1980). No attempt has been made to allow for
LU
K
possible thicknesses of sediments eroded prior to
subsequent periods of deposition and hence the
figures must be treated as speculative estimates
of the regional burial depths for each sand.
Since the chronological age of the sands is well
1000 J l , , I ' ' ' ' l
documented (Geological Society Special Report
0 200 volumes cited above and Anderton et al. 1979), it
Ma (b.p.)
is considered a more satisfactory parameter for
detailed comparison between the sands. The
FIG. 2. Estimate of the maximum depth of burial
(based on regional stratigraphy) versus chronological chronological age does, of course, represent a
age. The plot suggests that there is an approximately time-scale and, in terms of slow diagenetic
linear relationship between maximum burial depth processes requiring extended time periods, such
and age for these deposits. comparisons will be germane. It is also possible,
32 S. N . P a l m e r & M . E. B a r t o n

O-- 9C m ,Beach

iB
o..

100-- W
"F
RANGE OF MIN. REMOULDED
POROSITIES

G-!
I
I
200
I i [ I iI I I I I I
30 40 50

Porosity %
FIG. 3. Mean porosities of sands versus chronological age. The curve shows an initially rapid decrease in
porosity with time, followed by a continuing, slow reduction over a long time period. The Folkestone (F) and
Woburn (W) sands both have an approximate age of 100 Ma and have been separated purely for clarity.

as suggested by Fig. 2, that the chronological age above. Thus, as seen in Fig. 3, apart from the
of the sediments may represent a scale of modern Beach and the Pleistocene Sand, the
generally increasing depth of burial. existing in situ porosities are considerably re-
duced (by about 12-13% on average).
Comparisons of in situ porosities with re-
Porosity compacted (or 'remoulded') porosities are com-
mon practice in soil engineering (Kolbuszewski
The natural intact or in situ porosities of the 1948). Various techniques can be used to obtain a
sands, as measured by standard petrographic range of porosities from a maximum possible at
and soil engineering techniques, are given in one end of the scale to a minimum (without grain
Table 2. They are plotted versus age in Fig. 3 fracturing or crushing) at the other. In this study
which shows a rapid decrease in porosity from the minimum porosity was achieved using a
the recent beach sand (with a porosity equal to device applying a lateral vibration during sedi-
the deposited value) to the Mesozoic sands, mentation (Barton & Brookes, in press); giving
where significant porosity reduction has oc- values in the range from 38 to 41%. The existing
curred. With sands such as these (well sorted, in situ porosities, as shown in Fig. 3, are
quartz dominated, matrix-free) the in situ porosi- significantly less than these values (by about 4-
ties can usefully be compared with (i) the likely 5% on average).
original deposited porosity and (ii) the minimum Associated with these reductions in porosity
porosity obtained in the laboratory by re- are changes in the microfabric and the degree of
compacting the disaggregated grains (producing lithification.
a remoulded, dense sand deposit).
Original depositional porosities of sands are
generally in the range of 40-50% and more Microfabric
specifically 46-50% for fine sands such as those
of this study (Pryor 1973, Friedman & Sanders Standard petrographic techniques show all the
1978, North 1985). There are apparently no sands to have remarkably similar microfabrics,
records of denser packing being achieved in i.e. high intergranular porosities, grain-support-
sands by the normal, natural processes of ing, matrix-free, uncemented micro-textures.
sedimentation and thus no reason to believe that Similarly to the reduction in porosity, the
the depositional porosities of the sands detailed numbers of grain contacts per grain (Taylor
here were any different to the figures quoted 1950) increase with age, as shown in Fig. 4.
 Porosity reduction in UK uncemented sands 33
0-- Concomitantly, a change in the type of contacts
occurs, principally as a reduction in the numbers
of tangential contacts with an increase in straight
and concavo/convex contacts (Fig. 5). The
microfabrics are illustrated in the photomicro-
T
graphs (Fig. 6) and SEM photographs (Fig. 7). A
small number of sutured contacts are seen in the
Cretaceous and Jurassic sands and although a
100-- F few of these are clearly inherited (in polycrystal-
line grains) most appear to be primary diagenetic
features. A search has been made, using cathodo-
luminescence petrography, SEM and X R D tech-
niques, for evidence of cementation but none has
been found by the authors in any of these sands.
!
/ Confirmation of a lack of quartz overgrowths in
2OO the Grantham, Folkestone and Barton Sands has
I I r 1 been made by GAPS Geological Services. Grain
4 3 2 1 0
N~- of Contacts per Grain contact areas and grain surface textures seen
under the SEM in the older sands (Fig. 6b) show
FIG. 4. Number of grain contacts per grain (contact features strongly suggestive of pressure solution.
index) versus age, illustrating an increase in the state

100
of packing with time. Both the reduced porosities and the types of

Tangential
Long Strt.
Short St(t.
Long C - C
o,<,~ Short C - C
~ ............... Sutured

S SS "J 3Z

_ ..jj.<-/
ss . / T
Y Z
_ ~s ._.._._..__.--------~ ~ ' ~ " ~ T ~ s
IS
.......SC

__ I~S ...--- ' ~ ..~.~"

~..~Is/- ~ ....--. IC
IS IS ~ _._...~ I.~.,S C SC __.--S c--- SC
~__~..~.~. .... Ic IC ~_----------/~~'-"--'-" .~S

%0 k I i 1 l i I 1 1 i I I ~ I I
0 50 100 150
Ma (b.p.)
FIG. 5. Grain contact morphology versus chronological age. The plot illustrates a progressive decrease of
tangential contacts and increase in planar-type contacts with age. The latter causes gradual increase in the
grain-grain contact areas and the state of fabric interlocking with age. *The Woburn and Folkestone sands
both have an approximate age of 100 Ma, and have been separated purely for clarity.
34 S. N. Palmer & M. E. Barton

FIG. 6. Photomicrographs of five of the sands to illustrate the progressive porosity reduction, increase in the
interlocking of the fabric (tightness of packing) and change in contact type morphologies: (a) modern-day
beach sand, with high porosity, very low number of grain contacts per grain, tangential contacts dominating;
(b) Barton Sand; (c) Folkestone Sand, tightness of packing increasing, planar contacts important; (d)
Kellaways Sand; (e) and (f) Grantham Sand, marked porosity reduction, high number of grain contacts per
grain, interlocked fabric, predominantly straight and concavo/convex contacts giving high grain-grain contact
areas. With the exception of the modern-day beach sand, which was impregnated in situ, all the sands were
block-sampled and impregnated in the laboratory using the Drip Method and the epoxy Epo-tec 301, spiked
with the dye Waxoline Blue AP-FW in a 1~ wt concentration (Palmer & Barton 1986). (Scale in (f) applies to
all photomicrographs.)

grain contacts are consistent with features A quantitative measure of lithification can be
known to be produced by pressure solution. obtained either from the shear strength
measured on dry samples in the direct shear box
(Fig. 8) or by taking values of unconfined
Lithification compressive strength (Fig. 9). Samples for these
tests were prepared by hand carving (Fig. 10).
Lithification is thought of in geological terms as The increase in age, and concomitant decrease in
a change from unconsolidated to indurated rock. porosity, is associated with an increase in
 Porosity reduction in UK uncemented sands 35

FIG. 7. SEM photomicrographs of (a) Barton (scale bar = 40 p.m) and (b) Grantham Sand (scale bar =
100 ~tm) showing the interlocked fabric, absence of cement and matrix. Features suggestive of pressure
solution are seen in the Grantham Sand.

lithification as measured by these tests of Discussion

strength.
The direct shear strength tests were conducted The sands studied clearly show a gradual series of
as slow, drained tests at normal stresses ranging diagenetic changes. Admittedly the degree of
from 50 to 900 kPa. The samples gave approxi- diagenetic change can be regarded as small--
mately linear envelopes over this stress range and thus we may compare, for instance, the consider-
show increasing values of shearing resistance ably altered but more deeply buried Brent Sand
angle (4r and increasing values of the cohesive of the North Sea with the comparably aged
intercept (c'), with age of the sand, correlating Grantham Sand as shown here. Comparisons of
with porosity reduction and an increase in the porosity with depth have been made by Atwater
number of grain contacts per grain. The beach & Miller (1965), Hsii (1977), Selley (1978),
sand, which could not be prepared for shear box Zieglar & Spotts (1978), Magara (1980) and
testing of the in situ fabric, and the Pleistocene Loucks et al. (1984) among others, but in all these
sand have values of cohesive intercept (c') equal cases the depth of burial is considerable and the
to zero. There is some evidence to suggest that at porosity much less (especially at the deeper
very low normal stresses the envelopes may levels) than found in the sands of this study. Even
become curvilinear but, owing to the usual so the porosities recorded in Table 2, and as
difficulty of obtaining reliable results in this shown in Fig. 3, are reduced below the original
range, further work will be required to substanti- deposited values for such sands and, therefore,
ate this. indicate a measure of diagenetic change.
Measurement of unconfined compressive Similarly we can compare the gradual changes
strength, using dry cylindrical samples in a in packing and particle contact types (as shown
triaxial testing machine, are given in Fig. 9 in Figs 3 and 4 respectively) with the well-
against values of grain-grain contact areas known and referenced study by Taylor (1950) of
(Dobereiner & de Freitas 1986) and numbers of the deeply buried Jurassic and Cretaceous sands
tangential contacts (a measure of fabric 'lock- in Wyoming. In this case, however, the changes
ing'). A further confirmation of the increasing in contact index and contact types are consider-
lithification with age is obtained together with ably more intense than observed in this study.
the relationship with the other diagenetic fea- Taylor concluded that porosity reduction, in-
tures--a decrease in tangential contacts and an crease in contact index, decrease in tangential
increase in grain contact areas. The most contacts and increase in concavo/convex and
'lithified' formation, the Grantham Sand, re- sutured contacts relate to an increase in compac-
mains geologically 'a sand' and geotechnically it tion. Selley (1982) and Burley et al. (1985) suggest
is best classed as an engineering soil. Although that the features in the Wyoming Sands are due
none of these sands are sufficiently indurated to to the development of quartz overgrowths. In
withstand coring, with the exception of the this latter case the observed changes would be
beach sand they are compact enough to be clearly of a different origin from the uncemented
sampled and transported as intact blocks in fabrics studied here. Other studies of grain
which the porosity and micro-fabric remain in contacts and packing parameters versus depth of
the in situ state (Barton et al. 1986a). burial include Beaudry (1950) and Hays (1951),
36 S. N. Palmer & M. E. Barton
2000--

///"
.d /
oY /
9/ ; + / /
t~
O_
,,i

I-O

9 /" .- ~"

/ o+, / /..
/ / , o..C
T j / /-. 4 . " ,
/ /',J ,/'.: /
/ r ///-- "
#
.../
. / ~ . ~ /
,..L~+,, /
. . / ./
/ / /@~ / ./

i
9 ./ /
/ "/ / // / -" ./ r
/ /// / /
,,...//i 4 ./t,./..I"

0 -
l I I I I I I I I I
0 1000
O'4n (kPa)
FIG. 8. Shear strength envelopes obtained from direct shear tests on dry samples using a standard 60 mm
square shear box. The envelopes are drawn as best-fit straight lines through the experimental points: there is
evidence that at low stresses the envelopes become markedly curvilinear. The strength of the sands increases
with the age of the deposits. The envelope of the beach sand is obtained from the maximum angle of repose in
the laboratory, since undisturbed samples for shear box testing could not be produced for this sand, it being a
truly unlithified deposit.

as reported by Wolf & Chilingarian (1976), of ly due to compaction (mechanical and/or chemi-
deeply buried sandstones showing cementation, cal) alone.
unlike the sands of this study. It would appear It is not intended to discuss here in detail
there is an absence of studies in the literature the reasons for the diagenetic changes observed
where porosity reduction in sand is unequivocal- in the sands studied, but simply to comment
 Porosity reduction in UK uncemented sands 37
10--
_

- -

- - GRANTHAM

-- - - K E L L A W A Y S
Ck _

__ ~ - - ' FOLKESTONE-

- -

- - W O B U R N

~THANET--
0.1~

- -

- - - - - BARTON

-- CRAG

0.01 - i
I ' I ' I ' I '
0 20 30 40 50

Grain Contact & Tan. Index (%)

Fro. 9. Unconfined or uniaxial compressive strength versus number of tangential contacts (tan. index) and
grain contact areas (grain contact). The plot illustrates that the unconfined compressive strengths of the sands
increase with age, and correlates to decreasing numbers of tangential contacts and increasing grain contact
areas, i.e. an increase in the interlocking.

upon the fundamental processes that may have the plastic deformation of ductile
been operative. The main diagenetic processes components.
which primarily control porosity reduction (ii) Mechanical compaction due to seismic
in sediments, and therefore also the develop- disturbances over a geological time-scale--
ment of microfabric in these sands (as causing the geometrical rearrangement of
discussed by Fuchtbauer 1967, Bjorlykke 1983 framework components.
and Houseknecht 1984) are: (iii) Chemical compaction--the intergranular
(i) Mechanical compaction due to overburden pressure solution (dissolution) at contacts
pressure--causing the rotation, reorienta- between particles, resulting in material
tion and fracturing of competent grains, and mass transfer.
38 S. N. Palmer & M . E. Barton

FIG. 10. Preparation of prismatic samples for direct shear box testing (left) and cylindrical samples (centre
and right) for confined and unconfined compressive tests. These samples are prepared by hand carving from
block samples such as that in the left background.

(iv) Cementation--the precipitation of authi- porosity reductions. The work of Selig (1963) and
genic mineral matter in interstitial pore the authors shows that the presence of an
space. overburden load inhibits the porosity reduction
Various studies indicate that the observed resulting from vibration.
porosities and microfabrics in the sands of this It is suggested that sands of this type, with
type could not have been obtained by simple depositional porosities of 47-48%, have under-
mechanical compaction. The fabrics produced gone only a small porosity reduction resulting
by laboratory compaction do not produce fea- directly from mechanical compaction owing to
tures consistent with the microfabrics observed the increasing stability produced in the structure
in natural sands and sandstones (Maxwell 1960). of the sand as porosity reduces. Furthermore it is
Berner (1971), Pettijohn et al. (1972), North considered that porosities equivalent to the
(1985) and preliminary studies by the authors all laboratory minimum values are unlikely to be
indicate that minimal porosity reduction will achieved diagenetically by mechanical compac-
occur by the mechanical mutual adjustment of tion alone at the relatively small burial depths of
grains under load. This is particularly so since these sands. Thus other diagenetic processes
the sands' compositions are dominated by pure, must have been operative to produce the dense
stable quartz, there being only very small state of packing observed, particularly in the
proportions of incompetent ductile clasts, thus Mesozoic sands for example. Chemical compac-
very much limiting porosity reduction by plastic tion and cementation are the other main pro-
deformation of grains (Nagtegaal 1978, Dodge & cesses which could be invoked to explain the
Loucks 1979). There is no petrographic evidence remaining porosity reduction in the sands.
to suggest that grain fracturing (caused by Petrographic, SEM and CL analyses reveal
mechanical compaction) significantly contribut- insignificant amounts of authigenic mineral
ed to porosity reduction. All the sands show phases (Table 2), so cementation does not
textural parameters, such as good or moderate account for the observed reductions in porosity.
sorting, which are similarly known to promote There is no evidence to suggest previous higher
the retention of porosity under load (Rittenhouse grade diagenetic changes: the observed porosi-
1971). The textural and mineralogical maturity ties must therefore be considered as primary
are widely recognized as major factors in porosities.
determining the mechanical compaction of a The simplest and most reasonable explanation
sediment (e.g. North 1985 and Selley 1985). of the porosity reduction and microfabric
Mechanical compaction due to seismic distur- changes therefore is that they are the products of
bance over geological time-scales is also not the effects of pressure solution. The geometries
considered as likely to affect the observed of the planar grain contacts, especially in the
 Porosity reduction in UK uncemented sands 39
older sands, are those widely known to result but further work is required to assess its relative
from congruent quartz dissolution. Although importance.
quantitatively the sands show relatively small The decrease in porosity, but more significant-
amounts of dissolution, minor pressure solution ly the progressive change of contact types and
can cause marked porosity reduction (Fficht- increase of grain contact areas with age, is
bauer 1967). SEM analysis of the Cretaceous and thought to explain the increase in the degree of
Jurassic sands shows micro-pitting and contact- induration. It is suggested, therefore, that these
hollows consistent with pressure dissolution. It is sands are primarily lithified by the physical
therefore tentatively proposed that, with refer- inter-locking of the detrital grains, mainly due to
ence to Fig. 3, initial rapid but small porosity the high states of packing obtained through
reduction is obtained by mechanical compac- pressure solution.
tion. Slower but more significant porosity loss is
obtained later by chemical compaction. A
similar conclusion for more deeply buried sands ACKNOWLEDGMENTS"The authors wish to express their
was presented by Ffichtbauer (1967). thanks to the following companies for giving permis-
It has been shown how the decrease in sion to visit and to sample at their sand-pits: Joseph
porosity and increase in grain contact areas Arnold & Sons Ltd; Barker-Mill Estates; S. E. Borrow
relates to an increase in lithification, and it Ltd; Ready Mixed Concrete (UK) Ltd; St Albans Sand
and Gravel Co. Ltd and in particular British Industrial
follows from the discussion above that this also
Sand Ltd.
results predominantly from pressure solution. We would sincerely like to thank Dr P. J. Gregson,
Any contribution to the lithification by the Department of Engineering Materials, Southampton
cohesion provided by simple clay bonding from University for the use of the SEM; Drs I. M. West and
the minor matrix (as discussed by Waldschmidst T. Clayton, Department of Geology, Southampton
1941 and Dapples 1972) and the negligible University, for the use of the XRD and CL apparatus
amounts of cement are thought to be minimal. and their helpful comments; Dr G. M. Power,
Evidence for this is also provided by the lack of Department of Geology, Portsmouth Polytechnic for
any correlation between strength and the percen- the use of the photomicroscope, and Mr P. G.
tage of either matrix or cement. Pressure Cambridge, Geological Society of Norfolk. Special
thanks are due to Messrs A. Brookes and Y. L. Wong
cohesion or intergranular welding of the detrital for technical assistance with sampling and testing and
grains (Dapples 1967, 1972, Pettijohn et al. 1972, for drawing the figures.
Fairbridge 1983, North 1985)probably contrib- Financial support for this research was provided by
utes to the lithification, possibly significantly, the Science and Engineering Research Council.

References
ANDERTON, R., BRIDGES, P. H., LEEDER, M. R. & BERNER, R. A. 1971. Principlesof Chemical Sedimento-
SELLWOOD,B. W. 1979. A Dynamic Stratigraphy of logy. McGraw-Hill, New York.
the British Isles. Allen & Unwin, London. BJORLYKKE, K. 1983. Diagenetic reactions in sand-
ATWATER, G. I. & MILLER, E. E. 1965. The effects of stones. In: PARKER, A. & SELLWOOD,B. W. (eds)
decrease in porosity with depth on future develop- Sediment Diagenesis, pp. 169-213. Reidel, Dor-
ment of oil and gas reserves in south Louisiana drecht, Holland.
(abs.) Bulletin of the American Association of BURLEY, S. D., KANTOROWICZ,J. D. & WAUGH, B.
Petroleum Geologists, 49, 334. 1985. Clastic diagenesis. In: BRENCHLEY, P. J.
BARTON, M. E. & BROOKES, A. (in press). Lateral & WILLIAMS, P. J. W. (eds) Sedimentology.
shaking during sedimentation--a new technique Recent Developments and Applied Aspects, pp. 189-
for obtaining the minimum porosities of granular 226. Geological Society Special Publication
soils. Ground Engineering. No. 18. Blackwell Scientific Publications,
--BROOKES, A., PALMER, S. N. & WONG, Y. L. Oxford.
1986a. A collapsible sampling box for the collec- CHILINGARIAN,G. W. 1983. Compactional diagenesis.
tion and transport of intact block samples of In: PARKER,A. & SELLWOOD,B. W. (eds) Sediment
friable uncemented sands. Journal of Sedimentary Diagenesis, pp. 57-167. Reidel, Dordrecht,
Petrology, 56, 540-41. Holland.
--PALMER, S. N. & WONG, Y. L. 1986b. A COPE, J. C. W., DUFF, L. L., PARSONS,C. K., TORRENS,
geotechnical investigation of two Hampshire H. S., WIMBLEDON,W. A. & WRIGHT, J. K. 1980.
Tertiary sand beds: are they locked sands? A correlation of the Jurassic rocks in the British
Quarterly Journal of Engineering Geology, 19, 399- Isles. Part Two: Middle and Upper Jurassic.
412. Geological Society of London, Special Report No.
, & (in prep.) Studies of a locked sand of 15.
Jurassic age in the U.K." the Grantham Forma- CURRY, D., ADAMS,C. G., BOULTER,M. C., DILLEY,
tion sand at Wittering, Cambridgeshire. F. C., EAMES,F. E., FUNNELL,B. M. & WELLS,M.
BEAUDRY, D. A. 1950. Pore-space reduction in some K. 1978. A correlation of Tertiary rocks in the
deeply buried sandstones. Unpublished Thesis, British Isles. Geological Society of London, Special
University of Cincinnatti. Report No. 12.
 40 S. N. Palmer & M. E. Barton

DAPPLES, E. C. 1967. Diagenesis of sandstones. In: MITCHELL, G. F., PENNY, L. F., SHOTTON, F. W. &
LARSEN, G. & CHILINGAR,G. V. (eds). Diagenesis WEST, R. G. 1973. A correlation of Quaternary
in Sediments, pp. 91-125. Elsevier, Amsterdam. deposits in the British Isles. Geological Society of
- - 1972. Some concepts of cementation and lithifica- London, Special Report No. 4.
tion of sandstones. Bulletin of the American NAGTEGAAL, P. J. C. 1978. Sandstone-framework
Association of Petroleum Geologists, 56, 3-25. instability as a function of burial depth. Journal of
DOBEREINER, L. & DE FREITAS, M. H. 1986. Geotech- the Geological Society of London, 135, 101-6.
nical properties of weak sandstones. Gbotechnique, NORTH, F. K. 1985. Petroleum Geology. Allen &
36, 79-94. Unwin, London.
DODGE, M. M. & LOUKS, R. G. 1979. Mineralogic PALMER, S. N. & BARTON, M. E. 1986. Avoiding
composition and diagenesis of Tertiary sandstones microfabric disruption during the impregnation of
along Texas Gulf Coast. Bulletin of the American friable, uncemented sands with dyed epoxy.
Association of Petroleum Geologists, 63, 440. Journal of Sedimentary Petrology, 56, 556-7.
DUSSEAULT, M. B. & MORGENSTERN, N. R. 1979. PETTIJOHN, F. J., POTTER, P. E. & SILVER, R. 1972.
Locked sands. Quarterly Journal of Engineering Sand and Sandstone. Springer-Verlag, Berlin.
Geology, 12, 117-31. PRYOR, W. A. 1973. Permeability-porosity patterns
FAIRBRIDGE, R. W. 1983. Syndiagenesos-anadia- and variations in some Holocene sand bodies.
genesis-epidiagenesis: phases in lithogenesis. In: Bulletin of the American Association of Petroleum
LARSEN, G. & CHILINGAR, G. V. (eds) Diagenesis Geologists, 57, 162-89.
in Sediments and Sedimentary Rocks, 2, 17-113. RAWSON, P. F., CURRY, D., DILLEY, F. C., HANCOCK,
Elsevier, Amsterdam. J. M., KENNEDY, W. J., NEALE, J. W., WOOD,
FRIEDMAN, G. M. & SANDERS,J. E. 1978. Principles of C. J. & WORSSAM, B. C. 1978. A correlation of
Sedimentology, Wiley, London. Cretaceous rocks in the British Isles. Geological
FUCHTBAUER, H. 1967. Influence of different types of Society of London, Special Report No. 9.
diagenesis on sandstone porosity. Proceedings of RITTENHOUSE, G. 1971. Mechanical compaction of
the 7th Worm Petroleum Congress, Mexico, 2, 353- sands containing different percentages of ductile
69. grains: a theoretical approach. Bulletin of the
HAYS, F. R. 1951. Petrographic analysis of deep well American Association of Petroleum Geologists, 52,
cores. Unpublished Thesis, University of 92-6.
Cincinnatti. SELIG, E. Z. 1963. Effect of Vibrationon Density of Sand.
HOUSEKNECHT,D. W. 1984. Influence of grain size and 2nd Pan-American Conference on Soil Mechanics
temperature on intergranular pressure solution, and Foundation Engineering, Rio de Janeiro, 1,
quartz cementation, and porosity in a quartzose 129-44.
sandstone. Journal of Sedimentary Petrology, 54, SELLEY, R. C. 1978. Porosity gradients in North Sea
348-61. oil-bearing sandstones. Journal of the Geological
HsO, K. J. 1977. Studies of Ventura Field, California Society of London, 135, 119-31.
II: Lithology, compaction, and permeability of -- 1982. An Introduction to Sedimentology (2nd ed.).
sands. Bulletin of the American Association of Academic Press, London.
Petroleum Geologists, 61, 169-91. -- 1985. Elements of Petroleum Geology. Freeman,
KOLBUSZEWSKI, J. 1948. An experimental study of San Francisco.
the maximum and minimum porosities of TAYLOR, J. M. 1950. Pore-space reduction in sand-
sands. Proceedings of the 2nd International Confer- stones. Bulletin of the American Association of
ence of Soil Mechanics and Foundation Engineering, Petroleum Geologists, 34, 701-16.
Rotterdam, 1, 158-65. WALDSCHMIDT, W. A. 1941. Cementing materials in
LOUCKS,R. G., MARIANNE,M. D. & GALLOWAY,W. E. sandstones and their probable influence on migra-
1984. Regional controls on diagenesis and reser- tion and accumulation ofoil and gas. Bulletin of the
voir quality in Lower Tertiary Sandstones along American Association of Petroleum Geologists, 25,
the Texas Gulf coast. In: MCDONALD, O. A. & 1839-79.
SURDAM, R. C. (eds) Clastic Diagenesis, pp. 15-45. WHITTAKER, A. (ed.) 1985. Atlas of Onshore Sedimen-
The American Association of Petroleum tary Basins in England and Wales: Post-Carboni-
Geologists. ferous Tectonics and Stratigraphy. Blackie,
MAXWELL,J. C. 1960. Experiments on compaction and London.
cementation of sand. In: GRIGGS, D. & HANDIN,J. WOLF, K. H. & CHILINGARIAN,G. V. 1976. Diagenesis
(eds) Rock Deformation. Geological Society of of sandstones and compaction. In: CHILINGARIAN,
America Memoir, 79, 105-32. G. V. & WOLF, K. H. (eds) Compaction of Coarse-
MAGARA, K. 1980. Comparison of porosity-depth grained Sediments, II, 69-444. Elsevier,
relationships of shale and sandstones. Journal of Amsterdam.
Petroleum Geology, 3, 175-85. ZIEGLAR, D. L. & SPOTTS, J. H. 1978. Reservoir and
MEADE, R. H. 1966. Factors influencing the early source-bed history of Great Valley, California.
stages of the compaction of clays and sand-- Bulletin of the American Association of Petroleum
review. Journal of Sedimentary Petrology, 36, Geologists, 62, 813-26.
1085-101.
S. N. PALMER & M. E. BARTON, Department of Civil Engineering, University of
Southampton, Southampton SO9 5NH, UK.
 Non-steady state microbiological diagenesis and the origin of
concretions and nodular limestones

R. Raiswell

S U M M A R Y : Morphological, chemical, isotopic and textural studies of pyritiferous

carbonate concretions from the Jet Rock (Lower Jurassic, England) are all consistent in
indicating an origin by anaerobic methane oxidation. Concretionary growth occurred in a
stratigraphically thin, sub-surface zone of uncompacted sediment, where methane (diffusing
from below) was consumed to stimulate a late phase of locally intense sulphate reduction,
which caused the precipitation of CaCO3 and isotopically heavy FeS2. The carbon isotope
composition of the concretionary carbonate was indistinguishable from that of sulphate
reduction (using sedimentary organic matter), due to the simultaneous diffusion of
isotopicaUy light C H 4 and isotopically heavy dissolved carbonate from the underlying
methanogenic zone. The alkalinity generated by anaerobic methane oxidation resulted in
precipitation being restricted to a thin, vertically confined zone beneath the sediment/water
interface. A pause in deposition then allowed continued precipitation in the form of
concretionary growth. Different stages in the evolution of isolated concretions into nodular
limestones depend on the magnitude of carbonate supersaturation in the zone of anaerobic
methane oxidation and the duration of the pause in sedimentation. Vertically confined zones
of carbonate precipitation may also be developed deeper in the sediment due to iron
reduction, a further pause in deposition can then produce isotopically heavy carbonate
concretions in the zone of methanogenesis.

The formation of carbonate concretions is one of associated with pyrite). Finally a tentative
the most ubiquitous but enigmatic features in the extrapolation will be made to explain how
diagenesis of ancient organic carbon-rich sedi- alternative diagenetic pathways lead either to
ments. Substantial progress has been made in early diagenetic concretions (as above) or to later
identifying the timing and nature of the concre- diagenetic concretions (formed by methanogene-
tionary growth process and stable isotopes have sis, giving a characteristically heavy 613C signa-
proved particularly valuable in unambiguously ture often associated with iron-rich carbonates).
showing that the principal authigenic carbonate
source is from the microbiological decay of
organic matter, either by sulphate reduction or
methanogenesis (Curtis et al. 1972, Hudson 1978, Previous studies
Pearson 1979, Coleman & Raiswell 1981, Gau-
tier 1982 and others). Concretionary growth, Localized cementation as carbonate concretions
however, is a localized phenomenon despite the has been envisaged as resulting either from
fact that every horizon in an ancient organic-rich isolated nucleation within an environment which
marine sediment must have passed through the is uniformly supersaturated with respect to
successive burial zones of sulphate reduction carbonates, or from the generation of localized
and/or methanogenesis. This poses a fundamen- supersaturation by the decay of concentrations of
tal question: precisely what factors localized organic matter at the concretionary site (Rais-
microbial decay to produce the stratigraphic well 1976). For example the excellent preserva-
confinement of concretionary horizons, and the tion of the fossils found inside concretions can be
siting of individual concretions? interpreted as suggesting that fossil cores are an
The present contribution will address these essential feature of concretionary growth, either
questions specifically for the occurrence of two because the fossil originally contained organic
concretionary horizons in the Jet Rock (Lower body material, or because the shell nucleated
Jurassic) of NE England. A model of concretion- carbonate precipitation.
ary growth in the Jet Rock will be proposed However a nucleation control is difficult to
which should be generally applicable to other reconcile with the extreme variability in the
early diagenetic carbonate concretions formed relative proportions of barren and fossiliferous
by sulphate reduction (those comprising iron- concretions which occur in some sediments
poor calcite with a light •13C signature, often (Weeks 1957) and with the coexistence of barren

From: MARSHALL, J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences,

Geological Society Special Publication No. 36, pp. 41-54.
41
42 R. Raiswell

and fossiliferous concretions within the same of organic soft-bodied material which could
horizons in which non-concretionary fossils also decompose entirely without leaving any residual
occur (Waage 1964). Furthemore in many cases tissue or skeletal material, and which was also
non-concretionary fossil nuclei are sufficiently thoroughly disseminated and intimately mixed
abundant in concretionary sediments such that with the host sediment. Here these circum-
the distribution of macrofossils is essentially stances are regarded as unlikely and instead it is
unrelated to concretions (e.g. Dickson & Barber proposed that the Jet Rock concretions grew in
1976, Raiswell 1976). These observations dem- thin zones below the sediment surface where
onstrate that fossil nuclei cannot always have alkalinity (and hence carbonate supersaturation)
been the principal control on concretion siting. was locally enhanced by the anaerobic oxidation
Instead, siting control is usually attributed to of dissolved methane.
the microbiological decay of local concentrations
of organic matter (Weeks 1957, Zangerl et al.
1969, Sass & Kolodny 1972, Dickson & Barber Approach
1976, Berner 1980). Experimental work (Berner
1968) has shown that the anaerobic decay of fish At Port Mulgrave, 11 km NW of Whitby
and clam body material can generate supersatur- (Yorkshire), the Jet Rock consists of fine-
ation with respect to carbonates, although grained, well-laminated sediment containing up
precipitation in fact occurs as Ca soap (which to 12% organic carbon (Raiswell & Berner 1985).
may later be converted to CaCO3). A well- Two laterally extensive concretionary horizons
documented example of siting control through (Howarth 1962) contain pyritiferous carbonate
local concentrations of organic matter occurs in concretions and are known as the Curling Stones
the Fayetteville Shale (Zangerl et al. 1969), and the Cannonball Doggers.
where concretionary growth occurred around The Cannonball Doggers range from flattened
coprolitic material and shells containing organic elongate bodies to regular ellipsoids (maximum
carcasses. Here the presence of organic matter diameter 30 cm) whereas the Curling Stones are
was clearly indicated by several lines of evi- commonly oblate spheroids up to 70 cm maxi-
dence; namely, the textural similarities with mum diameter. In each case the host sediment is
coprolite, the absence of sediment from inside differentially compacted around the concretion,
shell body chambers and, crucially, by the demonstrating growth and cementation in poorly
explosive brecciation of shells from the pressure compacted sediment (Raiswell 1976).
of gases generated during microbiological decay A single concretion was collected from each of
(whose escape was prevented by concretionary the horizons bearing the Cannonball Doggers
cementation). Similarly, concretions which en- (concretion UA) and the Curling Stones (concre-
compass intact skeletons of organisms such as tion UB). Each concretion was sampled by
fish (Weeks 1957) are also observed to preserve removing a slice from the centre, parallel to the
delicate features in tissue, body and other non- bedding plane. This slice was further sub-divided
skeletal material. Such concretions clearly owe to provide a representative edge-centre-edge
their origin to the microbiological decay of traverse. Further details of the sampling scheme
organic matter associated with the residual and analytical methodologies are given else-
skeleton. where~ here we are concerned with a new
However there are considerable difficulties in interpretation, integrating earlier chemical
accepting the decay of organic carcasses as the (Raiswell 1976), isotopic (Coleman & Raiswell
main control on the siting of concretions like 1981) and textural (Raiswell 1982) data in the
those in the Jet Rock, because there is simply no context of recent studies of anaerobic pore water
evidence to indicate the presence of any organic chemistry.
carcasses. Large, intact skeletal cores are uncom-
mon in the Jet Rock concretions, nor are there
any apparent associations with an organic
carcass. Instead the concretions consist of an Chemical, isotopic and textural
intimate mixture of intact and broken skeletal features of the Jet Rock concretions
debris, authigenic calcite and pyrite, and small
proportions (< 15%) of the host sediment (Rais- The two most extensively studied concretionary
well 1976). There is no textural evidence for the horizons in the Jet Rock (the Curling Stones, the
presence of organic tissues nor is there any field Cannonball Doggers) can be traced at the same
evidence for a local concentration of organic stratigraphic level over nearly 20 km distance
matter. through which there is little apparent variation
Collectively, these observations can only be in the size, shape or frequency of the concretions.
reconciled by proposing the existence of a mass Clearly the horizons are laterally persistent but
 The origin of concretions 43
stratigraphically confined. Evidence for their matter (at - 2 5 to -30%o). However, organic
early diagenetic origin in uncompacted sediment carbon contents inside the concretions show no
is conclusive. Field observations include the clear evidence for any local concentrations of
presence of uncrushed skeletal debris and the organic carbon, on an authigenic mineral (i.e.
compaction of host sediment laminae around the carbonate- and pyrite-free) basis. Thus the
concretions (Raiswell 1976). An early origin is concretions (UA 8.3 + 2.2~C, UB 8.6 + 3.7%C)
further confirmed by estimates of the host are more variable, but not statistically dissimilar,
sediment porosity at the time of concretionary in organic carbon content to the host sediments
growth, based on the assumption that the (UA 6.3 + 0.1%C, UB 9.0 + 0.6~C). Moreover
concretion cement replaces water-filled pore the isotopic compositions of the residual organic
space (Raiswell 1971). Porosity estimates of 77- matter inside the concretions (UA - 3 2 to
87~ were obtained from the Jet Rock concre- -33%0, UB - 2 6 to -27~o) and outside in the
tions which indicates an origin in the top few sediments (UA - 3 1 to -35%o, UB - 3 0 to
metres of sediment (Raiswell 1976). -31%o) are similarly light, probably due to the
The carbon isotope composition of the concre- selective preservation of a lipid fraction (Cole-
tionary carbonate is - 13 to - 16%o (Coleman & man & Raiswell 1981). These observations
Raiswell 1981), which has been interpreted as indicate that microbial concretionary growth
mainly indicating a mixture of marine-derived required an organic matter source which could
dissolved carbonate (at about 0%o) and carbonate decompose entirely to dissolved carbonate, with-
generated by sulphate reduction from organic out leaving any chemical or isotopic signature.

80
c9

r =-
0
.0
t_
03
0
v| 4 0
"7.
a.
ae

inc. E~---F F---~inc,E F inc.E ~-- F + E ---~ i n c . E

PYRITE TEXTURE

0
co
03

-10

-20

-30 1 I I ] 1 I I I I 9 I I I I I | I I I 1 1 1

1 2 3 4 5 6 7 8 9 Host 1 2 3 4 5 6 7 8 9 10 11 H o s t
Sediment Sediment
edge centre edge edge centre edge
UA UB

FIG. 1. Edge-centre-edge traverses in pyrite content, and (~345of pyrite for Jet Rock concretions UA and UB.
Host sediments contain mainly framboidal pyrite (F), as also does the central part of concretion UA. The
centre of concretion UB contains framboids overgrown by euhedral pyrite (E). Both concretions show a
trend of increasing euhedral pyrite towards their margin (after Coleman & Raiswell 1981).
44 R. Raiswell
Figure 1 shows values for the masses and (ii) Required an organic matter source which
isotopic composition of pyrite inside the concre- could decompose entirely to leave no chemi-
tions and host sediments (on a carbonate-free cal or textural trace.
basis). Pyrite in both host sediments occurs (iii) Occurred at a depth where the pore system
mainly as framboids, which have 634S = - 2 4 to contained high concentrations of isotopical-
-26%0. Framboids appear similarly abundant ly light carbonate, sufficient to produce
inside both concretions, where they are envel- carbonate supersaturation.
oped by a later phase of equigranular, euhedral (iv) Was uniquely characterized by the occur-
pyrite which is associated with concretionary rence of a renewed phase of local sulphate
growth. This pyrite nucleates on the framboids reduction, at a depth within the sediment
and steadily increases in abundance towards the where sulphate was isotopically heavy and
concretion margins (Raiswell 1982), as shown in hence depleted.
Fig. 1.
This combination of circumstances has been
The pyrite at the centre of concretion UA is recognized in studies of modern pore waters, and
exclusively framboidal and, after correction to a provides a unique fingerprint for the involve-
carbonate-free basis, its abundance is similar to
ment of anaerobic methane oxidation in concre-
that in the host sediment. However with increas-
tionary growth.
ing amounts of the syn-concretionary euhedral
pyrite, the bulk isotopic composition becomes
steadily heavier and reaches - 2 to -3%0 at the Anaerobic methane oxidation in
margin of the concretion. By contrast the
samples at the centre of concretion UB have modern sediments
some euhedral overgrowths on the framboids,
and also contain more pyrite than the host The presence of anaerobic methane oxidation
sediments on a carbonate-free basis (Fig. 1). The has been inferred from studies of pore water
presence of the euhedral pyrite causes the central chemistry in a variety of modern sediments
samples of concretion UB to be isotopically (Barnes & Goldberg 1976, Reeburgh 1976, 1980,
heavier ( ~ - 8%0)and more pyritiferous than the Martens & Berner 1977, Devol & Ahmed 1981,
host sediments. Apart from these central sam- Devol 1983). These studies show the following
pies, concretion UB shows essentially the same generalized features in the depth distributions of
trend towards isotopically heavy pyrites at the dissolved carbon and sulphur species and their
concretion margin ( - 5 to - 6%0) as the euhedral isotopic compositions. Distinct zones of steady
pyrite becomes more abundant. The margins of state diagenesis can be recognized in modern
both concretions contain 5-10 times more pyrite anaerobic sediments (Fig. 2), arising from the
than their host sediments. influences of sulphate reduction, methane oxida-
The trends in both concretions can be ex- tion and methanogenesis (Reeburgh 1980, Rice
plained on the same basis, where an initial phase & Claypool 1981) as described below.
of framboids formed uniformly throughout the
sediment prior to concretionary growth. The
isotopic composition of these pyrites indicates Sulphate reduction
that they formed at shallow depths, in a pore In the uppermost zone sulphate reduction
system with a free diffusive connection to
overlying seawater. These pre-concretionary pyr- 2CH20 + 8 0 4 2 - - ~ 2 H C O 3 - + H2S (1)
ites were passively occluded by the growing causes a steady decrease in dissolved sulphate
concretions; thus framboids of the same isotopic with depth. The instantaneous isotopic fraction-
composition and abundance can be recognized at ation between seawater sulphate (+20%0) and
the centre of concretion UA and in the host sulphide is commonly about - 4 0 to -60%0, so
sediments. Concretionary growth itself was that the pyrite being formed by reaction between
associated with formation of the euhedral pyrite dissolved H:S and detrital iron oxides is always
(6345 ranging up to -2%o), which increased the lighter than the residual sulphate (which itself
total pyrite in the concretions to levels signifi- progressively evolves towards heavier values).
cantly above those in the host sediments. The Thus later pyrites formed from a depleted
isotopically heavy pyrite formed in a pore system sulphate reservoir are isotopically heavy. Dis-
from which sulphate was depleted. solved carbonate (YCO:) concentrations in-
In summary, these data indicate that concre- crease as sulphate decreases and, because the
tionary growth: carbonate is derived from the oxidation of
(i) Was confined to a thin, subsurface zone of organic matter (613C = - 2 5 to -30%o), its
uncompacted sediment within a few metres isotopic composition decreases rapidly with
of the sediment surface. depth from surface values of 0%0 to approach the
 The origin o f concretions 45

~34 S or t~13 C
-80 -40 0 .400/00 0 20 40 60 Conc(mmol)
Depth ~34S Sulphate
1 / ~S042-)t/02_-~2 Reduction

iiii!!!i iiill i !ii!!!i::i ::::::i: ::i ill:: :ii::i ::: ] Anaerobic

~
~iii::ii::::!::iLi
N~: ~ii!::i::!::::::~
I~C / OxidatiOn
~iii Methane

(~13 C 3C H4 Methanogenesis
/'~ E \ I
FIG. 2. Steady state diagenetic zones in modern anaerobic sediments (depth scale in arbitrary units). Sulphate
Reduction Zone is characterized by [SO42-] ~ 0 to 28 mmol with 63"~Sat ~0 to -20%0; [Y~COz]> 2 mmol
with 613C at 0 to -25%0. Anaerobic Methane Oxidation Zone (shaded) has [SO42-] ~ 0 mmol with
634S --- 0%0; [YCO2] >>2 mmol with ~13C at -15 to -25%0; [CH~] low with 613C at -60 to -80%0.
Methanogenic Zone has [Y~CO2]>>2 mmol with ~513C~ 0%0; [CH~] increasing with 613C at -60 to -80%0.
After Reeburgh (1980) and Rice & Claypool (1981).

value of the organic phase. Note that sulphate water sediments (Rice & Claypool 1981). Meth-
reduction adds dissolved carbonate as HCO 3- ane production does not apparently become
and the consequent increases in alkalinity (to- significant until pore water sulphate is nearly
gether with that due to ammonia formation) depleted although the reasons for this are still
produce oversaturation with respect to CaCO3 under discussion (Claypool & Kvenvolden 1983).
(Berner et al. 1970, Sholkovitz 1973). The In any event the pore-water profiles show that
precipitation of the pre-concretionary pyrite by dissolved CH4 becomes significant only when
the reaction between iron oxides and H2S, sulphate is nearly depleted, also that concentra-
tions of dissolved carbonate subsequently in-
2Fe203 + 8SO42- + 15CH20
crease less rapidly, and are close to their
~4FeS2 + 15HCO 3- + 7H20 + OH- (2)
maximum levels (Claypool & Kaplan 1974,
also generates alkalinity (Coleman 1985). Clearly Reeburgh & Heggie 1977, Reeburgh 1980,
the processes occurring in the sulphate reduction Claypool & Kvenvolden 1980, Devol et al. 1984),
zone are important in producing oversaturation because CH4 is now being produced by CO2
of CaCO3, but most Ca 2+ removal from reduction. The isotopic trends in dissolved
porewaters by precipitation appears to occur at carbonate confirm these processes. The CO2
the base of the sulphate reduction zone, where removed to form methane is typically ~70%0
methane oxidation becomes significant, e.g. lighter than the 613C of the bulk dissolved
5 cm and below in Saanich Inlet sediments carbonate, due to a kinetic isotope effect (Clay-
(Murray et al. 1978, Devol 1983). pool & Kaplan 1974). Thus continuing methano-
genic CO,, reduction is indicated by increasing
Methanogenesis ~5~3C values of the residual dissolved carbonate.
The relative rates of metabolic CO, addition and
Below the sulphate reduction zone, the main
methanogenic removal will determine the pre-
microbiological process is methanogenesis
cise nature of the depth changes in the isotopic
which is often referred to as CO2 reduction
composition and concentrations of dissolved
(Claypool & Kaplan 1974). The reducing agent is
carbonate, but dissolved carbonate usually
usually assumed to be H2, which must be derived
reaches its minimum ~13C value at the start of
from organic matter.
methanogenesis, although concentrations of dis-
2CH20 + H20-~2CO 2 + 4H 2 (3) solved carbonate continue to increase (albeit
now more slowly), as shown in Fig. 2. Note that
4H2 + CO2~CH4 + 2H20. (4) dissolved carbonate is added as CO2 in the
Methanogenic bacteria can also use other sub- methanogenic zone (rather than as HCO 3- in the
strates but the above mechanism probably sulphate reduction zone) and hence alkalinity
accounts for most methane produced in deep will now decrease (Rice & Claypool 1981) unless
46 R. Raiswell
significant iron reduction accompanies methan- rates are enhanced also corresponds with maxi-
ogenesis (Coleman 1985), as shown below. mum rates of methane oxidation and occurs
within the top metre of sediment (Devol &
13CH20 + 2Fe203 + 3H20
Ahmed 1981, Devol 1983, Devol et al. 1984,
~6CH4 + 7HCO3- + 4Fe 2+ + OH-. (5)
Iverson & Jorgensen 1985).
In the Jet Rock, iron is almost completely The contributions of anaerobic methane oxi-
consumed to form the pre-concretionary pyrite dation to the isotopic composition of the dis-
(Raiswell 1982) and hence insufficient alkalinity solved carbonate in the porewater will initially
can be generated by iron reduction to offset the be exceedingly light, since the CH4 utilized will
production of CO2 by methanogenesis alone. be --- 70%0lighter than the CO: from which it was
derived. However the removal of light CO2 to
form CH4 is accompanied by a rapid increase in
Anaerobic methane oxidation
the 513C of the residual carbonate (see earlier).
A significant feature of Fig. 2 is that dissolved The overall isotopic composition of the pore
methane concentrations are low or absent in waters in the zone of anaerobic methane oxida-
near-surface sediments but increase with depth. tion will initially reflect the size and isotopic
It is argued that this type of profile can only composition of the pool of dissolved carbonate
result from CH4 consumption reactions which initially added from skeletal debris and seawater
act as a barrier to the transfer of methane across (at 0%~ and from sulphate reduction (at - 2 5 to
the sediment/water interface (Martens & Berner -30%o), which is modified by the rate of
1977, Reeburgh 1976, 1980, Devol 1983). In addition of light CO2 from anaerobic methane
sediments deposited from anoxic bottom waters oxidation, and by the rates of diffusion of
(as is believed to be the case for the Jet Rock; dissolved carbonate to, or from, adjacent
Raiswell & Berner 1985), removal cannot occur porewaters.
by bioturbation or aerobic oxidation and anaero- A more detailed attempt is made in the
bic consumption by sulphate reduction has been following section to model the fi~3C changes in
proposed. the zone of anaerobic methane oxidation. At this
point it is noted that despite the complexity of
CH4 + SO42--~HS - + HCO 3- + H20. (6)
the system the generalized sequences of Ree-
The reaction is thermodynamically favourable burgh (1980) and Claypool & Kvenvolden (1983)
and sulphate reducers capable of oxidizing CH4 show that the boundary between sulphate reduc-
have been cultured in the laboratory, but not tion and methanogenesis is marked by minimum
isolated from modern sediments (Reeburgh 513C values for dissolved carbonate ( - 1 5 to
1980). Observations of pore water chemistry -25%~ at maximum or near-maximum concen-
provide the strongest evidence for anaerobic trations of dissolved carbonate. This range of
methane oxidation and these have recently been 613C values is typical of those found in carbonate
successful in documenting the existence of a thin concretions, probably because anaerobic meth-
zone of locally enhanced sulphate reduction, ane oxidation increases carbonate supersatura-
occurring at the base of the sulphate reduction tion to the point of precipitation more rapidly
zone (Fig. 3). The zone where sulphate reduction than the 513C of the dissolved carbonate pool can

2- -3 -1
CH 4 0 x i d . Rate SO 4 R e d . R a t e (nmol.cm h )
0 0.4 0 4 8
i I i t t

Depth

i• 9
-__-7o-,-2-.... ~

::~::::::::::::
;

PRINCIPAL ZONE OF
:::!!) , i.~.0 . . . . i i i . i . i i . i i i . ANAEROBIC

20 ~ r ~ .......... ==== ===================================

t
FIG. 3. Correspondence between the depth profiles of rates of methane oxidation and rates of sulphate
reduction, indicating methane consumption by sulphate reducing bacteria (after Devol 1983).
 The origin o f concretions 47

be changed (see later). Atypical circumstances is small, relative to the mass of carbonate
where the (~13C of the dissolved carbonate can be precipitated. That this is also the case for the Jet
more rapidly altered may be responsible for the Rock concretions is demonstrated below.
very light 6~3C values recorded in some concre- The Curling Stones are separated by a mean
tions (e.g. -42%o, Hudson & Friedman 1978). nearest neighbour distance of approximately
Anaerobic methane oxidation clearly increases 0.47 m (Raiswell & White 1978). The Curling
alkalinity (equation 6), and thus may produce Stone analysed by Raiswell (1976) is of average
further sufficient increases in carbonate supersa- size and contains approximately 153 moles of
turation (over and above that resulting from the CaCO3. Typical anaerobic pore waters at the
earlier phase of sulphate reduction and pyrite base of the sulphate reduction zone contain
formation; see equations 1 and 2) to induce about 50 meq 1-1 of alkalinity (e.g. Berner et al.
precipitation. 1970, Sholkovitz 1973). Assuming that all alka-
In summary, these studies of modern sedi- linity is HCO3- and that its complete precipita-
ments indicate that anaerobic methane tion can be achieved, a supply zone of radius
oxidation: 23.5cm in a sediment of 85% porosity is
estimated to contain the following amounts of
(i) Is confined to a thin subsurface zone of
carbonate.
uncompacted sediment, within a metre of
the sediment surface. Zone thickness are (23"5)3 • 0.85 • 50 • 10-3
10-50 cm (Reeburgh 1983) and the upper 1000
surface lies about 10-100cm below the = 2.3 moles HCO3-.
sediment/water interface (Reeburgh 1976,
This is one to two orders of magnitude less than
1980, Devol & Ahmed 1981, Devol 1983,
that required for concretionary growth. Clearly
Iverson & Jorgensen 1985).
most carbonate must be derived from inputs
(ii) Utilizes dissolved or gaseous methane,
uniquely associated with concretionary growth,
which could be decomposed without leaving
furthermore the isotopic signature of the concre-
any residual phases in the sediment.
tions will reflect these inputs, rather than that
(iii) Occurs at a depth where the pore system
from the pool of dissolved carbonate arising from
contains high concentrations of isotopically
the pre-concretionary diagenetic processes.
light carbonate, and where the further
There are two kinetic fractions which are of
increases in alkalinity could be sufficient to
potential importance in the methane formation
cause carbonate precipitation.
and oxidation cycle. The first is that associated
(iv) Is uniquely characterized by the occurrence
with methanogenesis, where the CO2 removed to
of a renewed phase of sulphate reduction, at
form CH4 is around 70%o lighter than the CO2
a depth within the sediment where sulphate
from which it was derived. Continued removal of
is substantially depleted and isotopically
CO, thus leads to the residual dissolved carbon-
heavier.
ate becoming heavier (Fig. 2), and hence the
methane also evolves to heavier values. The
initial methane often ranges from - 70 to - 85%o
and porewaters found in the methanogenic zone
Carbon isotope sources to the zone (Claypool & Kaplan 1974) typically show that
of methane oxidation methanogenesis has continued to give CH4 at
- 6 0 to -70%0, which is associated with dis-
The previous section demonstrated that pore solved carbonate at 0 to +10%o. Note that
waters in the zone of anaerobic methane oxida- whatever the concentrations and isotopic com-
tion typically have carbon isotope compositions positions of the separate CH4 and CO2 phases,
in the range - 15 to - 25%o. Under steady state the bulk carbon phase (CO2 + CH4) will still
diagenesis this signature potentially reflects an retain the approximate isotopic signature of the
initial reservoir composition inherited from organic matter from which the CO2 and CH4
shallow diagenetic sources (marine-derived ma- were both initially derived (i.e. - 2 5 to -30%o).
terial and sulphate reduction carbonate) modi- A further kinetic fractionation is associated
fied by the addition of light dissolved carbonate with anaerobic methane oxidation, such that the
from anaerobic methane oxidation and by residual methane becomes heavier by 2-14%o;
diffusive fluxes. The usual treatment of carbon Whiticar & Faber (1986). Thus the bicarbonate
isotopes in diagenesis has been to assume that produced by oxidation should be this amount
the carbonates precipitated reflect the isotopic lighter than the CH4 from which it was derived.
signature of the input processes (i.e are indepen- In proposing methane as the organic carbon
dent of initial reservoir composition). In general source for concretionary growth, it is clearly
this is only true when the initial isotopic reservoir necessary to explain why exceedingly light 613C
48 R. Raiswell
values are not commonly found in carbonate carbonate removal in Saanich Inlet sediments at
concretions. It is suggested here that both CH4 depths both before and during anaerobic meth-
and dissolved carbonate diffuse upwards to the ane oxidation. In the following section a model of
zone of methane oxidation in such a way that concretionary growth is proposed which explains
their net isotopic signature ( - 2 5 to -30%0) is how the involvement of methane oxidation
not grossly altered. Once in the zone of methane imposes spatial constraints on the distribution of
oxidation, the light H C O 3- derived from CH4 concretionary growth.
mixes with the heavier transported H C O 3 - and
thus a light diagenetic phase is precipitated. Chemical and physical constraints on the
Flux measurements of dissolved carbonate distribution of concretionary growth
and CH4 suggest that this explanation is reasona- The features of anaerobic methane oxidation
ble. Sediments from the Cariaco Trench have already discussed are summarized in the follow-
equal or nearly equal fluxes of CH4 and dissolved ing model of concretionary growth (Fig. 4). The
carbonate from the methanogenic zone into the growth process is initiated at depth, in the
zone of methane oxidation (Reeburgh 1976). In methanogenic zone, where isotopically light CH4
Saanich Inlet, methanogenesis produces C H 4 and heavy dissolved carbonate are produced.
and CO2 at similar rates (Murray et al. 1978), Both diffuse upwards at rates which in most
however, although the existing carbonate pool cases are not sufficiently different to produce a
(mainly from sulphate reduction) enhances car- distinct isotopic separation. Thus the carbon
bonate fluxes over CH4, the overall 6~3C value phases overall more or less retain their original,
must still be light. In these cases the extent of organic matter, isotopic signature. In the zone of
fractionation within the methanogenic zone (and anaerobic methane oxidation, the H C O 3- de-
the oxidation zone) will be unimportant and the rived from CH~ mixes with the transported
net isotopic signature of the carbon phases may dissolved carbonate and the locally enhanced
be close to that of organic matter, provided alkalinity generates carbonate precipitation.
methane is completely oxidized to dissolved Iron may also diffuse upwards from the methan-
carbonate. The absence of CH4 in near surface ogenic zone to react with the sulphide produced
sediments (Fig. 2) indicates that this is generally by methane oxidation (Raiswell 1982).
the case, except perhaps where rapid methane The vertical confinement of concretionary
ebullition is caused by tidal hydrostatic pressure horizons arises in the first instance from the
changes in shallow marine sediments (Martens & chemical controls on the methane oxidation
Klump 1984). This flux evidence provides strong process, which is constrained upwards by lack of
support for the transportation of dissolved C H 4 (because C H 4 generation does not begin
carbonate plus CH4 without isotopic separation, until sulphate is nearly depleted) and downwards
as is also indicated by the 613C values of by lack of sulphate. Thus increased carbonate
dissolved carbonate in the zone of methane supersaturation is generated only in a vertically
oxidation (Fig. 2). In these circumstances the localized zone.
isotopic signature of sulphate reduction using The preceding phase of sulphate reduction
sedimentary organic matter and using C H 4 will may also be important in increasing alkalinity,
be indistinguishable. and hence carbonate supersaturation, to the
point where precipitation is close. Whether or
Synthesis not concretionary growth then occurs would
depend on sufficient further supersaturation
The two summaries demonstrate a remarkable being induced by anaerobic methane oxidation,
similarity between the chemical and physical despite the inhibiting effects of dissolved organ-
conditions required for concretionary growth ics and Mg 2§ on C a C O 3 precipitation. Anaero-
and caused by anaerobic methane oxidation. The bic methane oxidation appears to be more
carbon isotope evidence has also been shown to important at some locations than others (Devol
be consistent with the involvement of methane et al. 1984) and in some cases its magnitude may
oxidation in concretionary growth, and a causal be insufficient to cause carbonate precipitation.
link between the two processes is strongly The reasons for variations in the importance of
indicated. Further studies of modem porewaters anaerobic methane oxidation between locations
could add still more support if full porewater are not at present understood (Devol et al. 1984).
anlyses were made which would allow carbonate Within the zone of methane oxidation, rates of
saturation indices, and rates of carbonate preci- sulphate reduction may be seasonally and spa-
pitation, to be estimated and to be linked to rates tially variable (Reeburgh 1980, Devol & Ahmed
of anaerobic methane oxidation. No such data 1981, Crill & Martens 1983, Iverson & Jorgensen
have yet been published, although the data of 1985). Seasonal variations are probably unim-
Murray et al. (1978) do appear to demonstrate portant on the time-scale of concretionary
 The orig& of concretions 49

ZONE Water

Sediment
SULPHATE
REDUCTION

ANAEROBIC
METHANE
OXIDATION

At c o n c r e t i o n site

C H 4 + SO 2 - - - ~ H C O ; + H S - + H 2 0
-60 to-80 - 6 0 to - 8 0

HCO - + HCO -._,. 2HCO - PRECIPITATED

3 3 3 AS CaCO3
- 6 0 to 0 to -25 0/00 to-300/O0
-80 +10
METHANOGENESIS ~ ~ Diffusive flux
Spatially variable
HCO + CH
3 4

2CH O+2H O . - ~ 2CO + 4 H 1

2 2 2 2 t ~13 O/
CO 2 + 4 H 2 --~ CH + 2H O C ~-25 O0
4 2 t o - 3 0 0/O0
CO + H O ..., H C O - + H §
2 2 3

FIG. 4. A schematic representation of the chemical processes occurring at the site of a growing concretion,
where a relatively high flux of methane sustains high methane oxidation rates, and hence produces locally
enhanced carbonate supersaturation. Equal HCO 3- and CH4 fluxes from methanogenesis, unaccompanied by
other carbonate sources.

growth, but the spatial variations may determine expected between concretionary growth and
the siting of individual concretions. The most macrofossil distribution.
permeable vertical pathways will have the However the occurrence of anaerobic methane
highest rates of CH4 transport from the methan- oxidation is not, in itself, sufficient to cause
ogenic zone and into the zone of methane concretionary growth. An idealized model of
oxidation, and these are likely to produce the anaerobic methane oxidation can demonstrate
highest rates of sulphate reduction and are the that a further crucial requirement is a pause in
most probable sites of carbonate precipitation sedimentation (or an abrupt decrease in sedi-
and concretionary growth. Such sites will usually mentation rates). For example, assume that the
be randomly distributed, as indeed are the Jet depth variation in the carbonate saturation index
Rock concentrations (Raiswell & White 1978). roughly mimics the depth variations in the rates
The presence of favourable nuclei (e.g. macro- of methane oxidation (Fig. 5), and that the
fossils) near a potential site of concretionary shallower pore waters are on the threshold of
growth may initiate carbonate precipitation at a carbonate precipitation. Thus further supersa-
somewhat lower degree of supersaturation, but turation resulting from methane oxidation is
siting is mainly controlled by the generation of represented by the area of the shaded triangle.
local carbonate supersaturation through en- Under continuous sedimentation the area swept
hanced rates of sulphate reduction (in turn by the triangle is always the same throughout the
determined by minor vertical permeability varia- sediment column and a diffuse phase of carbon-
tions which produce point-to-point variations in ate cementation would result. By contrast, any
the CH4 flux from below). Any fine-grained sedimentological break stabilizesthe potential
sediment containing skeletal debris should have zone of carbonate precipitation at a fixed
at least some grains capable of nucleating distance below the sediment/water interface.
carbonate precipitation at a potential concretion Concretion morphology (Fig. 5) would then be
site, thus no consistent relationship should be a function of:
50 R. Raiswell

~. ~ii~%~!~.~>~:~.,'L ]
li~ ~:~'~,~-+;~#~'-.~

r Time

(o) Sat.incl.

Saturation Index 999 ooo OO0 )

Profile

~ ooo eeo ------ i i

lID

r Time

Sat.lnd
(b) 9 9 9

S a t u r a tPi roonf i l eI n d e x ~ 9 .. ooo ooo .o.ool

9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9

i- , . , , 9 ~ 9
. . . . .
9 9149 91499 ~ OOOl
9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9
9 9 9 9 9 9 9 9 , , ~ 9 9 9 9

FIG. 5. Schematic representation of the progressive variations in concretion morphology caused by changes in
sedimentation rates with time, and carbonate saturation index with depth. A complete break in sedimentation
(a) fixes the zone of anaerobic methane oxidation (AMO), and hence carbonate supersaturation, a few tens of
centimetres below the sediment/water interface. The vertical extent of the concretionary horizon is then
determined by the thickness of the zone of AMO and coalescence is encouraged by longer breaks in
sedimentation and high saturation indices. During periods of reduced sedimentation (b), the zone of AMO
may instigate brief periods of growth at several closely spaced horizons, with the size and frequency of
concretionary growth reaching its maximum extent when sedimentation rates are at a minimum.

(a) Relative changes in sedimentation rates. A fluxes of methane and dissolved carbonate are
complete absence of deposition would give nearly equal, such that their net isotopic signa-
the minimal possible vertical range between ture is approximately that of sedimentary or-
the highest and lowest concretions in that ganic matter. The net signature of the residual
horizon, slow deposition would give a longer carbon pool then also retains the same value.
vertical range. However, if the diffusive fluxes differ, then the
(b) Time. Longer sedimentological breaks carbonate precipitated in the zone of anaerobic
would encourage the growth of large concre- methane oxidation will become either lighter (if
tions, and ultimately would tend to favour the methane flux is larger) or heavier (if the
the coalescence of adjacent concretions into carbonate flux is larger) than sediment organic
diagenetic limestone beds. matter. Furthermore the cements will then also
(c) Shape of the saturation index versus depth evolve with time. In fact the Jet Rock concre-
profile. Sharp, narrow profiles would favour tions have edges which are little more than 1-2%o
a limited vertical extent to the concretionary heavier than their centres (Coleman & Raiswell
bed and would tend to produce coalescence 1981), which implies that methane and dissolved
into thin limestones. A wider profile would carbonate fluxes were little different.
disperse concretionary growth over a thicker Clearly representative horizons of concretion-
sequence of sediments. ary growth will tend to arise where sedimenta-
tion is episodic, and the closer steady state
The non-steady state conditions arising from a conditions are approached the less likely are
depositional hiatus would, of course, change the concretionary phenomena. A model of this type
nature of the isotopic gradients in Fig. 2. In the can explain cyclic limestone-shale rhythms, as in
methanogenic zone, both methane and dissolved the Lower Lias of southern England (Hallam
carbonate evolve to heavier values with time. In 1964) and the Ordovician of Norway (Gluyas
this model it has been assumed that the diffusive 1984). The limestone horizons in these sequences
 The origin of concretions 51
are isotopically attributable to mixtures of Alternative diagenetic pathways to methane-
marine-derived and sulphate reduction carbon- derived or methanogenic concretions
ate (Campos & Hallam 1979, Gluyas 1984) and
they differ from the Jet Rock concretions only in The above conclusions, it should be emphasized,
two important respects; their morphology and apply in detail only to the early diagenetic
the smaller amounts of a pyrite phase specifically calcareous concretions with an isotopically light
associated with anaerobic methane oxidation. cement, as distinct from the later concretions
The differences in morphology can be explained discussed by Curtis et al. (1972), Irwin et al.
by a suitable choice of the variables illustrated in (1977), Pearson (1979) and Gautier (1982), where
Fig. 5, such that concretionary growth occurred the carbonate phase is often iron-rich (sideritic
in thin zones and was of sufficient persistence to or ankeritic) and invariably isotopically heavy.
result in coalescence of the individual growth However the same characteristics of non-steady
centres. Less later pyrite is probably the result of state sedimentation may also favour the forma-
low amounts of reactive iron within the sedi- tion of these methanogenic carbonates, where
ment, after the end of the initial phase of the carbonate saturation index instead develops
sulphate reduction (Raiswell 1982, Coleman & a deeper maximum in the methanogenic zone.
Raiswell in preparation). The following tentative model outlines how this

EDUCTIONNNNNNNNNNNN\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\~
SULPHATE R

Porewaters very o v e r s a t u r a t e d Porewaters only slightly o v e r -

with calcium carbonate, sediment s a t u r a t e d with calcium carbonate,
high in sulphide,low in reactive sediment tow in sulphide,high
iron n reactive iron

CALCIUM CARBONATE
t
IA,ka,,n,y IA'k'"n"Y
, PPTED AS CONCRETIONS liii- l NO CARBONATE [ NO CARBONATE '1
<sotopically Light) ~ PPTED PPTED
~:~::~:~::::::::::::::::::~:~:X:~:~:~.~:`~:.>3:~:~:~:~:~:~X::~.~:~:~`:~:~:~:~:~:~3.~
Alkalinity Retained Retained
Depleted
S\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\L~
M ET H A NO G ENE S I

Little iron reduction.

t t
Little iron reduction,
t
Iiron reduction
alkalinity low and but possibly enough to /important. alkalinity
decreasing,carbonates increase alkalinity to [increases
undersaturated ppt. calcium c a r b o n a t e /
t ~ ) >20 I(Ca)/(Fe)
/ <20

CALCIUM CARBONATE ' "t!:itl~

IPPTED AS CONCRETIONS fiiit ,;o. .c#S ,o.
s~itlop ic aII y heavy) I (Isotopically heavy) ~

FIG. 6. Alternative diagenetic pathways leading to the location of a zone of carbonate precipitation either in
the zone of anaerobic methane oxidation, or the zone of methanogenesis. The role of iron is crucial in
determining whether, and where, sufficient alkalinity can be developed to give carbonate supersaturation. The
criteria of Berner (1971) is used as a general indication of how Ca/Fe ratios affect carbonate mineralogy.
52 R. Raiswell
can occur, by showing alternative diagenetic tial. This diagenetic pathway is exemplified by
pathways which could lead to early, isotopically the Jet Rock concretions.
light CaCO 3 formed by methane oxidation and The alternative pathway is where sulphate
later, isotopically heavy carbonates formed by reduction is less intense (and anaerobic methane
methanogenesis (Fig. 6). The role of reactive iron oxidation weak or non-existent). Here insuffi-
is crucial (Coleman 1985) and two end-member cient alkalinity is generated to result in early
cases can be so identified depending on whether precipitation, and relatively large amounts of
or not reactive iron is substantially depleted prior reactive iron (low DOP values) survive entry into
to burial into the methanogenic zone. the methanogenic zone. Porewaters entering this
Both sulphate reduction and anaerobic meth- zone are perhaps over-saturated with respect to
ane oxidation generate alkalinity as HCO 3- CaCO3, but insufficiently so to induce precipita-
(equations 1 and 6) and hence produce tion. Concretionary growth can now occur if iron
porewaters which become oversaturated with reduction (see above) generates alkalinity suffi-
respect to CaCO3 (e.g. Berner etal. 1970, Murray cient to offset the CO2 production by methano-
et al. 1978). The greater the intensity of these genesis. Iron-rich carbonate will result, with the
processes, the more alkalinity is generated and iron content (either iron-rich calcites, siderites or
the more likely is CaCO3 precipitation. However ankerites) reflecting the relative proportions of
the concomitant generation of H2S also con- Ca 2+ and Fe 2+ in the porewaters (values in Fig. 6
sumes iron to form pyrite, such that little reactive from Berner 1971), and the occurrence of calcite
iron survives an intense phase of sulphate or dolomitic-type minerals reflecting the abun-
reduction and methane oxidation. The extent of dance in sulphate (Coleman 1985). This pathway
iron depletion can be estimated by the parameter is very probable in freshwater sediments, where
Degree of Pyritization (DOP) defined by Berner low sulphate concentrations act to limit the
(1971) as extent of pyrite formation, with the result that
Pyrite Fe proportionally more iron survives burial into the
DOP = methanogenic zone. However the same pathway
(Pyrite Fe + HCl-soluble Fe)" can result in marine sediments, where isotopical-
Calculation of DOP values assumes that all ly heavy siderite concretions (Gautier 1982) are
pyrite Fe was originally present in a form found because rapid burial limited the time spent
reactive towards H2S, also that HCl-soluble Fe in the sulphate reduction zone, and hence iron
would so react, given sufficent time. In the Jet was subsequently still available to generate a
Rock, DOP values in excess of 0.85 occur and methanogenic alkalinity maximum.
hence little reactive iron survived passage
through the zones of sulphate reduction which
have the pre-concretionary pyrite. If CaCO3 Conclusions
concretions are subsequently precipitated as a
result of methane oxidation, then the porewaters
entering the methanogenic zone will be substan- Morphological features of pyritiferous carbonate
tially depleted in alkalinity and Ca 2+, because concretions in the Jet Rock show that growth
had to occur during early diagenesis and was
Ca 2+ + 2HCO3-~CaCO 3 + CO 2 + H20. (7) confined to a thin, subsurface zone of uncom-
Furthermore methanogenesis generates CO2, pacted sediment, within a few metres of the
rather than alkalinity, hence the porewaters will sediment surface. These concretions contain no
approach undersaturation with respect to evidence of formation around local concentra-
CaCO3, unless a new alkalinity-generating tions of organic matter, although their carbon
mechanism occurs. This cannot be iron reduc- isotope composition ( - 1 3 to -16%o) clearly
tion (Coleman 1985) indicates a biogenic source. Crucial evidence as
to the nature of that source is revealed by the
3H20 + 2Fe203 + CH20 concretionary pyrite, which is both isotopically
~HCO 3- + 4Fe 2+ + 7OH- (8)
heavier than the host sediment pyrite ( - 2 4 to
since insufficient reactive iron is now available -26%0) and also texturally distinct. Thus an
to produce the alkalinity needed to offset early phase of pyrite formation occurred
methanogenic CO: production. Thus the very throughout the sediment prior to concretionary
intensity of the early sulphate reduction and growth and the concretion site was subsequently
methane oxidation phases of diagenesis needed distinguished from the surrounding sediments by
to produce CaCO 3 precipitation will, in general, the occurrence of locally intense, later phase of
preclude further carbonate precipitation by sulphate reduction and pyrite formation which
producing porewaters of low alkalinity and occurred in a pore system now depleted in
sediments with little alkalinity-generating poten- sulphate.
 The orig& of concretions 53
Studies of modern sediments show that anaer- explains the preferential development of concre-
obic methane oxidation tions along certain bedding planes). A complete
absence of sedimentation (or a sharp reduction in
CH4 + S O 4 2 - ~ H C O 3 - -{- HS- + H 2 0 the sedimentation rate) is necessary for anaero-
bic methane oxidation to localize carbonate
exhibits all these necessary features of a concre-
precipitation for long enough to produce a
tionary growth process. It occurs in a thin
concretionary horizon. Concretions represent an
subsurface zone of uncompacted sediment and
initial stage in carbonate precipitation, and their
stimulates a late phase of renewed sulphate
evolution into nodular limestones is related to the
reduction in pore systems which are sulphate
duration of the break in sedimentation, and the
depleted. Anaerobic methane oxidation also
rapidity of cementation. Similar non-steady state
generates alkalinity which could result in carbon-
sedimentation conditions may also cause the
ate precipitation. The carbon isotope signals of
origin of later, methanogenic concretions, where
cements precipitated by methane oxidation
a deeper maximum in carbonate supersaturation
would be indistinguishable from those of simple
is developed due to alkalinity derived from iron
sulphate reduction, provided the methane diffus-
reduction.
ing up into the oxidation zone is accompanied by
the methanogenic-derived dissolved carbonate.
ACKNOWLEDGMENTS: Charles Curtis, John Hudson,
Anaerobic methane oxidation produces car- Don Gautier and Joe Macquaker are thanked for their
bonate from a dissolved organic source (which contributions on a recent visit to the Jet Rock. Joe
leaves no fossil trace within the concretion) and Macquaker and John Gluyas also provided most
is a stratigraphically-confined process (which encouraging reviews.

References
BARNES, R. O. & GOLDBERG, E. D. 1976. Methane anoxic coastal marine sediments. Limnology and
production and consumption in anoxic marine Oceanography, 28, 1117-30.
sediments. Geology, 4, 297-300. CURTIS, C. D., PETROWSKI, C. & OERTEL, G. 1972.
BERNER, R. A. 1968. Calcium carbonate concretions Stable carbon isotopes ratios within carbonate
formed by the decomposition of organic matter. concretions: a clue to place and time of formation.
Science, 159, 195-7. Nature, 235, 98-100.
--1971. Principles of Chemical Sedimentology. DEVOL, A. H. 1983. Methane oxidation rates in the
McGraw-Hill, New York. anaerobic sediments of Saanich Inlet. Limnology
- - 1 9 8 0 . Early Diagenesis, a Theoretical Approach. and Oceanography, 28, 738-42.
Princeton University Press. - - & AHMED,S. I. 1981. Are higher rates of sulphate
- - , SCOTT,M. R. & THOMLINSON,C. 1970. Carbon- reduction associated with anaerobic methane
ate alkalinity in the pore waters of anoxic marine oxidation? Nature, 291, 407-8.
sediments. Limnology and Oceanography, 15, 544- - - , ANDERSON,J. J., KUIVILA,K. & MURRAY,J. W.
9. 1984. A model for coupled sulfate reduction and
CAMPOS, H. S. & HALLAM, A. 1979. Diagenesis of methane oxidation in the sediments of Saanich
English Lower Jurassic limestones as inferred Inlet. Geochimica et Cosmochimica Acta, 48, 993-
from oxygen and carbon isotope analysis. Earth 1004.
and Planetary Science Letters, 45, 23-31. DICKSON, J. A. D. & BARBER, C. 1976. Petrography,
CLAYPOOL,G. E. & KAPLAN,I. R. 1974. The origin and chemistry and origin of early diagenetic concre-
distribution of methane in marine sediments. In: tions in the Lower Carboniferous of the Isle of
KAPLAN, ]. R. (ed.) Natural Gases in Marine Man. Sedimentology, 23, 189-211.
Sediments, pp. 97-139. Plenum Press, New GAUTIER, O. L. 1982. Siderite concretions--indica-
York. tions of early diagenesis in the Gammon Shale
-- & KVENVOLDEN,K. A. 1983. Methane and other (Cretaceous). Journal of Sedimentary Petrology, 52,
hydrocarbon gases in marine sediment. Annual 859-71.
Reviews in Earth and Planetary Sciences, ll, 299- GLUYAS,J. G. 1984. Early carbonate diagenesis within
327. Phanerozoic shales and sandstones of the N.W.
COLEMAN, M. L. 1985. Geochemistry of diagenetic European Shelf. Clay Minerals, 19, 309-21.
non-silicate minerals: kinetic considerations. Phi- HALLAM, A. 1964. Origin of the limestone-shale
losophical Transactions of the Royal Society of rhythm in the Blue Lias of England: a composite
London, 315A, 39-56. theory. Journal of Geology, 72, 157-69.
--• RAISWELL, R. 1981. Carbon oxygen and HOWARTH,M. K. 1962. The Jet Rock Series and Alum
sulphur isotope variations in concretions from the Shale Series of the Yorkshire Coast. Proceedingsof
Upper Lias of N.E. England. Geochimica et the Yorkshire Geological Society, 33, 381-422.
Cosmochimica Acta, 45, 329-40. HUDSON, J. D. 1978. Concretions, isotopes, and the
CRILL, P. M. & MARTENS, C. S. 1983. Spatial and diagenetic history of the Oxford Clay (Jurassic) of
temporal fluctuations of methane production in central England. Sedimentology, 25, 339-70.
54 R. Raiswell
--& FRIEDMAN, I. 1978. Carbon and oxygen & WHITE, N. J. M. 1978. Spatial aspects of
isotopes in concretions: relationship to pore-water concretionary growth in the Upper Lias of N.E.
changes during diagenesis. In: CADEK,J. & PACES, England. Sedimentary Geology, 20, 291-300.
T. (eds). Proceedings, Symposium on Water-Rock REEBURGH, W. S. 1976. Methane consumption in
Interaction, Czechoslovakia, pp. 331-9. Cariaco Trench waters and sediments. Earth and
IRWIN, M., COLEMAN, M. L. & CURTIS, C. D. 1977. Planetary Science Letters, 28, 337~14.
Isotopic evidence for source of diagenetic carbon- --1980. Anaerobic methane oxidation rate depth
ates formed during burial of organic-rich sedi- distribution in Skan Bay sediments. Earth and
ments. Nature, 269, 209-13. Planetary Science Letters, 47, 345-52.
IVERSON, N. & JORGENSEN, B. B. 1985. Anaerobic --1983. Rates of biogeochemical processes in
methane oxidation rates at the sulfate-methane anoxic sediments. Annual Reviews in Earth and
transition in marine sediments from Kattegat and Planetary Sciences, 11, 269-98.
Skagerrak (Denmark). Limnology and Oceanogra- & HEGGIE, D. T. 1977. Microbial methane
phy, 30, 944-55. consumption reactions and their effect on meth-
MARTENS, C. S. & BERNER, R. A. 1977. Interstitial ane distributions in freshwater and marine envi-
water chemistry of anoxic Long Island Sound ronments. Limnology and Oceanography, 22, 1-9.
sediments. 1. Dissolved gases. Limnology and RICE, D. D. & CLAYPOOL, G. W. 1981. Generation,
Oceanography, 22, 10-25. accumulation and resource potential of biogenic
- - & KLUMP, J. V. 1984. Biogeochemical cycling in gas. American Association of Petroleum Geologists
an organic-rich coastal marine basin: an organic Bulletin, 65, 5-25.
carbon budget for sediments dominated by sul- SASS, E. & KOLODNY, Y. 1972. Stable isotopes,
phate reduction and methanogenesis. Geochimica chemistry and petrology of carbonate concretions
et Cosmochimica Acta, 48, 1987-2004. (Mishash Formation, Israel). Chemical Geology,
MURRAY, J. W., GRUNDMANIS,V. & SMETHIC,M S. JR. 10, 261-86.
1978. Interstitial water chemistry in the sediments SHOLKOVITZ, E. 1973. Interstitial water chemistry of
of Saanich Inlet. Geochmica et Cosmochimica Acta, the Santa Barbara Basin sediments. Geochimica et
42, 1011-26. Cosmochimica Acta, 37, 2043-73.
PEARSON, R. 1979. Geochemistry of the Hepworth WAAGE, K. M. 1964. Origin of repeated fossiliferous
Carboniferous sediment sequence and origin of concretion layers in the Fox Hills Formation.
the diagenetic iron minerals and concretions. Kansas Geological Survey Bulletin, 169, 541-63.
Geochimica et Cosmochimica Acta, 43, 927-41. WEEKS, L. G. 1957. Origin of carbonate concretions in
RAISWELL, R. 1971. The growth of Cambrian and shales, Magdalena Valley, Columbia. Geological
Liassic concretions. Sedimentology, 12, 147-71. Society of America Bulletin, 68, 95-102.
-- 1976. The microbiological formation of carbonate WHITICAR, M. J. & FABER, E. 1986. Methane
concretions in the Upper Lias of N.E. England. oxidation in sediment and water column environ-
Chemical Geology, 18, 227-44. ments--isotope evidence. Organic Geochemistry,
- - 1982. Pyrite texture, isotopic composition and the 10, 759-68.
availability of iron. American Journal of Science, ZANGERL, R., WOODLAND, B. G., RICHARDSON,E. S.
282, 1244-63. JR & ZACHRY, D. L. JR. 1969. Early diagenetic
--& BERNER, R. A. 1985. Pyrite formation in phenomena in the Fayetteville Black Shale (Mis-
euxinic and semi-euxinic sediments. American sissippian) of Arkansas. Sedimentary Geology, 3,
Journal of Science, 285, 71 0-24. 87-119.

R. RAISWELL, Department of Earth Sciences, University of Leeds, Leeds LS2 9JT, UK.
 The application of a solution-mineral equilibrium model to the
diagenesis of Carboniferous sandstones, Bothamsall oilfield, East
Midlands, England
E. A. Warren

SUMMARY: Within the fluviatile reservoir sandstones of the Bothamsall oilfield,

authigenic cements of early quartz, kaolinite and illite predominate. Detrital feldspars show
varying degrees of alteration and muscovite micas are commonly altered to intergrowths of
kaolinite and illite.
Solution-mineral equilibria in the idealized system of K20-A1203-SiO2-H20 are used to
relate the simple silicate mineral assemblage observed to possible pore-fluid compositions.
Previous studies considered aluminium to be immobile in this system. Petrographic evidence
of the paragenetic sequence at Bothamsall indicates that aluminium is mobile, albeit over
short distances. This study therefore considers aluminium mobility in the solution-mineral
equilibria. Graphical plots of total aluminium versus pH for given potassium ion activity, at
a temperature of 298 K, are constructed to illustrate the stability fields of the phases
identified. These show that aluminium can be an important component, together with pH
and potassium-ion activity, in affecting mineral stabilities in this system.
The existence of several co-stable phases at Bothamsall considerably reduces the possible
permutations for pore-fluid chemistry; kaolinite-illite alteration of mica proves particularly
sensitive in this respect. The relative merits of two different pore-fluid models are compared:
one, where the pore-fluid composition is considered to be constant in an open system, and
another where the pore-fluid composition is assumed to vary in response to in situ mineralogy
in a closed system. The former would require a very restricted range of pore-fluid
compositions to result in the diagenetic modifications observed, whereas a wide range of
pore fluids could evolve in a closed system driven by feldspar dissolution. This pore-fluid
evolutionary model is preferred.

Introduction England. The example taken concentrates on the

early part of the paragenetic sequence, that
One useful aim of the study of diagenesis is to primarily involving silicates, which was followed
relate the authigenic minerals observed in the by oil emplacement and dissolution. These
rock to the chemistry of the pore fluid. Mineral subsequent events indicate a change in pore-fluid
precipitation sequences can then be examined in chemistry, so modern formation water data
terms of the chemical composition and evolution (Downing & Howitt 1969) are unlikely to be
of the pore fluid. accurate analogues of the early pore fluid.
The use of thermodynamic phase diagrams in
the interpretation of mineral systems at low
temperatures has long been recognized. When Petrography
combined with formation water data they can
form a powerful tool in the elucidation of The Bothamsall oilfield reservoir rocks studied
diagenetic sequences (e.g. Kaiser 1984). How- comprise the Crawshaw and sub-Alton sand-
ever, where reliable chemical compositional stones (Westphalian A, Upper Carboniferous).
information for formation water is non-existent, These have been interpreted as representing
or known to be unrelated to the minerals delta-front and low-sinuousity fluviatile deposits
considered, interpretation of the pore-fluid sys- (Hawkins 1972, 1978, Guion 1971).
tem from which diagenetic minerals precipitated The results presented here form part of a
can be rather difficult. This paper describes the detailed petrographic and chemical study of 100
construction of a solution-mineral equilibrium samples from 10 wells within the oilfield. The
model relevant to diagenetic systems, in which interpretations are based on extensive examin-
all components are assumed to be mobile. It is ation of thin-sections, secondary and backscatter
used to analyse possible pore-fluid evolutionary electron imaging (SEI and BEI) of rock chips
pathways of a simple paragenetic sequence and polished thin-sections and hot- and cold-
described from the Bothamsall oilfield (National cathodoluminescence. Semi-quantitative and
Grid Reference SK6674), Nottinghamshire, quantitative chemical analyses of mineral phases

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences,

Geological Society Special Publication No. 36, pp. 55-69. 55
56 E. A. Warren

TIME AFTER BURIAL

QUARTZ " ~ ?'- ~[?-_ ~ [] ~ '

KAOLINITE
ILLITE ~__~T___~
SIDERITE [] []
ANKERITE ~ []
FERROAN CALCITE []
PYRITE, ANATASE []
OIL [~-~
FELDSPAR
MUSCOVITE

PRECIPITATION ~ DISSOLUTION

FIG. 1. A diagenetic sequence for the Bothamsall oilfield. The relative timing of precipitation and dissolution
of the more important diagenetic phases is shown. Note the sharp division between silicate precipitation
(early), and subsequent precipitation of carbonates and oil emplacement. Two possible phases of quartz
overgrowths and illite precipitation relative to kaolinite are indicated.

FIG. 2. Muscovite mica. Alteration of detrital muscovite mica to laths of illite (I, light) and kaolinite (K,
dark). Authigenic kaolinite and illite infill remanent porosity. Note only kaolinite infills the adjacent, but non-
connecting pore. This may indicate the existence of different pore-fluid chemical compositions in adjacent
pores resulting in different cement sequences. Illite formation may be controlled by mica dissolution. Back-
scatter SEM micrograph (Oxford Polytechnic). Scale bar = 100 ~tm.
 Model application to sandstone diagenesis 57

were obtained using Energy Dispersive Spectral companied, or preceded, by substantial dissolu-
(EDS) X-ray analysis systems on scanning tion of both quartz and feldspar creating secon-
electron microscopes and Wavelength Disper- dary porosity. Although no phyllosilicate
sive Spectral (WDS) X-ray analysis on a micro- dissolution was recognized in this event, no
probe. The mineralogy and chemical compo- subsequent precipitation of silicates has been
sition of clay minerals were studied using an observed. In addition several other minerals,
Analytical Transmission Electron Microscope most notably siderite, pyrite and anatase were
(ATEM) in ultra-thin (c. 300 nm) sections, and observed but found to be of only local volumetric
powder X-ray diffractometry (XRD) of fine significance.
separates from whole rock samples. Potassium feldspar grains show varying de-
The sandstone samples examined range from grees of alteration and dissolution; partially
fine- to coarse-grained arkoses to quartz-arenites. altered feldspar grains are often associated with
Present porosity varies significantly both kaolinite and/or illite. The presence of abundant
between, and within, samples, but can be as high overgrowths tends to obscure grain textures, but
as 15% The main detrital phases are quartz, generally angular grains with point contacts were
alkali feldspar and muscovite. identified, indicating minimal pressure solution
The most abundant authigenic phases identi- prior to quartz cementation. Muscovite micas
fied are quartz overgrowths, kaolinite, illite and are frequently altered to complex intergrowths of
zoned ankerite. All authigenic phases are distrib- kaolinite and illite (Figs 2 and 3) (cf. Huggett
uted throughout the reservoir; none appears to 1984). Kaolinite appears to have grown by the
be controlled by the presence of unstable detrital displacement of the mica parallel to cleavage
phases, such as feldspar or muscovite. A general (Fig. 3).
diagenetic sequence of these minerals (Fig. 1) The precipitation sequences within the sili-
involves early silicates, followed either by oil cates are complex. Kaolinite commonly abuts
emplacement or, in the water zone, ankerite (cf. (Fig. 4), or has been partially enclosed by (Figs 6
Kantorowicz 1985). This event was ac- and 7), quartz overgrowths. Quartz thus both

FIG. 3. Detail of Fig. 2. Inset of mica intergrowth and authigenic cements. Illite lines mica and quartz grains.
Authigenic kaolinite rouleaux appear to have displaced earlier illite resulting in the 'mesh-work' texture seen.
Back-scatter SEM micrograph (Oxford Polytechnic). Scale bar = 10 p.m.
58 E. A. Warren

FIG. 4. Authigenic kaolinite and quartz.The quartz overgrowth partially encloses some kaolinite booklets,
whilst other booklets lie upon the overgrowth. This is interpreted as either two generations of quartz, with
kaolinite precipitation in between, or as continual precipitation of quartz, before and during kaolinite
formation. Lack of signs of dissolution (e.g. pitting) suggests that both phases were stable, i.e. fluid was
saturated with respect to both. SEM. Scale bar = 10 ~tm.

FIG. 5. Authigenic illite on kaolinite. Kaolinite rouleaux are coated by later illite of both platy and hairy
morphologies. No evidence of dissolution indicates that kaolinite was stable when illite precipitated. SEM.
Scale bar = 10 ~m.
 Model application to sandstone diagenesis 59

FIC. 6. Illite and kaolinite on quartz overgrowth. Note the hairy habit of illite. No evidence of dissolution is
apparent, which would indicate that the accompanying solution was saturated with respect to all three phases.
SEM. Scale bar = l0 pm.

FIG. 7. Ultra-thin section of authigenic silicates. Kaolinite plates (K) abutting, and partially enclosed by,
quartz overgrowth (Q). Later pore-filling illite (I) is delicately attached to kaolinite (arrowed). A solution
saturated in quartz, then quartz plus kaolinite, then quartz plus kaolinite plus illite (i.e. three co-stable phases)
is interpreted. TEM micrograph. Scale bar = 500 nm.
60 E. A. Warren

preceded and followed the start of kaolinite resulted in the precipitation of the silicates by
precipitation. This could be interpreted as either considering a chemical model based on the
two distinct generations of quartz precipitation, mineral solubility equilibria of the phases in-
or as a period during which two silicates, volved. The distribution of authigenic alumino-
kaolinite and quartz, precipitated. Cathodolu- silicates within the reservoir also indicates that
minescence did not resolve this aspect. Neither aluminium should be treated as a mobile species.
phase shows evidence of dissolution. This could
imply that the pore-fluid present was at least
A thermodynamic model
continually saturated with respect to both kaolin-
ite and quartz, although kinetic factors, such as The Gibb's phase rule considers phase, compo-
dissolution rate, may be important. Illite has nent and resulting degrees of freedom (or
been interpreted as both preceding (Fig. 3) and variance), F, within any given system through
post-dating kaolinite (Figs 5 and 7) in different the relationship:
areas of the oilfield. It is often intimately
P+F=C+2.
associated with kaolinite, to which delicate
attachments have been observed (Fig. 7), indi- Studies of real diagenetic systems (e.g. Kaiser
cating illite post-dating kaolinite. Elsewhere, 1984) often suffer from the problem that the
illite replacement of muscovite is often followed number of components greatly exceeds the
by the precipitation of pore-lining illite, which is number of phases present, thus resulting in non-
displaced by later pore-filling kaolinite (Fig. 3). unique solutions for the variance. This is true for
This indicates that supersaturation of the pore the Bothamsall sandstones, for which a rigorous
fluid with respect to illite and kaolinite occurred treatment must consider 10 or more components
at different times in different parts of the but far fewer phases. To overcome this problem
reservoir. Lack of dissolution textures (e.g. etch the number of components must be reduced in
pits), on quartz grains and overgrowths (Fig. 6) relation to the number of phases present. In this
indicates that the solution precipitating illite was case study only the early part of the diagenetic
also saturated with respect to quartz. sequence was considered--that restricted to
A very different relationship is observed silicate diagenesis. Furthermore, the considera-
between the authigenic silicates and ankerite. tion of'ideal' stoichiometric compositions for the
Authigenic ankerite rhombs commonly appear phases enabled the components to be limited to
to displace clay minerals, and quartz over- four: K20, A1203, SiO2, H20. This assumption
growths are often corroded to the grain. Pore would appear to be reasonable from published
fluids precipitating ankerite thus must have been analyses of quartz, kaolinite, muscovite and K-
undersaturated with respect to quartz. feldspar (e.g. Deer et al. 1962), but less so for
A pore-fluid history can be constructed by illite. A composition of K1.sA15.sSi6.sO2o(OH)4
considering the assemblages of authigenic was used for illite, which approximates to the
phases and their textures. The initial pore water average analysis determined directly for the
was probably a Carboniferous meteoric water, in Bothamsall material using Analytical T E M - -
accordance with the depositional interpretation. K i.2Al4.aMgo.2Feo.2Si6.sO2o(OH)4 (details to be
This was perhaps quartz saturated, but with published separately).
burial became supersaturated, resulting in quartz A variety of methods for modelling diagenetic
precipitation as grain overgrowths. Progressive systems has appeared in the literature. Experi-
supersaturation in kaolinite and illite followed mental investigations of solution-mineral equili-
during which the pore fluid remained saturated, bria (e.g. Surdam et al. 1984) have the obvious
or supersaturated, in quartz. The same fluid was advantage of enabling the sensitivity of phases to
also undersaturated with respect to K-feldspar components to be determined directly, but suffer
and muscovite, resulting in their dissolution or from problems of kinetics and the metastability
replacement by illite and/or kaolinite. Subse- of intermediate phases (May et al. 1986).
quently, the pore fluid became undersaturated Computer programs of solution models based on
with respect to quartz, resulting in dissolution of thermodynamics, e.g. MINEQL (Westall et aL
both grain and overgrowths. This was ac- 1976) and PHREEQE (Parkhurst et al. 1980),
companied by the emplacement of oil or the provide an elegant method of predicting fluid
precipitation of ankerite. Large-scale fluid evolution but require an initial estimate of the
movement must have occurred, as oil generation pore fluid composition. In the case of Bothamsall
within the reservoir is unlikely. This indicates this presents difficulties for it is unlikely that the
that the present-day pore fluid is of a different present-day formation water (data in Downing &
chemical composition to the pore fluids preced- Howitt 1969) should reflect the physicochemical
ing oil emplacement. This study attempts to composition of the possible Carboniferous wa-
examine the early pore-fluid chemistry which ters from which the authigenic silicate phases
 Model application to sandstone diagenesis 6I

\\
o-\
\
\
\
\
\
\
\
\
\
\
._1
< \
\ /
/
0
oT -5 \ /
0
._1
\ /
\ /
\ /
KAOLINITE AND SOLUTION /
/
\ /
/
\ /
\ /
\ /
SOLUTION ONLY N /
\ /

T -- 2 9 8 K
-lO -
a K. = 0 mol I -~

aH48i04-- 1 . 0 7 x 1 0 -4 mol I -I

- I I [ I
4 6 8 10

pH
FIG. 8. Solution-mineral equilibria in the system: Al203-SiO~-H20, T= 298 K, P = 1 bar, quartz-saturated
(after Curtis 1983). Activity of potassium ion is zero. Kaolinite will precipitate from any solution whose
chemical composition lies within the saturation curve. All solutions should precipitate quartz (quartz-
saturated). N.B. a(Al)v is the sum of the activities of the individual aluminium species calculated from the
free energy data of the solution-mineral equilibria.

precipitated, when subsequent diagenetic events were not applicable to the composition consi-
(e.g. dissolution and ankerite formation) indicate dered. Instead, a model (Tardy & Garrels 1974)
a change in fluid chemistry. was used to derive an empirical value for the
The approach used here follows that described Gibb's free energy of formation of
in Garrels & Christ (1965) and Kaiser (1984). - 1320 kcal mo1-1. For each solution-mineral
Solution-mineral equilibria for the phases and reaction the free energy for reaction, and hence
compositions considered were determined from the equilibrium constant, was calculated by
basic thermodynamic principles (Atkins 1978) summation of the free energy of formation data
and plotted in the system K20-A1203-SiO2-H20. of the species considered.
Thermodynamic data for the species were taken In the treatment of similar systems previous
from Robie et al. (1978). A problem existed in the authors (Garrels & Christ 1965, Kaiser 1984)
case of illite for available thermodynamic data have considered the aluminium species to be
 62 E. A. Warren
~ KAOLINITE
~ ILLITE
-'~ K-FELDSPAR

~ KAOLINITE + ILLITE

~ ILLITE + K - S P A R

~ MUSCOVITE + ILLITE

] MUSCOVITE * ILLITE + K - S P A R

1-~ MUSCOVITE + ILLITE + KAOLINITE

ILLITE

----- J KAOLINITE
...
/ ./" MUSCOVITE
A--.II.-- ".... ......... K - F E L D S P A R

o
5
O
J
"" "' ALL MINERALS
"". AND SOLUTION //~.~'~-~

SOLUTION ONLY

T:298K
-10
aK.:4x 10 -5mol I -~
aH4Si04-- 1.07 x 1 0 4 m o I I -I

1 I
4 6 8 10
pH
FIGS 9-12. Solution-mineral equilibria in the system: K,O-A1203-SiOz-H20 , T= 298 K, P = 1 bar, quartz-
saturated, and increasing potassium ion activity. Individual mineral saturation curves (lines) and stability
fields (shaded areas) are indicated. Note sensitivity of K-feldspar field with K § See text for discussion. N.B.
a(Al)x is the sum of the activities of the individual aluminium species calculated from the free energy data of
the solution-mineral equilibria.

effectively immobile. This has the advantage of Aluminium, together with potassium ions and
eliminating aluminium, so reducing the number pH, have thus been considered as independent
of components. It also means that the only components in this system. This presents prob-
charged species are potassium ions, hydrogen lems representing the system graphically as a
ions and hydroxyl; as charge must be conserved single dimension for each of the four components
the potassium activity and pH are thus interde- is required. The presence of several phases co-
pendent. This effectively reduces the compon- stable with quartz enabled the solution-mineral
ents one further and enables a four oxide system equilibria to be considered at quartz saturation.
to be plotted on a two-dimensional diagram (e.g. Several alternative graphical plots were possible.
K+/H + versus H4SiO4). A plot of total aluminium versus pH for different
At Bothamsall, however, the presence of activities of potassium ion typical of analysed
authigenic pore-filling kaolinite and illite distrib- porewaters (e.g. Downing & Howitt 1969,
uted throughout the reservoir suggests that Stumm & Morgan 1981) has been considered
aluminium is mobile, if only over short distances. here in Figs 8-12.
 Model application to sandstone diagenesis 63
~ KAOLINITE

~ ILLITE

: ~ " ~ K- FE LDS PAR

I:. ~1 KAOLINITE + ILLITE

[ ~ ILLITE ~- K-SPAR

~ MUSCOVITE + ILLITE

] MUSCOVITE § ILLITE + K-SPAR

]-~ MUSCOVITE § ILLITE + KAOLINITE

~-.... ILLITE
,.......
I- KAOLINITE
._,.--
i" MUSCOVITE
A
I-
--I .......... K-FELDSPAR
<
v
o
-5-
o7
0
._1

l
""..
"",.
r...
ALL MINERALS
AND SOLUTION

["1..

: :/
T =298K
-10
a K . = 4 x ] O -4 mol I-~
aH4Si04:1.07 • 10 -4 m01 I-I
I I
4 6 8 10

pH
FIG. 10.

One important component not considered is Discussion

temperature. The temperature increase associat-
ed with burial can affect mineral stability. The Amphoteric aluminium
solubility of many phases increases with tem-
perature which could alter the juxtaposition of Aluminium forms many ionic species in solution
the solubility curves in Figs 8-13. Unfortunately, of which three are dominant: A13+, AI(OH)2 +,
modelling the pore-fluid system considered here AI(OH)4-. These are responsible for the ampho-
at elevated temperatures (e.g. 100~ has been teric nature of aluminium in solution; aluminium
hampered by the sparse data base available for being soluble at both high and low pH but
many species, aluminium in particular (Hovey & markedly insoluble at neutral and near neutral
Tremaine 1986). The diagrams presented here pH. A graphical plot of total aluminium species
therefore only indicate the relative stability of versus pH thus results in a curve (Fig. 8). This
minerals and solution at 25~ They cannot be effect gives rise to the curve for kaolinite (Curtis
used to obtain absolute values for pore-fluid 1983), and also illite, muscovite and K-feldspar.
compositions. Values cited in the discussion are At high pH the solubility of all silicates is
merely used to highlight the areas of the increased further, due to the formation of
diagrams being referred to. charged ionic species of silica.
64 E. A. Warren

KAOLINITF

ILLITE

K-FELDSPAR

KAOLINITE § ILLITE

ILLITE § K - S P A R

MUSCOVITE + ILLITE
-..
MUSCOVITE + ILLITE + K-SPAR

MUSCOVITE + ILLITE + KAOLINITE

J ILLITE

KAOLINITE

MUSCOVITE
A
K-FELDSPAR

o
-5 "'....
0

ALL M I N E R A L S
".... AND S O L U T I O N

I'...

S O L U T I O N ONLY
J""i.

..... j
T = 298K \
-10
aK+=4x 10 -3 mol I -I
aH4Si04 = 1.07 x 10 -4 mol I -I

I I I
4 6 10

FIG. 11. pH

Phases are only stable in solutions whose Mineral precipitation, of kaolinite for exam-
compositions lie within the solubility boundary, ple, would be favoured in neutral pH solutions
or saturation curves derived from the solution- where solubility is lowest (Fig. 8). Conversely,
mineral equilibria. If a pore fluid lies outside the mineral dissolution would be most favourable in
solubility boundary of a phase, then that phase solutions of either high or low pH. A subtle
should dissolve as the solution is undersaturated. change in pH could thus drastically affect the
Similarly a mineral should precipitate only if the stability of the minerals, causing either dissolu-
pore-fluid composition lies within the solubility tion of pre-existing phases or precipitation of
boundary of that phase--the solution is then existing and additional phases, thus resulting in a
supersaturated in that phase. Kinetic factors, change in the bulk assemblage.
such as nucleation and growth, increase the
degree of supersaturation required from the
Potassium ion activity
phase diagram for precipitation to commence.
Dissolution is affected by activation energies in a The solubility of the potassium-bearing phases is
similar way. Thus the solution-mineral equilibria also sensitive to variations of potassium activity
only indicate where a particular phase is thermo- in solution (Figs 9-12). None of these phases is
dynamically stable. Kinetics will determine stable in a solution of zero K + activity, leaving
whether it actually dissolves or precipitates. kaolinite and quartz as the only stable phases
 Model application to sandstone diagenesis 65
" ~ KAOLINITE
r.... ,LL,TE
9 ~ K-FELDSPAR
" ~ KAOLINITE § ILLITE
~] ILLITE + K-SPAR

i ". ~ MUSCOVITE * ILLITE

A
I--
ii,il
-

i~
:

'"
-"
MUSCOVITE + ILLITE + K-SPAR
MUSOOV'T

---,--"'" KAOLINITE
.----9"-" MUSCOVITE
.,'',T +

--I i 'i" .......... K-FELDSPAR

". .4 ~
o
oT 5
O
,_1
"" ALL MINERALS
",, AND SOLUTION

SOLUTION ONLY

T -- 2 9 8 K ~::.:i~
-10
aK. = 4 x 10 -2 mol I-I
aH4Si04 ~ 1.07 x 1 0 4 mol I -~

I I L I
4 6 8 10

pH
F I G . 12.

(Fig. 8). With increasing potassium ion activity dered. Kaolinite is the most stable phase at very
in a neutral pH solution, illite becomes the most low potassium ion activity (Fig. 8), but with
stable authigenic phase. Illite and, to a lesser increasing amounts of potassium other phases
extent, muscovite only show a limited sensitivity become more stable. The behaviour of muscovite
to potassium-ion activity. K-feldspar, however, is similar to that of illite, but under none of the
is particularly sensitive; from being the most considered conditions is muscovite more stable
soluble phase at low activity (Fig. 9), its than illite. K-feldspar shows the most sensitivity
precipitation is favoured at high potassium ion to both pH and potassium ion activity. It appears
activity (Fig. 12). to be very soluble at low pH and low potassium
ion activity, with solubility markedly decreasing
as each increases.
Relative stability of minerals
The various solution-mineral equilibria illustrat-
ed (Figs 8-12) indicate that illite is the least The application to the Bothamsall sandstones
soluble, or most stable phase, over the widest The diagenetic history of the Bothamsall sand-
range of pH and potassium ion activity consi- stones can now be investigated in terms of the
66 E. A. Warren

solution-mineral equilibria of the phases. The the fluid may have become muscovite saturated.
preservation of point contacts and the lack of Thus the pore fluid may have approached
dissolution of quartz grains imply that initially a equilibrium with muscovite.
pore fluid saturated with respect to quartz
entered the sandstone. Quartz overgrowths fol-
lowed, indicating availability of supersaturated
K-feldspar
solutions. Detrital K-feldspars showed varying degrees of
Two different precipitation sequences are dissolution and no evidence was seen of K-
observed at Bothamsall: one in which illite post- feldspar overgrowths. This would indicate that
dates kaolinite (case 1, Figs 5, 6 and 7), the other such grains were in contact with a solution
where kaolinite post-dates earlier illite (case 2, undersaturated with respect to feldspar (barring
Fig. 3). Both kaolinite and illite precipitation kinetic factors). Feldspar, as already discussed, is
post-date the onset of quartz cementation. This very sensitive to potassium ion activity relative
indicates that the initial solution was probably to the other phases. Low potassium ion activity
undersaturated with respect to both kaolinite could cause feldspar dissolution (Fig. 9). Alterna-
and illite, perhaps because of an initial low tively, low pH solutions would also be undersa-
aluminium and potassium ion activity. turated irrespective of potassium activity, and
Where illite post-dates kaolinite (case 1), neutral pH solutions likewise if aluminium
aluminium activity may have increased from its activity was low (<10-7mo11-1). High pH
initial value, so that the kaolinite solution solutions, unless potassium activity was low,
boundary was crossed (see Fig. 8). Alternatively, would be supersaturated, and feldspar dissolu-
pH could have increased in a solution of tion would not occur (Figs 11 and 12).
moderate aluminium content, but low potassium
ion activity, again entering the kaolinite field
Summary
(Fig. 8). Illite saturation may then have been
reached in either situation by an increase in K § The precipitation of quartz, kaolinite and illite
activity in the solution. would be most favourable, in the simple example
Considering case 2, potassium ion activity considered, in solutions of low to neutral pH,
must have been sufficiently high with increasing or fixed aluminium and potas-
(> 10-6 mol 1-1) for illite to precipitate before sium activity. Feldspar dissolution could occur in
kaolinite. Kaolinite could then precipitate if the low pH or low potassium activity. Muscovite
solution entered the stability field for kaolinite alteration to illite and kaolinite could only occur
and illite (Figs 9-12). An increase in either pH (if in a restricted range of solution compositions.
initially low) or aluminium ion activity would The possible consideration of the mineral-
cause this to occur. solution equilibria has restricted the possible
The relative timing and magnitude of changes pore-fluid compositions from which the diagene-
in pH, aluminium and potassium ion activity tic sequences evolved. However, it is necessary to
could thus have a dramatic effect on the consider whether the pore-fluid chemistry itself
authigenic assemblage and account for the changed or whether the diagenetic sequence
variations observed (Fig. 1). Conversely, a could have resulted from a pore fluid of more or
particular observed assemblage could result from less constant chemical composition.
a number of different pore-fluid evolution
pathways.
Open system
Muscovite alteration In an open system the chemical composition of
The alteration of detrital muscovite grains to the pore fluid is controlled by a source external to
intergrowths of kaolinite and illite was frequent- the pore considered. The reactants are assumed
ly observed. The pore fluid in contact with the to be unlimited in supply, and products removed
muscovite must have been undersaturated with from the system, so that the resultant fluid
respect to muscovite, but saturated in both composition remains unchanged. The petrogra-
kaolinite and illite. Such solutions are restricted phic evidence of muscovite alteration and feld-
in pH and aluminium and potassium ion space spar dissolution indicates that the accompanying
(Figs 9-12). Aluminium activity must have been fluid lay outside the solution-saturation curves
sufficiently high (> 10-9 mol 1-1) for both to for these phases but within those of kaolinite and
precipitate, otherwise illite would be the only illite, the precipitated phases. Such compositions
stable phase. Potassium ion activity in excess of lie within a restricted region of low pH and
10-6 mol 1-1 must also be assumed for illite to potassium ion activity for the aluminium activi-
precipitate. The presence in many samples of ties commonly found in solutions (10 -1~ to
incomplete alteration suggests, moreover, that 10-8 mol 1-1) (Figs 9-12, 13).
 Model application to sandstone diagenesis 67

KAOLINITE

~ ILLITE

] K-FELDSPAR

~ KAOLINITE+ ILLITE

~.~ ILLITE + K-SPAR

~ MUSCOVITE * ILLITE

] MUSCOVITE * ILLITE -" K-SPAR

'_.
]~} MUSCOVITE ", ILLITE + KAOLINITE

".. ILLITE
,t"1 KAOLINITE

.----9""" MUSCOVITE
A I.-
.......... K-FELDSPAR
'" ....
o
~" --5
0
-.I

~ 1 1 1 I -.L--L/. . '.+ .. "

T-- 2 9 8 K .-::
-10
a K. = 4 x 10 -5 mol I-I
aH4Si04 = 1 . 0 7 x 1 0 4 m o I I -I

I 1 I I
4 6 8 10

pH
FIG. 13. Possible pore-fluid evolutionary pathways of a solution of initial composition (x) in a closed system.
Precipitation of illite starts at (1), kaolinite at (2). Muscovite alteration ceases at (3), fluid continues to evolve,
driven by K-feldspar alteration, to (y), where the solution is saturated with respect to all phases. The only
possible open system solutions would have a chemical composition within the kaolinite + illite shaded area.

The variations in the diagenetic sequence of Invoking the former merely removes the prob-
different parts of the reservoir, such as the lem of the cause elsewhere, and there is little
relative timing of kaolinite and illite, must substantial knowledge of the processes and
indicate that pore-fluid chemistry was not homo- kinetics of clay mineral precipitation. The
geneous within the reservoir. This could be diagenetic features observed at Bothamsall could
explained by the existence of several small cells have developed in an open system, but only from
behaving as isolated open systems within the a very restricted range of fluid compositions.
reservoir. Several problems exist, however, in the
consideration of open systems. First, the con-
straints on the composition of the solution
Closed system
require a source of hydrogen and potassium ions.
The breakdown of muscovite and K-feldspar are A closed system assumes no gain or loss of
possible sources. Second, progressive sequences material from the volume considered (it is
could only be produced via a change in fluid isochemical). This means that the fluid compo-
composition, or by a governing kinetic process. sition must evolve in response to the formation or
68 E. A. Warren

destruction of the solid phases. A fluid of A closed system could account for the cement
variable chemical composition could thus cross sequences observed and some of the spatial and
several mineral solubility boundaries as it re- textural evidence. The closed system argument
sponds to the ions produced from an unstable has the advantage that a range of pore-fluid
phase. A possible pore fluid evolution applicable chemistries could evolve to give the diagenetic
to Bothamsall is considered in Fig. 13. A solution features described. Initially these fluids would
of low pH and low potassium ion activity (x) have low pH, aluminium and potassium ion
would cause feldspar and muscovite to dissolve: activities; evolution would be driven by feldspar
and muscovite alteration which would cease on
KA13Si3010(OH)2 + 10H § saturation.
= K § + 3AI 3§ + 3H4SIO4
KAISi308 + 4H20 + 4H §
--- K + + A13+ + 3H4SIO4. Conclusions
Potassium, aluminium and silica would be Many common aluminosilicate minerals in silici-
released to the fluid and hydrogen ions con- clastic rocks show marked sensitivity not only to
sumed. Thus pH, aluminium, potassium and pH and potassium ion activity, but also to
silica activities would increase. The fluid compo- aluminium. This is due to the amphoteric
sition would therefore evolve until saturated with behaviour of aluminium in solution. The consi-
both muscovite and feldspar. When silica activ- deration of aluminium mobility, together with all
ity reaches quartz saturation, precipitation of other species, is therefore important when
quartz should occur. The increase in aluminium modelling the chemical evolution of pore fluids
and potassium ions in solution would cause in diagenesis. Graphical plots of aluminium
aluminosilicates to precipitate. In the situation versus pH and potassium ion activity, for various
considered in Fig. 13, illite precipitates first (1) silicates, have been presented for 298 K and
due to the potassium ion activity (as in case 2 illustrate the sensitivity these minerals have to all
previously considered). Kaolinite would precipi- these components. Whilst constructions at ele-
tate when the fluid composition evolves to the vated temperatures might be more appropriate,
kaolinite saturation curve (2). Precipitation of these diagrams do demonstrate the importance
both kaolinite and illite could then occur. At (3) of aluminium to mineral stability.
dissolution of muscovite would cease. Only K- A simple case example from the Bothamsall
feldspar would dissolve until the solution be- oilfield has been taken to illustrate the applica-
comes saturated (y). Alternatively, an increase in tion of the diagrams to diagenetic situations. The
aluminium activity, while potassium ion activity precipitation of the authigenic minerals (quartz,
remained low, would cause kaolinite to precipi- kaolinite and illite) and feldspar dissolution
tate before illite (case 1), the other diagenetic could result from a wide range of fluid composi-
sequence observed. No further change in pore tions. The alteration of muscovite mica to
fluid would occur as equilibrium has been kaolinite and illite is predicted to occur in
reached. chemically restricted fluid compositions. The
Alteration and dissolution of the grains would possible evolutionary pathways do not necessar-
continue in a closed system during precipitation ily preclude either an open or a closed system.
of other phases, because precipitation would Only a restricted range of pore-fluid composi-
consume the ions produced from dissolution and tions could have resulted, however, in the
so buffer the solution composition. Dissolution observed sequence from large-scale fluid move-
would stop only if precipitation of more stable ment in an open system. Chemical evolution of a
phases was hindered in some way, kinetic factors silica-saturated, low aluminium and potassium-
(nucleation, growth) perhaps. Alternatively a ion pore water in a closed system controlled by
physical factor, volume, may be important. If the feldspar and muscovite dissolution, would not
pore space is effectively filled by a phase, only result in the progressive precipitation of the
kaolinite for example, no more can precipitate. quartz, kaolinite and illite, but also approach
Feldspar or muscovite grains associated with equilibrium with the unstable phases. Thus
pores entirely filled by authigenic cement might muscovite and feldspar dissolution would cease,
thus suffer only partial replacement, whilst those leaving some grains only partially altered. Only
in pores that remained open after cementation subtle variations in initial pore-fluid composition
might suffer total obliteration. However, this are needed to generate differences in the relative
treatment clearly overlooks the problem of mass timing of precipitates in different areas of the
balance. This is particularly pertinent to closed reservoir. Thus evolution of a meteoric water in a
systems where chemicals are remobilized rather closed system driven by feldspar and, to a lesser
than introduced or removed in an open system. extent, muscovite alteration, seems to be the
 Model application to sandstone diagenesis 69
more likely mechanism for generating the early ACKNOWLEDGMENTS: Support for this research has
diagenetic modification of the sandstones. been generously provided by a British Petroleum
The relative importance of temperature, press- Scholarship. My supervisors, Charles D. Curtis (Shef-
ure and kinetics on the diagenetic assemblage field) and Jenny Huggett (BP) are thanked for many
remain to be evaluated. However, the inclusion hours of fruitful discussion and, together with Tim
Young, for reading through the manuscript. My thanks
of aluminium mobility in solution-mineral mod-
are due to Anton Kearsley and Oxford Polytechnic for
elling does appear to be significant in the the use of the Back-scatter Electron Microscope
interpretation of diagenetic histories in terms of facility, and to Mike Cooper for drafting the diagrams.
pore-fluid chemistry.

References
ATKINS, P. W. 1978. Physical Chemistry. Oxford R. C. (eds). Clastic Diagenesis. American Associ-
University Press. ation of Petroleum Geologists Memoir No. 37,
CURTIS, C. D. 1983. Link between aluminium mobility 195-216.
and destruction of secondary porosity. Bulletin of KANIOROWlCZ, J. D. 1985. The origin of authigenic
the American Association of Petroleum Geologists, ankerite from the Ninian Field, U.K. North Sea.
67, 380-95. Nature, 315, 214-6.
DEER, W. A., HOWlE, R. A. & ZUSSMAN,J. 1966. An MAY, H. M., KINNIBURGH,D. G., HELMKE, P. A. &
Introduction to the Rock-forming Minerals. Long- JACKSON,M. L. 1986. Aqueous dissolution, solubi-
man, London. lities and thermodynamic stabilities of common
DOWNING, R. A. & HOWITT, F. 1969. Saline ground- aluminosilicate clay minerals: kaolinite and smec-
waters in the Carboniferous rocks of the English tites. Geochimica et Cosmochimica Acta, 50, 1667-
East Midlands in relation to the geology. Quarterly 77.
Journal of Engineering Geology, 1, 241-69. PARKHURST, D. L., THORSTENSON, n . C. & PLUMMER,
GARRELS, R. M. & CHRIST, C. L. 1965. Solutions, L. N. 1980. PHREEQE--a computer program for
Minerals and Equilibria. Harper & Row, New geochemical calculations. U.S. Geological Survey,
York. Water Resources Invest. 80-96, NTIS Technical
GUION, P. D. 1971. A sedimentological study of the Report PB81-167801, Springfield, VA.
Crawshaw sandstone (Westphalian A) in the East ROBIE, R. A., HEMINGWAY,B. S. & FISHER, J. R. 1978.
Midlands coalfield area. Unpublished MSc Thesis, Thermodynamic properties of minerals and relat-
University of Keele. ed substances at 298.15 K and 1 bar 005 pascals)
HAWKINS, P. J. 1972. Carboniferous sandstone oil pressure and at higher temperatures. U.S. Geologi-
reservoir, East Midlands, England. Unpublished cal Survey Bulletin No. 1452.
PhD Thesis, University of London. STUMM,W. & MORGAN,J. J. 1981. Aquatic Chemistry:
-- 1978. Relationship between diagenesis, porosity an Introduction Emphasizing Chemical Equilibria in
reduction, and oil emplacement in late Carboni- Natural Waters (2nd. ed.). Wiley, London.
ferous sandstone reservoirs, Bothamsall Oilfield, SURDAM, R. C., BOESE, S. W. & CROSSLY,L. J. 1984.
East Midlands. Journal of the Geological Society, The chemistry of secondary porosity. In: MC-
135, 7-24. DONALD, D. A. & SURDAM, R. C. (eds). Clastic
HOVEY,J. K. & TREMAINE,P. R. 1986. Thermodynam- Diagenesis. American Association of Petroleum
ics of aqueous aluminium; Standard partial molar Geologists Memoir No. 37, 127-49.
heat capacities of Al3§ from 10 to 55~ Geochi- TARDY, Y. & GARRELS, R. M. 1974. A method for
mica et Cosmochimica Acta, 50, 453-9. estimating the Gibbs energies of formation of
HUGGETT, J. i . 1984. An S.E.M. study of phyllosili- layer silicates. Geochimica et Cosmochimica Acta,
cates in a Westphalian coal measures sandstone 38, 1101-16.
using back-scattered electron imaging and wave- WESTALL, J. C., ZACHARY,J. L. & MOREL, F. M. M.
length dispersive spectral analysis. Sedimentary 1976. MINEQL"a computer program for the calculatin
Geology, 40, 233-47. of chemical equilibrium composition of aqueous
KAISER, W. R. 1984. Predicting reservoir quality and systems. Water Quality Laboratory, Massachu-
diagenetic history in the Frio Formation (Oligo- setts Institute of technology, Technical Note No.
cene) of Texas. In: MCDONALD,D. A. & SURDAM, 18.

E. A. WARREN,Department of Geology, Mappin Street, University of Sheffield, Sheffield

$3 7HF, UK.
Petrology (including fluorescence microscopy) of cherts from the
Portlandian of Wiltshire, UK evidence of an episode of
meteoric water circulation

T. R. Astin

S U M M A R Y : Chert nodules from the Wockley Member of the Portland Stone (Upper
Jurassic) in Wiltshire, England are comprised of three stages of chert cementation and
sediment replacement: (1) opal-CT lepisphere cement and grain replacement, (2) length fast
chalcedony grain and matrix replacement and cement, and (3) quartz cement. These are
readily distinguished by their fluorescence, the first and last stages being non-fluorescent, the
second stage showing internal zonation changing from moderate to bright fluorescence with
time.
Some carbonate grain dissolution took place prior to silica precipitation.-The first stage
silica cement formed mainly as pore-linings, with minor replacement of grain margins. Its
botryoidal, lepisphere habit implies it was opal-CT, now inverted to quartz. Extensive grain
replacement took place while chalcedony cements formed. Chalcedony cements show an
increase in inclusions of brown organic particles with time, joined at a late stage by iron
oxide inclusions.
Silicification and carbonate dissolution occurred during a period of meteoric water
penetration, probably as a result of soil formation at the top of the Portland Stone. Silica was
apparently derived as opal-A from adjacent marine limestones, redistributed by ground-
water flow, and precipitated as opal-CT and chalcedony. The organic inclusions probably
come from the developing soil horizon, the iron oxides marking the migration of an oxidation
front away from the soil horizon into the sediment. The later quartz cement lacks inclusions
and was formed during later burial after stagnant groundwater conditions were re-
established.

This paper describes some cherts from the matrix, and as inclusions in cements. Organic
Portlandian (Upper Jurassic) of Wiltshire, Eng- matter can preserve the fine internal structure of
land (National Grid Reference ST 963 296) (Fig. biogenic grains, preserve grain boundaries (e.g.
1), which is an open to marginal marine sequence when included in micrite envelopes), highlight
of mainly limestones and sandy limestones. The the locations of diagenesis associated with
cherts come from a variably fossiliferous calcar- microbial activity, and reveal cement zonation.
eous, micritic sandstone, the Chicksgrove Plant All these effects have been seen and used to
Bed (Fig. 2), which contains terrestrial, lagoonal reveal primary textures of the cherts, to distin-
and marine faunal elements (Wimbledon 1976, guish between the stages of silica cementation
1980). The Plant Bed gives evidence of the and replacement, and to relate variations in the
earliest known non-marine conditions in the fluorescence of some of the silica cements to the
Upper Jurassic of southern England, and the types and distribution of their inclusions.
fauna includes an important mammalian as- The textural evidence has been used to relate
semblage. As part of a multidisciplinary study of chert formation in the Plant Bed with the period
the whole bed, including its provenance, strati- of emergence and soil formation which occurred
graphy and palaeontology (Wimbledon et al., in at the base of the overlying Purbeck Formation.
preparation), the cherts were investigated for
information on the depositional and early dia-
genetic environments in which the original Geological setting
sediment was deposited.
Chert textures in thin section are often Figure 2 gives the Portlandian sequence at the
difficult to interpret because of their small grain Chicksgrove Quarry. The Chicksgrove Plant
size, primary void fillings and neomorphosed Bed is about 60 cm thick and lies at the base of
grains with similar textures, and primary metas- the micritic Wockley Member. It rests erosively
table phases such as opal-CT having inverted to on marine glauconitic shelly sandstones and is
quartz at a later stage. overlain by a lagoonal micrite, followed by
Fluorescence microscopy brings out additional marine micritic limestones. The Plant Bed is
textural information in these cherts mainly about 8 m below the Great Dirt Bed, a well-
through its tendency to reveal the location of known palaeosol at the top of the Portland Stone
organic material in detrital grains and in the (Francis 1984).

From: MARSrlALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences,

Geological Society Special Publication No. 36, pp. 73-85. 73
74 T. R. Astin

[
o o o I
o o o o o o

t ~ ~ ~ i ~ "~ "~

CH/C/~SGROVE Q U A R R Y
~ PUROECK /. 6REENSAND, GAULT, U. GREENSAND
PORTLAND WEAI. DEN

FIG. 1. Location map for Chicksgrove Quarry. The left diagram shows the broad distribution of rocks in
southern England, with the boxed area expanded on the right. The map on the right is based on the British
Geological Survey 1:50 000 series geological maps.

Most of the chert nodules occur near the results in a greater colour range in the transmit-
bottom of the bed. They are typically 2-4 cm ted light than blue or blue-violet excitation, and
across and tend to be elongate along the bedding. the fluorescence was sufficiently strong in this
They have compact, dark grey centres and case to take advantage of this.
porous, friable, pale grey outer layers. Cherts also Two samples were analysed by semiquantita-
occur in the adjacent beds. The Wockley micrites tive XRD to determine the mineralogy of the
above contain scattered black, flint-like cherts cherts, and two thin sections were examined by
which replace the limestone. In the upper part of SEM with an EDX microanalyser to investigate
the underlying Tisbury Member, lenticular the composition of the inclusions.
patches and beds of crossbedded glauconitic
sands are totally chert cemented. Lower down,
silicification in the Tisbury Member is limited to
partial replacement of shells in sandy micritic Chert petrology
packstones.
Original sediment
Methodology The initial sediment comprised a wide variety of
detrital components with a mean grain size of
The chert nodules were sliced and impregnated fine sand (Wentworth scale). The largest grains
in araldite coloured with a green dye to make included unmicritized small gastropods, wood
porosity distinct in thin section. Twelve doubly fragments and a few large bivalve shell frag-
polished thin sections were prepared and exam- ments. The fine sand consisted of angular quartz,
ined by optical petrographic methods and inci- spores and carbonate, the latter mostly micri-
dent light excitation fluorescence. tized skeletal grains some of which are coated.
Both UV and blue-violet light excitation were The cherts included cemented and replaced
used for the fluorescence. The transmission grainstone, as well as replaced packstone with a
characteristics of the UV filters used are given in matrix of carbonate silt and mud which is
Fig. 3. Dravis & Yurewicz (1985) report that for difficult to resolve texturally.
carbonate rocks blue light excitation is better All the original non-silicate components (car-
than UV light excitation, because of the gener- bonate grains, organic grains, matrix) are now
ally greater fluorescence intensities resulting. In replaced by quartz. The replaced grains are made
the present study however, UV excitation was up of microcrystalline quartz or fibrous
found to be more effective. UV light excitation chalcedony.
 Chert petrology from the Wiltshire Portlandian 75

CHICKSGROVE ST 863 296

Fissile limestones
and palaeosol
Shelly micrite
a
z WOCKLEY
--I MEMBER
I-

t-
0
n
Gastropod micrite
0
] CH~'CHICKSGROVE PLANT
~
CH BED
C CH
CH
0
Glauconitic
TISBURY quartzose
0
13. MEMBER shelly
sands

.o_
CHICKSGROVE Bioturbated
0 MEMBER micrites

WARDOUR Sands,
MEMBER I silts
and muds

I ~ Mudstone [~ Bivalve shells

K ~ Sandstone r~l Gastropods
E
y ~ Limestone Wood
Crossbedding Chert
FIG. 2. The sequence present at Chicksgrove quarry (modified from Wimbledon 1976). The cherts described
come from the sandstone bed at the base of the Wockley Member.

The primary sedimentary texture of all the (2) length fast chalcedony, and (3) quartz (Table
chert is preserved unmodified by compaction. 1). They are most clearly seen in the largest pores
Thus delicate grains, including spores and which originate from gastropod chambers (Fig.
micrite envelopes defining secondary pores, are 4A, B, C).
preserved unflattened, and matrix-rich areas The first stage consists of partial to whole
often preserve pellet textures (Fig. 4A, C). This lepispheres (10-13 ~tm diameter) lining the pores
contrasts with the adjacent sediment in which and locally coalescing into botryoidal layers.
some compaction took place, as shown by shell They are now made of radiating microquartz
breakage and a denser matrix with no surviving crystals giving pseudo-extinction crosses
pelletal texture (Fig. 4K, L). between crossed polars. Their shape suggests
that they were originally opal-CT lepispheres,
and this is consistent with their sizes which lie
Cement stratigraphy
towards the upper end of those of opal-CT
Three main stages of silica cementation are seen lepispheres reported from deep sea cherts (see,
in the nodule centres, these being (1) lepispheres, e.g. Weaver & Wise 1972, Wise & Weaver 1974,
76 T. R. Astin
EXCITATIONFILTER

I~I eyepiece
'/Microscope
O.Z
Borrier
00"3
51 ~ /~ BARRIERFILTER
Hg vopour filter
lamp O;rt l
, i
), {rim) 400 500 600
plitter Wovelength
o

~Excitotion//)~ o~ 0.7
filter "~ 0-5 ~ 9s SPLITTER
J Objective

I'~ 0"1
T i , i ,
Specimen %(rim) 400 500 600

FIG. 3. An outline of the microscope arrangement for incident light fluorescence microscopy. The response
characteristics of the filters used for UV fluorescence in this study are shown.

Hein et al. 1978, Pisciotto 1981). Opal-CT is also pore-linings and up to 250 ~tm thick in the largest
reported as forming botryoidal layers (Robertson pores. Small pores are totally filled. In thin
& Hudson 1974). Neomorphosed opal-CT has section the earlier chalcedony is transparent and
been previously recognized in cherts for similar pale green, becoming increasingly opaque
reasons (Meyers 1977, Keene 1983). towards the younger pore-centres where it is full
The second stage cements are length fast, of brown to black inclusions. There has been
radial fibrous quartz (chalcedony), present as uncertainty about whether similar length-fast

FIG. 4. Photographs of the chert textures seen in thin section. In all cases the scale bar is 100 ~tm.
(A) Plain polarized light (PPL). Cross-section of a gastropod showing an internal chamber filled with a
combination of internal sediment (arrowed) and silica cement showing a prominent dark zone full of
inclusions.
(B) Crossed polarized light (XPL). This shows the quartz cement up to and including the inclusion-rich zone
to be fibrous (length fast) chalcedony, followed by a crystalline quartz cement.
(C) UV light fluorescence (UVF). This clarifies the distribution of internal sediment (sed), and shows the
presence of early hemispheres of non-fluorescent cement prior to the chalcedony (le). The chalcedony changes
from moderately to brightly fluorescent in the inclusion-rich zone. The late quartz is non-fluorescent.
(D) UVF. Another gastropod chamber cement sequence showing the presence of some brightly fluorescing
nuclei to some early cement lepispheres (a), the moderately fluorescing earlier chalcedony (b) passing into the
brightly fluorescing later chalcedony (c), and dark inclusions in the outer part of the bright zone of chalcedony
cement (d). These fluoresce a reddish-brown colour typical of iron oxides. (E) PPL. Outer part of a chert
nodule showing secondary voids (v) formed by opal sphere dissolution, now filled with green stained epoxy,
 Chert petrology from the Wiltshire Portlandian 77

replaced partly by early non-fluorescent silica (a),

and partly by fluorescent silica of the same
generation as the chalcedony cement (ch). (G) PPL.
Chert replaced sediment in which only larger quartz
grains (q), shell fragments (s) and spore grains (p) are
obvious. (H) XPL. This shows the microcrystalline
quartz texture typical of most of the chert with few
petrographic features visible. (I) UVF. This shows
much more of the original sediment texture, with
non-fluorescent quartz grains (q), replaced shells (s)
and spores (p). Areas of pore-filling cement are
revealed (ce), as are silt sized grains (si). (J) UVF.
Wood showing preservation of cell walls by
and showing up darker than the chalcedony cement moderately fluorescing silica and infilled cells with
(ch). The voids have a thin lining of quartz cement later bright silica (arrowed, ch) both of the same
(q). (F) UVF. A gastropod from the margin of a stage as the chalcedony cements. The wood has been
nodule showing voids (v) resulting from secondarily partly rotted prior to silicification as shown by the
dissolved lepispheres. The quartz lining the voids is presence of areas infilled with chalcedony but
non-fluorescent. The gastropod shell has been without cell structure preserved (rot).
78 T. R. Astin

FIG. 4 (continued). (K) PPL. Calcareous micritic sandstone of the Chicksgrove Plant Bed away from the chert
nodules. The sand includes shells (s) and a 'Raxella' type sponge spicule set in a carbonate matrix (m).
(L) XPL. Same view as (K) showing the voids (v) resulting from selective (presumed aragonitic) shell and
biogenic opal spicule dissolution, and sparry calcite (cal) partly filling some of the secondary pores.
chalcedony is a primary precipitate or a replace- weakly to non-fluorescent. They often have very
ment of opal-CT cement. Some authors prefer a bright fluorescent centres (Fig. 4C, D), seen as
replacement origin (Wise & Weaver 1974, brown particles 8-10 ~tm across in ordinary light.
Meyers 1977, though Meyers pointed out the The particles are therefore probably organic,
uncertainty) while others favour a primary perhaps the remains of microorganisms on
precipitate (Lancelot 1973, Keene 1983). Chal- which the opal-CT nucleated.
cedony is undoubtedly a primary precipitate in The earlier, pale green chalcedony shows
geodes (Mieke et al. 1984) and with no evidence moderate fluorescence, the intensity of which
to the contrary the Chicksgrove chalcedony is increases into the later chalcedony as the density
interpreted as a primary precipitate. Quartz fills of dark inclusions increases. The high intensity
the remaining space in the gastropod chambers, suggests that most of the brown inclusions are
increasing in crystal size toward the original pore organic. The youngest, innermost parts of the
centres. chalcedony contain inclusions which fluoresce
The three cement types have distinctly differ- red-brown (Fig. 4D), characteristic of haematitic
ent fluorescence intensities. The lepispheres are iron oxides. These are indistinguishable from
 Chert petrology from the Wiltshire Portlandian 79

TABLE 1. Summary of diagenetic textural evolution in the cherts, and the inferred controlling water
chemistry at each stage
TEXTURAL CHANGES INFERRED WATER CHEMISTRY
I
U.V.
STAGE DISSOLUTIONS I REPLACEMENTS CEMENTS FLUORESCENCE [Carbonate] [Organic
RESPONSE [Silica] [Oxygen] Matter]
i

Aragonite Microbial
I shells alteration of
wood
[ '~ ,-
A few shell
2 margins Opo~ ~ C T Dead
replaced by I~ ;
Opal--CT
Main Length - pale
replacement Dull
of carbonate fast green
grains and
3 matrix by Chalcedony brown Bright
Silica organtcs i.~ '~- i~ "

organics Bright + Red

+ Fe oxides Spots
Opal ~ CT in
nodule
margins
Micro- quartz Dead ]
organic matter in ordinary light. Late quartz is to that of the sediment inside the gastropod
mostly non-fluorescent. A few localized blue chambers (Fig. 4C, I). Within this matrix, many
spots are of uncertain origin. silt-sized detrital quartz and replaced carbonate
Two thin sections were examined by SEM grains are revealed for the first time (Fig. 4I).
with an energy dispersive X-ray (EDX) analysis The most likely precursor for the silicified
system. EDX analysis of the inclusion-rich matrix is lime mud, with some terrigenous silt.
chalcedony showed no major components other XRD analysis of the adjacent unsilicified sedi-
than silica. Minor local iron confirmed the ment shows that clay minerals form only a very
presence of some iron oxide inclusions and that small part of the sediment, while micrite is
most of the inclusions are organic and therefore abundantly present (Dr A. Parker, personal
undetectable by EDX. communication). Locally, the sediment within
Lepispheres and chalcedony cements fill all the gastropods is obviously pelleted, so much of
the primary pores between the finer grains in the the finer matrix may have originated as lime mud
nodule centres. The outer parts of the nodules, pellets.
however, are porous and friable (Fig. 4E, F). Wood fragments often show a thin zone of
Pores at the margin are partially to fully dark brown matrix at their margins in ordinary
cemented by pore-lining lepispheres followed by light. Borings filled with the same material
pale green and brown chalcedony. Most of the (diameter about 10~m) locally penetrate the
lepispheres dissolved after the chalcedony edge of the wood, truncating the primary cell
formed, leaving secondary pores with character- structure. No bark layer seems to be preserved.
istic spherical shapes. Where spheres dissolved, The marginal brown matrix has a fluorescence
a thin rim of non-fluorescent microcrystalline colour similar to, but brighter than, that of the
quartz is often present in the void (Fig. 4E, F). primary matrix (Fig. 4J).
This could be either undissolved silica from the Wood margins are interpreted as resulting
margins of the lepispheres which subsequently from microbial degradation (probably involving
recrystallized to quartz, or it could be a thin rim fungi and bacteria) during transport and contin-
of stage three quartz cement deposited after uing into early burial. Microbial organic matter
lepisphere dissolution. probably gives rise to the bright fluorescence.
Smaller amounts of the same organic matter
presumably occur in the fine sediment generat-
Matrix ing the similar fluorescence response.
The fine grained matrix is now a mass of
microcrystalline quartz crystals (1 and 10~tm) Grain replacement
within which primary grains, matrix and ce-
ments are difficult to distinguish in transmitted In ordinary light the grain margins are not
light. Fluoresence microscopy, however, clearly always visible (Fig. 4G), but UV fluorescence
distinguishes between these. Areas of silicified shows up distinctly the distribution of grains and
primary matrix show a fawn fluorescence similar their internal structures (Fig. 4I).
80 T. R. Astin

Some shell fragments and ooids became the typical sequence of events during the
secondary pores through carbonate dissolution silicification of wood (Scurfield & Segnit 1984).
prior to silicification, and were later filled with Most of the spores seem to have been replaced by
opal-CT lepispheres followed by chalcedony just the later more fluorescent chalcedony (Fig. 41).
as in primary pores. Along nodule margins, the All silica replacement took place before the
secondary pores have remained partially or even quartz cement formed.
wholly unfilled. Preferential leaching of arago-
nite grains is thought to give rise to these
secondary pores. The adjacent sediment also Discussion
shows selective replacement of carbonate grains
Silica source
with subsequent partial cementation by calcite
spar (Fig. 4K). In order to get the initial opal-CT precipitation
The replaced grains often show a fluorescence in the cherts, a high silica concentration in the
colour and intensity correlatable with the cement pore water is needed (Williams et al. 1985). This
sequence, revealing the relative timing of grain must necessarily be higher than would arise from
replacement. The fluorescence colours are identi- pore waters in simple equilibrium with quartz. In
cal with those of the cements, and some wood principle there are two ways of achieving this
and carbonate grains preserve the internal while still involving quartz dissolution.
organic structure, such as protein layers in First, quartz is much more soluble in strongly
bivalve shells and cellulose cell walls in wood alkaline (pH >i 10) water than under lower pH
(Fig. 4J). conditions (Volosov et al. 1972). The high silica
Two end-member types of replacement can be concentrations required for opal-CT precipita-
recognized. In one, quartz precipitation coin- tion at neutral to low pH could have arisen
cided with grain removal by dissolution, no large through quartz dissolving in alkaline solutions
voids forming at any stage. Original structural and then the pH changing. But strongly alkaline
organic material is usually preserved, and boun- waters are rare in nature and cannot have arisen
daries between replacive silica from different in the Chicksgrove situation because the abun-
episodes follow pre-existing internal boundaries dant carbonate would have buffered the pore
(Fig. 41). In the second type, a secondary pore water to pH 9.9 or below (Garrels & Christ 1965).
was formed as an intermediate stage. Organic Second, a high silica concentration could have
components of shells are not preserved within been achieved through evaporation. Evaporative
the replacive quartz, because organic material concentration is effective only at or close to the
degradation and removal took place at the same land surface, where it can give rise to soil-profile
time as carbonate dissolution and before silica silcretes (Summerfield 1982). Such silcretes have
replacement. Pore filling fabrics are present characteristic textures, including quartz grain
within the second types of grain, boundaries etching and replacement, grain fracturing, ex-
between different cement generations showing pansive fabrics, complex rug fills, and glaebule
no relation with former internal structures. formation (Summerfield 1982) which are absent
An intermediate type also occurs, some grains here. The Chicksgrove textures, on the other
showing partial loss of structure through partial hand, show that the cherts are a type of
dissolution prior to silicification. This has taken groundwater silcrete, formed in phreatic condi-
place, for example, in wood fragments, where the tions where evaporation could not play a direct
original wood structure has been locally lost, role in concentration.
suggesting some rotting of the wood before This argument suggests that quartz was not the
siiicification (Fig. 4J). primary source of the silica, and that a biogenic
opal-A source is likely, as indeed it is for the large
majority of cherts (see Wise & Weaver 1974,
Timing of grain replacement
Meyers 1977, Hein et al. 1978, Knauth 1979,
Some carbonate grains have been partially Williams et al. 1985 for just a few examples). As
replaced by non-fluorescent silica, typically at opal-A dissolves, opal-CT, being less soluble
shell margins (Fig. 4C, F, I). The silica was than opal-A, can precipitate and is kinetically
probably opal-CT. Most of the carbonate grains preferred to quartz precipitation from pore
and wood are represented by combinations of waters having moderately high silica concentra-
weakly and brightly fluorescent silica, the flu- tions (Williams & Crerer 1985).
orescence being the same as in the chalcedony A few sponge spicules are indeed found in the
cements. The cellulose cell walls of the woods Chicksgrove Plant Bed, possibly derived, with
were replaced first, typically by weakly fluores- other marine biogenic grains, from slightly older
cent silica, and the cells subsequently infilled rocks. The spicules are preserved as microquartz
with brightly fluorescent silica (Fig. 4J). This is within the cherts, and the presence of others is
 Chert petrology from the Wiltshire Portlandian 81
recorded by voids in the upper micritic part of line quartz cement which may have formed in
the Plant Bed (Fig. 4K, L). However, they were the nodule margins at this time, being less than
too few to have acted as the sole source of opal-A. the opal-A dissolved, suggests that ground water
One untested possibility is that sponge spicules was no longer contributing silica, the pore water
were dissolved from the overlying marine micri- outside the nodules becoming undersaturated in
tic limestones, perhaps with a minor contribu- silica with respect to quartz during opal-CT
tion from the marine glauconitic sandy lime- dissolution.
stones underneath. This is suggested by the However, the central pores became completely
overall distribution and types of Portlandian cemented with quartz, implying that higher
chert at Chicksgrove. The black flint-like cherts silica concentrations were maintained inside the
in the Wockley Limestone suggest a local silica nodules than outside. This must have been set up
source in this unit, and silicification decreases by silica diffusion from sites of opal-CT dissolu-
downwards below the Plant Bed through the tion in the nodule margins.
Tisbury Member, implying a source from above Opal-CT dissolution took place before opal-
rather than below, and transport of dissolved CT inversion to quartz, but both this and the
silica by downward moving ground water. associated quartz cementation could have taken
place later than the formation of chalcedony.

Inferred water chemistry during diagenesis Carbonate concentration

As shown by the shell replacement textures and
Silica concentration the presence of micrite envelope 'eggshell' grains
The change in cement type with time is from in the nodule margins, some dissolution of
opal-CT via length-fast chalcedony to quartz. carbonate must have taken place to form
The same sequence is found in Mississippian secondary pores before silicification (Table 1).
limestones (Meyers 1977). In both, therefore, the Aragonitic grains were probably lost preferen-
concentration of dissolved silica in the pore tially to form these secondary voids, so there is no
water was initially high, being saturated with need to infer carbonate undersaturation with
respect to opal-CT, and then decreased, finally respect to calcite prior to silicification.
falling below that leading to quartz precipitation The opal-CT is nearly all void filling cement,
(Table 1). with only a little grain replacement at this stage.
The amounts and distribution of the three This implies approximate carbonate saturation
cement types show that the rate of change of of the pore-waters at this stage. All the remaining
silica concentration varied. After initial opal-CT carbonate in the nodules was dissolved synchron-
precipitation the silica concentration fell, but ously with the formation of chalcedony when the
remained high enough for some time to cause the pore water became undersaturated with respect
growth of fibrous chalcedony, this being the to carbonate. This trend towards increasing
main type of silica in the nodules. To maintain a carbonate dissolution, coincident with silicifica-
high concentration, dissolved opal-A must have tion, demands the introduction of quite large
been transported into the bed. volumes of meteoric water (Fig. 5C).
Two likely mechanisms for the fall in silica Undersaturation is known to occur in at least
during opal-CT and chalcedony precipitation are two stages of a meteoric water flow path through
(1) progressive dilution of silica-rich pore waters carbonate rocks. Meteoric water, initially under-
by silica-poor water, and (2) progressive deple- saturated with carbonate on entering the sedi-
tion of the opal-A available for dissolution in the ment, tends towards equilibrium with carbonate
source rock (probably 'the overlying limestones). along the flow-path. Knauth (1979) pointed out
The first effect will have taken place following that undersaturation recurs in the mixing zone
exposure as meteoric water mixed with and with marine pore-water if suitable differences of
replaced the original marine pore water, the Pco2 exist.
second effect occurred during prolonged meteor-
ic water circulation (Fig. 5B, C).
Significance of inclusions in the chalcedony
The end of chalcedony precipitation could
have resulted either from opal-A depletion in the The chalcedony shows a systematic increase in
source, or from marine transgression of the the abundance of included brown particles from
exposed surface which stopped the circulation of early to late formed chalcedony. The intense
meteoric water. fluorescence of the inclusion-rich chalcedony
After precipitation stopped, opal-CT was proves that they are largely organic.
dissolved at the nodule margins and quartz Organic matter in the pore water during early
precipitated. The small amount of microcrystal- diagenesis is likely to have come partly from
8z T. R. Astin

A
I I I
I I WOCKLEY
1 I I ~ MEMBER
MARINE 1
9 I. - - . - - ~ . . . J _ . ~-- ~. --
I 4- P L A N T
9 " ..t . . . . . BED
. . I . . . . . . !
. I 9 ,
9 9 1. 9 .
,~TISBURY
9 , "1 ...----_1 MEMBER

I ~" r-1 ,'i mvl EMERGENCE

B i i i J
I 1 I i I
i / 1 I
__.L.._.~_. w/ .., z l
~ , , , ~ _ " .'. .
MIXING 9 "! "-
....
.~11~ . ' . ' q "
. I 9
9 9.
.I
ZONE
.I . . . . . L

C ~--J/~-, J'~ ~-~ 4 - S O I L

9 ~ " /~1
I I "~. j / FORMATION
I . /I / i

METEORIC ._~: ~ J - - ~ ' r

~ I * i~ ~ . . . . 9
I . . . . . . "1 9 .

9 ." 1 . I 9 .. I
9 , 9 .| , .. 9

9 9 "1 . . . . I

D o,
1 I L '*'PURBECK
t ~" ~~F -%'~- ~ I-: F O R M A T I O N
I I
MARINE i t J I i
MODIFIED ~ - "- 4 - i ' " - ." "'"t . "..,"
_ . . . ,":-
_ ":
. . . . ! . . . . . I
9 I. . . 9I .
9 .I , 9 I

FIG. 5. This diagram summarizes the stages of groundwater evolution inferred during early diagenesis and
chert formation. Stages A to D correspond approximately to stages 1 to 4 of diagenesis identified in Table 1.
A. Initially the pore water was marine, and slow aragonite dissolution could have taken place in the
sediments. Some flint-like cherts may have started forming in the Wockley Member micrites above the
sandstone bed. B. Upon emergence, meteoric water penetrated to the sandstone bed, first mixing with and
then displacing the marine pore water. The arrows show the meteoric water flow schematically. During this
stage, opal-CT lepisphere cements formed. C. Continued meteoric water circulation took place during soil
formation on the emergence surface. The sandstone is the most permeable unit and is likely to have caused a
change to lateral flow of the meteoric water at this level, limiting the amount of chert formation in the
decreasingly permeable Tisbury Member below. During this stage, chalcedony formation and carbonate
dissolution took place. D. Following marine transgression and the start of Purbeck Formation sedimentation,
marine pore waters were re-introduced and vigorous groundwater circulation ceased, bringing an end to chert
formation. At this or a later stage, some quartz cement formed in chert centres.

physical b r e a k d o w n of dissolving shells and be caused by dissolved organic c o m p o u n d s from

partly from microbial activity. Both processes these sources being trapped during growth.
must have operated before and during the early The a b u n d a n t organic inclusions in the later
stages of chalcedony formation. The m o d e r a t e c h a l c e d o n y show that a new source of organic
fluorescence shown by the early chalcedony, particles b e c a m e available in m u c h greater
unrelated to obvious particulate inclusions m a y concentrations than ordinarily occur during early
 Chert petrology from the Wiltshire Portlandian 83
diagenesis, they must have been introduced by and selective dissolution of some (probably
moving ground water. The most likely source is aragonitic) carbonate grains (Table 1).
the humic layer of a suitable soil. The palaeosol Knauth (1979) assumed that meteoric water
previously mentioned lies about 8 m above the would reach saturation with carbonate sooner
cherts (Fig. 2) and presumably had a suitable than with opal-A at the start of the flow path. But
humic layer (Francis 1984). this need not necessarily be the case; for example
The iron oxides shown by fluorescence micros- where meteoric water influx is episodic, as in a
copy to be present in the latest chalcedony (Fig. seasonal climate. In this situation, silica and
4D) suggest that the waters became oxidizing carbonate concentrations move towards satura-
when this formed. Here the postulated circula- tion in stagnant groundwater during the dry
tion of meteoric water at this time is strength- season. Influx of meteoric water during the wet
ened, because it would lead to progressively season dilutes the groundwater, which becomes
deeper penetration of oxygenated water along undersaturated in carbonate and opal-A, but
the flow path, away from the recharge zone (soil) remains saturated with respect to chalcedony
(Fig. 5C). close to the recharge zone.
The later quartz cement is clear and non-
fluorescent, implying that the pore waters no
longer contained significant amounts of dis-
solved or particulate organic matter. This con- Conclusions
trasts markedly with the immediately preceding
chalcedony, and records a quite different pore Early silicification resulted from meteoric water
environment when marine conditions were re- penetrating the Portlandian limestones at a time
established and stopped meteoric water penetra- of emergence and soil formation. The soil in
tion in the area (Fig. 5D). Later quartz cement question was probably what is now the Portland
could have formed significantly later. Dirt Bed, a palaeosol 8 m above. Meteoric water,
undersaturated with respect to opaline silica and
carbonate, dissolved biogenic opal from adjacent
Relation between meteoric water circulation and limestones and transported it into the Plant Bed.
chert formation Initially, the meteoric water mixed with little
modified marine water saturated with respect to
The cherts show very similar textures to those carbonate and biogenic opal (Fig. 5B) and
described from Mississippian limestones by precipitated opal-CT cement. Increasing vo-
Meyers (1977). He relates chert formation to an lumes of meteoric water led first to mixing with
opal-A source of silica combined with meteoric the saline water and then to complete replace-
water circulation, basing his argument on the ment (Fig. 5C). Hence the pore water became
timing of silicification relative to calcite cement undersaturated with respect to carbonate, result-
zones of proven origin from meteoric phreatic ing in dissolution of the carbonate in the nodules
water. and allowing extensive silicification. At the same
Knauth (1979) suggests that the formation of time, carbonate was leached from adjacent
replacive cherts in limestones requires special limestones. Silica concentrations dropped as
circumstances for the necessary undersaturation biogenic opal was depleted in its source rock,
with respect to carbonate in the meteoric water. leading to precipitation of chalcedony after opal-
He concluded that silicification is most likely in CT. As an organic-rich soil developed on the
the zone of mixing between previously carbon- exposed land, the ground water became enriched
ate-saturated meteoric and marine waters, where in organic particles. These formed inclusions in
local undersaturation could sometimes occur as a the later chalcedony. During the same episode an
result of differences in Pco2 between the two oxidation front migrated into the sediment,
types of water. giving rise to iron oxide inclusions in the late
In the Chicksgrove case, silicification is chalcedony cement. Quartz cements followed
inferred to have taken place close to the recharge after the soil horizon was buried by further
zone before the meteoric water became saturated sediment (Fig. 5D). They may have formed much
with carbonate, the recharge zone being the soil later during deeper burial, but probably before
now some 8 m above. This is consistent with the opal-CT inverted to quartz because some opal-
uncompacted fabrics of the sediments in the CT dissolution probably freed silica for the
cherts, compared with the compacted matrix quartz cement.
fabrics in adjacent sediments. Chert formation The study exemplifies how an early episode of
therefore took place fairly early in the diagenetic emergence and meteoric water circulation can
history, the only changes prior to silicification cause silica mobilization and silicification. It
being minor microbial degradation of the wood shows that this silicification can occur close to
84 T. R. Astin

the recharge zone. In the P o r t l a n d i a n of Wilt- Penny Williams helped me with some stimulating
shire this has led locally to the p r e - c o m p a c t i o n discussions about the textures of silcretes and their
preservation of fossils with obvious palaeontolo- genesis, and the extent to which these cherts might be
considered groundwater silcretes.
gical benefits.
I wish to thank those who led me to fluorescence
microscopy, whether knowingly or unknowingly, and
subsequently encouraged me to explore its possibili-
ACKNOWLEDGMENTS:Thanks to Bill Wimbledon of the ties; M. Talbot, J. A. D. Dickson and J. Dravis.
Nature Conservancy Council for asking me to investi- Thanks to the reviewers of the paper, W. J. Meyers
gate the cherts and explaining their geological setting. and G. Carson, who offered numerous helpful com-
The technical staff in the Department of Geology at ments for improvement of the manuscript, and my
Reading are thanked for their help. Ian Pryde took understanding of the subject. Professor P. Alien
much care in making the thin sections, Jim Watkins instructively helped improve my English style, though
helped prepare the photographs for publication and any poor construction and inaccuracies remaining are
Alan Cross helped prepare the diagrams. my own choice.

References

BUURMAN, P. 1972. Mineralisation of fossil wood. PISCIOTTO, K. A. 1981. Diagenetic trends in the
Scripta Geologica, 12, 1-43. siliceous facies of the Monterey shale in the Santa
DRAVIS, J. J. & YUREWICZ, D. A. 1985. Enhanced Maria region, California. Sedimentology, 28, 547-
carbonate petrography using fluorescence micros- 71.
copy. Journal of Sedimentary Petrology, 55, 795- ROBERTSON, A. H. F. & HUDSON, J. D. 1974. Pelagic
804. sediments in the Cretaceous and Tertiary history
FRANCIS, J. E. 1984. The seasonal environment of the of the Troodos Massif, Cyprus. In: HsO, K. J.
Purbeck (Upper Jurassic) fossil forests. Palaeogeo- & JENKYNS, H. C. (eds). Pelagic Sediments: on
graphy, Palaeoclimatology and Palaeoecology, 48, Land and Under the Sea. Special Publication of
285-307. the International Association of Sedimentologists,
GARRELS, R. M. & CHRIST, C. L. 1965. Solutions, 1, 403-36. Blackwell Scientific Publications,
Minerals and Equilibria. Harper & Row, New Oxford.
York. SCURFIELD, G. & SEGNIT, E. R. 1984. Petrification of
HEIN, J. R., SCHOLL,D. W., BARRON,J. A., JONES, M. wood by silica minerals. Sedimentary Geology, 39,
G. & MILLER, J. 1978. Diagenesis of late Cenozoic 149-67.
diatomaceous deposits and formation of the SUMMERFIELD,M. A. 1982. Silcrete. In: GOUDIE, A. S.
bottom simulating reflector in the southern Bering & PYE, K. (eds). Chemical Sediments and Geomor-
Sea. Sedimentology, 25, 144-81. phology. Academic Press, London.
KEENE, J. B. 1983. Chalcedonic quartz and occurrence VOLOSOV, A. G., KHODAKOVSKIY,I. L. & RYZENKO,
of quartzine (length-slow chalcedony) in pelagic B. N. 1972. Equilibria in the system SiOz-
sediments. Sedimentology, 30, 449-54. H,O at elevated temperatures along the lower
KNAUTI-I,L. P. 1979. A model for the origin of chert in three-phase curve. Geochemistry International, 9,
limestone. Geology, 7, 274-7. 362-77.
LANCELOT, Y. 1973. Chert and silica diagenesis in WEAVER, F. M. & WISE, S. W. 1972. Ultramorphology
sediments from the central Pacific. In: WINTERER, of deep sea cristobalitic chert. Nature, London,
E. L., EWlNG, J. I. et al. (eds). Initial Reports of the 237, 56-7.
Deep Sea Drilling Project, XVII, 377-405. US WILLIAMS, L. A. & CRERAR, D. A. 1985. Silica
Government Printing Office, Washington, DC. diagenesis, II. General mechanisms. Journal of
MEYERS, W. J. 1977. Chertification in the Mississip- Sedimentary Petrology, 55, 312-21.
plan Lake Valley Formation, Sacramento Moun- WILLIAMS,L. A., PARKS,G. A. & CRERAR, D. A. 1985.
tains, New Mexico. Sedimentology, 24, 75-105. Silica diagenesis, I. Solubility controls. Journal of
MIEKE, Ci., GRAETSDCH, H. & FLORKE, O. W. 1984. Sedimentary Petrology, 55, 301-11.
Crystal structure and growth fabric of length-fast WIMBLEDON, W. A. 1976. The Portland Beds (Upper
chalcedony. Physics and Chemistry of Minerals, 10, Jurassic) of Wiltshire. Wiltshire Natural History
197-9. Magazine, 71, 3-11.
 Chert petrology from the Wiltshire Portlandian 85
- 1980. Portlandian correlation chart. In: COPE, J. WISE, S. W. & WEAVER,F. M. 1974. Chertification of
W. C., DUFF, K. L., PARSONS,C. F., TORRENS, H. ocean sediments. In: HsT2,K. J. & JENKYNS, H. C.
S., WlMBLEDON, W. A. & WRIGHT, J. K. A (eds). Pelagic Sediments: on Land and Under the
Correlation of Jurassic Rocks in the British Isles Part Sea. Special Publication of the International
Two: Middle and Upper Jurassic. Geology Society Association of Sedimentologists, 1,301-26. Black-
of London, Special Report No. 15. well Scientific Publications, Oxford.

T. R. ASTIN, Department of Geology, University of Reading, Whiteknights, PO Box 227,

Reading RG6 2AB, UK.
Silicification fabrics from the Cenomanian and basal Turonian of
Devon, England: isotopic results
Greg A. Carson

S U M M A RY: Silicification fabrics within the Cenomanian and basal Turonian of Devon
may be classified by their morphology and petrological relationships. The replacement of
bioclasts by silica was often a two stage process; the early stage has 6180 values ranging from
30.0 to 32.4%0SMOW. The palaeotemperatures calculated from these are close to values for
Cenomanian sea water and thus the early silicification took place in marine porewaters and
before significant burial. The consistently lighter values for the later replacement (ranging
from 27.2 to 29.4%0) cannot be explained by an increase in burial depth. Partial
recrystallization in meteoric waters is believed to be the mechanism responsible. Isotopic
data suggest that the formation of siliceous nodules (chalcedony cemented sands) within the
arenaceous lower members of the Cenomanian was an early process, although petrographic
evidence points to possible (later?) recrystallization. The formation of late-stage drusy quartz
occurred entirely in meteoric porewaters.

Two types of silicification occur in the Cenoman- Location

ian of Devon: the replacement of fossils and the
The principal field locality is a small working
formation of siliceous nodules. Silicification of
sand pit (known as the White Hart Sand Pit) at
bioclasts is both geographically and temporally
Wilmington, south Devon (National Grid Refer-
common. It has been documented in the Ordovi-
ence SY 209 998). Of minor importance in this
cian brachiopods of Scotland (Curry 1986), in
study are several coastal exposures situated
Carboniferous corals in Belgium (Buurman &
between Branscombe and Haven Cliff (Fig. 1).
Van der Plas 1971), brachiopods in Ireland
The Cenomanian consists of a series of variably
(Brunton 1966, 1968, 1984) and New Mexico
indurated bioclastic and arenaceous packstones
(Meyers 1977), bivalves and brachiopods from
which, on the coast, may be divided into four
the Permian of Wyoming (Schmitt & Boyd 1981),
distinct members, each one separated by a
corals and bivalves in the Liassic of Morocco
prominent glauconitized and/or phosphatized
(Faug6res 1978), bioclasts in the Upper Jurassic
hardground (Fig. 2). Inland, the basal members
(Wilson 1966) and brachiopods, bivalves (Brown
of the succession change from indurated lime-
et al. 1969) and echinoids (Holdaway & Clayton
stones to a sequence of coarse, loose quartz sands
1982) in the Upper Cretaceous of southern
with occasional patchy carbonate cementation
England. In many cases cited in the literature,
the silicification was considered significant only and development of silicified nodules and cherts.
The Cenomanian in Devon shows extreme
in as much as the excellent preservation was of
variation in thickness over very short distances.
taxonomic value to palaeontologists (e.g. Brun-
ton 1984 and Curry 1986). In other instances, The thinnest sequence is 26 cm at Haven Cliff,
limited geochemical work was carried out (e.g. while the expanded succession at Wilmington,
Buurman & Van der Plas 1971), while a detailed just 11 km away, is nearly 9 m. The same scale of
hypothesis on the formation of different replace- variation is seen over a distance of less than
1.5 km--the Cenomanian is 37 cm thick on Beer
ment fabrics was established by Holdaway &
Clayton (1982) on petrographical data. In this Beach and over 11 m at the Beer Stone Adit in
study the work of Holdaway & Clayton (1982) is Hooken Cliff (Jarvis & Woodroof 1984). Smith
taken one stage further, both in more detailed (1957) has attributed this variation to intra-
observation of replacement fabrics and in using Cretaceous folding. However, it is more likely to
oxygen stable isotope methods in the interpreta- be due to intra-Cretaceous faulting (for fuller
tion of the timing of replacement. details, see Jarvis & Woodroof 1984).
The siliceous nodules in the Cenomanian of
Devon are essentially chalcedony cemented
Depth of burial
quartz sands. Similar chalcedony cementation
fabrics are well documented, from small-scale In order to assess the effect of any burial
cementation of lepispheres in flint (Clayton diagenesis, it is necessary to consider the burial
1982) to cementation of ooids as described by history of the sequence. Sediment deposition at
Wilson (1966). Wilmington after the Cretaceous was minor.

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences,

Geological Society Special Publication No. 36, pp. 87-102.
87
88 G. A. Carson
possible that deposition at Wilmington took
place during the remainder of the Senonian, but
the sediment was eroded during uplift. Using an
average deposition rate of 15 m/106 yr (Black
1953), the duration of the Senonian (excluding
the Maastrichtian) as 16 Ma (Van Hinte 1976)
and a minimum of 10% compaction (H~kansson
et al. 1974), a chalk pile of 216 m is produced.
Including the 45 m of observed Turonian on the
coast, this gives 261 m. It is, however, unlikely
that 260 m of burial was attained since this part
of England was in a structurally high setting
during the upper Cretaceous (Hancock 1975,
fig. 2).

Analytical methods
Standard reflected and transmitted light petro-
graphy were employed. Scanning electron mi-
croscopy (SEM) specimens were mounted on
aluminium stubs and sputter coated with Au/Pd.
They were investigated using a Philips 501 SEM
with an energy dispersive X-ray (EDAX) attach-
ment. Oxygen isotope analysis was carried out at
the NERC BGS Laboratory, London. The
oxygen was extracted from the manually ground
silica by bromine pentafluoride treatment using
a method similar to that described by Clayton &
Mayeda (1963). Sample outgassing took place at
550~ overnight, pre-fluorination was at room
temperature and reaction was at 450~ for 8 h.
Analysis was performed on a VG Micromass 903
mass spectrometer. Results are reported in the 6
notation as the %o deviation from SMOW. The
value for the international quartz standard (NBS
28) determined by the BGS laboratory is 9.6%0,
--" / ~ i \ Be~eHraVenCliff '~...~ "~ with a reproducibility of +0.2%0. The mean
beach deviation for duplicate sample analyses was
~eer Head "~e +0.45%0.
Hooken Cliff ,9,,~
Upper Greensand 3km
Chalk and Cenomanian
Petrography

FIG. 1. Location map of field area (after Silicification of bioclasts

Smith 1961).
Holdaway & Clayton (1982) identify three main
silicification fabrics in these fossils:
Following regional uplift, SW England was (1) A granular white crust, which they interpret
subaerially exposed during the Cenozoic (see as having formed when the rate of silica
palaeographic reconstructions in Anderton et al. supply outpaced carbonate dissolution.
1979). Thus an estimation of maximum burial (2) Beekite rings, where silica supply was limi-
depth for these sediments depends solely upon ted and outpaced by carbonate dissolution.
the amount of chalk deposition during the They suggest that a 'liesegang' diffusion
remainder of the Cretaceous (that is, the Turon- situation was achieved, causing the silica to
ian onwards). The maximum thickness of chalk be precipitated in concentric zones.
present in Devon today is about 98 m (Rawson et (3) Fine-scale replacement, which they believe
al. 1978), extending just into the Coniacian. It is was due to precipitation of silica keeping
 Silicification fabrics from the Cenomanian of Devon 89
White Hart sand pit, Wilmington [ tm
(composite sectioff) . . . . . . . Lo

] Chalk

Sandy/gtauconitic chalk

Loose sands

~ Nodular beds/mineralized

Beer Head ~ Gtauconitized or phosphatized

hardground
Indurated sediment

Burrows (Thalassinoides3 and

bosses of hardground
Pinnacles ~ Shell debris

Little Beach ~-~ Siliceous nodules

Hooken I ~ U//~}) \ \\\ Haven Cliff ~ Goethite (after pyrite)

[ ~ Ceriopora ramulosa

Pounds Pool

Cove H d o d . - . ~ ~ . . . . . . . . .

' ' -,=--------- 4 km . P t J ~= 11 km ,,

FIG. 2. Three stratigraphic profiles of the Cenomanian in Devon.

FIG. 3. Lutecite rosettes (R) replacing an oyster shell (S). Note the replacement of the vesicular shell structure
by silica (arrowed). Loose, Wilmington. Crossed nicols. 0.5 mm scale bar.
90 G. A. Carson

pace with slow calcite dissolution, and thus common in bivalves though they are also seen in
accurately preserving the original crystal rhynchonellids.
morphology.
Fine scale replacement. This tends to be concor-
They suggest that the crystal form and thus dant with the shell fabric and occurs frequently
fossil type is the controlling factor in determining in brachiopods and Inoceramus. The mode of
the type of silicification that develops. So, in replacement tends to vary in appearance depend-
echinoids, the large crystal size causes a granular ing upon fossil taxa. In Inoceramus there is
crust to form (dissolution of carbonate being slow clearly an inter-prism replacement, the prism
compared to the precipitation of silica); in boundaries acting as the principal pathways
bivalves (especially Exogyra), beekite forms (due through which carbonate dissolution and silica
to the rapid dissolution of the smaller calcite replacement took place (Fig. 4). Where this
crystals), while in brachiopods, a fine-scale process is most advanced, nearly all the original
replacement occurs (due to the shell prisms being internal structure of the shell is lost (Fig. 5). An
large enough to prevent very rapid dissolution, inter-crystalline replacement also occurs within
but not too large, preventing dissolution and belemnites (Fig. 6), where the silica follows the
causing the precipitation of a granular crust). concentric and radial structure of the guard. In
In the present study, the following replace- rhynchonellids, however, the smaller dimensions
ment fabrics were observed in thin section: of the individual fibres commonly promote their
complete silicification so that an incompletely
Chalcedony 'rosettes'. These are discordant, cut- replaced single fibre may be difficult to discern.
ting across the original calcite fabric of the shell, Figure 7 shows a completely replaced brachial
and occur as aggregates of chalcedonite and/or loop of a Kingella sp., where the silica closely
quartzine/lutecite (Folk & Pittman 1971) fibres. follows the original fibrous structure of the shell,
When they form as 'solitary' replacements, they In a partially replaced example (Fig. 8) the silica
are centred around a point, forming what must also follows the original shell fabric, but pos-
be an irregular sphere. However, most com- sesses a much more irregular surface.
monly they are gradational from finer grained The present detailed investigation demon-
quartz and possess hemispherical or botryoidal strates that the silicification fabrics by Holdaway
forms (Fig. 3). They are equivalent to cross- & Clayton (1982) are by no means mutually
sections through beekite. These fabrics are most exclusive:

FIG. 4. Partial silicification of Inoceramus sp. taking place preferentially along prism boundaries (small
arrows). The same phase of silicification also follows boundaries between sparry calcite cement crystals (large
arrows). Base of Hooken member, loose block at Humble Point. Plane polarized light. 0.2 mm scale bar.
 Silicification fabrics from the Cenomanian of Devon 91

FIG. 5. Silicification of Inoceramussp., more advanced than in Fig. 6.

Silicification is most intense towards the centre of the fragment (S), where no
remnants of original shell fabric are seen. Calcite still remains in the exterior
part of the shell (E) and as 'islands' (I) isolated by silica replacement on the
interior margin. Base of the Hooken member, loose block at Humble Point.
Plane polarized light. 0.2 mm scale bar.

FIG. 6. Silicification of a belemnite following the concentric (C) and radial

(R) structure of the guard. Wilmington sands (73 cm above the Small Cove
Hardground), Wilmington. Crossed nicols. 0.2 mm scale bar.

(1) Scanning electron microscopy reveals the At a macroscopic scale--one valve of a

fine-scale replacement in brachiopods. At a pectinid (sample L24) contains at least three
macroscopic scale, however, definite concen- texturally different forms of silica (Fig. 9): a very
tric beekite structures are visible. soft (chalk-like) pure white replacement occur-
(2) Beekite is observed grading into fine-grained ring as small (-~ 2 mm) blebs sporadically in the
quartz in Exogyra specimens. shell, a white beekite replacement (from which
(3) Different styles of silicification are observed individual zones were sampled) and a grey fine-
within individual specimens at different scale replacement. Frequently, the white beekite
scales: may grade into the grey replacement (i.e. the
92 G . A . Carson

FIG. 7. SEM photograph of a completely silicified brachial loop of Kingella

sp. Note the faithful replacement of the original shell fabric. The patch on
the left is a local development of drusy quartz. Loose, Wilmington. 0.1 mm
scale bar.

FIG. 8. SEM photograph of a partially silicified rhynchonellid. The original

calcite fibres (C) possess a fresh appearance, while the silicified prisms (S)
have a more irregular outline. Loose, Wilmington. 30 ~tm scale bar.

latter may possess concentric rings), but also the A drusy fabric of mega-quartz (Folk &
boundary may be abrupt. Pittman 1971), occluding primary porosity (Fig.
At a microscopic scale--within one fragment 12), is occasionally associated with replaced
of Inoceramus--replacive silica is observed in fossils. Crystal sizes vary from less than 100 p.m
two completely different styles: replacing along (near the margin) up to 800 ~tm (in the centre of
prism boundaries (Fig. 10) and forming chalce- the cavity).
dony botryoids that completely transect prism
boundaries (Fig. 11). Siliceous nodules
Silicification post-dates borings and early
calcite cementation (Fig. 4) and took place The occurrence of white siliceous nodules is
during early diagenesis. restricted to the sandy facies in the White Hart
 Silicification fabrics from the Cenomanh~n of Devon 93

FIG. 9. Silicified pectinid (after etching in dilute HC1). Most of the shell
consists of prominent beekite (A-J) and a dark fine-scale replacement (K-M)
with occasional blebs of soft, white silica (N-P. N is scraped from the
surface of a beekite ring). Loose, Wilmington.

FIG. 10. SEM photograph showing silica following prism boundaries in

Inoceramussp. Note the corroded prisms where silicification has occurred
(right) compared to the unaltered shell (left). Loose, Wilmington. 0.1 mm
scale bar.

Sand Pit at Wilmington. Here they occur in The nodules are composed almost entirely of
vague bands, the more prominent ones being chalcedony cemented quartz grains (samples
,-~ 1 m apart. Their morphology clearly reflects a 436-10, 436-14, 462) and do not show any
relationship to burrow structures within the evidence of an initial lepispheric or sponge
sediment. They vary in form from digitate to framework. The overall fabric is very similar in
irregular masses. In many cases they may still be style and scale to those found by Wilson (1966,
hollow, the interior space corresponding very fig. 3) except, instead of ooids, the chalcedony
well with the dimensions of Thalassinoides cements detrital quartz. Frequently the central
burrows found in the overlying chalk (Fig. 13). part of a nodule is composed of red chert,
94 G. A. Carson

FIG. 11. From the same specimen as Fig. 10, SEM photograph of a
chalcedony rosette cutting across prisms in Inoceramus sp. Loose,
Wilmington. 0.1 mm scale bar.

FIG. 12. Drusy quartz (D) occluding primary porosity within a silicified
rhynchonellid (R). The development of 'scalenohedral' calcite (C) on the
inside of the shell occurred prior to quartz precipitation. Cenomanian burrow
fill, 34 cm below the Small Cove Hardground, Wilmington. 0.5 mm scale bar.

showing a conchoidal fracture even on a micro- wide (in digitate forms) to greater than 8 cm
scopic scale. The chert core and hollow centre wide when replacing more idiomorphic
appear to be mutually exclusive (although one burrows.
loose specimen shows a small 'cavity' with (2) In one example (Fig. 14), a patch of
development of quartz druse). chalcedony, which did not contain any
Investigation has also shown the following quartz grains, was observed immediately
features: adjacent to the central chert, and was
continuous with an area within the chert
(1) H a n d specimen observation revealed that, which was also devoid of quartz grains.
despite the size of the nodule, the thickness (3) At certain levels in the sands the nodules do
of the exterior zone is always about the same, not contain a chert core but possess a distinct
--~1-3 cm. Chert cores vary from ~0.5 cm 'pink' central area. It is suggested that the
 Silicification fabrics from the Cenomanian of Devon 95

FIG. 13. Slightly compacted hollow Thalassinoidesburrow preserved as a

siliceous nodule. The chalcedony cement occurs as botryoids on the interior
surface. Wilmington Sands, Wilmington.

FIG. 14. Margin of chert core within a siliceous nodule. A patch of white
chalcedony, free of detrital quartz (N), is adjacent to a patch of chert, also
free of detrital quartz (C). This association suggests the alteration of the chert
core to the fabric seen in the rest of the nodule. See text for details. Loose,
Wilmington. 0.25 mm scale bar.

pink chalcedony represents a remnant of chalcedony overgrowing perfectly preserved

virtually complete alteration of the chert. authigenic 'dog-tooth' calcite (---20 ~tm long,
--- 5 ~tm wide) suggests some calcite precipitation
Scanning electron microscopy reveals the prior to chalcedony cementation.
common preservation of primary porosity and Thin section petrography shows that despite
occasionally moldic secondary porosity (Fig. 15), the macroscopic contrast between the chert core
which suggests the nodule formed prior to and the white chalcedonic nodule, there is no
aragonite dissolution. Silica cementation certain- apparent difference between the two textures
ly took place before burrow compaction (al- under crossed nicols, either in the type of
though the burrows do have a tendency to be cementation or in the cement/detrital quartz
slightly oval in cross-section, the longer diameter ratio. Under plane polarized light, however, a
being horizontal). However, the observation of very slight difference in refractive index is
96 G. A. Carson

F16. 15. SEM photograph of a siliceous nodule. Chalcedony cement (C) coats
detrital quartz (Q). The surface expression of the cement is seen as botryoids
(B). Note the presence of secondary porosity (S) and the preservation of
primary porosity (P). Wilmington sands (300 cm below the Humble Point
Hardground), Wilmington. 0.2 mm scale bar.

observed between the cement and quartz grains. In this study, the difference between the (~180 of
This difference is greater in the nodule than in the water and that of the silica is taken as being
the chert, where the two phases cannot be approximately the same as 1000 ln~ (Friedman &
distinguished. The most likely cause for the O'Neil 1977). A pre-glacial sea water value of
decreased refractive index of the chalcedony -1%o SMOW (Shackleton & Kennett 1975) is
cement is the presence of water inclusions (Folk assumed for the Cretaceous. The full list of
& Weaver 1952), probably due to a rapid results is given in Table 1.
precipitation of the cement (see below).
Another form of silicification is the replace- Silicified fossils
ment of carbonate within a hardground, seen
only in the Small Cove Hardground. This is also The isotopic values (Table 1) are related to the
found in association with Thalassinoides bur- type of replacement and do not appear to
rows, but is equally pervasive into the sediment. correlate directly with the fossil taxa. For
This occurrence will not be further dealt with example, brachiopod shells may be replaced by
here. two (most often mutually exclusive) forms of
silica--a very pure white variety often with a
vitreous lustre and, more frequently, a patchy
grey saccharoidal variety which on close inspec-
Isotopic results tion shows faint beekite rings. Both of these are
the 'fine-scale' replacement described by
As the quartz-water isotopic fractionation of Holdaway & Clayton (1982). The average values
oxygen is inversely proportional to temperature for these replacements are 32.4 + 0.06 (n = 2)
(Knauth & Epstein 1975), the 6180 value of the and 29.2%o + 1.28 (n = 2) respectively, indicating
silica reflects the relative temperature of silica that the two varieties formed in different
precipitation and the isotopic composition of the environments.
pore waters. Here, the empirical equation quoted In Exogyra a similar pattern emerges, al-
in Knauth & Epstein (1976) for quartz-water though in this case the different replacement
fractionation at low temperatures (extrapolated fabrics tend to occur in the same shell: frequent-
from the high temperature (>200~ data of ly, prominent pink beekite occurs near the umbo
Clayton et al. 1972) is used to calculate forma- and a grey fine-scale replacement occupies the
tional temperatures. rest of the shell, which may often show beekite
structures. Even though these two phases may be
1000 ln,cw.~v,~ = 3.09(106 T- 2) _ 3.29. within one specimen, the 61 s o values are similar
 Silicification fabrics from the Cenomanian of Devon 97

~S

.=b

:~~898~ 8~
+ I I I + 777777777T ?+

o o o o:.= ~r~ ~= o o o o o o o o o o o~= o o o o o o o o o o o o o o o o o o o o o o o o o o~

. . . . ~.~ r .'-~ ."~ .-~ .'-~ ."=~ .-~ .-~ .-~ . ~ .-~ .-~ ~ .,-~ ..-~ .,~ .,-: .~.,-:..-:~..--: ..-: ..--:_ ..-: .--: . , ~ . , ~ . , - ~ .,-:.,~ .,.-~..-=..-:~ ~ ~ ~ ~ ~ ~ ~

ii
@-~
,~ ~:'888 ~~ 7 ,~-,-=~,.o~,.~-oo-..o ~'',
o c~ ~ o
,~=,~~~ ~~ ~=~ ~-~,.,~~,,~, ~ ~,..,
~:~, ; ~ ~ i~..~ r ~ ~.~
~ ~ r ~I.'~ ~ .~'~ ,..~ ;>, ,.~ ~-, ~ .~

0 0 ~
~ , I::~, (m 0 0

.<
[r
98 G. A. Carson

FIG. 16. Silicified pectinid. A grey fine-scale saccharroidal replacement (with

a vague concentric structure) is truncated by prominent, white beekite.
Loose, Wilmington. 2 mm scale bar.

to the equivalent phases in brachiopods, analysis of each phase was carried out, the results
31.6 + 0.27 (n = 2) and 29.4%0 + 0.87 (n = 5) are comparable with those of the Exogyra and
respectively. Figure 16 shows a replaced pectinid brachiopods--the earlier replacement having a
(spec. L9) which is composed of only two types of 6180 of 30.6, the later one being 28.7%0.
silica--white porcellanous beekite rings, ~ 1- Further analysis was carried out on a different
2 mm diameter, and a later fine-scale saccharroi- pectinid (spec. L24--see above). In this speci-
dal grey variety, which on close inspection also men, the three distinct types of replacement were
possesses concentric rings. Though only one sampled (Fig. 9) and it was possible to separate
individual zones within a beekite 'disc'
(Holdaway & Clayton 1982). The 6180 values are
given in Table 2 and do not show any consistent
TABLE 2. Detail of isotopic results from pectinid spatial variation within individual structures.
L24. Lettered key refers to Fig. 9 Thus the growth of the rings would have been
6180 average la rapid relative to the burial of the specimen. The
average values of each of the two discs analysed
Individual A 28.6 in detail are both similar to values obtained from
beekite rings-- B 30.7 'whole-rock' ring analysis. Even though petro-
29.7 0.96 graphic evidence shows that different beekite
from exterior to C 29.3
interior D 30.3 disks have grown sequentially (Holdaway &
E 29.3 Clayton 1982), they must all have completed
F 30.2 their growth quite rapidly since they all possess
G 26.6 28.1 1.86 approximately the same 6180 values.
H 26.5 The average of all the beekite analyses from
Whole beekite I 29.9 sample L24 is 29.3%o_+ 0.79 (n = 4 ) which is
'discs' J 29.3 similar to the fine-scale grey replacement of
previously mentioned specimens. The later grey
Beekite average 2 9 . 3 (+0.79) replacement in sample L24 has an average value
Late grey K 27.7 of 27.6%0 + 0.36 (n = 3) and the very soft pure
replacement L 27.9 27.6 0.36 white replacement has an average value of
M 27.2 30.5%0 + 0.42 (n = 3). The textural relationship
shows the grey replacement post-dating the
N 30.8 beekite, which in turn post-dates the soft, white
Soft, white O 30.0 30.5 0.42
beekite P 30.5 replacement. Why the absolute values should be
offset from other specimens remains uncertain.
 Silicification fabrics from the Cenomanian of Devon 99
Siliceous nodules

The isotopic results from the chert core and the

Early repl.
exterior of a siliceous nodule are very similar
(24.4, 24.0 and ---23.0%0 respectively). Both n
phases contain detrital quartz with a mean value
of 11.5%o + 0.48 (n = 2), typical for quartz der-
ived from acid igneous sources (Taylor 1968, fig.
8) in this case, probably the Permian granites to
the north-west (although see discussion in Han-
cock 1969). In both the chert and the nodule,
quartz grains occupy ~ 35% of the rock. So on
adjusting the initial values, we get 3180 values of Late repl.
29.3 and 30.9 for the nodule, and 31.5%0 for the
chert in specimen 436, and a value of 34.6 for the
nodule L22B. Where chalcedony cement was n
isolated from a nodule (specimen 462), it gave a
mean value of 29.7 + 0.95%0 (n = 2).

0
26 3O
Discussion 8'80 sMow
Silicified fossils FIG. 17. Histograms of 6tso values obtained for
early and late replacements. Where duplicate
Except for specimen L24, the mean 6180 values analyses were made from one specimen, the mean
(early: from 30.6 to 32.4%0, late: from 28.7 to value was used. Stippled area indicates values
29.4%0) give temperatures of 17.2-24.4~ for the obtained from specimen L24.
earlier replacements and 30.0-32.9~ for the
later ones (using a sea water 6180 SMOW of
- 1%0). Lowenstam (1964) calculated Cenoman-
ian sea water temperatures varying from 17 to However, the values for the later replacement
29~ (depending upon latitude) assuming a sea (mean = 28.9%0 + 1.02 (n = 12)) pose a problem
water 6180 SMOW of 0 (Lowenstam & Epstein (Fig. 17). Consistently lower values were ob-
1954). Recalculation of Lowenstam's data using tained for the later replacement. The 6180
a sea water value of -1%o, gives a temperature difference would correspond to a temperature
range from 13 to 25~ These figures were increase of ~ 10~ with a minimum increase of
calculated on data from belemnites which are 8.5~ between the early and late phases in one
nektonic organisms, and thus would reflect fossil group. Four possibilities may be invoked to
temperatures at some distance above the sedi- explain this:
ment surface in the water column. However,
Cretaceous sea water can be assumed to have had
little vertical variation in temperature (due to (1) Formation or recrystallization at greater depth
lack of cold bottom currents), and thus the
temperature reflected by the belemnites would From the discussion of burial depth, it seems
be comparable to that at the sediment-water unlikely that these sediments were buried more
interface. than ~260 m. Using an average geothermal
gradient of Y35~ (measurement from
The presence of a sandy facies in a predomin-
Dorset--Bloomer et al. 1979), the temperature
antly carbonate succession inland at Wilmington
increase due to burial can only have been a
suggests a marginal environment, although it is
maximum of 8.7~ Thus the estimate of maxi-
possible that early Cenomanian local faulting to
m u m possible overburden ( ~ 260 m) can only just
the west contributed some of the sandy material
account for the minimum calculated temperature
(Jarvis & Tocher 1987). This, along with the
increase, 8.5~ (equivalent to 250 m of burial).
degree of condensation of the Cenomanian on
the present coast (implying a shallow water
environment), would further support a vertically
(2) Formation in a mixing zone
homogeneous water temperature. Thus the data
given for the earliest silica replacement are Knauth (1979) suggested the formation of chert
consistent with the idea of replacement at a in an environment where the mixing of sea water
shallow burial depth. and meteoric water would produce a solution
IO0 G. A. Carson

undersaturated with respect to calcite but super- further work on the associated carbonate phases
saturated with respect to quartz. This idea could might prove this to be a viable explanation.
be applied to the silicification of bioclasts and
would explain the light 6180 values. However,
Siliceous nodules
even though Wilmington represents a marginal
environment, evidence for such a mode of The ~180 values for the siliceous nodules and
formation (such as the association with dolomite cherts give an average temperature of 22.1~
formation) is totally lacking (cf. formation of (n = 5) (assuming a sea water value of -1%o)
spherulites, Elorza & Orue-Etxebarria 1985). which corresponds very well to values given for
the early fossil replacement. However, the
similarity of isotopic results between the chert
(3) Late recrystallization with &otopically light
core and chalcedony cemented margin is perhaps
meteoric waters
surprising in view of the petrographic evidence
Clayton (1984) illustrates that light 6180 values cited earlier, which suggests a difference in water
of certain types of flint are due to a combination content between the interior and exterior. There
of inversion of opal-CT to quartz at considerable are two possibilities to explain this difference.
depth and recrystallization with meteoric waters The first is that the siliceous nodule originally
during uplift. Even though they are all varieties crystallized as chert (as in the core) and that some
of hydrated silica, the distinctly different mor- later mechanism has exsolved the water on the
phological forms of silica replacement at Wilm- margins of the chert. Three possible mechanisms
ington suggest that one type (the later grey are: recrystallization at depth, recrystallization
replacement) may have been more susceptible to in meteoric waters or an initial faster precipita-
recrystallization than the others. Thus the unex- tion rate. Recrystallization at depth is unlikely
pected (~180 values of late replacement may not since it has already been demonstrated that these
reflect its original value but a reduced value due cherts could not have been buried to any great
to recrystallization. The question remains as to depth. Clayton (1984) has suggested recrystalli-
whether this took place at depth during burial or zation of the cortex of flint may be 'seeded' by an
in a meteoric water environment. influx of meteoric water. Yet, despite the effect
Meyers & James (1978) noted that even though of a possible meteoric phreatic phase on the fossil
the petrography suggested that many silica replacements, it is unlikely to have affected the
cements and replacements were precipitated nodules since the oxygen isotope values have not
early, the light 6180 values implied recrystalliza- been affected in any way. Thus we are left with
tion in the presence of phreatic groundwaters. the hypothesis that even though the precipitation
This was the case for the chalcedony, which of the cement has to be fairly rapid anyway (to
originally would have been in the form of banded form hydrated silica), the exterior precipitated
opal-CT (Meyers 1977), and the microquartz, somewhat more rapidly than the centre of the
which contain microspherules interpreted to be chert. Thus the silica in the exterior was slightly
relics of cristobalite lepispheres (Meyers 1977). more structurally disordered than that on the
The fabrics observed in the present study show interior, with slightly higher concentrations of
no signs of lepispheric structures, but since entrapped pore waters and was more prone to
isotopic exchange of the silica by atomic diffu- recrystallization. If this was so, the recrystalliza-
sion is unlikely, as is re-equilibration of opal-CT tion must have occurred at an early stage to
(at less than 1000 m--Knauth & Epstein 1975) or preserve the relatively heavy 6180 values. This is
quartz (Taylor 1968), it could be that the contrary to the formation of flints where silicifi-
replacement silica has undergone some sort of cation is believed to be more intense (and thus
recrystallization, possibly from an opal-CT rapid) towards the centre (Clayton 1984). How-
precursor. ever, flint formation in the chalk is initiated by
The recrystallization probably took place lepisphere growth so a diffusion gradient would
during uplift, after the end of the Cretaceous, be set up with silica nucleation centres more
when the Wilmington sands would have been closely spaced towards the centre of the flint
subjected to (phreatic?) meteoric water input. (Clayton 1984). In the sands, there is no evidence
for a lepispheric initiation of chalcedony precipi-
The geochemical data and petrographical tation and it is more likely that the initial zone of
observations do not conclusively point to any one precipitation is around the circumference of a
of the three possibilities, yet a tentative sugges- burrow than within it. This would also explain
tion is that recrystallization plays a role in why some of the burrows remain hollow.
explaining the light oxygen values. However, it Recrystallization during deeper burial within a
must be noted that formation within a mixing closed system (Pingitore 1982), preserving the
zone was only rejected on negative evidence-- heavy 6180 values, is unlikely because the sandy
 Siliqification fabrics from the Cenomanian of Devon I 01

facies would probably have allowed free Conclusions

porewater circulation.
The second possibility is that later exsolution The petrographic and isotopic evidence above
of structural water need not have occurred, the suggests that the replacement of bioclasts was
initial precipitation rate being rapid enough to essentially a very early diagenetic process, where
trap fluid in inclusions during growth. However, the restricted microenvironment within shells or
this would fail to account for the constant their fragments would initiate silicification (see
thickness of the white exterior of the nodules, Holdaway & Clayton !982). However, the later
unless the rate of growth and silica supply was replacements were more prone to a partial
the same vertically and horizontally throughout recrystallization which tOOk place much later in
the succession. the diagenetic history of the sediment, in a
meteoric phreatic environment. The formation
Drusy quartz in silicified fossils of drusy quartz tOOk place in this same environ-
ment. Siliceous nodules formed penecontempor-
All the drusy quartz analysed has a much lighter aneously with (or perhaps slightly earlier than)
value (average = 24.2%0 + 0.94 (n = 3), than the early fossil replacement. There is uncertainty as
enclosing replaced shell material. Assuming to whether the exterior of the nodules grew at a
formation from marine waters with a 6180 of faster rate than the interior or recrystallized at an
-1%o, the values for the drusy quartz give a early stage during the burial of the sequence.
formation temperature of 56.5~ If they had
formed from the same waters there would be a
34~ difference between the lightest fossils and ACKNOWLEDGMENTS:This work could not have been
the drusy quartz. This implies a difference in undertaken without the tuition on the fluorination line
depth of formation of over 1000 m and we have given by Peter Greenwood and Linda Thrift (BGS,
already calculated a maximum burial depth of London), and their preparatory work for optimizing
~ 260 m. The most likely explanation for these running conditions appropriate to these samples. Their
values is that the quartz formed from isotopically constant guidance and many fruitful discussions were
light meteoric waters in a phreatic environment an inspiration to this study. My gratitude also goes to
Jim Marshall and Chris Paul for their criticism and
(Meyers & James 1978). Late recrystallization help in increasing the eloquence of the text and to Chris
can be excluded since mega-quartz is believed to Clayton, Tim Astin and Baruch Spiro for their
be isotopically immune to groundwater exchange constructive reviews. Financial support was provided
(see evidence cited in Knauth & Lowe 1978, p. by NERC and the isotopic data are published with the
214). permission of the Director, British Geological Survey.

References
ANDERTON, R., BRIDGES, P. H., LEEDER, M. R. & BUURMAN,P. & VAN DER PLAS,L. 1971. The genesis of
SELLWOOD,B. W. 1979. A Dynamic Stratigraphy of Belgian and Dutch flints and cherts. Geologie en
the British Isles: a Study in Crustal Evolution. Allen Mijnbouw, 50, 9-28.
& Unwin, London. CLAYTON,C. J. 1982. Growth history and microstruc-
BLACK, i . 1953. The constitution of the Chalk. ture of flint. International Association of Sedimento-
Proceedings of the Geological Society of London, logists 3rd European Meeting, Copenhagen (Ab-
1499, 31-6. stract), pp. 105-7.
BLOOMER,J. R., RICHARDSON,S. W. & OXBURGH,E. R. --1984. The geochemistry of chert formation in
1979. Heat flow in Britain: an assessment of the Upper Cretaceous Chalks. PhD thesis, University
values and their reliability. In: CERMAK, V. & of London.
RYBACH,L. (eds) Terrestrial Heat Flow in Europe, CLAYTON, R. N. • MAYEDA,T. K. 1963. The use of
pp. 293-300. Springer-Verlag, Berlin. bromine pentafluoride in the extraction of oxygen
BROWN,G., CATT, J. A., HOLLYER,S. E. & OLLIER,C. from oxides and silicates for isotopic analysis.
D. 1969. Partial silicification of chalk fossils from Geochimica et Cosmochimica Acta, 27, 43-52.
the Chilterns. Geological Magazine, 106, 583--6. - - , O'NEIL, J. R. & MAYEDA, T. K. 1972. Oxygen
BRUNTON, C. H. C. 1966. Silicified productoids from isotope exchange between quartz and water.
the Vis6an of County Fermanagh. Bulletin of the Journal of Geophysical Research, 77, 3057-67.
British Museum (Natural History), Geology, 12, CURRY, G. B. 1986. Fossils and tectonics along the
173-243. Highland Boundary Fault in Scotland. Journal of
--1968. Silicified brachiopods from the Vis6an of the Geological Society of London, 143, 193-8.
County Fermanagh (II). Bulletin of the British ELORZA, J. & ORUE-ETXEBARRIA,X. 1985. An example
Museum (Natural History), Geology, 16, 1-70. of silicification in Gryphaea sp. shells from Lafio
-- 1984. Silicified brachiopods from the Vis6an of (south of Vitoria, Spain). 6th International Associ-
County Fermanagh (III). Bulletin of the British ation of Sedimentologists Regional Meeting, Lleida
Museum (Natural History), Geology, 38, 27-130. (Abstract), pp. 556-9.
IO2 G. A. Carson

FAUGI!RES, J.-C. 1978. Les Rides sud-rifaines. Evolu- billion years), Transvaal, South Africa, with
tion sedimentaire et structurale d'un bassin atlan- implications for secular variations in the isotopic
tico-mesogen de la marge africaine. PhD thesis, composition of cherts. Earth and Planetary Science
Part 3: Sedimentology. University of Bordeaux. Letters, 41, 209-22.
FOLK, R. L. & PITTMAN, J. S. 1971. Length slow LOWENSTAM, H. A. 1964. Paleotemperatures of the
chalcedony: A new testament for vanished eva- Permian and Cretaceous periods. In: NAIRN, A. E.
porites. Journal of Sedimentary Petrology, 41, M. (ed.). Problems in Paleoclimatology, pp. 227-
1045-58. 48.
- - & WEAVER,C. E. 1952. A study on the texture and - - & EPSTEIN, S. 1954. Paleotemperatures of the
composition of chert. American Journal of Science, post-Aptian Cretaceous as determined by the
250, 498-510. oxygen isotope method. Journal of Geology, 6 2 ,
FRIEDMAN, I. & O'NEIL, J. R. 1977. Compilation of 207-48.
stable isotope fractionation factors of geochemical MEYERS, W. J. 1977. Chertification in the Mississip-
interest. In: FLEICHER,M. (ed.)Data of Geochemis- pian Lake Valley Formation, Sacramento Moun-
try, Geological Survey Professional Paper 440- tains, New Mexico. Sedimentology, 24, 75-105.
KK. - - & JAMES,A. T. 1978. Stable isotopes of cherts and
H~ANSSON, E., BROMLEY, R. & PERCH-NIELSEN, K. carbonate cements in the Lake Valley Formation
1974. Maastrichtian chalk of north-west Europe-- (Mississippian), Sacramento Mts, New Mexico.
a pelagic shelf sediment. In: Hsf2, K. J. & Sedimentology, 25, 105-24.
JENKYNS, H. C. (eds) Pelagic Sediments: on Land PINGITOREJR, N. E. 1982. The role of diffusion during
and Under the Sea. Special Publication of the carbonate diagenesis. Journal of Sedimentary Pet-
International Association of Sedimentologists, 1, rology, 52, 27-39.
211-33. Blackwell Scientific Publications, Oxford. RAWSON, P. F., CURRY, D., DILLEY, F. C., HANCOCK,
HANCOCK,J. M. 1969. Transgression of the Cretaceous J. M., KENNEDY, W. J., NEALE, J. W., WOOD, C.
sea in south-west England. Proceedings of the J. & WORSSAM, B. C. 1978. A correlation of
Ussher Society, 2, 61-83. Cretaceous rocks m the British Isles. Special Report
-- 1975. The petrology of the Chalk. Proceedings of of the Geological Society, 9.
the Geologists' Association, 86, 499-535. SCHMITT, J. G. 8~ BOYD, D. W. 1981. Patterns of
HOLDAWAY, H. K. & CLAYTON, C. J. 1982. Preserva- silicification in Permian pelecypods and brachio-
tion of shell microstructure in silicified brachio- pods from Wyoming. Journal of Sedimentary
pods from the Upper Cretaceous Wilmington Petrology, 51, 1297-308.
Sands of Devon. Geological Magazine, 119, 371- SHACKLETON,N. J. & KENNETT, J. P. 1975. Paleotem-
82. perature history of the Cenozoic and the initiation
JARVIS, I. & TOCHER, B. A. 1987. Field meeting: the of Antarctic glaciation: oxygen and carbon iso-
Cretaceous of S.E. Devon, 14-16th March 1986. tope analysis in DSDP sites 277, 279 and 281.
Proceedings of the Geologists' Association, 98, 51- Initial Reports of the Deep Sea Drilling Project, 24,
66. 743-55.
- - & WOODROOF, P. B. 1984. Stratigraphy of the SMITH, W. E. 1957. The Cenomanian Limestone of the
Cenomanian and basal Turonian (Upper Creta- Beer District, south Devon. Proceedings of the
ceous) between Branscombe and Seaton, S.E. Geologists" Association, 68, 115-35.
Devon, England. Proceedings of the Geologists' --1961. The Cenomanian deposits of south-east
Association, 95, 193-215. Devonshire. Proceedings of the Geologists' Associ-
KNAUTH, L. P. 1979. A model for the origin of chert in ation, 72, 91-134.
limestone. Geology, 7, 274-7. TAYLOR, H. P. 1968. The oxygen isotope geochemistry
- - & EPSTEIN,S. 1975. Hydrogen and oxygen isotope of igneous rocks. Contributions to Mineralogy and
ratios in silica from the JOIDES DSDP. Earth and Petrology, 19, 1-71.
Planetary Science Letters, 25, 1- l O. VAN HINTE, J. E. 1976. A Cretaceous time scale.
- - & -- 1976. Hydrogen and oxygen isotope ratios Bulletin of the American Association of Petroleum
in nodular and bedded cherts. Geochimica et Geologists, 60, 498-516.
Cosmochimica Acta, 40, 1095-1108. WILSON, R. C. L. 1966. Silica diagenesis in Upper
-- & LOWE, D. R. 1978. Oxygen isotope geochemis- Jurassic limestones of southern England. Journal
try of cherts from the Onverwacht Group (3.4 of Sedimentary Petrology, 36, 1036-49.

GREG A. CARSON, Department of Geological Sciences, University of Liverpool, Brownlow

Street, PO Box 147, Liverpool L69 3BX, UK.
 Controls on the geometry and distribution of carbonate cements
in Jurassic sandstones: Bridport Sands, southern England and
Viking Group, Troll Field, Norway

J. D. Kantorowicz, I. D. Bryant & J. M. Dawans

S U M M A R Y: The petrography and diagenesis of calcite cements in the Lower Jurassic,

Bridport Sands (southern England) and Upper Jurassic, Viking Group sandstones (Troll
Field, offshore Norway) have been investigated in order to assess their geometry and effect
on hydrocarbon recovery.
In the Bridport Sands, sediment texture and mineralogy controlled carbonate cementa-
tion. Clay-rich fairweather sediments were weakly cemented and are now compacted.
Bioclast-rich storm deposits were stabilized mechanically by early fringing cements. During
burial bioclasts and fringing cements were replaced or dissolved, and pores were filled by
simultaneously precipitated ferroan calcite. Thus, cemented beds are laterally continuous for
several kilometres in the Bridport Sands as a consequence of the sheet-like geometry of the
storm beds in which they developed.
In the Viking Group sandstones, carbonate cementation was controlled by rate of burial.
Fringing cements formed locally during non-deposition or emergence. Cementation
continued with non-ferroan calcite incorporating bacterially derived bicarbonate generated
during prolonged residence in near surface zones of bacterial activity. Cement geometries
will reflect the distribution of emergent surfaces and the longevity of residence near the
surface.
These two cases demonstrate the potential for laterally extensive carbonate cements to
develop in shelf sandstones. The cements in these examples have different origins but in both
cases their distribution is related to the episodic nature of deposition in the shelf
environment.

Carbonate minerals of diverse origin occur as any controls on its distribution. This in turn
cements formed during the diagenesis of shallow- necessitates understanding the host sediment's
marine sandstones. Cement origins include pre- depositional, diagenetic and burial histories.
cipitation from sea water, the dissolution and This paper describes two case studies of carbon-
reprecipitation of bioclastic aragonite and cal- ate cementation in shallow-marine sandstones.
cite, and precipitation of bicarbonate liberated In the Bridport Sands carbonate cements are
during bacterial processes (see Hudson 1977). known to be laterally extensive, whilst in the
Precipitation has been recognized to occur Viking Group sandstones in the Troll Field their
rapidly on the seafloor (Nelson & Lawrence geometry is unknown.
1984, H~vland et al. 1985), during burial (Milli-
ken et al. 1981), and at depths of 3 km following Bridport Sands, onshore Dorset, UK
hydrocarbon emplacement (Kantorowicz 1985).
The effect of carbonate-cemented sandstones The Upper Liassic (Lower Jurassic) Bridport
on hydrocarbon reservoir development depends Sands outcrop in Dorset as spectacular cliffs
upon cement geometry. Nodular, laterally dis- between Bridport and Burton Bradstock. Inland
continuous cemented sands are only likely to the Sands have been cored in a number of
influence fluid flow locally, if at all, whilst boreholes, and form part of the reservoir in the
laterally continuous cemented sandstones may Wytch Farm oil field (Fig. 2a; Colter & Havard
present 'baffles' to fluid flow within a reservoir or 1981). The Bridport Sands were formed as a
divide (or 'compartmentalize') the reservoir. shallow marine bar (Davies 1967). Outcrop and
Whether this is beneficial or detrimental de- borehole studies support this interpretation, and
pends upon the orientation of the cemented enable a more detailed model to be proposed.
sandstone and the nature of the proposed Throughout southern Britain the Bridport Sands
development programme (Fig. 1). Thus, when comprise shallowing-upwards cycles of shelf
carbonate-cemented sandstones are encountered sands, capped locally by thin limestones. Each
in the subsurface it is desirable to be able to cycle contains fine to very fine grained clay-rich
predict their distribution. This requires an sandstone beds and fine grained bioclastic
understanding of the origin of the cement and sandstone beds. Clay-rich beds predominate at

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, IO 3

Geological Society Special Publication No. 36, pp. 103-118.
[04 J. D. Kantorowicz et al.
In ector Producer

(a)
Injector Producer

iiiiiiiiiiiiii ii!i!i!i:!i;i;i!i
FIG. 1. (a) Cemented sand hindering development by isolating 20~ of the oil zone in a separate accumulation.
Accelerated water-coning is also likely as a result of the orientation of the tight streak. (b) Cemented sand
aiding development by reducing possibility of bottom-water coning into the larger upper oil accumulation.
However, note reduced communication between the oil- and water-saturated parts of the reservoir.

the base, enclosing occasional thick bioclastic taining abundant silt-sized quartz grains and up
beds. The bioclastic sands become thinner and to 26~ detrital clay. The framework is matrix
more abundant towards the top of each cycle. supported. A variety of bioclasts was observed in
Most cemented beds are continuous over 90~ of the core but only belemnites were identified at
the 5 km long cliff section. Based on a combina- outcrop.
tion of sedimentological, palynological and ich- The sediments have undergone various dia-
nological data the clay-rich beds are interpreted genetic modifications. Pyrite, often in associ-
to have been deposited on the shelf during ation with organic matter, occurs as a trace
fairweather processes and the bioclast-rich beds percentage throughout and is the only authigenic
during storm processes. mineral present where abundant detrital clay
occurs. Where less detrital clay occurs, well-
developed grain-coating berthierine has formed.
Petrography Quartz overgrowths occasionally engulf berthier-
The Bridport Sands contain a wide variety of ine, but usually are developed only where
detrital and authigenic minerals, modal analysis berthierine is absent (Fig. 3). Part of the
of which describes a gradation between porous, remaining pore space was filled subsequently
clay-rich fairweather sediments and carbonate- with either ferroan calcite or ferroan dolomite.
cemented clay-poor bioclast-rich storm deposits Many of the bioclasts observed in thin section
(Table 1). In general, the fairweather sediments have undergone extensive grain interpenetration
are weakly cemented, and the storm deposits and dissolution.
tightly cemented. The pronounced differences
between them observed at outcrop may be
attributed to the effects of surface weathering Cemented sands
(Fig. 2b).
These are texturally mature, fine- to very fine-
grained bioclast-rich sandstones with a clast-
Clay-rich sands supported framework. The bioclasts present
These are porous, very fine- to fine-grained include brachiopod and bivalve shells which
subarkoses. They are texturally immature, con- contain extensive iron oxide-lined algal borings,
 Carbonate cements in Jurassic sandstones Io5

CHELTENHAM9I~ LEGEND:
GLOUCEST~S ~OUTCROP
D OIL FIELD " /~ ./ o ,c
4~ 80*E.OLE ~ MILE~

DURSLEY.,..1r
N

0 IOMILES
c.4,0/~ BATH I I

~
tli~ SCALE

,~- KNOLL MENDIPAXIS

GLAS~TTOoN.R~U~ SOUTHAMPTON

~ WYT~
(a) eNGLISH ~) CHANNEL

FIG. 2. (a) Location map and outcrop of the Upper Lias of Southern England. (b) Outcrop of the Bridport
Sands at Bridport, Dorset. Carbonate cemented sands are hard and resistant to weathering, and so stand out
relative to the more easily weathered uncemented sands.

gastropods, echinoderm plates, forams, ammon- The diagenetic history of these sediments is
ites, belemnites and bryozoa. Berthierine ooliths complex with widespread evidence of possibly
are common in the Marchwood Borehole but concomitant carbonate precipitation and disso-
were not observed in outcrop samples. lution (Fig. 4). Post-depositional modifications
IO6 J. D. Kantorowicz et al.
TABLE 1. Petrographical characteristics of the Bridport Sand
Cemented Clay-rich

Average clastic grain size Fine Fine

Sorting Good Moderate
Clastic grain to matrix ratio 3-4 : 1 2 :1
Texture Mature, clast supported Immature, matrix supported
Compaction effects None observed Deformed micas throughout
Porosity Mouldic and shrinkage Intergranular and
within ooliths microporosity

Typical modal mineralogy

Quartz and feldspar 20.0 46.33
Mica 0.33 7.33
Rock fragments 0.67 1.00
Bioclasts 40.33 8.0
Berthierine ooliths 8.33 --
Detrital clay 7.67 23.33
Calcite cement 21.67 3.67
Dolomite cement -- 3.0
Quartz overgrowth -- 1.33
Pyrite -- 1.33
Porosity 1.0 4.67

FIO. 3. Scanning electron photomicrograph, Bridport Sands, Marchwood-1, 3907" 2'. Grain-coating
berthierine locally inhibits quartz overgrowth. Scale bar = 10 ~tm.

began with widespread algal and sponge boring luminesce nor fluoresce. The cementation his-
and micritization (Fig. 5a). The bioclasts subse- tory is complicated because of the uncertain
quently formed nuclei for non-ferroan calcite timing of dissolution of aragonitic and high-Mg
fringing cements and overgrowths (Fig. 5b). The calcite bioclasts and early cements (Fig. 5c, d).
remaining intergranular porosity was cemented There is widespread evidence of grain interpene-
with a mosaic of blocky and occasionally tration and dissolution, whilst mouldic porosity
poikilotopic ferroan calcite. Where fewer bio- also often occurs. These mouldic pores may be
clasts occur, poikilotopic calcite becomes more lined with euhedral calcite rhombs growing
abundant. Neither fringing nor mosaic calcites inwards from the preserved micritic envelope, or
 Carbonate cements in Jurassic sandstones IO7

BIOCLAST-RICH SAND ]
I
Algal colonisation.
Oxidation of detrital iron oxides.

1 AND MICRITIZATION
i
Calcite precipitation from
I
interstitial depositional
pore waters.

FRINGING NON FERROAN CALCITE CEMENT

AND OVERGROWTHS ON BIOCLASTS

Increased overburden
pressu res.

I
Bioclast dissolution

MOULDIC POROSITY

Calcite precipitation incorporating

locally derived calcite.

T
Neomorphism of bioclasts,
overgrowths and early cements.
J

FERROAN CALCITE FILLING MOULDIC

AND INTERGRANULAR POROSITY, AND
REPLACING BIOCLASTS
1
Fracturing

I VEIN FILLING FERROAN CALCITE

FIG. 4. Flow chart of diagenetic modifications to cemented sands in the Bridport Sand.

completely engulfed by blocky calcite (Fig. 5e). Interpretation

Finally, much of the aragonite or high-Mg calcite
The carbon isotopic composition of all the calcite
was replaced by ferroan calcite during diagenesis
(Fig. 50. Isolated ferroan dolomite rhombs occur and dolomite analysed from isolated cements
and bulk samples is indicative of an originally
occasionally. Ferroan calcite cemented fractures
marine (sea water) origin (513C+2%o), with
occur in many cemented beds but the calcite has
minor bacterial influence possibly shifting the
not penetrated into intergranular pore space.
613C values towards -2%0 (see Hudson 1977).
The oxygen isotopic composition, however, is
indicative of precipitation at more elevated
Carbon and oxygen stable isotope analysis temperatures than expected near the sediment
surface and presumably reflects the composition
The results of stable isotope analysis are pre- of the paragenetically later and replacive ce-
sented in Fig. 6 and Table 2. ments rather than early fringing cements. The
108 J. D. Kantorowicz et al.
FIG. 5. Thin section photomicrographs, Marchwood-1 borehole. (a) Fine grained sandstone, with abundant
bioclasts. The bioclast B has been bored and micritized by algae and sponges. MWD-1, 3885'6". Scale
bar = 175 ~tm. (b) Bioclastic sandstone with a well developed non-ferroan calcite overgrowth (stained pink)on
an echinoderm plate, E, fringing cements on the other bioclasts present, and oxidized borings in places.
MWD-1, 3871'6". Scale bar = 175 ~m. (c) Fine grained sandstone with non-ferroan bioclasts in various stages
of diagenetic modification. Most bioclasts are fringed with calcite cement. The heavily bored bioclast B has
been dissolved leaving goethite lined borings within the mouldic porosity. MWD-1, 3885'6". Scale
bar = 175 pm. (d) Fine grained bioclastic sandstone. Grain interpenetration has resulted in quartz grains
penetrating through the dissolved bioclast. MWD-1 3861'11". Scale bar = 175 ~tm. (e) Ferroan calcite cements
this bioclast-rich sand. It occurs as an intergranular cement and within mouldic porosity in this case
precipitating after dissolution of the bored bioclast B. Only the goethite filled borings and micritized envelope
remain. MWD-1, 3871'6". Scale bar = 140 pm. (f) Bioclastic-rich sandstone. The echinoderm plate E and
overgrowth O have recrystallized to ferroan calcite whilst the non-ferroan brachiopod B and overgrowth have
not. MWD-1, 3871'6". Scale b a r = 140 ~tm.

818 0 ~DB

-5
I I I 1 813C PDB

INFLUENCE OF BACTERIAL
HCO3 (~13C -25) ~ IDEAL JURASSIC
SEAWATER
PRECIPITATES

BURIAL AND
D RECRYSTALLIZATION

C
-5
D
D C D

C
F
s /F
B BIOCLAST
F FRACTURE FILL CALCITE

F F S SPARRYCALCITE WITHIN BIOCLAST

F C CALCITECEMENT
F F
D DOLOMITECEMENT

FIG. 6. Carbon and oxygen stable isotopic compositions of calcite and dolomite from Bridport Sand. The
isotopic composition of calcite cement (613C 0 to - 1 . 5 and 6180 - 4 . 5 to -6.5) has been modified from a
probable Jurassic sea water composition by two processes. Bacterially derived HCO 3- with a &~3C
composition of - 2 5 has diluted the original marine bicarbonate to more negative values. Dissolution and
reprecipitation, and recrystallization during burial have led to a more negative shift in 5180 values. The
dolomite compositions (513C + 0.5 to - 1.5 and 5180 - 4 . 0 to -6.0) possibly reflect the release of magnesium
during calcite recrystallization.

shift from an assumed 6~80 -1.2%o ( S M O W ) of the following: (a) precipitation of the later
Jurassic sea water (Shackleton & K e n n e t t 1975) c e m e n t s during b u r i a l - - t h e m a x i m u m burial
w h i c h would have precipitated - 1 to -2%0 temperatures being 80-100~ (average vitrinite
(PDB) fringing cement, towards - 4 to -6%0 reflectance R0 = 1.12), (b) the influence of the
(PDB) calcite may, therefore, reflect one or more low-salinity (fresh?) g r o u n d w a t e r influx w h i c h
 . .~i, ~ . .,roll t ~ ~ , .~

r !~1~ ,7 .~. ~" 9 ~

 C a r b o n a t e c e m e n t s in Jurassic s a n d s t o n e s IO9

TABLE 2. Carbon and oxygen stable isotopic data" Bridport

Sands
Sample Details Calcite Dolomite

613C 5180 613C 6180

W e s t Bay-2 (1) Bioclast -1.73 -5.14

W e s t B a y - 2 (3) Bioclast -1.78 -5.26
Burton
B r a d s t o c k - 4 (1) V e i n fill - 0.92 - 7.33
,, -4 (2) ,, -0.73 -6.89
,, -2 ,, -0.40 -7.18
,, -1 (1) ,, -0.74 -7.25
H a m d o n Hill-1 (1) ,, - 1.24 - 8.67
,, -1 (2) ,, - 1.50 -9.38
,, -1 (3) ,, -0.54 -8.76
-1 (4) - 1.07 -9.30
Marclawood-1 "
3890' V e i n fill +0.51 -7.59
3890' Cements + 0.25 - 6.48 +0.50 -5.56
3863'11" V e i n fill -0.91 -9.44
3818'9" Cements -0.48 -5.62 -0.13 -4.92
3933'2" ,, -0.45 -6.53 -1.11 -5.51
3838' ,, -0.08 -6.50 -0.35 -5.62
3940'8" ,, - 1.66 -4.71 - 1.46 -4.08
3888'3" ,, +0.12 -6.50 +0.09 -6.64
3864' ,, - 1.40 - 7.07

Edmunds et al. (1982) believe to have modified sediment or its interstitial pore waters and are
the originally marine pore waters, or (c) given the often recorded by the carbon isotopic composi-
palaeoburial temperature and the presence of tions of authigenic carbonates. However, the
liquid hydrocarbons in the Bridport Sands at clay-rich sandstones here are dominated by the
Wytch Farm (Colter & Havard 1981) the non-carbonate products of suboxic iron reduc-
influence of thermally derived CO2 obviously tion. Iron silicates presumably formed instead of
cannot be ruled out. iron carbonates because produced bicarbonates
were able to diffuse out of the system, preventing
the build up of high carbonate alkalinity. Ferrous
Controls on carbonate mineral diagenesis in the iron silicates have low solubility relative to
Bridport Sands calcite and precipitated rapidly in the sediment.
The large crystals which formed reflect the
The cemented beds in the Bridport Sands longevity of suboxic iron reduction. The associat-
contain two generations of carbonate cement: ed pyrite may have formed subsequently, when
fringes or overgrowths on detrital grains and a the sediments were eventually buried into the
pore-filling mosaic. The mosaic cement formed sulphate reducing zone, or when upwardly
by the mobilization and redistribution of detrital mobile H2S rose from below. Any bicarbonate
and authigenic calcite and aragonite within the generated subsequently appears to have been
cemented beds themselves, and from the clay- swamped during cementation of the storm
rich beds into the cemented beds. deposits by the abundant carbonate derived by
dissolution of bioclasts. In summary, the diagen-
etic processes which operated in the fairweather
Berthierine sediments were driven by bacterial activity,
however, this is recorded by the presence of iron
The most favourable conditions for berthierine silicates and iron sulphides rather than carbon-
formation are believed to be prolonged iron ate cements with highly negative ~13C.
reduction in the suboxic zone (Coleman 1985,
Curtis 1985). This allows large crystals to form.
Fringing cements
Most marine organic-rich sediments pass
through oxic, suboxic and anoxic diagenetic Fringing cements are not particularly diagnostic
zones during shallow burial diagenesis (Claypool of any sedimentary environment, requiring only
& Kaplan 1974, Froelich et al. 1979). These are the presence of detrital carbonate grains as
characterized by distinctive reactions between nuclei (e.g. Jorgensen 1976, Nelson & Lawrence
organic matter and various oxidants in the 1984). They are most commonly found in
I Io J.D. K a n t o r o w i c z et al.

firmgrounds and hardgrounds and hence reflect Summary

episodic non-deposition (e.g. Wilkinson et al.
Carbonate cementation in the Bridport Sands
1985). In the Bridport Sands, bioclasts were reflects the original sediment texture. Clay-rich
originally heterogeneously distributed, being fairweather sediments are weakly cemented and
concentrated within the storm deposits. Non- have undergone compaction. Interbedded clay-
ferroan calcite or (? aragonite) fringing cements poor, bioclast-rich storm-deposited sediments
on the bioclasts mechanically stabilized the developed early fringing cements resulting in
sediment framework. Bioclasts in the clay-rich little compaction. Pores in these uncompacted
sediments have occasional overgrowths, but beds were filled by later cementation resulting
these were not concentrated enough to stabilize from redistribution of bioclast-derived carbon-
the framework and prevent compaction. The ate. Bacterial activity is only weakly reflected in
distribution of fringing cements therefore re- the carbonate isotopic compositions of these
flects the texture and mineralogy of the original
sediments. Diagnostic criteria to predict the
sedimentary deposits. There is no evidence of distribution of these cements may be found in
vadose cementation or of any other influence on their textural relationship to the host sediment.
fringing and overgrowth cementation. Mosaic calcite has formed where fringing ce-
ments were present. Fringing cements formed in
the clay-free, bioclast-rich sands. This leads to
Mosaic cements the prediction that the cements will be found in
the storm deposits because this is where the
After precipitation of the fringing cements the
bioclasts were concentrated originally. These
remaining intergranular pores in the storm
cemented beds are laterally continuous as a result
sediments were filled with calcite. Petrographi-
of the sheet-like geometry of the storm-deposited
cal evidence such as mouldic pores suggest that
sands.
the cements were derived by internal redistribu-
tion of calcite. Similarly carbon stable isotope
analysis only records minor bacterial influence
on the original marine isotopic composition. Viking Group, Troll Field, offshore
However, it is not possible to calculate how much Norway
carbonate may have been mobilized from the
fairweather beds since their initial carbonate The Troll Field straddles Norwegian North Sea
content cannot be estimated objectively. The Blocks 31/2, 31/3, 31/5 and 31/6, and lies
driving force during precipitation of the mosaic approximately 80 km WNW of Bergen (Fig. 7).
calcite was attempted equilibration of the origin- The Troll Field reservoir comprises 400 m of
ally detrital sedimentary assemblage with its Middle to Upper Jurassic Viking Group sand-
pore waters. This assemblage included stable or stones (Osborne 1985). The sedimentology of
metastable aragonite, high-Mg and low-Mg these shallow-marine shelf sands is complex,
calcite bioclasts. The initial fringing cements reflecting a number of transgressive and regres-
were non-ferroan and may have comprised sive cycles. Each cycle comprises coarsening
aragonite and calcite. During burial, pore waters upwards sands deposited during 'progradation'
became anoxic and temperatures increased. on to a shallow marine shelf. These sands are fine
Ferroan calcite filled pores and replaced now and micaceous at the base and coarsen upwards.
unstable bioclasts and their overgrowths. Thus The cycles are capped by coarse lag deposits
the carbonate mineral assemblage evolved dur- formed by winnowing and reworking of the
ing burial from a mixture of high and low-Mg 'progradational' sands during subsequent
calcite and aragonite towards ferroan calcite and transgression (Whitaker 1984, Osborne 1985).
dolomite. This evolution is also reflected by the Palynofacies studies indicate that these 'progra-
gradual shift to negative 6180 values. During dational' sands are shelf-sand ridges or bars,
diagenesis the clay-rich beds underwent compac- rather than coastal sediments (Whitaker 1984).
tion, progressively destroying intergranular The conventional microscopic, cathodolu-
microporosity. Patches of cement formed but the minescence (CL) and stable isotopic characteris-
bioclasts were dissolved, contributing to cementa- tics of the non-ferroan calcite from two thick
tion in the bioclast-rich beds. Most calcite cemented beds (1382-1383, and 1417.5-1419 m)
precipitated within the pores preserved by early in the core from Well 31/2-12, and several
fringing cementation of the storm deposits. The thinner beds from the same well are described
distribution o f the mosaic cements therefore here. The texturally mature host sediments are
reflects the texture and mineralogy of the described as bioclastic and non-bioclastic, the
sediments after the fringing cements pre- texturally immature sediments as micaceous. It is
cipitated. important to note that no specific lithofacies or
 C a r b o n a t e c e m e n t s in Jur as s ic s a n d s t o n e s III
I I
0o 20 4~ 60
k
k

-- 62o

t
ISTATFJORD~-~I
31/2

t
t

60~
3i/3
- -

!
!
t
t
NORTH t SEA
t
t
_ 59 ~ 3,/sV~ Stavang~

\
TROLl F I E L I ~
oo \
I k 2~
I
0 I00 200 km
I I I

FIG. 7. Troll Field location map.

mineralogy is cemented preferentially. The sam- not observed in the texturally mature sediments.
ple base used in this study is extremely small and, The picture is complicated by the amount of
as other cemented beds in the reservoir may have quartz corrosion but it may be suggested that
different origins, the conclusions reached here cementation took place before any significant
can only be extrapolated with extreme caution. compaction occurred.
In the texturally mature sediments non-fer-
roan calcite occurs as fringing and mosaic
cement. Fringing cement occurs as polygonal
Petrography calcite around bioclasts and detrital clastic
grains, and syntaxial overgrowths around
echinoderm plate fragments. Fringing cements
Conventional microscopy
are succeeded and enclosed by the more volume-
In the tightly cemented beds, non-ferroan calcite trically significant mosaic calcite. This mosaic
comprises up to 5 0 ~ bulk volume. Texturally, calcite can be classified according to crystal size,
the majority of the grains 'float' in the calcite or comprising very fine or micritic calcite (15 p.m in
are only in point contact. Deformed micas were length), fine calcite (50 ~tm), medium to coarse
1 I2 J . D . Kantorowicz et al.

blocky calcite, and poikilotopic calcite crystals commencing with fringing cementation, fol-
(over 3 mm long). lowed by a variety of brown or brown and yellow
zoned mosaic calcites. The sequence progresses
through yellow to dull orange luminescing calcite
Cathodoluminescence texture with contemporaneous fracturing.
CL microscopy revealed the fringing and mosaic
cements visible in transmitted light (Fig. 8c), as
well as primary calcite textures consisting of
polygonal and botryoidal zoned calcite which are Stable isotope analysis
not visible during conventional microscopy (F!g.
8a, b and d). The rhombs occur as pendants Comparing the isotopic compositions of the
hanging from detrital grains, or as a meniscus calcites from the Viking Group sandstones (Fig.
cement at grain contacts. Both primary textures 9 and Table 3) with the luminescence colour and
have now been replaced. CL microscopy also transmitted light textures revealed no systema-
reveals that microbrecciation occurred after the tic variation. It follows that porewaters with the
rhombs had formed. The mosaic cements were same stable isotopic compositions can precipi-
found to emit a variety of homogeneous or zoned tate apparently different calcites whilst appar-
luminescence patterns. However, there is no ently similar calcites can be precipitated from
correspondence between the luminescence pat- porewaters with widely differing isotopic
terns observed in the mosaic cements and the compositions.
mosaic textures observed during conventional
microscopy. The homogeneously luminescing
calcites display brown, bright yellow or dull-
orange luminescence (Fig. 8b, c). The zoned Interpretation
calcites comprise from three to 10 or 12 zones,
made up of either non-luminscent calcite, or In both the bioclastic and non-bioclastic beds the
brown, bright yellow and dull orange lumines- fringing and pendant calcites have carbon
cent calcite (Fig. 8e). Besides containing variable isotopic compositions indicative of precipitation
numbers of zones, the zoned calcites display no from sea water (see Table 3). The remaining data
systematic sequence of luminescence colour. fall into two clusters with distinct and highly
Zonation in carbonate cements can reflect negative 613C values - 2 1 to -31%o and - 4 1 to
variations in both the composition and the -47%0. These carbon isotopic compositions are
degree of saturation of the precipitating fluids as all indicative of bacterial sources of bicarbonate
well as crystal growth rate. Consequently, the (Hudson 1977, Irwin et al. 1977).
specific colours manifested here do not necessar- In the bioclastic bed, fringing calcite is post-
ily suggest that widely differing porewater dated by calcite with ~513C values of - 2 1 to
conditions existed during diagenesis (Ten Have -31%o (Fig. 10a). These values are indicative of
& Heijnen 1985). calcite incorporating bicarbonate generated by
In the micaceous beds CL microscopy re- sulphate-reducing bacterial processes. The cal-
vealed a complex fabric of grain breakage and cite is non-ferroan and encloses pyrite, confirm-
microbrecciation (Fig. 8f). ing that sulphate reduction had occurred. If
mobile H2S fixed the ferrous iron generated
during suboxic reduction, then these iron-poor
carbonates formed in the sulphate reduction
Diagenetic sequence of non-ferroan carbonate zone. The oxygen isotopic compositions of these
cementation
samples are consistent (around 6 ~sO - 1%0) and,
On the basis of conventional and CL microscopy assuming Jurassic sea water to have a 6180
it is possible to establish a diagenetic sequence SMOW composition of -1.2%o (Shackleton &

FIG. 8. Thin section photomicrographs, Troll Field. (a) and (b) In transmitted light only a mosaic of fine
calcite cement is visible. CL reveals a zoned polygonal cement which occurs as a pendant or meniscus coating.
Intergranular porosity is cemented with homogeneously luminescing cement. Fracture filling yellow calcite
also occurs. 31/2-12, 1418.5 m (5), XPL and CL respectively. (c) Bioclast fringed with brown and yellow zoned
calcite. Intergranular porosity is filled with homogeneous yellow or orange luminescing calcite. 31/2-12, 1417.8
m (lb), CL. (d) Brown and yellow zoned botryoidal calcite. Cement nuclei occur preferentially at the bases of
detrital grains. 31/2-12, 1418.5 m (5) CL. (e) Multiple zoned calcite cementing a texturally mature sandstone.
31/2-12, 1382 m. (f) Micaceous sand cemented with brown luminescing calcite, itself fractured and infilled
with yellow and then orange luminescing calcite 31/2-7, 1612 m. CL. Scale bars = 200 ~tm.
 k ~

i
 Carbonate cements in Jurassic sandstones I 13

~18 0 PDB

2o

Io
METHANIC
FERMENTATION

SEA WATER

- 50 . . - - - - - - - - - - - - ~ 40

OXODIZED METHANE
~ 0
I I~~
I0 20
813C PDB

AEROBIC OR
SULPHATE-REDUCING -IO
BACTERIA

F - BIOCLAST-RICH BEDS
H - NON-BIOCLASTIC BEDS
LATE CARBONATES -20

-3o

FIG. 9. Stable isotopic compositions of calcite cements, Viking Group Sandstones, Troll Field, classified
according to host lithology and compared with data from Hudson (1977).

TABLE 3. Stable isotopic compositions of calcite samples, Well 31/2-12 Troll Field, Norwegian sector,
North Sea
(a) Bioclast-rich beds

Core Isotope 613C 31 sO

Depth (m) sample sample %0 PDB %0 PDB Luminescence character Remarks

1378.5 2 18 -26.22 -1.54 Brown

1382.5 1 16 -21.88 -2.52 Brightly luminescing zoned yellow and dull orange
1382.6 3 11 - 26.00 - 1.65 Brown
1382.6 3 20 - 3.97 - 0.06 Bright yellow Bioclast
1382.7 4 15 -29.11 1.48 Brightly luminescing zoned yellow and dull orange
1382.8 5 4 + 1.32 - 0.54 Dull orange Bioclast
1384.0 1 3 -26.52 -0.21 Brightly luminescing zoned brown and yellow Calcite filling vein
1 12 - 31.54 - 1.68 Brown
(b) Non-bioclastic beds
1413.0 1 10 -47.15 - 1.38 Dull orange
1417.6 IA 5 -46.16 -0.90 Dull orange
1417.8 IB IA -32.59 -1.11 Bright yellow
1417.8 1B 1B +0.41 -0.61 Dull orange Bioclast + early cement
1417.8 1B 1C -41.45 -0.88 Dull orange
1418.0 2 6 -47.13 -0.54 Dull orange
1418.2 3 7 -41.68 -1.12 Dull orange
1418.3 4 8 -45.86 - 1.01 Dull orange
1418.5 5A 13 -31.46 -1.93 Zoned brown and dull orange, botryoidal zonation Early vadose cement nucleus?
1418.5 5B 14 -0.14 -0.15 Brightly luminescing zoned brown and yellow Early vadose cement
1418.8 6 9 -46.12 -0.93 Dull orange

Kennett 1975), formed at or near the sediment indicating that it incorporated bicarbonate gen-
surface. There is no systematic zonation through erated during bacterial oxidation of methane
this bed. (Fig. 10b). This occurs when biogenic and
In the non-bioclastic bed most of the calcite thermogenic methane serve as a substrate for
which encloses pendant cement or blankets continued bacterial activity. Some of the calcite
entire samples has a 513C value ( - 4 1 to -47%o) in these cemented beds has a 5~3C value of
Ix4 J. D. Kantorowicz et al.

Graphic log Sample ~13C ~I8 0

-5o -4o -3o -2o -io o -I0 0 I0
1382.5 rn o
l I I l I I I I I

o -.-,---r (r6) V V
I
I
I
?
"-" o
l
I
-,,-- 3 (~t, 20) & . . . . x AX
/
/
/
?
/
/
-"-- 4 (15) ~

5 (4) X X
n
1382.8

X - FOSSIL m - YELLOW LUMINESCENCE

ZONED CALCITE 0 DULL-ORANGE LUMINESCENCE

9 - BROWN LUMINESCENCE 9 FERROAN CALCITE

Graphic log Setup le ~ 13 C ~18 0

-50 -40 -30 -zo -no o io -io o io
1417,5 m I I I I I I I I I I

Io (5) 0 0
\ I
,
,

9
-
Ib (A,B,C)
/ 0~7 X (FOSSIL) X

I
// (6) 0

\
2 0

//
/ 9

""-- 3 (7) 0
I
0
--...<
9 o

4 (8)
/
0
I
O

)) 5 (13,14) ~7
(BLOCKY)
(VADOSE)
I
//
9 /
/ I
O
6 (9) 0
/ ~

1419.0

FIG. 10. (a) Stable isotopic composition of calcite cemented bed. Bioclastic bed, 31/2-12, 1382 m. (b) Stable
isotopic composition of calcite cemented bed. Non-bioclastic bed, 31/2-12, 1418 m.

- 31%o and an origin similar to the calcite in the Controls on carbonate mineral authigenesis in
bioclastic bed. The samples analysed here may the Viking Group sandstones
also have contained disproportionate mixtures of
pendant calcite (~13C=0%o) and isotopically The non-ferroan calcite in the Viking Group
light calcite (613C - 41%o). The oxygen isotopic sandstones in Troll comprises a fringing or
values are consistently around 3180 -1%o and possibly vadose pendant cement and nucleated
again it is suggested that precipitation occurred on these more extensive mosaic calcites that
near the seafloor. incorporated bacterially derived carbonate 9
 Carbonate cements in Jurassic sandstones I [5

These generations of cement are separated by sandstones one would expect to find evidence of
authigenic pyrite, kaolinite and K-feldspar. exposure towards the top of regressive sequences
or immediately beneath transgressive sands.
Additional evidence for emergence is seen in the
Fringing and pendant cement
micaceous sands which occur at the base of the
As discussed earlier, fringing cements do not sequence and display fabrics reminiscent of
appear to be particularly diagnostic of any intertidal hardground or calcrete development
sedimentary environment. By contrast, pendant (Fig. 8; Assereto & Kendall 1977).
or dripstone textures have only been described
from coastal beachrock or intertidal areas, where
Mosaic calcite
they form in the vadose or unsaturated zone
(Mfiller 1971). However, there is no reason why In the Viking Group sandstones, fringing calcite
they should not form in offshore environments is enclosed by a mosaic calcite. Thin section and
which undergo uplift and emergence. Einsele et oxygen stable isotopic evidence both suggest that
al. (1977) discuss the effects of emergence on cementation took place before significant burial
Holocene sediments from the continental shelf had occurred. The early cements may have
off Mauritania. They attribute carbonate lithifi- provided a nucleus for the mosaic or they may
cation of sediments 100 km offshore and now simply have maintained a porous framework
100 m below sea-level to subaerial exposure. The within which the mosaic developed. Whether
effects of emergence overprint marine sediments these cemented beds are nodular and discontin-
whilst subsequent transgression and reworking uous or continuous will depend on whether they
has removed any direct evidence of subaerial grew around dispersed nuclei or from a finite
exposure. It follows that in the Viking Group supply of bicarbonate (Fig. 11). In the bioclastic

TIME SLOW B U R I A L
A B U N O A N T SUPPLY OF OXIDANTS

-@
B"~-'~"~(~1..~. AEBoBICBAC~BB~ALs163163 Cementedlayers
A\\ ~'---~........-..
\

~ '\,\
\\
\ ,kN~Es SULp~4AIss163
\

|
...... - ...... \\\,

,,,

;'~p,Ms ['O~
ANAEBoB~C ME~'~4ANoGEN~C \\\
BAC3-~'B~AL LEGEND:
Aerobic processes:
~]Jj 513C 20 to -35 carbonate
(oxidation of methane, -40 to 50)

Sulphate reduction:
513C -20 to -25 carbonate
Methanogenic processes:
..~.~~~0~ ~ ' ~ Nod.... ~ 613C ,10carbonate

FIG. 1l. Paths A, B and C illustrate the effect of burial rate on the style of cementation which may develop in
marine sandstones. In path A an isotopically and mineralogically zoned nodule develops. More dispersed,
patchy cementation can also occur. With faster burial than shown in path A smaller zoned nodules will form
and more organic matter will be available to generate hydrocarbons. Paths B and C illustrate the isotopic and
mineralogical composition of cements that develop at slower burial rates. Path C is the extreme case in which
the sediment remains within the superficial bacterial zones and is completely cemented with an isotopically
homogeneous non-ferroan calcite. In constructing this diagram it is assumed that the boundaries between
successive bacterial reaction zones are maintained at constant depths. Obviously this is not always the case.
I I6 J.D. K a n t o r o w i c z et al.

beds there is no evidence of isotopic zonation in permeability of the sediment enabling methane
the mosaic cements (Fig. 10). After precipitation to rise into the bacterial zones above.
of the early cements these beds were buried into In the case of the non-bioclastic bed it is not
the sulphate-reduction zone, where they re- clear whether the growth banding seen in the
mained until tightly cemented. The thickness of non-bioclastic bed defines the size of the cemen-
the cemented bed and its uniform composition ted area, or the continued accretion of cement as
suggest that the sediments remained in this further methane was supplied. If this process
bacterial reaction zone for some time: a pathway occurred over a large area of the seafloor then the
between B and C in Fig. 11. nodular textures observed in the core could
In the non-bioclastic cemented beds, calcite reflect coalescence of cements around discrete
precipitated incorporating bicarbonate liberated pendant cement nuclei (a pathway between B
during sulphate-reducing bacterial processes and and C in Fig. 11). In both the bioclastic and non-
then during oxidation of methane (Fig. 10b), bioclastic beds it is not possible to establish from
giving rise to the apparent growth banding core data whether or not these areas of cement
observed in the cores. In a closed system, as will have coalesced to form continuous layers.
might exist within a shale-dominated sequence, This will depend on the longevity of the bacterial
this succession would lead to a nodule forming processes involved.
(Fig. 11, pathway A). However, several lines of
evidence suggest that the Viking Group sand-
stones may not be described in this way. First,
the cemented beds are interbedded with highly
Summary
permeable sands. Thus methane generated else- The cemented beds in the Viking Group sand-
where and rising through the sediment is more stones in Troll contain two distinct calcite
likely to have escaped to the surface than to have cements. First, a thin fringing or pendant cement
been trapped and fixed in the sulphate-reduction precipitated from sea water (during regression or
zone. Indeed, almost all the recorded recently- even temporary emergence) and, second, a
formed methanic cement in marine sandstones is mosaic calcite incorporating bacterially derived
found on the seafloor (e.g. Hovland et al. 1985, bicarbonate. To facilitate extensive cementation
Nelson & Lawrence 1984 and references there- the sediments must have remained near the
in). Textural evidence confirms that cementation sediment surface for some time whilst bacterial
took place prior to compaction. Second, the now processes operated. There is no direct evidence
replaced botryoidal texture is more likely to have that these cemented beds are nodular, the
been aragonite formed on the seafloor rather product of closed system diagenesis. Their
than during burial. Third, the oxygen isotopic diagenesis reflects two independent processes
compositions do not show any zonation or operating near the sediment surface. Thus,
fractionation. This is consistent with continuous similar cementation may occur in sediments
precipitation from porewaters with a uniform maintained near the sediment surface, first,
composition, rather than in a closed system or during periods of non-deposition, and possibly
during burial. Finally, it is not clear that even emergence and, second, within zones of
anaerobic oxidation of methane is actually intense bacterial activity.
possible as the porewater profiles in which this
process is inferred to take place are open to
alternative interpretation (see, e.g. Martens &
Berner 1974, Barnes & Goldberg 1976, Lovley & Discussion
Klug 1986, Whiticar & Faber 1986).
In the Viking Group sandstones, mosaic These case studies demonstrate the potential for
calcite precipitation began when the early- reservoir compartmentalization in shelf sand-
formed fringing cements were buried into the stones. Although the diagenetic histories differ a
sulphate-reduction zone. Continued precipita- number of controls on cement geometries have
tion, however, involved aerobic oxidation of been identified. The geometry of the cemented
methane. As the resulting cemented bed is over beds in the Bridport Sands reflects the sandbody
1 m thick, prolonged bacterial activity must have geometry: they are laterally extensive. In the
occurred. Thus it is likely that a temporary hiatus Viking Group sandstones in Troll the location of
occurred maintaining the sandstones near the the cements reflects two conditions and their
seafloor for some time. It follows that three geometry will depend on the extent and timing of
controls on the location of these cemented beds emergence of the original deposits. The question
may be identified: (a) the presence of a nucleus of of whether or not these controls on sandstone
early fringing or pendant calcite; (b) residence cementation can be used for predictive purposes
near the seafloor during precipitation; (c) the elsewhere is addressed below.
 Carbonate cements in Jurassic sandstones I I7

In the Bridport Sands early fringing cements ted layers. These examples show that some of the
developed around storm-deposited bioclasts. controls on the location of cements can be related
Fewer bioclasts were deposited in the clay-rich to the episodic nature of shelf sedimentation, and
fairweather sediments with the result that less to the textural and mineralogical variations
fringing cement developed. Thus it was the fast- which result. Laterally extensive cements de-
slow nature of storm versus fairweather deposi- velop quite simply because the specific sedi-
tion as well as the bioclast concentration in the ments in which they form, or the processes
storm deposits which predetermined the cement responsible for their formation, occur over
location. This cementation model may be widely particularly large areas. Similar cements will not
applicable in storm or other sediments in which develop in more heterogeneous systems where
potential cement nuclei are heterogeneously such controls are not laterally persistent. How-
distributed. In the Viking Group the fringing ever, laterally extensive cements can develop for
and pendant cements owe their origin to non- a variety of other reasons. For example, Gautier
deposition, if not emergence. Thus the location & Rice (1981) describe cementation at a water
of cementation reflects the stop-start nature of table in the Cretaceous Eagle Sandstone, a
sedimentation and burial of these sands. This coastal deposit in Montana. This cement cross-
style of cementation is analogous to firmground cuts depositional sedimentary structures. The
and hardground development and this model potential also exists for late cementation to form
may apply to almost any shelf sediments. laterally persistent barriers, for example at
Whilst the location of the cemented beds hydrocarbon-water contacts.
reflects large-scale sedimentological controls, the Laterally extensive cementation requires an
origin of the pore-filling mosaic calcite reflects abundant supply of cement. Here too there are no
smaller-scale controls. First, in the Bridport unique solutions, the Bridport Sands' cement, for
Sands the early framework cements maintained example, forming from internally redistributed
open pores which were cemented during burial calcite, the Viking Group's cement forming as a
by the remobilization of detrital and early result of bacterial activity. As can be shown from
authigenic calcite. Gradual equilibration of the the Viking Group data, the potential for exten-
sedimentary minerals occurred moving from sive cementation exists when distinct processes
aragonite, high-Mg and low-Mg calcite to fer- operate for prolonged periods. In the Bridport
roan calcite and dolomite. Bacterial activity Sands, processes have continued for long enough
influenced diagenesis as evidenced by the abun- to create continuous layers of totally cemented
dant berthierine but the bicarbonate liberated sand. In the case of the Viking Group sandstone
during bacterial activity has not significantly in Troll it is not possible to establish from this
influenced the isotopic composition of the data set the precise geometry of these extensively
cements. Interestingly, redistribution of detrital cemented beds and hence to predict whether or
bioclasts produces spherical nodules in more not laterally continuous layers have also de-
homogeneous sediments. Hence, only with an veloped here.
understanding of the relationship to host sedi- These case studies demonstrate the potential
ment texture is it possible to distinguish between for laterally extensive cementation to develop in
the two in a core. In the Viking Group shelf sandstones. However, cementation can
sandstones the fringing cement framework was occur for more than one reason and thorough
also filled by mosaic calcite, but the bicarbonate petrographical investigation is necessary to dis-
incorporated here was derived from bacterial tinguish between various possible controls. Only
degradation of organic matter. The potential for after establishing the nature of controls on
bacterially derived cements to develop obviously cementation is it possible to make accurate
exists in any organic-rich sediments subject to predictions about the geometry of cemented
slow and episodic burial. Second, the Bridport sands and their likely influence on hydrocarbon
Sands' cement appears to have evolved through recovery efficiency.
time, leading to a progressive depletion in
oxygen isotopic composition. In the Viking
Group the cements have homogeneous isotopic ACKNOWLEDGMENTS"We thank Shell Research BV,
values: during precipitation porewater composi- Shell International Petroleum Maatschappij, Norske
tions were relatively consistent. This style of Shell and partners in the Troll Field, and the Director
cementation can, therefore, be distinguished of the British Geological Survey (NERC), for permis-
sion to publish this paper. We thank Norske Shell,
from burial cements around similar fringing Poroperm Laboratories, Professor W. G. Mook (Gro-
cements in which isotopic zonation would be ningen), and Professor J. Thorez (Liege) for providing
expected. some of the data discussed. We thank our colleagues in
There are clearly many processes involved in Rijswijk for stimulating discussions. Errors of fact and
generating laterally extensive carbonate cemen- interpretation, however, are our own.
1 18 J . D . Kantorowicz et al.

References

ASSERETO, R. L. M. & KENDALL, C. G. ST. C. 1977. IRWIN, H., CURTIS, C. D. & COLEMAN, M. L. 1977.
Nature, origin and classification of peritidal tepee Isotopic evidence for source of diagenetic carbon-
structures and related breccias. Sedimentology, 24, ates formed during burial of organic-rich sedi-
153-210. ments. Nature, London, 269, 209-13.
BARNES, R. O. & GOLDBERG, E. n. 1976. Methane JORGENSEN, N. O. 1976. Recent high magnesian
production and consumption in anoxic marine calcite/aragonite cementation of beach and sub-
sediments. Geology, 4, 297-300. marine sediments from Denmark. Journal of
CLAYPOOL,G. E. & KAPLAN,I. R. 1974. The origin and Sedimentary Petrology, 46, 940-51.
distribution of methane in marine sediments. In: LOVLEY, D. R. & KLUG, i . J. 1986. Model for the
KAPLAN, I. R. (ed.). Natural Gases in Marine distribution of sulphate reduction and methano-
Sediments, pp. 99-139. Plenum Press, New York. genesis in freshwater sediments. Geochimica et
COLEMAN, M. L. 1985. Geochemistry of diagenetic Cosmochimica Acta, 50, 11-8.
non-silicate minerals: kinetic considerations. Phi- KANTOROWICZ, J. n . 1985. The origin of authigenic
losophical Transactions of the Royal Society of ankerite from the Ninian Field, UK North Sea.
London, A35, 39-56. Nature, London, 315, 214-6.
COLTER, L. V. & HAVARD,D. I. 1981. The Wytch Farm MARTENS, C. S. & BERNER, R. A. 1974. Methane
oilfield Dorset. In : ILLING, V. C. & HOBSON,G. D. production in the interstitial waters of sulphate-
(eds). Petroleum Geologyof North- West Europe, pp. depleted marine sediments. Science, 185, 1167-9.
494-503. Applied Science, London. MILLIKEN, K. L., LAND, L. S. & LOUCKS, R. G. 1981.
CURTIS, C. D. 1985. Clay mineral precipitation and History of burial diagenesis determined from
transformation during burial diagenesis. Philoso- isotopic geochemistry. Frio Formation, Brazos
phical Transactions of the Royal Society, A315, 91- County, Texas. American Association of Petroleum
105. Geologists Bulletin, 65, 1397~413.
DAVIES, D. K. 1967. Origin of friable sandstone MfDLLER, G. 1971. 'Gravitational' cement: an indica-
rhythms in the Lias of England. Journal of tor for the vadose zone of the subaerial diagenetic
Sedimentary Petrology, 37, 1179-88. environment. In: BRICKER,O. P. (ed.). Carbonate
EDMUNDS, W. M., BATH, A. H. & MILES, D. L. 1982. Cements, pp. 301-20. Johns Hopkins University
Pore fluid geochemistry of the Bridport Sands Press, Baltimore.
(Lower Jurassic) and the Sherwood Sandstone NELSON, C. S. & LAWRENCE, M. F. 1984. Methane
(Triassic) intervals of the Winterborne Kingston derived high-Mg calcite submarine cement in
borehole, Dorset. In: RHYS, G. H., LOTT, G. K. & Holocene nodules from the Fraser delta, British
CALVER, M. A. (eds). The Winterbourne Kingston Colombia, Canada. Sedimentology, 31, 645-54.
borehole, Dorset, England, pp. 149-63. Report OSBORNE, P. 1985. The Troll Field. Proceedings,
Institute of Geological Sciences, 81/3. Norwegian Institute of Technology Seminar on North
EINSELE,G., ELOUARD,P., HERM, D., KOGLER, F. C. & Sea Oil and Gas Reservoirs, Trondheim.
SCHWARZ, H. U. 1977. Source and biofacies of SHACKLETON,N. J. & KENNETT,J. P. 1975. Palaeotem-
Late Quaternary sediments in relation to sea level perature history of the Cenozoic and the initiation
on the shelf off Mauritania, West Africa. 'Meteor' of the Antarctic glaciation: oxygen and carbon
Forschungsergebnisse Reiche, C26, 1--43. isotope analysis in DSDP sites 277, 279, and 281.
FROELICH, P. N., KLINKHAMMER,G. P., BENDER,M. L. In: Initial Reports of the Deep Sea Drilling Project,
LUEDTKE, N. A., HEATH, G. R., CULLEN, D., 29, 743-55. Government Printing Officer, Wash-
DAUPHIN, P., HAMMOND, D., HARTMAN, B. & ington DC.
MAYNARD, V. 1979. Early oxidation of organic TEN HAVE, A. H. M. & HEIJNEN, W. 1985. Cathodolu-
matter in pelagic sediments of the eastern equator- minescence activation and zonation in carbonate
ial Atlantic: suboxic diagenesis. Geochimica et rocks: an experimental approach. Geologie en
Cosmochimica Acta, 43, 1075-90. Mijnbouw, 64, 297-310.
GAUTIER, D. L. & RICE, D. D. 1981. Comparison of WHITAKER, M. F. 1984. The usage of palynology in
conventional and low-permeability reservoirs of definition of Troll Field geology. 6th Offshore
shallow gas in the northern Great Plains. Society Northern Seas Conference and Exhibition,
of Petroleum Engineers Paper 9846. Stavanger.
HOVLAND, i . , TALBOT, M., OLAUSSON,S. & AASBERG, WHITICAR, M. J. & FABER,E. 1986. Methane oxidation
L. 1985. Recently formed methane-derived car- in sediment and water column environment--
bonates from the North Sea floor. In: Petroleum isotope evidence. Organic Geochemistry, in press.
Geochemistry in Exploration of the Norwegian shelf, WILKINSON, B. H., SMITH, A. L. & LOHMANN, K. C.
pp. 263-6. Norwegian Petroleum Society. 1985. Sparry calcite marine cement in Upper
HUDSON, J. D. 1977. Stable isotopes and limestone Jurassic limestones of southeastern Wyoming.
lithification. Journal of the Geological Society of Society of Economic Paleontologists and Mineralo-
London, 133, 637-60. gists, Tulsa. Special Publication, 36, 169-84.

J. D. KANTOROWICZ, I. D. BRYANT and J. M. DAWANS, Koninklijke/Shell Exploratie en

Produktie Laboratorium, Postbus 60, 2280 AB Rijswijk ZH, The Netherlands.
Magnesite formation in recent playa lakes, Los Monegros, Spain
J. J. Pueyo Mur & M. Ingl6s Urpinell

s u M M A R Y : Early diagenetic magnesite is at present forming at a depth of some 20 cm in

recent playa lakes in NE Spain. Magnesite formation can be considered as a result of two
factors: (a) an increase in CO2 activity caused by the decay of organic matter, and (b) the
presence of post-halitic brines, strongly concentrated in magnesium, that form in summer.
These two factors are mainly observed in the ephemeral salt-pan zone where the highest
accumulation of magnesite is found.

The Recent lakes studied are located in Los which affected the rnain Tertiary depressions of
Monegros in the central sector of the Ebro Basin, the Iberian Peninsula during the Miocene. It is
NE Spain (Fig. 1,A). They occur on evaporite the recycling of these saline materials, occurring
bearing Miocene and Oligocene sediments. selectively according to their solubility, which
These lakes are playa lakes and dry salt lakes; gives a high salinity to the waters flowing into the
they are small and have ephemeral ponds with lakes.
brines.
The lakes are found between 300 and 400 m
above sea-level and develop on depressions Evaporite precipitation
which originated through dissolution of Miocene
evaporitic levels, accompanied by surface defla- Different kinds of evaporitic minerals have been
tion. A great many of these lakes develop found. They originated through direct precipita-
asymmetrical shapes conditioned by the NW tion from the superficial waters as well as
dominant wind. Insolation and wind are the through interstitial or efflorescent precipitation
main factors affecting the evaporation of the from the interstitial brines.
waters. Rainfall in the area is low: about Crystallization from the surficial waters of the
350 mm yr -1 (Quirantes 1965). lakes develops as a response to their seasonal
evolution (Quirantes 1965, Pueyo 1978-79).
Composition and evolution of brines (a) Spring-summer precipitation: Crystalliza-
tion occurs in a concentric zoned distribution
Surficial brines are of C1--SO42--Na+-(Mg 2+) according to the carbonates-gypsum-halite
type (Pueyo 1978-79) and undergo strong sequence.
seasonal oscillations in concentration because of (b) Winter precipitation: Crystallization of
evaporation and progressive precipitation of sodium sulphate in the form of mirabilite.
mineral phases. In most lakes there is no surficial This mineral changes easily to thenardite by
water in the greater part of the year and even dehydration induced by wind and insolation.
during the entire year in dry periods. Efflorescent crystal-precipitates develop from
Interstitial brine accumulates in the eastern- interstitial brines carried up to the surface by
central area of the lakes (ephemeral salt-pan) and capillarity, with bloedite, thenardite and halite
in the western area where expandable clay being the minerals most commonly represented.
minerals, with high water retention capacity, are Evaporitic sediments are reworked by wind,
more abundant. The interstitial salinity increases waves, rain and organisms living in the water or
towards the lake surface as is usual in a playa at the dry lake bottom (cyanobacterial mats,
lake regime, and is controlled by evaporative small crustaceans, coleoptera, etc.).
pumping in the vadose zone during dessication
periods. An increase in CI-, Na + and Mg 2+
concentration has also been reported from the
leeside (SE) to the windward side (NW) of the
Mineralogy of the Quaternary
playa lakes. lacustrine deposits
Brine generation depends not only on climatic
conditions (aridness) but also on the influence of To study the Quaternary sediments, sampling
the evaporite-bearing substratum. Substrate eva- work in 10 playa lakes has been carried out
porites developed in intermontane conditions (Pueyo & Ingl6s 1986). One of these lakes (the

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, II9

Geological Society Special Publication No. 36, pp. 119-122.
120 J. J. Pueyo Mur & M. Inglbs Urpinell

~ G
L
i Km
t I

FIG. 1. Geographical situation of the studied playa lakes. P and G correspond to Pito and Guallar playa
lakes, mainly reported in this paper. P. Pito playa lake and core sampling positions. G. Guallar playa lake
and core sampling positions. AM sample (of Pito playa lake) and AG sample (of Guallar playa lake) are
located in the ephemeral salt-pan of the lakes (dotted line).

Pito pl~ya; Fig. 1,P) has been studied exhaus- shows a similar horizontal distribution pattern to
tively and taken as a model which can be applied that of gypsum, except for the top 20 cm of the
to the other ones. sediment, where magnesite is usually not found.
The Quaternary deposits that infill the lacus- Exceptions to this are the occurrence of minor
trine depressions are made up of detrital and amounts of surficial magnesite in the lakes where
evaporitic sediments and are up to 2 m thick. the sediments are frequently brine-filled. Minor
Gypsum, halite, thernardite, bloedite and mira- quantities of fibrous clays (sepiolite?), not
bilite all form by evaporation in the lakes, but detected in the Tertiary substrate have also been
gypsum is the only one which is preserved at any found in the salt-pan area. Carbonate and clay
depth in the sediment. It develops as lenticular distributions reflect the early diagenetic interac-
crystals and concentrates mainly in the salt-pan tions between interstitial brines and detrital
zone. Sometimes gypsum nodules several centi- sediments.
metres in size, formed by clusters of lenticular
crystals, can be found. The other evaporitic
minerals are dissolved during rainy seasons. Magnesiteformation---some
Gypsiferous and detrital sediments show op-
posite distribution patterns. Detrital deposits
remarks
originate from erosion taking place in areas Although magnesite has been found in all the
adjoining the playa lakes (up to a few kilo- lakes studied, Guallar playa lake (Fig. 1,G) has
metres). They are concentrated towards the been taken as a model for the study of magnesite
edges and the windward side of the playas and crystals due to the great abundance of this
are mainly represented by quartz, clay minerals mineral found. In the ephemeral salt pan of this
and carbonates. Illite, chlorite, smectite and lake, magnesite reaches 50~o of the water-
kaolinite form the clay fraction present in the insoluble fraction (carbonates + quartz + clays),
Quaternary sediments and the relative abun- which means approximately 15~ of the
dances depend on the composition of the sediment.
Tertiary substrate. Magnesite seems to have originated through
Carbonates present in the lake sediments are interstitial precipitation from phreatic brines.
calcite, dolomite and magnesite. Calcite is The reasons for suggesting this are the almost
dominant on the surface: it has a detrital origin. total absence of magnesium carbonate at the
At depth (some 20 cm), the dolomite content surface and the subhedral shape of the magnesite
tends to be higher and magnesite may appear crystals (Fig. 2c) which grow displacively in
(Fig. 2a), both minerals showing evidence of dominantly siliciclastic mud with some carbon-
being formed during early diagenesis. The ate content (30~ on average). Detrital compon-
highest magnesite concentration is found below ents are dominant at the surface but magnesite
the ephemeral salt-pan area where sometimes it rhombs are progressively more abundant with
may be the dominant carbonate. Magnesite depth (this has been established for the upper 40-
 Magnesite formation in recent playa lakes 121

AM PITO
o io 20
!
%
L I

o!I

~
(c)

AG GUALLAR
o Io 20%
!
| I

5 0 cm

(a) (b)

FIG. 2. (a) Distribution of minerals (carbonates and quartz) with depth at AM (Pito) and AG (Guallar)
sampling points. Bars represent weight percentage of different minerals in the water-insoluble residue (C:
calcite, D: dolomite, M: magnesite and Q: quartz). Samples have been taken at 5 cm intervals. (b) Map of
magnesite concentration (as per cent) in Pito playa lake, at a depth of 40-45 cm. The contours were calculated
using a microcomputer starting from an initial grid of 19 data points. (c) Scanning micrograph of magnesite
crystals. Ultrasonic pre-treatment and sieving have been used to remove the clay fraction. Guallar playa lake,
AG point at a depth of 40-45 cm.

50 cm). It is suggested that the magnesite could period subject to direct evaporation. This ena-
have precipitated interstitially as a response to bles the brines to reach the post-halitic stage with
the increase in the CO2 and CO3 z- activity due a strong magnesian concentration at the end of
to bacterial decay of algal and cyanobacterial summer.
matter in a similar way to that described by In the Guallar playa lake these conditions are
Perthuisot (1974) in E1 Melah sebkha (Tunis), in more obvious. For the last 10 years at least, a
the presence of strongly magnesic brines. Preci- significant growth of cyanobacterial mats, coex-
pitation occurs in the ephemeral salt-pan zones isting in summer with magnesian brines, has
(Fig. 2b) as they are the lowest areas of the playa been recorded (Pueyo 1978-79). Density and
lakes. These zones collect a greater amount of viscosity are high in these brines and water
rain water. This causes cyanobacterial mats to activity is low; such conditions favour magnesite
develop frequently, leading to high organic formation and prevent nucleation of metastable
(sapropelic) accumulation. During dry seasons hydrated magnesium carbonates, as suggested by
these areas have surficial brines for a longer Langmuir (1965) and Christ & Hostetler (1970).
I22 J. J. Pueyo Mur & M. InglOs Urpinell
References
CHRIST, C. L. & HOSTETLER, P. B. 1970. Studies in the tica actual en las lagunas saladas del /trea:
system MgO-SiO2-CO2-H20. II: The activity Bujaraloz, SS.stago, Caspe, Alcafiiz y Calanda
product constant of magnesite. American Journal (provincias de Zaragoza y Teruel). Revista Insti-
of Science, 268, 439-53. tuto Investigaciones Geol6gicas. Diputaci6n Provin-
LANGMUIR, n. 1965. Stability of carbonates in the cial Barcelona, 33, 5-56.
system MgO-COz-H20. Journal of Geology, 73, - - & INGLf~SURPINELL, M. 1986. Substrate minera-
730-54. logy, pore brine composition, and diagenetic
PERTHUISOT, J. P. 1974. Les d+p6ts salins de la sebkha processes in the playa lakes of Los Monegros and
E1 Melah de Zarzis: Conditions et modalit6s de Bajo Arag6n (Spain). I International Symposium on
la s6dimentation 6vaporitique. Revue de Gbogra- Geochemistry of the Earth Surface. Granada (Spain),
phie Physique et de GOologieDynamique, (2) 16, f. 2, 16-22 March 1986, in press.
177-88. QUIRANTES, J. 1965. Nota sobre las lagunas de
PUEYO MUR, J . J . 1978-79. La precipitacibn evapori- Bujaraloz--Sfistago. Geographyca, 12, 30-4.

J. J. PUEYOMUR and M. INGLF_,SURPINELL, Dep. de Geoquimica, Petrologia y Prospecci6n

Geol6gica, Gran Via, 585, 08007 Barcelona, Spain.
 Mixed-water dolomitization in a transgressive beach-ridge
system, Eocene Catalan Basin, NE Spain
C. Taberner & C. Santisteban

S U M M A RY : Dolomitization occurs at the top of alluvial fan deposits in the SE margin of

the Eocene Catalan Basin, NE Spain. The dolomitized rocks occur just below erosion
terraces which show a stepqike disposition on the continental sediments and on which
shallow marine facies (beach-ridge systems and reefs) are developed. Petrography shows that
diagenesis in the continental sediments mainly originated under meteoric water influence,
while in beach-ridge systems and reef carbonates there is evidence of early diagenesis in a
marine environment. All the nearshore sediments (from alluvial fan to reefs) record a distinct
geochemical pattern where a progressively more marine influence parallels the regional
trend in facies across the ancient shoreline. Petrographic and geochemical data suggest that
mixing of meteoric and marine waters in the sediments representing the continental-to-
marine transition caused dolomitization at the top of the alluvial fan deposits. Unlike most
mixing-zone dolomites, in this case mixing took place when shallow marine sedimentation
took place on erosional surfaces causing the inflow of marine waters that mixed with
meteoric porewaters.

Diagenetic dolomite occurs in several facies Sedimentary features

associations in the Vic area (SE margin of the
Eocene Catalan Basin, Figs 1 and 2). Some of
this dolomite formed during early diagenesis and The beach-ridge systems
has been recorded mainly in Eocene marine
sediments, in a transitional zone between the Levels interpreted as beach-ridge systems are
continental and marine deposits of the SE sector intercalated with, or found at the top of, alluvial
of the studied area (Figs 2 and 3). fan deposits in the southern and eastern margins
In this zone dolomitized sediments are gener- of the studied area (Figs 1 and 2). They are
ally confined to a narrow strip at the top of the Upper Lutetian-Lower Biarritzian in age, and
continental sediments, interpreted as alluvial fan their outcrops are arranged in a 50 km long
systems (Reguant 1967, Colombo 1980, Taberner narrow strip bounded by alluvial fan sediments
1982). The Middle Eocene transgression affected and marine platform facies associations (Fig. 2).
these deposits, so that nearshore sediments Beach system outcrops are distributed in three
developed over the alluvial fan facies associ- well-defined areas, corresponding to troughs
ations (Fig. 3). The basal sediments in the marine bounded by N N W - S S E trending fractures,
series are interpreted as transgressively-de- whose synsedimentary action during the Palaeo-
veloped beach-ridge systems. gene sedimentation in this area has been referred
Field and textural features of rocks in the to (Taberner 1982).
nearshore facies associations where dolomite has The position of beach levels are evidence of
been found may be taken as evidence of dolomite the location of ancient shorelines during the
development during early diagenesis prior to successive stages of the Eocene transgression.
compaction. Petrography and preliminary geo- These sediments originated after the reworking
chemical results point out the possibility of of continental supplies, and the areas where they
mixing-zone dolomitization. This mechanism are best recorded coincide with those of greater
has been widely referred to as producing dolomi- thicknesses of alluvial fan deposits.
tization in nearshore areas (Hanshaw et al. 1971, Beach deposits in the area studied are ar-
Badiozamani 1973, Land 1973a, b, Folk & Land ranged in tabular siliciclastic bodies. Their
1975, Ward & Halley 1985, among others). The lengths along the ancient shoreline can be in tens
cases usually discussed in the literature are those of kilometres and widths are up to 2 km. Beach
regarding dolomitization of marine nearshore levels are up to 4 m thick (when successive beach
sediments by mixing-zone water. In this paper units are amalgamated). Two distinct groups of
we show how the special features of the beach sediments in the Vic area have been
transgressive beach-ridge depositional system distinguished: (1) constructive depositional
may have given rise to the dolomitization beaches and (2) erosional beaches originating
observed at the top of the alluvial fan sediments. from marine reworking of alluvial supplies.

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 123

Geological Society Special Publication No. 36, pp. 123-139.
I24 C. Taberner & C. Santisteban

~ - ~-~1(studyarea ,1!II

1 : Albany'} Fault
2 : Amer Fault c
3 : Fabregues Fault
4 Seva
: Fault
5 ElCongostFault
_S A

/ B
\
\
A: Girones Empord6 Basin
B : Vic Basin ( location of the
studied area)
C: Igualada Basin

c ~ Alluvial and deltaic systems

O 5OKra
t .J '- : I 1
O: Olot
R: Rjpoll
- ] Poleozoic, Mesozoic 8, Neogene I : Igualada
(emergent areas in Eocene)

i' [ [ Paleogene sediments.

Fia. 1. Geological features of the Eocene Catalan Basin. The NNW-SSE fault system controlled the location
of minor basins during the Palaeogene as well as the distribution of deposits along the southeastern sector of
the Ebro Depression. The Vic sedimentation area (B) was bounded by El Congost and Amer faults. The
activity of these fractures during sedimentation brought about a more complete record of the Middle Eocene
transgressive stage in the southeastern margin of the Basin. The thickest sequences of beach sediments were
developed in trough areas bounded by NNW-SSE trending faults.

Most beach deposits in the study area corre- proximal areas on a horizontal to slightly
spond to the constructive depositional type. seaward-dipping terrace (Figs 3 and 4), above
Their external shapes and internal structures the continental deposits. The manner in which
may be best compared to the beach-ridge systems these erosional surfaces were formed indicates
described by Carter (1986) in Northern Ireland. dispersion of alluvial fan deposits and coincides
Depositional beaches are closely related to reef with the development of erosional beaches.
carbonates, corresponding to slightly deeper Thus, erosional terraces stemmed from the
sedimentation zones. In some cases, beach levels reworking of continental deposits while ero-
prograded over coral-reefs, but reefs are also sional beaches represent the lag products of
present on depositional beach systems as a reworking.
result of the transgression of the depositional The step-like arrangement of different beach
system. levels indicates the combined effect of transgres-
Erosional beaches are only represented by up sion with tectonic pulses, causing deepening in
to 50 cm thick lag deposits composed of cobbles the sedimentation area (Fig. 4). This process
and boulders, within a coarse-grained sandstone would proceed in the following manner (Ta-
matrix. These deposits formed over up to 20 m berner 1982):
long erosional terraces on top of the alluvial fan
continental sediments. (1) Rapid sinking of the sedimentation area and
Both kinds of beaches characterize the transi- submersion of the continental deposits, there-
tion between continental and marine deposits by developing an erosional terrace from the
during the Eocene (Lutetian-Biarritzian) marine effects of wave action.
transgression of the Catalan Basin. Beach units (2) Stabilization and progradation of a deposi-
occur in a step-like arrangement, each develop- tional beach over the erosional terrace
ing, whether erosional or depositional, in the surface.
 Mixed-water dolomitization 125

St. Feliu de

!:-. Rupit

::i.'. % Tavertet
Manlleu :i::.!:. . ...- Amer
9 ..': '.".'.
....::L :.;!. : ".:.~.....~

:.,..-...
:.7:
::.... . . :.,'
I
I I

" ~
9

:-5 :.: .".;i

:.:%: -.

9 ":ii:::ill
i:/
~::.: ..:.. :'.'..~
[ ~ Basinal marls

~Neorshore facies associations

Centellese'i: ~ Beach units

.9 ~:;;
~ Alluvial-fan sediments

~ Pre-Tertiarybasement
9 .. .i.': : .
, 9 ... , ...:.:

0 5Kin
t ,!

FIG. 2. Location of beach deposit outcrops in the southeastern margin of the Catalan Basin (Vic area).

Reef buildups associated with the beach systems systems, (3) reefs were dynamically related to
beach siliciclastic sediments, and (4) biological
Beach sediments are interbedded with reef communities, as well as siliciclastic sedimenta-
carbonates and offshore marls in the distal tion systems, showed similar trends during the
regions. Reefs are found overlying beach depos- marine transgression. The relationships between
its: they first developed in the frontal part of reefs and beach levels are discussed in detail in
beach-ridge units, and progressively overlapped Santisteban & Taberner (in press).
them. The transgressive migration of reefs Reef buildups of the studied area consist of
parallels that of beach-ridge systems (Fig. 3). fringing-reefs made up of corals and coralline red
This suggests that: (1) reefs grew on previously algae. Taberner & Bosence (1985) found these
drowned beaches, (2) they developed in front of coralline red algae indicative of shaded zones
and sometimes simultaneously with active beach where turbid waters restricted light passage into
I26 C. Taberner & C. Santisteban
reef developing areas. These turbid waters were
probably due to active siliciclastic sedimentation
0
within the beach-ridge systems. Therefore, reefs
would coexist with siliciclastic beach develop-
ment near the shore.
~

g >.2
Diagenetic features
The sediments we have been dealing with show
0
evidence of meteoric as well as marine influence
during early diagenesis.
This section describes the diagenetic features
that characterize each of the facies associations
making up the transgressive shore systems. As
basic units in these systems, we have considered:
(1) alluvial fan sediments, (2) transgressive beach
.~. ~ systems, and (3) coral and red-algal reefs,
developed in front of and on the beach ridges.

Alluvial fan sediments

"7,~ A haematitic matrix and haematitic coatings
No surrounding siliciclastic grains are characteristic
features of the alluvial fan sediments of the Vic
area. Iron oxides were probably derived from
iron-containing minerals; probably biotites and
amphiboles, which are the most abundant mafic
~ ~ components in the likely source area (granitic
and metamorphic rocks). The transformation of
biotites into iron oxides is common in the studied
continental rocks.
The iron oxides are likely to have formed once
sediments were stabilized and buried in a
phreatic environment with an oscillating water-
table, in a similar way to that suggested by
Walker & Waugh (1973), Walker (1976) and
~ o
~9 ~-~
Glennie et al. (1978). However, iron coatings can
also develop in meteoric phreatic conditions
(Kessler 1978) or in vadose conditions (Turner
1980). We have not found any evidence to favour
any individual origin for these coatings.
Primary porosity in these sediments was
probably negligible: the haematitic matrix filling
nearly all interparticle spaces. All residual pores
are now filled in by sparry calcite cement,
probably originating from precipitation in mete-
ee oric water conditions.
r ~.~ Alluvial fan sediments are grey at the top.
These grey sediments show sedimentary struc-
tures similar to those in the red ones. The upper
.~',~
contact of the grey sediments corresponds to the
alluvial fan top, whereas their lower contact is
irregular and cross-cuts the sedimentary struc-
tures of the red alluvial fan deposits. In some
cases there are patches of red sediments sur-
rounded by grey ones at the top of the alluvial
fan.
 Mixed-water dolomitization 127
/,Depositional beach
1 ,y .... ..... Beach progradotion

,~ Subsidence
~ ~ / (alluvialfan~
Erosion

Erosion surface Ascending sea level change

(non quantitative}
Erosional beach

3 J

2 Ero~onol beach ~ r f a c e

9 "~. ... ~__ --~a~"--'-: ~ // k.,Ullllll•lllaJ sedi,,rne,,,o

Depositionol beach

FIG. 4. Diagrams showing the development and relationships between depositional and erosional beaches as a
result of differences in subsidence rate during the Lutetian-Biarritzian transgressive stage. The step-like
disposition of these beach systems suggests that besides transgression, tectonic activity controlled their
occurrence.

Carbonaceous remains are commonly pre- lime and clay mud surrounding the siliciclastic
served in the grey sediments. This fact, together components.
with the grey colours themselves, may be Up to 125 ~tm, dolomite rhombohedra are
evidence of the influence of a reducing environ- limpid and euhedral, except for the smallest
ment. Iron oxides originating in meteoric phrea- crystals, where a form is difficult to assess. Very
tic conditions would have been reduced in this small (up to 10~tm) crystals are arranged in
zone. The mean thickness of the reduced homogeneous aggregates of sucrosic aspect (Fig.
material ranges from 0.5 to 1 m. 5a) surrounding the larger ones.
The matrix in the reduced section has less Dolomite rhombohedra have been dissolved in
terrigenous components than in the oxidized the interpenetration contacts with siliciclastic
equivalents. This may be related to a leaching components (Fig. 5b, c). Dolomites also accumu-
effect. Dolomite rhombohedra are abundant in late along pressure-solution surfaces which, in
the reduced material, increasing in abundance turn, affected them (Fig. 5d). These facts suggest
from the lower to the upper part of the section, that dolomitization occurred before the compac-
but never exceeding 10~o of the total rock tion of sediment during early burial.
volume. At the top, dolomite rhombohedra are
the main component of the matrix between the
siliciclastic grains (Fig. 5a). The transgressive beach systems
The dolomite crystals are from 1 to 325 ~tm in
size. The largest ones (> 125 ~tm) show the most Deposits of the transgressive beach systems show
perfect shapes and are usually zoned (Fig. 5b, c): evidence of cementation through the influence of
They have dusty nuclei and cleaner well-zoned marine waters during early diagenetic stages.
overgrowths, and usually include traces of the However, they also have textures that may be
original clay matrix (Fig. 5e). This suggests interpreted as having originated under the
poikilitic growth in an unlithified matrix. The influence of meteoric waters, also during early
largest crystals are generally found isolated in diagenesis.
I28 C. Taberner & C. Santisteban

~J .-,

.o .o

.~ 8 o".~ ~.~.

~'.~

~ ~ . ~ _

~= ~ ~

N~o .'C N
~ "~ "~ ..~ "~
9~ ~

~ 0~.0~-~

o~ 0 ~ 0 ~0
.~'---

~ .~

~.~ 0~0 ~

~ ~.~0~

~ 0 ~
~ a D O 0 ~

0~.~ ~ . ~

"~ ~ . -

9 -~

~do~N
Mixed-water dolomitization I2 9

oDO

._ 0 ~ C

~ ~.~
~ . ~ ~

- - , ~ ,.~ .=

~ o ~ ~.~'~

9"~ ~ = ~o~.

~.~ 0 ~ "~ ~-~

~.~ ~ "~ g ~o
I3O C. Taberner & C. S a n t i s t e b a n

Palisade fringes of low-magnesium calcite (Taberner & Bosence 1985) seen in the hetero-
crystals developed around siliciclastic grains. geneity of colours and components in the
These crystals are 35/am wide and up to 75 lam different generations of borings. Usually, the
long. They are prismatic in shape (Fig. 5f); and first matrix generations may be distinguished by
show blunted crystal terminations. Micritic darker colours and greater percentages of silici-
envelopes, surround these cements and probably elastic components, while the younger ones are
contributed to their preservation. lighter and purer. From this we can deduce that
The above-described cements were probably cement and/or internal sediment increase from
not originally low-magnesium calcite. This may the first to the later generations of borings. This
be inferred from the fact that they show evidence leads us to the hypothesis that the cement
of having been dissolved during early diagenesis. responsible for cementation of the micrite matrix
This is more obvious in cases where a micritic of these carbonates had high-magnesium calcite
envelope formed around the fringe of cement. mineralogy, analogous to submarine micrite
The space previously occupied by the fringe is cements from recent carbonate platforms (Bath-
now represented by a microsparitic cement, urst 1971 and James & Ginsburg 1979).
displaying two infilling generations, the first Continuous isopachous fringes of fibrous
with a drusy aspect (Fig. 5g, h). crystals are occasionally found infilling interpar-
The palisade cements possibly were composed ticle cavities (i.e. forams, bivalves, etc.). Crystals
originally of aragonite or high-magnesium cal- vary between 10 and 100 lam in length and are up
cite and therefore they may have formed in a to 1 ~tm in width. They are usually included in
submarine environment. Their textural features, larger low-magnesium calcite sparite crystals
dimensions and similarity to stubby marine (Fig. 6a) and occasionally their shapes are
cements described in recent environments by outlined by micrite relics in the sparry calcite.
Schroeder (1972), James et al. (1976) and James Neomorphic processes enabled the shapes to be
& Ginsburg (1979), lead us to assume that they preserved in the sparite cements. The occurrence
would have originally been high-magnesium of these processes lead us to assume that the
calcites. Their dissolution or neomorphism to previous mineralogies might have been high-
low-magnesium calcite is most likely to have magnesium calcite or aragonite. They particu-
occurred under the influence of meteoric waters. larly resemble recent aragonite cements de-
Scattered dolomite rhombohedra have also scribed by Macintyre et al. (1968), Shinn (1969)
been found growing in the matrix between and James et al. (1976). The fringes are affected
grains. Their textural features and dimensions by borings, and small fringes of the same kind of
are similar to those described above in the cement may develop in boring cavities. From
decoloured sediments at the top of the alluvial this we can assume they are of marine origin.
fan deposits. However, they are much less Former acicular cements are also present as
abundant. radial aggregates of crystals nucleated at one
point (Fig. 6c). They are particularly common in
the interlaminar porosities of red algal crusts
Coral and red algal buildups (Taberner & Bosence 1985), where coalescent
Reef deposits show evidence of having been aggregates are found (Fig. 6d).
early-cemented under the influence of marine Dimensions of acicular crystals composing the
waters; but again during early diagenesis these fan-like aggregates vary from 100 to 6001_tm.
cements would have been influenced by the They are up to 7 lam wide, and usually the largest
movement of meteoric waters. crystals are found in the most central part of the
Two kinds of cements of likely marine origin fan-like structure. The greatest variations in
are common in the reef carbonates: (a) micrite length of the acicular crystals are found in cases
cements, which were possibly originally high- where the fan-like structures are coalescent.
magnesium calcite, and (b) isopachous fibrous Isolated aggregates growing in free conditions
cements and botryoidal cements, both of which tend to have a semicircular external shape and
were probably originally of aragonitic the crystals tend to be the same size. A
composition. hemispheric shape may be deduced for these
The reef sediments are now composed of low aggregates from the different forms they show in
magnesium calcite, but textural features suggest thin-section (Fig. 6c, d).
that the micrite matrix was cemented by an early There is evidence that this kind of cement
high magnesium calcite cement: (1) borings developed in submarine conditions: it is bored
affecting the micrite matrix as well as crusts of together with hard skeletal components (coral-
red algae (Taberner & Bosence 1985) are line red algal crusts) and is found in cavities that
common (Fig. 6b); (2) the matrix shows different have also been filled in with lime mud, cemented
stages of boring, infilling and cementation in the marine environment. This type of cement
Mixed-water dolomitization 13

"~ 0 ~

g2 ~.

.~,~ ~
~ N

_9 2. ~,

~...~ ~ ' ~ = ~

~,,..~ ~ ~ ~'"~

. - ~ ~
~,~.. = :~.~.
~"U~ h "u"

,~ ~ ~ ~~
I3 2 C. Taberner & C. Santisteban

is common among marine carbonate deposits have been made of bulk samples. Microsampling
from the studied area, while it has never been was possible for sparitic cements. Separation of
found in the continental sediments. dolomite crystals has not been possible in the
The shapes and arrangements of the cements dolomitized rocks due to their small size,
described here are similar to those of recent proportion and scattered distribution in the
botryoidal aragonitic cements described by matrix. However, samples have been enriched in
Schroeder (1973), James et al. (1975) and dolomite by means of selective dissolution in cold
Ginsburg & James (1976). However, in this case 1 M acetic acid for half an hour. Original
original crystals forming the radial aggregates dolomite/calcite ratios are about 5~ (on average)
were neomorphosed to low-magnesium sparry and after treatment, about 60%
calcite which inherited the radial arrangement, For comparison sampling has also been car-
and undulose extinction of the precursor (Fig. 6c, ried out in a younger (Priabonian) barrier-reef
d). The undulose extinction might reflect the sequence, where 60m thick reef carbonates
retention of subcrystalline domains of possible represent the record of almost continuous marine
primary origin as proposed by Sandberg (1985). sedimentation. Petrographic and textural fea-
In our case the subcrystalline domains corre- tures in these deposits point out that early
spond to the original radial-acicular crystal diagenesis occurred in a marine environment.
fabric, observed through the presence of micrite Results from the barrier-reef deposits have been
relics. Acicular crystals arranged in a fan-like taken to represent the geochemical trends in
disposition would represent former submarine deposits mainly influenced by marine diagenesis.
botryoidal aragonite cements. Neomorphism Samples from this reef have been studied using
into low-magnesium sparry calcite might be the same methods as for rocks from the near-
related to later influence of meteoric waters. shore systems.

Trace elements
Isotopic composition and trace The trace element values for beach-ridge sys-
element distribution tems and associated sediments (BS), as well as in
barrier-reef carbonates (R) are shown in Figs 7
The diagenetic evolution of these nearshore and 8. These figures show the behaviour of
deposits will be better constrained on the basis of strontium, manganese and iron (Fe 2§ contents.
geochemical results, our preliminary results are These elements may give some idea of the
presented here. influence of meteoric waters (high Fe 2§ and
Carbon and oxygen isotopic composition and manganese contents) or a marine influence (high
trace element concentrations have been deter- strontium values) (Veizer 1983).
mined. The most useful trace element data are Manganese and iron show good correlation in
strontium, manganese, zinc and iron (Fe2+). Fe 2§ the samples of the depositional beach systems,
levels have been determined using colorimetric where the highest concentrations of both ele-
(1,10-Phenanthroline) methods, and the other ments have been found. The lowest manganese
elements through atomic absorption spectropho- and iron contents correspond to samples from
tometry of the acid (cold 10~ HCI) soluble dolomitized decoloured levels and to dolomite-
fraction. A Link System energy dispersive X-ray enriched samples. An increasing general trend
microanalysis system and JEOL JSM-840 SEM from decoloured levels towards coralgal reefs can
have been used to detect trace elements in be seen for both Mn and Fe 2+ (Fig. 7c, d).
individual crystals. Strontium enables discrimination of two well-
Oxygen and carbon isotopic composition has defined fields representing the beach system
been determined using a VG SIRA-9 mass facies associations and barrier-reef samples (Fig.
spectrometer. Extraction of CO2 from carbon- 7a, b).
ates was carried out according to the method It is also interesting to note that the highest
described by McCrea (1950). A 6180 correction values in zinc correspond to the nearshore
factor of -0.8%~ has been applied to compare sediments (100-170 ppm), while in the barrier-
dolomite with calcite results. The dolomite/cal- reef rocks this element is usually not detected.
cite ratio has been considered when correcting Sodium values in bulk samples of nearshore
for the possible differences in phosphoric acid facies associations range from 284 to 340 ppm.
fractionation of dolomite and calcite in the However, sodium presence has been easily
studied samples. Results are expressed with detected in dolomite crystals through energy
respect to PDB standard. dispersive microanalysis. This means that
Individual petrographic phases were generally sodium contents in dolomite crystals may be
too small to be separated, so in general analyses much higher.
 Mixed-water dolomitization I33
m Fe §176

5000 .".....
BS ,, _ ",

R
0.05 :/i..'"!ii

I.!.
0.02

.~. -D D~,
.

(o/ iI- "'~-/ , (b) s, . SI

2o~3o Sr 0.001 0.602 mSr
0.2-
m Fe++"
Fe § race

5000- .':aD
Rt ~ ."~176
j ": /
/ ~ .." - BS
0.05- BS ..""
/ o'."A / !
/ ".~" / 9 ".:

9I 9 e/ ..,
,*"" "A :"

I." i ,," 0.02

. 9 / ."
..."" ~ ...."

(c) ,D~;:;s / 600 (d) ':~"

O.()02 mMn 0.065
ldn
mCa
-6 ,~'3C - 4 -2 2
I I I I I i i i

-2

9" i'"'"'.
....-;; .)
8s . . - " " ~ ' " , " :g-,
,-" \
/ x
\

:,"- ~" " ~ ~ _ ..... -

9 S 9 R
__ _e_ /
".~" .-s

(e) ~ , i o -"2 (f)

-6 &~3C -4

~k Decoloured (oliuvioi-fan}
-8
9 Beach sediments S Sporite
, Corolgol reef D Dolomite -enriched
9 Barrier-reef

FIG. 7. Binary diagrams showing the geochemical pattern in the samples from the Vic area. BS" Beach-ridge
systems; R: Barrier-reef. Values in (a) and (c) are in ppm. In (b) and (d) Fe 2§ Sr and Mn values are
expressed in relation to Ca. As analyses were performed on bulk samples these ratios may give a clearer idea
of the real variation in the carbonate fraction.
I34 C. Taberner & C. Santisteban
mSr mMn m Fe+ +
mCo mCo m ~"o

"' . . . . . . . . . . " . I d ~ "Ik "" ~I~ d ~ d

"X" Decoloured alluvial-fan
BS ....~,-"" _ _ ..~"

.... ~. ~ --- -~t & Beech sediments

. . . . . . . . . . . . . . . . . . . -'-;~. =-. . . . . . ."t.:~ . . . . . . . . . . . . . . t Coralgol reef

9 Barrier- reef

d Dolomite enriched

<.._ S Sparite

-5 "4 "3 '.-2 "1

I0 I0 I0 I0 I0

FIG. 8. Distribution of the molar ratio values of Sr, Mn and Fe 2+ with respect to Ca in the samples analysed
in the Vic area. BS: Beach-ridge systems; R: Barrier-reef. Variations in beach system molar ratios differ from
those in the barrier-reef samples. Sr/Ca values are much lower in the beach rocks, while Mn/Ca and Fe2+/Ca
are higher. These relationships point to a meteoric influence in beach sediments and a marine signature in
barrier-reef deposits.

Our trace element data may be interpreted in +0.6%0 as 61SO%oPDB ranges from -5.8%0 to
terms of variation in marine or meteoric influ- - 3.2%0. The lowest values in oxygen and carbon
ence using the approach outlined by Veizer correspond to the decoloured facies, and there is
(1983). The high manganese, iron and zinc levels a progressive shifting towards heavier values in
and low strontium contents in the beach-system more marine samples (reefs). The trend in the
rocks may be taken as evidence of meteoric isotopic values (Fig. 7f) corresponds to the facies
water influence. The sodium concentrations in succession from continental to marine sediments
dolomite crystals may point to a marine influence in the beach system. This is also reflected in the
in the dolomitization process, however, Busen- diagram of Fig. 7(e), where manganese and 613C
berg & Plummet (1985) question the use of ratios show a good correlation in the beach
sodium as a marker for marine conditions. The system deposits.
trace element distributions (Fig. 7) clearly The heavy 613C values recorded from the reefs
demonstrate greater marine influence in the associated with beach levels and in reefs at the
barrier-reef samples, where strontium values top of the marine series may be taken as being a
may be up to 2100 ppm. primary marine signature. Lighter 6180 and
negative 613C values may represent the influence
of meteoric water (Land et al. 1975 and Anderson
Isotopic ratios
& Arthur 1983). The oxygen isotopic values
Isotopic ratios in samples from the beach might seem to be too light to be primary marine
sediments (BS), as well as from the barrier-reef values. It is interesting to note that Renard et al.
carbonates (R) are shown in Fig. 7(f). (1986) give 6180 values between -2%o and
6~3C%oPDB against manganese contents have -2.5%o for Eocene pelagic carbonates in North
also been plotted (Fig. 7e). These variables have Atlantic sites. General values for pelagic carbon-
been chosen as they can both be markers of ates proposed by Hudson (1977) are from 0%0 to
meteoric influence (613C%oPDB negative values, + 2%o. Thus it would seem that isotopic values in
Land et al. 1975, and Anderson & Arthur 1983, marine zones might have been lower during the
and high Mn contents, Veizer 1983). In our Eocene in areas connected to the Atlantic such as
samples, the diagram in Fig. 7(e) enables a good in the case of the Eocene Catalan Basin.
discrimination between sample sets.
Isotopic values are in general negative. Posi-
tive 613C values have been found only in samples
from the barrier-reef and in reefs related to the Mechanism of dolomitization
beach systems (Fig. 7e, f). The samples from the
barrier-reef plot as a cluster (Fig. 7f) whilst those Dolomitization is a widely reported process. One
from the beach systems show a linear relation- may find compilations in Land (1980, 1983) and
ship: 613C%oPDB values range from -3.5%0 to in Machel & Mountjoy (1986). One of the
 Mixed-water dolomitization 135

frequently suggested mechanisms is for dolomiti- 6180%oPDB from - 5%0 to + 3.5%o ( - 2%o to
zation in zones of mixed marine and meteoric -2.5%0 on average).
water. We suggest that this is the mechanism for There is no doubt that these dolomites were
dolomitization in our rocks based on the whole of formed during early diagenetic stages. This can
the petrologic, diagenetic and geochemical fea- be seen from their textural features and also from
tures observed in the sediments from the conti- the fact that there is a clear sedimentological
nental to marine transition. These features are: control in their distribution.
Taking into account the isotopic values and
(1) The dolomitized sediments are laterally and trace element contents, we do not think that they
vertically restricted to sediments represent- could have developed by burial-related mechan-
ing the continental to marine boundary. isms or be related to hypersaline conditions. If
(2) The dolomitized zone coincides with the these dolomites had originated from burial,
decoloured zone at the top of the alluvial-fan where brines would be related to formation
sediments. This zone is likely to have been waters or come from basinal shales, there would
decoloured through the influence of a reduc- be a greater development of dolomites in the
ing environment. basinal areas, (Dunham & Olson 1980); in that
(3) Evidence of cementation in a marine diagen- case, the calcite-dolomite distribution would be
etic environment in the overlying sediments the opposite to the present one. If these dolomites
(beach and reef levels) implies that there may were related to a hypersaline environment, trace
have been a transition zone with mixed element distribution would be different, isotopic
marine and continental influences in the ratios would be heavier and there would also b e
upper part of the continental succession. evidence of simultaneous precipitation of eva-
(4) Dolomite developed during early diagenetic porite minerals.
stages before compaction. Dolomite crystals We suggest that these dolomites originated in
show interpenetration contacts with silici- a mixed water environment in nearshore areas
clastic grains, and are affected by pressure- where marine waters invaded continental sedi-
solution surfaces (Fig. 5b, c, and d). ments with existing meteoric porewaters. The
(5) Dolomite rhombohedra are usually limpid necessary Mg 2+ supply would come from marine
although occasionally they have a nucleus waters, and dolomite formation would be
with matrix-inclusions. This kind of dolo- favoured by the dilution of marine waters by the
mite, with early stages being inclusion rich meteoric ones (Badiozamani 1973, Land 1973a,
and later stages of zoned growth, has been b and Folk & Land 1975). In this way the amount
found in other dolomites thought to have of dissolved salts in marine waters would have
formed in mixing-zones (Folk & Land 1975, diminished, and, according to Baker & Kastner
Land et al. 1975, Land 1983, Ward & Halley (1981) made it possible to overcome the inhibit-
1985 and Machel & Mountjoy 1986). ing effect of seawater sulphate.
(6) Trace element contents (high Mn, Zn and Dolomite isotopic compositions are lighter
Fe 2§ and low Sr contents) in the sediments of than those found in the mixing zone dolomites
the beach system might be evidence of studied by Zhao & Fairchild (this volume).
meteoric influence in all the facies. Accord- However the shift towards heavier values in 613C
ing to Dunham & Olson (1980) and Veizer and 6~sO from the decoloured sediments to the
(1983), low strontium contents are related to reefs associated with beach-levels gives the idea
solutions whose Sr~+/Ca 2§ ratios are lower of a seaward-decreasing influence of meteoric
than that of sea water. Land (1980) assumes waters both in lateral and vertical directions. The
that the low values in strontium in ancient progressive increase towards heavier isotopic
dolomites may be related to schizohaline values is considered to reflect flux gradients in
environments. High values in the other meteoric aquifers flowing in marine-influenced
elements might show the influence of mete- environments (Land et al. 1975 and Meyers &
oric waters (Veizer 1983). Lohmann 1985). It can be assumed that there
(7) Isotopic compositions for ~180%oPDB and was a flux of fresh ground waters from the
613C%oPDB are similar to the values quoted subaerially exposed areas of the alluvial fan
by various authors for mixing-zone dolomites towards the marine areas (Fig. 9). However, the
(Land et al. 1975, Dunham & Olson 1980 and inflow of marine waters in the Lutetian-Biarrit-
Zhao & Fairchild, this volume). However, it zian beach-ridge systems would be downwards
seems that isotopic values may not be because of the transgressive disposition of the
conclusive as Mattes & Mountjoy (1980) marine facies associations over the continental
have found values in the literature ranging in ones (Figs 4 and 9). The marine water flux
613C~oPDB from -9%0 to +4%0 (about direction would be opposite to that of the
-2.5%0 on average) and ranging in meteoric waters (Figs 9 and 10).
I36 C. Taberner & C. Santisteban
Beach sediments

1 ~ ,~, ~ ~ ~~~L.~ meteoric pheatic

~--~_~--~..~L,,~ I " ~ ' ~ -'~l~*J~ ~ continentalsediments

-+,.,o0 zooe

" / " " " " marine phreatic "~,~.

reduced zone Alluvial-fan

~ ~ ~ ~ - .~_\. .. ~
. . . . . . . . . ~,~__~..r'7 ~ ' - . Z , - ' ~ - , . ~ r - ~ . 4 . / ~. _.~..
. . . . . .

+. . o ~ ~ .i.-- -- + \ \
- . ~ - ' " ~ ~ _ non- reduced-patches

9 " " " I o + . ~

~ ~ ; .---~~--r~t I~,1 i r
+ I l-I "r 4" ' .edoced o.d do,om,,+zed zone

FIG. 9. Schemes showing progressive migration stages of sedimentary bodies in beach-ridge systems. This
migration caused chemical changes in porewaters. Previously oxidized continental sediments are reduced in
their uppermost part, just below beach levels. The reduced area also shows evidence of early dolomitization.

S o m e final c o m m e n t s there was both a sedimentary and palaeogeogra-

phic control in the formation and distribution of
The facies associations of the Middle Eocene dolomites at the top of the alluvial fan sediments.
nearshore sediments in the SE margin of the In such sediments it may be difficult to
Catalan Basin are closely interrelated in time and separate the marine diagenetic signatures from
space. This should have had significant effects the continental ones, because of oscillations and
on the diagenesis. Both meteoric-influenced and migrations of the diagenetic environments. We
marine-influenced diagenetic environments can find evidence of meteoric diagenesis as well
should have changed in time and space. The as of marine and mixed-water diagenesis. In this
mixing zone should also reflect these changes sense, Land (1980) has already explained the
(Fig. 9), but generally it should have been located difficulties in interpreting the isotopic values in
at the top of the alluvial fan deposits, just in the rocks subjected to mixing-zone processes. In our
zone representing the vertical transition from case, we are dealing with a diagenetic situation
continental to marine sediments (Figs 9 and 10). related to the deposition of marine facies
There is a coincidence in the distribution of associations on continental deposits. This situa-
dolomitized zones and beach-ridge systems tion would have brought about the modification
which suggests that dolomitizing fluids flowed in of meteoric porewaters by the inflowing marine
from the upper marine sediments. Therefore, waters (Figs 9 and 10). Most cases referred to in
 Mixed-water dolomitization I37

9 /
9..: ..i ~ MW

T : Tronsgression
MW: Mixing water zone
M: Meteoric recharge

FIG. 10. Diagenetic model of dolomitization proposed for nearshore complexes at the SE margin of the
Catalan Basin. Dolomitization at the top of the alluvial fan sediments developed during early diagenetic
stages as a consequence of the transgression which allowed for the inflow of marine waters into continental
sediments where existing porewaters had a meteoric origin. Recharge of meteoric waters would have been
through subaerially exposed continental deposits. Marine waters were mixed with the meteoric ones in the
zone corresponding to the decoloured levels.

the literature are the opposite to this and occur the mixing zone by the inflow of marine waters in
where marine waters are modified by incoming meteoric aquifers, establishing whether these
meteoric ones. Land (1980) also raises the different situations are responsible for variations
question of the problems involved in the in the isotopic composition or in the trace
extrapolation of models with different dimen- element distribution, or whether there is any
sions and sedimentation rates. In this case, as we evidence here for more conventional mixing
have already mentioned, there is a pulsating where marine fluids are displaced by meteoric
variation in the sedimentation rates. The ter- ones.
races represent zones, erosional for a long time,
which changed suddenly into sedimentation ACKNOWLEDGMENTS:We are indebted to Dr C. Pierre
areas with general marine features (Fig. 4). (Laboratoire de G6ologie Dynamique de l'Universiti6
These depositional zones are narrow and usually Pierre et Marie Curie, Paris VI), who performed
arranged over the continental deposits. The analysis of the isotopic composition in the studied
pulsating dynamic pattern in the sedimentary samples.
environments might have influenced the hydrol- We very much appreciated the help of R. Fontarnau,
C. Carulla and A. Dominguez for instruction in the use
ogy of porewaters and, consequently, the diagen- of the electron microscope and analysing system and
etic products. the 'Servei de Microscopia Electr6nica de la Universi-
The collection of more data will perhaps tat de Barcelona' for use of equipment and laboratory
enable us to detect variations in the meteoric and facilities. We also are grateful to the 'Servei d'Analisi
marine water-tables during the Middle Eocene Quimica de la Facultat de Geologia' and especially to
transgressive stage in the Catalan Basin. Some R. Marim6n, J. Sfinchez and E. Balart for their help
aspects that will also have to be considered to with trace elements. The help of J. Ros with the
achieve this are: (1) the variations in sedimenta- drawings was also of great value.
tion rates throughout the different areas; (2) the We wish to convey our deep appreciation to Dr
Marshall and Dr Fairchild for their constant support
area influenced by each successive mixing-zone; and continuing effort in reviewing this paper. Their
(3) the relative velocities of transgression-regres- comments were extremely valuable. We also wish to
sion and (4) to decide in each case if nearshore acknowledge Dr Nichols for his kind help on sedimen-
dynamics could have enabled the development of tological aspects.

References
ANDERSON, T. F. & ARTHUR, M. A. 1983. Stable BADIOZAMANI,K. 1973. Dorag dolomitization model,
isotopes of oxygen and carbon and their applica- Ordovician. Wisconsin. Journal of Sedimentary
tion to sedimentologic and paleoenvironmental Petrology, 43, 965-84.
problems. In: Stable Isotopes in Sedimentary BAKER, P. A. & KASTNER,M. 1981. Constraints on the
Geology. Society of Economic Paleontologists and formation of sedimentary dolomite. Science, 213,
Mineralogists. Short Course, 10, 1.1-151. 214-6.
 I38 C. Taberner & C. Santisteban
BATHURST, R. G. C. 1971. Carbonate Sediments and Economic Paleontologists and Mineralogists.
their Diagenesis, Elsevier, Amsterdam. Short Course, 10.
BUSENBERG, E. & PLUMMER, L. N. 1985. Kinetic and --, SALEM, M. R. I. & MORROW, D. W. 1975.
thermodynamic factors controlling the distribu- Paleohydrology of ancient dolomites. Geochemi-
tion of SO42- and Na § in calcites and selected cal evidence. American Association of Petroleum
aragonites. Geochimica et Cosmochimica Acta, 49, Geologists Bulletin, 59(9), 1602--25.
713-25. MACHEL, H. G. & MOUNTJOY, E. W. 1986. Chemistry
CARTER, R. W. G. 1986. The morphodynamics of and environments of dolomitization--a reapprais-
beach ridge formation; Magilligan, Northern al. Earth Science Reviews, 23(3), 175-222.
Ireland. Marine Geology, 73 (3/4), 191-214. MACINTYRE,J. G., MOUNTJOY, E. W. & D'ANGLEJAN,
COLOMBO, F. 1980. Estratigrafia y sedimentologia del B. F. 1968. An occurrence of submarine cementa-
terciario inferior continental de los Catal~mides. tion of carbonate sediments off the west coast of
Unpublished PhD Thesis, University of Barcelona. Barbados W.I. Journal of Sedimentary Petrology,
DUNHAM, J. B. & OLSON, E. R. 1980. Shallow 38, 660-3.
subsurface dolomitization of subtidally deposited MATTES, B. W. & MOUNTJOY, E. W. 1980. Burial
carbonate sediments in the Hanson Creek Forma- dolomitization of the Upper Devonian Miette
tion (Ordovician-Silurian) of Central Nevada. In: Buildup, Jasper National Park, Alberta. In:
ZENGER, D. H., DUNHAM, J. B. & ETHINGTON, ZENGER, D. H., DUNnAM, J. B. & ETHINGTON,
R. L. (eds). Concepts and Models of Dolomitization, R. L. (eds). Concepts and Models of Dolomitization,
pp. 139-61. Society of Economic Paleontologists pp. 259-97. Society of Economic Paleontologists
and Mineralogists. Special Publication, 28. and Mineralogists. Special Publication, 28.
FOLK, R. L. & LAND, L. S. 1975. Mg/Ca ratio and MCCREA, J. M. 1950. On the isotopic chemistry of
salinity: two controls over crystallization of carbonates and a paleotemperature scale. Journal
dolomite. American Association of Petroleum Geolo- of Chemistry and Physics, 18, 849-57.
gists Bulletin, 59, 60-8. MEYERS, W. J. & LOI-IMANN, K. C. 1985. Isotope
GINSBURG, R. N. & JAMES, N. P. 1976. Submarine geochemistry of regionally extensive calcite ce-
botryoidal aragonite in Holocene reef limestone. ment zones and marine components in Mississip-
Belize. Geology, 4(7), 431-6. plan limestones, New Mexico. In: SCHNEIDER-
GLENNIE, K. W., MUDD, G. C. & NAGTEGAAL,P. J. C. MANN, N. & HARRIS, P. M. (eds). Carbonate
1978. Depositional environment and diagenesis of Cements, pp. 223-39. Society of Economic Paleon-
Permian Rotliegendes sandstones in Leman Bank tologists and Mineralogists. Special Publication,
and Sole Pit areas of the U.K. Southern North 36.
Sea. Journal of the Geological Society of London, REGUANT, S. 1967. E1 Eocene marino de Vic. Memorias
135, 25-34. del Instituto Geolbgico y Minero de Espaha, LXVII,
HANSHAW, B. B., BACK, W. & DEIKE, R. G. 1971. A 1-330.
geochemical hypothesis for dolomitization by RENARD, M., BERTHENET, F., CLAUSER, S. & RIGHE-
ground water. Economic Geology, 66, 710-24. ~:)IS, G. 1986. Geochemical events (trace elements
HUDSON, J. D. 1977. Stable isotopes and limestone and stable isotopes) recorded on bulk carbonates
lithification. Journal of the Geological Society of near the Eocene-Oligocene boundary. Applica-
London, 133, 637-60. tion to the Contessa Section (Gubbio, Umbria,
JAMES, N. P. & GINSBURG, R. N. 1979. Morphology, Italia). In: POMEROL, C. & PREMOLI-SILVA, I.
sedimentology, organism distribution and Quar- (eds). Terminal Eocene Events, pp. 331~48.
ternary history. In: The Seaward Margin of Belize SANDBERG, P. 1985. Aragonite cements and their
Barrier and Atoll Reefs. International Association occurrence in ancient limestones. In: SCHNEIDER-
of Sedimentologists. Special Publication, 3. Black- MANN, N. & HARRIS, P. M. (eds). Carbonate
well Scientific Publications, Oxford. Cements, pp. 33-57. Society of Economic Paleon-
KESSLERII, L. G. 1978. Diagenetic sequence in ancient tologists and Mineralogists. Special Publication,
sandstones under desert climatic conditions. Jour- 36.
nal of the Geological Society of London 135, 41-9. SAYrISTEBAN,C. & TABERNER, C. (in press). Sedimen-
LAND, L. S. 1973a. Contemporaneous dolomitization tary models of siliciclastic deposits and coral reef
of Middle Pleistocene reefs by meteoric water, interrelation. In: Carbonate-Clastic Transitions.
North Jamaica. Marine Science Bulletin, 23, 64-92. Developments in Sedimentology. Elsevier,
--1973b. Holocene meteoric dolomitization of Amsterdam.
Pleistocene limestones, North Jamaica. Sedimen- SCHROEDER, J. H. 1972. Fabrics and sequences of
tology, 70, 411-24. submarine carbonate cements in Holocene Ber-
- - 1980. The isotopic and trace element geochemis- muda cup reefs. Geologische Rundschau, 61, 708-
try of dolomite: the state of art. In: ZENGER,D. H., 30.
DUNHAM, J. B. & ETHINGTON, R. L. (eds). Submaine and vadose cements in Pleisto-
- - 1 9 7 3 .

Concepts and Models of Dolomitization, pp. 87- cene Bermuda Reef-Rock. Sedimentary Geology,
110. Society of Economic Paleontologists and 10, 179-204.
Mineralogists. Special Publication, 28. SHINN, E. 1969. Submarine lithification of Holocene
- - 1983. The application of stable isotopes to studies carbonate sediments in the Persian Gulf. Sedimen-
of the origin of dolomite and to problems of tology, 12, 109-44.
diagenesis of clastic sediments. In: Stable Isotopes TABERNER, C. 1982 (1983). Evoluci6n ambiental y
in Sedimentary Geology, pp. 4.1-22. Society of diagen&ica de los dep6sitos del Terciario inferior
 Mixed-water dolomitization 139
(Paleoceno y Eoceno) de la cuenca de Vic. PhD Permian in Central West and South Europe, pp.
Thesis. Universidad de Barcelona. 240-82. NATO Advanced Study Institute Series:
- - & BOSENCE,D. W. J. 1985. Ecological sucession Series C. Mathematical and Physical Sciences.
from corals to Coralline algae in Eocene patch Reidel, Dordrecht, Holland.
reefs of N. Spain. In: TOOMEY& NITECKI (eds). - - • WAUGH, B. 1973. Intrastratal alteration of
Paleoalgology : Contemporary Research and Appli- silicate minerals in late Tertiary fluvial arkose.
cation, pp. 226-36. Springer-Verlag, Berlin. Baja California. Mexico. Geological Society of
TURNER, P. 1980. Continental red beds. In: Develop- America Bulletin. Abstracts with Program, 7, 853-
ments in Sedimentology, 29. Elsevier, Amsterdam. 4.
VEIZER, J. 1983. Chemical diagenesis of carbonates: WARD, W. C. & HALLEY,R. B. 1985. Dolomitization in
theory and application of trace element technique. a mixing zone of near-seawater composition. Late
In : Stable Isotopes in Sedimentary Geology, pp. 3.1- Pleistocene, northeastern Yucatan Peninsula.
100. Society of Economic Paleontologists and Journal of Sedimentary Petrology, 55, 407-20.
Mineralogists Short Course, 10. ZHAO, X. & FAIRCHILD,I. 1987. Mixing zone dolomiti-
WALKER, T. R. 1976. Diagenetic origin of continental zation of Devonian carbonates, Guangxi, South
red beds. In: FALKE, H. (ed.). The Continental China. This volume.

C. TABERNER, Departament de Geoquimica, Petrologia i Prospecci6 Geol6gica, Fctat. de

Geologia, Universitat de Barcelona, Gran Via 585, 08007 Barcelona, Spain.
C. SAYrISTEBAN, Departament de Geologia, Fctat. de Biologia, Universitat de Valencia,
C/Dr Moliner 50, Burjassot, Valencia, Spain.
 Early diagenetic phosphate cements in a turbidite basin

R. D. A. Smith

S U M M A R Y: Early diagenetic phosphate cements of late Llandovery age developed in

fine-grained turbidites deposited in the Welsh Basin. Bottom water conditions fluctuated
between dysaerobic and anaerobic. Phosphate enrichment of layers close to the tops of fine-
grained turbidites occurred very rapidly after turbidite deposition. Early phosphate cement
precipitation caused lithification of these layers prior to significant sediment compaction.
The cements consist of apatite crystallites (0.5-5 ~tm in diameter) which filled pore spaces
and also grew displacively in the cleavages of detrital phyllosilicates. Two possible
mechanisms for the localization of cements close to turbidite tops are (a) bacterial
concentration of phosphate in near-surface layers of transition between oxic and sulphidic
environments, and (b) adsorption of upward-diffusing dissolved phosphate from underlying
anoxic porewaters by ferric and manganese oxyhydroxides present above levels of oxygen
and nitrate exhaustion. Phosphate cements are absent from adjacent Lower Silurian shelf
facies deposited in fully oxic bottom water environments, and are also absent from turbidites
deposited in totally anoxic bottom waters. In the former case oxidation of labile organic
matter close to the sediment-water interface prevented a source of dissolved phosphate from
being buried, whereas in the latter case phosphate would have been free to diffuse out into
bottom waters.

Early diagenetic phosphate cements are a com- relationships to their enclosing event deposits.
mon feature of Lower Silurian fine-grained In this paper new fabric and mineralogical
turbidites in Wales (Cave 1979), but they have data are presented and integrated with palaeo-
not previously been studied in any detail. This geographical data to develop a qualitative model
study focuses on phosphate cements developed for the controls on marine apatite cement
in the deposits of a late Llandovery turbidite precipitation in the early Silurian Welsh Basin.
system of griestoniensis Zone age (Figs 1 and 2)
and forms part of a sedimentological basin
analysis of the Welsh Basin for this zone (Smith
1987a and in preparation). Methods
The main aims of this paper are to attempt to
answer the following questions: (1) What was the Samples collected from various mid-Wales expo-
development history of the cements relative to sures of griestoniensis Zone turbidites were
individual turbidite depositional events? (2) slabbed and polished in order to study the
What were the phosphate-concentrating pro- relationships between phosphate cements and
cesses and which factors controlled the local- individual turbidite and hemipelagic units. Co-
ization of phosphate cements within individual lours were determined with reference to a
turbidites? (3) Which factors controlled the Geological Society of America Rock Colour
large-scale (basin-wide) distribution of phos- Chart using the Munsell system.
phate cements? Epoxy resin-mounted thin sections of over 20
The griestoniensis Zone sequence is ideal for samples were prepared for study using optical
such a study in that (a) a good biostratigraphic microscopy, backscattered scanning electron
framework for correlation exists utilizing grapto- microscopy (BSEM) and energy-dispersive X-
lite zones in basinal facies and brachiopod zones ray microanalysis, (EDXRA), electron micro-
in shelf facies (Wood 1906, Cocks et al. 1984); (b) probe analysis (EMP) and cathodoluminescence
the late Llandovery palaeogeography is well- (CL). The sections were cut perpendicular to
known (Ziegler et al. 1968, Bridges 1975, Smith bedding and were polished using Buehler TEX-
1987a) and (c) the sequence contains numerous TET laps and 1 ~tm diamond paste abrasive.
reference levels, in the form of turbidite event For BSEM and E D X R A the specimens were
deposits, against which early diagenetic cement coated with 100A of carbon, mounted on
development can be considered. Since the aluminium stubs and examined in a Phillips
cements developed in a deep-water basin they 501B SEM equipped with a four element KE
escaped the winnowing and concentration pro- backscatter detector and a Link AN 10000 X-ray
cesses normally suffered by shallow water phos- microanalyser. A number of specimens were also
phate cements (e.g. Mullins & Rasch 1985, Cook studied in a Jeol JSM-840 SEM equipped with a
& Shergold 1986) hence preserving their spatial Jeol dipole solid state backscatter detector and a

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, I4I

Geological Society Special Publication No. 36, pp. 141-156.
I4Z R. D. A. Smith

SERIES STAGE ZONE | SL h TRACOR Northern TN2000 X-ray microana-

Z
lyser. BSEM produces images in which bright-
428Ma < crenulata
..... CT/ ness is proportional to mean atomic number. It
o>- griestoniensis enables petrographic examination of fine-
/
,.J grained rocks with far higher resolution than is
w
~" crispus // available using optical microscopy (Pye &
Krinsley 1984).
turriculatus /
>. / A cold-cathode luminoscope with an approxi-
tr z sedgwickii mately 25 kV generator built by W. Wilson and
ul < E R~'~ Dr J. A. D. Dickson (see Fairchild 1980) was
~> ~ convolutus / used for CL examination ot the samples.
O o / Bulk-rock mineralogies were determined by
~ argenteus
m /ET X-ray powder diffraction in a Phillips powder
z ~ /
magnus diffractometer using Cu K~ radiation. Minerals
.j /
--I triangulatus [ were identified by reference to the ASTM
/ Powder Diffraction File and data in Brindley &
<z cyphus / Brown (1980). The CO2 contents of the apatites
E< acinaces / were obtained from the positions of the (004) and
o / (410) peaks using the method of Gulbrandsen
atavus ] (1970).
/ Eight specimens were prepared for auto-
438Ma
--
~ acuminatus ] radiographic study with the aim of detecting any
FIG. 1. Biostratigraphy of the Llandovery Series and enrichments of uranium in phosphatic layers.
sea-level curve generalized from McKerrow (1979). CR39 plastic was attached to polished surfaces of
CT, continental trangression; ET, eustatic slabbed samples and exposed for five weeks. The
transgression; ER, eustatic regression. Absolute ages density of alpha tracks developed in the plastic is
are from the Harland et al. (1982) time-scale. proportional to the concentration of alpha

~ LLANDOVERY SERIES
~ griestoniensis ZONE
TURBIDITE S Y S T E M

BIOFACIES
E Eocoelia
P Pentameroides
S Costistricklandia N
C Clorinda
S
N Nereites

~S ~p

0 20km

f
i ~ I

FIG. 2. Location map showing outcrop patterns of the Llandovery Series and deposits of the griestoniensis
turbidite system, and the distribution of benthic biofacies (see Fig. 3) of griestoniensis Zone age. g denotes the
presence of pelagic faunas of graptolites. The distribution of shelly 'communities' of the shelf is after Ziegler
et al. (1968).
 Phosphate cements in a turbidite basin I43
particle-emitting elements (mainly uranium and evidence indicates a low latitude position in the
thorium). The plastic was then etched in 6 M southern hemisphere for the early Silurian Welsh
NaOH at 80~ for 18 hr to enlarge the alpha Basin (Piper 1985).
tracks and viewed down a petrological The preserved deposits of the late Llandovery
microscope. griestoniensis Zone turbidite system comprise
remnants of the distal part of a large (minimum
length of 100 km) elongate turbidite system. The
Palaeogeographical setting sequence consists of a thin-bedded fine-grained
(silt-mud) turbidite background punctuated by
The early Palaeozoic Welsh Basin was situated packets of thicker-bedded sandstones (sandstone
on continental crust close to the leading edge of a lobes). Palaeo-water depths of the order of 500-
northward-drifting microcontinental fragment, 1000 m have been estimated for the late Llando-
'Eastern Avalonia' of Soper et al. (1987). The very Welsh Basin (Smith 1987a).
basin geometry was strongly controlled by an Biofacies distributions indicate that the early
array of long-lived lineaments dominantly trend- Silurian basinal water column was stratified.
ing between NE-SW and N-S (Kokelaar et al. Dissolved oxygen levels of bottom water in the
1984, Woodcock 1984, Smith 1987a). Petro- basin were too low to support skeletonized
chemical studies of Welsh volcanic rocks indi- benthos and at times even excluded soft-bodied
cate that the basin occupied a back-arc setting benthos which enabled the preservation of fine
from early to late Ordovician (Arenig to Cara- laminations in hemipelagic facies (Smith 1987a,
doc) times (Kokelaar et al. 1984). The leading c). At present laminated sediments devoid of
edge of Eastern Avalonia may have become a macrobenthos are preserved beneath bottom
transform margin after the main, late Ordovi- waters containing less than 0.1 ml 1-1 dissolved
cian, volcanic shutdown (Leggett et al. 1983). oxygen (Savrda et al. 1984, Thomson et al. 1985).
However, documented strike-slip displacements By analogy with biofacies patterns in modern
on Welsh Lineaments affecting Lower Palaeo- oxygen-deficient basins, bottom water environ-
zoic rocks are all small (e.g. Fitches & Campbell ments in the late Llandovery Welsh Basin
1987) and were probably mainly generated fluctuated between dysaerobic and anaerobic
during the late Caledonian (Acadian) and Varis- conditions (cf. Rhoads & Morse 1971, Savrda et
can deformation maxima. During the Silurian al. 1984, Thomson et al. 1985, Edwards 1985,
the remnant back-arc basin began to be modified Savrda & Bottjer 1986). The Llandovery was a
by a phase of sinistral transpression, associated time of high sea-level (Fig. 1) with high organic
with Avalonia-Laurentia collision which cul- productivities giving rise to a strong tendency to
minated in early to mid-Devonian time (Smith oxygen-depletion in the water column (Leggett
1987a, b, Soper et al. in press). Palaeomagnetic 1980). Figures 2 and 3 show the distribution of

DISSOLVED /
7--'";"/
_777t__. ........................
y ,.,oo~ l
,%mP~~ PHOSPHATECEMENTS r '

NOSKELETONISEDBENTHOS COMMUNITIESOFSKELETONISEDBENTHOS
Nereltes ICHNOFACIES Clorinda Costistrlcklandia I Pentameroide$ Eocoelia

FIG. 3. Biofacies distributions with water depth in the late Llandovery Welsh Basin during a period of
dysaerobic bottom-water conditions. The schematic dissolved oxygen profile has been drawn by analogy with
the present-day Santa Cruz Basin (Edwards 1985).
144 R. D. A. Smith
biofacies for the griestoniensis Zone sediments more rarely in isolated concretions. Sheet and
and their interpreted significance in terms of a concretion thicknesses range from about 1 to
lateral gradient in palaeo-oxygenation. 10 mm. Turbidites commonly have paler col-
oured top layers, which range in thickness from
1 mm to 2 cm and are usually bioturbated.
Morphology and occurrence Bioturbation in these top layers may be either in
the form of pervasive microbioturbation asso-
The phosphate cements, when not reworked, ciated with burrows up to 1 mm in diameter, or
occur close to the tops of fine-grained (silt-mud) consist of isolated Chondrites burrow systems
turbidites (Fig. 4), or occasionally in the basal with burrow diameters of about 0.1-0.3 mm.
coarse silt divisions. Turbidites containing phos- Phosphate sheets o~"layers of phosphatic concre-
phatic layers range in thickness from about 1.5 to tions frequently occur at, or a small distance
10cm. Bouma (1962) Tc~, Td_~ or simply Te above, colour changes, which may be very subtle.
divisions are represented. Many of the finer- However, they may also occur without any
grained turbidites are very similar to examples obvious associated colour change. The phos-
described by Stow & Shanmugam (1980). Inter- phate cements are black when fresh, but weather
turbidite hemipelagic facies are either poor in white. Turbidite colours range from olive-black
organic carbon and bioturbated or richer in (Munsell colour 5Y 2/1) to light olive-grey
organic carbon and finely laminated. The phos- (Munsell colour 5Y 5/2). Phosphatic layers often
phate cements occur in bedding-parallel thin show evidence of physical disruption by biotur-
sheets, layers of strongly oblate concretions, and bating organisms, and wet-sediment faulting and

HEMI-
PELAGITE , . . ......
5Y4/1 [[~j~IO..YY6/..22 ~5NI 5/2 N6
TURBIDITE
~:~ N2

~ N2

Y 3/2
~..~ OY 4 / 2 2/1

~
HEMI-
LAG,TE 5Y 2/1 N3
5Y 4/1 N2
N2

5Y 2/1

~m:ammma

SCALE KEY
-lcm
~ 0
Laminated siltstone

Laminated mudstone
~

~
Bioturbation

Phosphate concretion

Mudstone

FIG. 4. Representative silt-mud turbidite-hemipelagite couplets with phosphatic layers. Colour designations
use the Munsell system notation.
 Phosphate cements in a turbidite basin I45

FIG. 5. Evidence for early unconsolidated stages of phosphatic layer development (a--d) and for early
lithification (e-g). (a) Disruption of a phosphatic layer (dark) by bioturbation. Apatite crystallites have been
mixed into the underlying part of the enclosing turbidite. (b) Wet-sediment microfaulting affecting a
phosphatic layer. Note the contrast in intensity of bioturbation above and below the phosphatic layer. (c)
Plastic deformation of a phosphatic layer within a mudstone rip-up clast present near the top of a sandstone
turbidite. (d) Truncation of a phosphatic layer at the edge of a mudstone rip-up clast. (e) Curving of laminae
in a siltstone turbidite around a phosphatic concretion. (f) An undeformed ripped-up phosphate concretion
present in a matrix-rich sandstone turbidite. (g) Silicified cone-in-cone concretion with a phosphatic sheet
core. All scale bars represent 1 cm.
I46 R. D. A. Smith

C C
SCALE COLOUR O M
(cm) CO Q QCM MF M M C
HP "_~-~..__~_~ 5Y 4/1 ~ . ~
Z_%-~.--r- 5Y 5/2
3- ~.~..~
_.-y~-. ~-

2- 5Y 3/2

(a) " ~

Q IQ
i
SCALE COLOUR
(cm) Q
5Y 2 / 1 oc,?o? A c . . Fc M c M

5Y 4/1

N2 ~

.... A

5Y 2/1",~ A
1-

O-

(b) ~ 4'5 4'0 3'0 2'0 1'0 5

FIG. 6. Turbidite-hemipelagite mineralogies shown by XRD traces (Cu K, radiation). Colour designations use
the Munsell system notation. HP, hemipelagite; A, apatite; C, chlorite; F, feldspar; M, white mica;
Q, quartz. Note the increased abundances of chlorite beneath phosphatic layers.
 Phosphate cements in a turbidite basin I47
turbidites associated with laminated hemipela-
gites, in which it occurs in the form of framboidal
aggregates as well as euhedra. Framboidal pyrite
aggregates are occasionally intergrown with
chalcopyrite and arsenopyrite.
Phosphatic layers contain up to 15~ yellow-
luminescing apatite (Fig. 7) in the form of 0.5-
5 ~m crystallites with hexagonal sections (Figs 8a
and 9a). The apatite crystallites formed an inter-
particle porosity-filling cement (Fig. 7), but also
grew displacively in detrital phyllosilicates as
shown by the splitting apart of white mica
lamellae in early diagenetic chlorite mica stacks
(Fig. 9b). Dimberline (1986) argues that at least
some of these stacks result from early diagenetic
modification of detrital biotites. Detrital smec-
tite is another probable precursor phase (Craig et
al. 1982). The apatites have negligible carbonate
content (using the peak pair method of Gul-
brandsen 1970), probably the result of carbonate
loss from an original francolite (carbonate-
fluorapatite) composition during burial diagene-
sis and incipient metamorphism since the turbi-
dites are presently dominantly of lower to upper
anchizone grade (cf. McClellan 1980, McArthur
1985, Smith 1987b). No uranium enrichments in
phosphatic layers were detected using
autoradiography.
Below colour changes, near the tops of
FIG. 7. Cathodoluminescence image of a 5 mm thick turbidites, there is frequently a marked increase
phosphatic layer showing the distribution of
luminescent apatite cement. in small chlorite-dominated chlorite-mica inter-
growths of the order of 5-30 ~m in diameter
(Figs 6 and 9c). Such colour changes thus appear
to be dominantly controlled by chlorite abun-
dances. Pyrite also increases in abundance below
folding (sensu Pickering 1987), either in situ or in colour changes in the darker, more organic
rip-up clasts present within thicker turbidite carbon-rich turbidites. Below colour changes in
sandstones (Fig. 5a, b, c). Phosphate sheets often such turbidites, it is present as framboidal
form cores to later diagenetic cone-in-cone aggregates (up to 150 ~tm diameter) as well as
concretions (Fig. 5g). euhedra ranging from about 2 to 15 ~m diameter
Phosphate cements have not been recorded (Fig. 8d). Above colour changes pyrite occurs as
from Lower Silurian shelf facies of the Midland scattered euhedra 2-5 ~m in diameter.
Platform (Bridges 1 9 7 5 and author's The interfaces between bioturbated hemipela-
observations). gites and underlying turbidites are often highly
disrupted by bioturbation. Bioturbated hemipe-
lagites have a very similar mineralogical compo-
sition to the turbidites, but contain fewer
chlorite-mica stacks. The laminated hemipela-
Microfabrics and mineralogy gites, however, also contain dolomite rhombs (up
to 60 ~tm diameter) which are usually ferroan, but
Turbidite mineralogies, determined using XRD, sometimes preserve non-ferroan dolomite cores
EDXRA and EMP, are dominated by an (Fig. 8). In addition, they contain more abundant
assemblage of quartz, chlorite and white mica pyrite in the form of framboidal aggregates (up to
(often intergrown in chlorite-mica stacks ranging 60 ~tm diameter) and euhedra (2-40 ~tm diam-
up to 150 ~tm maximum diameter), feldspar and eter). Dolomite rhombs have sharp crystal
pyrite (Fig. 6). Accessories include titanium faces and often have pyrite and quartz inclu-
dioxides, florencite, xenotime and zircon. Chlor- sions. These features suggest that they probably
ites have iron-rich (ripidolite) compositions. formed within the sediment as authigenic
Pyrite is more abundant in darker coloured precipitates.
I48 R. D. A. Smith

FIG. 8. (a) BSEM image of microfabric and mineralogy of a phosphatic layer. (b) BSEM image of microfabric
and mineralogy of a laminated hemipelagite. A, apatite; C, chlorite; F, ferroan carbonate; M, white mica; P,
pyrite; Q, quartz; T, titanium dioxide. Scale bars represent 10 p.m.

Discussion (1) Evidence for disruption of phosphatic layers

by burrowing animals (Fig. 5a). This indi-
cates that phosphate enrichment initiated
Early unconsolidated stages of phosphatic layer very soon after turbidite deposition since
development bioturbation is limited to the top few
millimetres or centimetres and turbidite
An early unconsolidated stage in the develop- repeat times were probably short. An order
ment of the phosphatic layers is indicated by a of magnitude estimate for the mean repeat
number of features: time between turbidite deposition events of
 Phosphate cements in a turbidite basin I49

FIG. 9. BSEM images illustrating turbidite microfabrics and mineralogies. (a) Apatite crystallites in a siltstone
turbidite. Note pressure solution embayments in detrital quartz grains. Scale bar = 50 ~tm. (b) Apatite
crystallites which have grown in a detrital phyllosilicate grain. Note the splitting apart of white mica lamellae.
Scale bar = 10 p,m. (c) High contrast image showing bright apatite-rich band and a marked enrichment of
bright chlorites below this band. Scale bar = 500 ktm. (d) Framboidal pyrite aggregate present in an organic-
carbon rich turbidite beneath the phosphatic layer. Scale bar = 100 ~tm. (e) Clustering of apatite crystallites
around pyrite euhedra in a phosphatic layer. Scale bar = 50 ~tm. (f) Organic inclusions present in an unusually
large apatite crystallite. The surrounding quartz and mica has been blacked out using contrast and brightness
controls. Scale bar = 10 ~tm. A, apatite; C, chlorite; M, white mica; P, pyrite; Q, quartz.
x5o R. D. A. Smith

about 180 years has been calculated for the (Fig. 5f). In these cases the phosphate
most proximal preserved parts of the turbi- concretions had clearly become lithified
dite system using the following assumptions: prior to the enclosing turbidite mudstone.
(a) the 13 graptolite biozones of the Llando- (3) Phosphatic sheets commonly form the cores
very Series (Cocks et al. 1984) represent to later diagenetic concretions with cone-in-
equal periods of time; (b) The duration of the cone structure (Fig. 5g). These are at present
Llandovery Series was 10Ma (using the composed of microgranular quartz, but are
Harland et al. 1982 time-scale); (c) the mean interpreted as having been originally com-
number of turbidites present per 10 m of posed of calcite which grew displacively into
section is 105, based on a 45 m measured still unlithified enclosing mudstones (Wood-
section near Blaen Twrch, NE Dyfed land 1964, Marshall 1982). Marshall's stable
(SN67304893) and (d) The succession of isotopic data indicated that his Jurassic study
griestoniensis Zone age near Blaen Twrch is examples of cone-in-cone structure were
about 400m thick (Smith 1987b). Since formed at depths of tens to hundreds of
assumption (c) is based on all turbidites in metres in overpressured sediments. In the
the Blaen Twrch section, including thick case of the griestoniensis Zone sequence,
sandstone turbidites, it will tend to lead to an however, since ripped-up concretions with
overestimate of the repeat times of the thin- phosphate cores have been found in sand-
bedded fine-grained turbidites which con- stone turbidites, at least some of these
tain in situ phosphatic layers. concretions may have formed at shallower
(2) Phosphatic layers are affected by wet-sedi- depths. In one example, a 35 cm x 12cm
ment faulting (Fig. 5b). concretion with a core consisting of a
(3) Plastic deformation of phosphatic layers phosphatic sheet (5 mm thick) occurs within
occurred in some ripped-up clasts of unlithi- a coarse to medium-grained sandstone turbi-
fled turbidite mudstone (Fig. 5c). Plastic dite (40 cm thick) near Ty-Poeth, Powys
deformation of phosphate layers associated (SO 866 072). Turbidity currents depositing
with loading of overlying sandstone turbi- such sand turbidites excavated scours up to
dites has also been observed. 10 m or more deep in the channel-lobe
(4) Phosphatic layers may be truncated at the transition zone (Mutti & Normark 1987) of
edges of rip-up clasts (Fig. 5d). the griestoniensis Zone turbidite system
(Smith 1986, 1987a). Hence this area is a
Many authors have reported similar early
possible source for such ripped-up cone-in-
unconsolidated stages of development in recent
phosphatic sediments (e.g. Burnett 1977, 1980, cone concretions.
Baturin & Bezrukov 1979, Burnett et al. 1983,
O'Brien & Veeh 1980). The features described
above may record the presence of either an Localization of cements near turbidite tops
amorphous calcium phosphate precursor phase,
or the presence of numerous isolated apatite There is still considerable uncertainty about
crystallites which would provide nuclei for many details of the mechanisms of phosphate
further growth on subsequent supply of enrichment and precipitation in sediments.
phosphate. However, there is general agreement on some
points. The source of phosphate for marine
phosphate cements is widely thought to be a by-
Early lithification product of the microbial degradation of organic
matter (e.g. Berner 1974, 1980, Burnett, 1977,
Early lithification of the phosphatic layers,
Baturin & Bezrukov 1979, Baturin 1982, De
either through crystallization of an apatite
Lange 1986). In anoxic conditions porewaters
mineral from an amorphous precursor, or
become greatly enriched in phosphate relative to
through precipitation of more apatite crystal-
phosphate concentrations in sea water (e.g.
lites, is indicated by the following features:
Nissenbaum et al. 1972, Sholkovitz 1973). Phos-
(1) In cases of cement development within phate may become enriched in sediments
laminated turbidite siltstones laminae clearly through a variety of processes, including the
curve around concretions (Fig. 5e). This following: (a) adsorption on to iron and manga-
indicates that lithification took place prior to nese oxyhydroxides (e.g. Berner 1973, Krom &
signifcant sediment compaction. Berner 1981, Borgaard 1983, Watanabe &
(2) Rip-up clasts of isolated undeformed phos- Tsunogai 1984, De Lange 1986, Schaffer 1986);
phate concretions with thin mudstone ven- (b) adsorption on to illite-mica (Bremner 1980);
eers are common in thick turbidite sand- (c) replacement of carbonate grains by apatite
stones of the griestoniensis Zone succession (e.g. Manheim et al. 1975, de Kanel & Morse
 Phosphate cements in a t u r b i d i t e b a s & I5I
1978, Bentor 1980, Berner 1980, Pr6vot & LucS.s Ferric and manganese oxyhydroxides would
1986); precipitation of phosphate minerals from still be present in thin 'post-oxic' (Berner 1981) or
porewaters (e.g. Berner 1974, 1980, Bentor 1980, 'suboxic' (Froelich et al. 1979) layers present
Burnett 1977) and (d) in situ concentration of above levels of nitrate exhaustion. Hence, phos-
phosphate in bacterial cells (O'Brien & Veeh phate cements not associated with bioturbated
1982, Williams & Reimers 1983). Apatite preci- top layers may have been localized close to levels
pitation may be favoured by relatively low pH of nitrate exhaustion at times when bottom
(Nathan & Sass 1981), low magnesium concen- waters contained insufficient dissolved oxygen
trations in porewaters (Martens & Harris 1970), for downward burning oxidation fronts to deve-
and the presence of catalysts such as the enzyme lop. Since nitrate reduction accounts for very
carbonic anhydrase (McConnell et al. 1961). The small proportions of the total oxidation of
following discussion attempts to constrain the organic matter in marine sediments (e.g. Mfiller
major controls on phosphate concentration and & Mangini 1980, Reeburgh 1983, Jones 1985) no
apatite cement precipitation in the Lower Silur- strong colour change due to organic matter
ian turbidites of the Welsh Basin. depletion would occur across such a level. Subtle
The origin of thin paler-coloured bioturbated colour changes appear to be due to the enrich-
top layers of some of the griestoniensis Zone ment of small chlorite-mica intergrowths, togeth-
turbidites is attributed to a downward migrating er with pyrite, in the lower parts of turbidites
oxidation front process analogous to the pheno- (Figs 6 and 9c). Such enrichments suggest that
menon recently documented and inferred from ferrous iron was being immobilized at an early
late Pleistocene to Recent turbidites of the stage in the sulphidic diagenetic environment
Madeira Abyssal Plain (Colley et al. 1984, Colley (Berner 1981) present below the thin weakly oxic
& Thomson 1985, Wilson et al. 1985, 1986, Jarvis or post-oxic surface layer.
& Higgs 1987, Thomson et al. 1987). Wilson et al. In cases where turbidites with phosphate
(1985) suggest that the well-known vertical cements are overlain by unbioturbated laminat-
sequence of diagenetic zones present in 'steady ed hemipelagites oxic or post-oxic conditions at
state' hemipelagic regimes (e.g. Froelich et al. the sediment surface were clearly only short-
1979) may still be present beneath downward- lived since the absence of bioturbation in the
burning oxidation fronts active in turbidites, but laminated hemipelagites indicates that a sulphi-
in a highly compressed form. Their sample dic diagenetic environment extended up to the
spacing, however, was not close enough to be sediment surface during their accumulation (cf.
able to demonstrate this. Oxidation fronts in the Maynard 1982). The ferroan dolomite rhombs
Welsh Basin turbidites migrated far shorter present in the laminated hemipelagites possibly
distances (rarely greater than 1 cm) than the grew at the top of the methanogenesis zone, with
Madeira Abyssal Plain examples probably pri- their non-ferroan cores precipitating in the
marily due to the low levels of dissolved oxygen overlying sulphate reduction zone. Within the
in the bottom waters. In addition, short turbidite sulphate reduction zone iron would be preferen-
repeat times would prevent large distances of tially incorporated in sulphides (cf. Pye 1985,
downward migration. The short migration dis- Burns & Baker 1987). The presence of minute
tances account for the failure to observe in the Chondrites burrow systems in some turbidite top
Welsh examples the marked uranium enrich- layers beneath laminated hemipelagites indi-
ments recorded at relict oxidation fronts in the cates seafloor dissolved oxygen concentrations
Madeira Abyssal Plain turbidites. only slightly higher than critical levels required
One possible explanation for the localization by macrobenthos for survival (Bromley & Ekdale
of phosphate cements at, or a short distance 1984, Savrda & Bottjer 1986). Under such
above, the interface between the paler top layers conditions it is difficult to envisage sufficient
and the lower parts of turbidites is that phos- inorganic concentration of phosphate in the
phate enrichment occurred at these levels due to temporarily weakly oxic or post-oxic surface
the adsorption of dissolved phosphate, diffusing layers significantly to influence phosphate ce-
upwards along a concentration gradient from ment localization. Another possibly important
underlying anoxic porewaters, by iron and factor in localizing phosphate cements close to
manganese oxyhydroxides present above the turbidite tops in these cases is phosphate
relict oxidation fronts. The phosphate was enrichment associated with high microbial
probably a by-product of the microbial degrada- growth rates resulting from nutrient cycling
tion of organic matter in the underlying anoxic across the oxic-anoxic transition layer (e.g.
interstitial waters. The turbiditic sediments may nitrate-nitrite cycling:Anderson 1982, Anderson
have been already enriched in phosphate in et al. 1982). Williams & Reimers (1983) suggest-
shallow marine environments prior to resedi- ed that phosphate cement development in the
mentation in the basin (cf. Krom & Berner 1981). Miocene Monterey Formation of California may
I52 R. D. A. Smith

have been controlled by the presence of non- beneath succeeding turbidites and their remains
,photosynthetic filamentous bacteria.-Such bac- would become available for degradation leading
teria, which obtain energy by oxidizing reduced to regeneration of phosphate. Apatite might
sulphur species, have been recently reported precipitate either directly, if porewaters were
from sediments off Peru and Chile (Gallardo depleted in magnesium (Martens & Harris 1970),
1977) and California (Soutar & Crill 1977, or via the ageing of an amorphous calcium
Williams & Reimers 1983). These forms are phosphate precursor (cf. Jahnke et al. 1982,
adapted to microaerophilic conditions present in Gulbrandsen et al. 1984). Clustering of apatite
thin zones of transition between overlying oxic crystallites around authigenic pyrite euhedra
and underlying sulphidic environments, exactly occasionally seen in BSEM images (e.g. Fig. 9e)
analogous to those inferred to have existed close can be interpreted as the product of phosphate
to the tops of many of the Welsh turbidites. release on reduction of ferric oxyhydroxides in
On subsequent burial of turbidites with phos- the sulphate reduction zone with synchronous
phate-enriched top layers, ferric iron in the iron precipitation of pyrite. It is also possible that
oxyhydroxides would be reduced and the ad- enzymes such as carbonic anhydrase, a common
sorbed phosphate released. In addition, bacteria constituent of bacteria (Veitch & Blakenship
active at oxic-anoxic interfaces would die in the 1963), may have catalysed apatite precipitation
sulphidic environment created during burial since this enzyme is known to stimulate the

FIG. 10. Schematic model for possible processes involved in the localization (a) and precipitation (b) of early
diagenetic phosphate cements in Lower Silurian Welsh Basin turbidites. Diagenetic environments are
designated using the terminology of Berner (1981).
 Phosphate cements in a turbidite basin I53
precipitation of carbonate-hydroxyapatite soon after deposition of turbidites in dysaero-
(McConnell et al. 1961). O'Brien & Veeh (1982) bic bottom water environments of the Welsh
speculate that the operation of an in vivo Basin.
mechanism in bacteria may prevent mineraliza- (2) Phosphatic layers became lithified early,
tion of bacterially 'fixed' phosphate until cell prior to significant sediment compaction.
death when such an inhibitor would cease to be (3) The localization of phosphate cements close
active, enabling mineralization to occur. Pos- to the tops ofturbidites is thought to be due to
sible support for such a scenario in the Welsh early enrichment of phosphate at these
Basin cements comes from the presence of levels. Two possible mechanisms are: (a)
organic inclusions, sometimes seen to be elongat- adsorption of dissolved phosphate diffusing
ed (with approximate dimensions of 2 x 0.3 ~tm), upwards from underlying anoxic interstitial
in some of the larger apatite crystallites (e.g. Fig. water by ferric and manganese oxyhydrox-
9f). The absence of carbonates in turbidite ides, and (b) concentration by bacteria
mudstones suggests that the pH was low enough adapted to the oxic-sulphidic transition
to prevent their precipitation. A relatively low layer. The initiation of cement precipitation
pH increases the stability of apatites relative to may have been catalysed by enzymes present
that of calcium carbonates (Nathan & Sass in bacterial cells.
1981). Figure 10 diagrammatically summarizes (4) The basin-wide control on the distribution of
the inferred, probably dominant, controls on phosphate cements appears to have been
early diagenetic apatite cement development in primarily bottom water oxicity. This is
the griestoniensis Zone turbidites. because bottom water oxygen levels (in
addition to sedimentation rate) control: (a)
the flux of labile organic matter, the major
Controls on basin-wide distribution source of phosphate, through the sediment-
of phosphate cements water interface and (b) the presence or
absence of iron and manganese oxyhydrox-
In Upper Llandovery sequences in Wales and ides, and certain environmentally specific
the Welsh Borderland phosphate cements ap- bacteria in layers close to the sediment
pear to be restricted to turbiditic deposits of the surface. These limit the escape of dissolved
oxygen-deficient Welsh Basin. The apparent phosphate into bottom waters.
absence of phosphate cements in the shelf facies Similar phosphate cements might be expected
is simply explained as the result of oxidation of in other successions deposited in dysaerob~c
labile organic matter close to the sediment-water turbidite basins.
interface, preventing the burial of a phosphate
source. This would have been enhanced by lower ACKNOWLEDGMENTS:This work has greatly benefited
sedimentation rates on the shelf. from numerous discussions with the following people:
Phosphate cements are also absent from the Max Coleman, Sue Colley, Peter Cook, Charles Curtis,
overlying Wenlock turbidite sequence which was Tony Dickson, Andrew Dimberline, Harry Elderfield,
deposited in totally anoxic (euxinic) bottom Cathy Goss, Ian Jarvis, Anton Kearsley, Ken Pye and
waters (Dimberline & Woodcock 1987). This John Thomson. Dick Cave first drew my attention to
probably results from the diffusive loss of the presence of the phosphate cements. Peter Cook,
Andrew Dimberline, and Jim Marshall suggested
dissolved phosphate out into the bottom waters useful improvements to the manuscript. I would like to
(cf. Watanabe & Tsunogai 1984). thank Dave Newling and Anton Kearsley for their
assistance with the SEM work and Tony Abrahams for
assistance with the XRD work. Tony Dickson and
Conclusions Alison Searle helped with the CL work. Dave Martell
helped with the autoradiography work. I acknowledge
(1) Phosphate enrichment in thin layers close to a NERC research studentship. Cambridge Earth
the tops of fine-grained turbidites occurred Science Series No. 977.

References
ANDERSON,J. J. 1982. The nitrite-oxygen interface at --, OKUBO, A., ROBBINS, A. S. & RICHARDS, F. A.
the top of the oxygen minimum zone in the eastern 1982. A model for nitrite and nitrate redistribu-
tropical North Pacific. Deep-Sea Research, 29, tions in oceanic oxygen minimum zones. Deep-Sea
1193-201. Research, 29, 1113--40.
I54 R. D. A. Smith
BATURIN, G. N. 1982. Phosphorites on the Sea Floor. COCKS, L. R. M., WOODCOCK,N. H., RICKARDS,R. B.,
Developments in Sedimentology. Elsevier, TEMPLE, J. T. & LANE, P. D. 1984. The Llandovery
Amsterdam. Series of the type area. Bulletin of the British
- - & BEZRUKOV,P. L. 1979. Phosphorites on the sea Museum (Natural History), 38, 131-82.
floor and their origin. Marine Geology, 31, 317-32. COLLEY, S. & THOMSONS, J. 1985. Recurrent uranium
BENTOR, Y. K. 1980. Phosphorites--the unsolved relocations in distal turbidites emplaced in pelagic
problems. In: BENTOR, Y. K. (ed.). Marine conditions. Geochimica et Cosmochimica Acta, 49,
Phosphorites--Geochemistry, Occurrence, Genesis, 2339~48.
pp. 3-18. Society of Economic Paleontologists and , WILSON, T. R. S. & HIGGS, N. C. 1984.
Mineralogists Special Publication, 29. Post-depositional migration of elements during
BURNER,R. A. 1973. Phosphate removal from seawater diagenesis in brown clay and turbidite sequences
by adsorption on volcanogenic ferric oxides. Earth in the North East Atlantic. Geochimica et Cosmo-
and Planetary Science Letters, 18, 77-86. chimica Acta, 48, 1223-35.
- - 1974. Kinetic models for the early diagenesis of COOK, P. J. & SHERGOLD, J. H. 1986. Proterozoic and
nitrogen, sulfur, phosphorus and silicon in anoxic Cambrian phosphorites--nature and origin. In:
marine sediments. In: GOLDBERG,E. D. (ed.). The COOK, P. J. & SHERGOLD, J. H. (eds). Phosphate
Sea, 5, 427-50. Wiley, New York. Deposits of the World, Volume 1, pp. 369-86.
-- 1980. Early Diagenesis. A Theoretical Approach. Cambridge University Press.
Princeton Series in Geochemistry. Princeton CRAIG, J., FITCHES, W. R. & MALTMAN, A. J.
University Press. 1982. Chlorite-mica stacks in low-strain rocks
--1981. A new geochemical classification of sedi- from central Wales. Geological Magazine, 119,
mentary environments. Journal of Sedimentary 243-56.
Petrology, 51, 359-65. DE KANEL, J. & MORSE, J. W. 1978. The chemistry of
BORGAARD, O. K. 1983. Effect of surface area and orthophosphate uptake from seawater onto calcite
mineralogy of iron oxides on their surface charge and aragonite. Geochimica et Cosmochimica Acta,
and anion-adsorption properties. Clays and Clay 42, 1335-40.
Minerals, 31, 230-32. DE LANGE, G. J. 1986. Early diagenetic reactions in
BOUMA, A. H. 1962. The Sedimentology of some Flysch interbedded pelagic and turbiditic sediments in
Deposits. Elsevier, Amsterdam. the Nares Abyssal Plain (western north Atlantic):
BREMNER,J. M. 1980. Concretionary phosphorite from Consequences for the composition of sediment
SW Africa. Journal of the Geological Society of and interstitial water. Geochimica et Cosmochimica
London, 137, 773-86. Acta, 50, 2543-61.
BRIDGES, P. E. 1975. The transgression of a hard DIMBERLINE, A. J. 1986. Electron microscope and
substance shelf: the Llandovery (Lower Silurian) microprobe analysis of chlorite-mica stacks in the
of the Welsh Borderland. Journal of Sedimentary Wenlock turbidites, mid Wales, U.K. Geological
Petrology, 45, 79-94. Magazine, 123, 299-306.
BRINDLEY, G. W. & BROWN, G. (eds) 1980. Crystal - - & WOODCOCK,N. H. 1987. The southeast margin
Structures of Clay Minerals and their X-ray Identifi- of the Wenlock turbidite system. Geological
cation. The Mineralogical Society, London. Journal, 22, 61-71.
BROMLEY, R. G. & EKDALE,A. A. 1984. Chondrites: a EDWARDS, B. D. 1985. Bioturbation in a dysaerobic,
trace fossil indicator of anoxia in sediments. bathyal basin: California Borderland. In: CUR-
Science, 224, 874-972. RAN, H. A. (ed.). Biogenic Structures: their Use in
BURNETr, W. C. 1977. Geochemistry and origin of Interpreting Depositional Environments, pp. 309-31.
phosphorite deposits from off Peru and Chile. Society of Economic Paleontologists and Minera-
Bulletin of the Geological Society of America, 88, logists Special Publication, 35.
813-23. FAIRCHILD,I. J. 1980. Stages in Precambrian dolomiti-
--1980. Apatite-glauconite associations off Peru zation, Scotland: cementing versus replacement
and Chile: palaeo-oceanographic implications. textures. Sedimentology, 27, 631-50.
Journal of the Geological Society of London, 137, FITCHES, W. R. & CAMPBELL,S. D. G. 1987. Tectonic
757-64. evolution of the Bala Lineament in the Welsh
--, ROE, K. K. & PIPER, D. Z. 1983. Upwelling and Basin. Geological Journal, 22, 131-53.
phosphorite formation in the ocean. In: SUESS, E. FROELICH, P. N . , KLINKHAMMER, G. P., BENDER,
& THIEDE, J. (eds). Coastal Upwelling and its G. R., LUEDTKE, N. A., HEATH, G. R. CULLEN,
Sediment Record, Part A, pp. 377-96. Plenum D., DAUPHIN, P., HAMMOND,D. & HARTMAN,B.
Press, New York. 1979. Early oxidation of organic matter in pelagic
BURNS,S. J. & BAKER,P. A. 1987. A geochemical study sediments of the eastern equatorial Atlantic:
of dolomite in the Monterey Formation, Califor- suboxic diagenesis. Geochimica et Cosmochimica
nia. Journal of Sedimentary Petrology, 57, 128- Acta, 43, 1075-90.
39. GALLARDO, V. A. 1977. Large benthic microbial
CAVE, R. 1979. Sedimentary environments of the communities in sulfide biota under Peru-Chile
basinal Llandovery of mid-Wales. In: HARRIS, subsurface counter current. Nature, 268, 331--2.
A. L., HOLLAND,C. H. & LEAKE, B. E. (eds). The GULBRANDSEN,R. A. 1970. Relation of carbon dioxide
Caledonides of the British Isles--reviewed, pp. 517- content of apatite of the Phosphoria Formation to
26. Special Publication of the Geological Society regional facies. United States Geological Survey
of London No. 8. Professional Paper, 700B, 9-13.
 Phosphate cements & a turbidite basin I55
--, ROBERSON,C. E. & NEIL, S. T. 1984. Time and 1961. Relation between inorganic chemistry and
crystallisation of apatite in sea water. Geochimica biochemistry of bone mineralization, Science, 133,
et Cosmochimica Acta, 48, 213-8. 281-2.
HARLAND, W. B., Cox, A. V., LLEWELYN, P. G., MCKERROW, W. S. 1979. Ordovician and Silurian
PICKTON, C. A. G., SMITH, A. G. & WALTERS, R. changes in sea level. Journal of the Geological
1982. A Geological Time Scale. Cambridge Univer- Society of London, 136, 137-45.
sity Press. Mi2LLER, P. J. & MANGINI, A. 1980. Organic carbon
JARVIS, I. & HIGGS, N. 1987. Trace element mobility decomposition rates in sediments of the Pacific
during early diagenesis in distal turbidites: Late Manganese Nodule Belt dated by Z3~ and
Quarternary of the Madeira Abyssal Plain, N 231Pa. Earth and Planetary Science Letters, 51, 94-
Atlantic. In: WEAVER, P. P. E. & THOMSON, J. ll4.
(eds). Geology and Geochemistry of Abyssal Plains, MULLINS, H. T. & RASCH, R. F. 1985. Sea-floor
pp. 179-213. Geological Society Special Publica- phosphorites along the Central California conti-
tion, 31. Blackwell Scientific Publications, nental margin. Economic Geology, 80, 696-715.
Oxford. MuTrI, E. & NORMARK, W. R. 1987. Comparing
JAHNKE, R. A., EMERSON, S. R., ROE, K. K. & examples of modern and ancient turbidite sys-
BURNETT, W. C. 1982. The present day formation tems: problems and concepts. In: LEGGETT, J. K.
of apatite in Mexican continental margin sedi- & ZUFFA, G. G. (eds). Deep Marine Sedimento-
ments. Geochimica et Cosmochimica Acta, 47, 259- logy: Models and Case Histories. Graham &
66. Trotman, London.
JONES, J. G. 1985. Microbes and microbial processes in NATHAN, Y. & SANS, E. 1981. Stability relations of
sediments. Philosophical Transactions of the Royal apatites and calcium carbonates. Chemical Geo-
Society of London, A315, 3-17. logy, 34, 103-11.
KOKELAAR, B. P., HOWELLS, M. F., BEVlNS, R. E., NISSENBAUM,A., PRESLEY,B. J. & KAPLAN, I. R. 1972.
ROACH, R. A. & DUNKLEY, P. N. 1984. The Early diagenesis in a reducing fjord, Saanich Inlet,
Ordovician marginal basin of Wales. In: KOKE- British Columbia, I, Chemical and isotopic
LAAR, B. P. & HOWELLS, M. F. (eds). Marginal changes in major components of interstitial
Basin Geology, pp. 245-71. Special Publication of water. Geochimica et Cosmochimica Acta, 36,
the Geological Society of London, 16. Blackwell 1007-27.
Scientific Publications, Oxford. O'BRIEN, G. W. & VEEH, H. H. 1980. Holocene
KROM, M. D. & BERNER, R. A. 1981. The diagenesis phosphorite on the East Australian continental
of phosphorus in a nearshore marine margin. Nature, 288, 690-2.
sediment. Geochimica et Cosmochimica Acta, 45, - - & -- 1982. Are phosphorites reliable indicators
207-16. of upwelling? In: SUESS, E. & THIEDE, J. (eds).
LEGGET, J. K. 1980. British Lower Palaeozoic black Coastal Upwelling, its Sediment record, pp. 399-
shales and their palaeo-oceanographic signifi- 419. Plenum Press, New York.
cance. Journal of the Geological Society of Lon- PICKERING, K. T. 1987. Wet-sediment deformation in
don, 137, 139-56. the Upper Ordovician Point Leamington Forma-
--, MCKERROW, W. S. & SOPER, N. J. 1983. A tion: an active thrust-imbricate system during
model for the crustal evolution of Southern sedimentation, Notre Dame Bay, north-central
Scotland. Tectonics, 2, 187-210. Newfoundland. In: JONES, M. E. & PRESTON,
MANI-IEIM, F., ROWE, G. T. & JIPA, D. 1975. Marine R. M. F. (eds). Deformation of Sediments and
phosphorite formation off Peru. Journal of Sedi- Sedimentary Rocks, pp. 213-39. Geological So-
mentary Petrology, 45, 243-51. ciety Special Publication, 29. Blackwell Scientific
MARSHALL, J. O. 1982. Isotopic composition of Publications, Oxford.
displacive fibrous calcite veins: reversals PIPER, J. D. A. 1985. Palaeomagnetism in the
in porewater composition during burial dia- Caledonian-Appalachian orogen: a review. In:
genesis. Journal of Sedimentary Petrology, 5 2 , GEE, D. G. & STURT, B. A. (eds). The Caledonide
615-30. Orogen ."Scandinavia and Related Areas, pp. 35-49.
MARTENS, C. S. & HARRIS, R. C. 1970. Inhibition of Wiley, London.
apatite precipitation in the environment by PR~VOT, L. & LucLs, J. 1986. Microstructure of apatite
magnesium ions. Geochimica et Cosmochimica replacing carbonate in synthesised and natural
Acta, 34, 621-5. examples. Journal of Sedimentary Petrology, 56,
MAYNARD, J. B. 1982. Extension of Berner's 'New 153-9.
Geochemical classification of sedimentary envir- PYE, K. 1985. Electron microscope analysis of zoned
onments' to ancient sediments. Journal of Sedi- dolomite rhombs in the Jet Rock Formation
mentary Petrology, 52, 1325-31. (Lower Toarcian) of the Whitby area, U.K.
MCARTHUR, J. M. 1985. Francolite geochemistry-- Geological Magazine, 122, 279-86.
compositional controls during formation, diagene- - - & KRINSLEY, D. H. 1984. Petrographic examin-
sis, metamorphism and weathering. Geochimica et ation of sedimentary rocks in the SEM using
Cosmochimica Acta, 49, 23-35. backscattered electron detectors. Journal of Sedi-
MCCLELLAN, G. H. 1980. Mineralogy of carbonate- mentary Petrology, 54, 877-88.
fluorapatite. Journal of the Geological Society of REEBURGH, W. S. 1983. Rates of biogeochemical
London, 137, 675-81. processes in anoxic sediments. Annual Reviews of
MCCONNELL, D., FRAJOLA, W. J. & DEANER, D. W. Earth and Planetary Sciences, 11, 269-98.
I56 R. D. A. Smith

RHOADS, D. C. & MORSE, J. W. 1971. Evolutionary and structures in fine-grained turbidites; comparison
ecologic significance of oxygen-deficient marine of recent deep sea and ancient flysch sediments.
basins. Lethaia, 4, 413-28. Sedimentary Geology, 25, 23-42.
SAVRDA, C. E. & BOTTJER, D. J. 1986. Trace fossil THOMSON, J., COLLEY, S. HIGGS, N. C., HYDES, D. J.,
model for the reconstruction of palaeo-oxygen- WILSON, T. R. S. • SORENSEN,J. 1987. Geochemi-
ation in bottom waters. Geology, 14, 3-6. cal oxidation fronts in NE Atlantic distal turbi-
- - , & GORSLINE, D. S. 1984. Development of a dites and their effects in the sedimentary record.
comprehensive oxygen-deficient marine biofacies In: WEAVER,P. P. E. & THOMSON,J. (eds). Geology
model: Evidence from Santa Monica, San Pedro, and Geochemistry of Abyssal Plains, pp. 167-77.
and Santa Barbara Basins, California continental Special Publication of the Geological Society of
borderland. American Association of Petroleum London, 31. Blackwell Scientific Publications,
Geologists Bulletin, 68, 1179-92. Oxford.
SCBAFFER, G. 1986. Phosphate pumps and shuttles in THOMSON, J. B., MULLINS, H. T., NEWTON, C. R. &
the Black Sea. Nature, 321, 515-7. VERCOU'rERE, T. L. 1985. Alternative biofacies
SHOLKOVlTZ,E. R. 1973. Interstitial water chemistry of model for dysaerobic communities. Lethaia, 18,
the Santa Barbara Basin sediments. Geochimica et 167-79.
Cosmochimica Acta, 37, 2043-73. VEITCH, F. P. & BLAKENSHIP,L. C. 1963. Carbonic
SMITH, R. D. A. 1986. Giant erosional structures in the anhydrase in bacteria. Nature, 197, 76-7.
channel-lobe transition zone of a Welsh Basin WATANABE, Y. & TSUNOGAI, S. 1984. Adsorption-
turbidite system. Abstracts of the annual meeting of desorption control of phosphate in anoxic sedi-
the British Sedimentological Research Group, ment of a coastal sea, Funka Bay, Japan. Marine
Nottingham. Chemistry, 15, 71-83.
- - 1 9 8 7 a . The griestoniensis zone turbidite system, WILLIAMS,L. A. & REIMERS,C. 1983. Role of bacterial
Welsh Basin. In: LEGGETT, J. K. & ZUFFA, G. G. mats in oxygen-deficient marine basins and
(eds). Deep Marine Clastic Sedimentology : Models coastal upwelling regimes: preliminary report.
and Case Histories, pp. 89-107. Graham & Geology, 11, 267-9.
Trotman, London. WILSON, T. R. S., THOMSON,J., COLLEY,S., HYDES,D.
- - 1987b. Structure and deformation history of the & HIGGS, N. C. 1985. Early organic diagenesis:
Central Wales Synclinorium, NE Dyfed: evidence the significance of progressive subsurface oxida-
for a long-lived basement structure. Geological tion fronts in pelagic sediments. Geochimica et
Journal, 22, 183-98. Cosmochimica Acta, 49, 811-22.
- - 1 9 8 7 c . Palaeo-oxygenation in a turbidite basin: - - , HYDES, D. J., COLLEY, S., COLKIN, F. &
reconstruction of lateral gradients and variations SORENSEN, J. 1986. Oxidation fronts in pelagic
through time using trace fossil evidence. Abstracts sediments: diagenetic formation of metal-rich
of the 8th European Meeting of the International layers. Science, 232, 972-5.
Association of Sedimentologists, Tunis. WOOD, E. M. R. 1906. The Tarannon Series of
SOPER, N. J., WEBB, B. C. & WOODCOCK,N. H. 1987. Tarannon. Quarterly Journal of the Geological
Late Caledonian (Acadian) transpression in Society of London, 62, 644-701.
North West England: timing, geometry and WOODCOCK, N. H. 1984. Early Palaeozoic sedimenta-
geotectonic significance. Journal of the Geological tion and tectonics in Wales. Proceedings of the
Society of London, in press. Geologists' Association of London, 95, 323-35.
SOUTAR, A. & CRILL, P. A. 1977. Sedimentation and WOODLAND,B. G. 1964. The nature and origin of cone-
climatic patterns in the Santa Barbara basin in-cone structure. Fieldiana: Geology, 13, 185-305.
during the 19th and 20th centuries. Bulletin of the ZIEGLER, A. M., COCKS,L. R. M. & MCKERROW, W. S.
Geological Society of America, 88, 1161-72. 1968. The Llandovery transgression of the Welsh
STOW, D. A. V. & SHANMUGAM,G. 1980. Sequence of Borderland. Palaeontology, 11, 736-82.

R. D. A. SMITH, Department of Earth Sciences, University of Cambridge, Downing Street,

Cambridge CB2 3EQ, UK.
 Mixing zone dolomitization of Devonian carbonates, Guangxi,
South China
Zhao Xun & Ian J. Fairchild

SUMMARY: Regional facies relationships indicate that dolomitization around the

Lower-Middle Devonian boundary in Guangxi region is largely restricted to platform
margin facies. In the Liujin area dolostones are represented by the Tangding and succeeding
Yingtang Formations. The latter comprises alternations of laminated fine-grained
dolostones (type A dolomite) and massive, more coarsely-crystalline dolostones (type B
dolomite). Type B is richer in marine fossils and other aUochems, and increases in
importance upwards. The Formation is bounded above by a disconformity marked by a
haematitic marl, interpreted as a subaerial erosion surface, and is succeeded by limestones
lacking dolomite.
Petrographic observations indicate that the two dolomite types represent the dolomitiza-
tion of originally different sedimentary facies by a single dolomitization process. The
dolomite is dominantly replacive, but also partly cementing. Variably negative ~13C and
6lsO values (interpreted as original) and covariation of these parameters with Sr and Na
points to dolomitization by mixed meteoric-marine fluids. Stratigraphic variation in
geochemical data is inconsistent with a single episode of subaerial exposure and therefore
points to repeated emergence.
Dominantly replacive dolostones can form by the mixing-zone mechanism given
sufficiently stable conditions.

Land (1980) stated that the essence of the evidence, particularly a sympathetic variation of
'dolomite problem' was the dolomitization of carbon and oxygen isotopes with values some-
platform carbonates. Of these, dolostones that what lighter than expected for marine waters.
contain a normal marine fauna are particularly Machel & Mountjoy's (1986) arguments are re-
intriguing. Following the documentation of examined in the light of this evidence.
Holocene and Pleistocene dolomite formation in The strata concerned are located in the
meteoric-marine mixing zones (e.g. Land 1973a, Guangxi region, South China (Fig. 1) where
b), this setting has been used by many authors to platform carbonates largely of Middle to Upper
explain examples of ancient platform dolomite. Devonian age succeed Lower Devonian non-
Recently, Machel & Mountjoy (1986) have marine siliciclastic deposits. Much research
vigorously disputed the effectiveness of mixing- effort has been expended on these rocks over
zone dolomitization and pointed out the equiv- several decades because of their associated
ocal origin of many alleged examples. They mineralization and their petroleum potential,
contrast the massive, replacive nature of many although this information is largely in unpub-
ancient occurrences of platform dolostones with lished reports of the Guangxi Geological Bureau
the abundance of dolomite cement in undisputed (but see also Hou 1978, Liao et al. 1979, Bai et al.
mixing-zone dolomitic rocks (e.g. Kaldi & 1982, pp. 22-6, Kuang 1982, Dineley 1984). This
Gidman 1982, Ward & Halley 1985). Indeed, article synthesizes information produced by the
where geochemical evidence has been available Guangxi Geological Bureau in collaboration
for alleged ancient mixing-zone dolostones, it has with the Chinese Academy of Sciences concern-
often been at best consistent with the proposed ing field relationships and palaeontology, and
model rather than strongly supporting it (e.g. the introduces new laboratory data obtained in the
rather heavy carbon and oxygen isotopes in a UK.
Carboniferous example described by Choquette
& Steinen 1980). In contrast, the majority of
ancient platform dolostones have rather negative
oxygen isotope values, and Land (1980) conclud- Methods
ed that crystallization or recrystallization of
dolomite in contact with meteoric and/or hot Having observed various samples in stained thin
waters must have occurred. In this paper we section, selected samples were studied by catho-
describe an example of a dolostone unit where doluminescence (CL) in polished section, X-ray
the field data suggesting dolomitization associat- diffraction (XRD), particularly for characteriz-
ed with emergence are backed by geochemical ing dolomite stoichiometry; the geochemistry of

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, I57

Geological Society Special Publication No. 36, pp. 157-170.
x58 Zhao Xun & I. J. Fairchild

the acid-soluble fraction was analysed by induc- analysed along with a laboratory standard
tively-coupled plasma emission spectrometry dolomite on a VG 903 mass spectrometer. A
(ICP); most of these samples were also analysed phosphoric acid fractionation factor of 1.01109
for carbon and oxygen isotopes and some were (Sharma & Clayton 1965) was used in calculating
studied in the Scanning Electron Microscope results, which are expressed with respect to the
(SEM), including a little quantitative analysis. PDB standard and are reproducible to 0.05%0.
Samples for XRD and ICP were drilled from
specimens and ground to 10-50 pm. XRD was
carried out at 40 kV and 20 mA using Cu K~ Field relationships
radiation. ICP analysis was carried out on solutes
from cold 10% HCI dissolution of samples at Studies of numerous cross-sections of Devonian
dilutions of 1:200 (50 mg sample in 10 ml final carbonates in the Guangxi region indicate that
solution) and using matrix-matched standards. several types of dolomite are present. One type
Accuracy and precision are only around 10 occurs beneath exposure surfaces and so a
relative % because of initial problems in estab- mixing-zone origin was proposed. The dolomi-
lishing standardization procedures. A larger tized strata are either reefs or bioclastic bank
(0.1-3 g) portion of each powder was also accumulations that have topographic relief,
dissolved in HCI for determination of insoluble making them more likely to have suffered
residue (IR) content. Comparison of XRD and subaerial exposure than surrounding strata.
ICP results indicate negligible calcite impurity in Three examples are given below.
the dolomite. First, in the Beishan section and associated
For oxygen and carbon stable isotope analysis, with a base-metal deposit is a stromatoporoid-
10 mg of each sample powder were reacted with algal reef within the Middle to Upper Devonian
100% phosphoric acid at 25~ for 72 hr and Dongganglin and Guilin Formations. The reef is

105'E / .. ' / ~- '"

u I /' ~.Hunan,.~.),r~
(
9\
Guelchow
.~
/z-v
.
I " %).
.
I f-~-.~"; .1
"\ ./ I. f * ' . . , l ~ " Gudang oGuilln./

( "-.....; ~e,~on ] ~' L---, Xper;t;dal zone ! .--..~

'~. ~.h Guangx~ J'z/h. . . . . . 1 f

YQnnan I " -~~ta~gj'~"~ / Guang~: ((~ i
.J~'Tongwa n - 0 ?' \ f'
carbonate platform \)/ ~._./)
v ~ ) N~Changtan~ I . /.1
L ,ujin /JLand)

VIETNAM ~C_H INA ,a h i"fOuangdon9 . ~ // .,.-/

I ~ ~X IW~UHqgcho u / L ~ T. . . . . tron ~../" /)) / / r , J G u a o g d o n g
zone/~ ~ //// .~
/ D e e p sea // ." / _. :"

~ South China
//,,/

Sea
.~..f ~

Nanning Wuhe Changtong L,ujin Litang Ton, wan Ertang Zhongpin Srpoi

lI:
i I I I J I __1 ~1 I
. ~
-- " I 9 I ' I I I I 500

~r T!ii

sandy shale limestone I~ ] ~ i limestone

FIG. 1. (a) Geography of the Guangxi region including location of line of section in (c). (b) Palaeogeography
of the Guangxi region around the boundary of Lower and Middle Devonian times. (c) Reconstruction of
lithofacies relationships of Lower and Middle Devonian shelf sediments in the Guangxi region. The Liujin
section provided the specimens used in the present study. An approximate time-line corresponding to the top
of the Yingtang Formation is shown as a heavy line.
 Dolomitization of Devon&n carbonates, South China [ 59

100 m thick of which the upper 50 m is dolomi- tions straddle the Lower-Middle Devonian
tized completely and the lower half partially. boundary. Palaeontological evidence indicates
Second, in the Gudang section, the 270 m that the Tangding Formation corresponds with
thick Dongganglin Formation is dolomitic the Upper Zlickovian stage of the European
throughout, and contains 40-75% transported Lower Devonian. The Yingtang Formation, for
bioclasts (stromatoporoids, echinoderms, ostra- which laboratory data are given in this paper,
codes, calcispheres, corals and brachiopods) and contains fossils indicative of the Dalejian to
some in situ stromatoporoids and crinoids in a Eifelian stages of the Middle Devonian, and the
lime-mud matrix. It passes laterally shorewards succeeding Dongganglin Formation is of Give-
into laminated dolostones with few fossils (calci- tian age (Dineley 1984).
spheres and tentaculitids). The top is marked by Figure 1(c) summarizes the stratigraphic rela-
a scoured surface and conglomerate with iron tionships. The dolostones at Liujin were located
oxide-stained pebbles. Pores in the dolomite at the margin of a carbonate platform (Fig. lb)
contain bitumen and there is much pyrite and separating restricted shelf carbonate facies
base-metal mineralization which also appears to (eventually passing landwards into terrigenous
post-date dolomitization. clastics) from slope-related cherty limestones.
The third example is from the regional The Tangding Formation is mostly composed
standard cross-section at Liujin where dolo- of dark grey fine-grained dolostones with milli-
stones of the Tangding and Yingtang Forma- metre-scale lamination, birdseye structure and

AVERAGE
CRYSTAL SIZE/Jm Sr Concentration ppm

Dongganglin
S0 200 500
(~gI"''''' 50 60 70 80 90
.... 100 110 120

EM.MS
M1 y // ' A
~L imestone
/, /
Nj2g
~'11 +In ~ T y p e B dolomite
i /,
I /
/
i ~ Type A dolomite
"1/1/
Nj 28 - - f I [~--] Intractast
/ /
Yingtang ~ Bioclast
EM.
Nj 2 4
Nil8 ~
{+) [ - ~ Pellet
k9 §

Nj9 [] +
~] Crinoid
@/
Nj7 ~ z /
[] + ~ - ~ Stromatoporoid
i

7~--~ Tentaculiles

Nil
/ /
< 9
l'J -1.1-I ~ Brachiopod
/ _Z /
/ / ,
i I / /
--~ Coral
1oI ~ Gastropod
Sample location

Tangding
EM.
/ i i i i i
.i i i i i ] i .J-J

FIG. 2. Stratigraphy of the Yingtang Formation in the Liujin cross-section together with the stratigraphic
distribution of Sr (key as in Table 1).
I6o Zhao Xun & I. J. Fairchild

local silicification including length-slow chalce- and southern Gueichow, Bai et al. 1982) where a
dony and dolomitized calcarenites. The sedi- prominent disconformity is found.
ments contain a fauna dominated by nektonic/ The conclusion drawn from the field data was
planktonic tentaculitids, ostracodes and ammon- that the type B dolostones were formed by
oids, with the addition of echinoderms and mixing-zone dolomitization associated with ex-
gastropods higher up. posure at the top of the Yingtang Formation. The
The Yingtang Formation (Fig 2) likewise type A dolostones were thought to have been
contains some laminated dolostones, here re- dolomitized prior to emergence. The laboratory
ferred to as type A dolomite. These dolostones data discussed below point to a mixing-zone
contain fossil moulds and show synsedimentary origin for both types of dolomite, but with
brecciation in places. However, increasingly repeated exposure during deposition of the
upwards, the sediments consist of more massive, Yingtang Formation in addition to the final
porous and coarsely-crystalline dolostones, re- episode of emergence.
ferred to as type B dolomite. Although the fabric-
destructive nature of the dolomitization makes
recognition of primary characteristics difficult, Petrography
bioclasts can be recognized: dominantly crinoids
and stromatoporoids, with subordinate bryo- The type A dolostones are completely dolomi-
zoans, gastropods, brachiopods and corals. tized carbonate mudstones or wackestones, with
There are also intraclasts and peloids (possibly some local dolomite-cemented fossil moulds
originally ooids). Some sinusoidal lamination
and normal and reverse grading suggestive of
intense wave or current action are recognizable
in places. Overall, the Yingtang Formation
appears to represent a bank or shoal with
intercalations of finer sediments representing
more sheltered environments. It is notable that
base-metal deposits co-exist with the dolomite,
especially on the upper part of the onshore side of
the bank.
The top of the Yingtang Formation is marked
by a discontinuity surface with a haematitic marl
layer (Kuang 1982) followed upwards by a
conglomerate of carbonate pebbles and then
dolomite-free limestones with a diverse open-
marine fauna (Fig. 2). This time-horizon was also
marked by uplift in the continental sediments in
the northern basin margin (northern Guangxi

FIG. 4. Sample Nj-1 seen in plane polarized light (a)

and under CL (b). Scale bar 100 jam. Dark area,
labelled A in (a) is type A dolomite seen in (b) to
have identical luminescence characteristics to the
earlier growth stages (labelled B) of the zoned type B
FIG. 3. Typical inclusion-rich type A dolomite as dolomite. The latter terminates against a vug at the
seen in plane polarized light (sample UA). Moulds of top of the photographs (ragged edges may be an
probable tentaculitids are present. Scale bar 100 jam. artefact).
 Dolomitization of Devonian carbonates, South China I6I
(Fig. 3) and lamination. The matrix has crystal with type A in some samples, indicating that the
sizes typically of a few tens of microns, and two dolomite types represent original sedimen-
contains a high density of inclusions of fluid, tary layers of different composition. Fossil-rich
calcite, iron ?oxide, pyrite, clay (identified by layers (presumed original grainstones or pack-
qualitative analysis on SEM) and very finely- stones) are always preserved as type B dolomite
disseminated organic carbon. Such dolomite also which is fabric-destructive (Fig. 5b). Conversely,
makes up the matrix of a sedimentary breccia in type A dolomite seems to represent replacement
one sample. of micritic or dolomicritic layers rich in organic
The type B dolostones (Figs 4 to 6) are carbon.
composed of more coarsely-crystalline dolomite, A combination of observations in transmitted
zoned by density of inclusions (mostly of calcite light and by cathodoluminescence reveals the
and fluid, Fig. 5a). Type B is also interlaminated growth history of the dolomite. In a given sample

FIG. 5. Type B dolomite. (a) Inclusion-rich dolomite

succeeded by inclusion-free growth (sample Nj-1)
viewed by backscattered electron imagery. Inclusions
are of calcite (pale) and of fluid (seen as holes). Scale
bar 10 p,m. (b)Typical fabric-destructive dolomite.
Portion of crinoid stem recognizable since it is
preserved as a few large dolomite crystals (sample
M1). Scale bar 100 ~tm.
162 Zhao Xun & I. J. Fairchild

type A has the same luminescence characteristics CL zonation is invariably shown by the
as the earlier growth stages of type B dolomite dolomites, although commonly only in the later
(e.g. Fig. 4). Typically, CL is very dull red, stages of growth. Nj-28 and Nj-29 show an
becoming brighter near crystal boundaries, al- apparently identical zonation of later growth
though a more brightly-luminescing core is stages, but other samples differ from each other
shown in specimens Nj-24 and M1 (see Fig. 2 for in varying degrees and show no correlation if
stratigraphic location). Half to 95% of each inclusion density as well as luminescence intensi-
sample is represented by early growth stages of ty are considered (e.g. compare Figs 4 and 6).
dolomite (types A, B or both) as just described. Nj-29 displays a haematitic rind (?with clay)
Later stages of growth of type B dolomite on the margins of dolomite crystals which grew
generally show a reduction in the proportions of into former vugs (Fig. 6) and the final event in
inclusions, locally with a sharply-defined and this and samples Nj-7 and Nj-24 was a calcite
narrow inclusion-free zone followed sharply by filling of the vugs.
more inclusion-bearing dolomite and then finish- Sample M5, just below the disconformity at
ing with inclusion-free growth. Since some of the the top of the Formation, displays small intersti-
inclusions are calcite crystals (Fig. 5a) far larger tial areas of undolomitized limestone, as well as
in size than calcites known to have formed by later vug-fills of calcite.
exsolution of dolomite (Wenk & Zhang 1985), it In summary, the dolomite of the Yingtang
can be deduced that the calcite inclusions are Formation is dominantly replacive. In a given
relics of the original sediment, now replaced. At sample dolomite crystals started to grow simulta-
some stage the inclusion-free growth became neously in both carbonate mudstones and wacke-
cementing rather than replacive since the dolo- stones rich in organic carbon and in other
mite crystals border cavities in several samples. lithologies. Crystal nuclei were more closely
spaced in the carbonate mudstones and wacke-
stones so that growth was concluded earlier to
yield type A dolomite with its relatively fine
crystal size. In other lithologies type B dolomite
continued to grow and eventually became
cementing. Although some vugs are primary
cavities in fossils, most are not obviously related
to the primary rock fabric and so are probably
secondary. In general the zonation of the crystals
is not identical in different samples, therefore
growth of the dolomite was not necessarily
synchronous throughout the Formation. Follow-
ing dolomitization, haematite rinds and calcite
cements formed locally.

Geochemistry
Table 1 shows the data and Figs 7 to 10 illustrate
the most important geochemical relationships.
Subsamples of types A and B are distinguished
on the plots. In some cases it was possible to
subsample areas of dolomite dominated by
relatively earlier (type A or early zones of type B)
or later (type B) growth stages. Wherever
subsamples differing in relative age of crystal
growth are available they are connected by an
arrow on the plots to show the geochemical
trends with time for each sample.
One clear feature of the diagrams is the
complete overlap of the chemical data for types
FIG. 6. Sample Nj-29 seen in plane polarized light A and B dolomite and the conflicting trends of
(a) and under CL (b). Scale bar 100 ~tm. Dolomite variation from early to late stages of growth in
crystals projecting into vugs are coated by haematite different samples. This indicates that type A and
(arrowed in a) and succeeded by calcite cement (c). all stages of type B dolomite can potentially be
 Dolomitization of Devonian carbonates, South China 16 3
TABLE 1.

mole%
Type A Type B Crystal 513C (~180 Sr Na Fe Mn CaCO 3 in
Sample dolomite dolomite size, ~tm % %0 ppm ppm ppm ppm dolomite %IR

UA + 40 -3.26 -4.24 76 215 972 128 55.7 5.29

Njl-1 + 52 -2.41 -2.62 78 215 198 53 55.7 1.31
Nil-2 Early and late 131 -2.81 -3.70 79 210 445 65 53.2 4.43
Nj7-1 Early 86 -2.42 -4.57 74 195 196 59 54.2 0.32
Nj7-2 Late 242 - 3.58 - 5.50 67 179 196 70 56.9 0.25
Nj9-1 + 63 -2.32 -4.23 88 196 437 80 52.8 2.72
N j9-2 Early and late 157 -2.32 -4.51 81 164 394 56 52.6 0.71
NjI8-1 + 66 75 188 287 69 52.8 2.02
NjI8-2 Early 156 68 227 704 73 52.6 0.90
N j24 (+) 99 - 1.17 -4.93 120 215 408 91 54.2 1.18
N j28 + 68 78 167 512 130 52.9 5.32
N j29-1 + 69 -0.84 -2.34 93 255 433 80 51.7 5.90
Nj29-2 Early and late 131 -0.81 -1.53 96 255 289 81 52.4 8.82
Ml-1 Early 86 56 175 323 59 52.3 0.29
MI-2 Late 202 65 167 175 56 52.4 0.18

-0 13 C PDB
%0
•+ + ----~1--1
--1
(+) (+)

--2

//
~]4-+ j +

--3
4-
El []

--L

ppm Sr 0 ~180pD B
I I I I I I u6 I I I I n
120 100 80 60 LO 2o - -5 -L -3 -2 -1
-20
-49
4- Type A dolomite -60

[] Type B dolomite -80

-100
-120
-140

-160 El
-180

-2O0
(+) (+) + D ~--___._ +
[] -220
-240
D+ ppm +. ,~z3
-260
Na

FIG. 7. Plot o f 51ac, 6180, Sr a n d N a to illustrate their covariation. K e y as T a b l e 1 ; significance o f

correlations listed in T a b l e 2.
I64 Z h a o X u n & I. J. Fa ir c h ild

TABLE 2. Correlations between para- The data show a clear positive covariation of
meters expressed as the Pearson 613C, Sr, 6180 and Na (Fig. 7), inverse correla-
correlation coefficient (calculated by tion of ~13C and Ca content of dolomite (Fig. 8),
the methods of Lemaitre 1982), with positive correlation of Fe and Mn (Fig. 9) and
significance level in parentheses positive correlation of IR with Mn (Fig. 10), Sr,
(only correlations significant at min- Na, and 6180 (Table 2). It is notable that crystal
imum 90% level included) size shows no correlation with geochemical
parameters. This is in agreement with the lack of
Significance level overall geochemical trends with crystal growth,
Covariant
parameters withoutNj-24 with Nj-24 as already noted.
The correlation of IR with Na could be
~13C-Sr +0.92(99) +0.81(97) explained by some leaching of these elements
t$1aO-Na +0.88(98) +0.81(97) from insoluble residue (Veizer et al. 1977),
Sr-IR +0.75(99) +0.40(<50)
Fe-Mn + 0.71(98) + 0.70(98) although the correlations with Sr, Mn and 6180
t$13C-%CaCO3 -0.82(96) -0.73(92) are unexpected. However, considerable geo-
~3C-6~so +0.81(96) +0.60(66) chemical variation occurs in samples with low
Na-IR +0.64(95) +0.61(94)
~lsO-Sr +0.80(95) +0.19(<50) content o f l R (e.g. Fig. 10). Also, the IR of Nj-29
Sr-Na + 0.62(93) + 0.51(78) is clearly post-depositional haematite, different
IR-Mn +0.61(92) +0.57(89) from the IR of other samples. Removal of this
~sO-IR +0.75(90) +0.76(94) data point removes the statistically significant
correlation of IR with Na and 6180.

Isotopes
explained by a single process with reversible
geochemical characteristics. The assumption is made in the following
The data show clear covariations between discussion that the dolomites preserve their
several variables (Table 2). One sample (Nj-24) original isotope values. This is warranted by
does not fit these trends as far as 3180 is their preserved elemental zones. In contrast,
concerned. This Sample contains some clear many ancient finely crystalline dolomites seem to
dolomite veins with bright luminescence, similar have undergone some depletion in 180, presuma-
to veins (separated from a younger Devonian bly due to recrystallization (Land 1980). Hence
sample in another cross-section) which yielded we do not compare the Yingtang dolomites with
3180 of - 12. Although such veins do not occupy Palaeozoic fine-grained dolomites, but with well-
more than 10% of the sample, as seen in thin preserved Devonian calcitic components which
section, it is possible that a higher density of allow the isotopic composition of ancient sea
veins was met with when drilling out the sample water to be inferred. Recent work on Devonian
powder. The oxygen isotope composition of Nj- brachiopods (Poppet al. 1986, Veizer et al. 1986)
24 is thus set aside in the following discussion. suggests that if one assumes no temperature shift

513 r,
~/o2 PDB

+--,--El
-1-
I+1

-2-
El+

-3-

§ ""~ n
-4-

mote % Ca CO3 in dolomi~'e

FIG. 8. Plot of 613C versus Ca content of dolomite (see also Tables 1 and 2).
 Dolomitization of Devonian carbonates, South China I65
130 - ppm + +
120 - Nn
110_

100_

go- (+1
SO-
//
. . . . . ~ []
70-

60_

50_

40-

30-

20-

10-

o 16o 26o 360 ~o 5~o 66o 700 800 900 I000

p pm Fe
FIG. 9. Plot of Mn versus Fe (see also Tables 1 and 2).

%
Insoluble
Residue
(IR)

§247

/
/ §

E
E
s'o 6'o 7b a'o 9'o 16o 1'1o 1~o 1'3o
pprn pin

FIG. 10. Plot of IR versus Mn (see also Tables 1 and 2).

in low-latitude seas, the hydrosphere would have Marine Devonian carbonates have carbon
been depleted in x80 by about 3%0 compared isotopic composition around zero to + 1 (Holser
with today: hence sea water &180 would have 1984). Since the Yingtang dolomites are isotopi-
been of the order of - 4 . Since there is no cally negative they must have incorporated
evidence that evaporites ever formed in or excess 12C from organic or atmospheric sources.
directly above the Yingtang Formation, hyper- Given the conclusions of the last paragraph, a
saline fluids need not be considered here. Using freshwater source of excess 12C would have been
the dolomite-water fractionation equation advo- available if the carbonate bank became emer-
cated by Land (1983) and removing the Sharma gent. A meteoric lens would have developed with
& Clayton (1965) correction to the data, if sea isotopically negative carbon arising from the
water were the dolomitizing agent, palaeotem- dissolution of carbonates by rainwater charged
peratures would fall in the range 22-37~ One with atmospheric CO2, and perhaps by CO2-
can also consider the isotopic range of fluids incorporation into groundwater from decay of
corresponding to the dolomites' isotopes under land plants colonizing the emergent surface. Two
isothermal conditions. If the temperature was other possible mechanisms for 13C-depletion are
20~ the fluids would have ranged from around introduction of isotopically light carbon to
- 4 to - 8 SMOW. These data seem most porewater by bacterially mediated reactions such
consistent with near surface dolomitization by as sulphate reduction, and introduction of light
mixed meteoric-marine waters, with a dominant CO2 by thermal breakdown of organic matter.
marine component. These are discussed below.
I66 Zhao Xun & I. J. Fairchild
A positive covariation of oxygen and carbon hydrological situations, but is an independent
isotopes can arise in different circumstances. variable in general. Also some Na and Sr has also
The mechanism we prefer is that of mixing of probably been derived from fluid inclusions
meteoric and marine waters (e.g. Allan & and/or by leaching of IR. Although quantitative
Matthews 1982, Ward & Halley 1985). The slope modelling is not thought appropriate in this
of the covariation will vary in different circum- instance, qualitatively the good correlation of
stances, depending on both the isotopic compo- both Sr and Na with 5180 and 313C is best
sition of the groundwater and the degree of attributed mostly to salinity variations of the
inheritance of carbon isotopes from precursor dolomitizing fluids.
carbonate, the latter perhaps being relatively The good correlation of Fe and Mn with each
small in this case. An alternative mechanism other and not with other geochemical parameters
arises during burial diagenesis where increasing indicates a separate control on their abundance.
temperatures, leading to progressively more Chemical zonation of Mn is indicated by the
negative oxygen isotopes in cements, are ac- luminescence zonation given that Fe is well
companied by increasing input of isotopically below the concentrations required to quench
light carbon to pore waters by decarboxylation luminescence (Fairchild 1983). One sample was
(e.g. Dickson & Coleman 1980, Tucker 1982). microprobed and Fe-zoning was found to be
This is thought inapplicable to the Yingtang present. Complex Mn or Fe-Mn zonation is
dolomites for the following reasons: (1) there is known from inferred phreatic (Meyers 1978) and
no relation between growth stage and isotopic mixing-zone (Choquette & Steinen 1980) carbon-
values: (2) oxygen isotope values are too heavy; ates and is likely to result principally from Eh
(3) the sharp top to the dolomitized strata variations in the source area, related to changing
suggests near-surface dolomitization; (4) vari- water-table position. Values of iron in particular
able amounts of burial diagenetic dolomite are are too high to have formed under oxygenated
present in higher parts of the Devonian sections, conditions, but are compatible with mixing-zone
they differ in being more coarsely crystalline fluids.
(average around 200 ~tm) and more negative The reasoLs for departures from dolomite Ca-
oxygen isotope values (12 analyses average Mg stoichiometry have been much debated, but
- 8%0). are not resolved. Despite the expectation that
The existence of the covariation is also an dilute water 'limpid' dolomite would be stoichio-
argument against incorporation of much carbon metric because of ideal conditions of slow growth
via sulphate reduction since sulphate availability (Folk & Land 1975), such dolomite can be
will be least in the least saline fluids. The limited extremely calcian (Ward & Halley 1985). Eva-
range of isotopic variation contrasts with dolo- poritic dolomite forming in modern sabkhas
mites in mixed carbonate-siliciclastic organic- with demonstrably high pore water Mg/Ca ratios
rich sediments (e.g. Baker & Burns 1985, may also be deficient in Mg (e.g. Patterson &
Behrens & Land 1972) where bacterial reactions Kinsman 1982), although there is some evidence
are important in supplying carbon for carbonate. that within a given setting, higher Mg/Ca in
Nevertheless, it is recognized that the depletion porewater is associated with more stoichiometric
in carbon is more than might be expected for a dolomite (Sass & Katz 1982). Ancient evaporitic
mid-Palaeozoic mixing zone, and a subordinate dolomites tend to be near-stoichiometric whilst
source from bacterial reactions (probably includ- non-evaporitic ones are more variable and more
ing Fe- and Mn-reduction) may be required. Ca-rich on average (Lumsden & Chimahusky
1980). In general, one would expect that fine-
grained and poorly-crystalline calcian dolomites
Elemental chemistry
would tend to recrystallize and become more
The Sr and Na values are relatively low for stoichiometric during diagenesis, while original-
dolostones and typical of examples interpreted as ly coarser dolomite may retain its original
non-evaporitic or mixing-zone in origin (e.g. chemistry and preserve its original growth zones.
Land 1973a, Choquette & Steinen 1980, Churnet The Yingtang dolomites show a wide range in
& Misra 1981). However, the absolute values are Ca content from 51.7 to 56.9 mole% CaCO3.
not direct indicators of salinity since Sr and Na Sample Nj-1 analysed by microprobe showed
may be variably inherited from precursor car- intra-crystalline variations of up to 1.7 mole%.
bonate and Na incorporation is now known to be Since the type B dolomites preserve fine zona-
related to lattice defects (Busenberg & Plummer tion, these chemical variations are probably
1985). Defect density depends on speed of original. Ward & Halley (1985) noted consider-
growth of crystals which is a function of able variations in stoichiometry with growth in
supersaturation of solution. Supersaturation may Pleistocene Yucatan mixing-zone dolomite cry-
be a simple function of salinity in individual stals. The Yingtang dolomites show a good
 Dolomitization of Devonian carbonates, South China I67

inverse correlation of ~13C with Ca content: the tion were apparently associated with more rapid
more 'marine' samples are more stoichiometric. dissolution of calcite leading to cementation
Since sea water would be expected to have a instead of replacement by dolomite. Vuggy
much higher Mg/Ca ratio than fresh water on an porosity is most abundant near the top of the
emergent carbonate shoal, this relationship is the Formation, presumably related to the final
one predicted by Sass & Katz (1982). Note, episode of emergence. It is possible that all the
however, that the least stoichiometric dolomites leaching throughout the Formation relates to this
described by Sass & Katz (1982) were inferred to time. The patchy nature of the later calcite
have formed in more saline waters, but with cementation suggests an origin related to early
Mg/Ca lowered by dolomitization in a diageneti- phreatic processes, rather than later burial
cally isolated system (Fig. 12). Hence, since cementation.
relatively low Mg/Ca can occur in waters of both The extent of mixing-zone dolomitization in
low and high salinities, so can Ca-enriched the underlying Tangding Formation is unknown.
dolomite. The greater lateral persistence of this Formation
(lower half of dolostone lens of Fig. 1c) and the
occurrence of length-slow chalcedony suggests
Discussion that penecontemporaneous dolomitization may
be important in this formation.
The stratigraphic evidence that dolomitization This paper contains inferences which appear
was near-surface, the fact that open-marine to contradict the statements of Machel &
sediments on palaeotopographic highs were Mountjoy (1986) concerning the efficiency of
preferentially dolomitized, and the isotopic val- mixing-zone dolomitization: they stress the
ues and their co-variation are the main lines of incomplete nature of the dolomitization and the
evidence used to support the hypothesis of predominance of dolomite cement. Based on
dolomitization by mixed meteoric-marine fluids. isotopic evidence one can distinguish two kinds
However, the initial hypothesis that this could of Neogene mixing-zone dolomites: those with
relate entirely to the final episode of emergence is highly negative isotopic values (particularly for
contradicted by the stratigraphic variation of carbon) which appear to be forming in near
613C, 6180, Sr and Na (palaeosalinity indicators freshwater solutions with Mg supplied perhaps
in this case) which do not show uniform by diffusion (Land 1973b, Magaritz et al. 1980)
lightening of isotopes or decreases in minor and those with heavier isotopic values indicating
elements upwards (Fig. 2, Table 1). This points either a high proportion of sea water in the mixed
to the necessity for a series of periods of fluid, or else evaporation (Land 1973a, Sibley
emergence during the deposition of the Yingtang 1980, Ward & Halley 1985). The Yingtang
Formation. Although such emergence surfaces dolostones, although intermediate, seem closer
have not yet been found, they need never have isotopically to the second type which also
been prominent and fieldwork to date has not contains the examples of more complete dolomi-
been geared to looking for such features. The tization. It seems likely that these Neogene
existence of repeated emergence removes the examples do not fully illustrate the range of
necessity for large-scale erosion at the top of the potential mixing-zone dolostones. They formed
Yingtang Formation, but some erosion (?up to a during times of geologically brief sea-level fall in
few tens of metres) must have occurred to the chaotic eustatic pattern of the Neogene. They
account for the presence of largely dolomitized do include examples of replacive dolomite with
strata at the surface when the succeeding crystals on the 100 ~m scale (Sibley 1980, 1982),
limestones were deposited. but show highly variable fabrics and degrees of
The overall change from inclusion-rich to dolomitization probably related to their unstable
inclusion-poor (ultimately cementing) dolomite history. In the Yingtang example we require
is common throughout the Formation although, overall more stable conditions allowing dolomiti-
as suggested above, this must be at least partly zation to proceed to completion. Periodic emer-
diachronous. The lack of fine-scale mimic gence caused by sedimentation leading to the
replacement and presence of calcite inclusions production of freshwater lenses above calcite
suggest that dolomitizing solutions were only saturation and highly undersaturated for dolo-
weakly supersaturated for dolomite (few nuclea- mite could lead at shallow depth to mixing-zone
tion sites) and not greatly undersaturated for dolomitization in solutions slightly supersaturat-
calcite. Sibley (1982) argued for dolomitization ed for dolomite and slightly undersaturated or
of low-magnesian calcite by solutions slightly oversaturated (Sibley 1982) for calcite leading to
oversaturated for calcite to yield the inclusion- the observed textures. It should also be borne in
rich cores common to much replacive dolomite. mind that the dolomitization may have been
Later stages of growth in the Yingtang Forma- aided by factors other than the thermodynamic
I68 Zhao Xun & I. J. Fairchild

YINGTANG MODEL (this paper) I SOREQ MODEL (Sass & Katz,1982)

I
Shallow do[omitization with d o m i n a n c e of replacive dolomite
Mixing zone dolomitization Dolomitization close to sediment surface
few
s SHOAL A metres

DIFFUSION or REFLUX i ? EVAPORATION

"~.~.~"-~ ~ y ;metres 9 L A T E R A L -~F-- ~, few metres
NIXING ZONE FLOW
Isotopic variation reflects s a l i n i t y t r e n d s
No stratigraphic trends in geochemical data because:
Repeated emergence Fluids s h o w repeated diagenetic
evolution and replenishment

Isotopes too light for marine fluids Isotopes consistent with m a r i n e - b a s e d fluids
Leaching and dolomite cementation present No leaching or dolomite cementation

Sr,Na show limited variation controlled by salinity Sr,No v a r i a t i o n large owing to evolution of
diagenetica[ty- isolated fluids

>.. SALINE Mg/Ca fluid increases with >- SALINE . ~ ~ C a decreases with diagenetic
i.- isolation
MARINE depth at a z MARINE
- time u
-J FRESH ~ <c F R E S H
bo
Ca -rich Stoichiometric Ca -rich Stoichiometric
M g / C o Dolomite Mg/Ca Dolomite

FIG. I I. Comparison of the mixing-zone model for the Yingtang Formation with the model for the
Cretaceous Soreq Formation of Israel as described by Sass & Katz (1982). Note that the Na modelling of
these authors is invalidated by the work of Ishikawa & Ichikuni (1984) and Busenberg & Plummer (1985).

changes resulting from mixing. For example, it type of dolomitization. Although several similar-
could be argued that kinetic triggers such as ities exist, clearly two different models are
sulphate reduction or removal or organic inhibi- required to explain these two dolomitic forma-
tars were important. Unfortunately the available tions. This study confirms the necessity of
data do not allow us to evaluate these studying not only the numerical value of geo-
possibilities. chemical parameters but also the relationships
The shallow subtidal model advocated by between them in order to assess the dolomitiza-
Machel & Mountjoy 0986) may well explain tion mechanism (Sass & Katz 1982).
many platform dolomites, but suffers even more
from the lack of latter-day analogues than does
the mixing-zone model since described examples
of Holocene shallow marine dolomitization have ACKNOWLEDGMENTS:Zhao Xun is supported by the
variable carbon isotopic values indicating a link Chinese Government and the British Council. Isotope
with early bacterial reactions, and hence a work was carried out at the NERC Stable Isotope
different model The lack of a stable sea-level Facility at the British Geological Survey, London
probably also explains the restricted range of where Charles Branch and Baruch Spiro gave valuable
subtidal dolomite occurrences. instruction and assistance. ICP work was carried out at
Figure 11 summarizes the main characteristics Kings College, London: the help of Alison Warren and
of the model of dolomitization of the Yingtang Nick Walsh is gratefully acknowledged. At Bir-
mingham, Graham Hendry is thanked for his patient
Formation by repeated emergence, and makes a
assistance with XRD and advice on analytical proce-
comparison with the model of Sass & Katz (1982) dures, Paul Hands for rock-sectioning, and Carl
based on the Cretaceous Soreq Formation of Burness for draughting. We are grateful to Baruch
Israel. The latter is used by Machel & Mountjoy Spiro, Julian Andrews, Maurice Tucker and Jim
(1986) as a prime example of the shallow subtidal Marshall for their comments on the manuscript.
 Dolomitization of Devonian carbonates, South China 169
References
ALLAN, J. R. & MATTHEWS, R. K. 1982. Isotope Guangxi Geological Institute (in Chinese with
signatures associated with early meteoric diagene- English abstract).
sis. Sedimentology, 29, 791-817. LAND, L. S. 1973a. Contemporaneous dolomitization
BAI, S.-L., JIN, S.-Y. & NING, Z.-S. 1982. The Devonian of Middle Pleistocene reefs by meteoric water,
Biostratigraphy of Guangxi and Adjacent Areas. North Jamaica. Bulletin of Marine Science, 23, 64-
Beijing University Press House (in Chinese). 92.
BAKER, P. A. & BURNS, S. J. 1985. Occurrence and Holocene dolomitization of Pleistocene
- - 1 9 7 3 b .

formation of dolomite in organic-rich continental limestones, North Jamaica. Sedimentology, 20,

margin sediments. American Association of Petro~ 411-24.
leum Geologists Bulletin, 69, 1917-30. - - 1980. The isotopic and trace element geochemis-
BEHRENS,E. W. & LAND,L. S. 1972. Subtidal Holocene try of dolomite: the state of the art. In: ZENGER,
dolomite, Baffin Bay, Texas. Journal of Sedimen- D. H., DUNHAM, J. B. & ETHINGTON, R. L. (eds).
tary Petrology, 42, 155-61. Concepts and Models of Dolomitization, pp. 87-110.
BUSENBERG, E. & PLUMMER, L. N. 1985. Kinetic and Society of Economic Paleontologists and Minera-
thermodynamic factors controlling the distribu- logists Special Publication, 28.
tion of SO42- and Na + in calcites and selected - - 1983. The application of stable isotopes to studies
aragonites. Geochimica et Cosmochimica Acta, 49, of the origin of dolomite and to problems of
713-25. diagenesis of clastic sediments. In: ARTHUR,
CHOQUE'VrE, P. W. & STEINEN, R. P. 1980. Mississip- M. A., ANDERSON,T. F., KAPLAN, I. R., VEIZER,
pian non-supratidal dolomite, Ste, Genevieve J. & LAND, L. S. (contributors). Stable Isotopes in
Limestone, Illinois Basin: evidence for mixed- Sedimentary Geology, pp. 4-1-22. Society of Eco-
water dolomitization. In: ZENGER, D. H., DUN- nomic Paleontologists and Mineralogists Special
HAM,J. B. & ETHINGTON, R. L. (eds). Concepts and Publication, 10.
Models of Dolomitization pp. 163-96. Society of LEMAITRE, R. W. 1982. Numerical Petrology. Elsevier,
Economic Paleontologists and Mineralogists Spe- Amsterdam.
cial Publication, 28. LIAO, W.-H., Snu, H.-G., WANG, C. Y., CAI, C. Y.,
CHURNET, H. B. & MISRA, K. C. 1981. Genetic RUAN, Y.-P., MU, D.-C. & LU, C.-L. 1979.
implications of the trace element distribution Description of some Standard Devonian Sections in
pattern in the Upper Knox carbonate rocks, South-west China, pp. 221-9. Press House of
Copper Ridge District, East Tennessee. Sedimen- Science (in Chinese).
tary Geology, 30, 173-94. LUMSDEN,D. N. & CHIMAHUSKY,J. S. 1980. Relation-
DICKSON,J. A. D. & COLEMAN,M. L. 1980. Changes in ship between dolomite non-stoichiometry and
carbon and oxygen isotope composition during carbonate facies parameters. In: ZENGER, D. n.,
limestone diagenesis. Sedimentology, 27, 107-18. DUNHAM, J. B. & ETnlNGTON, R. L. (eds).
DINELEY,D. L. 1984. Aspects of a Stratigraphic System ." Concepts and Models of Dolomitization, pp. 123-37.
the Devonian. Macmillan, London. Society of Economic Paleontologists and Minera-
FAIRCHILD,I. J. 1983. Cathodoluminescence of natural logists Special Publication, 28.
calcites and dolomites: new data and review. MACHEL, H.-G. & MOUNTJOY, E. W. 1986. Chemistry
Sedimentology, 30, 579-83. and environments of dolomitization--a reapprais-
FOLK, R. L. & LAND, L. S. 1975. Mg/Ca ratios and al. Earth Science Reviews, 23, 175-222.
salinity: two controls over crystallization of MAGARITZ, M., GOLDENBERG,L., KAFRI, O. & ARAD,
dolomite. American Association of Petroleum Geolo- A. 1980. Dolomite formation in the seawater-
gists Bulletin, 59, 60-8. freshwater interface. Nature, 287, 622-4.
HOLSER, W. T. 1984. Gradual and abrupt shifts in MEYERS, W. J. 1978. Carbonate cements: their
ocean chemistry during Phanerozoic time. In: regional distribution and interpretation in Missis-
HOLLAND, H. D. & TRENDALL, A. F. (eds). sippian limestones of southwestern New Mexico.
Patterns of Change of Earth Evolution, pp. 171-205. Sedimentology, 25, 371-99.
Springer-Verlag, Berlin. PATTERSON,R. J. & KINSMAN,D. J. J. 1982. Formation
Hou, H. 1978. The Devonian system in South China. of diagenetic dolomite in coastal sabkha along
In: CHINESEGEOLOGICALACADEMY(ed.). Confer- Arabian (Persian) Gulf. American Association of
ence on Devonian in South China, pp. 214-40. Petroleum Geologists Bulletin, 66, 28-43..
Geological Press House (in Chinese). PoPP, B. N., ANDERSON,T. F. & SANDBERG,P. A. 1986.
ISHIKAWA,i . & ICHIKUNI,i . 1984. Uptake of sodium Textural, elemental and isotopic variations among
and potassium by calcite. Chemical Geology, 42, constituents in Middle Devonian limestones,
137-46. North America. Journal of Sedimentary Petrology,
KALDI, J. G. & GIDMAN, J. 1982. Early diagenetic 56, 715-27.
dolomite cements; examples from the Permian SASS, E. & KATZ, A. 1982. The origin of platform
Lower Magnesian Limestone of England and the dolomites: new evidence. American Journal of
Pleistocene carbonates of the Bahamas. Journal of Science, 282, 1184-213.
Sedimentary Petrology, 52, 1073-86. SHARMA, T. & CLAYTON, R. N. 1965. Measurement
KUANG, G.-D. 1982. The Devonian cross-sections of of O18/O ~6 ratios of total oxygen and of
Liujing and Loufu in the Guangxi region. In: carbonates. Geochimica et Cosmochimica Acta, 29,
Devonian Cross-sections in Guangxi, pp. 1-47. 1347-53.
I7O Zhao Xun & I. J. Fairchild
SIBLEY, D. F. 1980. Climatic control of dolomitization, VEIZER, J., FRITZ, P. & JONES, B. 1986. Geochemistry
Seroe Domi Formation (Pliocene), Bonaire, N. A. of brachiopods: oxygen and carbon isotopic
In: ZENGER, D. H., DUNHAM,J. B. & ETHINGTON, records of Paleozoic oceans. Geochimicaet Cosmo-
R. L. (eds). Concepts and Models of Dolomitization, chimica Acta, 50, 1679-96.
pp. 247-58. Society of Economic Paleontologists --, LEMIEUX, J., JONES, B., GIBLING, M. R. &
and Mineralogists Special Publication, 28. SAVELLE, J. 1977. Sodium: palaeosalinity indica-
-- 1982. The origin of common dolomite fabrics: tor in ancient carbonate rocks. Geology, 5, 177-9.
clues from the Pliocene. Journal of Sedimentary WARD, W. C. & HALLEY,R. B. 1985. Dolomitization in
Petrology, 52, 1087-1100. a mixing zone of near-seawater composition, Late
TUCKER, i . E. 1982. Precambrian dolomites: petro- Pleistocene, Northeastern Yucatan peninsula.
graphic and isotopic evidence that they differ Journal of Sedimentary Petrology, 55, 407-20.
from Phanerozoic dolomites. Geology, 10, 7-12. WENK, H.-R. & ZHANG, F. 1985. Coherent transforma-
tions in calcian dolomite. Geology, 13, 457-60.

ZHAO XUN* & IAN J. FAIRCHILD, Department of Geological Sciences, University of

Birmingham, PO Box 363, Birmingham B 15 2TT, UK. *Permanent address: Bureau of
Geology and Mineral Resources of Guangxi, 1, Jianzheng Road, Nannin City,
Guangxi, China.
Diagenesis of carbonate cements in Permo-Triassic sandstones in
the Wessex and East Yorkshire-Lincolnshire Basins, UK: a
stable isotope study
A. H. Bath, A. E. Milodowski & B. Spiro

S U M M A R Y : The 13C/12Cand 180/160 ratios of carbonate cements in Permo-Triassic

fluvio-deltaic sandstones differ between the Wessex and East Yorkshire-Lincolnshire
Basins. The isotopic data indicate different conditions of deposition and early post-
depositional carbonate precipitation which supports petrological and mineralogical studies
of the same sandstones focussed on the development of reservoir properties. Isotopic
compositions of carbonate cements vary across the Wessex Basin; they indicate early
derivation of dissolved carbonate during recrystallization of detrital limestone. Within the
basin, calcite cements show little variation in 613C but considerable differences in ~lso
between early calcrete calcites and isotopically-lightlater calcites; 6180 decreases to values
close to isotopic equilibrium with present water while ~13C is not in isotopic equilibrium with
bicarbonate in present-day brines. Two distinct generations of replacive dolomite, early non-
ferroan and later ferroan, are found in the basin centre and have different ~180 but similar
~13C values. 34S/32Sand 180/160 ratios of anhydrite cements could indicate a remote
Zechstein source of sulphate or a mixed marine sulphate-sulphide source. The Permo-
Triassic sandstones of the East Yorkshire-LincolnshireBasin, in contrast, contain dominantly
dolomite cement. ~18O and ~13C values are relatively heavy and indicate early dolomitization
of calcite over large areas. ~180 values of minor calcites indicate that dedolomitization is
continuing in the present-day groundwater regime. The isotopic differences between carbon-
ate cements in the two basins is tentatively attributed to the differing early post-depositional
conditions and to the differences in carbonate chemistry between calcite dissolution/precipita-
tion reactions and dolomitization reactions.

The Wessex and East Yorkshire-Lincolnshire accumulation of clays and limestones as further
Basins are onshore structures in southern and basinal subsidence occurred along growth faults.
eastern England respectively (Fig. 1). They are Maximum burial depth of basal Triassic sand-
both filled with Permian and Mesozoic sedi- stones is thought to have been around 2.5 km in
ments which reach maximum thicknesses of over the fault-bounded Winterborne Kingston trough
3000 m in the Wessex Basin and over 2000 m in near the centre of the basin; this is only slightly
the East Yorkshire-Lincolnshire Basin. more than their present depth (Fig. 3).
The Triassic Sherwood Sandstone in the The Permo-Triassic sandstones in the East
Wessex Basin (Fig. 2) was derived probably from Midlands, Lincolnshire and East Yorkshire are
sources to the south following early Permian fluvial deposits on t h e eastern England Shelf
subsidence which was variable and strongly (Fig. 4). The relatively stable eastern England
controlled by faulting. These sedimdnts, predo- Shelf is separated from the southern North Sea
minantly arenaceous until deposition of the late Basin, in which subsidence and deposition were
Triassic Mercia Mudstone, were laid down under more rapid, by the Dowsing Fault Zone (Fig. 4).
fluviatile or sub-aerial conditions in an arid or Subsequent subsidence of the shelf has resulted
semi-arid climate (Henson 1970, L o t t & Strong in burial and cover by Triassic clays and marls;
1982). Thicknesses of the Sherwood Sandstone the Permo-Triassic strata dip eastwards regu-
exceed 100 m in much of the basin, and reach a larly from outcrop running N-S through Not-
maximum of over 300 m. Detailed study of tinghamshire and Yorkshire. The depths at
depositional facies has shown extensive and which the Permo-Triassic sandstones occur on
complex lateral as well as vertical variation, the east coast of England are between 1100 and
suggesting transport and deposition by braided 1900 m. Here, as represented by the Cleethorpes
river systems (Henson 1970, Lott & Strong 1982). borehole, the Triassic Sherwood Sandstone is up
Triassic sedimentation culminated in the deposi- to 400 m thick, whereas the Basal Permian Sands
tion of the Mercia Mudstone. This represents a are around 25 m thick (Fig. 5).
low energy evaporitic environment in which silts Because of the potentially favourable aquifer
and clays were deposited, possibly with the properties of these two sandstone sequences and
occasional development of playas. Subsequently, their burial depth, they have been assessed for
in the Jurassic, marine transgression led to their potential as low enthalpy geothermal

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, I73

Geological Society Special Publication No. 36, pp. 173-190.
[74 A. H. Bath et al.

CARLISLE
BASIN . . . . . . -I

S
i Area s h o w n in
~--- Figure 4

EAST Y O R K S H I R E
A N D LINCOLNSHIRE
BASIN

W E S T LANCASHIRE L
~ND CHESHIRE B A S I N S

@ WORCESTER
BASIN
9 /

~, l>
~f . . . . . . . . . . .
~ ~ 1 % W E S S E X BASIN

Area s h o w n in 0 50 100 150

Figure 2 t I I f
km
FIG. 1. Locations of the Wessex and East Yorkshire-Lincolnshire sedimentary basins; the other three
Mesozoic basins in England and Wales are also shown.

energy sources (Downing & Gray 1986). This study of factors determining the reservoir pro-
paper describes the stable isotopic evidence for perties of the Permo-Triassic sandstones (Milo-
the environments of formation and for sources of dowski et al. 1986, 1987, Strong & Milodowski,
carbonate and sulphate cements; the study was this volume). The Permo-Triassic sequences of
part of a broader petrological and mineralogical both the Wessex Basin and North Sea Basin are

\ X ~e~- \ London

"'~ "" "~" ~-- Mere Fault

--4J ~ /2 " """ ~ - - ~ ~ ~ "~

Ii ,'
i'TAUNTON ....... 600 - -" "- "-. " " t ~ -uo_
- - . . . . . -. -.. {, ~"
/ / 1000 "" A " SOUTHAMPTON t
,. .- .... ~(---='r. L;re~ ...... .. . .~Western Esplanade Portsdown -

', ', "",'- _ ~'00~'~0 W nterborne'~, k~ k-l. ij - (. .

I--I00~ ) ~ ~ '~ ' ,'---'K ngston ', "~ ~ " - ~ ~).,,C~l~ ~/"~/'" ~ ~ -

/9s / ~ " - , B O U R N E M O U ~ ~ ~ ~ v

0 10 2
I I I 400 500
I I

FIG. 2. Locations of drillcore samples from the Wessex Basin. Samples were also studied from the Sherwood
Sandstone outcrop on the western edge (shown stippled). Contours show depth (metres below sea-level) to top
of the Sherwood Sandstone Group (after Downing & Gray 1986).
 Carbonate cement diagenesis, UK [75
SOUTH DEVON WlNTERBORNE MARCHWOOD
COAST KINGSTON

MERCIA
oos,oN ,iii:t
ii
I ;ii!iiiiiiii.i
i;i il
.... . ..... :
,
iN120m -".':'i'i:"~ '"': / 725m
t O'IEERSANDSTONE

:.:,-"':':'::.:.:: / / OLD RED

' ~ .......... / / SANDSTONE
~ooo(
~[BUDLEIGH SALTERTON >ooo o o< ~ o ( N30m
~"
L PEBBLEBEDS o ooo

).':. i)''"
.....
. . .::::":2 ~ 4 6 / ~ ~

AYLESBEARE
GROUP

KEY
Predominantly feldspathic sandstones
with subordinate conglomerates
Quartz-lithic muddy sandstones
Predominantly conglomerates
with subordinate sandstones
I Cored interval

FIG. 3. Simplified stratigraphic sections showing thicknesses and lithologies of the Triassic Sherwood
Sandstone Group across the Wessex Basin from the western outcrop in South Devon to the Marchwood
geothermal well on the eastern side of the basin; the Winterborne Kingston sequence represents the
depositional centre of the basin.

important locally as hydrocarbon reservoirs--in and anhydritic mudstones and siltstones are also
the onshore Wytch Farm oilfield and in offshore present in wells drilled close to the basin
gasfields respectively. depocentre, particularly in the lower part of the
sequence. Calcite cement forms nodular expan-
Petrological background sive calcrete concretions at the tops of some
cycles in the marginal parts of the basin, e.g.
The petrology and diagenetic history of the outcrop on the western rim and boreholes at
Sherwood Sandstone in the Wessex Basin is Southampton on the eastern rim.
described in detail by Strong & Milodowski (this The major diagenetic processes identified by
volume). In summary, the formation has a lower Strong & Milodowski (this volume) are: (i) early
conglomeratic unit referred to at outcrop on the cementation by calcite and anhydrite inhibiting
western margin of the basin as the Budleigh further compaction followed by local later-stage
Salterton Pebble Beds (Fig. 3). This is overlain removal of anhydrite; (ii) variable degree of
by a thick sequence of feldspathic sandstones, compaction depending on framework grain
the Otter Sandstones at outcrop, exhibiting a (quartz, feldspar) characteristics and the devel-
series of fining-upwards cycles often with ero- opment of overgrowths, followed by later-stage
sional conglomeratic bases (Edmonds & Wil- partial dissolution of some framework grains,
liams 1985). Cementation by carbonate is a and (iii) subsequent invasion of this 'secondary
major feature of the sandstone throughout the porosity' by late-stage cements (calcite, dolomite,
basin; discrete thin bands of anhydrite (< 1 m) anhydrite). Earlier studies of drillcore indicated
I76 A. H. Bath et al.
I

, / I t I I I

/
!
i-.. .
#,
@ -
'; CLEVEILAND BASIN .~ @O
,, ,// / x . ~ .......... / / , a 0 ~

:: ::~ ...... - ..... "',, ', I T ', "7", ....,.wt~,",9 I

-... -.. ,, !~ ,...... -. ~.o,,,ol
' i ~ ......... "'" MAR,KE' T" " ~ ~'I" "" ""'", ~ 9
"i WEIGHTON
' , - \ X4 ~', "'] -t~ougn
x: ::::: : ::: 9
', [ BI'OCK
,
;\
, 9 , , ,
,' gas j - j \
I : ', , ; \ ,' fie d .

'h : P ,; ~~-,4_O-..Cle et h 6 rP e s - ~ ", >

-400

3amston ',

" ~
i '\,

' L.
, ,
..' I " ",..
i ",
NOTTINGHAM ,' ',,
,/ // '~

Y'LoNDON PLATFORM
5OO
1

FIG. 4. Locations of drillcore samples from the East Yorkshire-Lincolnshire Basin and the Sherwood
Sandstone outcrop. Contours show depth (metres below sea-level) to top of the Sherwood Sandstone Group
(after Downing & Gray 1986).

that framework grain (K-feldspar) dissolution in The Sherwood Sandstone in the East York-
some zones in late diagenesis has compensated shire-Lincolnshire Basin is dominated by feld-
for porosity reduced by compaction (Knox et al. spathic sandstones although the proportion of
1984). On the other hand, Burley (1984) inter- K-feldspar is lower than in the Wessex Basin
preted the present porosity distribution as result- (Milodowski et al. 1987; Fig. 6). Carbonate
ing from the removal of much of a widespread cementation is only weakly and patchily de-
early compaction-resisting carbonate cement. As veloped, dolomite being the major cement
a result of further basin-wide petrological accompanied by only minor patches of calcite.
studies, Milodowski et al. (1986) have suggested The varying cementation makes the sandstone
that widespread early anhydrite cementation friable, particularly at outcrop where carbonates
might also account for resistance to compaction, and feldspars show evidence of extensive dissolu-
being subsequently dissolved in many parts of tion. Anhydrite occurs as a major cement near
the basin. the top of the buried sandstone, but is absent at
 Carbonate cement d&genesis, UK I77
CLEETHORPES
BOREHOLE

MERCIA MUDSTONE
GROUP

"1" 1100m[
0%

SHERWOOD SANDSTONE
GROUP

FELDSPAR 51% LITHIC

FIG. 6. Triangular diagram showing approximate

1498m proportions of quartz, feldspar and lithic grains in
ESKDALE GROUP the Triassic Sherwood Sandstone samples in this
STAINTON DALE GROUP
study.
TEESSIDE GROUP

AISLABY GROUP Sherwood Sandstone Group in the two basins are

summarized in Fig. 7, which illustrates the
similarities and differences in this respect
DON GROUP between the two basins.

T1866m
1 8 9 1 rn
:3:.'i!":-:':':-~".?.': BASAL PERMIAN SANDS
Samples and methods
Four suites of drillcore samples from the Wessex
Basin (Fig. 2), and a drillcore from the Clee-
COAL MEASURES
thorpes well in NE Lincolnshire (Fig. 4) were
available for study. The Cleethorpes well and the
Winterborne Kingston, Marchwood and Wes-
tern Esplanade (near and in Southampton
respectively) wells in the Wessex Basin, were all
drilled specifically for the investigation of geo-
FIG. 5. Simplified stratigraphic sequence in Permo- thermal potential in the deep aquifers (Downing
Triassic sediments intersected by the Cleethorpes
geothermal well in the East Yorkshire-Lincolnshire & Gray 1986). In addition, a drillcore from a
Basin. Samples of feldspathic sandstone with hydrocarbon well at Wytch Farm was studied. In
dolomite cements from the Sherwood Sandstone both basins, samples of outcropping Permo-
Group and from the Basal Permian Sands were Triassic sandstone were collected for compari-
studied. The intervening five groups are evaporitic son; outcrops occur only on the western rim of
units within the Zechstein sequence. Key as in the Wessex Basin running N-S through Devon
Fig. 3. and Somerset (Fig. 2), whilst Sherwood Sand-
stone outcrops from Nottinghamshire north-
outcrop. The Basal Permian Sands at depth in wards into East Yorkshire (Fig. 4). Two drill-
the basin comprise massive poorly-sorted sand- cores from or close to outcrop in eastern England
stones with minor conglomerates and interbed- were also studied: at Wistow in South Yorkshire
ded siltstones. The sandstones have around 5- and Gamston in Nottinghamshire.
10% K-feldspar, and both these and the conglo- Samples for stable carbon and oxygen isotopic
merates are variably cemented by dolomite and analysis of carbonates were selected from zones
anhydrite. Dolomite in the conglomerates is for which full petrographic and mineralogical
clearly replacive of pre-existing carbonate; the information was available (Milodowski et al.
bulk of the dolomite throughout the sequence is 1986, 1987, Strong & Milodowski, this volume).
non-ferroan, but late-stage ankerite overgrowths Cathodoluminescence microscopy was particu-
are observed rimming the dolomite. larly valuable for identifying different genera-
The major diagenetic events which affect tions of carbonate, detecting mineral zonation,
carbonate and sulphate cementation of the as well as estimating the carbonate homogeneity.
[78 A. H. Bath et al.
Deposition Present

INCREASING BURIAL

Early Later

Calcite pptn I 1 ?FTODI J

mmmm

I Basin centre only

Dolomite pptn

w-Basin centre only

1 I j
Anhydrite pptn

Anhydrite I?
dissolution ? ~

Dedolomitization

Hydrocarbon Wytch Farm only

t" J
emv'acementl'~
+ Early Cretaceous .."Eocene

D D D D D ~ D O D D
Compaction effects
m n i m i | | | g m

t Wessex m E.Yorks-Lincs

FIG. 7. Summary comparison of major diagenetic events which affect carbonate and sulphate cementation in
the Sherwood Sandstone Group in the Wessex Basin and East Yorkshire-Lincolnshire Basin. (1Data on
timing of hydrocarbon emplacement into the Sherwood reservoir at Wytch Farm from Colter & Havard 1981.)

Due to the fine grain-size it was not feasible to Appendices 1 and 2 for the Wessex Basin and
separate calcite from dolomite, or to obtain East Yorkshire-Lincolnshire Basin respectively.
samples of discrete cements from complex The preparation of samples for stable isotope
samples. Most of the samples analysed had a analysis follows that described by McCrea
dominant single generation of cement. (1950). Small amounts of whole-rock samples
Early- and later-diagenetic calcites, as inferred were ground for 15 minutes in a ball mill, and 10-
by petrographic relationships, are clearly charac- 100 mg subsamples, depending on the estimated
terized by dull and bright orange cathodolu- carbonate content, were taken for 13C/12C and
minescence respectively. In addition dolomite is 180/160 analysis. Carbon dioxide was extracted
distinctive for its dull red luminescence. There- from samples by reacting with excess 100%
fore these characteristics were used to select orthophosphoric acid in v a c u o at a constant
appropriate samples and to estimate the amounts temperature of 25.2 + 0.05~ 6~80 was calculat-
of particular carbonate generations. However, ed using the constants 0~18co,-~aicit~= 1.01025
detailed petrographic and scanning electron (Friedman & O'Neil 1977) a/ad ~%O2-dolomit~
microscope studies in parallel with the cathodo- = 1.011 (Sharma & Clayton 1965).
luminescence observations suggested that the Isotopic analysis was performed on a VG-903
latter gave only a coarse classification; it is likely triple collector mass spectrometer. Results are
that the so-called early and later carbonates are quoted on the conventional del (6) scale in per
themselves subdivided into precipitates formed mil (%0) deviation from the PDB standard.
under ranges of environments within these broad The normal analytical precision of duplicate
classifications. Brief descriptions of the samples analyses is better than +0.1%o for both 613C
according to the above observations are listed in and 6180.
 Carbonate cement diagenesis, U K I79

Results and discussion Outcrop

Wessex Basin Carbonates in the outcrop samples of the

Sherwood Sandstone fall mostly into two groups
The stable isotopic compositions of total carbon- according to isotopic compositions (Fig. 8). A
ate in samples from the Wessex Basin are listed distinctive group (five analyses: 1, 2, 4, 5, 8) have
in Table 1. These results are plotted as 613C high ~13C ( - 2 . 7 to -1.2%o; B in Fig. 8) which
versus 6180 in Fig. 8, which includes data from suggests provenance from Carboniferous or
the previous study of Marchwood drillcore by Devonian detrital limestone (A in Fig. 8). Most
Knox et al. (1984). Isotopic ratios (62H, 61sO) of of these samples come from the Budleigh
present-day brines in the deep Sherwood Sand- Salterton Pebble Beds and contain discrete
tone are given in Table 2, which also shows the limestone pebbles; in one case a pebble has been
613C of dissolved bicarbonate in brine at separated and analysed (sample 1). The data
Marchwood. indicate recrystallization in a meteoric water
environment (either the early post-depositional
meteoric water which might have had 61sOrho
TABLE 1. 13C/12C and 1 8 0 / 1 6 0 analyses o f total
around - 4 to -6%0 vs SMOW at around 20-
carbonate content of samples from the Sher-
30~ or present meteoric water with ~ 8 0 H o
wood Sandstone in the Wessex Basin around - 6 to -7%0 at 10~ The ~13C ~f
~13C t~18O dissolved carbonate must have been dominated
by detrital calcite with a possibility of exchange
No. Sample number %o v s P D B only with atmospheric CO: (6~3C ~ -7%o), and
Western Outcrop with negligible contribution from CO2 of soil
1 HW03 ( c a l c i t e p e b b l e s ) -1.26 - 5.65 organic derivation. These conditions might have
2 HW03 (matrix) - 2.63 -7.22 occurred in the highly permeable Pebble Beds
3 MV01 -7.38 -9.49 due to rapid meteoric water flux. This process
4 MV02 - 2.67 -8.33
5 WS02 - 1.73 -6.20
continues presently at outcrop where the flux and
6 BL01 - 8.90 -6.09 hydrochemistry of groundwater in the Pebble
7 LB01 - 8.97 -5.33 Beds contrasts with conditions in the Otter
8 LB02 - 2.04 - 7.72 Sandstone (Walton 1982).
9 LB03 - 11.23 - 4.94
10 LBV01 - 7.76 - 5.78
A second group of samples (6, 7, 10), from the
Otter Sandstone formation, have isotopic com-
Winterborne Kingston drillcore positions ( - 9 to -7.5%0 613C, - 6 to -5%0
11 WK01 - 6.69 -11.33
12 WK06 -7.66 -8.44 61sO) similar to those found previously for
13 WK09 -6.83 - 11.62 calcrete carbonate in Marchwood drillcore
14 WKll -7.70 - 10.09 (Knox et al. 1984). These widespread early
15 WK18 -7.63 -1.13
calcretes formed by precipitation in fine- to
16 WK19 -7.45 - 1.00
17 WK22 -8.99 -0.43 medium-grained sediments in the shallow sub-
surface (evident from the expansive nature of the
Marchwood drillcore
18 E59535 -6.24 -10.11 cement and lack of compaction of the sediment).
19 E59536 (cornstones) -5.16 -2.08 Some organic activity in the soil zone, even in the
20 E59536 (bulk) -4.81 -2.84 semi-arid Triassic environment, would have
21 E59538 -7.62 -6.44
-8.41
generated CO2 which would have had 613C
22 MW 4 -7.76
23 MW 9 -8.45 -5.15 between -25%0 and -13%o, depending on the
24 MWl9 -7.37 - 12.96 plant types (Smith & Epstein 1971). The follow-
25 MW22 -8.98 -5.86 ing evaporation-reaction-exchange model de-
26 MW23 -8.33 -7.78
-6.14
scribes the processes and possible isotopic
changes (A to D in Fig. 8): (1) (~18OH20i n c r e a s e d
27 MW24 -8.93
Western Esplanade (Southampton) drillcore
28 E59553 - 5.54 -9.24 in this groundwater by up to 6%0 due to
29 E59557 - 7.07 -7.64 evaporative fractionation. (2) Dissolved carbon-
30 E59559 -6.84 - 7.08 ate in this groundwater had v a r i a b l e (~13CHco3
31 E59562 -6.24 -9.41
32 E59564 -6.78 - 7.68
depending on the behaviour of the dissolved
33 E59565 -6.37 -9.16 carbonate system as open or closed with respect
Wytch Farm X14 drillcore to evaporative loss of CO2 or to exchange with
34 E59540 - 5.43 -4.80 the reservoir of organically derived CO/ in the
35 E59542 -5.67 -4.78 soil zone. (3) Reaction of detrital carbonate with
36 E59544 -5.16 -6.66 dissolved soil z o n e C O 2 would give ~13CHco3
37 E59546 -6.57 -5.01
38 E59551 -7.24 - 6.42 between -13%o and -6%0 depending on the
isotopic compositions of carbon sources (Smith
I80 A . H . B a t h et al.

?A
513C %oPDB
-14 -12 -10 ~ - 8 -6 -4
I I ' ' I

l "J9 20

09
~ 0 3 7 5
09
~E J
~ -'" WAD \ 36 i
6180
%oPDB

Calculated compositions
-lO
of calcite in equilibrium
with p r e s Z y brine

i-12
G i

KEY
-14
0 Western Rim outcrop
9 Winterborne Kingston
9 Marchwood(D Knox et al., 1984)
9 Western Esplanade
9 Wytch FarmX14

FIG. 8. Plot of 613C against 6~so for samples from the Sherwood Sandstone in the Wessex Basin. Some
analyses previously reported by Knox et al. (1984) are also shown. Cross-hatched areas denote inferred
compositions of detrital calcite and of calcite in equilibrium with present-day brine. Stippled areas indicate
dolomite compositions. The arrows illustrate generalized isotopic evolution pathways. A--inferred
compositions of detrital ?Carboniferous/Devonian limestone; B---early recrystallized calcite at the basin rim,
trending towards C possibly due to mixing with later calcites precipitated from solutions expelled towards rim
during compaction; D--widespread early calcrete development in basin: E---early calcite occurring locally
(Wytch Farm) with possible minor contribution of CO2 from degradation (fermentation) of organics; F--
localized dolomitization of calcrete in sabkha developed at basin centre (Winterborne Kingston); G--
widespread later carbonates precipitated through maximum burial; H--localized later dolomite precipitated at
basin centre (Winterborne Kingston).

& Epstein 1971, Wigley et al. 1978, Fontes & atmospheric COz (-7%0 ~13C) and dissolved
Garnier 1979). (4) Evaporative outgassing of carbonate would result in precipitating carbon-
CO2 would increase 6 ~3CHco, of remaining ates with ~13C around +3%0 (~]3CaCO~-CO2
dissolved carbonate (Emrich etal. 1970). 1.0102 at 20~ Emrich et al. 1970). This is
The net result of these processes is that considerably higher than observed 613C values
carbonate precipitating from these solutions will and suggests that exchange with atmospheric
have 613Ccaco3values between - 11%o and -4%0, CO2 was not a dominant process in the forma-
or higher if CO2 loss occurs. Exchange between tion of early calcretes in this basin. This range of
 Carbonate cement diagenesis, U K 18 I

TABLE 2. 18O/16O, 2 H~1H and 13C/12 C analyses of brines from the Sherwood Sandstone reservoir in the
Wessex Basin
~518O ~2H
13C Salinity
Source of sample % vs SMOW %o vs PDB (mg 1-1 C1)

Winterborne Kingston (Triassic Drill Stem Test) -2.7 -24 180,000

Dorset (Triassic DST) -2.5 -30 160,000
Marchwood (Triassic pump-test) - 3.1 -33 - 15.8 63.000
Western Esplanade (Triassic pump-test) -2.2 -35 - 15.5 76.000
Dorset (Triassic drillcore water) -3.9 -34 59,000
Dorset (Triassic DST) -1.2 -25 170,000

possible isotopic behaviours in evaporative car- Fig. 8). Dolomitization within a sabkha or playa
bonate formation is illustrated by examples of environment is inferred to have taken place here.
modern calcretes (Talma & Netterberg 1983, The distinct isotopic composition suggests that
Salomons et al. 1978) and by active sabkha the aridity and groundwater discharge typical of
dolomitization in the Arabian Gulf (McKenzie continental sabkha conditions were developed
1981). only at the centre of this basin during the
The one anomalous composition of outcrop Triassic.
material (sample 3; -7.4%o ~3C and -9.5%0 Those samples (11, 13, 14) containing signifi-
~ 8 0 ) is of calcite coating a fracture surface in cant anhydrite and predominantly ferroan dolo-
the Pebble Beds and suggests precipitation in mite have a dramatically different 6lso ( - 10 to
different groundwater conditions. Its petrologi- - 12%o) but similar ~13C ( - 6 . 7 to -6.8%0). This
cal identification as a fracture-filling calcite lighter ferroan dolomite is replacive and must be
cutting across the dominantly early calcite- late stage, since the light ~180 values indicate
cemented Pebble Beds and the isotopic similarity formation at significant burial depths near the
to calcite cements from depth in the basin present maximum (H in Fig. 8). The remaining
suggest that this might have precipitated from sample (12) has an intermediate composition,
groundwater expelled towards the basin rim reflecting the mixture of early calcite and later
during burial and compaction. The trend of dolomite which is shown by petrography.
isotopic compositions (B to C in Fig. 8) in the It is concluded from the combined petrological
bulk of calcite cements at outcrop might be due and isotopic evidence that zones having high
to minor components with this origin. original contents of calcrete-type carbonates
were not penetrated by the fluids responsible for
anhydrite cementation. The anhydrite was possi-
Winterborne Kingston bly derived during burial from evaporites in the
This represents the thickest sequence in the overlying Mercia Mudstone or from syndeposi-
basin depocentre (Fig. 3). The cementing phases tional anhydrite near the basin depocentre--the
here are both anhydrite and carbonate, which is sulphur and oxygen isotope data are presented
predominantly replacive dolomite (Strong & later. Thus the early calcrete distribution is
Milodowski, this volume). Samples 15, 16 and 17, thought to have controlled subsequent diagenesis
which have very high (50-80~) contents of and permeability: the zones with sparse early
dolomite, are from the base of the Sherwood cementation allowed sulphate solutions and Mg-
Sandstone and have little or no anhydrite; they rich dolomitizing solutions to penetrate
have distinctively heavy 3180 (0-1%0)with ~ 13C subsequently.
values between - 9 and -7%0. The high 6180
indicates that the dolomite was precipitated Marchwood and Western Esplanade
from freshwater-derived solutions which had
been isotopically fractionated due to evapora- The present data cover a wider range than those
tion. It is thought unlikely that the 3180 could previously reported for Marchwood by Knox et
have been due to seawater, although the presence al. (1984), which are included in Fig. 8. A narrow
of marine brines locally during late Triassic range of values for calcrete calcite 3180 (of about
deposition of the overlying Mercia Mudstones -5%0 and ~13C from - 9 to -8%0) probably
cannot be definitely excluded. However the light represents the evaporation and exchange process
values for ~3C indicate that this was controlled already described (D in Fig. 8). Higher 3180 and
by the carbon isotopic composition of pre- 613C values (about - 2%0 and - 5%0respectively)
existing calcite--probably calcrete similar to which are found in analyses (19 and 20) from a
that found elsewhere in the basin (i.e. D to F in hand-picked calcite clast ('cornstone') and the
I82 A . H . B a t h et al.

corresponding bulk sample must indicate relati- cements in the Marchwood and Southampton
vely rare early post-depositional conditions with samples. The difference might reflect relative
isotopically heavier (3180 from - 3 to -4%0) differences in the contributions of different
water. Enrichment of 3180 and 313C with this inorganic carbon components (i.e. atmospheric
trend during evaporative precipitation of calcite and soil zone CO2, dissolved detrital carbonate)
has been observed in some modern calcretes at the time of formation. It is also possible that
(Salomons et al. 1978). this difference indicates a minor isotopically
The calcrete compositions occur as an end- heavier CO2 contribution resulting from bacte-
member of a mixing series of isotopic composi- rial fermentation of organic matter. The appar-
tions (i.e. D to G in Fig. 8), in which the samples ent absence here of carbonate cements with
might be situated according to their proportional lower 3180, and hence representing deeper
contents of the early eodiagenetic calcrete and burial stages, could be explained by the 'block-
the later mesodiagenetic calcite (see Strong & ing' action of hydrocarbons which were generat-
Milodowski, this volume). These positions rough- ed in Jurassic strata and migrated into the
ly correspond, in most cases, to the proportions Sherwood reservoir in the Wytch Farm area
estimated according to the contrast between the during the later Jurassic/early Cretaceous (Colter
nodular calcrete and the pore-filling luminescent & Havard 1981).
later calcite (Appendix 1). The samples dominat- Wytch Farm drillcore, like that at Winter-
ed by this late calcite (samples 18, 24, 28, 31, 33) borne Kingston, is characterized by anhydrite
have 3t80 mostly around -10%o and t~13C cement, up to 15% The two samples (38, 36) with
around -6%0; sample 24 from Marchwood has lowest anhydrite contents also have slightly
the lowest 3180 (-13%o). It is concluded that lower 3180 values ( - 6 . 4 and -6.7%0); the low
these calcites have precipitated at burial tem- abundance of early anhydrite cement in these
peratures up to 70~ if a basinal brine with zones might have permitted groundwater pene-
similar isotopic composition to the present tration and calcite cementation at a slightly later
(3t80,2o~-3%0 vs SMOW; Table 2) is as- stage of burial than elsewhere. In summary,
sumed. This is close to the present-day tempera- therefore, the sequence of diagenetic processes
ture at this depth in the basin, which according inferred in this part of the basin is (i) early post-
to burial curves in Milodowski et al. (1986) could depositional calcite cementation possibly with a
have been reached from the late Cretaceous to minor contribution from organic breakdown in
the present. adjacent strata, followed by (ii) anhydrite cemen-
The composition of calcite which would be tation which limited both groundwater flows and
in isotopic equilibrium with this present- further carbonate cementation, followed by (iii)
day Marchwood brine at the formation tempera- migration of hydrocarbons into the formation
ture of 70~ has been calculated, using the which prevented the deposition of the later,
values in Table 2 (6~80,2o ~ - 3%0vs SMOW and deep-burial, calcite cements found elsewhere in
313CHco3 ~~ " - - 16%o), #s CaCO3_H20~ 1.021 and the basin.
#3 C a C O 3 - H C O 3 ~"~
1.0036 (Emrich et al. 1970) at
70~ This calculation gives 318C a C O 3 ~ "~
- 12%o vs East Yorkshire-Lincolnshire Basin
PDB a n d 313caco3~ - - 12%0. This calculated 313C Dolomite is the dominant carbonate cement in
is very different to the observed calcite values, all samples studied from this basin (Milodowski
though 3180 is similar to some late-stage calcite et al. 1987). This contrasts with the Wessex Basin
cements (Fig. 8), therefore it is tentatively in which calcite was the dominant carbonate
concluded that the observed data show no with the exception of samples from Winterborne
evidence of present-day precipitation of cements Kingston. Two suites of samples were studied
in the deep Wessex Basin. There is no immediate from outcrop (including a shallow driUcore from
explanation for the apparent change in 3~3CHco3 Wistow), and from the well at Cleethorpes.
values between the solution which precipitated (Appendix 2). The stable isotope analytical
late-stage calcites and the present-day brine. results are shown in Table 3, and are illustrated
in Fig. 9.
Wytch Farm
Triassic sandstone outcrop in South Yorkshire-
These drillcore samples contain both anhydrite
Nottinghamshire
and calcite cements--cathodoluminescence sug-
gests mixtures of late and early calcite genera- The four outcrop samples and five samples from
tions (Appendix 1). Three samples (34, 35, 37) Wistow drillcore show a range of composition
have similar 3180 but slightly higher 313C values from - 3.6 to - 1.1~o 3 j 3C and - 5.9 to - 2.4~o
(E in Fig. 8) compared with the early calcite 3180. They fall close to the range found in a
 C a r b o n a t e c e m e n t diagenesis, UK I83
TABLE 3. 13C/12C and 1 8 0 / 1 6 0 analyses of total because the dolomitization takes place without
carbonate content of samples from the Permo- the necessity of additional CO2 influx:
Triassic sandstones in the East Yorkshire-
Lincolnshire Basin 2CACO3 + Mg 2+ ~CaMg(CO3)2 + Ca 2+
•13 C (~18O
whereas calcite dissolution and recrystallization
is driven by change in CO2 activity (with
No. Sample number %o vs PDB resultant influence on 613C due to isotopic
exchange with CO2):
Cleethorpes drill core (Triassic Sherwood Sandstone)
1 CL(SH)I (bulk sample) +0.03 -4.36 C a C O 3 + C O 2 -t- H20 ~ C a 2+ -k- 2 H C O 3 - .
2 CL(SH)I (calcite nodules) +0.04 - 12.79
3 CL(SH)6 +0.19 -4.22 The 6180 values for the early post-depositional
4 CL(SH)7 +0.34 -3.38 dolomite fall in a slightly higher range than those
5 CL(SH)13 +0.28 -4.15 for calcrete calcites in the Wessex Basin. This
6 CL(SH)20 -0.61 -4.06
difference might be due to the variation in
Cleethorpes drillcore (Basal Permian Sands) oxygen isotope fractionations between calcite or
7 CL(BP)3 - 1.23 -4.43
8 CL(BP)4 + 0.05 - 2.58 dolomite and solution (around 2%0; Fritz &
9 CL(BP)9 - 2.34 - 2.07 Smith 1970). It is also possible that this differ-
10 CL(BP) 11 - 3.00 - 2.20 ence indicates palaegeographical effects or vari-
Wistow drillcore (Sherwood Sandstone) ably higher extents of evaporative enrichment of
11 WSS-2 - 3.59 - 4.49 6~8OH,o in water recharging the early sub-surface
12 WSS-3 - 3.09 - 4.26 of the East Yorkshire-Lincolnshire Basin. The
13 WSS-7 -2.78 -3.85
14 WSS-9 -2.95 -4.04 surface conditions might have been similar to
15 WSS-14 -2.77 -4.45 those observed in modern continental sabkhas in
Outcropping Permo-Triassic sandstone in South Yorkshire which dolomite precipitates have high 613C and
16 789-1V -2.15 -2.35 6180 values (McKenzie 1981). However it is
17 178-1HXl -2.96 -5.91 doubtful whether this surface phenomenon alone
18 1207-1V - 1.08 - 2.90 could account for pervasive dolomitization
19 526-H121 -3.29 -5.15
throughout a sequence on this scale (Machel &
Mountjoy 1986).
Although isotopic analyses of separated calcite
previous study of Sherwood Sandstone drillcore have not been possible, their approximate
samples from shallow burial depth at Gamston in isotopic compositions are inferred from the trend
Nottinghamshire (Edmunds et al. 1982; also of data in Fig. 9 (B to D). The estimated t~13Cand
shown for comparison in Fig. 9). The isotopic, 6180 values correlate with values which repre-
chemical and mineralogical information on the sent isotopic equilibration with present-day
Gamston samples and co-existing groundwater groundwater which has 6~8OH2o ~-8.5%0 vs
was interpreted as indicating that the carbonate SMOW and t,,s~3CHCO3,'~ ~ - 13 to - 10%o (Table 4;
comprises isotopically heavy dolomite with Bath et al. 1979, Edmunds et al. 1982). This tends
minor amounts of isotopically light calcite which to confirm that these calcites are precipitating
is actively precipitating as a product of dedolo- ('dedolomite calcite') actively in the present
mitization (Edmunds et al. 1982, Evans et al. groundwater regime by the reaction:
1984, Bath et al. in press). The present outcrop
samples similarly have a spread of isotopic CaMg(CO3)2 + CO2 + H20
compositions representing mixtures between >.CaCO3 + Mg 2+ + 2HCO3-.
dolomite and subordinate calcite (positively
i d e n t i f i e d in 178-1HX1 and WSS-2
Cleethorpes : Sherwood Sandstone
and traces in others).
Petrographic observations show that the dolo- These carbonates are recognized petrographical-
mite here is an early-stage cement, in some cases ly as predominantly early replacive dolomite,
replacing pre-existing clasts (Appendix 2; Milo- and have higher 613C than the outcrop samples
dowski et al. 1987). Isotopic data suggest that the (Fig. 9). The source of these higher values is not
dolomite formed soon after deposition, probably clear--possibly different extents of atmospheric
in an environment influenced by Mg-enriched equilibration, or different early diagenetic condi-
groundwater evolving from nearby Permian tions affecting the carbonate system (A to C in
Magnesian Limestone. These early dolomite Fig. 9). In marked contrast, a nodule of late
cements have inherited the 613C of precursor calcite-cemented sandstone extracted from
detrital (?Carboniferous) limestones (>0%0 CL(SH)I has a very light 6180 value of - 12.8%o
613C) more directly than the bulk of the early which represents isotopic equilibrium with pre-
calcite cements in the Wessex Basin. This may be sent groundwater (-6.9%o; Table 4) at the
I84 A. H. Bath et al.
?A

5'3C%~ ~i~
-7
I
-6
i
-5
l
-4
I
-3 -2
1
-I,.C.
L.,.I.;,;.,,.I/ / / / / ~ +~
..... :-..,........-.-.~
l.'.'.'&i ...... :-:.>:.:.."
s( - " ' . ' . ' . ' . ' . ' - ' . ' . ~ , . . . . . . . .-.
/ " L ". . . . -. .>. > : ' > : 2I ~: : : i :. :. :. . . . ;. . . . . . - " : ' : ' : ' >
~.:7~:::: ::::. =================' =
~.....
" ......
========:..==
...................... ... .... ..........~;~ -

110 :: :::. 9 :.: : :: : : : ~ : . : S

.., ' "..-.'.'-'- 9
: : : : :~
..... ..... . . . . . . . . . ...f

., .... :-:.:.:.: : :-.',~,= -: :.: :/~--:-:.:.i! ~-

/<:::i:i:i::::::::::::::::::::::::::::::::::::::::::::::::::
/.: . . .... .:':-:..'.'---.-
,~ ........ ~ l w , , , . . . . . . . . . . . ...-~
...... .-:::':-'-.:.:.--:,~ 9 ~:
/ - .... :-:~:'""::'1112-:.-":'::::'::-. ~ 9 [:
A ..... "~11::'...-
...............:.:.:.: : . . .~15 :':':':-.']
. . . . :.:.:.:.:.'~ \~:

./" " ' : b~s ' :::::7

V~ ?D A j ///01;/ -6

5180
~ ? D Inferred compositions
%oPDB
' / / / / / / / / / / / / / / / / / / ~ ;fa Ima~ ? t I~;ll/Some

Basal Permian Sands

/ -10
Calculated compositions
of calcite in equilibrium
KEY with present-day brines
O Outcrop in S. Yorkshire N
-12
9 Wistow borehole
Sherwood Sandstone
9 Cleethorpes well: Sherwood Sst.
9 Cleethorpeswell: Basal Permian Sst.
-14
A Gamston borehole (Nottinghamshire outcrop)

FIG. 9. Plot of ~13C against 6180 for samples from the Sherwood Sandstone in the East Midlands and the
East Yorkshire-Lincolnshire Basin. Analyses of samples from the Basal Permian Sands in the Cleethorpes
drillcore are also shown. Stippling indicates ranges of dolomite compositions. The open triangles show
previously reported compositions of carbonate in drillcore samples from Gamston. Cross-hatched areas denote
inferred compositions of detrital calcite and of dedolomite calcites in outcrop and shallow buried sandstone;
the latter were estimated by progressive acid-leaching experiments on the Gamston samples (Edmunds et al.
1982, Evans et al. 1984). The arrows illustrate generalized isotopic evolution pathways. A---estimated
composition of detrital ?Carboniferous limestone; B and C-- ubiquitous early dolomitization of detrital
calcite, varying compositions in west (outcrop) and east (Cleethorpes) respectively; D--inferred compositions
of calcite due to late and present-day dedolomitization in shallow groundwater in west of basin; E--sparse
calcite due to late dedolomitization in deeply buried sandstones (Cleethorpes).

observed temperature of 50~ however, the Cleethorpes : Basal Permian S a n d s

613C of dissolved carbonate is not known, and
the relationship remains uncertain. It is possible The four samples analysed have scattered 313C
that this is a late-stage or active precipitate (C to values which fall across the field for other data
E in Fig. 9); the contrast in 613C with the from this area (Fig. 9). Two samples are
inferred composition of dedolomite calcite at dolomite-cemented conglomerates in which up
outcrop is thought to be a consequence of the to 20~ of the cement is a late-stage ankeritic rim.
depth and the absence of active circulation of However there is no direct correlation of anker-
meteoric water with isotopically-light dissolved ite proportions with isotopic compositions.
bicarbonate. Progressively less negative 613C Petrographic evidence suggests that these dolo-
values for dissolved bicarbonate have been mites and ankerites formed at different stages in
observed in groundwater down-gradient in the the burial history, and this could account for the
East Midlands Sherwood Sandstone aquifer range in (~13C. They have similar or slightly
(Bath et al. 1979; Edmunds et al. 1982). higher fi~80 values relative to the Sherwood
 Carbonate cement diagenesis, UK I85
TABLE 4. 180/160, 2H/1H and 13C/12Canalyses of groundwaters in Permo-
Triassic aquifers in the East Midlands and the East Yorkshire-Lincolnshire
Basin (Bath et al. 1979, Downing et al. 1985, Parker et al. 1986)
6180 6:H
613C HCO3 Salinity
Source of sample %ovs SMOW %ovs PDB (mg 1-1 CI)
Modern recharge to Notts -8.6 to -7.7 -58 to -50 -13.5 to -10.3 18 to 174"
Sherwood Sst aquifer
Late Pleistocenewater in -9.8 to -9.1 -66 to -57 -10.6 to -8.5 6 to 21
Notts/Lincs SherwoodSst
Modern recharge to S Yorks -8.2 to -8.0 -56 to -51 -11.4 18 to 120"
Sherwood Sst aquifer
Recent recharge in confined -7.8 to -7.1 -57 to -46 -17 9 to 22
SherwoodSst in S Yorks
Sherwood Sandstoneaquifer -6.9 -51 n/a 45200
in CleethorpesWell
Basal Permian Sands aquifer +0.4 -12 n/a 133 000
in CleethorpesWell

* Higher C1 values here attributed to contaminantsources affectingrecharge.

TABLE 5. 34S/32Sand 180/160 analyses of sulphate mineral phases in the Triassic

sandstones in the Wessex Basin
Sample number 634S di180 Notes
Winterborne Kingston drillcore
WK11 +7.85 + 14.53/+ 14.88 c. 5-10% anhydrite
Wytch Farm X14 drillcore
E59542 +9.30/+9.80 + 15.43/+ 15.54 c. 10-15%anhydrite
E59546 + 10.84 + 15.35 c. 5-10% anhydrite
t534Svalues are in %0variation with respect to the Canyon Diablo meteoritetroilite standard, CD.
6180 values are in ~ooversus SMOW.

Sandstone dolomites, which is noteworthy in well has already been reported: +7.5 to +8.7%0
view of the sharply contrasted values for 6180 in 6345 and + 11.6%o 6180 (Bath & Darling 1981).
present-day brines in the two formations (Table As expected, the sulphur isotope compositions
4: Downing et al. 1985). This suggests that these suggest that the dissolved sulphate is derived
early dolomites precipitated in broadly similar from the solid anhydrite.
environments, since then the two formation The 634S values are lower than those found in
brines have evolved independently. marine sulphate deposits through the stratigra-
phical record (Claypool et al. 1980). The mini-
mum of 634S in evaporites are in Permian
Stable isotopic composition of Zechstein sulphates, for which values are around
anhydrite cement + 10%o (Claypool et al. 1980, Taylor 1983). It is
suggested that the low 634S values of anhydrite in
Three 348/32S and 180/16 O analyses were carried Sherwood Sandstone indicate that the sulphate
out on the diagenetic anhydrite occurring in the was originally derived by remobilization of
Sherwood Sandstone at Winterborne Kingston Zechstein sulphate. However, the Wessex Basin
and Wytch Farm in the Wessex Basin. The has no Zechstein deposits in the Permo-Triassic
objective was to investigate the source and sequence, therefore a remote source with trans-
diagenesis of the anhydrite. The results are given port by surface- and/or groundwater into the
in Table 5; these are in the range of +7.8 to basin is inferred. Alternatively the anhydrite
+ 10.8%o 634S and + 14.5 to + 15.5%o 6180. The may have been produced from a mixture of
isotopic composition of dissolved sulphate marine sulphate and oxidized sulphides possibly
(1400 mg 1-1) in the brine from the Marchwood from the overlying Mercia Mudstone. Taylor
I86 A. H. B a t h et al.

(1983) has interpreted 33"S data for Mercia with relatively high 3180 indicates an evaporitic
Mudstone sulphate horizons in the East Mid- environment.
lands as indicating locally varying proportions of The isotopic compositions of calcite cements
marine and continental brine inputs to the late in a drillcore from the Wytch Farm Sherwood
Triassic depositional environment in that area. Sandstone oil reservoir differ slightly from those
The 3180 values of the anhydrite (c. + 159/o at other sites; this is attributed to processes
3180) are significantly heavier than the Zech- involving the inhibition of groundwater flux in
stein minimum (c. + 10%o) in the 6lSOso' time- the formation, first by early anhydrite cementa-
curve (Claypool et al. 1980). This might also tion and later by hydrocarbon influx.
indicate that the anhydrite contains some non- The restricted range of 613C values suggests
Zechstein sulphate; alternatively it could indi- that the carbonate system in the Wessex Basin
cate that there had been some oxygen-isotope has remained virtually closed over its burial
exchange between water and dissolved sulphate history since the early formation of calcretes.
resulting in an enrichment of 180 in the sulphate. The later carbonate cements have been derived
The rate at which this exchange might take place by remobilization of the pre-existing cements
under these conditions is uncertain as it depends with little change in 613C.
on pH as well as temperature. Present evidence Permo-Triassic sandstone samples from the
suggests this would be very slow even at 70~ at East Yorkshire-Lincolnshire Basin are dominat-
which the equilibrium isotopic fractionation ed by early dolomite precipitation. These dolo-
~tSso,_,2o is about 1.021 (Chiba & Sakai 1985). mites are characterized by relatively heavy 313C
However, isotope exchange seems unlikely since and 3180 ratios. This indicates early dolomitiza-
the anhydrite precipitation is attributed to an tion of pre-existing detrital calcite in sabkha
earlier stage of burial diagenesis when tempera- conditions, possibly associated with a wide-
tures and therefore rates of exchange would have spread flux of Mg-enriched groundwater into the
been lower. basin coupled with arid conditions as ground-
water discharged through the low-relief surface.
The difference in 313C values between the two
basins may reflect differences in the carbon
Conclusions chemistries of the two reactions: calcite recrys-
tallization and dolomitization. The resultant
Inevitably the models to interpret the stable 313C value depends both on the composition of
isotopic evolution of the meteoric water diagene- precursor carbonate and also on the contribution
tic environments are somewhat speculative. of external CO2 (atmospheric or soil zone) to the
They are inferences based on evidence from reaction.
modern meteoric water aquifers and evaporite A further difference between the two basins is
environments (calcretes and sabkhas). However found in the effective degrees and compositions
the scale of the processes, both in time and in of later carbonate cementation. The petrological
space in the Triassic rocks, and the uncertain and isotopic evidence indicates that the Wessex
detail of the Triassic palaeoenvironment intro- Basin has abundant calcite precipitated at
duce much uncertainty into these comparisons. various stages of burial up to the maximum: the
The Sherwood Sandstone in the Wessex Basin isotopic compositions form a series between the
is cemented predominantly with calcites whose early and deep-burial calcites (which, however,
isotopic compositions suggest precipitation in a are not in carbon isotopic equilibrium with
range of burial environments. This range extends present formation fluid). This apparently contin-
from early post-depositional calcrete formation uous process of precipitation in a closed carbon-
in freshwater to later deep-burial calcite recrys- ate system might be driven simply by the
tallization in basinal brines. The carbonate for pressure and temperature changes and compac-
early calcretes was probably derived by dissolu- tion during burial.
tion of detrital Carboniferous or Devonian The only isotopic evidence for carbonate
limestone fragments; the 313C values also indi- reaction after early dolomite cementation in the
cate a minor contribution from soil-zone organic- East Yorkshire-Lincolnshire Basin is that for a
derived CO2 except in calcite found in the progressive and still active dedolomitization due
present outcrop for which the data suggest to the influx of fresh groundwater through the
recrystallization without CO2 exchange. This outcropping rim of the basin, bringing with it
outcrop of the Wessex Basin marginal facies has isotopically light carbon from superficial
abundant recrystallized detrital carbonate which sources. This zone is therefore subjected at
is still being dissolved in the present groundwater present to enhanced dedolomitization, and any
regime. An exception to calcite cementation is calcite precipitated has a characteristically light
found in the basin's depocentre, where dolomite carbon and oxygen isotopic composition. The
 Carbonate cement diagenesis, UK I87
isotopic evidence for this process in the deeper ing the deepest burial conditions suggests that
basin--only one isotopically light calcite positi- the influx of hydrocarbon prior to the deepest
vely identified--is limited, although there is burial stage somehow inhibited further calcite
petrological evidence for sparse late calcite. The precipitation in both oil and water zones.
groundwater flux into the deeper basin over its
geological history would have been limited.
ACKNOWLEDGMENTS: This study was carried out
However, the light isotopic composition of the within the UK Geothermal Energy Assessment Pro-
present brine indicates a direct influence of ject, funded by Department of Energy, project man-
meteoric water here, in contrast to the isotopical- ager R. A. Downing. The assistance of G. Strong and S.
ly heavy evolved brines in the Wessex Basin. Holloway in the petrographical study and the structural
Finally, there is no substantial evidence that interpretation respectively is acknowledged. C. Branch
the organic reactions preceding hydrocarbon carried out C and O isotope analyses; M. E. Patton
migration into the Sherwood reservoir at Wytch (Department of Earth Sciences, University of Water-
Farm have greatly influenced isotopic composi- 1oo, Canada) carried out S and O isotope analyses. I.
George assisted with sample preparation. The paper
tions of carbonate cements. These carbonates
benefitted from reviews by Charles Curtis, Greg
have (~13C values slightly higher than elsewhere, Samways and Jim Marshall. The manuscript was typed
but the difference is small and its significance is by P. Bath and V. Jones. This paper is published with
not clear. However, the absence of late diagene- permission from the Director, British Geological
tic calcites with isotopic compositions represent- Survey (NERC).

References

BATH, A. H. & DARLING,W. G. 1981. Stable isotope Geothermal Potential of the U.K. British Geological
analyses of water, carbonate and sulphate from Survey, Keyworth.
the pump-test of Marchwood No. 1 borehole. - - & GRAY, D. A. (eds). 1986. Geothermal energy--
Stable Isotope Technical Report No. 9, British the potential in the United Kingdom. Report,
Geological Survey (unpublished). British Geological Survey. HMSO, London.
--, EDMUNDS, W. M. & ANDREWS, J. N. 1979. EDMONDS, E. A. & WILLIAMS,B. J. 1985. The geology
Palaeoclimatic trends deduced from the hydroche- of the country around Taunton and the Quantock
mistry of a Triassic sandstone aquifer, United Hills. Memoirs, British Geological Survey, Sheet
Kingdom. In: Isotope Hydrology 1978, pp. 545-68. 295.
IAEA, Vienna. EDMUNDS, W. M., BATH, A. H. & MILES,D. L. 1982.
- - , MILODOWSKI,A. E. & STRONG,G. E. In press. Hydrochemical evolution in the East Midlands
Fluid flow and diagenesis in the East Midlands Triassic sandstone aquifer, England. Geochimica
Triassic sandstone aquifer. In: GOFF, J. C. & et Cosmochimica Acta, 46, 2069-82.
WILLIAMS,B. P. J. (eds). FluidFlow in Sedimentary EMRICH, K., EHHALT, D. H. & VOGEL, J. C. 1970.
Basins and Aquifers. Special Publication of the Carbon isotope fractionation during the precipita-
Geological Society of London. Blackwell Scienti- tion of calcium carbonate. Earth and Planetary
fic Publications, Oxford. Science Letters, 8, 363-71.
BURLEY, S. D. 1984. Patterns of diagenesis in Sher- EVANS, G. V., OTLET,R. L., WASSELL,L. L. & BATH,
wood Sandstone Group (Triassic), United King- A. H. 1984. Verification of the presence of carbon-
dom. Clay Minerals, 19, 403-40. 14 in secondary carbonates within a sandstone
CHINA, H. & SAKAI,H. 1985. Oxygen isotope exchange aquifer and its hydrological implications. In:
rate between dissolved sulfate and water at Isotope Hydrology 1983, pp. 577-90. IAEA,
hydrothermal temperatures. Geochimica et Cosmo- Vienna.
chimica Acta, 49, 993-1000. FONTES, J.-C. & GARNIER, J.-M. 1979. Determination
CLAYPOOL, G. E., HOLSER, W. T., KAPLAN, I. R., of the initial 14C activity of the total dissolved
SAKAI,H. • ZAK,I. 1980. The age curves of sulfur carbon: a review of the existing models and a new
and oxygen isotopes in marine sulphate and their approach. Water Resources Research, 15, 399-413.
mutual interpretation. Chemical Geology, 28, 199- FRIEDMAN, I. & O'NEIL, J. R. 1977. Compilation of
260. stable isotope fractionation factors of geochemical
COLTER,V. S. & HAVARD,D. J. 1981. The Wytch Farm interest. In: FLEISCHER,M. (ed.) Data of Geochem-
oilfield, Dorset. In: Petroleum Geology of the istry, Chapter KK. US Geological Survey Profes-
Continental Shelf of North West Europe, pp. 494- sional Paper 440-KK.
503. Institute of Petroleum, London. FRITZ, P. & SMITH,D. 1970. The isotopic composition
DOWNING, R. A., ALLEN, D. J., BIRD, M. J., GALE, of secondary dolomites. Geochimica et Cosmochi-
I. N., KAY, R. L. F. & SMITH, I. F. 1985. mica Acta, 34, 1161-73.
Cleethorpes No 1 Geothermal Well--a prelimi- HENSON, M. R. 1970. The Triassic rocks of South
nary assessment of the resource. Investigation of the Devon. Proceedings of the Ussher Society, 2, 172-7.
I88 A. H. Bath et al.
KNOX, R. W. O'B., BURGESS,W. G., WILSON, K. S. & PARKER, J. M., FOSTER, S. S. D., SHERRATT, R. &
BATH, A. H. 1984. Diagenetic influences on ALDRICK, J. 1985. Diffuse pollution and ground-
reservoir properties of the Sherwood Sandstone water quality of the Triassic sandstone aquifer in
(Triassic) in the Marchwood geothermal borehole, southern Yorkshire. Report Series, British Geologi-
Southampton, England. Clay Minerals, 19, 441- cal Survey, Vol. 17, No. 5, HMSO, London.
56. SALOMONS, W., GOUDIE, A. & MOOK, W. G. 1978.
LOTT, G. K. & STRONG, G. E. 1982. The petrology and Isotopic composition of calcrete deposits from
petrography of the Sherwood Sandstone (?Middle Europe, Africa and India. Earth Surface Processes,
Triassic) of the Winterborne Kingston borehole, 3, 43-57.
Dorset. In: RHYS, G. H., LOTI, G. K. & CALVER, SHARMA,T. & CLAYTON,R. lq. 1965. Measurements of
M. A. (eds). The Winterborne Kingston Borehole, O18/O TM ratio of total oxygen carbonates. Geochi-
Dorset, England, pp. 135-42. Report Series, 81/3. mica et Cosmochimica Acta, 29, 1347-53.
Institute of Geological Sciences, London. SMITH, A. N. & EPSTEIN, S. 1971. Two categories of
MCCREA, J. M. 1950. On the isotopic chemistry of 13C/12C ratios for higher plants. Plant Physiology,
carbonates and a paleotemperature scale. Journal 47, 380-4.
of Chemical Physics, 18, 849-57. STRONG, G. E. & MILODOWSKLA. E. 1987. Aspects of
MCKENZlE, J. A. 1981. Holocene dolomitization of the diagenesis of the Sherwood Sandstone of the
calcium carbonate sediments from the coastal Wessex Basin. This volume.
sabkhas of Abu Dhabi, U.A.E. : a stable isotope TALMA, A. S. & NEll"ERBERG, F. 1983. Stable isotope
study. Journal of Geology, 89, 185-98. abundances in calcretes. In: WXLSON, R. C. L.
MACHEL, H.-G. & MOUNTJOY,E. W. 1986. Chemistry (ed.). Residual Deposits, pp. 221-33. Special
and environments of dolomitization--a reapprais- Publication of the Geological Society of London,
al. Sedimentology, 23, 175-222. 11. Blackwell Scientific Publications, Oxford.
MILODOWSKI, A. E., STRONG, G. E., WILSON, K. S., TAYLOR, S. R. 1983. A stable isotope study of the
ALLEN, D. J., HOLLOWAY,S. & BATH,A. H. 1986. Mercia Mudstones (Keuper Marl) and associated
Diagenetic influences on the aquifer properties of sulphate horizons in the English Midlands. Sedi-
the Sherwood Sandstone in the Wessex Basin. mentology, 30, 11-31.
Investigation Geothermal Potential U.K. British WALTON, N. R. G. 1982. A detailed hydrogeochemical
Geological Survey. study of groundwaters in the southwest England
, , , HOLLOWAY,S. & BATH, A. H. 1987. Triassic Sandstone aquifer. Report Series, 81/5.
Diagenetic influences on the aquifer properties of Institute of Geological Sciences, London.
the Permo-Triassic sandstones in the East WIGLEY, T. M. L., PLUMMER,L. N. & PEARSON, F. J.
Yorkshire-Lincolnshire Basin. Investigation 1978. Mass transfer and carbon isotope evolution
Geothermal Potential U.K. British Geological in natural water systems. Geochimica et Cosmochi-
Survey. mica Acta, 42, 1117-40.

A. H. BATH & A. E. MILODOWSKI, Fluid Processes Reserach Group, British Geological

Survey, Keyworth, Nottingham NG12 5GG, UK.
B. SPIRO, Mineral Science and Isotope Geology Research Group, British Geological Survey,
64 Gray's Inn Road, London WCIX 8NG, UK.

Appendix 1
Descriptive notes on carbonate cements in samples of Triassic sandstones from the Wessex Basin
(estimates of carbonate contents by cathodoluminescence microscopy)

No. Sample Notes (including National Grid Reference)

Western outcrop
1, 2 HW03 Holywell Quarry, Milverton (ST 1269 2703); Pebble Beds conglomerate with limestone
pebbles, zoned early calcite cements
3 MV01 Milverton (ST 1160 2465); Pebble Beds; calcite on fracture surface
4 MV01 Milverton; Pebble Beds; c. 35% calcite, virtually all early calcite with later calcite filling
rims
5 WS02 Wolston Quarry (ST 0945 4015); Pebble Beds; pure vein calcite with two zones: earlier is
brecciated followed by later luminescent calcite
6 BL01 Bishops Lydyard (ST 1699 3021); Otter Sandstone; c. 30% calcite, some pellets;
early : later calcite ~ 50 : 50
7 LB01 Ladram Bay (ST 0975 8520); Otter Sandstone; highly calcareous concretion nodule, major
part is luminescent calcite
8 LB02 Ladram Bay; Otter Sandstone; c. 25% calcite, virtually all non-luminescent calcite with
occasional rims of brighter calcite
 Carbonate cement diagenesis, UK I89
9 LB03 Ladram Bay; Otter Sandstone; calcrete sandstone, c. 20-25~ calcite mostly later
later : early ~ 80 : 20
10 LBV01 Langford Budville (ST 1235 2218); Otter Sandstone; mostly calcrete and few pellets, rims
of later calcite; early : later ~ 80 : 20
Winterborne Kingston drillcore (S Y 8470 9796)
11 WK01 2316.3 m. Total carbonate c. 10% Carbonate pellets with minor later carbonate; mostly
replacive and pore-filling dolomite; < 5% anhydrite cement
12 WK06 2321.2 m c. 50-60~ carbonate; early detrital calcite cores (50~) surrounded by later
calcite (c. 30~) and replacive dolomite (c. 20~); < 5 ~ anhydrite cement
13 WK09 2324.7 m. c. 10~ total carbonate; mainly dolomite with minor later calcite; dol:
calc. ,~ 95 : 5; < 5~ anhydrite
14 WKll 2328.1 m. c. 20~ carb; micrite pellets with early calcite cement rims and abundant later
calcite partly replaced by dolomite; pellets :late-calcite : dolomite ~ 40:40 : 20; 5-10~
anhydrite
15 WK18 2428.2 m. c. 60~ carbonate, all dolomite replacing clasts and cement (50:50)
16 WK19 2428.7 m c. 50~ carbonate, all dolomite replacing clasts and cement
17 WK22 2434.4 m. Layered calcareous mudstone with total replacement by dolomite (70-80~ of
total)
Marchwood drillcore (SU 3991 1118)
(E samples from Upper Unit--feldspathic; MW samples from Lower Unit--lithic)
18 E59535 1684.9 m. c. 25~ luminescenet later calcite, very small amount of non-luminescent earlier
calcite
19, 20 E59536 1684.4 m. 25-30~ calcite, virtually all earlier calcite with some pellet calcite (80:20)
21 E59538 1690.5 m < 10~ of calcite, mostly earlier calcite
22 MW4 1696.3 m. 25-30~ calcite; multiple zones earlier calcite and later calcite (20:80)
23 MW9 1699.7 m. c. 30% calcite; dominant earlier and later calcites (80:20); fine-grained
24 MW19 1707.3 m. c. 10~ calcite, all later calcite; fairly coarse-grained
25 MW22 1709.6 m. c. 30~ calcite, mainly earlier plus later calcites (80:20)
26 MW23 1714.2 m. c. 5 ~ later calcite
27 MW24 1715.2 m. Abundant carbonate-cemented pellets, mainly earlier calcite cement
Western Esplanade drillcore (SU 4415 1120)
28 E59553 1735.2 m. Highly calcareous (c. 25% total calcite); mainly later calcite cement with pellets
and earlier cement (70 : 15 : 15)
29 E59557 1740.4 m. Carbonate clasts with stellate earlier non-luminescent calcite and later calcite
cement (50:50)
30 E59559 1740.9 m. Cornstone pellets with stellate early and later intergranular calcite, patchy;
25:35:40; c. 10% total carbonate
31 E59562 1752 m. c. 15% total calcite; mainly later calcite cement with earlier calcite and pellets
(80:20)
32 E59564 1754.1 m. c. 25% total calcite, well-cemented, pellets abundant and often stellate;
later: early calcite and pellets ,~ 50 : 50
33 E59565 1757.9 m. c. 25-30% total calcite, mostly earlier cement; early:later calcite ~ 70:30
Wytch Farm X14 drillcore (SY 9804 8526)
34 E59540 1598.4 m. Mostly anhydrite cemented (c. 10%) with sparse (c. 5%) calcite cement; mostly
later with early calcite and some pellets (80:20)
35 E59542 1601.3 m. c. 10-15% calcite and 10-15% anhydrite; mostly calcite pellets with some later
and earlier calcite; (70 : 25 : 5)
36 E59544 1611.4 m. 5-10% calcite and < 5% anhydrite cement; later calcite on clast cores (50:50)
37 E59546 1631.2 m. c. 10% total calcite and 5-10% anhydrite; clasts with clast calcite cores and
later calcite rims (60:40); very minor earlier calcite rims
38 E59551 1649.8 m. c. 10% total calcite; carbonate-cemented pellets, probably later calcite; no
detectable anhydrite
I90 A. H. Bath et al.

Appendix 2
Descriptive notes on carbonate cements in samples of Permo-Triassic sandstones from the East
Yorkshire-Lincolnshire Basin (estimates of carbonate contents by cathodoluminescence microscopy)

No. Sample Notes (including National Grid Reference)

Cleethorpes drillcore (TA 30237 07090); Sherwood Sandstone

1 CL(SH)IA 1111.9 m. 5-10% dolomite replacing ferruginous clasts and rare replacive late calcite;
patchy minor anhydrite
2 CL(SH)IB 1111.9 m. Nodule of late calcite-cemented sandstone from CL(SH)IA
3 CL(SH)6 1115.5 m. 5-10% replacive dolomite; patchy later calcite cement, 2% and anhydrite
4 CL(SH)7 1116.3 m. Silty sandstone with minor anhydrite/gypsum; 25-30% replacive and pore-
filling dolomite
5 CL(SH)13 1208.5 m. 10-15% replacive dolomite
6 CL(SH)20 1317.0 m. 10% replacive dolomite
Cleethorpes drillcore (TA 30237 07090); Basal Permian Sands
7 CL(BP)3 1869.7m. Mostly early non-ferroan dolomite (10%) overgrown by minor later ankerite
8 CL(BP)4 1870.7m. Abundant (20-40%) replacive and pore-filling dolomite cements with ankeritic
overgrowths
9 CL(BP)9 1884.0 m. Ferruginous siltstone with interstitial fine-grained dolomite, c. 10%.
10 CL(BP)I 1 1890.7 m. Similar to CL(BP)4; 10-15% vuggy replacive calcite
Wistow drillcore (SE 57433 35619); Sherwood Sandstone
11 WSS-2 39.7 m. c. 5% replacive dolomite; rarer later poikilotopic calcite (2-5%)
12 WSS-3 49.1 m. 10% replacive dolomite
13 WSS-7 75.8 m. 10% dolomite; patchy later calcite (c. 1%) often enclosing or replacing dolomite
14 WSS-9 83.8 m. 10-15% dolomite replacing clasts; localized later poikilotopic calcite
15 WSS-14 106.6m. Ferruginous siltstone with 10% interstitial dolomite rhombs
Outcropping Permo-Triassic sandstones
16 789-1V SK 606 898. Coarse siltstone, 10-15% interstitial dolomite and trace late calcite
17 178-1HX1 SK 7063 9972. c. 8% replacive dolomite, patchy late calcite c. 3%
18 1207-1V Se 6451 2432. c. 10% dolomite replacing detrital clasts
19 526-H121 SE 4583 5587. c. 5% dolomite and sporadic nodules of late calcite
 Six million year diagenetic history, North Coles Levee, San
Joaquin Basin, California
J. R. Boles

S U M M A RY : North Coles Levee is an uppermost Miocene turbidite sandstone reservoir in

the San Joaquin Basin. The sands have undergone progressive burial and a complex history
of siderite, dolomite and calcite cementation. Precipitation of calcite cement bands occurred
between 40 and 80~ with the last cementation event corresponding to about 2.5 Ma np.
Between this time and the present, extensive dissolution of plagioclase and calcite resulted
from influx of acid during kerogen maturation. Emplacement of 500 million barrels of oil
closely followed or was contemporaneous with these events. Very late stage compaction
effects including albitization of plagioclase and dolomite crystallization in crushed biotite
have resulted from cement removal and/or fluid pressure drops during release of the gas cap.

This paper is a brief synthesis of the diagenetic North Coles Levee

history of a late Tertiary oil field in California.
Results and discussions presented here are given The North Coles Levee (NCL) reservoir is a
in greater detail in earlier papers (Boles & series of deep marine turbidite sandstones
Johnson 1983, Boles 1984 and Ramseyer & Boles (Webb 1981). Foraminifera in surrounding mud-
1986) and papers in preparation (Boles & stones indicate the sandstones were deposited at
Ramseyer, in press). A significant aspect of the bathyal depths during the latest Mohonian stage
history is the relatively small time interval (Lagoe, in press). Thus the diagenetic history
(<2.5 Ma) which encompasses dissolution of described in this paper occurred within the past
framework grains and cements as well as 7 Ma.
hydrocarbon emplacement. These effects are The main producing interval, between 2.6 and
consistent with the idea that porosity enhance- 2.9 km, is a series of amalgamated medium to
ment in the reservoir results from chemical coarse sandstones containing an estimated 500
evolution of the porewaters during or slightly million barrels of oil in place at the time of
prior to hydrocarbon emplacement (Surdam et discovery. Average reservoir porosity is about
al. 1984). The short time interval for these events 15% based on core plug measurements. Average
is surprising considering the amount of mass permeability in core plugs is about 10 md with
transfer involved. values as high as 150 md in some zones. The
reservoir sands have quartz :feldspar ratios of
1 : 1 and plagioclase:K-feldspar ratios > 1 : 1
(Ramseyer & Boles 1986). Sand isopach maps
Geological setting and facies analyses indicate they were derived
from the Jurassic-age Sierran granodiorite
The San Joaquin Basin contains more than 7 km batholith, 6 0 k m to the NE (Webb 1981).
of largely upper Miocene and younger shale and Plagioclase/K-feldspar ratios in the sands are
sandstone (Fig. 1). The North Coles Levee oil similar to Sierran granodiorites although the
field is located in the central basin area where the sands are enriched in quartz relative to Sierran
section has subsided continuously since Upper sources, presumably due to weathering and
Miocene time. In this area the depositional transport processes.
sequence changes from deep marine facies in the The reservoir at 2.6 km burial depth originally
Upper Miocene to shallow marine and non- produced fluids at 285 bars pressure at 105~
marine facies in the Plio-Pleistocene (Fig. 2). Pressures today are maintained at higher levels
The basin has proven recoverable hydrocarbon due to water injection in the central part of the
reserves in excess of 7.8 billion barrels of oil and field (Fig. 3). Original formation water at the
11.2 trillion cubic feet of gas in more than 40 time of field discovery was NaC1 rich with total
producing fields (Callaway 1971). The young age dissolved solids somewhat less than in sea water
of the basin, relatively simple structural history, (Table 1). There is no evidence of meteoric water
and abundant subsurface data allows consider- influx into this part of the section and the lower
able control to be placed on timing and physical salinities of these pore waters relative to sea
conditions of diagenesis. water is attributed in part to clay dehydration

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, I9I

Geological Society Special Publication No. 36, pp. 191-200.
192. J. R. Boles
I \
\ \
/ \\
\
\\ ,// \ \
\
\
x \
~. ,
\ -,, \\ x, O
\ O
,,, ,,,
"\ \ f

\\ \
'x \
\ \~::,
\
,% 0 x

N \
%o (~ ',\
0 i

d~
% t o,\
! c
t ~ o ~ " 7" SIERRA
\ \ NEVADA
\ \ ,.
X \ \ ] i
\ \ NORTH CGLES LEVEE ~ ~ X
C <"<7<<

L \,\
I ,\

F O " "",.

R N -.
"x
I A "?

o 50
l L , A , I
KILOMETERS

FIG. 1. Location of San Joaquin Basin and North Coles Levee oil field. Dashed lines show Tertiary
sedimentary rock thickness (metres) from Callaway (1971). Cross-section X-X' shown in Fig. 2.
SIERRA
NEVADAS
~ ~
\'" M'/~~ ' ~ ~ ~ COLES
NORTH \ x'
--- " - %U"~..-~ ~ ~ V E E

o
-""-,..,,~e ~ f ////
I ////
I , /

-2.2,,1"2"
/ X' ~ . ~ JI Miocen e I
\ \ "-\\- .... , t
/# ++++++~
++++++ + .
,//t/ +++o+++~++
i +++++ + + +
\ \\ "..... , :..;;'." ,

" ~,// oGee ~

20 km IO k m
i i

500 m
/ o%,(-:.:.:."
IOOO m
:.:.:."
FIG. 2. Cross-section of southern San Joaquin Basin showing location of Steven sandstones in the North
Coles Levee oil field. Modified from California Oil and Gas Fields (1982 publication, California Division of
Oil and Gas).
 6 M a diagenetic history, North Coles Levee I93

Gas Cap Release Water Flood

Expansion Gas Injection of G a s C a p __'~.~
5 --
<St Cr Cr
Hydrostatic Pressure
_f_
C>

I
E3 3 CO 2
-- 0 Discovery /~850 psi Injection
,-I x of Field
IL -

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

_ Bubble Point Pressure

Z Or~ 2
0 n.
!-
<
:E 1

0
i,I.
0 I 1 1 1 I
1940 1950 1960 1970 1980

DATE
FIG. 3. Fluid pressure history o f Stevens sandstone (2591 m) at N o r t h Coles Levee. D a t a f r o m Arco,
Bakersfield.

reactions (Ramseyer & Boles 1986). An unusual provide a mechanism for moving low solubility
aspect of the water is that alkalinity is dominated components (e.g. Surdam et al. 1984, Crossey et
by organic acids rather than carbonic acid al. 1986).
species. The organic acids are believed to form
complexes with metals (e.g. AI, Si) and hence
Diagenetic history
TABLE 1. Comparison of seawater composition
with evolved porewater at North Coles Levee The following section briefly describes the
(mg /--1). Seawater composition from Drever diagenetic phases in Stevens sandstones in their
(1982). Porewater HC03, acetate and pro- order of occurrence. These phases have formed
pionate from R. C. Surdam Laboratory at
University of Wyoming. Remaining analysis Ca
by ARCO Research Laboratory
Well NCL 14-31
Sea water at 2641 m

Na + 10 760 9300
K+ 399 200
Ca ++ 411 790
Mg +§ 1290 21
B+3 4.5 120
Si +4 0.5-10 44
Fe <1 1.9
Mn <1 0.7
Sr 8 49
C1- 19 350 14 000
NO3-- n.d. 120
SO4-- 2710 35
HCO3- 142 96 )
Acetate n.d. 3730 Fe 5o Mg
Propionate n.d. 896
pH 8.2-8.4 6.94 FIG. 4. Siderite compositions (mole %) from N o r t h
TDS 35 000 29 403* Coles Levee. D a t a from m i c r o p r o b e analyses o f five
samples in well no. N C L 488-29 b e t w e e n 2727 a n d
* Other elements analysed but not reported here. 2756 m depth.
I94 J. R. Boles

FIG. 5. Dolomite rhomb (D) attached to detrital grain (G) and surrounded by stained calcite (C). Well no.
NCL 488-29, 2727 m.

during a single burial cycle at conditions ranging Dolomite rhombs attached to detrital grains
from those at the late Miocene marine sediment- appear to post-date the siderite and predate most
water interface to present burial depths. calcite cement (Fig. 5). Dolomite is usually a
Siderite is one of the earliest diagenetic phases minor cement in the reservoir but several
occurring as 10 pm yellowish crystals attached to intervals (less than 1 m thick) contain 15-30
detrital grains and as scattered crystals in pore volume %. The dolomite has from 5 to 15 mole %
spaces and frequently enclosed by later carbon- Fe substitution and like many Tertiary dolomites
ate cements. Microprobe analyses indicate is Ca rich with Ca :Fe + Mg > 1 : 1 (Fig. 6).
appreciable Mg substitution in the siderites (Fig. Calcite is the major pore-filling cement in the
4), similar to that observed in marine siderites reservoir and is concentrated into thin zones
from other areas (Matsumoto & Iijima 1981). (0.2-1 m thick) with up to 30% calcite. Where
occurring with dolomite, calcite surrounds the
Fe euhedral dolomite crystals implying calcite
formed after dolomite (Fig. 5). The calcite has
5O
Fe :Mg ratios of about 1:1 and from 1 to 10 mole
% Fe + Mn + Mg substitution for Ca (Fig. 7).
Detailed trace element and isotopic analyses of
these and other carbonates is presented in a
forthcoming paper where it is shown that the
high Fe and Mg content of calcites from NCL is
distinctive relative to calcite compositions from
other depositional settings in the San Joaquin
Basin (Boles & Ramseyer, in press).
Temperatures for NCL calcite precipitation

Ca
/,,. 75
,.-\ 5o ~/Vlg
are calculated from their oxygen isotopic compo-
sition and estimated values of pore fluid oxygen
isotopes (see Boles & Ramseyer, in press). The
51sO values for 18 calcites range from +21.2 to
DOLOMITE +25.9 ( m e a n = +23.3). Pore water oxygen
FIG. 6. Early dolomite compositions (mole %) from values were estimated with the assumption that
North Coles Levee cores. Data from microprobe 51sO varied linearly between an initial value of 0
analyses of l0 samples in well no. NCL 488-29 at the marine sediment-water interface (4-10°C)
between 2716 and 2768 m depth. to the present-day value of + 2.8 at 2600 m burial
 6 Ma diagenetic history, North Coles Levee I95

Fe~Mn

a.

Ca 95 Mg
Fe

b.

Mn 50 Mg

FIG. 7. Calcite compositions (mole %) from North Coles Levee cores. Data from microprobe analyses of 20
samples in wells no. NCL 88-29, 487-29, and 488-29 between 2069 and 2768 m depth.

depth (105~ Results indicate that calcite range (e.g. Curtis et al. 1986). Fluid inclusions
precipitation occurred between about 40 and are generally very small in these calcites but a
80~ Carbon isotopes for these calcites are sample from 2727m (present temperature
generally light (average 613CpDB = --4.2) indi- 107~ has inclusions with fluid-gas homogeniza-
cating precipitation in the zone of catagenesis in tion temperatures of 57-62~ and an isotopic
accordance with the calculated temperature precipitation temperature of 70-79~ This
I96 J. R. Boles

agreement is reasonable considering the isotope curred but it is difficult to estimate the volume of
temperature estimate is an average for the calcite calcite cement removed. It is probable that the
in the rock and the samples contain a complex more porous horizons were never completely
cementation history judging from compositional cemented with calcite, and intergranular poros-
variation from closely spaced microprobe ity has been enhanced there by dissolution of
analyses. some calcite.
Illite/smectite (I/S) mixed-layer clays are abun- The dissolution of plagioclase and calcite
dant in some sandstone intervals and in sur- appear to be synchronous events. Authigenic
rounding shales. Much of this clay was probably kaolinite is found in sandstones containing
a smectite-rich detrital component of the turbi- porous feldspars and is believed to conserve A1
dity currents. The I/S clays generally show from plagioclase dissolution on the scale of a thin
expandibilities from about 30 to 50~ smectite section (Boles 1984). The kaolinite is absent in
layers with up to 80~ expandibility occurring in tightly cemented sandstones but occurs in porous
some samples (Ramseyer & Boles 1986). Expan- samples as interstitial crystal clusters. Calcite
dibilities tend to be higher (up to 10~ smectite cement in contact with these crystals has ragged
layers) in calcite-cemented sandstones than grain boundaries suggesting the kaolinite now
uncemented samples, reflecting slowing of the occupies space once cemented with carbonate
smectite to illite reaction at low permeability (Fig. 9). It is difficult to imagine a long residence
conditions (see fig. 5 of Ramseyer & Boles 1986). time of aluminum in solution from a dissolving
Porosity enhancement occurred after carbonate plagioclase, hence the kaolinite is believed to
cementation. Calcic plagioclase and calcite were form contemporaneously with plagioclase disso-
the important phases to be dissolved, whereas lution. This implies calcite dissolution has also
albite, K-feldspar and dolomite were unaffected. occurred contemporaneously or prior to these
Plagioclase porosity forms up to one-third of the events. Acids derived from the maturation of
total porosity in the rock and has developed after organic matter have been suggested as the cause
carbonate cementation (Boles 1984). Plagioclase of porosity enhancement in sandstones (e.g.
in calcite cemented rocks is not replaced by Curtis 1978, Surdam et al. 1984, Crossey et al.
calcite nor does it have dissolution porosity, 1986). It is interesting to note that the dissolution
whereas plagioclase in adjacent uncemented event at NCL occurred between 80 and 105~
sandstones exhibits extensive porosity (Fig. 8). and this temperature range corresponds closely
Calcite cements adjoining pore space have with a marked increase in organic acids in
ragged surfaces indicating dissolution has oc- porewaters of the San Joaquin Basin (Carothers

FI~. 8. Plagioclase remnant (P) with opaque residual hyrocarbon. Arrows denote original grain boundary.
Well no. NCL 488-29, 2727 m. Plane light. Note extensive porosity 'O'.
 6 Ma diagenetic history, North Coles Levee I97

FIG. 9. Kaolinite (K) with calcite cement (C). Note ragged texture on calcite cement (arrow) suggesting
kaolinite has filled pore space after some calcite dissolution. Well no. NCL 488-29, 2727 m, partially crossed
nichols. Photograph taken in porous sandstone within 2 mm of cemented-uncemented interface.

& Kharaka 1978). The overlap of the peak in Authigenic pyrite, although not volumetrically
organic acid content of these porewaters with the important, is seen in fractures which cut across
apparent timing of aluminosilicate and carbon- detrital grains and kaolinite. A possible source
ate dissolution at N C L shows the probable for the Fe may be calcite dissolution (see Fig. 7)
influence of kerogen maturation on porosity although this source would also release Mg, the
enhancement. sink for which has yet to be identified. The

FIG. 10. Stress-induced albitization. Fresh plagioclase (P) is albitized (A) in halo about stressed contact
(arrow) with resistant grain (G). Bedding approximately parallel to feldspar twinning. Well no. NCL 487-29,
2752.5 m, crossed nicols.
I98 J. R. Boles

FIG. 11. Euhedral very late stage dolomite rhomb (D) in crushed biotite (B). Well no. NCL 488-29, 2738.8 m.

sulphur for this late-stage pyrite is probably formation water where these carbonates occur is
sourced from hydrocarbons and associated pro- not known precisely because the water in the
ducts of kerogen maturation. central reservoir is an injected mixture of
Late-stage compaction is responsible for very original formation water (6180 = +2.8) and a
late diagenetic changes in these sandstones small volume of water from shallow wells (6180
including albitization of plagioclase along about - 8 ) . Nevertheless the similarity between
stressed grain contacts and growth of dolomite these values and those predicted from the
rhombs along cleavage flakes of crushed biotites isotopic calculations demonstrate the recent
(Boles & Johnson 1983, Boles 1984). These origin of these dolomites.
effects are restricted to porous sandstones that
have undergone compaction. The 'stress-indu- Timing of diagenesis
ced' albitization occurs as an albitization halo in
plagioclase grains in contact with other resistant A burial history curve for the Stevens sandstone
detrital grains (Fig. 10). These contacts are at North Coles Levee shows the very recent
usually concavo-convex and the resistant grain is timing of the above described diagenetic events
commonly quartz. Presumably high silica activi- (Fig. 12). The major carbonate cementation
ties at the stressed interface promote the albitiza- (calcite) occurred between 40 and 80~ which
tion reaction. corresponds to 4.6 to 2.5 Ma BP. Thus the
The dolomite-biotite occurrence is explained dissolution event, kaolinite crystallization, hy-
by development of a high pH halo adjacent to drocarbon maturation, and late stress effects are
biotite cleavage surfaces as a result of H § compressed into a time interval of less than
attraction to the mica surface (Boles & Johnson 2.5 Ma. The hydrocarbon volume alone in-
1983). This carbonate association with the volves an estimated 500 million barrels emplaced
crushed biotite is remarkably selective as in a into a reservoir with an average effective
number of samples it is the only carbonate permeability of 1-10 md. The key point here is
occurrence in the rock (Fig. 11). The dolomite that the carbonate cements in these sandstones
has a 6180 SMOW of + 19.9. Assuming that the are relatively late events. Thus it follows that
dolomite formed at the present reservoir tem- subsequent diagenetic events are compressed
perature of 107~ the calculated pore fluid into a rather narrow time window.
composition from which it precipitated is The stress-induced diagenetic processes are
between + 2 and - 0 . 4 depending on the clearly post-carbonate cementation and result
assumed dolomite-water fractionation curve from reservoir compaction. Removal of the
(Katz & Matthews 1977 and Fritz & Smith 1970). calcite cement during the dissolution phase may
The oxygen isotopic composition in the present have caused the collapse of the reservoir at
 6 Ma diagenetic history, North Coles Levee I99

o Carbonate
20 ntation
0
i._

,-i
40
i._ Late
0 Dissolution, Carbonate
60 .,~~ . . Kaolinite and Cementation,
F_.

~
~ " ~ Oil Migration Pyrite
I- 80 izatlon
m

!._
100

120 I I 1 I I I I
6 4 2 Present

Million Years
FIG. 12. Burial history (temperature) of Stevens sandstone at North Coles Levee in relation to diagenetic
events. Burial history based on stratigraphy shown in Fig. 2 and present geothermal gradient. Carbonate
cementation temperatures (circles) based on isotopic data and assumptions on isotopic fluid composition (see
text). Exact sequence of events from 80 to 105~ is uncertain.

hydrostatic fluid pressure conditions. A n alterna- ACKNOWLEDGMENTS:This paper is taken from a large
tive explanation is that these stress effects are study encompassing a number of oil fields in the San
even more recent, possibly human-induced. A Joaquin Basin. The North Coles Levee study could not
fluid pressure history curve for N o r t h Coles have been done without the complete cooperation of
Levee reveals a very low fluid pressure condition ARCO Oil and Gas Co., in particular ARCO geologist
(1000 psi) recently existed (about 1965) for a Bill Bazeley and engineers Art Tinnemeyer, Don
Puckett, Ken Kuch and Tony Marino. Funding for this
short time period w h e n the gas cap was released
project includes ARCO Geoscience Group and NSF
from the reservoir (Fig. 3). Most of the cores used Grant EAR77-13705 (Boles). The assistance of UCSB
in this study were cut after 1965 and these show personnel in preparing this manuscript including Dave
the most pronounced compaction effects. H e n c e Pierce (photographic expertise), Karen McCormack
the possibility exists that the observed stress- (computer support), Dianne Griffin (drafting), and
related processes of dolomite precipitation and Richard Coffman (proof-reading) is gratefully
albitization occurred within the last 20 years! acknowledged.

References
BOLES, J. R. 1984. Secondary porosity reactions in the CROSSLY, L. J., SURDAM,R. C., & LAHANN,R. 1986.
Stevens sandstone, San Joaquin Valley, Califor- Application of organic/inorganic diagenesis to
nia. In: MCDONALD, D. A. & SURDAM, R. C. porosity prediction. In: GAUTIER, D. L. (ed.).
(eds). Clastic Diagenesis, pp. 217-24. American Roles of Organic Matter in Sediment Diagenesis, pp.
Association of Petroleum Geologists Memoir, 147-55. Society of Economic Paleontologists and
37. Mineralogists Special Publication, 38.
- - & JOHNSON,K. S. 1983. Influence of mica surfaces CURTIS, C. D. 1978. Possible links between sandstone
on pore-water pH. Chemical Geology, 43, 303-17. diagenesis and depth related geochemical reac-
- - & RAMSEYER,K. In press. Diagenetic carbonate in tions occurring in enclosing mudstones. Journal of
a Miocene sandstone reservoir, San Joaquin the Geological Society of London, 135, 107-17.
Basin, California. American Association of Pet- - - , COLEMAN,M. L. • LOVE,L. G. 1986. Pore water
roleum Geologists Bulletin. evolution during sediment burial from isotopic
CALLAWAY, D. C. 1971. Petroleum potential of San and mineral chemistry of calcite, dolomite and
Joaquin Basin, California. In: CRAM, I. M. (ed.). siderite concretions. Geochimica et Cosmochimica
Future Petroleum Provinces of the United States-- Acta, .50, 2321-34.
their Geology and Potential, pp. 239-53. American DREVER, J. I. 1982. The Geochemistry of Natural
Association of Petroleum Geologists Memoir, 15. Waters. Prentice-Hall, Englewood Cliffs, NJ.
200 J. R. Boles
FRITZ, P. & SMXTH, D. C. W. 1970. The isotopic RAMSEYER, K. & BOLES, J. R. 1986. Mixed-layer
composition of secondary dolomite. Geochimica et smectite/illite minerals in Tertiary sandstones and
Cosmochimica Acta, 14, 1875-7. shales, San Joaquin basin, California. Clays and
KATZ, A. & MATTHEWS,A. 1977. The dolomitizationof Clay Minerals, 34, 115-24.
CaCOD: an experimental study at 252-295~ SURDAM, R. C., BOESE, S. W. • CROSSEY,L. J. 1984.
Geochimica et Cosmochimica Acta, 41, 297-308. The chemistry of secondary porosity. In: MC-
LAGOE, M. B. In press. The record of sea level and DONALD, D. A. & SURDAM, R. C. (eds). Clastic
climatic fluctuations in an active margin basin: Diagenesis, pp. 127-75. American Association of
The Stevens sandstone, Coles Levee area, Califor- Petroleum Geologists Memoir, 37.
nia. Palaios. WEBB, G. W. 1981. Stevens and earlier Miocene
MATSOMOTO, R. & IIJIMA, A. 1981. Origin and turbidite sandstones, southern San Joaquin Val-
diagenetic evolution of Ca-Mg-Fe carbonates in ley, California. American Association of Petroleum
some coalfields of Japan. Sedimentology, 28, 239- Geologists Bulletin, 65, 438-65.
59.

J. R. BOLES,Department of Geological Sciences, University of California, Santa Barbara,

CA 93106, USA.
 Trace-element source and mobility during limestone burial
diagenesis an example from the Middle Jurassic of eastern
England
D. Emery

S U M M A RY: Two phases of burial calcite from the Lincolnshire Limestone (Bajocian) of
central Lincolnshire have been analysed for their trace element (Fe, Mn, Mg, Sr) and
strontium isotopic compositions.
The earliest, non-ferroan, brightly luminescent phase has the following mean trace
element concentrations: Fe-- 919 ppm; Mn-- 116 ppm: Mg-- 1778 ppm; Sr-- 327 ppm.
The later, ferroan, dully luminescent phase has the following mean compositions: Fe--
3580ppm; Mn-- 225 ppm; Mg-- 1400ppm; Sr-- 318ppm. The strontium isotopic
composition of the two phases is indistinguishable, with a mean 87Sr/86Sr ratio of 0.70820.
This value is considerably more radiogenic than host Bajocian marine carbonate
(SVSr/S6Sr = 0.70725).
The Fe and Mn of the burial calcites were sourced by iron and manganese oxyhydroxides
associated with shales bounding (and internal to) the Lincolnshire Limestone. The relatively
shallow maximum burial depths (550 m) and low temperatures (maximum of 42~ attained
by the formation in the area of study preclude major clay or sandstone diagenetic
transformations as potential sources for Fe and Mn. These trace elements moved into the
Limestone in solution via compaction driven cross-formational flow from bounding shale
aquicludes.
The Sr and Mg trace element components of the burial cements were derived from: (i)
remobilized Bajocian marine carbonate, and (ii) remobilized Carboniferous marine
carbonate. These sources satisfy the requirements of a linear Sr versus Mg relationship
observed for the burial cements, that is, a constant Sr/Mg ratio with time, but highly variable
absolute Sr and Mg contents. These sources also satisfy the near constant 87Sr/86Sr
composition of the burial calcites.
The burial hydrology required by the trace element budget above is a convective flow
system with cross-formational elements as normal bleeder faults connecting the Carbonifer-
ous Limestone and the Lincolnshire Limestone. Additional compaction-driven flow
occurred from bounding shale formations (chiefly the underlying Grantham Formation and
to a lesser extent the Upper Estuarine 'Series') to the Lincolnshire Limestone. Up-dip flow
within the fully confined Limestone provides an enformational limb to the convective
system.

During burial of sedimentary sequences, re- composition of burial cement phases described
actions can occur such that large quantities of by Stueber et al. (1984), Faure & Szabo (1986)
cations (and anions) become redistributed in new and Emery et al. (1987) is considerably more
and modified phases throughout the sequence. radiogenic than that of the host marine
With regard to limestone burial diagenesis, the carbonate.
predominant newly formed phase is sparry Many potential sources for these cations exist
calcite cement often with subordinate dolomite. within bounding or intraformational clastic or
In many carbonate sequences it is apparent that clay-rich sequences; the relationship between
a large proportion of this burial calcite can be ferroan calcite and ferroan dolomite burial
derived from the host depositional (usually cements and shale sequences bounding the
marine) carbonate, by pressure dissolution or by carbonate of interest has been noted by many
leaching. It is not possible in many cases, authors (Oldershaw & Scoffin 1967, Scholle &
however, to account for all the calcite cement by Halley 1985, Schofield & Adams 1986). Other
remobilization of pre-existing material. Further- studies (notably Stueber et al. 1984) have invoked
more, a source external to the depositional the breakdown of rubidium-containing clastic
carbonate is necessitated for (i) trace elements phases (K feldspar and mica) to explain the
such as iron and manganese shown by many higher 87Sr/86Sr ratios of the burial calcites
authors to be present in burial calcite phases relative to their host carbonate.
(Oldershaw & Scoffin 1967, Scholle & Halley Whilst the circumstantial evidence for extra-
1985); (ii) strontium, as the strontium isotopic formational derivation of diagenetic components

From: MARSI-IALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 20I

Geological Society Special Publication No. 36, pp. 201-217.
202 D. Emery
is difficult to dispute, it is clear that a more porewater chemistry and precipitated phase
specific relationship between cation source, based on distribution coefficients. Such work is
transport pathways and sink needs to be estab- only of practical value in Recent or Pleistocene
lished. To this end, the burial history of the carbonates, where the aquifer system and its
sediments of interest and its bounding forma- attendant flow properties can be evaluated in
tions needs to be well-constrained and thereby its detail, and where the dissolving and precipitat-
thermal evolution and the attendant importance ing phases within the aquifer can be unequivo-
(or otherwise) of clastic and clay mineral cally identified. For details of such work the
diagenesis (e.g. Curtis 1983, Burley, Kan- reader is referred to Veizer (1983) for an excellent
torowicz & Waugh 1985). review.
This type of source-specific chemical model-
ling can be applied to the Lincolnshire Lime-
stone, a Middle Jurassic carbonate which, in the
study area, underwent shallow burial to 550 m. Geological setting
This limestone displays a distinctive suite of
burial carbonate phases, from earliest non- The Lincolnshire Limestone (Bajocian) is the
ferroan calcite to ferroan dolomite and latest major carbonate formation in the Middle Juras-
ferroan calcite. The calcite phases, on which the sic of eastern England, stretching from North
most data are available, will be discussed relative Humberside, through its thickest development of
to a burial budget of the following elements: Fe, 40 m at Boothby Pagnell in Lincolnshire to a
Mn, Mg, Sr. It is important to emphasize that feather-edge, probably determined by a combi-
this is not a study on the relationship between nation of original thinning and pre-Bathonian

GRIMSBY

NORTH SEA

OUTCROP

STUDY

AREA

GRANTHA~ THE WASH

PETERBOROUGH

; 2Y 25kin
KETTERING

FIG. 1. Outcrop (cross-hatched) and subcrop of the Lincolnshire Limestone in England.

 Limestone burial diagenesis, E. England zo3

70 ~ Lincoln

DQ /
/
FE

I' /
/o
HQ
~MQ /
/'
60
';sa~GB ,"GF
'
,/
GJ/ GC oGG
G D,--- ";"
,GA

GE H G
50
% HB
HE~ ,,

~ Sleaford HC
it.
CH. LQ
40 Ja~
/
)E/'

rantham 5000M

30 I l I

90 00 10 20
National Grid

FIG. 2. Distribution of core and quarry material used in the present study. Hatched line indicates western
limit of outcrop. HQ, Harmston Quarry; CQ, Coleby Quarry; LE, Leadenham Quarry; HF, North Rauceby
Core; CA, Castle Quarry, Ancaster; CH, Copper Hill Quarry, Ancaster; LQ, Lyte's Quarry; GQ, Gregory's
Quarry; RQ, Ropsley Quarry; DQ, Dunston Quarry; MQ, Metheringham Quarry; BQ, Blankney Quarry; GB,
Kirby Green Core; G J, Ashby Core; GC, Digby Core; GD, Bloxholm Core; GH, Cottage Farm Core; HG,
Pear Tree Farm Core; HE, Leasingham Core; JA, Aunsby Core; FE, Dunston Fen Core; GF, Timberland
Core; GG, Walcot Core; GA, Dorrington Core; HB, Asgarby Core; HC, Great Hale Core; JD, Bicker Core;
JE, Donington Core.

erosion, in mid-Northamptonshire (Fig. 1). The tilted and eroded. Deposition resumed during
study area for the present work and the distribu- the Bathonian, initially with the non-marine de-
tion of the core and outcrop material is illustrated posits of the Upper Estuarine 'Series', but later
in Fig. 2. The Lincolnshire Limestone shows a with renewed marine transgression. The uncon-
very complex internal stratigraphy and a wide formity between the Lincolnshire Limestone and
range of carbonate lithologies (Ashton 1980, Upper Estuarine 'Series' is represented by six
Emery 1986). These can be interpreted as ammonite Zones.
representing the landward migration of an The structure of the Lincolnshire Limestone is
offshore barrier-bar complex (now represented simple, dipping gently to the east at an average of
by the upper Lincolnshire Limestone) across a about 1~ beneath later Jurassic strata, themselves
protected lagoon and back-barrier (lower and overlain by Pleistocene and Holocene Fen
middle Lincolnshire Limestone: Ashton 1977 deposits. It is an important aquifer in an area of
and in Tucker 1985). This in turn transgresses low rainfall and intensive agriculture. Its hydro-
across non-marine coastal marsh deposits of the geology has been studied by Downing et al.
underlying Grantham Formation (Fig. 3). (1977) and Edmunds & Walton (1983) and cores
After the deposition of the Limestone there taken for this purpose have been sampled in the
was a regional regression and the Formation was present study, as well as available outcrops.
204 D. Emery
AGE UNIT

i]l[i
UPPER C R E T A C E O U S
m,;
I1 I I 1 I CHALK
I I I

LOWER CRETACEOUS SANDS,

LOWER C R E T A C E O U S
CLAYS, IRONSTONES

- - = KIMMERIDGE CLAY
UPPER J U R A S S I C

AMPTHILL CLAY

OXFORD CLAY
~ I I I I --
- B L I S W O R T H LST. KELLAWAYS BEDS
[:_=--=-~_~:5=]
-

MIDDLE JURASSIC -- UPPER E S T U A R I N E "SERIES"

. . . . . .
-- LINCOLNSHIRE LIMESTONE
F~:=~:I -- - GRANTHAM FORMATION &
_ _ .--
NORTHANTS. SAND IRONSTONE

LOWER JURASSIC LIAS C L A Y S

_ _ _ _

- - - - _

TRIASSIC TRIASSIC MUDSTONES

FIG. 3. Simplified stratigraphic column for the Mesozoic of the East Midlands.

Pre-burial diagenesis of the i.e. below the influence of shallow groundwater

circulation and above the zone of low-grade
Lincolnshire Limestone metamorphism.
Diagenesis of the Lincolnshire Limestone began
in the marine environment, with abundant The burial history of the
microboring and precipitation of a variety of
cements including major hardground cements Lincolnshire Limestone
(Marshall & Ashton 1980, Emery 1986). Pyrite
was the major non-carbonate precipitate within The Lincolnshire Limestone became buried
the marine realm. Localized syndepositional underneath further Middle Jurassic shales and
subaerial exposure of the Formation gave rise to minor limestones (Fig. 3), the majority of which
major aragonite dissolution and precipitation of were deposited in the marine environment.
a distinctive phase of meteoric cements (Emery During the Upper Jurassic, major thicknesses of
1986). the Oxford, Ampthill and Kimmeridge Clays
Post-depositional diagenesis commenced dur- accumulated, with particularly rapid deposition
ing folding and exposure of the Limestone in the of the Kimmeridge Clay (Fig. 4). Minor uplift
Bajocian, with the establishment of a largely and erosion of the Kimmeridge Clay blanket
unconfined meteoric aquifer system with atten- probably occurred in pre-Cretaceous times, but
dant aragonite dissolution and precipitation of a the precise amount of emergence is difficult to
Formation-wide correlatable cement sequence, evaluate (Fig. 4). Slow burial of the sequence
consisting of euhedral, distinctively zoned (under under Lower Cretaceous sands, shales and
cathodoluminescence) low magnesium calcite as ironstones persisted until deposition of the chalk,
echinoderm overgrowths, in biomouldic porosity when rapid, differential burial occurred (Fig. 4).
and rarely as intergranular fringe cements With the termination of Chalk deposition, the
(Emery 1986). Thereafter the Limestone became sequence underwent uplift related to end-Creta-
entirely confined beneath the Upper Estuarine ceous inversion tectonism (Glennie & Boegner
'Series' and moved into the realm of burial 1981). The Lincolnshire Limestone was ultimate-
diagenesis as defined by Scholle & Halley (1985), ly re-exposed during the Miocene (Peach 1984).
 Limestone burial diagenesis, E. England 205
MILLIONYEARSB. P.
190 150 100 ,
50 i

0
100
200 ~ 2 3 "~6.7(28.0) /
300
400
500
600
~ 42.0)

FIG. 4. Burial curve for the Lincolnshire Limestone within the study area. Broken lines represent uncertain
emergence. Temperatures were calculated from equation 2 of Andrews-Speed et al. (1984). Temperatures in
parentheses were calculated from oxygen isotopic composition of the burial calcites (Emery 1986). All
temperatures are in ~

The burial curve of Fig. 4 depicts temperatures Non-ferroan calcite

attained by the Limestone in the study area
This phase forms up to 5~ of the total rock by
during burial, calculated according to equation 2
volume. It displays an irregular distribution,
in Andrew-Speed, Oxburgh & Cooper (1984).
occurring as isolated poikilotopic crystals, im-
These temperatures represent likely maxima for
persistent lenses at least of the dimensions of
the Limestone in the study area due to heat
individual thin-sections and contiguous bands.
conduction alone through the sedimentary pile
The crystals form both drusy and poikilotopic
overlying the Limestone. The two temperatures
fills to intra- and intergranular porosity.
in parentheses are those attained by the Lime-
This phase stains a characteristic pink with
stone assuming heat transfer by fluid convection
mixed potassium ferricyanide and Alizarin Red-
through the Limestone during burial. These
S stain (Dickson 1966), indicative of a low iron
latter temperature determinations are based on
content (Dickson 1966), although a mauve tinge
6180 compositions of the burial cement phases;
may be present in some examples. Under
the rationale and calculations behind these
luminescence non-ferroan calcite is generally
estimates are detailed in Emery (1986).
brightly luminescent with complex intracrystal-
In summary, the maximum depth of burial of
line zonation in rare examples.
the Lincolnshire Limestone in the study area did
This calcite clearly post-dates major aragonite
not exceed 550m, the maximum temperature
dissolution and compaction of allochems and
attained by the Limestone was probably 42~
earlier cement generations.
and it is clear that the Formation was confined
between major clay-rich sequences during burial.
Ferroan dolomite
The most rapid burial of the formation occurred
(i) during deposition of the Kimmeridge Clay Dolomite is present both as cement and replacive
and (ii) during deposition of the Chalk. phase and forms a minor component (approxi-
mately 2~) of the total rock by volume in the
study area. Its origin intervening major phases of
calcite precipitation is problematic and beyond
Burial cements of the Lincolnshire Limestone
the scope of the present paper.
The cements of the Lincolnshire Limestone
Ferroan calcite
precipitated in the burial regime consist of (i)
non-ferroan calcite, (ii) ferroan dolomite and (iii) Ferroan calcite is the most volumetrically signifi-
ferroan calcite. cant diagenetic precipitate in the Limestone,
206 D. Emery
forming up to 25% of the whole rock by volume. allow direct comparison with the data of Burke et
It forms poikilotopic and drusy fills to intra- and al. (1982) which, while employing this very low
intergranular porosity and is of ubiquitous normalization value for the NBS 987 standard,
distribution. remains the most comprehensive data set for the
This phase stains a characteristic blue (Dick- Jurassic.
son 1966), and may display weakly ferroan or The results from ICP analysis are summarized
non-ferroan complex intracrystalline zonation. It in Table 1. Full, tabulated results of the Sr
is dully luminescent, with patchy intracrystalline isotopic data are given in Emery (1986) and
non-luminescent zones; however, no concentric Emery et al. (1987).
zonation, suggestive of a change in fluid chemis-
try with time, is observed.
The ferroan calcite precipitation post-dates Discussion
further compaction of the Limestone, as evi-
denced by fracture of the non-ferroan calcite
Iron concentrations
phase. The absolute timing of cement precipita-
tion is considered to have been during deposition The mean values and ranges for iron concentra-
of the Upper Chalk (Emery 1986). This assertion tions in the burial cement phases are given in
is based chiefly on stable isotopic evidence and Table 1.
burial history reconstructions. The important The presence of iron within calcite cements
feature of this cement is the relatively short depends on at least the following independent
timespan of cement precipitation and the ab- factors; the Eh and pH of the environment of
sence of concentric intracrystalline chemical and precipitation and the supply of iron. An impor-
isotopic zonation. tant result in all the Lincolnshire Limestone
The non-ferroan calcite is also considered, burial cements is the invariable presence of iron.
albeit more speculatively, to have been precipi- If this iron is substituting for Ca :+ in the calcite
tated over a relatively short timespan, and the lattice, the environment of precipitation of all
above point also applies to this phase. the burial calcites must have been reducing. A
control sample of (Recent?) speleothem calcite
from the Lincolnshire Limestone was analysed
Analytical methods and results with the burial calcites and, as anticipated,
contained no iron, at least justifying contamina-
(i) Inductively-coupled argon plasma spectro- tion-free sampling. The variation of iron content
metry (ICP). Samples of calcite were extracted in the burial cements is thus ascribed to a
using a stout needle from thin-sections and slabs. variation in the iron supply to the growing crystal
0.1 g of sample were dissolved in 10 ml of quartz- faces, not simply the total iron supply. This
distilled 1.75 M HC1. The samples were analysed qualification is necessary because the total iron
on a Phillips 50 MHz ICP. Results obtained (see supply to the porewaters may have been avail-
Table 1 for mean and range) are expressed with a able to other minerals besides calcite. Co-
+ / - 3~ confidence. Details of this technique precipitation of pyrite in a reducing environment
are available in Thompson & Walsh (1983). will selectively remove Fe z+ from the pore-
(ii) Strontium isotopic compositions were waters, with consequent precipitation of an iron-
determined by P. C. Smalley at the University of free or iron-poor calcite (Curtis 1977).
Oslo. Full details of this technique are available
in Emery (1986) and Emery et al. (1987).
Analysed SVSr/S6Sr ratios were adjusted to a Non-ferroan calcite
value for the NBS 987 standard of 0.71014 to There is no textural evidence for co-precipitation
of pyrite with the non-ferroan calcite phase, but
such evidence may be difficult to ascertain,
TABLE 1. All values are in parts per million. particularly in the absence of intergrowth
NFC--non-ferroan calcite; FC--ferroan fabrics. Indirect evaluations of the importance
calcite of pyrite precipitation at this stage can be
Fe Mn Mg Sr
considered through the sulphate budget of the
aquifer and the carbon isotopic composition of
Mean coneentrat~ns the non-ferroan cements:
NFC 920 120 1780 330 During the precipitation of the non-ferroan
FC 3580 230 1400 320
calcite, the Lincolnshire Limestone is shown to
Range have been a confined aquifer, probably buried
NFC 140-1680 80-230 600-2770 90-840
FC 2030-4830 160-390 540-2890 80-650
beneath Upper Jurassic Clays (Fig. 4). Under
such circumstances it would be difficult to
 Limestone burial diagenesis, E. England 207

replenish the supply of sulphate to the suggested by many authors, notably Oldershaw
porewaters. Any marine-derived sulphate would & Scoffin (1967). These workers demonstrated
probably have been long exhausted by bacterial that ferroan calcites precipitated in limestone
reduction and subsequent pyrite formation on sequences adjacent to shale bands displayed a
shallow burial within the marine realm (Berner decrease in Fe content away from the shales. My
1980). Supply of sulphate from meteoric waters is own general observations of British Jurassic
probably untenable; the presence of sulphate in carbonate sequences (Blisworth Limestone, Cor-
the present Lincolnshire Limestone porewaters allian) show a predominance of pore filling
(Edmunds & Walton 1983) is due to a combina- ferroan cements, presumably related to juxta-
tion of pyrite oxidation at or near recharge and posed argillaceous sequences. By contrast,
agrichemical poisoning of the aquifer. Neither of Moore & Druckman (1981) described non-
these situations would have pertained to the ferroan burial cements from a clay-poor carbon-
Lincolnshire Limestone aquifer during precipi- ate sequence. Bounding the Lincolnshire Lime-
tation of the non-ferroan calcite. stone are two clay-rich units, the Grantham
Consideration of the mean and range 6 ~3C of Formation and the Upper Estuarine Series (Fig.
the non-ferroan calcite demonstrates that it 3), and there are considerable thicknesses of clay
differs little from that of the marine carbonate within the Limestone itself (e.g. the Kirton Shale
(Fig. 5). There appears to have been no input of Member; Ashton 1980). There appears to be no
~3C depleted organogenic carbonate to the difficulty regarding clay-rich sources for the iron
calcite, as would be expected during bacterial of the burial cements. XRD analysis of clay-rich
sulphate reduction (Berner 1980, Irwin et al. sequences within the Limestone (no analyses of
1977). the clastic formations outside the Limestone
It is thus concluded that sulphate was absent have been made) show that its clay mineralogy
from the diagenetic system at this time, and consists chiefly of kaolinite, but there is certainly
pyrite co-precipitation with the non-ferroan sufficient smectite, illite and biotite present, all
calcite did not occur. The absence of major of which will contain structurally bound iron.
quantities of iron at this time is further interpret- Furthermore, smectite and illite have high cation
ed as a function of low overall iron supply to the exchange capacities resulting from charge imba-
porewaters. lances in the crystal structure or at its surface
(Carroll 1959, Rosler & Lange 1972). Dissolution
or recrystallization of these phases could poten-
Ferroan calcite
tially yield iron to the porewaters of the
The iron content of both burial cement phases Lincolnshire Limestone aquifer as Fe 2+ under
must be sourced from outside the carbonate- reducing conditions. However, conditions fa-
dominated portion of the Limestone, or indeed vourable for clastic diagenesis as above must also
from outside the Limestone itself. External have pertained during burial of the Limestone.
sources for iron in ferroan calcites have been Such diagenetic events would not be anticipated

/-7>- . . . . 8-,
,,# !
/" ]

..'";~ o./
fs ,~~
, ~ ~ ~: ,o

"'~ : A A A sS
MARINE FIELD
,$1 C
NON-FERROAN BURIAL CALCITE

-1( -9 -18 - '7 -=6 '5 - '4 -3' -2' - I1 00

8180
FIG. 5. The stable isotoic composition of the non-ferroan burial calcites compared to a marine field (defined
by diagenetically unaltered bivalves) from the Lincolnshire Limestone. Note the significant depletion in 180
of the burial calcites compared to the marine field, but the negligible depletion of 13C. Data from Emery
(1986).
208 D. E m e r y

at the relatively shallow burial depths and low correlation between low iron content, higher
temperatures attained by the Limestone during Mn/Fe ratio and bright luminescence exists.
burial, according to the scheme of Curtis (1983) Similar correlations have been described within
and Burley et aL (1985). It is concluded that individual studies by a number of authors (see
clastic and clay mineral diagenetic events were of Fairchild 1983), but an aggregation of all results
limited importance in cation supply to the burial reveals a random scatter (Dickson, personal
porewaters of the Lincolnshire Limestone. communication, 1986), albeit with brightly lumi-
Another potential source, also associated with nescent cements displaying higher Mn/Fe ratios.
clay minerals, may be as iron oxide films on the Such data scatter could be a function of variable
surface of the clays, or as poorly ordered, instrumentation, or subjective attribution of the
amorphous iron oxy-hydroxides. These sources, terms bright and dull. Despite this unavoidable
suggested as possible sources for iron in sedimen- subjectivity, it is widely accepted that high
tary ironstones, are particularly attractive as they Mn/Fe ratios are prerequisite for bright lumines-
may also provide a complementary source of cence of carbonates, although Ten Have &
manganese as oxide films, also present on clay Heijnen (1985) have suggested that luminescence
surfaces, as well as reactive, amorphous manga- intensity is controlled by the Mn concentration,
nese oxy-hydroxides (Curtis & Coleman 1986). and not by the Mn/Fe ratio. Both these possibili-
Furthermore, reducing conditions will liberate ties are borne out by the data on Lincolnshire
the iron and permit its transport as Fe2+; no Limestone burial cements.
clastic or clay mineral diagenetic events are
required. It is thus suggested that the iron and
Paired behaviour of iron and manganese
manganese of the burial cements was supplied
from oxide films on the surface of clay minerals The Fe versus Mn cross-plot of Fig. 6 provides
and/or from poorly ordered amorphous oxy- valuable information besides its implications for
hydroxides. calcite luminescence. A broadly linear correla-
tion between Fe and Mn of the non-ferroan and
ferroan burial cements is exhibited. Trends for
Manganese concentrations both the non-ferroan and ferroan calcites are
parallel and the regression line drawn through
The mean value and ranges for manganese
the points intercepts the Mn concentration axis
concentrations in the Lincolnshire Limestone
before the Fe concentration axis. The simplest
burial cements are detailed in Table 1, and
interpretation of this result is that Mn 2+ will be
display relatively low concentrations compared
incorporated into the calcite lattice before Fe 2+,
to abundant published data (e.g. Fairchild 1983,
as Mn is more readily reduced from its tri- or
Ten Have & Heijnen 1985). Incorporation of
tetravalent state than iron is reduced from its
manganese into the calcite structure necessitates
trivalent state (Curtis & Coleman 1986). This
its reduction to the divalent state, Mn 2+. This
assertion predicts that any iron-free calcite
again demands reducing conditions, but not to
precipitated will have manganese concentrations
the same low Eh as required for iron incorpora-
of ~ 70 ppm, although the presence of iron in all
tion into calcite (Oglesby 1976, Froelich et al.
burial phases analysed implicates conditions of
1979). However, the invariable presence of
sufficiently low Eh during burial such that
reducing conditions as demonstrated by the iron
ferrous iron was variably mobilized.
evidence will have maintained manganese in the
The broadly linear trend for Fe versus Mn of
divalent state throughout the precipitation of the
Fig. 6 demonstrates that for each additional
burial calcites, and since sulphate reduction is
portion of calcite precipitated the Mn/Fe ratio
assumed to have been minimal, Mn in the
therein was constant. This further implies that
porewaters will have been incorporated into the
calcites. the Mn/Fe ratio of the porewaters was constant
with time, as the concentration of a cation in
calcite is a linear function of its concentration in
the precipitating fluid assuming equilibrium
Implications for the luminescence of burial
cements partitioning (Mclntire 1963).
The constancy of the Mn/Fe ratio of the
Figure 6 depicts a cross-plot for Fe versus Mn porewaters with time was maintained, however,
concentrations in the late spars. Open triangles not by an unchanged absolute supply of Fe and
are samples of non-ferroan calcite which exhibit Mn, but by an increased rate of supply of both
bright luminescence. Not all the samples of non- cations which was balanced to maintain the
ferroan cements were analysed for luminescence Mn/Fe ratio, according to the simple linear
characteristics. No ferroan calcites (solid tri- relationship. There are two possible explanations
angles) displayed bright luminescence. A simple for this observed relationship: (a) The Fe and
 Limestone burial diagenesis, E. England 209

8000

7000

6000 - SJ S

jS

S
SS
5000 S
Sp
S

,s ~
A ss
E ss S
s
O.
ft.
4000
v A sS
@ s ~ SSS
1.1.
s
3000 ssS t
t / t
t/" /
2000

f sS
ss J

1000 sS S

si~ S

sr ~ A i Jr I I I
0
0 100 200 300 400

Mn (ppm)
FIG. 6. Cross-plot of Fe versus Mn concentrations in the burial calcites. Open triangles represent non-ferroan
burial calcites, solid triangles represent ferroan burial calcites.

Mn of oxide films and/or amorphous oxy- sequence is demonstrated both by petrography

hydroxides, the proposed source for these ca- and burial curve reconstruction (Fig. 4) and, in
tions, were increasingly mobilized on progressive the absence of contradictory data, option (b) is
burial. (b) The mobilization rate of Fe and Mn preferred.
was constant, but porewaters to which these
cations were being contributed (i.e. argillaceous Magnesium concentrations
units within and/or outside the formation)
formed an increasing proportion of the total The mean value and ranges for magnesium
Limestone porewaters through increasing cross- concentrations in the burial cements are shown
formational flow from the shale bands to the in Table 1. Magnesium will be incorporated into
Limestone as a function of progressive burial. the calcite lattice independently of redox poten-
Both possibilities relate to this simple cation tial of the environment of precipitation, as it
budget route: exists naturally as a divalent ion of an appropri-
ate ionic radius. It is an important component of
Mn/Fe oxides, hydroxides(Fe, ran)
marine calcite precipitates where it may substi-
>Porewaters(Fe, Mn) >CalcitetFe, M,)
tute up to 18 mol~ MgCO3 in the lattice as high
Distinction between the two possibilities is magnesium calcite. High magnesium calcite is
difficult; however, progressive burial of the generally considered to contain from 4-18 mol~
210 D. Emery
4

o
o
o

X
E
~'
o. 2
A
AA
9
A A 9

A A
/\

J
L i i I I I
0 1 2 3 4 5

Fe (ppmxlO00)

FIG. 7. Cross-plot of Mg versus Fe for the burial calcites. Large triangles represent mean concentrations, key
as for Fig. 6.

MgCO3 (Chave 1954). Magnesium may also greater. The mean Mg concentration of the non-
substitute into the aragonite lattice. Since the ferroan calcites is also slightly greater than that
Lincolnshire Limestone is composed dominantly for the ferroan calcites (Table 1).
of marine carbonate, a ready internal source of Consideration of the magnesium concentra-
Mg was available for incorporation into the tions in isolation, however, does not permit a
burial cements on remobilization, an important satisfactory evaluation of their source. Circum-
distinction from iron and manganese where an stantial evidence suggests that Mg could be at
external source of cations was necessary. This least partially derived from the host marine
is not to imply that high magnesium calcite carbonate; this assertion is best verified by
persisted into the burial regime; earlier diagene- comparison of the Mg concentrations with Sr.
tic events, particularly meteorically influenced
diagenesis, transformed much (and possibly all) Strontium concentrations
of the original marine high magnesium calcite
into low magnesium calcite. However earlier Strontium is potentially of great utility in the
cements and marine components still contain an chemical modelling of carbonate diagenesis
appreciable component of magnesium, up to (Veizer 1983), and has been used with success
7000 ppm in some echinoderm fragments ana- particularly in modelling diagenesis in the
lysed by energy-dispersive electron microprobe meteoric realm (Garish & Friedman 1969,
(Emery 1986). Kinsman 1969). Like magnesium, it exists as a
The magnesium content of the burial cements divalent ion, and is incorporated into the calcite
is variable and is best illustrated by a magnesium lattice independent of redox potential of the
versus iron cross-plot (Fig. 7) which shows no environment of precipitation. A further impor-
distinctive correlation between the two cations. tant property is its abundance in marine carbon-
However, for the relatively limited range in Fe of ate phases, particularly aragonite. A profusion of
the non-ferroan calcites, compared to the ferroan data exist on Sr concentrations in recent marine
calcites, the range in Mg is proportionally carbonates, notably after Kinsman (1969) and
 Limestone burial diagenesis, E. England 211
10

9 -

7
o A A
O 6
T-
X

E
o. 5

I I I i I
0 l 2 3 4 5
Fe (ppmx 1000)

FIG. 8. Cross-plot of Sr versus Fe for the burial calcites. Large triangles represent mean concentrations, key as
for Fig. 6.

reviewed by Veizer (1983) and it is clear that Sr/Mg ratio in the burial porewaters with time,
despite massive Sr loss from depositional arago- and that Sr and Mg were probably sourced by the
nite, largely by dissolution in the meteoric realm, same type of material. By contrast with the Fe
Sr would still be of sufficient concentration in versus Mn cross-plot, there is no absolute
marine carbonates and earlier cement phases to increase in the Mg and Sr contents with time (see
provide a major potential source of strontium for also Figs 7 and 8). (c) Any material(s) ultimately
the burial cements. Figure 8 depicts an Sr versus sourcing the Sr and Mg of the burial calcites must
Fe cross-plot for the burial calcites. From the have exhibited an invariant Sr/Mg with time,
form of this plot, it is clear that there is no although the absolute concentrations of Sr and
distinctive correlation between Sr and Fe. Table Mg may have shown considerable, unsystematic
1 further emphasizes the similarity of the mean variation with time.
and range of Sr concentrations for the non- The above constraints require a Sr and Mg
ferroan and ferroan calcites. The form of Fig. 8 source for the burial calcites of constant Sr/Mg
is, however, very similar to that of Fig. 7. A ratio, but one which would allow the absolute
direct plot of Sr versus Mg (Fig. 9) clearly concentrations of both these cations to vary with
demonstrates the origin of this similarity, in that time. The original marine carbonate of the
Sr and Mg describe a positive linear correlation. Lincolnshire Limestone apparently satisfies
The regression line through the points of Fig. 9 these constraints; furthermore, it has clearly
has the simple formula Sr(concentra- been removed by pressure-dissolution and
tion)= 0.21 Mg(concentration). This relation- normal dissolution processes during burial
ship has several significant implications: (a) The diagenesis.
intercept of the regression line is at the origin. If Bajocian marine carbonate is indeed the
This would be expected for two elements whose source of calcite for the burial cements, then the
incorporation into calcite is independent of STSr/86Sr of the burial cements ought to be
redox potential. (b) The linear relationship identical to that of original, diagenetically
between the two elements implies a constant unaltered, Bajocian marine carbonate, i.e. their
212 D. Emery
9

ss S
s
s
sS
sS

s S SSSSSJ /
6 s 9S
A
0 s

s9 A
0 9 9162
5
ss s 9 ~
X 94, J
9 9 SS
E
Q. 4
Q. sS
s
9 ss S

3- 9 A 9 ,x zx zx

sss SS

s 9S
s

9s SS
It I I
0 1 2 3 4
Mg ( p p m x l O 0 0 )
FIG. 9. Cross-plot of Sr versus Mg for the burial calcites.

mutual 87Sr/86Sr ratio should be approximately priate; a source of additional radiogenic Sr is

0.70725 (Burke et al. 1982). required in the first instance. However, this
additional radiogenic Sr must still satisfy the
Strontium isotopes criteria required by the Sr and Mg concentra-
tions of the burial cements, in that its Sr/Mg ratio
The details of the Sr isotopic composition of the must be constant and identical to that of the
Lincolnshire Limestone are presented in Emery remobilized Bajocian marine constituent. Fur-
(1986) and Emery et al. (1987). For the purposes thermore, the absolute Sr and Mg concentrations
of the present argument, Fig. 10 clearly demon- of the calcite must be allowed to vary consider-
strates that there is a considerable separation ably, yet the near-constant STSr/a6Sr ratio of the
between the Sr isotopic composition of Bajocian cements must remain unaffected.
marine material, as represented by brachiopods These constraints probably rule out clay or
from the Lincolnshire Limestone (mean clastic minerals as a source of Sr and Mg for the
STSr/86Sr=0.70725), and the two phases of burial cements. Analysis of the Sr isotopic
burial calcites (mean 87Sr/S6Sr--0.70820). Fur- composition of clay mineral phases in the
thermore, the strontium isotopic composition of Lincolnshire Limestone (Smalley et al. 1985),
the two phases of burial calcite is almost and consideration of the Sr isotopic composition
indistinguishable. of associated K feldspars (Emery 1986) shows
The model involving remobilization of Bajo- that these potential contributors are too radio-
cian marine carbonate alone is clearly inappro- genic, and that the Sr/Mg ratios of these phases is
 Limestone burial diagenesis, E. England 2I 3
h

LATE SPARS
FREQUENCY

I ININ
o
97 0 7 0
' , I ,
.7075
,, ,
.7080
,,,,
"'i
ININtNJN'
.7085
J I t t~
.7090

87 86
Sr/ Sr

FIG. 10. The Sr isotopic composition of Lincolnshire Limestone brachiopods and burial cements.

probably inappropriate. In corroboration, it has support this. At least the following possibilities
already been stated that major clay and clastic exist to account for this: (i) Clay mineral
mineral diagenesis is probably inappropriate at transformation and clastic mineral breakdown
the burial depths and thermal regime exper- reactions in bounding the intraformational
ienced by the Lincolnshire Limestone in the shales were unimportant at these elevated tem-
study area. peratures. (ii) Mineral sinks other than burial
However, in the more deeply buried area to the calcites are preferable to the cations of interest at
east, the Limestone may have attained tempera- these temperatures. (iii) Cations generated from
tures as high as 70~ with similarly elevated the shales at these temperatures enter the
temperatures affecting the bounding shale for- Limestone porewaters but are circulated out of
mations. Under such circumstances, clay min- the Limestone without entering the study area.
eral transformation and clastic mineral break- In the absence of data on burial cement phases
down reactions may have been expected to further down-dip to the east it is impossible to
contribute ions to the burial porewaters of the constrain these possibilities.
Limestone. However, the present data on all A possible alternative source for Sr and Mg
elemental concentrations in the burial cements, may be another marine carbonate within the
particularly the Sr-isotopic information, does not stratigraphic column of the East Midlands. Such

0.71001

| Range of late spars ......

0.7090 ............ .........
~:~.x ~'~iii~........ ~ili~
87Sr 0.7080 '~

. . . . ..... V

0"7070-

- Brachiopoda

0.7060 i I
o 160 260 3b0 460 5oo 600
Million years b.p.

Range of Late Spars Range of Marine Waters (Burke et al., 1982)

FIG. I 1. The variation of 87Sr/86Sr in Phanerozoic marine waters (after Burke et al. 1982). Brachiopod and
late spar results from this study are illustrated.
214 D. E m e r y
87 86
S r / Sr
PROPORTION OF ADDED Sr NEEDED TO CHANGE
ADDED

O.719 87Sr/86Sr FROM 0 . 7 0 7 2 5 TO: 0.7080

o.,1o /t'/,
0.713

O.710
0.7085
0.7082
0.707 0.7080
O .2 .4 .6 .8 1.0
PROPORTION ADDED

FIG. 12. Proportion of additional radiogenic Sr required to change the 87Sr/86Sr from 0.70725 (marine
components) to burial calcite 87Sr/86Sr compositions.

material would satisfy the requirement of an quoted problem of carbonate supply to buried
Sr/Mg ratio which is near-identical to Bajocian Limestones may also be overcome by these
marine carbonate. Figure 11 illustrates the sources.
evolution of the 87Sr/86Sr of marine waters with
Phanerozoic time (Burke et al. 1982). The range
in the strontium isotopic composition of the A scheme for trace-element mobility
burial calcites has been added to this curve. It is
evident from this construction that Carbonifer- during Lincolnshire Limestone burial
ous marine carbonate and other Lower Palaeo- diagenesis
zoic marine carbonates may have an STSr/86Sr
ratio which is sufficiently elevated to account for Figure 13 depicts the essential elements of the
the Sr isotopic composition of the calcite. model. Outstanding problems remain with re-
A further constraint which must be satisfied is gard to the identity of the Palaeozoic carbonate
one which allows the absolute concentrations of contributing Sr and Mg, and its essential
Sr and Mg in the burial calcites to show hydraulic continuity with the buried Lincoln-
considerable variation whilst maintaining a shire Limestone. The most obvious and proximal
broadly constant Sr isotopic composition. This source of Palaeozoic carbonate in the East
can also be satisfied by remobilizing Palaeozoic Midlands stratigraphic column is Carboniferous
marine carbonate, in that a large variation in the (Dinantian) marine carbonate, which satisfies all
contributions of Bajocian and Palaeozoic car- the requisite chemical and isotopic criteria.
bonate will still maintain the required Sr isotopic Seismic sections across the East Midlands Basin
composition of the burial calcites (Fig. 12: see into the southern North Sea demonstrate the
also Emery 1986, Emery et al. 1987). presence of normal faults cutting both Carboni-
Remobilization of older carbonate may also ferous Limestone and Lincolnshire Limestone.
account for a component of the calcium carbon- On burial, these faults would probably have
ate required by the burial cements. This mass- acted as major fluid conduits, as fluid expelled
balance problem (Bathurst 1975) is a recurrent from the sedimentary sequence attempted to
one in carbonate diagenesis; remobilized car- bleed upwards along zones of high permeability.
bonate from elsewhere in the stratigraphic A tentative suggestion is that fluid migration up
column on the scale envisaged above may be one faults comprises a cross-formational limb of a
solution to the difficulty. convective system. Once in the permeable
In summary, it therefore appears possible to Lincolnshire Limestone, the fluid would have
account for all the criteria pertinent to the Mg migrated up-dip, hence an enformational limb to
and Sr chemistry, and the Sr isotopic compo- the system which can also be demonstrated by
sition of the Lincolnshire Limestone burial the oxygen isotopic composition of the burial
cements by invoking contributions of these trace cements (Emery 1986). Briefly, as Fig. 4 demon-
elements from host Bajocian marine carbonate strates, the temperatures recorded by the precipi-
and Palaeozoic marine carbonate. The oft- tated ferroan calcite phase during maximum
 Limestone burial diagenesis, E. England 215
Approx 200km

(o) x7 V

t I,?
l
-]
Kimm. Clay - -
Top Lincs. Lstn.

t
Top Grantham Fm --
Top Carb. Lstn

~~'~,,
I

I1
I
,i, 'r, Car'
I I
,,,, b
I

I i 1 I
I
i_. st', \n',,', . "I?!:~Fi Z , : ~ : , , ~
I -- T

,
I
,
I
,
F

I'1
r~

I / I llll~l III III

I rll
~_
I I IIIII
I I
II
Itl I
I
, g Ill
;, 1,, ;, ,,.
(b) V

I ~;;..::.!,.:,;
Chalk - -
Top Lincs Lstn

19
g
FIG. 13. A model for trace element source and mobility during precipitation of Lincolnshire Limestone burial
cements. Small to large triangles represent increasing burial. Arrowheads represent increasing contributions of
trace elements. Light shading illustrates shale interval (Kirton Shale Member) within the Lincolnshire
Limestone. (a) Cation source for non-ferroan cements. (b) Cation sources for ferroan cements.

burial are considerably elevated over those due to nent would account for transfer of Fe and Mn
that conduction through the overlying sedimen- from the shales into the Lincolnshire Limestone
tary pile alone. The most effective way of aquifer during burial diagenesis.
enhancing precipitation temperatures above the
ambient conductive temperature is by up-dip
fluid flow through the Limestone; this is further
Conclusions
confirmed by the progressive down-dip ~so (i) Two phases of burial calcite are present
depletion of both burial calcite phases. Whether within the Lincolnshire Limestone, a non-
an enformational limb was present within the ferroan early phase and a later ferroan
Carboniferous Limestone during Lincolnshire phase.
Limestone burial calcitization is pure (ii) The Fe and Mn trace element component
speculation. of the burial cements is considered to be
Another component of cross-formational flow derived from Fe and Mn oxy-hydroxides
incorporated into the model is from bounding associated with bounding and intraforma-
(and intraformational) shales, chiefly the under- tional shales. Burial depths and tempera-
lying Grantham Formation and to a lesser extent tures associated with Lincolnshire Lime-
the Upper Estuarine 'Series'. This flow compo- stone are too low to account for derivation
216 D. Emery
of these cations from clastic or clay mineral sufficient to provide the necessary calcite
diagenesis. of the burial cements, thereby overcoming
(iii) Redox conditions during burial are be- the recurrent problem of carbonate source
lieved to have been consistently reducing, during burial diagenesis.
permitting the mobilization of Fe and Mn (ix) A speculative fluid circulation system
as divalent cations and their subsequent implied by the cation source evidence
incorporation into the burial calcites. requires fluid movement from the Carbon-
(iv) Increasing Fe and Mn contents of the later iferous Limestone up bleeder faults (the
ferroan calcite phase was the result of cross-formational limb of a convective
increasing cross-formational flow from the system) into the Lincolnshire Limestone
bounding and intraformational shales on (the enformational limb of the system). A
increasing burial. further cross-formational compactional
(v) Sulphate was probably largely absent from flow component is provided by fluid flow
the burial diagenetic regime. from shales bounding (and internal to) the
(vi) Mg and Sr concentrations in both the Limestone.
burial calcites display a positive linear
relationship. No absolute increase in the ACKNOWLEDGMENTS: The author is grateful to the
concentration of either cation is observed following institutions for their time and use of
in the burial calcites with time. equipment: The Department of Geology, King's
(vii) The 87Sr/86Sr ratio of both burial calcites is College, London; The Institute for Energy Technology,
indistinguishable (=0.70820). This isoto- Kjeller, Norway; The National Laboratory for Mass-
pic composition is considerably more Spectrometry, University of Oslo, Norway.
radiogenic than host Bajocian marine The author would also like to thank the following
carbonate. individuals for technical assistance and useful discus-
sion: Dr J. A. D. Dickson, Dr P. C. Smalley, Dr N.
(viii) The Sr and Mg of the burial calcites is Walsh, Cathy Lewin, Carol Birney, Alison Searl and
believed to have been derived from (a) Julian ('Sid') Fowles.
remobilized Bajocian marine carbonate I am grateful for the support of a NERC studentship
and (b) remobilized Carboniferous marine held at the University of Cambridge, and wish to thank
carbonate which mixed in variable propor- Texaco Inc., Houston, for support given to the
tions. These carbonate sources may be Cambridge carbonate group.

References
ANDREW-SPEED, C. P., OXBURGH, E. R. & COOPER, Geological Society of London Special Publication,
B. A. 1984. Temperature and depth dependent 18. Blackwell Scientific Publications, Oxford.
heat flow in the western North Sea. Bulletin of the CARROLL, D. 1959. Ion exchange in clays and other
American Association of Petroleum Geologists, 68, minerals. Bulletin of the American Association of
1764-81. Petroleum Geologists, 70, 754-79.
ASHTON, M. 1977. The stratigraphy and carbonate CRAVE, K. E. 1954. Aspects of the biochemistry of
environments of the Lincolnshire Limestone (Ba- magnesium. Journal of Geology, 62, 266-83.
jocian) in Lincolnshire and parts of Leicestershire. CURTIS, C. D. 1977. Sedimentary geochemistry:
Unpublished PhD thesis, University of Hull. environments and processes dominated by the
- - 1980. The stratigraphy of the Lincolnshire Lime- involvement of an aqueous phase. Philosophical
stone Formation (Bajocian) in Lincolnshire and Transactions of the Royal Society of London, 286A,
Rutland (Leicestershire). Proceedingsof the Geolo- 353-72.
gists' Association, 91, 203-23. --1983. Geochemical studies on the development
BATHURST, R. G. C. 1975. Carbonate Sediments and and destruction of secondary porosity. In:
their Diagenesis. Elsevier, Amsterdam. BROOKS, J. (eds). Petroleum Geochemistry and
BERNER, R. A. 1980. Early Diagenesis, a Theoretical Exploration of Europe, pp. 113-25. Geological
Approach. Princeton University Press. Society of London Special Publication, 12. Black-
BURKE, W. H., DENISON,R. E., HETHERINGTON,E. A., well Scientific Publications, Oxford.
KOEPNICK, R. B., NELSON,H. F. & OTTO, J. B. - - & COLEMAN,M. 1986. Controls on the precipita-
1982. Variation of seawater 87Sr/86Sr throughout tion of early diagenetic calcite, dolomite and
Phanerozoic time. Geology, 10, 516-19. siderite concretions in complex depositional se-
BURLEY, S. D., KANTOROWICZ,J. D. & WAUGH, B. quences. In: GAUTIER,D. L. (ed.). Roles of Organic
1985. Clastic diagenesis. In: BRENCHLEY,e. J. & Matter in Sediment Diagenesis, pp. 23-34. Society
WILLIAMS, B. P. J. (eds). Sedimentology: Recent of Economic Paleontologists and Mineralogists
Developments and Applied Aspects, pp. 189-226. Special Publication, 38.
 Limestone burial diagenesis, E. England 217
DICKSON, J. A. D. 1966. Carbonate identification and geology. Geochimica et Cosmochimica Acta, 27,
genesis as revealed by staining. Journal of Sedi- 1209-64.
mentary Petrology, 36, 491-505. MOORE, C. H. & DRUCKMAN,Y. 1981. Burial diagene-
DOWNING, R. A., SMITH, D. B., PEARSON, F. J. sis and porosity evolution, Upper Jurassic Smack-
MONKHOUSE,R. A. & OTLET, R. L. 1977. The age over, Arkansas and Louisiana. Bulletin of the
of groundwater in the Lincolnshire Limestone, American Association of Petroleum Geologists, 65,
England, and its relevance to the flow mechanism. 597--628.
Journal of Hydrology, 33, 201-16. OGLESBY, T. W. 1976. A model for the distribution of
EDMUNDS, W. M. & WALTON, N. R. G. 1983. The manganese, iron and magnesium in authigenic
Lincolnshire Limestone--hydrochemical evolu- calcite and dolomite cements in the Upper
tion over a ten-year period. Journal of Hydrology, Smackover Formation in East Mississippi. Unpub-
61, 201-11. lished MSc thesis. University of Missouri.
EMERY, D. 1986. The diagenesis of the Lincolnshire OLDERSHAW, A. E. & SCOFFIN, T. P. 1967. The source
Limestone (Bajocian) in Lincolnshire. Unpub- of ferroan and non-ferroan calcite cements in the
lished PhD thesis. University of Cambridge. Halkin and WeDlock Limestones. Journal of
--, DICKSON,J. A. D. & SMALLEY,P. C. 1987. The Geology, 5, 309-20.
strontium isotopic composition and origin of PEACH, n. 1984. Some aspects of the hydrology of the
burial cements in the Lincolnshire Limestone Lincolnshire Limestone. Unpublished PhD thesis.
(Bajocian) of central Lincolnshire, England. Sedi- University of Birmingham.
mentology, 34, 795-806. ROSLER, H. J. & LANGE, H. 1972. Geochemical Tables.
FAIRCHILD, I. J. 1983. Chemical controls of cathodo- Elsevier, Amsterdam.
luminescence of natural dolomites and calcites: SCHOFIELD, K. & ADAMS, A. E. 1986. Burial dolomiti-
New data and review. Sedimentology, 30, zation of the Woo Dale Limestones Formation
579-83. (Lower Carboniferous) Derbyshire, England.
FAURE, G. & SZABO, Z. 1986. Isotopic studies of Sedimentology, 33, 207-19.
mineral cements in North American Sandstones: SCHOLLE, P. A. & HALLEY, R. B. 1985. Burial
the Beria Sandstone of Ohio. In: RODRIGUEZ diagenesis--out of sight, out of mind ! In: SCHNEI-
CLEMENTE, R. & FENOLL HACH-ALI, P. (eds). DERMANN, N. & HARRIS, P. M. (eds). Carbonate
Abstracts, Geochemistry of the Earth Surface and Cements, pp. 309-34. Society of Economic Paleon-
Processes of Mineral Formation. Granada. tologists and Mineralogists Special Publication,
FROELICH, P. N., KLINKHAMMER, G. P., BENDER, 36.
M. L., LUEDTKE, N. A., HEATH, G. R., CULLEN, SMALLEY, P. C., RAHEIM, A., DICKSON, J. A. D. &
n., DAUPHIN, P., HAMMOND,D., HARTMAN,B. & EMERY, n . 1985. A strontium isotopic study of the
MAYNARD, V. 1979. Early oxidation of organic Jurassic Lincolnshire Limestone Formation, Eng-
matter in pelagic sediments of the eastern Equa- land. American Association of Petroleum Geologists.
torial Atlantic: suboxic diagenesis. Geochimica et Research Conference Abstracts. New Orleans,
Cosmochimica Acta, 43, 1075-90. 1986.
GAVISH, E. & FRIEDMAN, G. M. 1969. Progressive STUEBER, A. i . , PUSHKAR,P. & HETHERINGTON,E. A.
diagenesis in Quaternary to late Tertiary carbon- 1984. A strontium isotopic study of Smackover
ate sediments: sequence and time scale. Journal of brines and associated solids, Southern Arkansas.
Sedimentary Petrology, 39, 1-32. Geochimica et Cosmochimica Acta, 48, 1637-49.
GLENNIE, K. W. & BOEGNER, P. L. E. 1981. Sole Pit TEN HAVE, T. & HEIJNEN, W. 1985. Cathodolumines-
inversion tectonics. In: ILLING, L. V. & HOBSON, cence activation and zonation in carbonate rocks:
G. D. (eds). Petroleum Geology of the Continental an experimental approach. Geologie en Mijnbouw,
Shelf of North-West Europe, pp. 110-20. Heyden, 64, 297-310.
London. THOMPSON, A. & WALSH, N. 1983. A Handbook of
IRWIN, H., CURTIS, C. D. & COLEMAN, i . 1977. Inductively Coupled Plasma Spectrometry. Blackie,
Isotopic evidence for the source of diagenetic London.
carbonate during burial of organic rich sediments. TUCKER, M. J. 1985. Shallow-marine carbonate facies.
Nature, 269, 209-13. In: BRENCHLEY,P. J. & WILLIAMS,B. P. J. (eds).
KINSMAN, D. J. J 1969. Interpretation of Sr 2+ concen- Sedimentology : Recent Developments and Applied
trations in carbonate minerals and rocks. Journal Aspects, pp. 147-69. Geological Society of London
of Sedimentary Petrology, 39, 486-508. Special Publication, 18. Blackwell Scientific Pub-
MARSHALL, J. D. & ASHTON, M. 1980. Isotopic and lications, Oxford.
trace-element evidence for submarine lithification VEIZER, J. 1983. Chemical diagenesis of carbonates:
of hardgrounds in the Jurassic of eastern England. theory and application of trace-element tech-
Sedimentology, 27, 271-89. nique. In: Stable Isotopes in Sedimentary Geology.
MCINTIRE, W. L. 1963. Trace element partition co- Society of Economic Paleontologists and Minera-
efficients--a review of theory and application to logists. Short Course, 10.

D. EMERY, Department of Earth Sciences, University of Cambridge, Downing Street,

Cambridge CB2 3EQ, UK. Present address: Sedimentology Branch, BP Petroleum
Development, Britannic House, London EC2Y 9BU, UK.
 Wilcox sandstone diagenesis, Texas Gulf Coast" a regional
isotopic comparison with the Frio Formation
Lynton S. Land & R. Stephen Fisher

S U M M A RY : The burial diagenesis of Eocene Wilcox sandstones differs significantly

from the burial diagenesis of other Gulf Coast terrigenous clastic formations. Differences
between the onshore Wilcox and Oligocene-Miocene Frio Formations are larger than
regional variations within either formation. Younger, offshore units have undergone less
diagenetic alteration than either of the older, onshore units, whereas Mesozoic sandstones
are generally more extensively altered.
Cementation by quartz and calcite was the first diagenetic modification of volumetric
significance to affect either the Wilcox or Frio Formations. The average 6180 of Wilcox
quartz cement is approximately + 25%0SMOW, in contrast to + 31%ofor the Frio. In both
formations, quartz cement is more abundant in the overpressured zone. Calcite has an
average 6180 of - 10.8%oPDB in the Wilcox Formation (compared to -7.2%0 for the Frio).
87Sr/86Sr values suggest that calcite was derived from coeval nannofossils, carbonate rock
fragments of Mesozoic age, or by local input of fluids from underlying Mesozoic carbonates.
Most calcite cement (and therefore quartz which pre-dates calcite cement) was apparently
emplaced prior to extensive alteration of detrital silicates which would have released
strontium with a high 87Sr/86Sr ratio.
The most common carbonate cement in the Wilcox Formation is ankerite which replaces
calcite with increasing depth. The 613C of ankerite is essentially identical to that of calcite at
the same depth, and is also similar to Frio and younger carbonate cements. In contrast, 6180
values of Wilcox ankerite (average -9.8%0) indicate emplacement at higher temperature or
from more depleted water than was true of Wilcox calcite. 87Sr/86Sr values for Wilcox
ankerite are very radiogenic (>0.7100), indicating that ankerite emplacement occurred
during or after active silicate diagenesis. Both ankerite and dolomite are very uncommon
phases in Frio and younger sandstones despite massive smectite stabilization, suggesting
that the conversion of smectite to illite was not the source of iron and magnesium for the late
cements.
The volume of secondary porosity is similar in both Wilcox and Frio sandstones.
Albitization and K-feldspar removal from both formations are essentially complete below
3000 m and essentially no unaltered detrital feldspar occurs below that depth.
Differences between the Gulf Coast formations are attributed to different geothermal
gradients, differences in the basinal sediments over which the units prograded, and to the
changing nature of connate fluids in the units themselves. Diagenesis of the Wilcox
Formation, like other Gulf Coast terrigenous wedges, is understandable only in terms of
interaction with underlying units during the large-scale evolution, in both time and in space,
of the Gulf Coast diagenetic system.

The Eocene Wilcox and Oligocene-Miocene analysis of Wilcox cements and compare cemen-
Frio Formations constitute the two largest clastic tation conditions in the Wilcox and Frio
units beneath the on-shore Texas Gulf coastal Formations.
plain (Fig. 1). The extensive collection of The emphasis of this paper, as in the regional
sandstones amassed by the Burea of Economic study of the Frio Formation, is on the volumetri-
Geology, University of Texas at Austin, as part cally important diagenetic reactions: quartz
of the geothermal energy programme (Loucks et cementation, carbonate cementation, feldspar
al. 1984) formed the basis for diagenetic studies stabilization, and secondary porosity develop-
of the Frio Formation (Land & Milliken 1981, ment. Clay cements other than kaolinite are of
Milliken et al. 1981, Land 1984). This same lesser volumetric significance in most Wilcox
collection formed the basis for this study, sandstones, and clays could not be isolated from
supplemented by detailed study of the central the epoxy-impregnated samples available to this
part of the Texas Gulf Coast (Fisher 1982, Fisher study.
& Land 1986) and previously published data In addition, we attempt to explain the ob-
(Boles 1978, Boles & Franks 1979, Boles 1981). In served regional differences between the diagene-
this paper we present the results of isotopic sis of Wilcox and younger formations. We briefly

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 219

Geological Society Special Publication No. 36, pp. 219-235.
220 L. S. Land & R. S. Fisher

NW SE
COASTLINE) _ _
0
",,
~ "-.
Io ~ k,.--., ": GULF
2 U "- OF
~ ~ '......~ MEXICO

KILOMETERS'
6,

.......

8
"%",.,,.
. . . . ". . . . . . . . . . EOCENE

~ " ' " PALEOCENE

10 "~"~~',,,,~ U. CRETACEOUS

CRETACEOUS
'~--,t2;
"-~-~RA S SiC

FIG. 1. Very schematic cross-section through the central part of the Texas Gulf Coast. We have modified the
cross-section constructed by Bebout et al. (1982) by extrapolating the Gulf basinal sediments documented by
Winker & Buffler (in press) beneath the continental Tertiary section. The (locally drastic) effects of salt and
growth-fault tectonics have been ignored. This cross-section schematizes the relationship between thick sands
(stippled) and growth faults, and between the offlapping Wilcox and Frio Formations and the thick underlying
basinal Mesozoic marls, especially when the 23 x vertical exaggeration is taken into account.

NORTHERN
TEXAS

CENTRAL
TEXAS ~N

SOUTHERN
0 4O 80
TEXAS Mdes

FIG. 2. Location of wells from which samples were obtained for this study. Arbitrary boundaries between
'northern', 'central' and 'southern' Coastal Gulf Texas are also shown.
 Wilcox sandstone diagenesis, Texas Gulf Coast 221

compare the Wilcox Formation to underlying Diagenetic reactions

and overlying units to relate differences to the
overall diagenetic evolution of the Gulf Coast Diagenetic reactions are discussed in the tempo-
clastic wedge. ral order in which they occurred in the sand-
stones as determined by petrographic relation-
ships. Quartz, calcite and kaolinite precipitation
Methods preceded the development of secondary porosity,
some of which was subsequently occluded by
Thin section stubs of 75 samples from 29 wells ankerite and another generation of kaolinite.
(Fig. 2) formed the sample base. Strontium for
isotopic analysis was isolated from carbonate
cements by reacting washed, powdered rock
Quartz cementation
samples in 0.2 M HC1 for less than five minutes, The isotopic composition of Wilcox quartz
to minimize leaching of strontium from the overgrowth cements was determined by bulk
silicate phases, followed by conventional cation analysis of the sand-sized quartz fraction of
exchange chromatography. Strontium isotope sandstones having varying amounts of quartz
ratios were normalized to 0.7080 for the Eimer cement (Fig. 3). Data are presented in Table 1
and Amend standard. Analytical precision is (sandstones from which both quartz and feld-
better than 0.0002. Otherwise, the analytical spars were analysed isotopically), Table 2 (sand-
techniques used were identical to those outlined stones analysed solely for quartz), and in Fisher
by Land (1984). The same cautions which were & Land (1986).
advanced in that study with regard to the Wilcox sand containing no quartz over-
possible biases introduced by the use of this growths, or sand from which quartz overgrowths
sample suite apply to this study as well. Wells are were removed with HF (Qz sand - - cmt in Table
from areas of active hydrocarbon exploration 1) ranges from + 10.2 to + 16%o in 6180 (Fig. 3).
and areas of geothermal prospects, and may not This range is larger than is characteristic of
be representative of Wilcox sands from other younger units (10.6 to 14%o), but the average
areas. Sampling of the available core may also value of 13%0 for detrital Wilcox quartz is similar
have been biased toward the more altered for all Tertiary units we have analysed (Gold
(cemented or porous) rocks. Samples shallower 1984, Land 1984, Milliken 1985).
than about 1500 m are generally unavailable If Wilcox quartz sand is assumed to have a
from either formation. 61so of + 13%o, then statistical analysis of all

.50

~ §
overgrowths
total quartz .20

*
.i0

,r162162 ,~,,r A ,l:." d

lo i2 ~4 I~ i8
(~180 Quartz (%0)

FIG. 3. The fraction of quartz present as quartz overgrowths, determined from point-counts, plotted versus
the 6]80 of sand-sized quartz. Points along the bottom of the diagram include both sands having no
overgrowths, and values obtained by analyzing the residue after 'stripping' quartz sands of their overgrowths
with HF. Short line segments extrapolate between quartz sand including overgrowths (Qz sand), and quartz
sand minus cement (Qz sand - - cmt) from the same samples to values for pure overgrowths. The
extrapolation from average sand having a 6180 of 13%oto pure overgrowths (25%0+ 4%0) includes data from
this study (Tables 1 and 2), and from Fisher & Land (1986).
222 L. S. Land & R. S. Fisher
TABLE 1. Summary of quartz and feldspar isotopic analyses
Depth 6180 6180 6180 6180 Per cent 6180 Per cent Per cent 87Sr/S6Sr
(m) total qz + qz qz qz feldspar* feldspar ab feldspar
qz feldspar sand sand cmt
sand ----cmt

2357 15.2 15.7 16.0 10.2 11.8 7.1 67 0.7227

3008 15.1 17.1 15.9 81 0.7142
3031 15.7 16.8 16,2 15.7 10.4 21.2 11.9 90
3194 16.0 17.6 17.5 15.9 26.3 18.0 19.0 99 0.7127
3363 14.7 15.1 15.1 13.3 9.6 15.1 26.1 97
3445 14.7 16.1 15.8 13.8 16.3 19.4 8.3 97
3656 14.6 15.5 15.0 14.3 10.2 16.8 28.0 98
3976 15.0 15.6 15.1 I 1.1 22.4 6.8 98
4006 14.7 15.9 16.2 16.0 9.0 14.1 14.5 98
4112 14.7 16.1 15.4 14.7 23.8 19.1 18.8 98 0.7146
4297 14.6 16.7 16.2 15.6 14.9 19.7 14.2 99 0.7166
4344 13.3 15.1 14.6 13.5 15.7 17.1 20.2 98 0.7163

* Calculated.
6180 values are in %0relative to SMOW. Total quartz was determined after fusion of the whole rock in NaHSO4 and removal of
feldspars in H2SiF6 (Syers et al. 1968). Fine-grained quartz from rock fragments is thus included. In order to exclude such fine-
grained quartz, the > 62 ~tm (sand-sized) fraction of an aliquot of the rock was isolated by sieving following NaHSO 4 digestion,
analysed for 6180 before removal of feldspar (qz + feldspar sand), and after removal of feldspars (qz sand). The 6180 feldspar was
calculated from these two values and the % feldspar obtained by electron microprobe analysis of fused beads of the qz + feldspar
sand fraction. Quartz overgrowths were then stripped from the qz sand fraction in HF, to yield samples of quartz sand presumably
representative of the sand before cementation (qz sand - - cmt). From these values, and the % qz cmt determined by point counts,
the r of the overgrowths was calculated by material balance.

available data permits extrapolation to an aver- individual samples of their quartz overgrowths
age value for Wilcox quartz overgrowths of (see Milliken et al. 1981, for the procedure) are
+ 25%0 ( + 4%0, SMOW). An average of + 24%0 averaged.
( + 5%0) is obtained if only those values (Table 1 ; Interpretation of quartz cementation in both
Fisher & Land 1986) obtained by stripping the Frio and Wilcox Formations is complicated
by the fact that the cements are zoned. Alternat-
ing zones of subtle orange luminescence with
TABLE 2. Summary of isotopic analyses of quartz euhedral outlines are barely visible through the
Depth 6'sO total 6'80 quartz % quartz light microscope when excited by 15 kV electrons
(m) quartz sand cement on the electron microprobe, and document a
complex history of pore-filling. The cause of the
1591 12.7 13.6 0
1619 11.1 (12.0) 4.0
luminescent zoning and whether or not it is
1819 14.1 15.1 0.5 accompanied by isotopic zoning, is unknown.
2552 14.3 15.0 1.3 Some euhedral overgrowths terminate in pore
2572 13.8 14.7 0 space and therefore quartz overgrowth cementa-
2603 14.1 14.9 4.1
2656 14.7 15.1 2.9
tion could have taken place over an extended
2718 13.2 14.1 0 period of time, and minor amounts could even be
2750 13.5 11.1 taking place from contemporaneous fluids.
2796 15.5 16.3 3.6 Despite more variability in data from the
2914 15.4 (16.3) 16.0
2943 14.7 15.3 10.9
Wilcox Formation as compared to the Frio
2950 14.7 15.6 5.2 Formation, it is clear that Wilcox quartz over-
2975 14.7 (15.6) 20.0 growths are significantly depleted in ~sO relative
2998 14.2 15.6 1.8 to the value of + 31%o (Table 3) obtained from
3035 13.8 14.7 2.0
3045 14.7 15.6 5.0
Frio sandstones (compare Fig. 3 with fig. 3 in
3270 15.5 16.4 5.0 Land 1984). Therefore Wilcox quartz over-
3350 16.2 24.0 growths must have formed at a somewhat
3421 15.7 (16.6) 22.0 different temperature and/or from water of
3429 14.9 2.2
3591 17.0 21.5
somewhat different 6~80 than was observed in
3658 14.1 (15.0) 30.0 the Frio Formation.
4608 14.3 (15.2) 30.0 The average amount of quartz cement in
Wilcox sandstones (4.6 %) is approximately twice
Notation as in Table 1. Values for quartz sand in parentheses
were calculated from the average difference (0.92%0) between
that found in average Frio sandstones (2.6%).
total quartz and quartz sand for samples (n = 25) from which The average percentage of quartz cement in-
both analyses were available. creases with depth (Fig. 4), as is also true of the
 Wilcox sandstone diagenesis, Texas Gulf Coast 223
TABLE 3. Comparisonof Wilcox and Frio diagenetic phases
Wilcox Frio

Quartz cement
Volume ~ of rock 4.6 + 2.5 (462) 2.6 + 1.4 (645)
6180 SMOW 25 + 4%o (60) 31 + 1.5%(27)
Carbonate cement
Dominant mineralogy Ankerite Calcite
calcite of total carbonate 13 88
Other minerals Calcite Dolomite
Volume ~ of rock 3.3 + 1.2 (462) 5.3 + 2.9 (445)
6180 calcite PDB - 10.8 + 2.1%o (21) - 7 . 2 + 1.1%o (68)
51sO ankerite PDB - 9 . 8 + 1.1%o (70) --
513C carbonate PDB - 4 . 6 + 2.0%0 (91) - 4 . 0 _+ 2.6%0 (68)
Secondary porosity
Volume % of rock 10 12
Primary feldspar
total feldspar of rock 23 27
% K-feldspar of total feldspar 40 30
~ A n in unaltered plagioclase 5 20
Diagenetic feldspar
31aO albite SMOW 17 + 3.0%0 (11) 17%o

+ values are given as one standard deviation, with the number of samples in parenthesis. Data from
the Wilcox Formation illustrated in Figs 3 to 8, and presented in Tables 1 and 2. Data from the Frio
Formation from Milliken et al. (1981) and Land (1984).

Frio Formation9 In the case of the Frio Forma- (fig. 5 in Land 1984)9 In the Wilcox Formation,
tion, an increase in the degree of quartz the relationship between the degree of quartz
cementation corresponds approximately with cementation and overpressure is not as sharp,
the top of the present-day overpressured zone but the gross relationship still exists. As in the

% Quartz Cement
0 5 10 15 2 0 25 50

9_ , ~

mean = 46 %

thousand
feet lo
8'
~i " :~""
**
"1 g~

"

. ,
~

9 ". .

12" ' :

14..." 9 ....

FIG. 4. P e r c e n t a g e quartz c e m e n t versus depth. Q u a r t z c e m e n t a t i o n , as in the Frio F o r m a t i o n , is more

c o m m o n below m o d e r n geopressure which ranges between a b o u t 2400 m (8000 feet) a n d 3900 m (13 000) feet)
in the study area. T h e line is a m o v i n g average o f the raw data o f Loucks et al. (1984).
224 L. S. Land & R. S. Fisher
2.5 presence of ankerite as shallow as 1800 m and
calcite as deep as 3300 m document a more
gradational boundary between the two types of
Total Carbonate cement in this regional study than was observed
1.5 at 2700 m by Boles (1978) in a three county area
in South Texas. We agree with Boles that
to ankerite characteristically replaces calcite in the
tO deeper sands. We do not agree that a 'reaction
front' occurs regionally over a narrow depth
o 0.5 interval which corresponds to the transformation
c-
O of smectite to illite. Nor do we agree on the
source of components for the ankerite cement.
Electron microprobe analyses of Wilcox cal-
cite cements indicate that they are relatively
pure, containing more than 98 ~o CaCO3, similar
E to Frio calcite cements. Variations in compo-
1.5
C sition are small, and do not correlate with depth.
As in the case of calcite cements in the Frio
o l Formation, luminescence is uniform and zoning
is rarely observed. No systematic variation in the
0.5 degree of carbonate cementation as a function of
depth is apparent (Fig. 6). In this respect as well,
Frio and Wilcox sandstones are quite similar, but
0 5 10 15 2O 25 50 the average amount of carbonate cement in
Wilcox sandstones (3.3 ~o) is slightly less than the
% Cement
amount found in average Frio sandstones (5.3 ~o
FIG. 5. Log of the number of point-counts versus the - - Table 3).
percentage total carbonate (top) and quartz (bottom) Oxygen isotopic values for Wilcox calcite
cement in Wilcox sandstones. Unlike the Frio average -10.8%o (Fig. 7), compared with an
Formation, quartz cementation is more uniform in average of - 7.2%0 observed for the Frio Forma-
Wilcox sandstones. Like the Frio, however, most
sands, especially the matrix-rich ones, are tion (Table 3). As in the case of the Frio
uncemented, and very few sands are pervasively Formation (fig. 10 in Land 1984), no significant
cemented. Raw data from Loucks et al. (1984). relation between depth and 5180 is observed on a
regional basis.
Carbon isotopic values for Wilcox calcite and
Frio Formation, most sands, especially matrix- ankerite cements are indistinguishable from
rich ones, contain little quartz cement, and very each other (Fig. 8), and are similar to those found
few sands in either Formation are extensively in the Frio Formation (Table 3). Values between
quartz-cemented (Fig. 5). The distribution of - 2 and -6%0 are typical of deeper sandstones,
quartz cement in Wilcox sands is somewhat more whereas shallower sandstones range between
uniform than in Frio sands (compare Fig. 5 with about -3%0 and -14%o. The reason for this
fig. 2 in Land 1984). More uniformly cemented distribution of values in both formations is not
rocks are produced as cementation proceeds, entirely clear, and several alternative scenarios
probably because as cement begins to occlude can be constructed (Lundegard & Land 1986). If
porosity and reduce permeability, fluid flow is the source of CaCO3 for sandstone cements is
directed into more permeable (less cemented and carbonate nannofossils in contemporeous shales,
matrix-rich) sands. More intense early cementa- or CaCO 3 from older formations as suggested by
tion by a more dynamic flow system, or a longer the strontium isotopic compositions, then calcite
duration of cementation, may account for the cements should have a 513C near 0%0. Because
more uniform and more intense quartz cementa- the cements are depleted in 13C with respect to
tion in the Wilcox as compared to the Frio inorganic carbon, between l0 and 5 0 ~ organic
Formation. carbon having a ~13C of - - 2 6 % 0 must have been
introduced into the diagenetic system in order to
yield the measured t~13C values. The exact
Calcite cementation organic reaction responsible cannot be identified
with assurance, however. Lundegard et al. (1984)
Calcite is the dominant carbonate cement in showed that decarboxylation reactions could not
shallow Wilcox sandstones, whereas ankerite is provide sufficient CO2, or provide CO2 of the
more characteristic of deeper sandstones. The appropriate isotopic composition in the correct
 Wilcox sandstone diagenesis, Texas G u l f Coast 225

% Carbonate Cement
0 5 10 15 2 0 25 :50

t i" "~ " . ,

8
thousand
feet lo t"'.'...
~
..~ ". ~
~ 9~
i". -"

12

14
~ ~ ~

FIG. 6. Per cent carbonate cement (calcite + ankerite) versus depth. Similar tO the Frio Formation, no
systematic relationship with depth is observed. The line is a moving average based on the raw data of Loucks
et al. (1984).

proportions. Hydrous pyrolysis in underlying lower than that of Eocene sea water does not
formations might supply the carbon found in appear to be sufficiently large to account for the
carbonate cements (-6%0) as the result of the measured values. Most analyses of bulk feldspar
release of 13C-depleted methane (-46%0) from or clay from the Gulf Coast yield 87Sr/86Sr ratios
kerogen (-26%0). Simultaneous oxidation of greater than about 0.71 (Perry & Turekian 1974,
kerogen and reduction of ferric iron and/or Morton 1983, Table 1 and work in progress).
sulphate might also have released 13C-depleted Some calcite samples in the Wilcox and Frio
H2CO3 into the evolving pore water. Decompo- Formations are slightly more radiogenic (en-
sition of dissolved acetate (Carothers & Kharaka riched in 87Sr) than Eocene sea water. Therefore,
1978) is another possible source for 13C-depleted some calcite emplacement could have taken
carbon in the shallower samples. place from solutions in which silicate reactions,
Strontium isotopic analyses of Wilcox calcite such as the transformation of smectite to illite, or
cements range from 0.7070 to 0.7091 compared feldspar dissolution or replacement, had begun.
to values for Eocene sea water between 0.7076 The generally non-radiogenic nature of Wilcox
and 0.7078 (Burke et al. 1982). The measured (and Frio) calcite indicates that either extensive
values strongly suggest that most of the CaCO 3 silicate diagenesis had not yet taken place at the
has been derived from coeval nannofossils, from time calcite was emplaced, or that pore water
Mesozoic carbonate rock fragments (rare in strontium was dominated by strontium derived
these rocks), or from fluids equilibrated with from the dissolution of contemporaneous nanno-
underlying Mesozoic strata (Fig. 9). We cannot fossils, Mesozoic rock fragments, or formation
account for 87Sr/S6Sr ratios less than those waters from underlying units (Morton & Land
characteristic of Eocene sea water (about 0.7077 1987).
- - Burke et al. 1982) unless material has been
derived from older marine carbonate or evapor-
Feldspar diagenesis
ite formations. Young volcanic debris which
might contribute strontium with low 87Sr/86Sr The conversion of all detrital feldspar to albite,
ratios is rare in Wilcox sandstones. Based on or its dissolution to form secondary porosity, is
currently available data, the reservoir of stron- grossly similar in both Wilcox (Fig. 10) and Frio
tium from silicates having an 87Sr/86Sr ratio sandstones (fig. 11 in Land 1984). The zone in
226 L. S. Land & R. S. Fisher

x x
x
9 x
9 x
29 6
oe x x
c1 x 9 x
9 x~
X X9
(23
-o 8 2 (I)
"t3
2J- 8, ..-0-
o9 9 9 eI x "0 %: 9 22"
9 x ~ 9 I x
x 9 o'l 9
o x ~9 9 9 x x xt
o 10 e9 O '50
o O e o x 9 o)~ 9
X X
Xx
9 0
3 '-" 10
9 9 ell
x 0
o o~ x
0
0
, 9
0 x 3
t~0
.,,_,,

4 d) 12 Or)
fl)
...,+
14 v
'4
e9 0 9149149 9
14 9 o~ 9

16 9 eeee

-6 -8 4o -i2
6180 (%o) 16
0 -5 -10 -15
FIG. 7. 6]so of Wilcox calcite (crosses) and ankerite
(dots) cements versus depth. Calcite cements are & 13C (%o)
more variable in their isotopic composition
(standard deviation is 2.1%o)than ankerite (standard FIG. 8. ~13C of Wilcox calcite (crosses) and ankerite
deviation is 1.1%o). Average Wilcox calcite (-10.8%o) (dots) cements versus depth. As in the case of the
is depleted relative to Frio calcite cement (-7.2%0, Frio Formation, a minimum value of about -2%0 is
Table 3). Wilcox ankerite (-9.8%~ must have observed, and shallower cements tend to have a
formed under conditions significantly different from wider range in isotopic composition and to be more
Wilcox calcite, assuming a 3.8%0 fractionation depleted in ]3C.
between the CO2 extracted from calcite and the CO:
extracted from a co-genetic ankerite. Data include
those of Boles (1978) and Fisher (1982). The scatter
is not significantly reduced if present-day
temperature, rather than depth, is plotted. tions, a value of + 17%0 ( + 3 %0) will be assumed.
Wilcox and Frio authigenic albite (Milliken et al.
1981) is apparently similar in isotopic compo-
sition, suggesting that both formations may have
which K-feldspar is dissolved and plagioclase approached similar conditions of temperature
albitized extends between about 2500 and and 6180 water at the time of albitization.
3500m, approximately 500 m shallower than
was observed in the Frio Formation. The amount
Ankerite
of primary detrital feldspar in the two formations
is similar (Loucks et al. 1984, their figs 14 and The oxygen isotopic composition of Wilcox
16), the only major difference being that Wilcox ankerite cement ( - 9 . 8 + 1%o, Table 3) is less
plagioclase was initially much more albitic, variable than is true of calcite ( - 1 0 . 2 + 2%0).
having an average composition of An5 as T h e 6180 of ankerite cement, as in the case of
compared to An2 9 in the Frio. This difference calcite cement in both the Frio and Wilcox
probably reflects not only somewhat different Formations, displays no systematic variation
source rocks for the two formations, but the long with depth (Fig. 7). Our observations confirm the
time interval available for depositional recycling ionic compositional trend in ankerite, but not the
of Wilcox detritus on the craton prior to the decrease in 61 sO with increasing depth observed
Laramide orogeny. Table 1 presents data from locally by Boles (1978, his fig. 8). In contrast to
which 6180 values for authigenic albite have calcite cements in both the Frio and Wilcox
been calculated [6180 f e l d s p a r = ( ( 1 0 0 * 6 1 8 9 Formations, significant variations in ankerite
qz + feldspar sand]-((100-%feldspar)*6180 qz ionic composition do occur with increasing
sand))/~feldspar]. Recognizing the relatively depth. Based on statistical analysis of 32 sam-
large errors involved in these sorts of extrapola- pies, two trends have been found.
 Wilcox sandstone diagenesis, Texas Gulf Coast 227

.7095- I I I I
1

.7090-

.7085

87S~r .7080

.7075

.7070 -

.7065 1 ',1 '1 '1

TERTIARY CRETACEOUS JURASSIC

FIG. 9. 87Sr/S6Sr of Wilcox (black bars) and Frio (open bars) calcite cements superimposed on the post-
Triassic variation of 87Sr/86Sr in sea water presented by Burke et al. (1982). We interpret these data to
indicate that some of the cations for cementation of both formations were derived from Mesozoic marine
carbonate strata, either by remobilization of rock fragments in the sandstones, or as the result of fluids which
transported components into the Tertiary strata. Calcite (and quartz) emplacement seems to have taken place
prior to extensive silicate (feldspar and clay) stabilization reactions which would have released radiogenic
strontium (87Sr).

5 1 r 3"

6 --$~
x

CD
"o 8
"(3
$
~9 9 9
9
9 (D
9 ~o 30
~' I0 3
v
v
II --

12

13 --*

14 38 samples --~
1591 grains

15
50 20 I0 I0 20 50 40 50
% An in plagloclase % Or in total feldspar

FIG. 10. Composition of Wilcox feldspars as a function of depth. Lines through plagioclase compositions
are + 1 standard deviation in An content based on about 80 analyses per thin section. All sands below about
3000 m (10 000 feet) are at least partially albitized and have suffered extensive K-feldspar dissolution. Most
sands above about 2600 m (8500 feet), in contrast, are essentially unaffected by feldspar stabilization
processes. Crosses from Boles (1981).
228 L. S. Land & R. S. Fisher
mole %
eTSr/eSSr
CaC03 FeCO 3
.710 .715
850 . . . . 5.5 . . . . 6,0 15 25 i | . . . . !

o
4
o 10
9 ~
0
9
0 9
o 12
0

.....

14 9 9 e e e

FIG. 11. Stoichiometry and 875r/S6Sr of Wilcox ankerite as a function of depth. Less stoichiometric and less
radiogenic phases are more typical of shallower samples, although considerable scatter exists.

First, deeper ankerite tends to be more but the relative importance of feldspar versus
stoichiometric (Ca1.oFeo.52Mgo.48(CO3):) and clay mineral stabilization reactions has not yet
more enriched in 878r, whereas shallower anker- been determined. Almost all ankerite cement
ite tends to be less stoichiometric (Cal.2o- (Fig. 11) is less radiogenic than available
Feo.24Mg0.56(CO3)2) and less enriched in STSr analyses of either diagenetic illite or detrital
(Fig. 11). Boles (1978) attributed the variations in feldspars (Perry & Turekian 1974, Morton 1983,
ankerite composition to precipitation at differ- Table 1 and work in progress). Therefore
ent temperatures; less stoichiometric phases strontium isotopic compositions may reflect a
were hypothesized to be precipitated at lower mixture of the limited amount of strontium
temperatures, as is well documented for dolo- available from the silicate phases relative to the
mite. If this explanation applies on a regional abundant strontium available from dissolution or
basis, then the lack of correlation between recrystallization of marine carbonates or calcite
ankerite composition and 61sO requires that the cement.
large temperature differences between 2500 and
4500 m (where temperatures differ by approxi-
mately 70~ C today) must have been just offset by Synthesis
large variations in 61so water (see Fig. 12) to
account for ankerite of similar isotopic compo- Figure 12 shows the locus of temperature and
sition over a large depth range. Alternatively, 6180 water values calculated for average Wilcox
vertical changes in pore-fluid chemistry may quartz and calcite, and for average ankerite and
have been involved. The less radiogenic nature albite. As was the case in the regional study of the
of shallower ankerite might be explained by Frio Formation (Land 1984), the use of epoxy-
progressive dilution of radiogenic strontium by impregnated thin section stubs precluded any
less radiogenic strontium (from the calcite which attempt to obtain isotopic values for authigenic
was being replaced) as rising fluids supplied the kaolinite (but see Fisher & Land 1986) or
iron and magnesium from underlying units. chlorite.
Evolution of the pore fluids toward more Quartz and calcite were the first volumetrical-
calcium-rich, less radiogenic, ~80-depleted (be- ly important cements to form, and based on
cause of precipitation of l SO-enriched mineral petrographic relations, quartz commonly pre-
phases) compositions along vertical flow paths cedes calcite (Fig. 13). The isotopic data on both
could also account for the compositional varia- quartz and calcite suggest precipitation under
tions observed in Wilcox ankerite. similar conditions of temperature and 6180
The source of radiogenic strontium incorpor- water, in agreement with the paragenetic rela-
ated in the ankerite is undoubtedly from the tions seen in thin section. The 180-depleted
reaction of detrital K-bearing silicate minerals, nature of both Wilcox quartz and calcite cements
 Wilcox sandstone diagenesis, Texas Gulf Coast 229

140
9).?......
-.'l -': 9 L'.'. 1.
albite
,~

".).))i))i))i))i))i))........ a nk e r it e
i00

Temp.
(~
60
.========i=~='=:==i=~=:=i=i=:=i=i=!=.!.~.i.:..
====================.........
q u a rtz i.;'i.i'i.?i;.i.?........,
i/I

20

6 4 2 0 -2 -4

6180water

FIG. 12. Locus of temperature and 6 ~80 water values for average Wilcox quartz and calcite, and ankerite and
albite. Because of the range of about + 4%0 in values for quartz, more variable conditions of emplacement are
indicated for some of the quartz cement, but the average isotopic chemistry of quartz and calcite cement is in
accord with the petrographic relationships (Fig. 13). The shallowest depth of occurrence of calcite and quartz
today is similar, at approximately 85~C, and although ankerite occurs at shallower depths, it becomes
dominant at depths corresponding to a temperature of about 120~C. The preferred conditions of emplacement
are indicated by the positions of the mineral-pair labels.

relative to Frio cements (Table 3) must be due fig. 9 in Suchecki & Land 1983). Because only
either to emplacement of Wilcox cements at eight water samples from geopressured wells in
higher temperature or to water depleted in 180 the Wilcox Formation have been obtained (Fig.
(or both). Application of the quartz-calcite 14), we are forced to argue by analogy with the
geothermometer to the averages listed in Table 3 Frio Formation. If analogy with the Frio
predicts co-precipitation of Wilcox cements at a Formation is justified, and water having a 61sO
temperature of 68 ~ C from water having a 6180 of approximately +5%o was involved, then
of - 1~o. quartz and calcite in Wilcox sandstones were
The geothermal gradient in the Wilcox Forma- both emplaced at temperatures 20 - 30~ C hotter
tion today, approximately 43 ~ C km-1 in south- than was true in the Frio Formation (compare
ern Gulf Texas (Bodner 1985) and approximately Fig. 12 with fig. 4 in Land 1984). Higher thermal
33~ C km -1 in central and northern Gulf Texas, gradients in Wilcox sands could also account for
is considerably higher than in the Frio Forma- the shallower emplacement of quartz cement in
tion (approximately 25~ km -1. These gradi- the Wilcox Formation, and in later albitization
ents have probably prevailed throughout most of at shallower depths.
the history of both Formations. The effects of the If the reaction of smectite to illite controlled
thermal event which accompanied rifting of the the 6180 of the porewater, then the strontium
Gulf in Jurassic time would presumably have isotopic composition of the porewater would
dissipated during the more than 108 years that have been affected by that reaction as well. But
elapsed prior to the onset of Wilcox deposition. calcite cements are considerably less radiogenic
A Tertiary geothermal gradient in the Wilcox than the detrital clays. Therefore, clay transfor-
Formation higher than that observed at present mations either postdated calcite (and quartz)
is currently unsupported. cementation, or the large amount of strontium
If quartz and calcite cements precipitated (thousands of ppm) released from contempora-
from sea water buffered by clay mineral reac- neous nannofossils, and provided by water from
tions in shales, then the 6180 of the water must underlying strata, overwhelmed the small
have been within about 2%0 of +5%0. This amount (hundreds of ppm) of radiogenic stron-
conclusion is substantiated both by direct analy- tium involved in the reaction of smectite to illite
sis of diagenetic illite and present-day formation and the stabilization of detrital feldspars.
water from the Frio Formation (Fig. 14), and by The alternative to emplacement of Wilcox
quantitative modelling (compare Fig. 14 with quartz and calcite cements at higher temperatures
230 L. S. L a n d & R. S. Fisher

FIG. 13. Back-scattered electron image of calcite cement (C) which post-dates quartz overgrowth cement (Q).
The 878r/86Sr of the calcite cement is 0.7082, suggesting that it was derived primarily from dissolution of
coeval nannofossils. The fact that non-radiogenic calcite post-dates quartz in this sample (from 2251 m,
Karnes County) also suggests that extensive silicate reactions, which should involve radiogenic strontium
having an 8VSr/86Srratio no lower than 0.711, were not the source of SiO2 for quartz cementation. Scale bar is
10 ~tm.

than was characteristic of the Frio Formation, is quartz cementation of Mesozoic sands (Dutton
that the water which emplaced Wilcox quartz 1986, McBride et al. in press) suggests that
and calcite was several per mil lighter than the meteoric water may have replaced connate sea
water which emplaced Frio cements. Meteoric water prior to the development of over-pressure,
water is unlikely to have gravitationally re- if the hydrothermal systems which dominated
charged Wilcox deltaic sediments because it is the early diagenesis of Mesozoic sands continued
doubtful that sufficient head could have been into the Eocene.
generated in the low relief Wilcox deltaic plain. Alternatively, water near SMOW in compo-
Deep, continuous recharge is inconsistent with sition may have been derived by compaction of
the almost certain early establishment of thick carbonate-rich Mesozoic strata over which
geopressure in the thick, marine, pro-delta muds the Wilcox Formation prograded (Fig. 1). Con-
(Sharp & Domenico 1976). The initial burial of nate porewater in underlying strata would need
connate meteoric water is difficult to reconcile to have been compacted away before extensive
with the deposition of these dominantly marine high-temperature calcite-water reactions took
sediments, but the involvement of large volumes place, however, because such reations enrich the
of hydrothermally circulated meteoric water in water in 180 even more than the + 5%0 observed
 Wilcox sandstone diagenesis, Texas Gulf Coast 231

&*s0 0

o 5 10 15 20 25
I t | i 9 - i . . . . l
2O i . , i i

o
x

o
o
o
~ ~ 1 7 o6
6O
o
%
x I o
o
o

~ 100 +""" b o

o
o
o o
9 + "*'- x

o
. o
140 o
9 +%'+ 9
o 9 9 o

-4- o
+ x
180 x

FIG. 14. 6180 of modern Wilcox formation water below hard geopressure (crosses), Frio Formation water from
all pressure regimes (dots), and the < 1 ~tm fraction (essentially pure smectite-illite) of Frio shales (circles) and
Wilcox shales (crosses). Data from Kharaka et al. (1977), Yeh & Savin (1977), Morton (1983), Jackson (in
preparation) and Morton & Land (1987). Horizontal lines a and b indicate the equilibrium fractionation
between illite-smectite and water at 80~ and 100~ (14.8 and 12.7%o, respectively) using the equation of
Eslinger & Savin (1973). At all depths, the most depleted Frio clays are nearly in equilibrium with the most
enriched waters observed today at the same depth, even above present-day geopressure (approximately line b
in many areas). Wilcox diagenetic illite, in contrast, is depleted in 180 approximately 4%0 relative to Frio
diagenetic illite, as is also observed for calcite and quartz cements in the respective sandstones (Table 3).

in terrigenous units (Land & Prezbindowski centimetre of average Wilcox sandstone. Free
1981). convection appears to be the only viable mecha-
Jackson (in preparation) has found Wilcox nism available to transport such large volumes of
diagenetic illite to have a 6180 of approximately water. Free convection in the rapidly deposited,
+ 14%o compared to values no lighter than undercompacted, sands and muds, at tempera-
+ 18%o for the Frio Formation (Fig. 14). Illite tures near 80 ~ C, with water having 6180 near
values of + 14%o are difficult to explain unless a 0%0 can account for the isotopic chemistry of
source of depleted water is invoked. The data on Wilcox quartz and calcite cements. Free convec-
quartz and calcite cements in the sandstones, and tion appears to be feasible in thick sandy sections
diagenetic illite in the shales, therefore currently pressurized to near lithostatic pressures (Blan-
support Wilcox quartz and calcite cements being chard & Sharp 1985), and in more shale-rich
emplaced by water having a 6180 near SMOW in strata if extensive microfracturing occurs. Some
composition. In contrast, Frio cements were components for both Frio and Wilcox calcite
apparently emplaced by smectite-buffered sea cements seem to have been derived from
water having a 6180 near +5%0. underlying Mesozoic marine carbonates.
The mechanism for cement emplacement was The relative isotopic and minor element
probably silica and CO2 loss from ascending homogeneity, and lack of distinct trends with
(cooling and devolatilizing) pore fluids. The depth in both Wilcox and Frio calcite cements
volumes of cements involved could not have also requires an explanation. Cements having
been transported on a 'once through' basis by the relatively uniform oxygen isotopic and trace
limited volume of water available from compac- element chemistry might be emplaced near the
tion of underlying strata. The volumes of quartz top of a convecting geopressured system, at
and calcite cement involved in Wilcox burial relatively constant temperature (corresponding
diagenesis are similar to those involved in Frio to the temperature at which geopressure is
burial diagenesis, and the calculations presented established) and 61 sO water (corresponding to a
by Land (1984) will not be repeated here. If those mix of rock-buffered connate water, rock-buf-
assumptions and calculations are correct, then at fered mineral dehydration water, and water
least one litre of water was required to emplace expelled from underlying formations). In each
the volume of cement observed in each cubic unit, the geopressured zone would undoubtedly
231 L. S. Land & R. S. Fisher
have swept up through the thick clastic wedges non-radiogenic cements (calcite, and by infer-
as burial took place with time (Sharp & ence, quartz) and a late generation of radiogenic,
Domenico 1976). Cements could have been l sO-depleted ankerite as postulated by Boles &
deposited near the top of the cell, where Franks (1979, their fig. 9).
maximum temperature gradients could induce It is tempting to invoke fluids derived from
quartz precipitation and maximum pressure underlying Mesozoic strata to account for the
gradients could induce calcite precipitation replacement of calcite by ankerite in the immedi-
(Wood & Hewett 1984). The cements would have ately overlying Wilcox Formation. In some ways
been buried more deeply within the geopressured the late burial diagenesis of the underlying
cell as sedimentation continued. Because they Jurassic and Cretaceous platform carbonates
remained essentially in ionic (but not isotopic) (Moore & Druckman 1981, Prezbindowski 1985,
equilibrium with the pore fluids, they did not Woronick & Land 1985) and the diagenesis of
react until either acids were introduced into the Mesozoic sandstones (McBride 1981, Suchecki
system so that wholesale dissolution occurred, or 1983, Dahl 1984, Dutton 1986, McBride et al. in
until the ionic composition of the pore fluids press) more closely resembles some aspects of
changed drastically so that replacement took Wilcox sandstone diagenesis than does the
place. In the case of the Wilcox Formation, if a diagenesis of younger Gulf Coast terrigenous
more stable detrital clay suite required higher units. All the Mesozoic units contain somewhat
temperature to initiate reaction (Bruce 1984), radiogenic late ankerite or ferroan dolomite
and/or if water isotopically depleted in oxygen cements and replacement phases which are
was characteristic of the connate porefluids, then essentially absent in post-Eocene formations,
the more depleted cements in the Wilcox and the source for which is not clear. Weaver &
Formation relative to cements in younger forma- Beck (1971) hypothesized the transport of iron,
tions can be explained. More Wilcox cements are magnesium (and potassium) from underlying
now found above the top of the present overpres- units into the Tertiary section. Fluids derived
sured zone than is true of the Frio Formation, from the Mesozoic carbonates (which have
possibly because the pressure is beginning to 87Sr/86Sr ratios similiar to late phases in the
dissipate in the older unit. Mesozoic strata - - Stueber et al. 1984, Woronick
The major mineralogical (as opposed to isoto- & Land 1985) might also have supplemented
pic) difference between Wilcox and younger hydrocarbons generated in situ from Wilcox
sandstones is the late emplacement of ankerite in shales (Jenden & Kaplan 1984).
the Wilcox Formation. The reason for this
difference may be variations in the history of the
underlying strata, not in the history of co-
genetically deposited basinal sediments. The Conclusion
Wilcox Formation is the only Tertiary unit in the
Texas Gulf Coast in which ankerite is a major Wilcox and Frio sandstones differ significantly
carbonate cement. Ankerite is extremely rare in in their diagenetic history. Although systematic
post-Eocene sandstones relative to the amount of diagenetic variation can be documented within
calcite present (Gold 1984, Land 1984, Milliken each formation from north to south along the
1985, Land et al. 1987) despite the immense mass Texas Gulf Coast (e.g. Loucks et al. 1984),
of smectite which has been converted to illite in greater differences exist between the two Forma-
the younger units. tions than exist within either. In general, younger
Therefore it is questionable whether the Gulf Coast units are less diagenetically altered
reaction of smectite to illite provided iron and than either the Wilcox or Frio Formations (Gold
magnesium for ankeritization (Boles & Franks 1984, Milliken 1985). In contrast, older clastic
1979). It is possible, of course, that Wilcox units, although not as volumetrically extensive or
smectite was different from smectite deposited in as widely distributed, are more pervasively
the younger formations. The more volcanic-rich, altered and are altered at shallower depths. A
westerly source for major south Texas Frio growing number of isotopic analyses (Suchecki
deposition contrasts with the more northerly 1983, Franks & Forester 1984, Dutton 1986,
cratonic source for the Wilcox (Winker 1982, McBride et al. in press, work in progress) suggest
Loucks et al. 1984). Chlorite and pyrite occur as that the cements in older units are commonly
late diagenetic phases in both formations, but even more depleted in 180 than are the cements
their importance as sinks for iron and magne- present in Wilcox sands. Apparently, both
sium have not yet been quantitatively assessed in increased temperature and deeper penetration of
either. The reaction of a single phase containing meteoric water characterized the diagenesis of
radiogenic strontium (smectite to illite) could not Mesozoic sandstones early in the diagenetic
possibly account for both an early generation of history of the Gulf sediment wedge.
 Wilcox sandstone diagenesis, Texas Gulf Coast 233
Currently available data all suggest that the ACKNOWLEDGMENTS; We wish to thank Bob Loucks
Gulf has evolved toward less extensive alteration and Bob Morton of the Bureau of Economic Geology
for amassing the samples on which this study was based
of younger sands, alteration of younger sands at
and permitting us access to the point-count data and
progressively greater burial depths, and toward the thin section stubs. We have relied almost exclusive-
generally lower temperature diagenesis with time ly on the point-count data obtained by Loucks and his
(Land et al. 1987). The changing abundance and colleagues for Figs 4, 5 and 6. Numerous people have
chemical nature of the authigenic phases, and contributed to this study with their critical comments
the potential for large cross-formational fluxes of and encouragement. We especially want to thank the
material (as suggested by the 87Sr data and the members of FOG (Friends of the Gulf), a seminar
necessity for vertical transport of liquid hydro- group of the last few years. Paul Blanchard, Bill
carbons into immature strata, for example), Galloway, Paul Gold, Tim Jackson, Paul Lundegard,
Wendy Macpherson, Earle McBride, Kitty Milliken
demand that diagenetic studies of Gulf Coast
and Jack Sharp were all active in directing our thinking
units, and sedimentary basins in general, be and/or improving our prose. Steve Franks provided an
placed in a regional context. The burial diagene- especially constructive review, as did several other
sis of any particular unit can only be understood reviewers, including Fred Longstaffe, Sam Savin, Jim
as part of a much larger scale (in both time and Boles, and Steve Crowley.
space) diagenetic system. Considerably more Laboratory support for this work was provided by
data on sandstone, carbonate, evaporite and the National Science Foundation, EAR-7824081 and
shale diagenesis will be needed before the EAR-8121009. Larry Mack and Rosemary Capo did
diagenetic evolution of the Gulf Coast syncline most of the strontium isotopic analyses, and continuing
strontium isotopic analyses of Gulf Coast rocks and
can be understood to the point that it can be used
waters constitutes Larry's PhD dissertation. Addi-
for predictive purposes, as a model for other tional support was provided by the Mobil's Dallas
basins, or applied to uplifted and exhumed Research Laboratory, Texaco USA, and the Geology
sedimentary wedges. Foundation of the University of Texas at Austin.

References

BEBOUT, D. G., WEISE, B. R., GREGORY, A. R. & CAROTHERS,W. W. & KHARAKA,Y. K. 1978. Aliphatic
EDWARDS, M. B. 1982. Wilcox sandstone reser- acid anions in oil-field waters and their implica-
voirs in the deep subsurface along the Texas Gulf tions for the origin of natural gas. American
Coast. Bureau of Economic Geology, University of Association of Petroleum Geologists Bulletin, 62,
Texas, Austin, Report of Investigations, 117. 2441-53.
BLANCHARD, P. E. & SHARP, J. i . JR 1985. Possible DAHL, W. M. 1984. Progessive burial diagenesis in
free convection in thick Gulf Coast sandstone lower Tuscaloosa sandstones, Louisiana and Mis-
sequences. Southwest Section American Association sissippi (Abstract). Clay Minerals Society Annual
of Petroleum Geologists Transactions, 11, 6-12. Meetings, p. 42. Louisiana State University De-
BODNER, D. 1985. Heat variations caused by ground- partment of Geology, Baton Rouge, Louisiana.
water flow in growth faults of the South Texas DUTTON, S. P. 1986. Diagenesis and burial history of
Gulf Coast basin. MA thesis, University of Texas the Lower Cretaceous Travis Peak Formation,
at Austin. East Texas. PhD Dissertation, University of Texas
BOLES, J. R. 1978. Active ankerite cementation in the at Austin.
subsurface Eocene of southwest Texas. Contribu- ESLINGER, E. & SAVIN, S. M. 1973. Mineralogy and
tions to Mineralogy and Petrology, 68, 13-22. oxygen isotope geochemistry of the hydrothermal-
- 1981. Active albitization of plagioclase, Gulf
-
ly altered rocks of the Ohaki-Broadlands New
Coast Tertiary. American Journal of Science, 282, Zealand geothermal area. American Journal of
165-80. Science, 273, 240-67.
-- & FRANKS, S. G. 1979. Clay diagenesis in Wilcox FISHER, R. S. 1982. Diagenetic history of Eocene
sandstones of southwest Texas: implications of Wilcox sandstones and associated formation wa-
smectite diagenesis on sandstone cementation. ters, South-Central Texas. PhD Dissertation, Uni-
Journal of Sedimentary Petrology, 49, 55-70. versity of Texas at Austin.
BRUCE, C. H. 1984. Smectite dehydration - - its relation -- & LAND, L. S. 1986. Diagenetic history of Eocene
to structural development and hydrocarbon ac- Wilcox sandstones, South-Central Texas. Geochi-
cumulation in the Northern Gulf of Mexico basin. mica et Cosmochimica Acta, 50, 551-62.
American Association of Petroleum Geologists Bulle- FRANKS, S. G. & FORESTER, R. W. 1984. Relationships
tin, 68, 673-83. among secondary porosity, pore-fluid chemistry
BURKE, W. H., DENISON, R. E., HETHERINGTON,E. A., and carbon dioxide, Texas Gulf Coast. In:
KOEPNICK, R. B., NELSON, H. F. & OTTO, J. B. MCDONALD, D. A. & SURDAM,R. C. (eds). Clastic
1982. Variation of seawater 87Sr/86Sr throughout Diagenesis, pp. 63-80. American Association of
Phanerozoic time. Geology, 10, 516-9. Petroleum Geologists Memoir, 37.
234 L. S. Land & R. S. Fisher
GOLD, P. B. 1984. Diagenesis of middle and upper --, LAND, L. S. & LOUCKS, R. G. 1981. History of
Miocene sandstones, Louisiana Gulf Coast. MA burial diagenesis determined from isotopic geo-
thesis, University of Texas at Austin. chemistry, Frio formation, Brazoria County, Tex-
JACKSON, T. J. in preparation. Diagenesis of (Eocene) as. American Association of Petroleum Geologists
Wilcox sandstones and shales, Texas Gulf Coast. Bulletin, 65, 1397-413.
PhD Dissertation, University of Texas at Austin. MOORE, C. H. & DRUCKMAN,Y. 1981. Burial diagene-
JENDEN, P. D. & KAPLIN, I. R. 1984. Maturation of sis and porosity evolution, Upper Jurassic Smack-
organic matter in Paleocene-Eocene Wilcox over, Arkansas and Louisiana. American Associ-
group, South Texas: Relationship to clay diagene- ation of Petroleum Geologists Bulletin, 65, 597-
sis and sandstone cementation (Abstract). Ameri- 628.
can Association of Petroleum Geologists Bulletin, MORTON, J. P. 1983. Rb-Sr dating of clay diagenesis.
68, 492. PhD Dissertation. University of Texas at Austin.
KHARAKA,Y. K., CALLENDER,E. & CAROTHERS,W. W. MORTON, R. A. & LAND. L. S. 1987. Regional
1977. Geochemistry of geopressured waters from variations in formation water chemistry, Frio
the Texas Gulf Coast. In: Proceedings of the Third Formation (Oligocene), Texas Gulf Coast. Ameri-
Geopressured-Geothermal Energy Conference, pp. can Association of Petroleum Geologists Bulletin, 71,
121-65. University of Southwest Louisiana, La- 191-206.
fayette, Louisiana. PERRY, E. A. JR & TUREKIAN,K. K. 1974. The effects
LAND, L. S. 1984. Diagenesis of Frio sandstones, Texas of diagenesis on the redistribution of strontium
Gulf Coast: a regional isotopic study. In: MC- isotopes in shales. Geochimica et Cosmochimica
DONALD, D. A. & SURDAM, R. C. (eds). Clastic Acta, 38, 929-35.
Diagenesis, pp. 37-62. American Association of PREZBINDOWSKI,D. R. 1985. Burial cementation, is it
Petroleum Geologists Memoir, 37. important? A case study, Stuart City Trend, South
- - & MILLIKEN, K. L. 1981. Feldspar diagenesis in Central Texas. In : SCHNEIDERMANN,N. & HARRIS,
the Frio formation, Brazoria County, Texas Gulf P. M. (eds). Carbonate Cements, pp. 241-64.
Coast. Geology, 9, 314-8. Society of Economic Paleontologists and Minera-
, & MCBRIDE, E. F. 1987. Diagenetic logists Special Publication, 36.
evolution of Cenozoic sandstones, Gulf of Mexico SHARP, J. M. JR & DOMENICO, P. A. 1976. Energy
sedimentary basin. Sedimentary Geology, 50, 195 transport in thick sequences of compacting sedi-
225. ments. Geological Society of America Bulletin, 87,
- - & PREZBINDOWSKI,D. R. 1981. The origin and 390-400.
evolution of saline formation water, Lower Creta- STUEBER, A. M . , PUSHKAR, P. & HETHERINGTON, E. A .
ceous carbonates, South-Central Texas, U.S.A. 1984. A strontium isotopic study of Smackover
Journal of Itydrology, 54, 51-74. brines and associated solids, southern Arkansas.
LOUCKS, R. G., DODGE, M. M. & GALLOWAY, W. E. Geochimica et Cosmochimica Acta, 48, 1637-
1984. Regional controls on diagensis and reservoir 49.
quality in Lower Tertiary sandstones along the SUCHECKI,R. K. 1983. Isotopic evidence for large-scale
Texas Gulf Coast. In: MCDONALD, D. A. & interaction between formation waters and clastic
SURDAM,R. C. (eds). Clastic Diagenesis, pp. 15-46. rocks (Abstract). Geological Society of America
American Association of Petroleum Geologists Annual Meeting, Indianapolis, Indiana, p. 701.
Memoir, 37. & LAND, L. S. 1983. Isotopic geochemistry of
LUNDEGARD,P. D. & LAND, L. S. 1986. Carbon dioxide burial-metomorphosed volcanogenic sediments,
and organic acids: Their origin and role in Great Valley sequence, northern California.
diagenesis, Texas Gulf Coast Tertiary. In: GAU- Geochimica et Cosmochimica Acta, 47, 1487-
TIER, D. L. (ed.). The Roles of Organic Matter in 99
Diagenesis, pp. 129-46. Society of Economic SYERS, J. K., CHAPMAN, S. L., JACKSON, M. L., REX
Palentologists and Mineralogists Special Publica- R. W. & CLAYTON, R. N. 1968. Quartz isolation
tion, 38. from rocks, sediments and soils for determination
, & GALLOWAY, W. E. 1984. The problem of oxygen composition. Geochimica et Cosmo-
of secondary porosity: Frio formation (Oligocene), chimica Acta, 32, 1022-5.
Texas Gulf Coast. Geology, 12, 399-402. WEAVER, C. E. & BECK, K. C. 1971. Clay water
MCBRIDE, E. F. 1981. Diagenetic history of Norphlet diagenesis during burial: How mud becomes
Formation (Upper Jurassic), Rankin County, gneiss. Geological Society of America Special Paper,
Mississippi. Transactions of the Gulf Coast Associ- 134.
ation of Geological Societies, 31, 347-51. WINKER, C. D. 1982. Cenozoic shelf margins, North-
--, LAND, L. S. & MACK, L. E. In press. Diagenesis western Gulf of Mexico. Transactions of the Gulf
of feldspathic eolian and fluvial sandstones, Coast Association of Geological Societies, 32, 427-
Norphlet Formation (Upper Jurassic), Rankin 48.
County, Mississippi, and Mobile County, Ala- & BUFFLER, R. T. in press. North-south cross-
bama. American Association of Petroleum Geolo- section, Western Gulf of Mexico. In: Gulf of
gists Bulletin. Mexico Ocean Margin Drilling Program, Regional
MILLIKEN, K. L. 1985. Petrology and burial diagenesis Atlas Series, No. 6. Marine Science International,
of Plio-Pleistocene sediments, northern Gulf of Woods Hole, Massachusetts.
Mexico. PhD Dissertation. University of Texas at WOOD, J. R. & HEWETT, T. A. 1984. Reservoir
Austin. diagenesis and convective fluid flow. In: Mc-
 Wilcox sandstone diagenesis, Texas Gulf Coast 235
DONALD, D. A. & SURDAM, R. C. (eds). Clastic Cements, pp. 265-75. Society of Economic Paleon-
Diagenesis, pp. 99-110. American Association of tologists and Mineralogists Special Publication,
Petroleum Geologists Memoir, 37. 36.
WORONICK, R. E. & LAND, L. S. 1985. Late burial YEn, H. W. & SAVINS. M. 1977. Mechanism of burial
diagenesis, Lower Cretaceous Pearsall and Lower metamorphism of argillaceous sediments: 3. O-
Glen Rose formations, South Texas. In: SCHNEI- isotopic evidence. Geological Society of America
DERMANN, N. & HARRIS, P. M. (eds). Carbonate Bulletin, 88, 1321-30.

LYNTONS. LAND.Department of Geological Sciences, University of Texas at Austin, Austin,

Texas 78713, USA.
R. STEPHENFtSHER. Bureau of Economic Geology, University of Texas at Austin, Austin,
Texas 78713, USA.
Regional dolomitization of subtidal shelf carbonates: Burlington
and Keokuk Formations (Mississippian), Iowa and Illinois
David C. Harris & William J. Meyers

S U M M A R Y : Cathodoluminescent petrography of crinoidal limestones and dolomites

from the Mississippian (Osagean) Burlington and Keokuk Formations in Iowa and Illinois
has revealed a complex diagenetic history of calcite cementation, dolomitization,
chertification and compaction. Dolomite occurs abundantly in subtidal, open-marine facies
throughout the study area. Three luminescently and chemically distinct generations of
dolomite can be recognized regionally. Dolomite I, the oldest generation, is luminescent,
thinly zoned, and occurs mainly as a replacement of lime mud. Dolomite II has dull red
unzoned luminescence, and occurs mainly as a replacement of dolomite I rhombs. Dolomite
III is non-luminescent, and occurs as a syntaxial cement on, and replacement of, older
dolomite I and II rhombs.
Petrography of these dolomite generations, integrating calcite cement stratigraphy,
chertification and compaction histories has established the diagenetic sequence. Dolomites I
and II pre-date all calcite cements, most chert, intergranular compaction and stylolites.
Dolomite III precipitation occurred within the calcite cement sequence, after all chert, and
after at least some stylolitization. The stratigraphic limit of these dolomites to rocks older
than the St Louis Limestone (Meramecian) suggests that dolomitization took place before or
during a regional mid-Meramecian subaerial unconformity.
A single dolomitization model cannot reasonably explain all three generations of dolomite
in the Burlington and Keokuk limestones. Petrographic and geochemical characteristics
coupled with timing constraints suggest that dolomite I formed in a sea water-fresh water
mixing zone associated with a meteoric groundwater system established beneath the pre-St
Louis unconformity. Dolomite II and III may have formed from externally sourced warm
brines that replaced precursor dolomite at shallow burial depths. These models therefore
suggest that the required Mg for dolomite I was derived mainly from sea water, whereas that
for dolomites II and III was derived mainly from precursor Burlington--Keokuk dolomites
through replacement or pressure solution.

A fundamental problem in carbonate diagenesis the dolomites relative to other diagenetic events,
has been the origin of widespread dolomites, which in turn constrains their absolute ages. The
particularly those in subtidal, open marine study is based on 175 thin sections chosen from
depositional facies. In recent years research on about 385 samples collected from 25 measured
dolomites has become increasingly more sophis- sections throughout the outcrop belt of southeast-
ticated, using a wide range of geochemical and ern Iowa and adjacent Illinois (Fig. 1). These
crystallographic approaches (e.g. Zenger et al. samples were spaced at about 0.3-1 m intervals
1980, Reeder 1981, Saller 1984, Banner et al. in throughout the preserved thickness of the Burl-
press). Essential to using these approaches is the ington and Keokuk Formations, which averaged
establishment of a detailed petrographic and about 20 m (range = 8-50 m).
stratigraphic framework. We report herein one
such regional petrographic study of dolomites in
the Osagean Burlington and Keokuk Formations Stratigraphy and depositional facies
in southeastern Iowa and adjacent Illinois, a
study which is part of a broader project on the In southeastern Iowa, the main units investigat-
geochemistry and petrology of regional dolo- ed were the Osagean Burlington and Keokuk
mites in these strata (Harris 1982, Smith 1984, Formations (Fig. 1). Additionally, the underly-
Cander 1985, Kaufman 1985, Prosky & Meyers ing Kinderhookian and overlying Meramecian
1985, Banner 1986, Daniels 1986, Banner et al. in units were sampled in order to put constraints on
press). timing of calcite cements and dolomites.
The main approach in this petrographic study
has been the use of cathodoluminescence, which
Burlington and Keokuk Formations
has allowed us to: (1) identify three major,
regionally extensive stages of dolomite precipita- The Burlington and Keokuk Formations were
tion, (2) establish the relative ages of the three part of a broad shallow shelf that flanked the
dolomite types, and (3) to establish the timing of south side of the Transcontinental Arch during

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 237

Geological Society Special Publication No. 36, pp. 237-258.
238 D. C. Harris & W. J. Meyers
-4

L
~ii!iii!??!i!!ii??
:H::::;:::::;::::

OUTCROP OF
MISSISSIPPIAN I

~ L L INOIS

MISSOURI ,,,x
,G~XOv"
30 m

i
20

10

OZARK ' "

DOME ?

I I
0 100km

FIG. 1. Palaeogeographic map of the Burlington Shelf, and stratigraphic section of Osagean and Meramecian
rocks in Iowa and western Illinois. The outcrop belt of Mississippian carbonates is indicated by the shaded
pattern. Black circles and triangles indicate locations of measured sections in the Burlington and Keokuk
Formations and in Meramecian strata, respectively, used in this study. A possible Burlington-Keokuk
equivalent shoreline facies, studied by Sixt (1983), occurs as indicated to the NW of the study area.

Osagean time. These shelf carbonates extend kuk, especially in mud-supported rocks. Bedding
westward into Kansas, and thin toward the SE in typically varies from about 30 to 80 cm thick and
Illinois (Fig. 1), and toward the south in is defined by stylolites, shaly partings, and
Arkansas, Kansas, and Oklahoma to a deeper contacts between contrasting depositional
starved shelf facies comprising a condensed textures.
sequence of phosphatic and cherty carbonates The Burlington and Keokuk therefore com-
and shales (Harris & Parker 1964, Lane 1978, prise strictly subtidal shelf, open-marine facies,
Lane & DeKeyser 1980). lacking ooids, stromatolites, fenestral carbonates
The Burlington and Keokuk Formations in the and other indicators of sea-level/peritidal facies.
study area range up to about 50 m thick and The nearest possible peritidal facies are the
comprise skeletal limestones and dolostones. The Lower Humboldt beds of the Gilmore City
Keokuk differs from the Burlington mainly by Formation in north-central Iowa recently inter-
containing more shaly carbonates, but comprises preted to be shoreline equivalents of the Burling-
essentially the same depositional facies. These ton (Sixt 1983) (Fig. 1). Relevant to the diagene-
carbonates are dominated by crinoids, fenestrate tic history of the Burlington and Keokuk, there is
bryozoans and brachiopods, and vary in deposi- no substantive physical evidence for intraforma-
tional textures from grainstones to mudstones. tional subaerial disconformities, nor for subaer-
Locally they contain glauconite pellets and fish ial exposure at the Burlington-Keokuk contact.
teeth horizons. Chert nodules and lenses are We also see no evidence for submarine hard-
common throughout the Burlington and Keo- grounds within grain or mud-supported rocks.
 Carbonate dolomitization, Iowa and Illinois 239
Younger Mississippian formations and brachiopods. Dolomite pervasively replaced
lime mud in mud-supported textures. In cement-
Meramecian rocks in the study area are up to
free skeletal packstones, dolomite occurs as a
about 45 m thick, and consist in ascending order
'matrix' between grains (Fig. 2c). In cement-
of the Warsaw Shale, Spergen Formation, St
bearing packstones, dolomite occurs as geopetal
Louis Formation and the St Genevieve Lime-
fabrics on tops of grains and at the bottoms of
stone (Fig. 1). These in turn are overlain
axial canals of crinoids (Fig. 2d). This geopetal
unconformably by Pennsylvanian (Desmoine-
distribution reflects the original distribution of
sian) strata which cut down through progress-
lime mud. In some cement-bearing grain rocks,
ively older units towards the east.
dolomite occurs as clumps of rhombs near the
The Warsaw conformably overlies the Keo-
centres of inter-crinoid interstices, and as iso-
kuk and consists of fossiliferous calcareous
lated single rhombs or clusters of a few rhombs
shales and shaley dolomites. The Spergen For-
encased in calcite cements. In the first case, the
mation conformably overlies the Warsaw and
clumps of rhombs near the centres of intergranu-
consists of limestones and dolomites that are
lar interstices seemingly float in calcite cement
locally shaly, sandy and oolitic. The St Louis
(Fig. 3a), a distribution that could be interpreted
Formation consists of fine-grained, generally
as late, post-cement precipitation of dolomite.
unfossiliferous lime mudstones containing col-
However, based on dolomite-cement timing
lapse(?) breccias and shale beds. The breccias
discussed below, these clumps of dolomite are
are interpreted as evaporite solution-collapse
interpreted as replacements of micritic grains,
breccias based on the presence of evaporites in
possibly peloids or bryozoans. In the second case,
the subsurface in Illinois (Carlson 1979).
rhombs encased in calcite cement occur in
The contact between the Spergen and St Louis
contact with fossil grains in both geopetal and
has been interpreted as an unconformity based
non-geopetal distributions, and as rhombs com-
on the apparent overlap of St Louis rocks on to
pletely encased in calcite cement (Fig. 3b). These
Kinderhookian strata in north-central Iowa, and
rhombs may be dolomite cements anchored on
the absence of the Warsaw and Spergen in this
grains (or other rhombs) in a third dimension, or
area (Carlson 1979). The recent work of Sixt
may be replacements of former microcrystalline
(1983) implies that this angular discordance is
calcite grains or mud.
not real, because the rocks previously identified
Porosity in the Burlington-Keokuk carbon-
as St Louis in north-central Iowa are probably
ates is restricted to dolostones and to clumps of
Burlington equivalents. In spite of this there is a
dolomite within the grain-supported textures,
regional unconformity at the base of the St Louis
mainly as intercrystalline porosity (Figs 2a, b, 4c,
(Lane, personal communication) in southeastern
5c, 8b), as fossil moulds, and more rarely as intra-
Iowa, which may represent the Meramecian
rhomb pores.
eustatic sea-level drop shown on published global
sea-level curves (Vail et al. 1977).
In summary, the Spergen and St Louis
Formations represent shallowing, and restriction Cathodoluminescence petrography
of the former open-circulation deeper shelf of the
Cathodoluminescence petrography shows that
Burlington-Keokuk sediments. Relevant to the
there are three major regionally extensive gen-
diagenetic history there were at least two
erations of dolomite--dolomite I, II and I I I - -
unconformities that could have resulted in
each having a distinct cathodoluminescent sig-
freshwater diagenesis of the Burlington-Keokuk
nature and distinct geochemistry. All three occur
rocks: the pre-St Louis and the pre-Pennsylvan-
in mud- and grain-supported textures.
ian unconformities.
Dolomite I is the single most abundant
dolomite type and is characterized by finely
Dolomite petrography zoned orange luminescence (Fig. 4a, b). The thin
concentric zoning is due to variations in intensity
Burlington-Keokuk dolomites consist mainly of and colour of luminescence from yellowish-
euhedral to subhedral rhombs that average 70- orange to brownish-orange. Although the fine
100 pm in size (range = 30-200 pro) (Fig. 2a, b). zoning in dolomite I occurs throughout the study
The rhombs have unit extinction and are usually area, the details of the zonal stratigraphy vary
inclusion-poor, or have inclusion-rich cores and between measured sections. Dolomite I is Ca-
clear rims. These inclusions comprise opaque rich (54.5-56.5 mole% CaCO3), contains 800-
phases, CaCO3, and fluid inclusions (Smith 1500ppm Mn, 400-8000ppm Fe and has
1984). 8~Sr/86Sr near Mississippian sea water (0.7076)
The dolomite selectively replaced lime mud (Prosky & Meyers 1985, Banner el al. in press).
and bryozoans, and only rarely replaced crinoids Stable isotopes of dolomite I are relatively heavy,
z4o D. C. Harris & W. J. Meyers

" •~ ~'~

rn .~ ~

~ . ~

~o~
~_~ .~ "0 .~

~.~

n~

~ ~o
~'~
o .-.~.

~~~~
 Carbonate dolomitization, Iowa and Illinois z4I

FIG. 3. Dolomite from the Burlington and Keokuk Formations. (a) Dolomite occurring as clumps of rhombs
between grains (D) interpreted as a replacement of micritic skeletal or peloidal grains. Scale = 500 Ixm. (b)
Dolomite rhombs (D) encased and isolated in calcite cements in crinoidal grainstone. Rhombs may be either
cements, or replacements of small amounts of micritic calcite. E = euhedral rhomb, part of which has replaced
crinoid. Scale = 250 lam.

with m e a n 6180 of -0.4%o P D B ( n = 3 3 , brown (Fig. 4c, d). Dolomite II most c o m m o n l y

range=-2.2 to +2.5%0), and m e a n 613C of occurs as the interiors of r h o m b s that have
+2.5%0 P D B (n = 33, range = - 0 . 9 to +4.0%0) dolomite I rims, in which case the contacts
(Banner 1986). between the two are irregular and cut across
Dolomite II, the second most a b u n d a n t gener- zones within dolomite I (Fig. 4c). The geometries
ation, is characterized by unzoned red lumines- of the dolomite II cores are often irregular, and
cence that varies from moderate red to dark red- volumetrically range from a small percentage to
242 D. C. Harris & W. J. Meyers

0 ~ o~"~ O.....k
,-~ 0 ,.4 ~.-~"

' ~ 1 1 ~
~ ,~ ~ " ~ .~

, ~ ~-~

~ ~ ,.-.,..,
9~, ~ ~ ~ .~-
~ . ~ .~

="~ o ~'~

~ - -~
.~ L,~.~

~.~oo_~

N'.~,

~.~-~
 Carbonate dolomitization, Iowa and Illinois 243

occupying nearly the entire rhomb (Fig. 4c). intrarhomb pores (hollow cores and selectively
Dolomite II also occurs as rhombs without dissolved zones) occur in dolomite I, and are rare
dolomite I rims, in which cases it commonly in dolomite II. These features imply that dolo-
contains small brightly luminescing patches mite I was less stable than dolomite II. Thus,
(Figs 4d, 5d), some of which are similar in there was a chemical 'motive' for the replace-
luminescence colour to dolomite I. Geochemical ment of dolomite I by dolomite II, and for the
characteristics of dolomite II differ markedly common selective replacement of dolomite I
from those of dolomite I. Dolomite II is nearly cores.
stoichiometric (51-52 mole~ CaCO3), and con- The third volumetrically most important gen-
tains 1000-2600 ppm Mn and 1600-43 000 ppm eration, dolomite III, is non-luminescent, is
Fe, the darker luminescing varieties having the strongly ferroan, and is unzoned (Fig. 5a).
greater Fe and Mn contents (Prosky & Meyers Dolomite III is present on most dolomite I
1985). Dolomite II contains radiogenic Sr rhombs (Fig. 5a) to some degree, and occurs
(8~Sr/S6Sr = ---0.7094), mean 3180 of -3.8%o more rarely on dolomite II (Fig. 5d). Most
PDB ( n = 31, range = - 0 . 2 to -6.6%0 PDB), commonly dolomite III comprises syntaxial rims
and mean 313C of +2.7%0 PDB (n = 30, range on dolomites I or II, with the contact being either
= - 1.0 to +4.1%o PDB) (Banner et al. in press, conformable (Fig. 5a) or irregular (Fig. 5a, d).
Banner 1986). The irregular contacts with dolomite I cut across
The irregular geometries of the dolomite I- the fine zoning in dolomite I, and .commonly
dolomite II contacts could conceivably be ex- deeply embay the dolomite I rhombs (Fig. 5a). In
plained by either dolomite I or dolomite II being some cases dolomite III occurs in the centres and
the younger. Specifically, zoned dolomite I rims as zones within dolomite I rhombs (Fig. 5b), in
could be interpreted as syntaxial overgrowths on which cases their contacts are irregular and cut
corroded dolomite II rhombs or, conversely, across the zones of dolomite I. In the latter case,
zoned dolomite I rims could be interpreted as the both the inner and outer contacts of dolomite III
chemically more resistant relics after replace- zones with dolomite I host are unconformable
ment of cores by dolomite II. We favour the (Fig. 5b). In some samples whole rhombs of
second interpretation, that dolomite II replaced dolomite III are present, some of which contain
dolomite I, an interpretation first proposed by irregular remnants of zoned dolomite I (Fig. 5c).
Prosky (in preparation) and Banner (1986), for Geochemically, dolomite III is the most Fe- and
the following reasons. First, some of the small Mn-rich, with about 52 800-89 400 ppm Fe and
brightly luminescing patches within dolomite II 2500-8200ppm Mn, and contains about
(Fig. 4d) are most likely relics of dolomite I. 53 mole% CaCO 3 (Prosky & Meyers 1985).
Secondly, dolomite II rhombs are about the same We interpret the above described irregular
size as dolomite I rhombs, even where they occur contacts as dolomite III having replaced dolo-
in the same sample (Fig. 4c), and dolomite II mite I or II (Fig. 5a, d), and the conformable
dolostones have porosities comparable to those contacts as dolomite III having grown as
of dolomite I dolostones. The presence of rhombs syntaxial cements on dolomite I (Fig. 5a) or on
composed totally of dolomite I in a dolomite II dolomite II (Fig. 4d). The dolomite III rhombs
dolostone implies that dolomite I fluids moved containing relics of dolomite I and the associated
through these rocks for the same length of time as dolomite III rhombs (Fig. 5c) are interpreted as
through dolomite I dolostones. If, in the same dolomite III having completely replaced dolo-
sample, dolomite I rims were syntaxial over- mite I rhombs.
growths on corroded dolomite II cores, then In summary, the three major dolomite types
these rims should be equal in thickness to about represent progressive replacement; dolomite I
one-half of the width of the co-existing totally was replaced by dolomite II, and was replaced
dolomite I rhombs. In such a case, the dolomite and overgrown by dolomite III; dolomite II was
II plus dolomite I overgrowths should be signifi- replaced and overgrown by dolomite III (Fig. 6).
cantly larger than the co-existing dolomite I The most extensively replaced dolomite, dolo-
rhombs, and the dolostones should have lower mite I, is the least stoichiometric of the three.
intercrystalline porosity than dolomite I dolo- Furthermore, the replacement was accomplished
stones. Therefore, the fact that dolostones domi- by progressively more iron-rich dolomites. In the
nated by dolomite II have about the same above described cases where a younger dolomite
rhombs sizes and porosities as those dominated generation has replaced an older dolomite, we
by dolomite I is best explained by dolomite II see no definitive evidence for the older dolomites
being a replacement. Third, dolomite I is more having experienced a major moulding stage. For
Ca-rich than dolomite II, and commonly the example, we see no evidence of collapsed hollow
cores of dolomite I are more Ca-rich than their dolomite I rhombs pre-dating replacement by
rims (Prosky, in preparation). Fourth, most either dolomite II or dolomite III.
944 D. C. Harris & W. J. Meyers

~~ ~o
~ o~

~o~ 0 ~'~

I~ 0 ~'-~ II

~,.~. .~H~
 Carbonate dolomit&ation, Iowa and Illino& 245

FIG. 6. Summary of Burlington-Keokuk dolomite petrography. Dolomites II and III represent progressive
replacement and overgrowth of older dolomite I or II rhombs by younger, more stoichiometric and ferroan
dolomite phases. Disconformities between one dolomite generation, and a younger generation may represent a
discrete period of dolomite dissolution, or may have resulted directly from the replacement process.

Timing of dolomitization cement stratigraphy in the skeletal packstones

and grainstones of the Burlington-Keokuk rocks
Critical to constraining the possible diagenetic (Harris 1982, Harris & Meyers 1985, Cander
environments of dolomitization is determining 1985, Kaufman 1985, Daniels 1986, Kaufman et
the age of each dolomite generation relative to al. in press). This cement stratigraphy is essential
the diagenetic and burial history of the Burling- to identifying the paragenetic sequence, includ-
ton-Keokuk strata. This was accomplished ing the timing of the dolomites. The cements are
through four approaches: (1) timing of dolomites dominated by crinoid-syntaxial calcites that
relative to calcite cements, (2) stratigraphic contain as many as six major zones (Fig. 7a, b, c,
distribution of dolomites, (3) timing of dolomites d). From oldest to youngest, these comprise: zone
relative to compaction, and (4) timing of dolo- 1, non-ferroan, moderate to bright luminescence
mites relative to chertification. with many subzones; zone 2, non-ferroan, non-
luminescent with luminescent hairline subzones;
zone 3, non-ferroan, moderate luminescence;
Dolomite-calcite cement relationships
zone 4, non-ferroan, non-luminescent with lumi-
Cathodoluminescence petrography has estab- nescent hairline subzones; zone 5, ferroan, dull
lished a detailed and regionally extensive calcite fairly uniform luminescence with no or few
246 D. C. Harris & W. J. Meyers

= ~
o oo ~ ~

~-~ ~ ~ ~.~ ~ o

0 IZ~.,_, ~ "~t'.-~--- ,..

I ~'~- ~ T ~ ,-- ~- 0 ~

~,~ ~ . ~ ~ ~,.~

~, .,-,.~ ,~.-~ ~,-~ ~ ~ ...~

,,~, 1~ ~,.~.~ ~.~ ~'~ ~,~,,~

~:~ ~ O ~ 0 r 0 .~

9- ~ ~ ~ ~- i...;. ~ ' - _ ~

.0~ ~>.-.~
~- ~~ , . ~ ~ . c~~ . - .~- ~, ~
~-~

9. . . ~ ~ 0.~ ~ ~ ~ ~'=

~ ~ i: ~'C~ ~'~ ~- -

~ ~ ~ 0 ~ ~ 0 ~ .~ ~

"F: ~ ~ i~ ~-:-~- II o~.

9 ;;~ ~ ~ L~,,,I "~ ~-~

~ ~ o ~o,.-.,~ o _~ E
 Carbonate dolomitization, Iowa and Ill&ois 247
broad subzones; zone 6, non-ferroan, moderate tized relics of the grains (Fig. 7b), and the
uniform luminescence distinctly brighter than generally excellent fit of the dolomite rhombs
zone 5. to the re-entrants in the early cements (Figs 4d,
These six zones have been interpreted as non- 7a, c).
marine cements based on their petrography and In some instances, rhombs sharply truncate
geochemistry. They lack the characteristics of major cement zones and hairline zones (Fig. 7c),
shallow warm-water marine cements, such as with no deflection or pinchout of zones adjacent
columnar or prismatic crystal morphologies, to rhombs. These rhombs have the same lumines-
microdolomite inclusions, and post-cement ma- cent signature as rhombs which block calcite
rine internal sediment. Furthermore, they are growth in other areas within the same samples,
low in Mg (mean Mg for zone 1 = 1600 ppm; and therefore the apparent truncation geome-
zone 2 = 7 2 5 p p m ; zone 3 = 6 3 5 p p m ; zone tries are interpreted as dolomite pre-dating
4 = 320 ppm; zone 5 = 890 ppm; zone calcite cements. The apparent truncation of
6 = 600 ppm; Grams, in preparation). Addition- calcite cement zones by dolomite rhombs can be
ally, cement-rich rocks contain no intraclasts of interpreted as dolomite either pre-dating, or
grainstones even within beds that experienced post-dating and replacing cement. On the other
slow deposition (glauconitic and fish bone hand, the deflection of cement zones by dolomite
horizons). Finally, the sequence of cathodolu- rhombs is readily explained by dolomite pre-
minescent zones, probably reflecting fluctuat- dating cement, and is difficult to explain by
ing Eh during precipitation, is consistent over dolomite post-dating cement. The irregular con-
large areas and throughout relatively large tact between dolomite I and non-luminescent
stratigraphic intervals. These features are diffi- cement in Fig. 7(c) is best interpreted as the
cult to reconcile with a model of cementation by cement peripherally replacing dolomite I. If
marine porewaters during progressive burial. dolomite I had replaced the non-luminescent
The six cement zones have been interpreted as cement, the dolomite rhomb would be likely to be
being pre-St Louis Formation in age. This age euhedral where it embays the calcite crystal, as is
determination is based on the presence of a the usual geometry where dolomite replaces
cement stratigraphy within St Louis grain rocks single-crystal crinoids (Figs 3b, 4a).
that differs from that in the Burlington-Keokuk The geometric relationships between dolomite
strata, and that is found syntaxially on zones 5 II and the calcite cement zones are the same as
and 6 in a few places in Burlington-Keokuk those between dolomite I and the cements (Fig.
rocks. Had the Burlington-Keokuk rocks been 4d). However, because dolomite II is a replace-
completely uncemented during cementation of ment of dolomite I, we cannot definitively
the St Louis Formation, the St Louis cements establish its age relative to the early cements
should have been precipitated directly on Burl- (zones 1 to 4). Dolomite II commonly is encased
ington-Keokuk crinoids. in calcite cement zone 5 in grain rocks (Fig. 4d)
Having established the calcite zonal strati- and in dolostones, and was rarely replaced by
graphy, the relative age of each dolomite can be calcite cement 5. These relationships date dolo-
interpreted from the geometric relationships mite II as older than calcite cement 5. Although
between cement zones and dolomite rhombs. For we favour dolomite II being older than cement
example, luminescent zones in calcite cements zone 4, we cannot rule out dolomite II having
are commonly deflected or pinch out near a replaced dolomite I after cement zone 4 and
dolomite I rhomb (Fig. 7a) and, in many cases, before cement zone 5.
where rhombs are in direct contact with crinoids Several lines of evidence indicate a post-zone
they completely prevent the growth of cement 4, pre-zone 5 age for dolomite III. Dolomite I
(Figs 4d, 7b). In these cases rhombs were not rhombs partially encased in early calcite cements
covered with cement until the late zone 5 or 6 have dolomite III rims developed only on
precipitated. These geometric relationships are surfaces not in contact with cement zones 1 to 4
interpreted to indicate that dolomite I deflected (Fig. 7d). These dolomite III rims are surrounded
or blocked the growth of early cements (zones 1 by cement zones 5 or 6 (Fig. 7d). The dolomite III
to 4), and therefore pre-dated calcite cement rims thus precipitated before zone 5 cements,
zone 1. Early cements did not grow over the otherwise the necessary porosity would not have
rhombs, presumably due to an inhibition of been available. There is no evidence to suggest
calcite nucleation on dolomite substrates. An- that the dolomite Ill rims are a mould-fill or a
other possible interpretation is that the dolomite calcite cement replacement after zone 5 cemen-
I rhombs replaced grains (peloids, bryozoans) tation. Dolomite III is often encased in calcite 5
that deflected the cements, in which case where it replaced and overgrew dolomite I and II
dolomite I could be younger than cement zone 4. in grain rocks (Figs 4d, 5a, c) and in dolostones
Arguing against this is the absence of undolomi- (Fig. 5d). Dolomite III also post-dates the
z48 D. C. Harris & W. J. Meyers

OD~o ~ 0Q

~ "-I "
O,-, ~,.1~ -~,~

. ~ ,,-~, ~ ~

~ m

,.-.'? 0 ~ 0 --

~.,~ o ~. "~

~ ~ ~.~ o .-~
 Carbonate dolomitization, Iowa and Illinois 249
formation of mouldic porosity in mud-supported crinoids is ruled out since lime mud would have
dolomite textures, where it forms overgrowths on been removed by pressure solution between
dolomite I rhombs which border crinoid(?) grains.
moulds (Fig. 8a). In these cases, dissolution of Stylolites always truncate dolomite I and II
the fossil preceded the growth of dolomite III rhombs, but at least some are older than dolomite
rims, and later zone 5 cements filled the mould. III. Evidence for this is the presence of dolomite
III rims on truncated surfaces of dolomite I (or
II) rhombs at stylolites (Fig. 9a).
Stratigraphic distribution of dolomite The above features indicate that dolomite I
and II rhombs have undergone inter-rhomb
Since all calcite cements in the Burlington- pressure solution, and were therefore formed
Keokuk are interpreted to have formed before early in the burial history of the Burlington-
mid-Meramecian St Louis time, the three dolo- Keokuk strata. This is consistent with the
mite generations are also interpreted to have interpretation that they were one of the earliest
formed before this time. Dolomite from the diagenetic events. Dolomite III post-dates inter-
Spergen Formation at Coppock Quarry (CP in granular pressure solution and at least some
Fig. 1), which immediately underlies the pre-St stylolitization.
Louis unconformity, contains rhombs with lumi-
nescence identical to dolomite II in the Burling-
ton-Keokuk rocks. St Louis and St Genevieve
lime mudstones above the unconformity in the Dolomite-chert timing
same quarry are undolomitized. Their lack of Timing of chertification has been interpreted
dolomite when compared to totally dolomitized from the petrographic relationships between
muds in the Burlington-Spergen sequence im- cherts and calcite cements in grainstone and
plies that dolomitization of the latter formations packstone beds. Most commonly, chertification
occurred before St Louis time. followed incipient calcite cementation, as indi-
cated by chalcedony coating and pseudomorph-
ing dog-tooth calcite crystals on crinoids. These
Dolomite-compaction relationships calcite cements are as old as zone 1; rarely,
The above observations imply that dolomitiza- outside chert nodules but in close proximity to
tion was one of the earliest major diagenetic them, chalcedony coats cements as young as zone
events affecting Burlington-Keokuk rocks. If 4. Fractures that cut chert nodules are filled with
this is correct then the dolomites would be cement zone 5. Rarely, chert nodules show no
expected to pre-date some of the compaction evidence of cementation before chertification.
processes. The relationship between mechanical Therefore, chertification began before calcite
compaction and dolomites is ambiguous, but cementation in a few cases, but in most nodules
commonly dolomite was involved in chemical began after incipient cementation. Chertifica-
compaction, as shown by sutured contacts tion occurred as late as zone 4 calcite cementa-
between rhombs of dolomite I (Fig. 8b) and tion outside of nodules, but in no instance post-
between rhombs of dolomite II. We have not dates zone 5. Therefore, any dolomite that pre-
observed sutured contacts between dolomite III dates chertification must pre-date zone 5.
rhombs. Pressure-solution contacts also occur With regard to dolomite-chert relationships,
between dolomite I rhombs and calcite cements most information comes from mud-supported
(Fig. 8c). In the latter, the calcite cement has textures. Within chert nodules, a complete
embayed the rhomb. The exact correspondence spectrum from undolomitized to totally dolomi-
of zone 2 with the embayed part of the dolomite tized textures are seen frozen by microquartz
rhomb makes a growth relationship interpreta- replacement. Least common are nodules that
tion highly unlikely, and dates the pressure contain no evidence of dolomite rhombs in
solution as post-zone 2. microquartz. These samples are from the oldest
Pressure solution also occurred between part of the largest chert nodules and represent
rhombs and crinoid grains. Compaction of lime mud that was replaced by microquartz
dolomitized packstones has resulted in fitted, before any dolomitization. Most commonly chert
pressure-solved contacts between crinoids, with contains scattered dolomite rhombs floating in
dolomite rhombs caught between grains (Fig. microquartz. In most cases rhombs are partially
8d). In these cases, dolomite rhombs embay the preserved (Fig. 9b), but others are ghosts, having
adjacent crinoids along pressure solution con- been nearly totally replaced by microquartz; the
tacts, and involve both dolomite I (Fig. 8d) and only dolomite remaining is minute dolomite
dolomite II, but never dolomite III. Dolomitiza- inclusions (Fig. 9c) which show up under
tion of lime mud after chemical compaction of cathodoluminescence. In some dolostones micro-
250 D. C. Harris & W. J. Meyers

~ .o~ ~

~ ~ ~--~.~ ~

9- ~ +-, ~ +-, ~_.~ .,,~

.~o,~,~ ~

~ ~,,-~ ~.~

~.~'~, ~.~ ,-

~ r ~ 0,'~ ~

, - ~ u'~--
~9 ~.-~ .~#

o ,. R= "~Zm

oo~, ~.~
 Carbonate dolomitization, Iowa and Ill&ois 25I
quartz chert filled inter-rhomb pores and re- fication of lime muds began before dolomitiza-
placed some rhombs, indicating that chertifica- tion, and argues against dolomitization begin-
tion followed complete dolomitization. ning on the seafloor.
The fact that dolomite rhombs occur in cherts
does not in itself prove that dolomitization
preceded chertification, and could be interpreted
conversely. The presence of rhombs pseudomor- Diagenetic sequence and summary of
phically replaced by microquartz (Fig. 9c) does, timing of dolomitization
however, prove that chertification was the later
event. Under cathodoluminescence, rhombs in The diagenetic sequence of the Burlington-
microquartz matrix are irregular, have corroded Keokuk carbonates is summarized here in order
edges, and are deeply embayed, the microquartz- to place dolomitization into the overall diagene-
dolomite contacts cutting across zoning (Fig. 9b). tic history, and thus to constrain the possible
These geometries are best explained as chert environments and models for dolomitization.
having partially replaced dolomite rhombs. Note The diagenetic sequence is summarized in Figs
that this chertification of rhombs involved both 10, 11 and 12. Figure 10 shows the relative timing
dolomite I and II, but never dolomite III. of each diagenetic event; the length of events and
Dolomite III occurs in cherts only in fractures, the time between events is arbitrary. Not all
where it has syntaxially overgrown fracture- events can be precisely fitted into the sequence,
truncated dolomite I and II rhombs (Fig. 9d). and their placement is marked by question
These fractures are filled with zone 5 calcite marks. Figures 11 and 12 are diagrams summa-
cement, which encases the dolomite III syntaxial rizing diagenetic sequences in grain-supported
overgrowths (Fig. 9d). Dolomite III is not seen and mud-supported rocks respectively.
on rhombs encased in chert, and because it forms Based on the foregoing evidence, dolomitiza-
on rhombs truncated by chert fractures, it is tion was one of the earliest diagenetic events to
interpreted as being a post-chertification event. affect the Burlington-Keokuk sediments, pre-
This timing correlates well with dolomite I I I - ceded only by some chertification. Most dolomi-
calcite cement relationships since it has been tization (I, II) occurred before all of the major
shown that dolomite III post-dates zone 4 calcite cementation, and before most of the
cement. chertification. Its early age is consistent with
In summary, chert-dolomite relationships dolomite being involved in inter-rhomb pressure
show that dolomite I and II formed before most solution. Based on our sampling of younger
chertification and therefore before zone 5 cal- Mississippian strata, the stratigraphic distribu-
cites, a timing relationship that agrees with that tion of dolomite and post-dolomite cements
determined from the dolomite-cement zones implies a pre-St Louis age for the dolomites.
relationships. This also implies that some cherti- Other features that are consistent with this

KINDER-
HOOK BURLINGTON & KEOKUK ST LOUIS

DOLOM ITI ZATION I "E

?[~1
DOLOMITE
DISSOLUTION ?Fr/'A?
PRE-I I II III IV V Vl
CALCITE
CEMENTATI ON fl~l? F//"A I ~ 17/7~ F'~ F'/~
CALCITE 1773
DISSOLUTION

CHERT IFICATION I
MECHANICAL
COMPACTION P[ I
CHEMICAL
COMPACTION [ INTERGRANULAR I
?I STYLOLITIC
DEDOLOMITIZATION
Pl'--
TI ME
FIG. 10. Sequence of diagenetic events for Burlington and Keokuk rocks in the study area. The relative
timing of each event is shown, and uncertain relationships are indicated by question marks. The length of
events and the time between events are arbitrary.
252 D. C. Harris & W. J. Meyers

DO2OMITE

C R I N O I D S ~ DOLOMITE '~']E~

(o) (b)
PRECIPITATION OF DOLOMITEI REPLACEMENTOF DOLOMITE I
BY DOLOMITE Z

• -CALCITE
PRESSURE CEMENT
ION ZONES

~-/ (c)
INITIAL CHEMICAL COMPACTION PRECIPITATION OF CEMENT
AFFECTING GRAINS B DOLOMITE ZONES 1 THROUGH 4

(e)
PRECIPITATION OF DOLOMITETIT
AS OVERGROWTHS8L REPLACEMENTS,
FOLLOWED BY CALCITE CEMENTS 5 8t 6
FIG. 11. Summary of diagenetic sequence in crinoidal grainstones. In (a) and (b), dolomite rhombs may either
be cements, or may have replaced microcrystalline components such as small peloid or bryozoan grains, or in
some cases small amounts of lime mud.
 Carbonate dolomitization, Iowa and Illinois 253

~' MICROQTZ
CHEIT k ' "
' LIME MUD ~ ,

7 , , / -
i

(a) (b)
PRE-DOLOMITE CHERTIFICATION DOLOMITE I REPLACING LIME MUD
FOLLOWED BY DOL.TF REPLACING DOL.T

,/

(c) (d) ~P()RE

CHERTIFICATION OF SOME PERVASIVE DOLOMITIZATION OF LIME
D O L O M I T I Z E D MUD ROCK MUDS. CHERTIFICATION OF SOME.

. : 4"CALCITE 5 :: .fPORE

/ (e) DOL.TITJ
DOL TrT
FRACTURING OF CHERT FOLLOWED BY
DOL. TITFOLLOWED BY CALCITE 5
IN FRACTURES & SOME I N T E R -
C R Y S T A L L I N E PORES.

FIG. 12. Summary of diagenetic sequence in mud-supported rocks. In (d) and (e), undolomitized lime mud has
been removed by dissolution during or after initially more pervasive dolomitization. It should be noted that
some beds remained completely composed of dolomite I, while others were nearly entirely replaced by
dolomite II before chertification.

Mississippian age for cherts and dolomites a,e as post-karst, which of course differs from the
the presence of chert and dolomite clasts in pre- timing of zones 5 and 6 proposed here.
Pennsylvanian palaeo-karst in the Burlington- If our pre-St Louis age is correct it constrains
Keokuk rocks of central Missouri (Daniels burial depths to less than 100 m during dolomiti-
1986). Petrography of calcite cements associated zation of the Burlington-Keokuk strata and to
with these palaeokarst features has led Daniels less than about 30 m during dolomitization of the
(1986) and Kaufman et al. (in press) to interpret Spergen sediments. If the dolomites were post-
cement zones 1 to 4 as pre-karst and zones 5 and 6 Spergen but pre-Pennsylvanian, it would imply
254 D. C. Harris & W. J. M e y e r s

burial depths less than about 500 m, based on freshwater systems could have been established
the thickness of Meramecian and Chesterian with only modest relative sea-level drops. Water
strata from central western Illinois (Willman table elevations of only about 2.5 m could have
et al. 1975). generated a freshwater lens 100 m thick (based
on the Ghyben-Herzberg relation), thick enough
to extend stratigraphically through the Burling-
Models for dolomitization ton. During late Mississippian, the proposed
fresh waters conceivably could have had re-
Accepting the pre-St Louis age for the dolomiti- charge areas in the La Salle Arch, or the
zation, we rule out deep burial models involving Transcontinental Arch. Considering that shal-
elevated burial temperatures. Similarly, the low water facies are common in the Spergen and
absence of peritidal facies within the Burling- younger strata, there could also have been short-
ton-Keokuk rules out synsedimentary tidalite lived emergences during their deposition that
models. Such models are particularly difficult to have not been clearly recognized, and that could
invoke considering that proposed Burlington- have resulted in flushing of the underlying
equivalent shoreline facies of the Lower Hum- formations with fresh water.
boldt beds in north-central Iowa are largely non- Although the mixing-zone model has many
dolomitic (Sixt 1983). We consider below four attractive aspects, it does not explain the three
other models: meteoric-marine mixing zone different generations of dolomite. It is not at all
waters, post-depositional hypersaline waters, clear why the three chemically distinct dolomites
marine or modified marine waters, and deep I, II and III should all form in the same general
subsurface brines that migrated into shallow diagenetic environment, nor why they should
burial settings. For background on these and show a progressive increase in iron. The C, O
other dolomite models, the reader should see the and Sr isotopic compositions of dolomite I is
reviews by Morrow (1982) and Machel & consistent with a mixing-zone model, but the
Mountjoy (1986). depleted O and radiogenic Sr isotopic composi-
tions of dolomite II are difficult to reconcile with
Mixing-zone model this model. It is possible that each episode of
dolomitization represents a separate sea-level
Several facts support a mixing-zone origin for event, separated by periods of sedimentation.
Burlington-Keokuk dolomites. Most apparent is Thus, the younger the dolomite, the greater the
the dolomitization of subtidal, open-marine burial depth during mixing zone dolomitization.
sediments which contain no supratidal, hypersa- Another problem with invoking the mixing-
line indicators, and which is regional in extent. zone hypothesis is the presence of two-phase
Also consistent with this model is the early fluid inclusions in the dolomites and calcite
timing of dolomitization. Dolomite probably cements. These have been studied by Smith
formed after some burial as implied by some of (1984) and Smith et al. (1984) who interpreted
the chert pre-dating dolomite I. This is based on them as primary inclusions. Homogenization
analogy with deep sea cherts, which do not form temperatures are in the range of 90-100~ and
on the seafloor, but begin forming only after final melting temperatures of frozen inclusions
some shallow burial (yon Rad & Rosch 1974, indicate bulk salinities in the range of 100 000-
Wise & Weaver 1974). On the other hand, 200 000 ppm. If these are truly pristine primary
dolomites (at least dolomite I) formed before inclusions, they clearly rule out a simple
calcite cementation. This in turn suggests that seawater-freshwater-mixing-zone.
dolomitization occurred before establishment of
a regional fresh groundwater system, since the
Hypersaline reflux model
calcite cements are interpreted as freshwater
precipitates (Harris 1982). During the subaerial Another possible model invokes the 're fluxing' of
exposure that resulted in the calcite cements, a brines from the St Louis or other post-Keokuk
mixing-zone would have formed during flushing Mississippian units. The St Louis Limestone can
of the sediments by fresh water, and this be eliminated as a likely source of brines, even
brackish water would have swept through the though it contains solution collapse breccias
sediments ahead of the freshwater lens. Thus, which are interpreted as former evaporites. Since
this mixed water would have been the first non- Burlington-Keokuk dolomite is restricted to pre-
marine water to permeate the sediments. The St Louis formations, and St Louis rocks are
fact that there was no calcite cementation during undolomitized in the study area, this model
precipitation of dolomite I and II is consistent would imply that dolomitizing fluids were
with this, as is the absence of dolomitized fossils formed during or immediately after Warsaw or
(they are preserved as moulds). Significant Spergen time. We have seen no evidence of
 C a r b o n a t e dolomitization, I o w a a n d Illinois 255

evaporitic or restricted marine depositional dolomitization took place under some shallow
conditions in the Warsaw. On the contrary, the burial, assuming analogy with extant deep sea
Warsaw contains open marine fossils. We cannot nodular cherts (e.g. yon Rad & Rosch 1974). If
rule out evaporitic facies having been present in this is true, then dolomite I could have formed
the Spergen Formation. Although we have not under a few metres of sediment in sea water of
seen evidence of evaporites in the Spergen, we normal salinity. Marine porewaters in lime muds
have not studied the unit in detail, and it is tend to be reducing due to bacterial decay of
possible that evaporites existed in upper Spergen organic matter, and likewise tend to involve
rocks and were removed by pre-St Louis erosion. bacterial reduction of sulphate with attendant
A check for 'occult' gypsum or anhydrite in production of Fe sulphide. Baker & Kastner
Burlington and Keokuk dolomites, using the (1981) have suggested that the removal of
methods of Beales & Hardy (1980), found no sulphate catalyses dolomite formation. This
evidence of gypsum or anhydrite inclusions. If chemical setting would conceivably result in Fe-
indeed refluxing hypersaline brines were the poor, Mn-bearing dolomites such as dolomite I.
main dolomitizing agent then we would expect to It would also involve bacterial decay of organic
see some gypsum or anhydrite moulds or calcite matter and therefore generate a wide range of
pseudomorphs in the pervasively dolomitized 613C values, including some very light values
mud rocks of the Burlington and Keokuk. An (Irwin et al. 1977, Hudson 1977). We do not see
argument for a brine reflux model involves the light C isotopic signatures for either dolomite I or
interpretation of Chowns & Elkins (1974) of II. Furthermore, the wide range of O isotope
geodes from Warsaw and upper Keokuk strata as values for dolomites I and II, many depleted in
replacements of anhydrite nodules. The strongest 180, cannot be explained by a seawater model at
evidence they present for this is their recognition shallow burial, nor can the radiogenic Sr in
of abundant length-slow chalcedony (a common dolomites II. Another problem with a simple
replacement of evaporites; Folk & Pittman marine model is the replacement of dolomite I by
1971), and relict anhydrite inclusions in quartz dolomites II and III. A marine model would
crystals in the geodes. imply rather uniform porewater chemistry, yet
The hypersaline model is consistent with the these dolomites probably precipitated from
early timing of the dolomites, and with the high waters chemically different from one another.
salinity fluid inclusions. The luminescent zoning A model invoking marine waters with slightly
of dolomite I is probably compatible with elevated salinities refluxing from younger forma-
precipitation from proposed refluxing and down- tions, such as proposed by Simms (1984), would
soaking brines which could have had low enough be consistent with the early timing of dolomites,
Eh's to incorporate Mn and Fe. As with the but not with the ranges of O isotopes for dolomite
mixing-zone model, dolomites II and III are not I and for dolomite II, nor with the radiogenic Sr
readily explained by the reflux model. It is not in dolomite II.
clear what might have led to the multiple
replacements. Furthermore, the high Fe contents
Basinai brine model
of dolomite II and III argue against their
precipitation in high sulphate, low Eh brines. The recognition of fluid inclusions containing
Had this been the situation, most of the Fe highly saline brine and having elevated homog-
probably would have been tied up in sulphides enization temperatures raises the possibility that
(Frank et al. 1982). Additionally, the wide range extraneous hot brines moved through the Burl-
in O isotope values for dolomite I, some of them ington-Keokuk sediments and dolomitized them
depleted in 180, are difficult to reconcile with (Smith 1984). Possible sources for these are
hypersaline evaporitic waters. Similarly, the subsurface brines expelled from nearby basinal
radiogenic Sr in dolomite II is difficult to areas, such as the Illinois Basin, by compaction,
reconcile with the reflux model, since the bulk of gravity, or tectonic related processes (Cathels &
the dissolved Sr in a refluxing brine would likely Smith 1982, Leach et al. 1984, Bethke 1986).
be from Mississippian sea water. These brines were possibly similar to and
genetically related to brines thought to have
formed Mississippi Valley type Pb-Zn deposits
Marine or modified marine waters
(review by Sverjensky 1984). In support of this
The early pre-cement and pre-compaction age idea is the widespread occurrence of vugs in
ibr dolomites I and rI, and the Osagean 'marine' Burlington, Keokuk and Warsaw strata, many of
Sr isotopes of dolomite I are consistent with which contain sphalerite, quartz, coarse calcite,
dolomitization on or near the seafloor. On the pyrite and, more rarely, kaolinite and saddle
other hand, the presence of some chertification dolomites, minerals that are common in Missis-
of lime muds before dolomitization suggests that sippi Valley-type ore deposits. In Iowa we do not
z56 D. C. Harris & W. J. Meyers
know the precise age relationships between these and was responsible for the replacement of most
vug minerals and the Burlington-Keokuk dolo- of the lime mud. Dolomite II has a dull red
mites and calcite cements, however, work in luminescence, is unzoned, and occurs mainly as a
Illinois and Missouri has shown these vug-filling replacement of dolomite I rhombs. Dolomite III
phases to cut cements as young as calcite 5 is non-luminescent, and occurs as a syntaxial
(Cander 1985, Kaufman 1985, Kaufman et al. in cement on, and replacement of dolomites I and
press). This young age is consistent with the II.
occurrence of sphalerites in lower Pennsylvanian (3) Dolomite petrography indicates that dolo-
strata in the Missouri-Iowa-Illinois region mite I and probably dolomite II pre-date all
(Leach 1973) and with the late Pennsylvanian- calcite cements. Luminescent zoning within
early Permian age proposed for the Mississippi calcite cement crystals is deflected and inhibited
Valley-type ores (Wu & Beales 1981). by the presence of these rhombs. Dolomite I and
Considering the above data, this model would II pre-date most chert, intergranular chemical
require an early phase of brine expulsion, before compaction, and stylolites. Dolomite III post-
most of the mineralization, and would require dates cement zone 4 and pre-dates zone 5. It
that high temperature fluids moved over long formed after all chert, after the formation of
distances at shallow burial depths (less than moldic porosity in dolomite I, and after some
100m). The marine Sr isotope and relatively stylolites. The stratigraphic restriction of dolo-
heavy O isotope signatures of dolomite I are not mite to pre-St Louis Formation units suggests
consistent with this model. On the other hand, that dolomitization occurred before or during the
the radiogenic Sr and 1sO-depleted isotope pre-St Louis unconformity.
signatures of dolomite II are consistent with a (4) A single dolomitization model cannot
brine model. Similarly, the increasing Fe con- reasonably explain all three generations of
tents of the younger dolomites (II and III) might Burlington-Keokuk Formation dolomites. Tim-
be explained by pulses of progressively more Fe- ing constraints coupled with petrographic and
rich brines. With regard to the fluid inclusions, geochemical characteristics suggest that dolo-
two-phase fluid inclusions similar to those in the mite I formed in a sea water-freshwater mixing-
dolomites are found in all the calcite cement zone associated with a meteoric groundwater
zones, and the two-phase fluid inclusions in system established beneath the pre-St Louis
dolomite I are similar to those in dolomite III unconformity. Dolomite II and III, on the other
(Smith 1984, Smith et al. 1984). Considering the hand, may have formed from externally sourced
differences in timing, cathodoluminescence and brines at moderate temperatures. These pro-
geochemistry between many of these phases, it posed brines became more Fe-rich with de-
raises the possibility that the fluids in the creasing age, and the required Mg was derived
inclusions, and their homogenization tempera- mainly from the precursor dolomite that was re-
tures are secondary. In summary, fluid inclusions placed.
provide evidence that high temperature saline
brines moved through the Burlington-Keokuk
sediments, but it is still an open question whether
these are samples of the same fluids that formed
the calcite cements and dolomites.

ACKNOWLEDGMENTS:This paper is derived from a MS

thesis by the senior author while a graduate student at
the State University of New York at Stony Brook.
Financial support from Amoco Production Company,
Research Center, Tulsa, for fieldwork and thin sections
is gratefully acknowledged. We would like to thank
Conclusions Brian F. Glenister, R. C. Hager and the Iowa
Geological Survey for their help during fieldwork.
Raid Brothers Quarries (Burlington, Iowa) and Kaser
(1) The distribution of dolomite in the Burling- Construction Company (Des Moines, Iowa) granted
ton and Keokuk limestones is controlled primar- permission to work in their quarries. Special thanks go
ily by distribution of lime mud-rich depositional to A. C. Kendall and W. J. Meyers for their initial
textures. Lime mud is pervasively replaced conception of the Burlington project, and their contin-
throughout the study area. ued support and encouragement. We thank John
Miller, Jim Marshall, and an anonymous reviewer for
(2) Three regionally extensive generations of the thoughtful and helpful suggestions for improving
dolomite can be recognized by their cathodolu- the manuscript.
minescent characteristics. Dolomite I, the oldest This paper does not necessarily represent the views
generation, is luminescent, usually thinly zoned, 3f the Standard Oil Production Company.
 Carbonate dolomitization, Iowa and Illinois 257

References
BAKER, P. & KASTNER, M. 1981. Constraints on --& MEYERS, W. J. 1985. Carbonate cement
dolomite formation. Science, 213, 214-6. stratigraphy of Burlington Limestone (Osagean)
BANNER, J. L. 1986. Geochemical constraints on the of Iowa: evidence for Eh gradients in a regional
origin of Burlington-Keokuk dolomites. PhD Mississippian paleD-groundwater system. Bulletin
Thesis. State University of New York at Stony of the American Association of Petroleum Geologists,
Brook. 69, 263 (Abstract).
--, HANSON, G. N. & MEYERS, W. J. in press. HARRIS, S. E. & PARKER, M. C. 1964. Stratigraphy of
Determination of initial Sr isotope compositions the Osage Series in southeastern Iowa. Iowa
of dolostones from the Burlington-Keokuk Fms. Geological Survey Report of Investigations I.
of Iowa, Illinois and Missouri: constraints from HUDSON, J. 1977. Stable isotopes and limestone
cathodoluminescence and glauconite paragenesis. lithification. Journal of the Geological Society of
Journal of Sedimentary Petrology. London, 133, 637-60.
BEALES, F. W. & HARDY, J. W. 1980. Criteria for the IRWIN, H., CURTIS, C. & COLEMAN, M. 1977. Isotopic
recognition of diverse dolomite types with an evidence for source of diagenetic carbonate
emphasis on studies of host rocks for MVT ore formed during burial of organic rich sediments.
deposits. In: ZENGER, D. H., DUNHAM, J. B. & Nature, 269, 209-13.
ETHINGTON, R. L. (eds). Concepts and Models of KAUFMAN, J. 1985. Diagenesis of the Burlington-
Dolomitization, pp. 197-213. Society of Economic Keokuk Limestones (Mississippian), eastern Mis-
Paleontologists and Mineralogists Special Publi- souri. MS Thesis. State University of New York at
cation, 28. Stony Brook.
BETHKE, C. i . 1986. Hydrologic constraints on the --, CANDER, H. S., DANIELS,L. D. ,~ MEYERS, W. J.
genesis of the Upper Mississippi Valley mineral in press. Calcite cement stratigraphy and cemen-
district from Illinois Basin brines. Economic tation history of the Burlington-Keokuk Forma-
Geology, 81, 233-49. tion (Mississippian), Illinois and Missouri. Journal
CANDER, H. S. 1985. Diagenetic history of the of Sedimentary Petrology.
Burlington-Keokuk Limestone, Illinois. MS The- LANE, H. R. 1978. The Burlington Shelf (Miss.) north-
sis. State University of New York at Stony Brook. central United States. Geologica et Palaeontolo-
CARLSON, i . P. 1979. The Nebraska-Iowa region. In: gica, 12, 165-76.
CRAIG, L. C. & VARNES,K. L. (eds). Paleotectonic - - & DE KEYSER,T. L. 1980. Paleogeography of the
Investigations of the Mississippian System in the late early Miss. (Tournaisian 3) in the central
United States,--Part I: Introduction and regional and southwestern U.S. In: FOUCH, T. D. &
Analyses of the Mississippian System, pp. 107-14. MAGATHAN, E. R. (eds). Paleozoic Paleogeog-
United States Geological Survey Professional raphy of West-Central United States, West-Cen-
Paper, 1010-F. tral U.S. Paleogeography Symposium I, pp. 149-
CATHLES, L. i . & SMITH, A. T. 1983. Thermal 62. Rocky Mountain Section of the Society of
constraints on the formation of Mississippi Val- Economic Paleontologists and Mineralogists,
ley-type lead-zinc deposits and their implications Denver.
for episodic basin dewatering and deposit genesis. LEACH, D. L. 1973. A study of the barite-lead-zinc
Economic Geology, 78, 983-1002. deposits of central Missouri and related mineral
CHOWNS, T. M. & ELKINS, J. E. 1974. The origin of deposits in the Ozark region. PhD Thesis. Univer-
quartz geodes and cauliflower cherts through the sity of Missouri, Columbia.
silicification of anhydrite nodules. Journal of --, VIETS, J. G. & ROWAN, L. 1984. Appalachian-
Sedimentary Petrology, 44, 885-903. Ouachita orogeny and Mississippi Valley-type
DANIELS, L. D. 1986. Diagenesis and paleokarst of the lead-zinc deposits. Geological Society of America
Burlington-Keokuk Formation (Mississippian), Program with Abstracts, 97th Annual Meeting, p.
Missouri. MS Thesis, State University of New 572 (Abstract).
York at Stony Brook. MACHEL, H. G. & MOUNTJOY, E. W. 1986. Chemistry
FOLK, R. L. & PITTMAN, J. S. 1971. Length-slow and environments of dolomitization--a reap-
chalcedony: a new testament for vanished evapor- praisal. Earth-Science Reviews, 23, 175-
ites. JournalofSedimentary Petrology, 41, 1045-58. 222.
FRANK, J. R., CARPENTER, A. B. & OGLESBY, T. W. MORROW, D. W. 1982. Diagenesis II. Dolomite--part
1982. Cathodoluminescence and composition of II: Dolomitization models and ancient dolostones.
calcite cement in the Taum Sauk Limestone Geoscience Canada, 9, 95-107.
(Upper Cambrian), southeast Missouri. Journal of PROSKY, J. L. in preparation. Trace element geochem-
Sedimentary Petrology, 52, 631-8. istry of regional dolomites in the Burlington-
GRAMS, J. in preparation. Trace element geochemistry Keokuk Formations (Mississippian) of Iowa and
of calcite cements in the Burlington-Keokuk Illinois. MS Thesis. State University of New York
Formations, Iowa and Illinois. MS Thesis. State at Stony Brook.
University of New York at Stony Brook. - - & MEYERS, W. J. 1985. Nonstoichiometry and
HARRIS, n. C. 1982. Carbonate cement stratigraphy trace element geochemistry of the Burlington-
and diagenesis of the Burlington Limestone Keokuk dolomites. Society of Economic Paleontolo-
(Miss.), S.E. Iowa, W. Illinois. MS Thesis, State gists and Mineralogists Annual Midyear Meeting
University of New York at Stony Brook. Abstracts, p. 73 (Abstract).
 258 D. C. Harris & W. J. Meyers
REEDER, R. J. 1981. Electron optical investigation of level. In: PAYTON, C. E. (ed.) Seismic Strati-
sedimentary dolomites. Contributions to Minera- graphy-Applications to Hydrocarbon Exploration,
logy and Petrology, 76, 148-57. pp. 83-97. American Association of Petroleum
SALLER, A. H. 1984. Petrologic and geochemical Geologists Memoir, 26.
constraints on the origin of subsurface dolomite, VON RAD, U. & ROSCH, H. 1974. Petrography and
Enewetak Atoll: an example of dolomitization by diagenesis of deep-sea chert from the central
normal seawater. Geology, 12, 217-20. Atlantic. In: Hs/2, K. J. & JENKYNS, H. C. (eds).
SIMMS, M. 1984. Dolomitization by groundwater-flow Pelagic Sediments . on Land and Under the Sea, pp.
systems in carbonate platforms. Transactions of the 273-99. Special Publication of the International
Gulf Coast Association of Geological Societies Association of Sedimentologists, 1. Blackwell
XXXIV, 411-20. Scientific Publications, Oxford.
SIXT, S. 1983. Depositional environments, diagenesis WILLMAN, H. B., ATHERTON, E., BUSCHBACH, T. C.,
and stratigraphy of the Gilmore City Fm. (Miss.) COLLINSON, C., FRYE, J. C., HOPKINS, M. E.,
near Humboldt, north central Iowa. M S Thesis. LINEBACK,J. A. & SIMON,J. A. 1975. Handbook of
University of Iowa, Iowa City. Illinois Stratigraphy. Illinois State Geological
SMITH, F. O. 1984. A fluid inclusion study of the Survey Bulletin, 95.
dolomite-calcite transitions in the Burlington and WISE, S. • WEAVER,F. 1974. Chertification of oceanic
Keokuk Limestones (Mississippian), S.E. Iowa, sediments. In: HsO, K. & JENKYNS, H. (eds).
W. Illinois. MS Thesis. State University of New Pelagic Sediments: on Land and Under the Sea, pp.
York at Stony Brook. 301-26. Special Publication of the International
--, REEDER, R. J. & MEYERS, W. J. 1984. Fluid Association of Sedimentologists, 1. Blackwell
inclusions in Burlington Limestone (Middle Mis- Scientific Publications, Oxford.
sissippian)---evidence for multiple dewatering Wu, Y. & BEALES, F. W. 1981. A reconnaisance study
events from Illinois Basin. Bulletin of the American by paleomagnetic methods of the age of mineral-
Association of Petroleum Geologists, 68, 528-29 ization along the Viburnam Trend, southeastern
(Abstract). Missouri. Economic Geology, 76, 1879-94.
SVERJENSKY,D. A. 1984. Oil field brines as ore-forming ZENGER, D. H., DUNHAM, J. B. & ETHINGTON, R. U
solutions. Economic Geology, 79, 23-37. (eds) 1980. Concepts and Models of Dolomitization.
VAIL, P. R., MITCHUM, R. M. & THOMPSON, S. 1977. Society of Economic Paleontologists and Minera-
Seismic stratigraphy and global changes of sea logists Special Publication, 28.

DAVID C. HARRIS, Standard Oil Production Company, 5400 LBJ Freeway, Suite 1200,
Dallas, TX 75240, USA.
WILLIAM J. MEYERS, Department of Earth and Space Sciences, State University of New
York, Stony Brook, NY 11794, USA.
 The diagenesis of the Great Estuarine Group, Middle Jurassic,
Inner Hebrides, Scotland

J. D. Hudson & J. E. Andrews

S U M M A R Y: The Bathonian Great Estuarine Group consists of sandstones, silty shales

and mainly shelly limestones, deposited in micro-tidal brackish lagoons in a warm, seasonal
climate. The rocks show, in general, a lack of intense early diagenesis or deep-burial
diagenesis, and are thus suitable for the study of shallow-burial diagenetic processes. Early
diagenetic changes can be directly linked to depositional environment. They were only
important on the lagoon margins, where cyanobacteria flourished in schizohaline 'algal
marsh' settings and dolomite formed in response to evaporation of low-salinity lagoonal
waters. During burial to a few hundred metres, and after initial compaction, cementation by
ferroan calcite occurred. This was pervasive in limestones and formed large concretions in
sandstones. Quartz and feldspar grains in sandstones were marginally corroded by the
porewaters that precipitated calcite, and some feldspars had previously or concurrently
suffered partial solution. Aragonitic mollusc shells were replaced by calcite in permeable
rocks but remained unaltered in shales. Widespread volcanism in the Palaeocene buried the
area beneath basaltic lavas and dykes were intruded. These events produced little
metamorphic or burial-diagenetic change, except in close proximity to minor intrusive
contacts or in the vicinity of plutons. The evolution of porewaters can be monitored in the
carbon and oxygen isotopic composition of diagenetic calcites. After initial burial, they were
dominantly of meteoric derivation. Vitrinite reflectance measurements confirm the mild
thermal history of most of the sediments, and show the local effects of heating by igneous
intrusions.

The Great Estuarine Group, a paralic sequence middle two are the most important in the present
of sedimentary rocks, is an attractive subject for context, but we want to stress the interdepen-
diagenetic studies because it is lithologically dence of all stages.
varied and fossiliferous, thus containing a wide
variety of sedimentary particles of differing
diagenetic potential. In most of its outcrop area, The Great Estuarine Group
it has suffered only shallow burial, and some of
the 'unstable' minerals have survived virtually The islands of the Inner Hebrides (Fig. 1) expose
unaltered (e.g. shell aragonite and smectitic Mesozoic, mainly Jurassic, rocks that rest on
clay). Elsewhere, the rocks have been subjected Precambrian and Palaeozoic basement and are
to higher degrees of diagenetic alteration and overlain by Tertiary volcanics: they are also cut
even, in some areas, thermal metamorphism. by Tertiary intrusions. A similar succession is
Because it was mainly deposited in waters of less found in the submarine area of the Minch
than normal salinity, sub-littoral synsedimentary between the mainly Precambrian rocks of the
lithification was limited. It therefore provides a Scottish mainland and the Outer Hebrides.
useful contrast to some other sequences where Although direct evidence of palaeo-shorelines is
early marine cementation was pervasive. The lacking, we believe that the Inner Hebrides-
name of the Group emphasizes that the deposi- Minch region approximates to a depositional
tion, as well as the diagenesis, of the sediments Minch Basin, bordered on either side by swells
involved the interplay of at least two types of that were, at least in part, fault-bounded (Hud-
water, marine and meteoric. Although not all son 1983). The basin is crossed by a major fault,
aspects of Great Estuarine Group diagenesis the Camasunary Fault (Fig. 1), that itself had
have been studied as thoroughly as we would episodes of movement during the Mesozoic as
like, we do now have enough diverse information well as subsequently, but was probably largely
to attempt a synthesis of the diagenesis of the quiescent during the time that most concerns us.
Group as a whole, and to relate this to its burial It divides the Minch Basin in the broad sense
history. into the Sea of the Hebrides basin to the NW,
In this paper we consider the different stages and the Inner Hebrides basin to the SE separated
in the evolution of these rocks: derivation, by a palaeo-high in central Skye. The main
deposition, early diagenesis, burial diagenesis, Jurassic exposures follow the N N E - S S W trend
metamorphism, uplift and erosion. Of these the of the basin and its margins and hence the

From: MARSHALL,J. D. (ed.), 1987, Diagenesisof Sedimentary Sequences, 259

Geological Society Special Publication No. 36, pp. 259-276.
260 J. D. Hudson & J. E. Andrews

!
N

Rhum
. .

/~J! ~Iuu
. . .

.4 Outcrop
i I
c/ ~MUCK ~-~ Major Tertiary
L~ ( __ / ~
~ranamurc~ ~-~..Igneous
"~
Centres

FIG. 1. Map showing outcrop of the Great Estuarine Group (black), and distribution of the major Tertiary
igneous centres, northern Inner Hebrides. Fault line marked 'C' shows the position of the Camasunary fault.

depositional strike. There is limited evidence of limestones sandy or marly. Nevertheless it is of

westward thinning on Skye, and of southward interest to compare the approximate proportions
thinning from a depocentre in North Skye. of the major rock types. Based mainly on the logs
The Great Estuarine Group (Fig. 2) is a presented by Hudson & Harris 0979) and Harris
paralic episode in an otherwise marine Jurassic & Hudson (1980), the proportions for the main
succesion. During the accumulation of some outcrop (Trotternish, Strathaird, Eigg) are given
280 m of sediment (present, compacted thick- in Table 1. The Group is sandier than the world
ness) the depositional surface was never more average of sediment types (e.g. Garrels &
than a few metres from sea-level, yet despite the Mackenzie 1971, Pettijohn, Potter & Siever
great variety of facies seen in vertical sequences, 1972), but is probably a reasonably representa-
lateral variability is remarkably small (apart tive sample of epicontinental, near-source sedi-
from deltaic sands) along the 90 km of outcrop mentary rock.
from North Skye to Muck, and probably
originally further. Because of the lack of good
Derivation
zone fossils the diachroneity of facies cannot be
accurately assessed, but it was probably small The composition of the coarser sand detritus was
(Hudson 1980). investigated by Hudson (1964) and Harris
0984). In the Sea of the Hebrides basin much of
the sand is feldspathic (5-25~ of total grains).
The sediments
Orthoclase and microcline are roughly equal in
The sediments of the Great Estuarine Group are average abundance, plagioclase only minor. In
lithologically diverse but not well differentiated: the Inner Hebrides basin feldspar forms less than
most of the shales are more or less silty, many of 5~ of the coarse detritus. Heavy minerals are
the sands argillaceous and carbonaceous, the dominated by the stable species zircon, tourma-
 Great Estuarine Group diagenesis, Inner Hebrides 26I

Z
STRATIGRAPHY SCHEMATIC INTERPRETATION LEGEND
LOG
0
_1
-I
f I t
,r STAFFIN
Belemnite Sands Mb
0 B A Y FM
upper u s t r e a Mb
~Skud~urgh Fm
..-.. __~_c ~
"/
Shallow marine
Fluvial :iliiS.a,e
Sandstone
~ ~ Freshwater lagoon
"

KUmaluag Fm
.. - .f~
i I Limestone
i
Marine-brackish i I
Duntulm Fm
~ ,.~ ~-, lagoon
i i

Tabular
cross beds
Trough
cross beds
GREAT ~ Concretions
ESTUARINE V a l t o s Sst Fm ,-.. Fluvial delta
GROUP Algal beds
Oysters
Brachiopods
Lealt 9 ~ ~) r, Bivalves
Shales Lonfearn Mb ~ ~ , Brackish-marine Gastropods
..... lagoon n~- Belemnites
Fm Kildonnan Mb
~, Ammonites
Elgol Sst Fm Fluvial delta Ostracodes
-- ? - - Branchiopods
"~Cullaidh I ] ~ ~ Shallow marine
BEARRERAI( Shale Fm sand sheet
S S T FM I I
clay sand
(grain size)

FI~. 2. The stratigraphy of the Great Estuarine Group (after Harris & Hudson 1980) including a schematic
graphic tog of the Group (Trotternish) and palaeoenvironmental interpretations. The black and white scale
bar divisions represent ~ 25 m.

TABLE 1. Proportionsof different lithologies in the ultimate source and the Jurassic basin (Old Red
Great Estuarine Group Sandstone, Permo-Triassic?) remains unclear.
Studies on Sm-Nd model ages of detritus in the
Total Shale/ Sand- Lime- Kilmaluag Formation of the Isle of Muck (Inner
thickness silt stone stone Hebrides basin) are consistent with a largely
(m) ~ ~
Dalradian source (Andrews et al. in press) while
Trotternish 264 43 42 15 studies on Sr in the dolostones favour an
Strathaird 137 68 18 14 important ancient limestone component in the
Eigg 129 50 39 11 hinterland. Studies of clay mineralogy have
revealed an abundance of illite-smectite mixed
Weighted mean
(Composite (530) 51 35 14 layer clays (Andrews 1987), probably signify-
thickness) ing a volcanic source. This is not seen in the
sandstone-dominated formations. Most prob-
The weighted mean has been computed by adding total ably the source area was, from time to time,
thicknesses, thus emphasizing the thicker Trotternish mantled by fine-grained volcanic ash, which was
succession. transported into the basin along with the fine-
grained weathering products of the bedrock.
line and rutile, but with substantial proportions The sourcelands were thus within a few tens
of metamorphic minerals. Garnet is most abun- of kilometres of the basin of deposition. They
dant in the north and staurolite in the south, with were, for long periods, deeply weathered, with a
a few samples rich in epidote from NW Skye. soil mantle containing a volcanic-derived com-
These data indicate a dominant source from the ponent, and well vegetated. At these times they
Scottish mainland. A minor Lewisian source delivered fine-grained sediment to the basin. At
contributed in the NW, probably from an Outer other times they were sharply uplifted, delivering
Isles landmass ~(Hudson 1964). The r61e of coarse, unweathered sand with terrestrial plant
intermediate sources between a metamorphic debris as a minor but important part of the load,
262 J. D. Hudson & J. E. Andrews
via short, rapidly flowing rivers and their deltas supply and the open sea. Hence salinity
(Harris 1984). varied between 0 and 35%o,and the chemistry
of all but the most dilute waters was
controlled by the marine component. In
Deposition
'closed' lagoons the chief supply was fresh
The detrital sediment, and factors such as depth water, and its concentration by evaporation
of water, occasional exposure, conditions of produced waters chemically and isotopically
turbulence, turbidity and rate of sedimentation different from sea water, with different
were not very different between the Great faunas and different diagenetic sequences.
Estuarine Group and many shallow-water ma- This became particularly important in the
rine formations. Yet the rocks are different, and Kilmaluag Formation (Andrews 1985).
this is because of the predominantly brackish-
water deposition of the Group.
(1) The brackish waters favoured a narrow
Early diagenesis
range of taxa of shelled invertebrates, mainly The conditions of temperature, salinity etc., that
molluscs, but high productivity meant that we have listed (Table 2) were fundamental in
these few flourished exceedingly (Hudson controlling those diagenetic reactions that took
1963, 1980). Some of these shells occur place on the lagoon floor or immediately beneath
abundantly in shales and in rock-forming it, where pore waters were closely related to
abundance in limestones, in almost monoty- depositional waters and re-working resulted in
pic shell beds. Apart from their obvious the same particle repeatedly experiencing 'depo-
quantitative importance, they give excellent sitional' and 'diagenetic' effects. This stage
opportunities for comparative diagenetic corresponds to early, penecontemporaneous,
studies, both between types of shell and synsedimentary, or eo-diagenesis as used by
between shales and limestones. Oysters various authors. In the present case it is useful to
(Praeexogyra hebridica) in the Duntulm For- distinguish between the lagoon floor, on which
mation were primary low-Mg calcite. Prae- the sediments were under almost perpetual,
mytilus strathairdensis in the Kildonan Mem- though shallow, cover of water from lagoon
ber, Lealt Formation, was composed of margins where exposure was important.
nacreous aragonite. Neomiodon spp., were As in most sedimentary sequences, the coarse-
composed of crossed-lamellar aragonite. grained clastic sediments show the least effect of
High biological productivity is also revealed early diagenesis. Quartz (and even feldspar) have
by the abundance of phosphatic fish and low diagenetic potential for alteration by low
reptile remains and by an organic carbon temperature dilute waters and relatively mobile,
content of 1-2~ in many of the shales. organic-poor porewaters did not precipitate
(2) The well-preserved mollusc shells, together authigenic minerals. For converse reasons the
with other evidence, allow some estimates to shales show more obvious diagenetic effects, and
be made of the water types present during in the Great Estuarine Group they are linked
deposition, particularly their isotopic compo- with the limestones which commonly occur as
sition and temperatures (Tan & Hudson thin beds within them. Nevertheless, on the
1974, Hudson 1980). The results are summar- lagoon floor early diagenetic effects were less
ized in Table 2. An important distinction obvious than in many contemporaneous marine
between 'open' and 'closed' coastal lagoons sediments. In particular, cementation was inhib-
(Hudson 1980) is worth re-stressing. In 'open' ited: little early carbonate cement formed and no
lagoons, which dominated the lower forma- hardgrounds are found, nor was there much
tions of the Group, free connection was at all obvious dissolution of aragonite or calcite. On
times maintained between a freshwater the lagoon margin, however, or where run-off

TABLE 2. Setting of Great Estuarine Group deposition

Palaeolatitude 35~176
Temperature 15-25~
Water depth 0-10 m (?); micro-tidal
Salinity 'open' lagoons 0 (fresh water) to 35%0 (sea water)
'closed' lagoons 0 (fresh water to evaporated low SO~ waters
61sO of water (%0, SMOW) - 6 (fresh water) to - 1 (sea water) (increased by evaporation)
~13C of ppd carbonate (%0, PDB) - 3 (fresh water) to + 3 (sea water)

Compiled from papers referred to in the text, particularly Tan & Hudson (1974) and Hudson (1980).
 Great Estuarine Group diagenesis, Inner Hebrides 263

and desiccation affected closed lagoons, more

obvious changes took place: calcite and dolomite
could form.

The lagoon floor: clays and silts

The main clay mineral assemblages were deter-
mined by derivation rather than diagenetic
changes (Andrews 1987), but iron-rich minerals
are strongly linked to depositional environment.
This is shown by the colour of the rocks.
Typically, marine-brackish mudrocks are pale to
dark grey, due to pyrite formed by sulphate
reduction in the porewaters, and to organic
matter. Freshwater mudrocks are generally pale
green to dark greenish or brownish-grey. Near
absence of pyrite presumably allows the greenish
tints to be revealed. Finally, and related to the FIG. 3. Photomicrograph of a limestone from the
filling of the lagoon, the silts of the Skudiburgh Lonfearn Member, Lealt Formation, Trotternish,
Formation are typical red-brown alluvial 'red- Skye. Molluscan shell fragments are variably coated
with micrite envelopes or oolitic sheaths. That in the
beds', with grey-green reduced mottlings (An-
centre shows 'algal' borings; the one beneath it a
drews 1985). fractured micrite envelope.

Lagoon floor carbonates

These are confined to the Lonfearn Member,
Most of the limestones are thin ( < 0 . 5 m ) Lealt Formation.
laterally extensive (tens of metres minimum,
The shells in the limestones are frequently
sometimes several kilometres) molluscan shell
intensely 'bored', probably by cyanobacteria.
beds, interbedded with shales or silts. All
The borings are filled with either fine-grained
transitions from shell layers to shell limestones
calcite or pyrite. Shell beds of types 3, 4 and 5
can be found and the faunas of the limestones are
often contain a mixture of marginally bored and
the same as those of the shales that enclose them.
pyritized shells with almost pristine ones, thus
This argues against long-distance transport. The
showing how winnowing mixed shells which had
shell beds to not show the characteristics of storm
a long residence time on the lagoon floor, or even
deposits (Aigner 1985): no erosional bases, no
within the sulphate reduction zone of the
internal grading. They are probably the result of
sediments, with newly dead ones. Many of the
repeated winnowing events and illustrate the
bored fragments also show well-developed mi-
interplay of depositional and diagenetic pro-
crite envelopes or coatings, beneath which the
cesses. The Neomiodon shell beds of the Lealt
deeper borings extend for a few tens of microns
Formation, and the finer-grained facies of the
(Fig. 3). Oolitically coated grains in the Lonfearn
Valtos Formation, can be divided into different
Member, Lealt Shales, retain radial fabrics.
types in general similar to those from the
These ooids are now composed of ferroan
Purbeck Formation of southern England de-
dolomite, and their high ~lsO values were
scribed by E1-Shahat & West (1983). Some of the
interpreted by Tan & Hudson (1971) as showing
major types are:
that the dolomite is early diagenetic. By analogy
(1) Shales or silts with scattered shells, articulat- with modern radial ooids from a broadly analo-
ed but not in life position. gous depositional setting, the Laguna Madre,
(2) Shales with shell layers, mostly whole but Texas (Land et al. 1979), it is likely that the
disarticulated valves, convex-up. Lonfearn Member ooids were Mg calcite origin-
(3) Thin (few centimetres) shelly limestones ally; however, we note that radial fabric is not a
containing both articulated shells and single single criterion with which to distinguish former
valves, as well as shell fragments. carbonate mineralogy (Richter 1983). Dolomiti-
(4) Similar shelly limestones with a preponder- zation presumably took place beneath the sul-
ance of broken and rounded shell fragments, phate reduction zone, allowing Fe z+ to become
and generally finer-grained. incorporated in the dolomite.
(5) Limestones with rounded shell fragments Coated grains also sparsely occur in the
and abundant coated grains (ooliths) (Fig. 3). Duntulm Formation, Isle of Muck, where ?Cor-
Z64 J. D. Hudson & J. E. Andrews

bula shells display radial oolitic coatings of non- adjacent lagoon, followed by desiccation, result-
ferroan calcite in what appears to have been ed in the precipitation of some gypsum and
a storm winnowed horizon. MgCO3 is c. brecciation of the calcified algal 'heads' forming
0.70 mol~ enriched in the ooid cortex relative to an 'algal gravel'. Intense bacterial diagenesis in
the burial diagenetic ferroan spar which prob- organic-rich carbonate sediments produced a
ably means that the ooids were originally spectrum of diagenetic textures (cement fringes,
composed of a Mg calcite. spherulites, microspars) quite unlike anything
seen in the subaqueous deposits of the lagoon
floor.
Carbonate concretions
Septarian calcite concretions up to about 1 m The closed lagoon margin: dolostones of the
across and 20 cm thick occur at several horizons,
Kilmaluag Formation
most notably near the middle of the Lealt
Formation in the upper Kildonan and lower In the Great Estuarine Group depositional
Lonfearn Members. The beds concerned are dolomite is virtually confined to the low-salinity,
shale-limestone alternations with the concre- closed lagoon facies of the Kilmaluag Forma-
tions occurring in shales, but commonly close to tion; a fact in good general accord with the
limestone beds. The shales have faunas varying sulphate-inhibition theory of dolomite formation
from bed to bed from fresh water to marine- (Baker & Kastner 1981). Within the Kilmaluag
brackish. The ~i13C values ( - 6%0) of the concre- Formation, dolomite occurrences correlate
tions are low but not extreme, and can be strongly with evidence of evaporation and
explained as derived from mixtures between desiccation. Evidence from carbon and oxygen
sulphate reduction bicarbonate and bicarbonate isotopes agrees with an origin from evaporated
derived from the dissolution of fossils incorpor- meteoric waters, with dolomite forming via
ated in the shales. Their 6180 values (-2%0) dolomitization of a pre-existing calcite (probably
confirm their shallow-burial origin (Tan & Mg calcite) mud during seasonal episodes of
Hudson 1974). Calcite which filled the septarian evaporation (Andrews et al. in press). Once
cracks was introduced later, but the timing of the again, the variable conditions of the lagoon
opening of the cracks is unknown (cf. Astin margin promoted more active diagenetic change
1986). More rarely, sideritic concretions a few than the constant conditions of the lagoon floor.
centimetres long occur in the freshwater facies of Once formed, the aphanitic Kilmaluag dolomite,
the Kilmaluag Formation. enclosed within fine-grained rocks, escaped
At this diagenetic stage the rocks, apart from substantial diagenetic change during further
concretions, were not lithified, nor had aragonite burial.
been transformed to calcite. The limestones were
still porous and the muds little compacted. Pores
The final stage of lagoon closure: alluvial
were filled with water closely related to the water
calcretes of the Skudiburgh Formation
of deposition, with loss of much sulphate and
some carbonate. Within the mottled clays of the Skudiburgh
Formation white-grey nodular concretions, typi-
cally 2-8 cm in diameter occur. These are usually
The open lagoon margin: algal marsh deposits,
cracked in 'septarian style' and the ground mass
Duntulm Formation
of non-ferroan calcite microspar contains spher-
If the lagoon floor was diagenetically quiet, the ulitic structures of probable bacterial origin.
lagoon margin, whose deposits are only sparsely These concretions have been interpreted as
represented in the record, was very active. This is calcrete nodules (Andrews 1985). Their stable
especially well shown in the 'algal beds' of the isotope geochemistry, 613C values c. -10%o,
Duntulm Formation (Hudson 1970, Andrews 6180 values - 3 to -4%0, suggest formation from
1986). The setting was that of a supra-littoral meteoric derived soil-water which may have
fringe to a micro-tidal marine-brackish lagoon: undergone some evaporation.
the fact that the open lagoon was not far away is
shown by marine fossils washed on to the supra-
littoral deposits. The setting was classically Burial diagenesis
schizohaline. Most growth of the cyanobacter-
ium (blue-green alga) Cayeuxia probably took The deposition and early diagenesis of the Great
place during periods of abundant fresh water Estuarine Group certainly took place during
flushing from rainwater and run-off, and during Bathonian times on or just below the lagoon
this time its sheath was calcified. Periodic floor, so the time and place of the processes are
inundation of the marsh by sea water from the well defined. With burial the sediments entered a
 Great Estuarine Group diagenesis, Inner Hebrides 265

more obscure phase in which the relative, and mainly on North Skye. Although variations
still more the absolute, timing and depth of occur across the basin, it will apply in general
burial at which different events happened are terms throughout.
hard to discern. As most diagenetic changes are Deposition of the Group was terminated by
fluid-rock interactions, the nature of the the transgression of the Staffin Bay Formation in
porewaters is also critical. We can look for clues, the latest Bathonian (Bradshaw & Fenton 1982).
mostly to relative order of events, in the Slow but relatively steady subsidence ensued
sediments themselves, and to some extent recon- during the deposition of the Staffin Shale
struct the burial history from independent Formation through the Callovian, Oxfordian
evidence. and Lower Kimmeridgian (Sykes 1975), to a total
As any sediments are buried, certain changes observed thickness of 120 m. We do not know
take place which distinguish burial from near- how much sediment was deposited above this,
surface diagenesis. Light is excluded, tempera- and before an important phase of pre-upper
ture stabilizes and then gradually increases, Cretaceous erosion (Hudson 1983), but a maxi-
biological activity apart from bacteria ceases and mum of 500 m seems generous. The Upper
compaction starts. Adjacent beds that may have Cretaceous marine strata are thin (few metres)
experienced very different conditions at and and discontinuous, and there is no evidence of
immediately after formation are now subject to former great thicknesses in the region. There was
similar processes. Slow processes such as com- a further brief period of erosion before the onset
paction or convection-driven water flow have of volcanism in the Palaeocene. The earliest
time to take effect. It does not follow, however, volcanics are palagonite tufts, but most of the
that burial diagenesis is a one-way sequence of lavas are continental, with frequent evidence of
changes. Uplift can reverse the temperature weathering between lava flows. The preserved
trend, and new sources of water can traverse the thickness of lavas on Skye is about 600 m
sediment due, for instance, to hydrologic flow (Emeleus 1983); a former maximum thickness of
down aquifers in response to uplift of nearby 900 m (G. P. C. Walker, personal communica-
land, or to changes in sea-level. tion) is used on Fig. 4. Since high lava fields may
Because burial diagenesis is the hardest of the have been built up, depth below land surface
stages to define and understand, it is convenient rather than sea-level is shown. Most of this
to approach it by excluding other effects. As well activity was probably concentrated in a short
as working forwards from deposition and early period of time from about 60 to 58 Ma; the lavas
diagenesis, we can work backwards from the on Eigg and Muck may be older, about 63.5 Ma
present. The effects of modern weathering on our (Dagley & Mussett 1986), and the Sgurr of Eigg
rocks are rather easy to recognize and hence to pitchstone is the youngest well-dated igneous
avoid during sampling. The Tertiary igneous rock in the area at 52+ 1 Ma, early Eocene
activity that affected our region may have had (Dickin & Jones 1983, Dagley & Mussett 1986).
subtle effects related to heat flow and depth of This later activity was probably not widespread
burial--we discuss this below--but the influence in its effects whereas the earlier volcanic episode
of contact metamorphism near minor igneous could have involved a regional increase in
intrusions is easy to recognize and avoid. The heatflow (Emeleus 1983, p. 389).
regional effects of the Skye plutonic centre and The uplift and erosional history of the area
its hydrothermal aureole are also clearly distinct after the igneous activity and prior to the
from burial diagenesis and well recognized from Pleistocene glaciations is poorly known. Wat-
independent evidence (Taylor & Forester 1971). son's (1985) review suggests that the Palaeocene
This process of elimination leaves us with the was a time of general uplift in Scotland, and that
results of 'normal' burial diagenesis, which we substantial deformation of the lavas took place
believe were mostly completed during the period shortly afterwards. Oligocene non-marine sedi-
of initial burial of the sediment pile. In support of ments rest on lavas in the Canna basin. Most
this conclusion, blocks of lithified Jurassic probably the present land outcrop of the lavas,
sediments, principally limestones, occur in Terti- and the associated plutonic centres, were uplifted
ary volcanic explosion breccias in North Skye, during the same episode of deformation, that is
Strathaird and Muck. They do not differ in field prior to the late Oligocene.
appearance from equivalent lithologies in nearby
in situ outcrops. Compaction
The finer-grained rocks have suffered substan-
Burial history tial compaction and some of the shales of the
Figure 4 shows a simplified version of the burial Lealt Formation are well-laminated 'paper sha-
history of the Great Estuarine Group based les' containing crushed bivalve shells. More silty
266 J. D. Hudson & J. E. Andrews

TIME (MY) THEORETICAL

160 140 100 60 0 GEOTHERMS
0 1 , I , | , l , I J I = I i l , I 20 20
t,

...: ...~....o: Io'~

L- ~--~::.:o<:o:.o:.-. 4"5 @
OJ. vp~ ~'-"~176 t '~,~ I.':I"~
500 _- "%7%~ ~ /:.:) ~,

40 6o
I,~" .
/:r ~

1000 .-:~ :-T.:~V/ 50 80

:' '.:>" ;i JURASS,C

~-=L~176 3 0 ~ C KM
I -lOO
BURIAL 60 52 PALAEOCENE
DEPTH (M) Palaeocene - 1
Volcanism 60~ KM

FIG. 4. Postulated burial history for the Great Estuarine Group, based on north Skye. For discussion, see text.
Two geotherms are shown. The 'Jurassic' geotherm is a conventional 30~ km -1 assumption and probably
applies except during Palaeocene volcanism. At that time the geothermal gradient was higher, and an
arbitrary doubling to 60~ km-1 is shown.

and calcareous shales, on the other hand, show Several observations bear on their
much less compaction, and all fine-grained rocks interpretation.
are soft, porous and friable when not close to
igneous contacts. Meade (1966) and Rieke & (1) Sandstones less than 1 m thick are nearly
Chilingarian (1974) (e.g. fig. 17, p. 42) show that always fully cemented whereas thicker ones
a large range of compaction values are to be are concretionary.
found in argillaceous rocks at depths of burial of (2) In the upward-coarsening deltaic sand units
a few 100 m, depending on grain size, compo- that typify the Valtos Sandstone Formation,
sition and other properties of the shales con- the lower parts of the sand bodies are flat-
cerned. Qualitatively, the rocks are similar to bedded, and contain carbonaceous and
those of the Jurassic of the East Midlands of argillaceous partings. These are partially
England, which have never been buried to a cemented by concretions bounded by bed-
depth of more than about 500 m and are in a ding planes, though rounded in plan view.
stable region not affected by volcanic activity. The classic ellipsoids occur in the upper
Limestones show some evidence of pressure parts of sand bodies where permeability is
solution, with microstylolites sometimes de- more homogeneous (Fig. 5a).
veloped between adjacent allochems. More ex- (3) The uppermost part of a sand body is often
tensive stylolite development characterizes some a coarse sand or granule conglomerate
of the 'algal beds' of the Duntulm Formation, about 1 m thick, which is usually also rich in
where dissolution of Cayeuxia nodules has taken Neomiodon shell debris. This is typically
place adjacent to clay seams. Compaction phe- fully cemented, and half-ovoid concretion-
nomena in sandstones are discussed below. ary masses appear as though 'suspended'
below it.
(4) There may be a more general correlation
Concretionary cementation of sandstones
between the occurrence of shells in the
By far the most striking diagenetic phenomenon sandstones and adjacent shales and the
in these rocks is the occurrence of very large (up occurrence of concretions. Some of the
to more than 1 m radius) ellipsoidal calcareous more densely concretionary sands are also
concretions (doggers, nodules) in the Valtos shelly, whereas the coarsest and thickest
Sandstone Formation. These are well seen at the sand on Eigg (Division C, Hudson & Harris
type locality in Trotternish but even better on 1979) is only sparsely concretionary, and the
Eigg. Elgol Sandstone Formation, which is unfos-
 Great Estuarine Group diagenes&, Inner Hebrides 267

FIG. 5. (a) Calcite cemented concretions in the Valtos Formation, Valtos cliff section, Trotternish, Skye. The
concretions are about 1.5 m in diameter. Note contrast between 'flat' lower concretion and rounded upper
one. (b) Compaction of bedding in uncemented sandstone around a concretion. Valtos. Hammer is 38 cm
long.

siliferous and enclosed in sparsely fossilifer- concretions. The pressure dissolution of shells
ous shales above and below, is scarcely must clearly have happened before the concre-
cemented by calcite at all. tions formed (observation 7). Compaction con-
(5) In some units, notably the upper part of the tinued after concretion growth also, at least
Valtos Sandstone Formation in Trotternish, around some of the concretions (observation 6).
concretions are very rich in whole or broken This phase of compaction was accompanied or
Neomiodon shells, whereas the uncemented preceded by dissolution of shells outside the
sand outside the concretions is unfossilifer- concretions (observation 5). Petrographic obser-
ous, showing no trace of the shells that it vations (8) suggest that at the time of concretion
must formerly have contained. formation Neomiodon shells were composed of
(6) Whereas sedimentary lamination and cross- aragonite, as they still are in adjacent shales. This
bedding look similar at first, whether would explain the scarcity of nucleation points
within concretionary masses or outside for calcite (hence the large crystals) and the
them, in some instances at least bedding is enclosing of several then-aragonitic shells in one
deflected around concretions. About 15- poikilotopic calcite crystal. The chemistry of the
20~ compaction has been observed in the concretions also agrees with a burial diagenetic
few examples measured (Fig. 5b). origin, and contrasts with that of early-diagene-
(7) In thin sections from within the concretions tic septarian concretions of the shales in the
one commonly observes pressure-solution Great Estaurine Group (Tan & Hudson 1974),
of shells against quartz and feldspar grains,
and sometimes shells have been reduced to
mere films (Fig. 6).
(8) In the field, on bright days such as are,
surprisingly to some, common in the Hebri-
des, the calcite cements glisten with lustre-
mottling, revealing crystal sizes up to
centimetres across. This is confirmed in
thin section, where poikilotopic calcite
crystals may enclose hundreds of sand
grains and several Neomiodon shells. Fre-
quently, more than one shell forms part of a
single large poikilotopic crystal continuous
from cement to neopmorphic replacement.
(9) The calcite of the cements is always ferroan.
(10) The isotopic composition of the calcite is FIG. 6. Photomicrograph of a calcite-cemented
generally within the same range as that of sandstone from a concretion (cf. Fig. 3). A
limestones within the Great Estuarine Neomiodon shell, about 0.4 mm thick, shows pressure
Group (Tan & Hudson 1974). solution against quartz grains; this must have
happened before the sandstone was cemented. The
Taken together, these observations clearly replacement calcite in the shell is slightly pseudo-
establish the burial diagenetic origin of the pleochroic. Valtos Formation. Plane-polarized light.
268 J. D. Hudson & J. E. Andrews
and with that of carbonates clearly related to the Valtos Formation sandstones on Eigg. The
Tertiary igneous activity (see below). dykes are typically about 1 m wide, and them-
The formation of these concretions after initial selves have now largely weathered away. Their
burial but before deepest burial allows some baked margins, about 0.5 m wide, stand proud,
further observations and speculation concerning and are fully cemented at the actual contact (Fig.
their growth in relation to other events. Berner 7). Away from this, cementation is concretion-
(1980, p. 124) showed that sands will pass ary, with numerous small, more or less coalescent
through a concretionary stage on their way to concretions typically about 2 cm across, and
becoming fully cemented, rather than being gradually dying away into the surrounding
cemented via a 'cementation front': essentially sandstone. Evidently the short-lived intense
because fluids can continuously flow past a heating associated with dyke intrusion induced
growing concretion. A concretionary sand thus rapid precipitation (or re-distribution) of calcite,
results when the supply of CaCO 3 runs out before and precipitation was in the form of many small
cementation is complete: the more concretionary concretions. Slow, steady precipitation during
sand bodies in the Great Estuarine Group, such burial diagenesis produced, by contrast, the
as Division E, Valtos Sandstone Formations, fewer large concretions described above. The
Eigg, are about 20% cemented, as recent work by isotopic composition of the tiny concretions is
M. Wilkinson (University of Leicester) shows. indicative of precipitation from heated waters
These considerations say nothing about the size (see below).
of concretions (many small or few large ?): in
these sandstones, as in many others, large
concretions are the rule. Berner (1968) gives Mineral stability
formulae for the rate of concretion growth, and Thin sections of the large concretions show
using his recommended values for the various marginal corrosion of quartz and feldspar grains
parameters involved (some of which, such as against calcite cement, and cracks within grains
degree of CaCO 3 supersaturation, are poorly are calcite-filled and possibly enlarged. Apart
known), the time for formation of a spherical from this corrosion, many K-feldspar grains
concretion of radius 1 m can be calculated. A appear unaltered in thin section. However,
pure CaCO 3 concretion growing by diffusion recent SEM observations show that decomposed
only in stationary porewater would take feldspar grains also occur, and many of these are
4.55 Ma to form. If growth in ground-water plagioclase (M. Wilkinson, personal communi-
flowing at 3 m yr -1 (Bathurst 1975) is considered cation).
the time is reduced to 1.44 Ma. These figures are Certain heavy minerals show differential
only approximate, but they do indicate that these preservation in cemented versus uncemented
concretions could have taken in the order of 1 sands: kyanite, staurolite, sphene, epidote (Hud-
Ma to form. This is a small part of the time son 1964, Harris 1984). The data may be
available, if that is reckoned to be the period of summarized to give a stability ranking (Table 3).
late Jurassic and early Cretaceous initial burial, Comparison with Morton's (1984) data clearly
say 40 Ma. shows that loss of these minerals is related to
A very different geometry of cementation burial diagenesis rather than weathering. Be-
occurs alongside Tertiary basaltic dykes that cut cause of the differential preservation in concre-

FIo. 7. (a) Cemented margins of a weathered

basaltic dyke in Valtos Formation sandstone. Isle of
Eigg. Dyke is about 1 m wide. (b) Detail of a dyke
margin, showing small concretions in sandstone.
Same locality. Hammer head 15 cm.
 Great Estuarine Group diagenesis, Inner Hebrides 269
TABLE 3. Stability ranking of heavy minerals under different conditions of alteration

Morton (1984) This work

'Deep burial' 'Weathering' Great Estuarine

pH > 7, saline pH < 7, fresh Group

Unstable: Olivine, pyroxene Olivine, pyroxene

Andalusite, sillimanite Amphibole m

Amphibole Sphene Amphibole

Epidote Apatite Epidote
Sphene Epidote, garnet Sphene
Kyanite Chloritoid, spinel Kyanite
Staurolite Staurolite Staurolite
Garnet Kyanite Garnet
Apatite ] Andalusite ]
Chloritoid ~- Sillimanite
Spinel .J Tourmaline ]
Stable: Rutile, tourmaline, zircon Rutile, zircon Rutile "]
Tourmaline ~-
Zircon )

tions, most of this diagenesis must have taken der it pseudopleochroic (Sandberg & Hudson
place since concretion growth. The near-absence 1983).
of the highly unstable minerals that must have (4) Praemytilus commonly retains its aragonitic
been available at source, notably amphiboles, is mineralogy and nacreous shell micro-struc-
best explained by loss during source weathering, ture even in fully-cemented sparites, al-
transport, or pre-concretion diagenesis (Hudson though these are only a few centimetres thick
1964). and enclosed in shales. Evidently nacreous
microstructure was more resistant to replace-
ment than crossed-lamellar, possibly because
Cementation and neomorphism in shells and
of its higher organic content (Hudson 1967)
limestones
and layered nature.
The commonest type of limestone in the Great (5) Praeexogyra shells, being composed of low-
Estuarine Group is a molluscan biosparite or Mg calcite, are very resistant to diagenetic
biosparrudite, but finer-grained limestones also alteration. Their foliate microstructure is
occur, and mollusc shells occur in sandstones and always well-preserved, even when the shells
shales as well as in limestones. We summarize have been compacted and cemented into a
here some of the variations related to lithology 'micro-breccia'.
and original shell mineralogy.
Spar cements and shell neomorphism
Shell preservation
As recorded above, shelly limestones generally
(1) Shells enclosed in shales still generally retain lack early cements. The predominant cement is a
their original mineralogy and fabric: foliate post-compactional ferroan spar, resulting in
calcite in the case of Praeexogyra hebridica, fabrics identical to those of the 'compacted
nacreous high Sr (4800 ppm) aragonite in biosparrudites' of E1 Shahat & West (1983, fig.
Praemytilus strathairdensis, crossed-lamellar 6). Fabrics comparable to their 'early-lithified
aragonite, lower in Sr (3700 ppm) in Neomio- biosparrudites' (El Shahat & West 1983, fig. 7)
don spp, to name the three commonest are rare. The ferroan spar of the cement is
bivalves involved. essentially identical to that which replaces
(2) In thin, argillaceous 'earthy' limestones even aragonitic molluscs, except that Sr is enhanced in
Neomiodon may retain chalky aragonite (cf. the replaced shells. Single crystals commonly
E1-Shahat & West 1983), Praemytilus and transect shell-margins and post-date fracturing
Praeexogyra retain mineralogy and fabric. of the shells (Sandberg & Hudson 1983). The
(3) In all purer limestones Neomiodon is replaced isotopic and elemental compositions of the spar
by sparry calcite, but the calcite may retain are those typical of burial cements (Table 4;
aragonite and organic inclusions which ren- Andrews 1986, Sandberg & Hudson 1983).
27o J. D. Hudson & J. E. A n d r e w s

TABLE 4. Chemical characteristics of burial spar crospar, to 'beef' veins. This suggests that
cements microspar neomorphism and 'beef' veining were
contemporaneous, and the ferroan nature and
Elemental composition (ppm) carbon and oxygen isotopic composition of the
Mg < 2500 calcite involved shows that both were burial
Sr 200-600 diagenetic. Marshall (1982) and Stoneley (1983)
Mn c. 1000 discuss the possibility that such veins were
Fe 6000-13 000
formed during episodes of over-pressuring of the
Isotopic composition, %o PDB shales, to help account for their displacive
613C 0 to +2*
6130 - 10 to - 6 growth.
Marshall's example from the Valtos Forma-
* More negative ~13C values (to -4.6) occur in algal tion of Eigg showed a sharp change in chemistry
limestones in which early diagenetic calcites are also and isotopic composition at a petrographically
more negative (Andrews 1986). Data from Andrews recognizable boundary, as well as gradual
(1986), Sandberg & Hudson (1983) and Tan & Hudson changes along the lengths of the fibres. These
(1974). results are best explained by the introduction of
new sources of porewaters, as well as gradual
Microspars porewater evolution (Marshall 1982).
Most of the fibrous calcite veins appear to
Most of the finer-grained limestones are micro-
belong to 'normal' burial diagenesis. However it
sparites rather than micrites. They have been
is noticeable that in Strathaird there are many
most thoroughly studied from algal limestones of
such veins, some not parallel to bedding. These
the Duntulm Formation (Andrews 1986). Micro-
may relate to the Tertiary hydrothermal system
spar is typically composed of 4-30 I-tm, inclusion-
of that region. The example from Strathaird
rich calcite in equant to loaf-shaped crystals.
studied by Marshall (1982), however, is a burial-
Crystals are often separated by clay seams and
patches (Andrews 1985, fig. 13). Molluscan shell diagenetic one. Its isotopic composition suggests
it pre-dated the Tertiary heating and survived it
debris, peloids and microfossils appear to 'float'
without isotopic change.
in the microspar. It is undoubtedly of neomor-
phic origin (Bathurst 1975, pp. 484-6).
Both ferroan and non-ferroan microspars
occur. The non-ferroan variety is particularly
The effects of Tertiary volcanism
common in argillaceous limestones and probably and burial
formed in early diagenesis, perhaps within the
sulphate reduction zone where Fe was incorpor- More than 1 m or so from minor intrusions such
ated into pyrite. It is often transitional to ferroan as basaltic dykes, and more than a few kilo-
microspar which presumably formed under metres from the major plutonic centre of Skye,
deeper-burial conditions where iron was avail- metamorphic effects are far from obvious,
able for incorporation in the carbonate. In its despite the fact that the whole area was buried
turn this microspar can pass into pseudo-spar, beneath hundreds of metres of basaltic lavas, and
true void-filling spar, or into fibrous displacive that these contain zeolite minerals deposited by
calcite. circulating hot waters (Walker 1970).
The 6180 values of microspars ( - 5 to -9%0) We have investigated some of the problems
overlap with those of true spars but are generally involved by studying rocks from outcrops cate-
somewhat heavier. 613C values are lighter ( - 0 . 2 gorized as follows:
to -6.7%0) reflecting the incorporation of some
organic-derived carbon (Andrews 1986) (cf. Figs (1) Outcrops with few or no minor intrusions,
9a and 10). not close to major sills, and lacking signs of
baking, e.g. Duntulm and Staffin Bay, Skye;
Fibrous calcite veins Kildonan, Eigg.
(2) Outcrops in areas with many minor intru-
Many of the shales include beding-parallel layers sions, including thick (10 m + ) sills as well as
of fibrous calcite ('beef'), a few millimetres to a dykes, in which there is pervasive baking of
few centimetres thick. The thicker examples may the country rocks, e.g. Lealt area, Skye.
show cone-in-cone structure. Commonly, these (3) Strathaird, which is within 5 km of the
layers occur above or below thin shelly lime- margin of the Skye plutonic complex, and
stones or calcareous sandstones, or they may within its hydrothermal aureole (Taylor &
occur within limestones along shale partings. In Forester 1971).
such cases transitions can be observed from (4) Samples from within about 0.5 m of an
normal microspar, to fibrous-and-displacive mi- igneous contact.
 Great Estuarine Group diagenesis, Inner Hebrides 27I
In category 1, shales are soft and crumbly, usefully be considered in the light of our
smectitic clay minerals are preserved, and shale alteration categories, and of the burial history
colours are either pale grey, greenish, or reddish- (Fig. 4).
brown, depending on depositional facies. Ara- The Ro values show most samples from
gonitic fossils are preserved in shales, and retain category 1 below 0.5, with seven of them below
depositional 6180 values. 0.3. Some higher values, up to 0.8, are from
In category 2, shales are harder and more outcrops also yielding values <0.3. With one
brittle, but fissile, and are darker coloured. exception, all category 4 samples have Ro > 2,
Former aragonitic fossils are normally either and those from Strathaird >1.5. Category 2
represented by moulds only, or replaced by samples are, as expected, transitional, ranging
calcite. from 0.5 to 1.3.
In Strathaird, category 3, all shales are dark There is an interesting exception to these
grey or black, and have lost fissility. Aragonitic simple findings. The Duntulm Formation out-
fossils are never preserved. The average 6180 of crop in Lon Ostatoin, Trotternish, has crumbly
limestones, especially fine-grained ones, is green shales with preserved aragonitic fossils
lowered by about 5%o compared to their facies (Tan & Hudson 1974, p. 120). Yet R0 values
equivalents elsewhere (Tan & Hudson 1971, range from 0.7 to 1.3 (Fig. 7), like those from
1974). Smectitic clays are lacking (Andrews category 2 shales. The section at Lon Ostatoin
1987). contains two sandstones cemented with calcite of
In category 4, all these conditions are accentu- 6180 - 15 and - 19%o (Tan & Hudson 1974), and
ated, and even calcitic Praeexogyra shells may be is situated 30 m below a picrite-dolerite sill about
recrystallized, but these effects die away rapidly 100 m thick.
outwards from the intrusion in areas otherwise There is much difference of opinion on the
belonging to category 1 (Tan & Hudson 1971, p. interpretation of Ro values in terms of burial
763, fig. 2). Smectitic clays are lacking, the history and temperature. Many workers, e.g.
crystallinity of illite is increased, and some Waples (1980) believe that Ro is determined by a
chlorite is formed. sediment's total history, with an approximate
doubling of the rate of maturation for every 10~
rise in temperature (Lopatin theory). Price
Vitrinite reflectivity (1983), however, holds that maximum tempera-
We have limited data on vitrinite reflectivity on ture reached effectively defines R0 in sedimen-
samples from throughout the outcrop (measured tary basins. A similar conclusion as regards
by R. Lee and B. S. Cooper, University of water-dominated hydrothermal systems was
Newcastle upon Tyne, in 1975). There is no reached by Barker (1983).
obvious correlation with position in the basin, Lopatin calculations (Waples 1980) and com-
stratigraphical position, or inferred distance parison with data from the continuously subsid-
beneath the basalts. The results (Fig. 8) can most ing basin of the North Sea (Lerche et al. 1984)

~Unaltered sections 9 Slightly 'baked'

sections 9
9 igneous contacts
ILIUnaltered sections I~ Strathaird ,
t_l (Lon Ostatoin) ~ away from intrusions

Catagenesis Metagenesis
No
of Obs

.2 .~ ,.o i.~
i
2.0
i
2.~
L
31o 31~ 4.'o
,.o
i
..~

FIG. 8. Histogram of vitrinite reflectivity values (Ro, ~) classified according to apparent degree of thermal
alteration of shales. For discussion see text.
272 J. D. Hudson & J. E. Andrews
confirm the mild thermal history of our category (above) to late cavity-filling dolomites and
1 samples. calcites (Andrews 1986). However none of these
Barker's (1983) graph should be applicable to cements are volumetrically important, and se-
the hydrothermal aureole of the Skye plutonic veral are merely the partial fillings of voids left
centre (Taylor & Forester 1971). The Ro values over from earlier cementation. No doubt Terti-
from Strathaird suggest temperatures of approxi- ary hot waters traversed other porous rocks too,
mately 200-240~ This agrees with an estimate but without precipitating any carbonate.
of at least 200~ from spore colouration (Riding The only region in which pre-existing calcites
1984). (but not dolomites) appear to have re-equilibrat-
ed with heated waters is Strathaird (Tan &
Hudson 1971, 1974). This is consistent with the
Tertiary calcite cements
> 200~ temperatures probably reached there.
Although most cements in the limestones and
sandstones are burial diagenetic products, there
are several examples which we relate to Tertiary Evolution of water chemistry
heating either by spatial association, like the
small concretions along dykes mentioned above, Since the formation and diagenesis of the
or because they have unusually low 6180 values. reactive minerals represents a series of water-
The lightest apparently 'normal-diagenetic' cal- mineral interactions, with or without the inter-
cites are some calcite concretions at -13.7%o, vention of organisms, it is useful to try to trace
but the overwhelming majority of diagenetic the evolution of the waters concerned. Most of
calcites are not lighter than -10%o. Values less our information concerns the carbonates. We
than -15%o are therefore regarded as probably can only work back from the carbonate to the
related to Tertiary hot water. They are geo- water by making assumptions. In many cases we
graphically widespread, from North Trotternish can only do it at all because some metastable
to Muck. They range from fossil fills (like those carbonates have persisted through later events,
recorded from the underlying Bearreraig sand- with their compositions largely unchanged. The
stone by Marshall 1981) to apparently normal rocks in general are not, and never have been
cements of thin sandstones (like those in Lon equilibrium assemblages.
Ostatoin associated with anomalous Ro values, The carbon isotopic composition of a carbon-
see above), and from dyke-margin concretions ate (Fig. 9) should be a direct reflection of the

'Sulphate Reduction' 'Marine

v
'Decarboxylation' Carbonate' -'-~'Fer mentation'~--

Hot
Cements
Sst Concretions ] Burial
Septarian Lst & Sst Cements
Septarian
I Diagenetic
Concretion
Body Concretion Spar ~. Cements
}M~crites
Dolomite Dolomite
Calcite Dolo 1
micrites[ Early
(_29)__~Loch Bay Duntulm Fm Algal I Diagenesis
j ,sis j
Lealt Fm

------I Fossils

-15 -10 - 0
'
5 '
613

FIG. 9. Summary of carbon isotopic compositions of carbonate fossils and cements (313CpDB).
Characteristic 6~3C ranges for different types of carbonate (Curtis 1977) are shown. Most fossils have 613C
values near the marine range. Early diagenetic calcites are lighter, reflecting 'soil gas' and 'sulphate-reduction'
carbon sources. Burial diagenetic cements have 613C similar to fossils. Data from Tan & Hudson (1974) and
Andrews (1986).
 Great Estuarine Group diagenesis, Inner Hebrides 273
613C of the dissolved carbonate, fractionations deposition ranged between about 15 and 25~
and temperature effects on them being small Applying these temperatures to other shell
(Anderson & Arthur 1983). The oxygen isotopic analyses allows the 6180 of water in the lagoons
composition depends on both 6180 of water and to be calculated (Fig. 10). Early diagenetic
temperature, so a unique solution is not normally calcites, probably precipitated during meteoric-
possible, but as we discuss, interpretations can water flushing on algal marshes of the lagoon
be constrained within reasonable limits. margins (Andrews 1986) have 6180 very similar
During deposition, the main agents of carbon- to that of freshwater fossils. Both approaches
ate formation within the lagoons were molluscs. suggest that the isotopically lightest fresh water
Both aragonite and calcite were produced by in the lagoon had a (~aso of about -5%0. The
different forms. ~13C values are related to heaviest lagoonal waters were probably associ-
palaeoecology, being generally lighter in fresh- ated with evaporation, either on closed-lagoon
water shells (Tan & Hudson 1974). 6180 reflects margins where dolomites formed (Andrews et
salinity in 'open lagoonal' situations, both being al., in press) or possibly even in the lagoons
determined by seawater-freshwater mixing, but themselves (some Praeexogyra analysed by Tan
the situation is again not simple (Tan & Hudson & Hudson 1974).
1974). The molluscan faunas themselves are Early diagenesis just within the sediment was
certainly a better guide to palaeosalinities than the time of maximum bacterial metabolism of
their isotopic compositions. However, molluscs organic matter, and this is reflected in the very
from just above the Great Estuarine Group wide range of 613C values in the precipitated
(Staffin Bay Formation and basal Staffin Shales) carbonates. Algal limestones are consistently
are fully marine and give apparently reliable light in carbon (Tan & Hudson 1974, Andrews
palaeotemperatures using conventional assump- 1986). Septarian concretions also have light
tions (Tan et al. 1970). Assuming ~ s o of ocean carbon and normal-depositional oxygen. Many
water was -1%o in the Jurassic (Shackleton & of the early diagenetic calcites are relatively Mg-
Kennett 1975), and allowing for possible season- rich, suggesting that they were Mg calcites
al growth (Hudson 1968), temperatures of shell originally. The good preservation of these cal-

/_o~'/ /
Temperature ~ / <>#// /
I 0 10 20 30 40 50 0'~
g8 w SMOW

Tertiaryl
Meteoric I
Water I

Burial I
Diagenetic I
Water I

-' .....? ; 2 -
Freshwater
-2 ....: S

Marine-Brackish
Water

g~)CPDB
2 0 -4 -6 -8 -10 -12 -14 -16 -18 -20
Marine-Brackish /
Fossils .... /
Freshwater
Fossils & Burial (? Warm)
Early Diagenetic Diagenetic Spars Hot Cements
Carbonate
FIG. 10. Interpretation of oxygen isotopic composition of carbonates (318Oc, PDB) in terms of temperature
and isotopic composition of water (618Ow, SMOW). Temperature lines correspond to equilibrium precipitates
(6180~) from water (61sOw) according to the standard palaeotemperature equation (Anderson & Arthur 1983).
MBF = marine-brackish fossils, FWF = freshwater fossils, BDS = burial diagenesis spars, HWC = hot-water
cements. Data from Tan & Hudson (1974) and Andrews (1986). Discussion in text.
274 J. D. Hudson & J. E. Andrews

cites, even when very fine-grained as in 'algal' lighter, or warmer, formation waters. It should
tube calcifications, shows that they were not be emphasized that, although meteoric-derived,
aragonite: indeed we have no evidence of these burial diagenetic waters were probably by
inorganic aragonite precipitation. All the early no means 'fresh'. They were certainly reducing,
calcites are non-ferroan, being formed either in Fe-rich, and probably more or less saline.
oxic environments or in the sulphate reduction Polyphase growth of fibrous calcite veins shows
zone. Although these early diagenetic carbonates that more than one groundwater type traversed
are very informative, they are volumetrically these rocks.
minor. Again we have an obscure interval before the
There ensued a period in which carbonates heating associated with Palaeocene volcanism
were evidently not precipitated, and this inclu- apparently triggered the precipitation of small
ded initial compaction of the rocks, fracturing of quantities of calcite and dolomite cement. At this
shells, and pressure solution of aragonite against time the porewaters could have been isotopically
quartz in sandstones. Then, for unknown rea- lighter in oxygen, because Palaeocene rainwater
sons, calcite precipitation started again, giving probably had a 6180 down to - 12%o (Taylor &
the main phase of burial cementation. This was Forester 1971). However by the time it got into
pervasive throughout the Group affecting shales the Jurassic rocks (and assuming it got there at
(except when very impermeable), limestones, all), it would probably already have exchanged
and sandstones. The importance of cross-forma- with the heated rocks it traversed. Figure 10
tional water flow is indicated by the ferroan assumes 6180 between - 7 and -10%o for
nature of the calcites, because Fe (and Mn) Palaeocene groundwater, and this implies tem-
cannot have been obtained from within the purer pertures around 100~ for the precipitation of
sandstones and limestones. The formidable the lightest calcites we observed. These calcites
nature of the cementation problem at this stage is often only partially fill the pores in which they
indicated by Berner's (1980, p. 124) calculation formed, and are quantitatively minor compared
that at least 300 000 pore volumes are needed. to normal burial cements.
The 613C of the late spars is virtually identical to After this, very little seems to have happened
that of the commonest fossils, strongly suggest- to the Great Estuarine Group rocks. They were
ing that dissolution of fossil carbonates was the uplifted, faulted, jointed and exposed to the
main source of carbon, rather than organic modern erosional cycle. But the late Quaternary
reactions (Fig. 9). The 6180 values (Fig. 10), glaciations stripped off any weathering mantle
assuming a maximum temperature of formation that may once have enveloped them, and the
of 35~ (500m burial, 30~ km -1 geothermal rocks we now see are, apart from trivial recent
gradient, 20~ surface temperature) imply 61sO weathering, probably much as they were in the
of porewaters of between - 4 . 5 and -7%o (Fig. Eocene. Indeed, as we have stressed how small
10). These overlap with meteoric water values the Tertiary igneous effects were, some of them
calculated from freshwater fossils and early are probably much as they were in the
freshwater precipitates. The slightly lighter Cretaceous.
values are also reasonable because much of the
meteoric water which entered the sediments
during burial no doubt flowed down porous ACKNOWLEDGMENTS: We thank Jim Harris for use of
aquifers, having been precipitated on the hilly information on heavy minerals from his thesis, and our
hinterland rather than in the basin itself. There several co-workers, particularly Francis Tan, Jim
are several potential aquifers within the Jurassic Marshall, Tony Fallick, Joe Hamilton, Rob Raiswell
and Mark Wilkinson, for discussions in the field and
succession, both within the Great Estuarine
laboratory. JEA thanks NERC for a studentship held
Group itself and beneath it (Hudson 1983). The at the University of Leicester and for the use of
lightest supposed burial cements, from some of facilities at SURRC, East Kilbride and the University
the sandstone concretions, have 6180 values of Edinburgh. We thank Stuart Haszeldine and Gill
down to -13.7%o. These imply somewhat Harwood for helpful reviews.

References
AIGNER, T. 1985. Storm Depositional Systems. Lecture Economic Paleontologists and Mineralogists, Tul-
notes in Earth Sciences, 3. Springer-Verlag, sa, Oklahoma.
Berlin. ANDREWS,J. E. 1985. The sedimentary facies of a late
ANDERSON, T. F. & ARTHUR, M. A. 1983. Stable Bathonian regressive episode: the Kilmaluag and
isotopes of oxygen and carbon and their applica- Skudiburgh Formations of the Great Estuarine
tion to sedimentologic problems. In: ARTHUR,M. Group, Inner Hebrides, Scotland. Journal of
A. et al. (eds). Stable Isotopes in Sedimentary the Geological Society of London, 142, 1119-
Geology. Short Course Notes No. 10, Society of 37.
 Great Estuarine Group diagenesis, Inner Hebrides 275

- - 1 9 8 6 Microfacies and geochemistry of Middle

. - - 1 The petrology of the sandstones of the
9 6 4 .

Jurassic algal limestones from Scotland. Sedimen- Great Estuarine Series, and the Jurassic palaeo-
tology, 33, 499-520. geography of Scotland. Proceedings of the
- - 1987. Jurassic clay-mineral assemblages and their Geological Association of London, 75, 499-
post depositional alteration: upper Great Estuar- 528.
ine Group, Scotland. Geological Magazine, 124, -- 1967. The elemental composition of the organic
261-71. fraction, and the water content, of some recent
, HAMILTON,P. J. & FALLICK, A. E. In press. The and fossil mollusc shells. Geochimica et Cosmochi-
geochemistry of early diagenetic dolostones from a mica Acta, 31, 2361-78.
low-salinity Jurassic lagoon. Journal of the Geologi- - 1968. The microstructure and mineralogy of the
-

cal Society of London, shell of a Jurassic mytilid (Bivalvia). Palaeonto-

ASTIN, T. R. 1986. Septarian crack formation in log),, 11, 168-82.
carbonate concretions from shales and mudstones. - 1970. Algal limestones with pseudomorphs after
-

Clay' Minerals, 21, 617-31. gypsum from the Middle Jurassic of Scotland.
BAKER, P. & KASTNER, M. 1981. Constraints on the Lethaia, 3, 11-40.
formation of sedimentary dolomite. Science, 213, - 1980. Aspects of brackish-water facies and faunas
-

214-6. from the Jurassic of north-west Scotland. Proceed-

BARKER, C. E. 1983. Influence of time on metamor- ings ojthe GeologicalAssociation of London, 91, 1 &
phism of sedimentary organic matter in liquid- 2, 99-105.
dominated geothermal systems, western North - - 1 Mesozoic sedimentation and sedimentary
9 8 3 .

America. Geology, 11, 384-88. rocks in the Inner Hebrides. Proceedings of the
BATHURST, R. G. C. 1975. Carbonate Sediments and Royal Society of Edinburgh, $3b, 47-63.
their Diagenesis (2nd edn). Elsevier, Amsterdam. - & HARRIS, J. P. 1979. Sedimentology of the Great
-

BURNER, R. A. 1968. Rate of concretion growth. Estuarine Group (Middle Jurassic) of North-West
Geochimica et Cosmochimica Acta, 32, 477-83. Scotland. In: Sedimentation Jurassique W. Euro-
-- 1980. Early Diagenesis. a Theoretical Approach. poem pp. 1-13. Association des Sedimentologistes
Princeton University Press. Francais, Publ. Speciale no. 1.
BRADSHAW, M. J. & FENTON, J. P. G. 1982. The LAND, L. S., BEI-IRENS,E. W. & FRISHMAN,S. A. 1979.
Bajocian 'Cornbrash' of Raasay, Inner Hebrides: The ooids of Baffin Bay, Texas. Journal oJ
palynology, facies analysis and a revised geologi- Sedimentary Petrology, 49, 1269-78.
cal map. Scottish Journal of Geology, 18, 131 45. LERCHE, I., YARZAB, R. F. & KENDALL, C. G. STC.
CURTIS, C. D. 1977. Sedimentary geochemistry: envi- 1984. Determination of paleoheat flux from vitri-
ronments and processes dominated by the involve- nite reflectance data. Bulletin of the American
ment of an aqueous phase. Philosophical Transac- Association oj Petroleum Geologists, 68, 1704-
tions of the Royal Society of London, Series A, 286, 17.
353-72. MARSHALL, J. O. 1981. Zoned calcites in Jurassic
DAGLEY, P. & MUSSETT,A. E. 1986. Palaeomagnetism ammonite chambers: trace elements, isotopes and
and radiometric dating of the British Tertiary neomorphic origin. Sedimentology, 28, 867-87.
Igneous Province: Muck and Eigg. Geophysical - 1982. Isotopic composition of displacive fibrous
-

Journal of the Royal Astronomical Society, 85, calcite veins: reversals in pore-water composition
221-42. trends during burial diagenesis. Journal of Sedi-
DICKIN, A. P. & JONES, N. W. 1983. Isotopic evidence mentary Petrology, 52, 615-30.
for the age and origin of pitchstones and felsites, MEADE, R. H. 1966. Factors influencing the early
Isle of Eigg, NW Scotland. Journal of the stages of the compaction of clays and sands--
Geological Society of London, 140, 691-700. review. Journal of Sedimentary Petrology, 36,
EL-SHAHAT, A. & WEST, I. M. 1983. Early and late 1085-101.
lithification of aragonitic bivalve beds in the MORTON, A. C. 1984. Stability of detrital heavy
Purbeck Formation (Upper Jurassic-Lower Cre- minerals in Tertiary sandstones from the North
taceous) of southern England. Sedimentary Geo- Sea Basin. Clay Minerals, 19, 287-308.
logy, 35, 15-41. PETTIJOHN, F. J., POTTER, P. E. & SILVER, R. 1972.
EMELEUS, C. H. 1983. Tertiary igneous activity. In: Sand and Sandstone. Springer-Verlag, New York.
CRAIG, G. Y. (ed.). Geology of Scotland, pp. 357- PRICE, L. C. 1983. Geologic time as a parameter in
97. Scottish Academic Press, Edinburgh. organic metamorphism, and vitrinite reflectance
GARRETS, R. M. & MACKENZIE,F. T. 1971. Evolution of as an absolute palaeo-geothermometer. Journal oj
the Sedimentary Rocks. Norton, New York. Petroleum Geology, 6, 5-38.
HARRIS, J. P. 1984. Environments of deposition of RICHTER, D. K. 1983. Calcareous ooids: a synopsis. In:
Middle Jurassic sandstones in the Great Estuarine PERYT, T. M. (ed.). Coated Grains. Springer-
Group, N.W. Scotland. Unpublished PhD Thesis. Verlag, Berlin.
University of Leicester. RIDING, J. B. 1984. The palynology of the Tobar Cearn
-& HUDSON, J. D. 1980. Lithostratigraphy of the
- Siltstone Member, Staffin Shale Formation (Ju-
Great Estuarine Group (Middle Jurassic), Inner /
rassic, Callovian/Oxfordian), Strathaird, southern
Hebrides. Scottish Journal of Geology, 16, 231-50. Skye. Reports of the British Geological Survey, 16
HUDSON, J. D. 1963. The recognition of salinity- (10), 1-5.
controlled mollusc assemblages in the Great RIEKE, H. H. & CHILINGARIAN,G. V. 1974. Compaction
Estuarine Series (Middle Jurassic) of the Inner of Argillaceous Sediments. Developments in Sedi-
Hebrides. Palaeontology, 6, 318-26. mentology, 16. Elsevier, Amsterdam.
276 J. D. Hudson & J. E. Andrews
SANDBERG, P. A. & HUDSON, J. D. 1983. Aragonite - - & 1974. Isotopic studies of the palaeoecology
relics in Jurassic calcite-replaced bivalves. Sedi- and diagenesis of the Great Estuarine Series
mentology, 30, 879-92. (Jurassic) of Scotland. Scottish Journal of Geology,
SHACKLETON,N. J. & KENNETT,J. P. 1975. Palaeotem- 10, 91-128.
perature history of the Cenozoic and the initiation & KEITrt, M. L. 1970. Jurassic (Callovian)
of Antarctic glaciation: Oxygen and carbon paleotemperatures from Scotland. Earth and Plan-
isotope analyses in the DSDP sites 277, 279 and etary Science Letters, 9, 421-6.
281. In: Initial Reports of the Deep Sea Drilling TAYLOR, H. P. JR & FORESTER, R. W. 1971. Low-O 18
Project, 29, 743-55. US Government Printing igneous rocks from the intrusive complexes of
Office, Washington. Skye, Mull and Ardnamurchan, western Scotland.
STONELEY, R. 1983. Fibrous calcite veins, overpres- Journal of Petrology, 12, 465-97.
sures, and primary oil migration. Bulletin of the WALKER, G. P. L. 1970. The distribution of amygdale
American Association of Petroleum Geologists, 67, minerals in Mull and Morvern (western Scotland).
1427-8. In: MURTY, T. V. V. E. R. K. & RAO S. S. (eds).
SYKES, R. M. 1975. The stratigraphy of the Callovian- West Commemorative Volume, pp. 181-94. Univer-
Oxfordian stages (Middle-Upper Jurassic) in sity of Sagar, India.
northern Scotland. Scottish Journal of Geology, 11, WAPLES, n. 1980. Time and temperature in petroleum
51-78. formation: application of Lopatin's method to
TAN, F. C. & HUDSON, J. D. 1971. Carbon and oxygen petroleum exploration. Bulletin of the American
isotopic relationships of dolomites and co-existing Association of Petroleum Geologists, 64, 916-26.
calcites, Great Estuarine Series (Jurassic), Scot- WATSON,J. V. 1985. Northern Scotland as an Atlantic-
land. Geochimica et Cosmochimica Acta, 37, 755- North Sea divide. Journal of the Geological Society
67. of London, 142, 221-43.

J. D. HUDSON, Department of Geology, University of Leicester, Leicester LE1 7RH, UK.

J. E. ANDREWS, School of Environmental Sciences, University of East Anglia, Norwich
NR4 7T J, UK.
 Oxygen-isotope studies of clastic diagenesis in the Lower
Cretaceous Viking Formation, Alberta: implications for the role
of meteoric water
Fred J. Longstaffe & Avner Ayalon

S U M M A R Y : The oxygen-isotope compositions and paragenetic sequence of diagenetic

minerals from the Lower Cretaceous Viking sandstone and conglomerate from south-central
Alberta, Canada have been used to identify changes in porewater composition during
diagenesis and to relate these changes to major geological events within the western Canada
sedimentary basin. Large-scale influx of meteoric water has played an important role in
diagenesis of this unit, especially following uplift of the basin in early Eocene time.
Diagenetic phases include early kaolinite (6180 SMOW = +24.9 to + 28.2 %o), siderite
(618OSMOW= +22.0 to +23.9%0; 613CPDB = -6.3 to -1.3%o), calcite (6180
SMOW = +26.0 %0; 6~3C PDB = -0.4 %0) and chlorite (6180SMOW = +10.7 -
+13.4%o), followed in order by dolomite (318OSMOW = +19.3 - +22.1%0; 613C
PDB = -5.9 to -2.6%o), calcite (618OSMOW = +13.9 - +15.9%o; 613CPDB =
-8.4 to -6.1%o), ankerite (6180 SMOW = + 16.3 - + 17.9 %0; 313C PDB = - 12.4 to
-3.5 %o), kaolin group minerals (3180 SMOW = + 12.0 - + 15.7 %o), and illite (6180
SMOW = + 14.2 - + 15.7 %0) and illite/smectite (6180 SMOW = + 13.8 - + 16.5 %0)-
Quartz overgrowths (6180 SMOW = +16.2 - +27.4%0) began crystallizing relatively
early during burial diagenesis and continued to form at least up to the onset of late diagenetic
formation of clay minerals.
The interpretation of these results is that shallow diagenesis, early in the burial history
(glauconite, pyrite, calcite, chlorite), occurred largely in the presence of seawater-derived
fluids, although a freshwater influence is indicated where siderite and(or) early kaolinite
cements are abundant. As compaction and burial diagenesis proceeded (diagenetic chlorite,
quartz overgrowths, dolomite), the porewaters became enriched in 180 due to water/rock
interaction. Burial diagenesis was terminated in the early Eocene by uplift related to the
major Laramide Orogeny. Recharge of the basin by low 180 meteoric water occurred at this
time. The meteoric water then became involved in the formation ofdiagenetic quartz, calcite
and ankerite (and the dissolution and albitization of feldspar) at or near maximum burial
temperatures, and in the crystallization at lower temperatures of kaolin group minerals, illite
and illite/smectite as the post-Eocene erosion progressed.

In this paper we report the results of an oxygen- ing their origin (e.g. sea water, meteoric water)
isotope study of diagenetic minerals from the and modification during diagenesis. Diagenetic
Lower Cretaceous Viking Formation over an minerals should obtain oxygen-isotope signa-
area of about 20000 km 2 in southwestern tures characteristic of the porewater and the
Alberta, Canada (Fig. 1). Similar studies of temperature at which the crystallization oc-
diagenetic minerals from Upper Cretaceous curred. The fi180 values of these minerals can
sandstones in the western Canada sedimentary then be used to reconstruct some aspects of
basin have shown that waters of meteoric origin porewater evolution throughout diagenesis, pro-
have played an important role in their diagenesis vided that the paragenetic sequence of diagene-
(Longstaffe 1983, 1984, 1986, Ayalon & Long- tic minerals can be determined, and that the
staffe, in press). The purpose of this study is to oxygen-isotope compositions of the minerals
determine whether meteoric water has been have remained substantially unchanged since
involved on a regional scale in the diagenesis of crystallization.
Lower Cretaceous sandstones from the Viking At temperatures typical of sedimentary envi-
Formation, and to deduce how the oxygen- ronments, most minerals do not experience
isotope composition of porewaters in these rocks significant oxygen-isotope exchange with water.
varied in response to depositional environment, Such exchange occurs only during mineral
burial, and the subsequent uplift and erosion of dissolution and precipitation, the potential for
the basin that resulted from the major Laramide which increases as temperature rises. Of the
Orogeny in early Eocene time. phases most commonly formed during diagene-
The oxygen-isotope compositions of formation sis, quartz is the least affected by such re-equili-
waters provide important information concern- bration (Yeh & Savin 1977). Exchange of oxygen

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 277

Geological Society Special Publication No. 36, pp. 277-296.
278 F. J. Longstaf/e & A. Ayalon
114 ~
I 1 I I 1 I I I [ I
29 9

N_
27 9
_ 28 9
25 t 2 3 26 9
40 - 2 0 % ~ 9 22
21 24 9
SASK. i MA,. ,,-~',, ! 16 9 9 18 9 19e
15 9 Z
14 9 --<
12 9 9 __.
13 11 9 ~
8

~ U . s . A .
""~. ~.~--~.~

O
79 9 10

6e
3:
I--
I.I.
,'7
52~

I--

30
30 2o
le
0 Miles 30

0 Km 45

CALGARY---.....~
25 I 1 i 1 1 JgP"~ t 1
5 28 24

RANGE

FIG. 1. Location of study area and sample locations. Numbers refer to samples listed in Tables 1 and 2.

between water and structural sites in most clay referred to these studies for a comprehensive
minerals also seems to be unimportant until listing of other important works concerning the
temperatures well in excess of 100~ have been Viking Formation, and for detailed discussions
attained (Yeh & Savin 1976, 1977, Eslinger & of the stratigraphy and sedimentology of the
Yeh 1981). Due to a greater ease of recrystalliza- unit.
tion, oxygen-isotope exchange can affect carbon- During the earlier part of late Albian time, the
ate minerals (calcite more so than dolomite) at interior of western North America was inundat-
lower temperatures than silicate minerals (Clay- ed by transgressive seas from the north and, to a
ton 1959); the possibility of isotopic exchange lesser extent, from the south (Williams & Stelck
between calcite cement and porewater during 1975, Weimer 1983), resulting in the deposition
subsequent stages of diagenesis should not be of marine shales (Joli Fou Formation). A
totally ignored. Within these constraints, how- regression towards the end of late Albian time
ever, the oxygen-isotope compositions of most marked the beginning of Viking deposition, the
diagenetic minerals should provide a valuable sediment being derived dominantly from the
record of the temperatures and water ~ 1 8 0 rising Cordillera to the west.
compositions prevailing during their formation. Hein et al. (1986) have characterized the
Viking as a coarsening upward sequence of
interbedded, fine-grained sandstone and shale,
Geological background fine- or medium-grained sandstone, and chert-
pebble conglomerate and pebbly sandstone.
The Lower Cretaceous Viking Formation (Slip- Within the study area, the Viking Formation can
per 1918) has been extensively studied because of be up to 60 m thick. Three main depositional
its importance as a hydrocarbon reservoir. Most events have been proposed by Hein et al. (1986):
recently, the sedimentology of the study area has (i) progradation of a shoreline-attached, clastic
been discussed by Grant (1985), Robb (1985), wedge during a regression, (ii) cut-and-fill of
Dean (1986) and Hein et al. (1986); the reader is channels that have dissected the clastic wedge,
 I s o t o p e s t u d i e s in the V i k i n g F o r m a t i o n 279
and (iii) marine reworking of the sediment into Standard Mean Ocean Water (SMOW) for
sheet-like deposits of sandstones and conglomer- oxygen (Craig 1961) and the belemnite Belemni-
ates on the offshore shelf. tella americana from the PeeDee Formation
During the later portion of the early Turonian, (PDB) for carbon (Craig 1957). To calculate the
a second major transgression occurred, ending oxygen-isotope results for calcite and dolomite,
Viking deposition; the overlying Lloydminster values of 1.01025 and 1.01110, respectively, were
and Colorado shales were deposited at this time. employed for the total carbonate CO2-phos-
The first phase of the Laramide Orogeny began phoric acid extracted CO2 fractionation-factor
in the late Cretaceous, causing thrusting and (a) at 25~ (modified after Sharma & Clayton
uplift along the eastern Cordillera (Taylor et al. 1965). The calcite ~ was used in all calculations
1964, Dickinson & Snyder 1978). To the east, for siderite using the ~ (25~ recently reported
continental deposition dominated through the by Rosenbaum & Sheppard (1986) for siderite
late Cretaceous and into the early Tertiary as the would lower the values reported here by about
sediment was shed from the Cordillera into the 1.36%o. The dolomite ~ was used in all calcula-
downwarped basin containing the Viking For- tions for ankerite; using the new a (25~
mation (Beaumont 1984). Maximum burial of reported by Rosenbaum & Sheppard (1986) for
the Viking Formation probably occurred in the ankerite would lower the values reported here by
late Palaeocene or early Eocene (Taylor et al. about 0.66 %o. These differences are insufficient
1964, Hitchon 1984). At this time the major pulse to affect the sense of the arguments and
of the Laramide Orogeny (early Eocene) resulted conclusions that follow. CO2 was initially ob-
in extensive deformation (overthrusting) in the tained from the siderite and ankerite by reaction
eastern Cordillera and significant uplift of the at 25~ for 21 days; these experiments were
sedimentary basin in Alberta; extensive erosion repeated on fresh samples at 50~ for periods of
of the accumulated Tertiary and Upper Creta- 16, 24, 48 and 72 hr. While the CO2 yields
ceous rocks ensued (Taylor et al. 1964, Beaumont increased at the higher temperatures, no signifi-
1981). cant differences in the oxygen (temperature
corrected) or carbon-isotope compositions of the
carbonates were noted, an observation in agree-
Analytical procedures ment with Becker & Clayton (1972).
An oxygen-isotope CO2-H20 fractionation
Thirty-eight core samples of the Viking Forma- factor of 1.0412 at 25~ has been used through-
tion have been examined (Fig. 1) using optical out this study. During the course of the experi-
microscopy, scanning electron microscopy, ener- ments a value for silica standard NBS-28 of
gy dispersive spectrometry, X-ray diffraction, + 9.62 + 0.12%o was obtained. Calcite standard
inductively coupled plasma spectrometry and NBS-20 gave an average 618OSMOW =
oxygen- and carbon-isotope geochemistry. All of +26.50+0.02%0 and 613CPDB = - 1.17
the analytical procedures, including the mineral _+0.01%o. Except where noted otherwise, the
separation techniques used to obtain subsamples analytical results reported are precise to better
for isotopic analyses of the diagenetic minerals than _+0.2 %o- Higher uncertainties arise when
(except quartz overgrowths), are discussed in the isotopic composition of a phase has been
Ayalon & Longstaffe (in press) and are not calculated from that of a binary mixture (e.g.
repeated here. chlorite and, in some samples, kaolinite; see
To separate the quartz overgrowths, samples Longstaffe 1986 and Ayalon & Longstaffe, in
were crushed using a mortar and pestle, and press).
disaggregated in distilled water using an ultra-
sonic probe at low power. The samples were then
dispersed in distilled water, and the <2 ~tm Results
fraction removed using standard sedimentation
techniques. The samples were dried and the 125- The petrology and diagenesis of the Viking
250 mesh size-fraction obtained by sieving. Formation have been discussed by many
Quartz was extracted from this fraction follow- workers, including Gammel (1955), Glaister
ing the method of Syers et al. (1968) and Jackson (1958), Thomas & Oliver (1979), Amajor (1980),
(1979). The quartz overgrowths were then isolat- Foscolos et al. (1982), Dean et al. (1984), Grant
ed from the detrital quartz following the method (1985), Robb (1985), Dean et al. (1985) and Dean
of Lee & Savin (1985), and examined for (1986). A detailed discussion of the diagensis of
authigenic morphology and purity using the several oil fields within the Viking Formation is
scanning electron microscope. given in Dean et al. (in preparation). Preliminary
The oxygen- and carbon-isotope data are stable isotope results have been reported by
presented in the usual f-notation relative to Longstaffe & Ayalon (1986).
280 F. J. Longstaffe & A. Ayalon
Three main rock types have been studied: 60%), chert fragments (5-20%), rock fragments
conglomerate, medium- and coarse-grained (mostly shale, 2-25%), and feldspar (mostly
sandstone, and fine-grained sandstone; their plagioclase, 1-6%). In some samples, dolomite
framework-grain petrography is summarized grains are present. The cements are comprised of
here largely after Grant (1985), Robb (1985) and carbonate minerals (siderite, dolomite and an-
Dean (1986). The conglomerates can be classi- kerite; calcite is rare), quartz and clay minerals.
fied as quartzarenites or sublitharenites (Folk Unlike the conglomerates and coarse-grained
1968), and are comprised of framework grains sandstones, diagenetic illitic clay minerals are
(75-85%), cement (5-15%) and porosity (3-15%). more abundant than kaolin group minerals
Framework grains consist of chert fragments (Table 1).
(45-80%), quartz (5-25%), rock fragments (most- Most of the detrital silica has 61so values
ly shale, trace-15%) and feldspar (mostly plagio- between + 14.0 and + 16.9%o (Table 2), charac-
clase, 1-12%). Ragged flakes of kaolinite (detri- teristic of a sedimentary origin. Higher values
tal?) are present in some samples. The dominant ( + 17.6 to + 20.0%0, Table 2) were obtained for a
cements are quartz and clay minerals (kaolin few samples of conglomerate and coarse-grained
group clay minerals > illitic clay minerals > sandstone, probably reflecting the higher propor-
smectite; Table 1). Minor siderite, dolomite and tions of chert fragments versus quartz grains in
ankerite cements are also present. these rocks.
The medium- and coarse-grained sandstones The 6180 values of the < 2 ~m clay fraction
are quartz arenites or sublitharenites, and are from two shale samples are + 13.5 and + 17.1%o
composed of detrital grains (70-80%), cement (8- (Table 2); values of + 17 to +20%o have been
25%) and porosity (3-13%). Framework grains reported for < 2 r t m clay fractions of Viking
consist of quartz (5-50%), chert fragments (20- shales, mudstones and argillaceous sandstones
65%), rock fragments (mostly shale, 0-14%) and located further to the east in Alberta (Longstaffe
feldspar (mostly plagioclase, 2-30%). Detrital (?) 1983). These values are similar to those obtained
kaolinite occurs in some samples. The main for other Cretaceous shales from the western
cements are quartz and clay minerals (kaolin Canada sedimentary basin (Longstaffe 1983,
group and illitic clays, minor smectite; Table 1). 1984, 1986, 1987), and are also typical of detrital
The fine-grained sandstones are sublitharen- clay mixtures from marine sediments (Savin &
ites, and are composed of framework grains (55- Epstein 1970a, b).
80%), cements (10-50%) and porosity (trace- Of particular interest is that the < 2 ~tm clay
10%). The framework grains are quartz (35- size-fractions from the Viking sandstones and

TABLE 1. Relative percentage of clay minerals in the < 2 #m size fraction

Sample 1 Depth (m) Rock type Kaolin group Chlorite Illite 2 Smectite

2 2977.6 Shale 32 5 63
4b 1870.6 Fsst 3 13 87
4c 1872.6 Fsst 5 10 85
5a 1367.8 CongP 47 18 35
5b 1373.4 Fsst 32 68
7 2719.1 Fsst 18 6 76
8a 2731.4 Csst 5 60 4 35 1
8b 2734.6 Congl 63 3 23 11
10 2174.5 Shale 18 10 68 4
11 1903.5 Msst 6 14 10 76
12 2386.9 Msst 5 30 65
13 2290.6 Fsst 21 10 69
14b 1987.4 Fsst 15 10 75
16b 2404.6 Fsst 5 5 90
18 1904.5 Csst 60 5 23 12
19a 1544.6 Csst 75 5 20
21 2257.8 Congl 91 9
22 2229.7 Congl 40 15 45
25 2294.3 Congl 31 5 60 4
26b 1923.3 Congl 47 8 30 15
28 1909.6 Fsst 30 8 62

1 See Fig. 1 for locations of samples. 2 Includes illite/smectite. 3 Fine-grained sandstone. * Conglomerate.
5 Coarse-grained sandstone. 6 Medium-grained sandstone.
 Isotope studies in the Viking Formation 28I

conglomerates, which are comprised mostly of stable isotope studies of the siderite are in
authigenic clay minerals and some silica, are also progress (Connolly, personal communication,
180-rich (+15.1 to +20.8%o; Table 2). This Staley, personal communication).
result contrasts markedly with the much lower
Pyrite: Pyrite is a common though volumetrically
61sO values ( + 9 to + 14%o) obtained for similar
unimportant early diagenetic phase in some fine-
materials from Upper Cretaceous sandstones in
grained sandstones. It occurs as framboid-like
Alberta; these low values have been interpreted
aggregates of small (0.5-1 lam) closely packed
to indicate the involvement of a sizeable fraction
crystallites (Fig. 2b) on silica grains, or as
of low-~80 meteoric water in the crystallization
individual, larger (4 ~tm) euhedra associated with
of the diagenetic clays (Longstaffe 1983, 1984,
the framboidal pyrite.
1986, 1987, Ayalon & Longstaffe, in press).
Chlorite: Authigenic, Fe-rich chlorite is volu-
Diagenetic minerals metrically minor in the samples studied (Table
1). It occurs as small rosettes or fan-shaped
The Viking Formation has experienced a variety clusters of pseudohexagonal crystals that have
of diagenetic processes, including compaction, developed directly upon detrital grains, begin-
cementation, grain alteration and dissolution. ning early in the diagenetic history of the rock.
Most pores have been at least partially filled by Chlorite from < 0.2 and (or) 0.2-0.5 ~tm mixtures
authigenic minerals. Diagenetic minerals that with illite/smectite was calculated to have ~180
have been observed in this study include glau- values of +10.7 to +13.4%o (Table 2) (for
conite, iron oxide, siderite, pyrite, Fe-chlorite, procedures, see Longstaffe 1986 and Ayalon &
quartz, calcite, albite, dolomite, ankerite, kaolin Longstaffe, in press); these values are precise
group minerals (kaolinite, dickite), illite, illite/ only to _+1%o.
smectite and smectite.
Calcite: Calcite is very uncommon in these rocks.
Iron oxide." Trace amounts of iron oxide occur as It occurs both as a comparatively early cement
very thin coatings on some detrital grains. These between relatively uncompacted grains, coated
'dust rims' characteristically mark boundaries by clay cements (Fig. 2c), and as a later phase,
between detrital silica and authigenic quartz where it overlies earlier cements such as siderite
overgrowths and were probably inherited during and quartz, but is itself overlain by authigenic
sedimentation. kaolin group minerals and illitic clay. The few
samples of the later calcite have lower 6~80
Glauconite: Glauconite is normally present in
only minor quantities as rounded or oval-shaped SMOW (+13.9 to +15.9%o, Table 2) and
613C PDB ( - 8.4 to - 6.1%o, Table 2) values than
grains (200-400~tm), and as matrix material
the one sample of early calcite (~180 SMOW =
compacted between rigid detrital grains. It is
+26.0%0, 613C PDB = -0.4%0; Table 2, no. 13)
most likely of very early diagenetic origin. The
that was analysed. Further isotopic studies of the
glauconite occurs generally in offshore rather
than nearshore portions of the study area. calcite are in progress (Connolly, personal
communication, Staley, personal communi-
Siderite: Siderite is most abundant in the finer- cation).
grained rocks, comprising up to 60~ of some
fine-grained sandstones; however it is patchy in Quartz: Silica is the most abundant authigenic
distribution (Dean 1986). Where such relation- cement in the Viking Formation; some pores
ships can be observed, most siderite appears to have been completely filled by diagenetic quartz
have formed rather early in diagenesis, prior to crystals, which can vary in length from 2 to
most other diagenetic cements (Fig. 2a). In some 60 p.m. Quartz overgrowth formation appears to
samples, siderite has filled all pore space, have occurred throughout much of the diagenetic
effectively preventing subsequent formation of history of the rock. Some early kaolinite is
other diagnetic phases. Framework grains in engulfed by quartz overgrowths. An authigenic
such samples are normally less compacted than origin for much of this clay is suggested by its
in other areas (Dean 1986). well crystallized morphology (Fig. 2d, e). In other
The 6~sO SMOW and (~13C PDB values of the samples, quartz overgrowths are post-dated by
siderite range from + 22.0 to + 23.9%o, and - 6.3 diagenetic kaolin group minerals (Fig. 2I") and
to -1.3%o, respectively (Table 2). Even lower illite or illite/smectite (Fig. 2g).
618OSMOW values (+18.1%o; 613C PDB = A wide variation in ~180 was obtained for the
-4.6%0) have been reported by Dean (1986) for diagenetic quartz (+ 16.2 to +27.4%0, Table 2).
siderite from the southwesternmost portion of The fine- and medium-grained sandstones have
the study area, where a rooted sedimentary facies ~5180 values in the lower end of this range (+ 16.2
has been identified (Hein et al. 1986). Detailed to + 20.8%0); the coarse-grained sandstones and
282 F. J. Longstaffe & A. Ayalon

I I I

I I I I

tt~ ~ tt~ r
I I I I

~6vz ~5

~ eq
r r

r< v- ,6vz

~5 ~ ~ eq
r r ~ r

9 ~

v- cq ~.
,6 r--

a.a,~ a,~ ~,~a,~a, ~ t ~ ~ a,a.a,u,

eq

~A
Isotope studies in the Viking Formation 283

e~
9 ~

~ ~ ~~oo ~~~ooo o
284 F. J. Longstaffe & A. Ayalon

FXG. 2. (a) Diagenetic siderite (Sid), calcite (Cc), kaolinite (K) and illite/smectite (1). Scale = 20 p,m. (b) Early
diagenetic pyrite (Py) on detrital quartz. Scale = 2 pm. (c) Diagenetic calcite (Cc) filling pore and coated by
illite and (or) illite/smectite (1). Scale = 10 ~tm. (d, e) Kaolin group minerals (K) engulfed by authigenic
quartz (AQ). Scale = 20 pm. (f) Authigenic kaolin group minerals (K) filling pores; quartz overgrowths (AQ)

conglomerates tend to have higher 3180 values Dolomite/ankerite: Late diagenetic carbonate
(+17.7 to +27.4%0). This difference may indi- cement in the Viking Formation is comprised
cate that some grain-size dependent condition mostly of ankerite and (or) dolomite. Ankerite
affected silica diagenesis. That selective conta- cement coats authigenic quartz, replaces detrital
mination by detrital quartz or chert fragments is feldspar and dolomite, and fills pores. In turn, it
wholly responsible for this shift in values is less is coated by authigenic clay minerals, typically
likely, but cannot be ruled out entirely. illite or illite/smectite (Fig. 2h). In some samples,
the Ca-rich nature of the ankerite suggests that it
Feldspar: Diagenetic reactions involving feld- has directly replaced calcite (Dean 1986). In
spar (mostly plagioclase) include alteration, others, earlier calcite cement is replaced by
dissolution and the formation of albite over- diagenetic dolomite, which can be rimmed or
growths. Alteration includes the partial or partly replaced by ankerite. Further studies are
complete replacement of feldspar by carbonate in progress (Connolly, personal communication).
and (or) kaolinite; Dean (1986) has also noted The ankerite has 6180 SMOW and 613C PDB
albitization of plagioclase. Dissolution of plagio- values of +16.3 to +17.9%o and - 1 2 . 4 to
clase grains has also been observed, together -3.5~o, respectively (Table 2). The diagenetic
with, in some cases, the appearance of authigenic dolomite is slightly richer in both ~so and 13C
albite overgrowths on the partially dissolved (61sOSMOW = +19.3 to +22.1%o; 613C
grain. The relative timing of plagioclase dissolu- PDB = - 5 . 9 to -2.6%0, Table 2). Unaltered
tion and authigenic albite crystallization is detrital dolomite from the Viking Formation and
unclear; the two processes may be related (Boles overlying Upper Cretaceous units such as the
1982, Land 1984). Milk River sandstone has generally higher 61 sO
 Isotope studies in the Viking Formation 28 5

line the pore walls. Scale = 40 ram. (g) Authigenic illite/smectite and (or) illite (1) coating quartz overgrowths
(AQ). Scale = 4 ~tm. (h) Authigenic ankerite (Ank) filling pore and coated by illite and (or) illite/smectite (1).
Scale = 10 I.tm.

SMOW values ( + 26 to + 28%0; Longstaffe 1983, 'ragged' flakes) has 6180 values of +24.9 to
1984, unpublished data). Some contamination of + 28.2%0 (+0.5 where determined from quartz-
the diagenetic dolomite samples by detrital kaolinite mixtures; Ayalon & Longstaffe, in
dolomite may have occurred. press). The later diagenetic material has much
lower values (+12.0 to 15.7%o; +0.5). Some
Kaolin group minerals: Kaolin group minerals
cross-contamination of samples between early
(kaolinite and dickite) are particularly abundant
and late kaolinite may have occurred.
as late diagenetic pore-fillers in the coarse-
grained sandstones and conglomerates (Table 1), Illite, illite/smectite and smectite" The 10/k clay
where they occur as loosely packed booklets of minerals are common diagenetic phases in the
pseudohexagonal flakes (10-25 lam in diameter, Viking Formation, especially in the fine- and
Fig. 2f). In addition, as noted previously, some medium-grained sandstones (Table 1). They
well crystallized, earlier formed kaolinite has occur as mixtures of illite (> 90~ non-expanda-
been engulfed by quartz overgrowths (Fig. 2d, e). ble layers) and illite/smectite (65-80~ non-
In other samples, more poorly crystallized (sub- expandable layers). Their most common modes
hedral) flakes of kaolinite are associated with of occurrence are as (i) webby to mat-like
diagenetic quartz. These flakes may be detrital; aggregations that coat grains and earlier cements
alternatively, authigenic kaolinite can attain (Fig. 2c), and (ii) filamentous material with
such a morphology due to dissolution during later curled projections that is attached to grains and
stages of diagenesis (Hurst 1980). earlier cements, and extends into pore space
Two populations of6180 values were obtained (Fig. 2g). The illitic clay minerals appear to be
for the kaolinite group minerals (Table 2). The among the last major authigenic cements to
early kaolinite (mostly euhedral, but also some form. They coat most previous cements, includ-
286 F. J. Longstaffe & A. Ayalon
ing some of the late diagenetic, pore-filling bracket precisely from the petrographic record
kaolin group minerals (see, e.g. Dean 1986). alone. However, most literature suggests that
However, the presence of some earlier diagenetic large-scale dissolution and albitization of feld-
illitic clays cannot be ruled out from petro- spar begins at relatively high temperatures (e.g.
graphic data; such material can be very difficult 110-120~ Boles 1982). The interval suggested
to recognize, especially if it has become matted in Fig. 3 is compatible with such observations
on grain surfaces. The illite/smectite and illite (see below).
obtained from <0.2 or 0.2-0.5 ~m size-fractions
have 6180 values of +13.8 to +16.5%o and
+ 14.2 to + 15.7%o, respectively (Table 2).
Discussion
Smectite is relatively uncommon, occurring
The petrographic and oxygen-isotope results for
mostly in the conglomerates or the coarse-
the diagenetic minerals can be combined to
grained sandstones as a late-stage pore-lining. It
deduce changes in porewater 3180 during depo-
has the characteristic honeycomb-like mode of
sition, burial, uplift and erosion of the Viking
aggregation (Wilson & Pittman 1977).
sandstones and conglomerates. The relationship
Paragenetic sequence-summary: It is difficult to between the measured 6180 of each diagenetic
define the order of diagenetic mineral formation mineral and the possible combinations of water
for an individual sample uniquely. However, the 6180 and temperature are shown in Fig. 4(a, b).
petrographic evidence from the core samples Our goal is to fit a porewater evolution curve to
suggests that the paragenetic sequence proposed this grid in a manner that intersects the range of
in Fig. 3 is generally applicable throughout the possible temperatures and water 3180 values in
study area. Very early (synsedimentary) forma- an order consistent both with the observed
tion of glauconite was followed during shallow paragenetic sequence (Fig. 3) and with geologi-
burial by the crystallization of pyrite and (or) cal events in the Viking Formation.
siderite. Chlorite diagenesis also appears to have Some uncertainties in this approach should be
begun relatively early; its timing relative to acknowledged. The reliability of the limiting
siderite or pyrite crystallization is as yet unclear. curves varies with our knowledge of the oxygen-
Some calcite and kaolinite cement also formed isotope equilibrium fractionation factor (~) for
early in diagenesis. Quartz crystallization oc- mineral-water pairs at low temperatures. For
curred throughout much of diagenesis, followed minerals such as calcite it is well understood;
by ankerite, authigenic kaolin group minerals much less is known for phases like Fe-chlorite.
and diagenetic illite, illite/smectite and (or) As discussed earlier, the potential for isotopic re-
smectite. At least some diagenetic illitic clay equilibration of phases like calcite (i.e. dissolu-
postdates other authigenic clay minerals. tion-reprecipitation), and the efficacy of mineral
The timing of feldspar dissolution and the separation, could also affect the position of these
formation of albite overgrowths is difficult to curves.

Diagenetic History Viking

Glauconite, Iron-Oxide BL

Siderite, Pyrite

Chlorite

Calcite

Quartz Overgrowths

Feldspar Solution & A m B - -

Overgrowths

Dolomite, Ankerite __ML.. _

Dolomite Ankerite
Kaolin Group ~ k , _ m m m

IIlite, I l l i t e / S m e c t i t e , '~__?_ _ ? __ .,~,. __

Smectite

Time ~,

FIG. 3. General paragenesis of diagenetic minerals in the Viking Formation.

 Isotope studies in the Viking Formation 287

+6

+4

~ "~ oo,~
o

~ -2

o~ -6 / / / ~ ~ ~ S Viking
/ / a~.o"~/~'J Maximum 6'80 -
-8 ,:,..ene.c,.'.nera,s _

/ / / , , , , , , , , , !
-10
0 20 40 60 80 100 120 140
(o) Temperature ~C

0 -2

O -4
r

-6 / ~ /.~,,"//jr. ~o."" Viking -

I ~ . , , ' ~ / . . , ~ . ' - " _. . . . ~,.. -
/ ~ ~ ' ~ " Minimum 0 '~O __
-8 / ~ / / / ~ ~ , ~ " "" Diagenetic Min e rals-
-10 ~ 1 ! / I" f J I I I I 1 I
0 20 40 60 80 100 120 140

(b) Temperature oC
FIG. 4. ~5~sOof porewater versus temperature for diagenetic minerals from the Viking Formation. The curves
for the minerals are shown for the maximum and minimum 6~sO values of each phase using the following
equations (T = degrees K): (1) A~sO chlorite-H20 = 1.56 (106) T -2 4.7 (Wenner & Taylor 1971). (2) 103 In
- -

illite/smectite-H20 = [(2.43 + 0.24E)](106) T -2 - 4.82 (Eslinger & Savin 1973, Yeh & Savin 1977; E = ~o
expandable layers in illite/smectite). (3) 103 ln~ calcite-H:O = 2.78(106)T -2 - 2.89 (Friedman & O'Neil 1977
after O'Neil et al. 1969); also used for siderite; no correction for Fe-content was made. (4) 103 ln~ kaolinite-
H20 = 2.5(10 ~ T -2 - 2.87 (Eslinger 1971, Land & Dutton 1978). (5) 103 ln~ quartz-H20 = 3.38(106) T -2 -
2.90 (Friedman & O'Neil 1977, after Clayton et al. 1972). (6) 103 ln~ dolomite-H20 = 3.14(106) T -2 - 2.0
(Land 1983). (7) 103 ln~ ankerite-H:O = 2.78(10 ~ T -2 -t- 0.32 (Dutton & Land 1985).

Present c o n d i t i o n s in the V i k i n g F o r m a t i o n F r i e d m a n 1969, S c h w a r t z et al. 1981 a n d

p r o v i d e an e n d - p o i n t for the p o r e w a t e r e v o l u t i o n u n p u b l i s h e d data). T h e p r e s e n t d e p t h of the
curve. A v a i l a b l e 6 1 s o values for f o r m a t i o n V i k i n g s a n d s t o n e s a n d c o n g l o m e r a t e s analysed
waters range f r o m - 7 to - 4 ( H i t c h o n & in this study ranges from a b o u t 1300 m in the east
288 F. J. L o n g s t a f f e & A . A y a l o n

to 2700 m in the west. Assuming an average From his data, Hacquebard (1977) calculated
surface temperature of 5~ and an average that the maximum sediment thickness eroded
geothermal gradient of about 24~ km -1 (Hit- since the Tertiary ranged from 3000 m in WSW
chon 1984), formation temperatures of 40-70~ Alberta to 800 m in the ENE. Similar trends,
can be calculated. These estimates are similar to though generally lower values (1900-900 m) were
corrected bottom-hole temperatures available for reported by Nurowski (1984) from his studies of
petroleum fields in the Viking Formation. the equilibrium moisture content of Upper
Estimation of the maximum temperatures Cretaceous coals. Using a crustal flexture model
reached during burial is more problematical. The together with Hacquebard's (1977) data, Beau-
palaeogeothermal gradient and (or) the thickness mont et al. (1985) suggested that between 1600
of sediment removed from the basin since and 1900 m of overburden were removed from
maximal burial in the early Eocene has been the study area.
examined by many workers, including Magara From the available information on palaeo-
(1976), Hacquebard (1977), Beaumont (1981), geothermal gradients, maximum burial depths,
Majorowicz & Jessop (1981), Hitchon (1984), and estimated average early Eocene surface
Nurkowski (1984), Beaumont et al. (1985), temperatures (+16~ Piel 1971), a range of
Majorowicz et al. (1985) and Longstaffe (1986). maximum burial temperatures between 100 and
According to Hitchon (1984), major changes in 150~ can be calculated for the study area. The
the geothermal gradient accompanied the second optimum estimate lies in the range of 110~
stage of the Laramide Orogeny; the geothermal 130~ with the temperature maximum decreas-
gradient across Alberta ranged from ~30~ ing to the ENE. This estimate is compatible with
km -1 in the west to ,~23~ km -~ in the east that indicated by the smectite to illite transfor-
during the late Palaeocene. By the Eocene, the mation in Viking shales from the study area
pattern was reversed, increasing from ~21~ (Dean 1986).
km -1 in the west to ~27~ km -~ in the east, a
subdued version of the present situation. In
Hitchon's (1984) model, the major (i.e. second
Early (shallow) diagenesis (stage I)
pulse) Laramide Orogeny produced a topo- The porewater evolution curve proposed to
graphic high (maximum relief) in western explain the paragenetic sequence of diagenetic
Alberta. minerals, and their 6180 values, is shown in Fig.
The resulting high potentiometric surfaces led to 5, together with appropriate portions of the
regional recharge of cold meteoric water in the curves from Fig. 4. Given the shallow marine
WSW, flowing to the ENE in the basin. This setting for most of the Viking sediments, the
gravity-induced flow of cold meteoric water most probable starting composition for
reversed the geothermal gradient pattern estab- porewater during low-temperature formation
lished by dominantly compactional flow prior to (~20~ of early diagenetic phases (chlorite,
the early Eocene. Fluid flow, more so than the calcite B; Fig. 5a) was sea water (6180 = 0%0).
thickness (or rate of erosion) of the Phanerozoic The water may have been slightly more l SO-poor
section, became the dominant control upon the (e.g. 6180 = -1%o), given possible early
geothermal gradient and heat flow, at least for contributions by meteoric water (see below),
Cretaceous and Tertiary sandstones and conglo- potential variations in the primary 6~80 values
merates of the basin. of entrapped bottom waters (Savin & Yeh 1981)
Hitchon's (1984) conclusions are based partly or, less likely, the effects of interaction and
upon calculations of palaeogeothermal gradients exchange with underlying material (Lawrence et
made by Hacquebard (1977), using vitrinine al. 1975). The r value of the early calcite
reflectance (coal rank) and the equilibrium ( - 0.4%0) is typical of marine carbonates derived
moisture contents of coals. More recently, Beau- from CO2 of inorganic origin (Land 1980).
mont et al. (1985) studied organic reactions in The 6180 values of the early kaolinite (kaolin-
shales through Alberta; they also concluded that ite B, Fig. 5a) are compatible with crystallization
since the early stages of the major Laramide from sea water (6180 = 0 + 2%0 at 20~
Orogeny, the palaeogeothermal gradient in- However, geochemically it is much more likely
creased systematically away from the Foothills, that the kaolinite formed in the presence of a
but noted that the gradient may have been more significant fraction of fresh water (e.g. Bjorlykke
similar to that of the present day (27~ km-~ in et al. 1979, Hurst & Irwin 1982); waters with
the Foothills to 40~ km- 1in the Edmonton area). 8180 values of - 5 to -2%0 are indicated at 5~
In general, however, no reversal in the heat-flow (Fig. 4)" by 15~ water with 6180 = - 3 to 0%0 is
pattern has occurred since the early Eocene required.
(Hitchon 1984, Majorowicz et al. 1985); the It is even more difficult to envisage the
hydrodynamic regime in Alberta has remained formation of the early diagenetic siderite from
fundamentally unchanged since that time. sea water alone (i.e. 5180 = 0%0). First,
 Isotope studies in the Viking Formation 289

crystallization temperatures in excess of 50~ entered (stage II, Fig. 5b), in keeping with the
would be required (still higher temperatures paragenetic sequence (Fig. 3).
result, following Rosenbaum & Sheppard 1986). The extent to which the porewater became
Second, formation of early diagenetic siderite is enriched in 180 during compaction and burial is
geochemically much simpler in brackish to difficult to determine. Involvement of very 180-
freshwater environments (Berner 1971, Gautier rich waters derived from Palaeozoic carbonates
1982, Postma 1982). seems most unlikely; instead, the maximum
The isotopic data support the involvement of 6180 of the porewater probably did not exceed
meteoric water at low temperatures in the + 6%0, because of the buffering of the water
formation of the siderite. First, its relatively high composition by clay mineral reactions (e.g.
613C values (average - 3%0;Table 2) suggest that Suchecki & Land 1983). Lower values, like those
isotopically very light CO2 normally associated shown in Fig. 5(b) (+2 + 1%o) are even more
with thermal decarboxylation of organic matter likely. First, the field for authigenic dolomite
(e.g. Irwin et al. 1977) was unimportant in cannot be intersected at maximum burial tem-
siderite crystallization. Instead, the 613C values peratures by water compositions much greater
are typical of many early, shallow diagenetic than +3 + 1%o; second, maximum 6180 values
siderites, and of CO2 of dominantly inorganic for water of ,~ + 3%0 can be calculated from a
origin. Some contribution by 13C-rich CO2 combination of fluid inclusion temperatures and
derived by bacterial fermentation at relatively the oxygen-isotope compositions of associated
low temperatures (<50-60~ cannot be ruled diagenetic quartz from other Lower Cretaceous
out entirely (Dean 1986). Second, from the sandstones and conglomerates in Alberta which
oxygen-isotope data, siderite crystallization at have reached much higher temperatures
temperatures between 20 and 40~ would require (~190~ during burial (Tilley & Longstaffe
water with 6180 of - 7 to -2%0 (Fig. 4). Even 1986).
lower 6180 values for the water would result
following Rosenbaum & Sheppard (1986); only Maximum burial and uplift (stage III)
slightly higher values are anticipated from an
experimentally determined equilibrium oxygen- While the oxygen-isotope requirements for dolo-
isotope, siderite-water relationship. The even mite can be satisfied at or near temperatures
lower 6180 for siderite reported by Dean (1986) associated with maximum burial, the range in
from a rooted zone (i.e. barrier island, delta, 6180 of the diagenetic quartz and subsequent
shoreline) in the southwestern portion of the carbonate cements (calcite, ankerite) cannot be
study area further indicates that fresh water has explained unless a significant lowering in
been important in siderite genesis. porewater 6180 occurred at, or shortly following,
A possible implication of these conclusions is maximum burial of the unit (Fig. 5b). This
that despite the shallow marine setting proposed change could only be accomplished by a perva-
by Hein et al. (1986) for the Viking sediments, sive influx of 1ow-180 meteoric water (6180
fresh water was introduced into at least some -15%0, see Taylor 1974). Such an occurrence is
sedimentary facies at certain localities, early in entirely consistent with the geological develop-
diagenesis. This influx may have been achieved ment of the basin; maximum burial coincided
through hydraulic connection to freshwater with maximum relief which, in turn, led to a
aquifers or by recharge of locally (and temporar- major, regional influx of fresh water into the
ily ?) emergent zones; either process could help to Cretaceous sandstones and conglomerates (Hit-
account for the patchiness of Siderite occurrence chon 1984). The exact slope of this portion of the
in the Viking sandstones. porewater evolution curve (Fig. 5b, stage III) is
idealized; a net cooling effect arising from uplift
and the recharge of cold meteoric water can be
expected.
Burial diagenesis (stage II) The oxygen-isotope results for diagenetic
With increasing depth of burial, the 6180 of the quartz from the coarse-grained sandstones and
dominantly marine porewater can be expected to conglomerates overlap much of this part of the
increase due to dissolution of rock fragments, the porewater evolution curve, as well as the earlier
possibility of isotopic exchange with phases like stage involving compactional waters (stage II,
calcite and (or) contributions from the dewater- Fig. 5b). In contrast, diagenetic quartz from the
ing of clay minerals (Clayton et al. 1966, Hitchon fine- and medium-grained sandstones has 6180
& Friedman 1969, Suchecki & Land 1983). values confined almost entirely to this low-lSO
Following such a porewater evolution path with segment of the curve. The reason for this
increasing temperature, the fields for early observation is not entirely clear; one explanation
diagenetic calcite and chlorite are left and those is that some of the finer grained rocks had better
for authigenic quartz and then dolomite are hydraulic connections to adjacent units (as
290 F. J. Longstaffe & A. Ayalon

I I I I 1 I 1 I I I [ I I

+6 _

- _~,,~ ~,~-~'~ ~:~,~'~'~

. ~q,.6. ~ r .,,.e* L~,~ .1"~
+4

+2
0
=E
u) 0
o
o
o -2 ........ / .:/ I-- +-- ~....M
o
s
-4
0
@o

oO
-6 - ,~ @ " i Viking
x~'~,~,q"~ Stage I
-8 -

-100-
20 40 60 80 100 120 140
(o) Temperature ~C

I I I I [ l 1 I 1 1 I ~ I

+6
~ - _ O Oat.t ,

+4 z ,:\
+2-
O
0-
o _.---,.~- / oo~,* oo,O ' ~ . ~ ~ /
,=,
-2 --
o
s / / / ~--"" i ~, ~ ~ ~,~,~ "k~'~l'V
O -4 --
@o /

1
oO / /
-6 --
Viking
-8 Stage II & III
L

I I J I L I J t [ L L J I
-10 140
0 20 40 60 80 100 120
Cb) Temperature ~C

FIG. 5.

proposed earlier for siderite formation). Another tic dolomite, which is interpreted to have formed
explanation is that dissolution of chert fragments mostly from evolved, compaction-derived
released abundant silica for diagenesis only in porewater at maximum burial, has 613C values
the coarser grained rocks. Quartz cementation in ( - 5.9 to - 2.6%o; average = - 4.1%o; Table 2)
the finer grained sandstones may have required very similar to the early diagenetic siderite. Such
externally derived silica, carried into the unit by values are more weighted towards inorganic (e.g.
meteoric water. The Fe required for ankeritiza- pre-existing calcite) than organic carbon reser-
tion of calcite could have been transported in the voirs. Low-~3C CO2, such as can be derived by
same manner. Still other scenarios are possible; thermal decarboxylation of organic matter, ap-
no unequivocal explanation is available at pears to have been relatively unimportant. In
present. contrast, the few samples of the later diagenetic
The carbon-isotope values of the carbonate calcite and ankerite, which apparently formed at
cements show some interesting trends. Diagene- high temperatures in the presence of a sizeable
 Isotope studies in the Viking Formation 29I

I I I [ I I [ I I I ! I [ -'i
+6

+4 nl
A

+2
0
=E
r o
m

o"" -2
O r " -, ......v. '..v %".: 3"~
s ~ ~ . , ~ .~.e~,,,' e " - _
O -4-
~0
~ ~~ r~N. o r~ .~
~ ~ x, ~ i ~ ~:*~. . . . , ~. .x~ ' _
k.~
-6 -
_ Yo~'~H ~ V ' , ~ Viking -
-8 -
q,~-="-- /-Present Conditions St e IV & V
,~,.
_

-10
0 20 40 60 80 1 oo 120 140
(c) Temperature ~C

FIG. 5. Idealized porewater evolution path (6180 of water versus temperature) for the Viking sandstones and
conglomerates. The curve has been deduced from the paragenetic sequence (Fig. 3) and the oxygen-isotope
compositions (Fig. 4a,b) of the diagenetic minerals. (a) Stage l, early (shallow) diagenesis; (b) stage II, burial
diagenesis and stage III, maximum burial and uplift; (c) stage IV, erosion and stage V, present day.

fraction of meteoric water, have significantly illite/smectite (stage V, Fig. 5c). The occurrence
lower 61sC values ( - 12.4 to -3.5%0; average = of illitic clays as the last major diagenetic phase
- 7.1%o) which demand involvement of organi- (Fig. 3) makes it at least possible that their
cally derived CO2. Meteoric water may have formation is continuing at present. Porewaters
helped to deliver CO2 (as well as Si, Fe?) which with suitable K§ +, for example, could have
was released into shale porewaters by the evolved from the originally fresh water as
maturation of organic matter, and expelled into water/rock interaction progressed (e.g. leldspar
the coarser grained units. Alternatively, the dissolution), and the ionic strength of the
surface-derived meteoric water may have origin- porewater increased. Compositional variations
ally dissolved significant quantities of low-lSC (e.g. ~ illite) in the illite/smectite are controlled
soil CO2. by water chemistry rather than temperature in
the Viking sandstones and conglomerates (Dean
et al., in preparation). However, the very fine-
Erosion (stage IV)
grained nature of the illitic clays, together with
The crystallization of the late stage, authigenic the present uncertainty over the exact mecha-
kaolin group minerals (kaolinite A, Fig. 5C) and nism by which the smectite to illite transforma-
illitic minerals largely post-dates the formation tion occurs, leaves open the possibility that
of diagenetic quartz and associated carbonate oxygen isotopic re-equilibration with present
cements (Fig. 3). To accommodate these obser- formation water has occurred. At least some of
vations and the 6180 values of the authigenic the diagenetic illitic clay may have formed
clay minerals requires that the porewater sub- earlier, at higher temperatures from more 180-
stantially declines in temperature. The most rich porewater. Radiogenic isotope studies
logical trajectory is towards the present range of (Rb/Sr; K/Ar) to resolve the timing of illitic clay
temperature and porewater 61sO (stage V, Fig. formation are in progress.
5c). Such cooling can be expected as a result of
the substantial erosion that followed the early
Eocene uplift of the Viking Formation. Concluding remarks
The porewater evolution curve ends within the
field known for present conditions in the study One objective of this paper is to demonstrate that
area; this range overlaps with much of the the oxygen-isotope compositions of diagenetic
isotopic data for the authigenic illite and minerals can provide a valuable record of the
292 F. J. Longstaffe & A. Ayalon
temperatures and porewater 6~80 compositions directly to changing geological conditions during
prevailing during their formation. Figure 6 the evolution of the basin. Our conclusions are
summarizes the model deduced for diagenetic summarized in Fig. 7.
minerals from the Viking Formation. Shallow diagenesis, early in the burial history
We also wished to determine whether meteoric (stage 1; diagenetic glauconite, pyrite, calcite,
water had been involved on a regional scale in chlorite) occurred mostly in the presence of
the diagenesis of the Viking sandstones and seawater-derived pore fluids, although a fresh-
conglomerates and, more specifically, to show water influence is indicated where siderite and
whether porewater compositions could be related early kaolinite cement are abundant. As compac-

f I I I r I I I I I 1 I I ]
+6 In

+4 ~-Quart~. ~
.~------ll ~._J
A --~-~ .~-.~-.. ..+.-----...:..:.-'~~.+..
. ~
I
-I-2 ~ + ~-~ ~arC~ :/~:::/://: ~:: v: ~Oo!om!te:~ ~ /!
O ~.~,~ . +./...:.-..../././..:/.:-....:::../..
~i.U......./...;/./..:.:.:-.:~:~ ~.~.:.-.'../..::..:/~ / /
=E ~.~dhlorite~.". i i~/.i-/.:/.-i. ;/ : . i ~ . ~ . ~ .~ ~-" o. .t.--~~,(~e r / -1
r 0 ~,,'~:...'.'-'-.-....;.?~.........'..'...'..'.-'.- v~ .... .~.,~ .~ /?'...../..'.'.

oo
.,,O~,~.,,e::.../..~::..~..~~=" 7 /
r, ~,..+. -..~" " /
.<.o~..::./..?:.:.:/ / 7
~ ~ . , - - ' . .~. . - - . " .....~. .: ~ r ....- ~\~.r
-../ / / /- 1
,o
-2 _ . , ~.~....:......:"...:.-./..!./.-/.~,~......i:.:~" / - I
o / ~.7"~..;'.... .' .~'"
__ .-".' .; ~." '
.~.~........:...:.:..:...::........~=,o~',n,,~..-.:::..~..~,"~~
"~ " . ' - ; " . .'.'.' . .'.:.' . ' . " . . ..".' .'.'~" . " . ~
~v
r l

s ,~i" ~-.: . ' '" "."."..".-. '.".'."-".'.".';;"..-.".".".:/."-~ ' ~

_

o -4 -
9 _~:~.i~'~ ~??2iS~'~:.....:.::....:.:.".i....~
OO
.,p.

-6 +:/ ~ v
Viking
-8 Summary

L L I I I I I I I I I I I
-10
0 20 40 60 80 100 120 140

Temperature oC
FIG. 6. Idealized porewater evolution path ((~180 of water versus temperature) for the Viking sandstones and
conglomerates: summary.

I I I I I 1 I I 1 I I
+4 ~ 110 ~ - 1 3 0 ~ C
lli-Early Eocene, Major Laramide
+2 If-Burial ~\~o~ OrogenY,and
ReliefMaximum Burial
A
l
O o Marine

20+ ~ C ? IV-Post Eocene Erosion

-2
0=
o"
O -4
,I- tBrackish V-Present -
O -6 20+~ C Conditions
OO
"11'=
ao 40 ~ -70 ~ C-
-8 LLate Albian - -

Deposition

-10 I t t I i i I [ I I i
100 80 60 40 20 0
Time (Ma)
FIG. 7. Idealized porewater evolution path ( ( ~ 1 8 0 of water versus time) for the Viking sandstones and
conglomerates. The exact shape of the curve from stage III to stage V is uncertain (see text).
 Isotope studies in the Viking Formation 293
tion and burial diagenesis proceeded (stage II, Upper Cretaceous units; Longstaffe 1983, 1986,
diagenetic chlorite, quartz overgrowths, dolo- Ayalon & Longstaffe 1985, in press); (ii) the
mite), the porewaters became enriched in 180 fraction of low -180 meteoric water mixed with
due to water/rock interaction. This diagenetic pre-existing formation water was smaller in the
style was halted by the major Laramide Orogeny Viking Formation than in the Upper Cretaceous
in the early Eocene. Maximum burial and relief units, and (or) (iii) meteoric water which pene-
(due to uplift) was accompanied by recharge of trated the Viking Formation had greater oppor-
Tertiary and Cretaceous rocks in the basin by tunity for water/rock interaction and ~80-enrich-
low-180 meteoric water (stage III, feldspar ment (e.g. longer travel path, higher
dissolution, albitization, and diagenetic quartz, temperatures, more time) than did fresh water
calcite and ankerite). The influx of surface- entering the stratigraphically higher units.
derived fresh water continued as the post-Eocene The relative importance of these possibilities
erosion progressed (stage IV, diagenetic kaolin and many others not considered here is not yet
group minerals, illite, illite/smectite) until the well understood. For example, the rate at which
porewater attained its present temperature and meteoric waters entered the subsurface following
composition (stage V, diagenetic illite, uplift of the basin and in turn became enriched in
illite/smectite ?). cations and 180 due to water/rock interaction, is
The model proposed for the diagenesis of the unknown. Hence, the exact shape of the 6180
Lower Cretaceous Viking sandstones and cong- water versus time curve from stages III to V (Fig.
lomerates implies large-scale movement of 7) remains speculative. Radiogenic and deuter-
meteoric water, at least within the Cretaceous ium-isotope studies of the illite and illite/smectite
portion of the western Canada sedimentary may provide greater insight into this portion of
basin. That meteoric water has been important the porewater evolution curve in the western
on a basin-wide scale is further substantiated by Canada sedimentary basin.
the even lower 6180 values of equivalent min-
erals from Upper Cretaceous sandstones within
the same locality (Longstaffe 1983, 1986, Ayalon ACKNOWLEDGMENTS:The capable assistance of Ms D
& Longstaffe 1985, in press). T h e more ~SO-rich Caird in the laboratory is gratefully acknowledged as
nature (by about 3-5%0) of both early and late are numerous discussions with M. Dean. Grants in aid
diagenetic minerals from the Viking Formation of this research were provided by the Natural Sciences
and Engineering Research Council of Canada
may have resulted from a combination of factors, (NSERC A7387) to FJL. Helpful reviews of an earlier
including: (i) the starting porewater in the version of this paper were provided by K. Bjorlykke
Viking sandstones and conglomerates and cong- and P. J. Hamilton. Financial aid for travel to the
lomerates was dominantly sea water, rather than Symposium from its organizers and industrial sponsors
brackish to fresh water (as was the case in some is acknowledged with great pleasure.

References
AMAJOR, L. C. 1980. Chronostratigraphy, depositional C. T. (eds). Siliciclastic Shelf Sediments, pp. 163-
patterns and environmental analysis of sub- 77. Society of Economic Paleontologists and
surface Lower Cretaceous (Albian) Viking reser- Mineralogists, Special Paper 34.
voir sandstones in central Alberta and part of - - , BOUTILIER,R., MACKENZIE,m. S. & RULLKOT-
southwestern Saskatchewan. PhD thesis, Univer- TER, J. 1985. Isomerization and aromatization of
sity of Alberta, Edmonton, Canada. hydrocarbons and the paleothermometry and
AVALON,A. & LONGSTAFFE,F. J. 1985. Diagenesis and burial history of Alberta foreland basin. American
pore-water evolution of the basal Belly River Association of Petroleum Geologists Bulletin, 69,
sandstone, Alberta: mineralogic, petrographic 546-66.
and stable isotope studies (abstract). Geological BECKER, R. H. & CLAYTON, R. N. 1972. Carbon
Society of America, Program with Abstracts, 17, 516. isotopic evidence for the origin of a banded iron-
--&-- in press. Oxygen-isotope studies of dia- formation in Western Australia. Geochimica et
genesis and porewater evolution in the western Cosmichimica Acta, 36, 577-95.
Canada sedimentary basin: evidence from the BERNER, R. A. 1971. Principles of Chemical Sedimento-
Upper Cretaceous basal Belly River sandstone. logy. Elsevier, New York.
Journal of Sedimentary Petrology. BJORLYKKE, K., ELVERHOI,A. & MALM, A. O. 1979.
BEAUMONT, C. 1981. Foreland basins. Geophysical Diagenesis in Mesozoic sandstones from Spitsber-
Journal of the Royal Astronomical Society, 65, 291- gen and the North Sea--a comparison. Geolo-
329. gisches Rundschau, 68, I 152-71.
BEAUMONT, E. A. 1984. Retrogradational shelf sedi- BOLES, J. R. 1982. Active albitization of plagioclase,
mentation: Lower Cretaceous Viking Formation, Gulf Coast Tertiary. American Journal of Science,
central Alberta. In: TILLMAN,R. W. & SIEMERS, 282, 165-80.
294 F. J. Longstaffe & A. Ayalon
CLAYTON, R. N. 1959. Oxygen isotope fractionation in stable isotope fractionation factors of geochemical
the system calcium carbonate-water. Journal of interest. In: FLEISCHER,M. (ed.). Data of Geochem-
Chemical Physics, 30, 1246-50. istry (6th edn.). United States Geological Survey
- - , FRIEDMAN, I., GRAF, D. L., MAYEDA, T. K., Professional Paper, 4 4 0 - K K .

MEENTS, W. F. & SHIMP, N. F. 1966. The origin of GAMMEL, H. G. 1955. The Viking Member in central
saline formation waters, 1. Isotopic composition. Alberta. Bulletin of the Alberta Society of Petroleum
Journal of Geophysical Research, 71, 3869-82. Geologists, 3, 63-9.
- - , O'NEIL, J. R. & MAYEDA, T. K. 1972. Oxygen GAUTIER, D. L. 1982. Siderite concretions: indicators
isotope exchange between quartz and water. of early diagenesis in the Gammon Shale (Creta-
Journal of Geophysical Research, 77, 3057-67. ceous). Journal of Sedimentary Petrology, 52, 859-
CRAIG, H. 1957. Isotopic standards for carbon and 71.
oxygen and correction factors for mass-spectro- GLAISTER, R. P. 1958. Petrology of the Blairmore
metric analysis of carbon dioxide. Geochimica et sandstone. Bulletin of the Alberta Society of
Cosmochimica Acta, 12, 133-49. Petroleum Geologists, 6, 43-9.
- -1961. Standards for reporting concentrations of GRANT, S. K. 1985. Geologic study of the Viking
deuterium and oxygen-18 in natural waters. Formation, Harmattan East Field. MSc thesis.
Science, 133, 1833-4. University of Alberta, Edmonton, Canada.
DEAN, M. E. 1986. Diagenesis of the Viking Forma- HACQUEBARD,P. A. 1977. Rank of coal as an index of
tion, south-central Alberta. MSc thesis. University organic metamorphism for oil and gas in Alberta.
of Alberta, Edmonton, Canada. In: DEROO, G., POWELL, T. G., TISSOT, B. &
, LONGSTAFFE,F, J. & HEIN, F. J. 1984. Mineralo- MCCROSSAN, G, (eds). The Origin and Migration of
gical controls on reservoir quality in the Harmat- Petroleum in the Western Canadian Sedimentary
tan Oil Field, Lower Cretaceous Viking Forma- Basin, Alberta, pp. 11-22. Geological Survey of
tion, south-central Alberta (abstract). Geological Canada Bulletin, 262.
Association of Canada, Program with Abstracts, 9, HEIN, F. J., DEAN, M. E., DEIURE, A. M., GRANT,
57. S. K., ROBB, G. A. & LONGSTAFFE,F. J. 1986. The
, & 1985. Regional patterns of diagene- Viking Formation in the Caroline, Garrington
sis in the Viking Formation, south-central Alberta and Harmattan East fields, western south-central
(abstract). Thirty-sixth annual technical meeting, Alberta: sedimentology and paleogeography. Bul-
Petroleum Society of CIM, p. 50. letin of Canadian Petroleum Geology, 34, 91-110.
~r -- In preparation. Diagenesis of the HITCHON, B. 1984. Geothermal gradients, hydrodyna-
Formation, south-central Alberta. Bulletin mics, and hydrocarbon occurrences, Alberta,
of Canadian Petroleum Geology. Canada. The American Association of Petroleum.
DICKINSON, W. R. & SNYDER, W. S. 1978. Plate Geologists Bulletin, 68, 713-43.
tectonics of the Laramide Orogeny. In: MArnEWS, & FRIEDMAN,I. 1969. Geochemistry and origin of
W. (ed.). Laramide FoMing Associated with Base- formation waters in the western Canada sedimen-
ment Block Faulting in the Western United States, tary basin--1. Stable isotopes of hydrogen and
pp. 355-66. Geological Society of America, oxygen. Geochimica et Cosmochimica Acta, 33,
Memoir, 151. 1 3 2 1 - 4 9 .

DUTTON, S. P. & LAND, L. S. 1985. Meteoric burial HURST, A. 1980. Occurrence of corroded authigenic
diagenesis of Pennsylvanian arkosic sandstones, kaolinite in a diagenetically modified sandstone.
southwestern Anadarko Basin, Texas. The Ameri- Clays and Clay Minerals, 28, 393-6.
can Association of Petroleum Geologists Bulletin, 69, - -& IRWIN, H. 1982. Geologic modelling of clay
22-38. diagenesis in sandstones. Clay Minerals, 17, 5-22.
ESLINGER, E. V. 1971. Mineralogy and oxygen isotope IRWIN, H., CURTIS, C. & COLEMAN, M. 1977. Isotopic
ratios of hydrothermal and low-grade metamor- evidence for the source of diagenetic carbonates
phic argillaceous rocks. PhD dissertation, Case formed during burial of organic-rich sediments.
Western Reserve University, Cleveland, United Nature, 269, 209-13.
States. JACKSON, M. L. 1979. Soil Chemical Analysis--Ad-
-- & SAVIN, S. M. 1973. Mineralogy and oxygen vanced Course (2nd edn). Published by the author,
isotope geochemistry of the hydrothermally al- Madison, Wisconsin.
tered rocks of the Ohaki-Broadlands, New Zea- LAND, L. S. 1980. The isotopic and trace element
land geothermal area. American Journal of Science, geochemistry of dolomite: the state of the art. In:
273, 240-67. ZENGER, D. H., DUYaAM, J. B. & ETHINGTON,
- & YEH, H. 1981. Mineralogy, O~8/O 16 and D/H
- R. A. (eds.). Concepts and Models of Dolomitiza-
ratios of clay-rich sediments from Deep Sea tion, pp. 87-110. Society of Economic Paleontolo-
Drilling Project site 180, Aleutian Trench. Clays gists and Mineralogists, Special Paper, 28.
and Clay Minerals, 29, 309-15. - -1983. The application of stable isotopes to studies
FOLK R. L. 1968. Petrology of Sedimentary Rocks. Pub- of the origin of dolomite and to problems of
lished by Hemphill's Book Store, Austin, Texas. diagenesis of clastic sediments. In: ARTHUR,
FOSCOLOS, A. E., REINSON, G. E. & POWELL, T. G. M. A. (organizer). Stable Isotopes in Sedimentary
1982. Controls on clay-mineral authigenesis in the Geology, pp. 4.1-4.22. Society of Economic Pale-
Viking sandstone, central Alberta, 1. shallow ontologists and Mineralogists, Short Course, 10.
depths. Canadian Mineralogist, 21), 141-50. 1984. Frio sandstone diagenesis, Texas Gulf
FRIEDMAN, I. & O'NEIL, J. R. 1977. Compilation of Coast: a regional isotopic study. In : SURDAM,R. &
 Isotope studies in the Viking Formation 295
MACDONALD, D. (eds). Clastic Diagenesis, pp. 47- brackish and freshwater swamp sediments. Ameri-
62. American Association of Petroleum Geolo- can Journal of Science, 282, 1151-83.
gists, Memoir, 37. ROBB, G. A. 1985. Sedimentology and diagenesis of the
-- & DUTTON, 1978. Cementation of a Pennsylvan- Viking Formation, Garrington Oil Field, south-
ian deltaic sandstone: isotopic data. Journal of central Alberta. MSc thesis. University of Alberta,
Sedimentary Petrology, 48, 1167-76. Edmonton, Canada.
LAWRENCE, J. R . , GIESKES, J. M . & BROECKER, W . S. ROSENBAUM,J. & SHEPPARD,S. M. F. 1986. An isotopic
1975. Oxygen isotope and cation composition of study of siderites, dolomites and ankerites at high
DSDP pore waters and the alteration of layer II temperatures. Geochimica et Cosmochimica Acta,
basalts. Earth and Planetary Science Letters, 27, 1- 50, 1147-50.
10. SAVIN, S. M. & EPSTEIN, S. 1970a. The oxygen and
LEE, M. & SAVIN, S. M. 1985. Isolation of diagenetic hydrogen isotope geochemistry of clay minerals.
overgrowths on quartz and grains for oxygen Geochimica et Cosmochimica Acta, 34, 25-42.
isotopic analysis. Geochimica et Cosmochimica -- & -- 1970b. The oxygen and hydrogen isotope
Acta, 49, 497-501. geochemistry of ocean sediments and shales.
LONGSTAFFE,F. J. 1983. Diagenesis, IV. Stable isotope Geochimica et Cosmoehimica Acta, 34, 43-63.
studies of diagenesis in clastic rocks. Geoscience & YEn, H. 1981. Stable isotopes in ocean
Canada, 10, 44-58. sediments. In: EMILIANI,C. (ed.). The Sea, volume
- - 1984. The role of meteoric water in diagenesis of 7, The Oceanic Lithosphere, pp. 1521-54. Wiley,
shallow sandstones: stable isotope studies of the New York.
Milk River aquifer and gas pool. In: SURDAM, R. SCHWARTZ, F. W., MUEHLENBACHS,K. & CHORLEY,
& MACDONALD, D. (eds). Clastic Diagenesis, pp. D. W. 1981. Flow-system controls of the chemical
81-98. American Association of Petroleum Geolo- evolution of groundwater. Journal of Hydrology,
gists, Memoir, 37. 54, 225-43.
- - 1986. Oxygen isotope studies of diagenesis in the SHARMA,Z. • CLAYTON,R. N. 1965. Measurements of
basal Belly River sandstone, Pembina I-Pool, O18/O 16 ratios of total oxygen of carbonates.
Alberta. Journal of Sedimentary Petrology, 56, 78- Geochimica et Cosmochimica Acta, 29, 1347-53.
88. SLIPPER, S. E. 1918. Viking gas field, structure of area.
- - 1987. Mineralogical and oxygen-isotope studies of Geological Survey of Canada, Summary Department
clastic diagenesis: implications for fluid flow in 1917, P a r t C, 8c.
sedimentary basins. In: HITCHON,B. (ed.). Hydro- SUCHECKI, R. K. & LAND, L. S. 1983. Isotopic
geology of Sedimentary Basins. Proceedings of the geochemistry of burial-metamorphosed volgano-
Third Annual Canadian/American Conference on genic sediments, Great Valley sequence, northern
Hydrogeology, National Water Well Association California. Geochimica et Cosmochimica Acta, 47,
(in press). 1487-99.
- - & AYALON, A. 1986. Oxygen-isotope studies of SYERS, J. K., CHAPMAN,S. L., JACKSON, M. L., REX,
diagenesis in clastic rocks from the Viking and R. W. & CLAYTON, R. N. 1968. Quartz isolation
Belly River Formations, Alberta (abstract). Terra from rocks, sediments and soils for determination
Cognita, 6, 109. of oxygen isotopes. Geochimica et Cosmochimica
MAGARA, K. 1976. Thickness of removed sedimentary Acta, 32, 1022-5.
rocks, paleopore pressure, and paleotemperature, TAYLOR, H. P. JR 1974. The application of oxygen and
southwestern part of western Canada Basin. hydrogen isotope studies to problems of hydro-
American Association of Petroleum Geologists Bulle- thermal alteration and ore deposition. Economic
tin, 60, 554-65. Geology, 69, 843-83.
MAJOROWlCZ, J. A. & JESSOP, A. M. 1981. Regional TAYLOR, R. S., MATHEWS, W. H. & KUPSCH, W. O.
heat flow patterns in the western Canadian 1964. Tertiary. In: MCCROSSAN, R. G. & GEMS-
sedimentary basin. Tectonophysics, 74, 209-38. TEa, R. P. (eds). Geological History of Western
- - , RAHMAN,M., JONES, F. W. & MCMILLAN, N. J. Canada, pp. 190-4. Alberta Society of Petroleum
1985. The paleogeothermal and present thermal Geologists.
regimes of the Alberta basin and their significance THOMAS, M. B. & OLIVER, T. A. 1979. Depth-porosity
for petroleum occurrences. Bulletin of Canadian relationships in the Viking and Cardium Forma-
Petroleum Geology, 33, 12-21. tions of central Alberta. Bulletin of Canadian
NURKOWSKI, J. R. 1984. Coal quality, coal rank Petroleum Geology, 27, 209-28.
variation and its relation to reconstructed overbur- TILLEY, B. J. & LONGSTAFFE, F. J. 1986. (~180 results
den, Upper Cretaceous and Tertiary Plains coals, for silicate and carbonate cements from Lower
Alberta, Canada. American Association of Petro- Cretaceous sedimentary rocks in the Alberta Deep
leum Geologists Bulletin, 68, 285-95. Basin, Canada (abstract). Terra Cognita, 6, 107.
O'NEIL, J. R., CLAYTON,R. N. & MAYEDA,T. K. 1969. WEIMER, R. J. 1983. Relation of unconformities,
Oxygen isotope fractionation in divalent metal tectonics, and sea level changes Cretaceous of the
carbonates. Journal of Chemical Physics, 51, 5547- Denver Basin and adjacent areas. In: REYNOLDS,
58. M. W. & DOLLY, E. D. (eds). Mesozoic Paleogeo-
PILL, K. 1971. Palynology of Oligocene sediments of graphy of West-central United States, pp. 359-76.
central British Columbia. Canadian Journal of Rocky Mountain Section, Society of Economic
Botany, 49, 1885-920. Paleontologists and Mineralogists.
POSTMA, D. 1982. Pyrite and siderite formation in WENNER, D. B. & TAYLOR, H. P. JR 1971. Tempera-
296 F. J. Longstaffe & A. Ayalon
tures of serpentinization of ultramafic rocks based clays in sandstones: recognition and influence on
upon 018/016 fractionation between coexisting reservoir properties and paleoenvironmental anal-
serpentine and magnetite. Contributions of Minera- ysis. Journal of Sedimentary Petrology, 476, 3-31.
logy and Petrology, 32, 165-85. YEn, H. & SAVlN, S. M. 1976. The extent of oxygen
WILLIAMS, G. D. & STELCK, C. R. 1975. Speculations isotope exchange between clay minerals and
on the Cretaceous paleogeography of North seawater. Geochimica et Cosmochimica Acta, 40,
America. In: CALDWELL, W. G. E. (ed.). The 743-8.
Cretaceous System in the Western Interior of North -- & -- 1977. Mechanism of burial metamor-
America, pp. 1-7. Geological Association of phism of argillaceous sediments: 3. O-isotope
Canada, Special Paper, 13. evidence. Geological Society of America Bulletin,
WILSON, M. D. & PITTMAN, E. D. 1977. Authigenic 88, 1321-30.

FRED J. LONGSTAFFE, Department of Geology, University of Western Ontario, London,

Ontario N6A 5B7, Canada.
AVNER AVALON, Geological Survey of Israel, 30 Malchei Israel Street, Jerusalem 95501,
Israel.
 Secondary porosity in hydrocarbon-bearing transgressive
sandstones on an unstable Lower Palaeozoic continental shelf,
Welsh Borderland
J. Parnell

S U M M A RY : Coarse sandstones were deposited during each of four transgressions across

the shelf of the Lower Palaeozoic back-arc Welsh Basin. Hydrocarbon shows record porosity
in the deposits of the Cambrian, Arenig, Caradoc and Llandovery transgressions.
Hydrocarbons migrated from two sources: from Lower Palaeozoic source rocks within the
basin, and later from Carboniferous rocks.
Diagenesis and reservoir potential is related to the mineralogy of the sandstones, which
reflects a balance between reworking of shelf sediment and an input from the eroded
substrate. The substrate included Precambrian volcanic and plutonic rocks formed at the
southern margin of the Iapetus Ocean. Quartz arenites are quartz-cemented or exhibit thick
pore-linings of clay which restrict permeability. They have only minor reservoir potential.
Arkoses do not exhibit macroporosity but have effective microporosity in the immediate
vicinity of source rocks. Subarkoses have optimal reservoir potential, consisting of primary
porosity about non-quartz grains (feldspars, glauconite) and secondary (intragranular)
porosity in feldspars and lithics. The secondary porosity formed by leaching of clay alteration
products in feldspars. Subarkoses retained sufficient framework stability to preserve the
secondary porosity, whereas arkoses experienced widespread alteration and framework
collapse. Limited secondary porosity after calcite dissolution may be due to surficial
weathering of the Lower Palaeozoic rocks during Devonian/Lower Carboniferous exposure.

An active continental margin presents a favour- make a major contribution to the siliciclastic
able setting for marine transgressive events sediment budget. Volcanic lithic fragments,
(Swift 1968). For example, several thick generally feldspathic, are highly reactive and
transgressive sandstones are recorded in the susceptible to hydration reactions (Surdam &
Cretaceous sequence of the Pacific coast of Boles 1979). The diagenetic products (clays,
North America (Bourgeois 1980). Rises in sea- zeolites, carbonates, oxides) tend to be less dense
level in the fore-arc region or marginal basin of a and therefore occupy a greater volume to the
destructive plate boundary are recorded by detriment of porosity. The reservoir potential of
transgressive stratigraphy across the adjacent such rocks is therefore generally poor (Nagtegaal
foreland. The transgressive sediments tend to be 1978, Galloway 1979). Nevertheless, there is
relatively coarse siliciclastic deposits, whilst fine- potential for generation of secondary porosity by
grained sediments are deposited in the outer leaching of volcanic grains if sufficient quartz is
shelf regions where an increase in water depth present to provide a stable framework (e.g.
accompanies the transgression. Outer shelf Loucks et al. 1979), and some volcaniclastic
mudrocks deposited during transgressive events reservoirs are known (reviewed by Seeman &
can contain substantial proportions of organic Scherer 1984). Petrographic studies of the sedi-
matter, dominated by types I and II kerogen ment in back-arc and fore-arc basins show them
(Demaison & Moore 1980). In favourable cir- to be mixtures of immature rock fragments and
cumstances, hydrocarbons generated from a reworked quartz-rich sediment (e.g. Ingersoll &
basinal/outer shelf mudrock sequence could be Suczek 1979). The relative proportions of these
reservoired in shelf sandstones (Galloway & components will influence the framework stabil-
Hobday 1983). The shelf mudstones may func- ity of the sediment.
tion as reservoir seals as well as source rocks. Transgressive deposits in the Lower Palaeo-
Brenner (1978) has described significant pet- zoic Welsh Basin have been examined to
roleum accumulations in shelf sandstones, pre- determine whether they function/have func-
served by updip stratigraphic pinchouts with a tioned as hydrocarbon reservoirs. The Welsh
shale seal, from the late Cretaceous Sussex Sand- Basin is a particularly interesting case for study
stone of Wyoming. because hydrocarbon migration and entrapment
However, another feature of unstable conti- can be envisaged from two distinct sources: (i)
nental margins is that volcanic material may maturation of Lower Palaeozoic source rocks,

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 297

Geological Society Special Publication No. 36, pp. 297-312.
298 J. Parnell

i.e. a source within the basin, (ii) migration from basement rocks outcrop in two main regions
Carboniferous reservoir rocks (containing hy- today, forming the topographic highs of the
drocarbons of Carboniferous origin) into uplifted Longmynd and the Malvern Hills (Fig. 1). These
Lower Palaeozoic rocks, i.e. a younger external regions were also topographic highs during the
source (further details in Parnell 1987). Lower Palaeozoic (Bridges 1975). The outcrops
Only transgressive shelf sandstones are dis- of Cambrian, Arenig and Caradoc rocks are all
cussed here. The basinal deposits include exten- in the two regions of basement outcrop. A map of
sive turbidite sandstones, many of which are the Longmynd region (Fig. 3) shows successive
petrographically greywackes. No porosity was onlap of Cambrian, Caradoc, Llandovery and
observed in the basinal sandstones. Carboniferous rocks on to the Precambrian, with
unconformities between each. Llandovery rocks
are more widely exposed (Fig. 4).
The Welsh Basin The mineralogy of the transgressive sand-
stones reflects a balance between reworking of
The Lower Palaeozoic Welsh Basin evolved as a sediment on the shelf and an input of new
back-arc basin at the southern continental sediment as the shelf substrate was eroded during
margin of the Iapetus Ocean. A comprehensive transgression. The balance has an important
review of tectonic setting, stratigraphy and bearing on the mineralogical and textural ma-
palaeogeography is given by Bassett (1980). turity (and hence diagenesis) of the sediments.
Towards the end of the Precambrian, the margin Whilst reworking enhanced the quartz content of
was a site of very active subsidence, plutonism, the sands, erosion of the substrate created a
volcanism and deformation. The Lower Palaeo- supply of lithic clasts, many of which were
zoic rocks accumulated on a basement of schists, susceptable to alteration. Sandstones of each
gneisses, low grade metasediments (Longmyn- transgressive episode contain abundant well-
dian) and a variety of igneous rocks dominated rounded quartz grains. In addition to shelf
by the calc-alkaline products of volcanic arcs reworking, a contribution from aeolian rework-
(Thorpe et al. 1984). In the Welsh Borderland the ing is suggested by records of dreikanters in the
basin shallowed to a shelf sea of varying width on Cambrian and Caradocian. Unstable lithic ma-
the western side of the Midland Platform (Fig. terial from the substrate included volcanic,
1). The sea transgressed across the shelf on plutonic and mudrock clasts. In particular,
numerous occasions, most importantly during Uriconian (Precambrian) basic to acid volcanic
the Lower/Middle Cambrian, Arenig, lower clasts occur in many of the coarse beds of the
Carodoc and Llandovery (Fig. 2). The shelf northern Welsh Borderland. An Ordovician
sediments are generally shallow-water siliciclas- volcanic basement in the western part of the
tic and carbonate deposits. In the basin, sedi- region also supplied unstable material.
mentation was more continuous, in the form of
pelagic muds/silts and turbiditic sands.
The basinal sediments include substantial Hydrocarbon shows
thicknesses of organic-rich sediment. Significant Hydrocarbon shows have been recorded in the
black shale deposition also occurred in outer deposits of all four transgressive episodes. De-
shelf areas following transgressive events, par- tails of occurrences in eight districts (Fig. 4) are
ticularly during and after the Caradoc and summarized in Table 1. The hydrocarbon shows
Llandovery transgressions (Leggett 1978, 1980). in Cambrian sandstones are described and
The basinal sediments experienced mild meta- discussed by Parnell (1987).
morphism to greenschist facies (Bevins & Row- Trecoed Beds sandstones occupy an ideal
botham 1983), and consequently organic-rich setting for a hydrocarbon reservoir from Lower
beds are overmature. Sediments on the inner Palaeozoic source rocks. They were deposited on
shelf have not experienced this degree of the outer shelf of the basin (James 1983) resting
metamorphism. Kelling (1978) estimates that the unconformably on Ordovician black shales and
shelf succession reaches 5 km thickness, whilst overlain by Upper Llandovery/Wenlock black
up to 15 km accumulated in the basin. Accord- shales (Jones 1949). They are overlapped by the
ingly, the organic-rich beds of the inner shelf overlying shales in a shorewards direction, i.e.
may have some residual hydrocarbon potential, the shales form a stratigraphic seal as well as a
particularly for gas. potential source rock (Fig. 5). The hydrocarbon
shows in the Folly Sandstone occur in the
Transgressive sandstone deposits same region, underlying bituminous Wenlockian
limestones. The Lower Cambrian sandstones of
Transgressive sandstones are exposed where the Malvern Hills are hydrocarbon-bearing
they onlap against basement rocks. Precambrian where they are faulted against Upper Cambrian
 Sandstone porosity, Welsh Borderland 299

0
I
Km 50
i
f
t

/ /

/ ~ ............<.,.
/ i
/ IRISH SEA -/fi:; ........ ~.-..':.-.,
/ PLATFOR/~" ://
/ / ~ /:'

. /
/

~:~ /

London"~-~i

............... Limit of Lr. Poloeozoic. Lr. CorQdoc shoreline.

, 9 9 [ombrimn shoreline. western limit of sub-
Llondovery unconformity.
tower Arenig shoreline. ---- major clcfive fou[fs.

FIG. 1. Location of Welsh Basin. Shoreline changed through time (Cambrian after Bassett, 1980, Arenig and
Caradoc after Anderton et al. 1978). L, M = Longmynd and Malvern Hills, regions of basement outcrop
which are/were topographic highs. PLF = Pontesford-Linley Fault, CSF = Church Stretton Fault. Box
indicates region of Fig. 3.

black shales (Parnell 1987). Both of these Palaeozoic and Carboniferous rocks in Shrop-
instances probably also had a Lower Palaeozoic shire had the same source (Robinson et al. 1986).
source. The Black Grit is similarly near the Carbonifer-
In the northern part of the Welsh Borderlands, ous outcrop (Fig. 3), but is overlain by dark
the shows in the Cambrian of Shropshire, Caradocian shales which are intruded by a
Stiperstones Quartzite, Hoar Edge Grit and Bog dolerite sill (Wedd 1932). Elsewhere in the Welsh
Quartzite probably had a Carboniferous source. Borderlands hydrocarbons were generated from
They occur close to unconformably overlying Lower Palaeozoic shales by the heat of igneous
Upper Carboniferous Sandstones (Figs 3 and 4) intrusions (Parnell 1983), and a Lower Palaeo-
that are locally impregnated with oil. Organic zoic source for the Black Grit hydrocarbons is
geochemical studies confirm that shows in Lower likely.
300 J. Parnell

LUDLOW

Siturion
WENLOCK

LLANDOVERY TRANSGRESSION
Bog Quartzite
FoLlySandstone
TrecoedBeds
CAR~DOC ~ 41
TRANSGRESSION
LLANDEILO Block Grit
HoarEdgeGrit

Ordovician
LLANVIRN L ~

TRANSGRESSION
Stiperstones Quartzite

TREMADOC

Upp~
Middte
Co,mbrion
TRANGRESSION
Comtey Sandstones
Precarnbrian 4 Hoi[ybush Sorldstone

FIG. 2. Generalized stratigraphic succession for the Welsh Borderland (from Earp & Hains 1971), showing
main transgressive episodes. Named stratigraphic units are hydrocarbon-bearing transgressive sandstones.
V = episodes of volcanic activity.

Sample analysis samples from the Hollybush Sandstone and

Black Grit, which are fine-grained hydrocarbon-
Sandstone samples, from surface outcrop, were
bearing arkoses and litharenites.
examined in thin section. Many samples are
The composition of Cambrian and Llandovery
from breached hydrocarbon traps and contain
sandstones clearly differs between the northern
reservoir bitumen/pyrobitumen. Samples not
and southern Welsh Borderlands. The sand-
plugged with hydrocarbons were impregnated
stones in the south are less mature. The propor-
with blue resin. Percentage porosity in hydrocar-
tion of acid volcanic rocks in the northern
bon-bearing sandstones, recorded in Table 1, is
basement is higher, contributing quartzose
porosity now occupied by hydrocarbon. These
grains to the sandstones there. In Llandovery
porosity values are therefore not exaggerated by
times, the sedimentary environments were dif-
surface weathering. Clay mineralogy was exam-
ferent with a predominance of marine reworked
ined and identified using scanning electron
sands in the north, but with a contribution from
microscopy with an EDAX facility. Analyses
regressive sands in the south (Bridges 1975).
were supported by X-ray diffraction analyses.

Quartz arenites
Quartz arenites were collected from the Cam-
brian sandstones of Shropshire, the Caradoc
Sandstone diagenesis Hoar Edge Grit and open marine Llandovery
sandstones. The quartz arenites are all medium
The relative proportions of quartz and unstable to coarse sandstones. They generally show less
grains is fundamental to the reservoir properties than 5% porosity. Porosity has been reduced by:
of the Welsh Borderland sandstones. Accord- (i) quartz overgrowths (Fig. 7a) which may
ingly, quartz arenites, subarkoses and sublithar- completely fill the pore-space, particularly in the
enites, and arkoses and litharenites (Fig. 6, Cambrian sandstones; (ii) pore-filling kaolinite,
classified according to Folk 1974) are discussed particularly in the Hoar Edge Grit, where quartz
separately. The samples plotted are medium to arenites are interbedded with arkoses and
coarse-grained sandstones, with the exception of subarkoses.
 Sandstone porosity, Welsh Borderland 3oi

k 5rif
Ord. MESOZOIC

% " . .

ga, J 9 9 9

-0

Sit. /0rd. ~ BOG, LONfMYND

COAL MEASURES

,oo
(WESTPHALIAN)
LLANDOVERY
-=/,
~ COSTONIAN
, .
, (CARADOC)
STIPERSTONES QUARTZITE 0 km 10
I
(ARENIG)
,-.._ -..
8"
:i]::.,';~ LOWER/MIDDLE CAMBR1AN
;-....;..:

/~?- PRECAMBRTAN
i ! 9

FIG. 3. Outcrops of transgressive deposits about the Longmynd topographic high, Shropshire. Cambrian,
Caradoc, Llandovery and Carboniferous sandstones successively onlap the Precambrian basement, with
unconformities between each.

Scanning electron microscopy shows that a shire in residual primary porosity (Fig. 7a); (ii)
thin grain coating of clay (variably illite or the Hoar Edge Grit, in which very minor traces
chlorite) predates the quartz overgrowths. At (less than 1%) occur in kaolinite-cemented
Rubery, the easternmost and possibly least quartz arenites and subarkoses.
buried exposure, cementation is incomplete and
the sandstone locally has 15-20% porosity.
Subarkoses or sublitharenites
However, very well developed pore-linings of
smectite/mixed-layer clay (Fig. 7b) severely limit The majority of sandstone samples can be
permeability. classified as subarkoses/sublitharenites. Glau-
conite is an important component of the Cam-
Hydrocarbon shows brian sandstones and a minor component of
Caradocian and Llandovery sandstones.
Limited hydrocarbon shows occur in quartz In many cases the subarkoses/sublitharenites
arenites, in" (i) Cambrian sandstones in Shrop- exhibit significant porosity due to the preserva-
302 J. P a r n e l l

' 3b

0 20 4O
i
' k'm

30-

.Rubery

treco d Trecoed Beds

~o Presteigne
Folly.Sandstone
0

I MolvernHilts
~ Hollybush Sandstone

Ltondovery outcrop
A HydrocQrbon shows in Moy Hitt
Lower Potoeozoic sandstones
FIG. 4. Llandovery outcrops in Welsh Borderland, and locations of hydrocarbon shows in Lower Palaeozoic
sandstones. BG = Black Grit (Caradoc), BQ = Bog Quartzite (Llandovery), CS = Comley Sandstones
(Cambrian), HEG = Hoar Edge Grit (Caradoc), SQ = Stiperstones Quartzite (Arenig). Box indicates region of
Fig. 3.

tion of residual primary porosity and the leach- grains with pores irregularly lined with clay to
ing of unstable grains. Porosity is particularly grains exhibiting 'clean' porosity.
associated with feldspar and glauconite grains The mineralogy of the clay mineral replace-
because: (i) overgrowths are not developed about ments is broadly related to the mineralogy of the
glauconite and only rarely about feldspar, (ii) original grains. Kaolinite replacements occur
glauconite grains develop fissures during glau- within feldspar grains, illitization is prominent
conitization (Odin & Matter 1981), (iii) altera- in feldspars and mudrock and volcanic clasts and
tion and leaching of feldspars gave rise to chlorite is a common replacement in basic
significant intragranular porosity. volcanic clasts (see Burley et al. 1985). Early clay
The honeycombed nature of the feldspar precipitation on grain surfaces pre-dated quartz
grains is strong evidence for post-depositional overgrowths. Where the clay layer was thin (not
leaching as they would rapidly disintegrate more than a few microns) overgrowths were
during sediment transport. Microprobe analysis readily developed. Thicker pore-lining clays
of altered feldspars in Cambrian sandstones apparently prevented quartz overgrowth. Thick
shows that sodic portions of perthitic inter- pore-linings are intact in rocks showing evidence
growths have been preferentially altered and for the leaching of clays from altered feldspars.
leached. In addition to the leaching of feldspar Pore-lining clays include illite and mixed layer
grains, the leaching of volcanic and mudrock illite/smectite, but are most commonly iron-rich
clasts and calcite bioclasts gave rise to porosity of chlorite. In several districts, particularly in the
up to 20%. The lithic clasts show varying stages southern Welsh Borderlands, the chlorites are
of replacement by kaolinite, illite or chlorite. The stained red by haematite/goethite due to altera-
replacive clays were locally leached to yield pore tion of the chlorite or later percolation of iron-
space. There is a gradation from feldspar or lithic rich fluids draining the Precambrian basement
grains with patches of replacive clay through rocks. The iron oxides severely limit microporo-
Sandstone porosity, Welsh Borderland 3o3

II
eL

~ o ~
o
8 8888~,8

o o

II
=
,A

L,,

r.~ o . - o_~ ~-~ ~

.=_
304 J. Parnell

o'3 & & x'~ illite grain-coating, quartz overgrowth and pore-
0 km ~ \I.AEsGwY..E filling kaolinite.
Calcite is present in three forms: detrital,
cement and replacement. Calcite bioclasts are
usually represented by moulds, but calcite is
preserved in fresh exposures such as road
cuttings and borehole samples. The mouldic
porosity is clearly not deeply penetrating. Sand-
stones above and below inner shelf limestones
WENLOCK ,\ ~
were susceptible to early carbonate cementation.
SHALES \ . ~ ORDOVIC!AN Sandstone immediately adjacent to limestone is
s~, ., \ " ~ SHALES still calcite-cemented. The transition from cal-
- \
cite-cemented sandstone to quartz-cemented
SANDSTONE sandstones is usually sharp and there are no
BURGHROAD \ X ~ ~ - - traces of calcite cement pre-dating overgrowths
II STATION \
in the quartzose sandstones. It can be assumed
that calcite cementation was limited to a zone
adjacent to the limestones. The Hoar Edge Grit
! I
is the only sandstone under consideration which
contains numerous limestone bands and traces of
WEN LOCK , ~ l l ~ l ~ l " ~ ' ~ calcite cement. Many Hoar Edge Grit subar-
ORDOVICIAN koses do not exhibit quartz overgrowths but are
LLANDOVERY
cemented by authigenic kaolinite. Their minus-
cement porosity is abnormally high (up to 25%)
and features including corroded grains, inhomo-
FIG. 5. Outcrop of outer shelf Llandovery sandstone
(Trecoed Beds) near Builth Wells. Sandstones are geneity of grains and over-sized pores suggest
sandwiched between Ordovician and Wenlock black pre-kaolinite secondary porosity. Kaolinite-
shales, and overstepped by the latter. Shales acted as cemented sandstone grades laterally into calcite-
a seal to hydrocarbons in the sandstone. cemented sandstone.

Hydrocarbon shows
sity within the pore-linings, and in some Llando- Hydrocarbon occurrences in the Comley Sand-
very rocks in the Malvern Hills they completely stones, the Stiperstones Quartzite and the Tre-
fill the remaining pore space. Some sandstones in coed Beds all depend on combined primary and
which cementation by quartz was incomplete secondary porosity. Hydrocarbons fill the pri-
contain pore-bridging kaolinite. For example, mary pores associated with grains which do not
Hoar Edge Grit samples can show a sequence of exhibit overgrowths, as noted above (Fig. 7c).

Q Q Q

"SQ N

G'I
_.~/S~
I

F L F L F L
CAMBRIAN ARENIG / CARADOC LLANDOVERY

FIG. 6. Compositional field of sandstones from transgressive deposits, Welsh Borderland. Subdivision into
quartz arenites, subarkoses, sublitharenites, arkoses and litharenites according to Folk (1974). End members
Q = monocrystalline and polycrystalline quartz including chert, F = total feldspars, L = lithics excepting
quartzose but including calcite bioclasts and glauconite. Interstitial material not counted. Plagioclase/total
feldspar ratio varies from 0.2 to 1.0. N, S = northern and southern Welsh Borderland, BG = Black Grit,
HEG = Hoar Edge Grit, SQ = Stiperstones Quartzite, SWG = Spy Wood Grit. Samples are medium/coarse-
grained, except fine-grained Cambrian (south) and Black Grit.
 Sandstone porosity, Welsh Borderland 305

FIG. 7. (a) Residual primary porosity between quartz overgrowths, infilled with hydrocarbon (black),
Cambrian Sandstone, Comley (field width 0.8 mm). (b) Pore-lining of mixed layer clay impregnated with iron
oxides, Llandovery quartz arenite, Rubery. Pore is filled with mounting resin (field width 1.2 mm). (c)
Glauconite grain with hydrocarbon in shrinkage fissure and in adjacent primary porosity, Cambrian
sandstone, Comley (SEM, field width 0.25 mm).
306 J. P a r n e l l

Fissuring in glauconite grains in the Comley dated quartz overgrowth and the precipitation of
Sandstones, is commonly occupied by hydrocar- pore-lining clays. The arkoses are substantially
bon (Fig. 7c and Parnell 1987). Hydrocarbons fill cemented and replaced by clay minerals, particu-
micropores between individual crystals in the larly chlorite.
boxwork fabric of glauconite grain surfaces (see
Odin & Matter 1981). Hydrocarbon filling of
Hydrocarbon shows
secondary pores within feldspars is also wide-
spread in the Cambrian sandstones (Fig. 8a) and Hydrocarbons impregnate arkosic sandstones in
in the Llandovery Trecoed Beds (Fig. 8b). This the Black Grit and Hollybush Sandstone. The
secondary porosity appears to be after leaching 6 m Black Grit is thoroughly impregnated with
of clays rather than calcite dissolution. The hydrocarbon at localities 2 km distant along-
gradation from replacement clay to clean pores strike. The Hollybush Sandstone is patchily
in feldspar grains shows that secondary porosity hydrocarbon-bearing at both extremes of a
is due to leaching of replacive clay rather than 1.2 km length outcrop, but only where faulted
direct leaching of feldspars. Feldspar grains in against black shales. Sandstone 100 m from the
the Cambrian sandstones exhibit alteration to fault is petrographically similar but not hydro-
chlorite and illite in quartz-cemented beds, but carbon-bearing (Fig. 10a,b). Grains appear to
are little altered in calcite-cemented beds. Secon- float in a hydrocarbon cement (Figs 8d and 10a)
dary porosity, and thence hydrocarbons, is because the hydrocarbons 'stain' the clay matrix.
limited to the quartz-cemented beds. Backscattered electron imagery of polished sur-
Fifty samples of Middle Cambrian sandstone faces show that the hydrocarbons fill micropores
(Upper Comley Group) from Comley, Shrop- (sensu law ; 5-20 ~tm pore size) in chloritic clays.
shire, were point-counted for quartz (percentage Hydrocarbons also fill pores in altered feldspars
of grains) and hydrocarbon (percentage of (Fig. 8d).
volume) (Fig. 9). The results show low hydrocar- Arkosic rocks elsewhere in the Welsh Border-
bon contents for quartz arenites, but significant lands are not impregnated with hydrocarbon.
values for quartz grain contents 75 to 90%. Few
samples were available with quartz percentages
less than 75% and none yielded hydrocarbons. Discussion
The role of alteration to clay minerals is most
obvious in the Stiperstones Quartzite, in which Diagenesis and reservoir potential are clearly
feldspar grains and feldspathic volcanic clasts related to lithological composition. Figure 11
are extensively altered to illite. The illite has been summarizes the main diagenetic changes experi-
subsequently leached out and the secondary enced by a quartz arenite, subarkose and arkose
porosity filled with hydrocarbon (Fig. 8c and during progressive burial and then uplift to the
Parnell 1987). In the most extreme cases clasts of surface. The consequences of early cementation
up to 1.5 cm width were completely leached by calcite are also figured. Hydrocarbon migra-
away to leave cavities supported by the surround- tion is indicated as following the leaching of
ing framework of quartzose clasts and cement. replacive clays and calcite cement (Fig. 11).
Hydrocarbons occupying secondary porosity
after dissolution of calcite occur in the Bog
Importance of secondary porosity
Quartzite, where they fill surficial mouldic
porosity (Whittard 1932), and as very minor Schmidt & McDonald (1979a) distinguish five
traces in the Hoar Edge Grit within kaolinitic textural origins of secondary porosity in sand-
cement. stones; fracturing, shrinkage, dissolution of
sedimentary material, dissolution of authigenic
cement and dissolution of authigenic replace-
ment. All five types are recognized and play a
Arkoses/litharenites role in hydrocarbon distribution in the Welsh
Arkoses were collected from Cambrian sand- Borderlands. Fracture-bound hydrocarbons are
stones in the Malvern Hills, the Caradoc Black widespread and are discussed elsewhere (Bath et
Grit and Llandovery sandstones in the southern al. 1986). Hydrocarbon shows occupying shrink-
Welsh Borderlands. The Cambrian and Caradoc age fissures in glauconite grains and secondary
arkoses include fine-grained sandstones. Labile porosity after dissolution of calcite cement,
grains such as volcanic and mudrock clasts were calcite bioclasts and clay replacements have all
deformed during compaction and squeezed been recorded above.
between rigid grains to reduce porosity. Compac- In the Comley Sandstones, Stiperstones Quart-
tion was also accommodated by mechanical zite and Trecoed Beds, where hydrocarbons are
fracturing of feldspars. The compaction post- relatively abundant, they predominantly occupy
Sandstone porosity, Welsh Borderland 3o7

Es

,.0

~az ~

~o-~
.~ ~,..~,.,,~

0 ,...~0 ~
"~.~ ,.~

~ ~.~
3o8 J. Parnell

FRAMEWORK SECONDARY
COLLAPSE POROSITY CEMENTATION
12-

10- -IO

-B
%
HYDROCARBON 99
-6

9 o9 4
9 o9

9 9 o9 9 9 o9 o9 -2
I ee 9 .1
traces 9 9 9149 trQces
0 1 A A 30

50 60 70 80 90 1oo

~ DRAINS = QUARTZ

FIO. 9. Hydrocarbon content of Cambrian sandstone (Upper Comley Group) samples, Comley, Shropshire.
Fifty samples were point-counted for quartz as a percentage of grains and hydrocarbon as a total volume
percentage. Maximum hydrocarbon content is in subarkoses, reflecting secondary porosity in feldspars
preserved by framework (Fig. 8a). Quartz arenites (Fig. 7a) and arkoses show low porosity.

FIG. 10. (a) Fine-grained ark9 Cambrian Hollybush Sandstone, with hydrocarbon-impregnated matrix
(field width 2.4 mm), which passes laterally into (b) sandstone with chloritic matrix not impregnated with
hydrocarbon (field width 2.4 mm).

secondary porosity. Primary porosity is limited Primary porosity in arkoses

because of quartz overgrowth, or pore-lining Primary porosity is subordinate in medium/
chlorite or illite. Hydrocarbons have not been coarse-grained sandstone, but may be the main
observed within pores with thick clay-linings, mode of porosity in the two fine-grained sand-
but do occur rarely within kaolinite cement. stones which bear hydrocarbons. The microporo-
Although the porosity of the chlorite-bearing sity measured in chlorite/hydrocarbon mixtures
sandstones is often higher because quartz over- does not exceed 20~m width. This is of the
growth has been inhibited, the pore-lining magnitude expected for primary intra-cement
restricted hydrocarbon migration and the poros- pores, whilst secondary intra-cement pores often
ity is isolated and therefore not effective. Pore- exceed this size (Schmidt & McDonald 1979b).
filling kaolinite clearly reduces porosity but still Although the sandstones are substantially re-
exhibits an effective microporosity because placed and cemented by chloritic clay, which is
permeability between pores is not so restricted. detrimental to porosity, hydrocarbon impregna-
 Sandstone porosity, Welsh Borderland 309

DIAOENESIS =- UPLIFT

QUARTZ
ARENITE

SUBARKOSE

ARKOSE

CALCITE-
CEMENTED
SANDSTONE

CALCITE CENENTATION. QUARTZCEMENTATION. PORE-LININGCLAY. ALTERATION OF LEACHING OF CLAY. LEACHING OF CALCITE.

FELDSPAR TO CLAY. FRAMEWORKCOLLAPSE. HYDROCARBON
REPLACIVE CALCITE HYDROCARBON MIGRATION.
MIGRATION.

~ Quartz~ Fetdspar~ CQtcite~ Pore_tiningc[ay

~ Kootinite[ ]Porosify
Hydrocarbon

FIG. 11. Diagenetic changes experienced by sandstones of different composition, through progressive burial
and uplift. Hydrocarbon migration into secondary porosity follows leaching of replacive clays and calcite
cement.

tion is accounted for by the very close proximity ing burial diagenesis'. However, in the Lower
of the source rock horizons (see above). Palaeozoic sandstones of the Welsh Borderlands,
where cementation is significant, the optimal
reservoir potential is developed in sandstones
Framework stability
containing a proportion of non-quartz grains.
Secondary porosity in sandstones is dependent The cementation in quartz arenites left only
on the alteration of grains and cement, but it is minor primary porosity and negligible secondary
not enhanced by any resultant mechanical porosity. Arkosic sandstones have poor porosity
compaction. For secondary porosity to be pre- except where in immediate contact with the
served, there must be a balance between leaching source rock. The optimum reservoir potential
of grains and retaining sufficient stability to was in the subarkoses, where secondary porosity
prevent pore collapse. The importance of sand- in feldspars is combined with a framework
stone framework stability to porosity and per- stability conferred by quartz grains and their
meability has been discussed by Nagtegaal cementing overgrowths. Clearly there must be a
(1978), who showed that quartz frameworks are limit to the porosity which can be supported
the most stable. Sandstones containing abundant before the framework collapses and porosity is
lithic clasts are the least stable because they destroyed. As the proportion of feldspars and
experience plastic deformation of mechanically lithic fragments becomes greater, mechanical
weak clasts. In sandstones free of allochthonous deformation, collapse/precipitation of cement-
cement, Nagtegaal (1978) concluded that 'pure ing clays becomes increasingly important and
quartz arenites have the optimal potential of the reservoir potential progressively lower. The
preserving high porosity and permeability dur- importance of the non-quartz component to
3IO J. P a r n e l l

reservoir potential is demonstrated by the data in (1986) report that organic-rich shales are unlikely
Fig. 9. to yield enough CO2 to generate secondary
porosity. Furthermore, whilst feldspars are
leached in the Trecoed Beds calcite bioclasts are
Significance of marginal basin setting still preserved. Acidic fluids would have leached
the calcite. Similar selective leaching of feldspars
Volcanic and plutonic rocks formed at the plate
relative to calcite is recorded by Heald & Larese
margin contributed material to the sandstones.
(1973). The origin of the secondary porosity is
Whilst products of Ordovician volcanism con-
therefore not clear.
tributed to the Llandovery sediments, volcanic
and plutonic basement rocks formed during the
late Precambrian evolution of the margin were a Shropshire : meteoric fluids
more important source. Clasts from Malvernian, The hydrocarbon shows from Shropshire occupy
Uriconian and Longmyndian basement are
secondary porosity after both feldspar (clay)
widely recognized. The time-scale of volcanicity, dissolution and calcite dissolution. Calcite is
erosion and diagenesis was much longer than preserved in the Trecoed Beds, so a different
reported in other studies of marginal basin mechanism is needed to explain its leaching at
sandstones (e.g. Davies et al. 1979, Galloway
Bog and in the Hoar Edge Grit. The calcite
1979, Surdam & Boles 1979, Vessell & Davies
mouldic porosity recorded above at Bog appears
1981). High heat flow in back-arc basins
to be surficial and yet contains hydrocarbons, so
normally accelerates diagenesis (Lee & Klein
at least some of the surficial porosity was not
1986). Where erosion of volcanic rocks made a
caused by recent exposure. Given that the Bog
substantial contribution to sediments they are hydrocarbons probably migrated from Carboni-
arkoses and litharenites and reservoir quality is
ferous rocks, the likely scenario is a post-Lower
negligible as expected. However, without an
Palaeozoic exposure of the sandstones, surficial
input from volcanic and plutonic rocks, succes-
leaching of calcite to yield secondary porosity,
sive transgressions would have reworked sedi-
burial beneath Carboniferous sediments and
ment almost entirely from pre-existing shelf
then ultimately hydrocarbon migration into the
deposits to yield abundant quartz arenites. This
Llandovery sandstones. The exposure and
occurred in regions where no volcanic or plutonic
weathering of Lower Palaeozoic rocks under a
basement was exposed, e.g. the Llandovery semi-arid Devonian/Lower Carboniferous cli-
transgression reworked sediments of the Cam-
mate would have facilitated secondary porosity
brian transgression at Rubery and the Caradoc
development. In the Hoar Edge Grit, evidence
transgression in east Shropshire. Where sedi-
for pre-kaolinite secondary porosity and the
ments incorporated a moderate input of unstable
gradation from kaolinite cement to calcite
grains from volcanic and plutonic sources, the
cement suggest that a calcite cement in the
sediment composition was such as to have some
sandstones was similarly leached and replaced by
reservoir potential as discussed above.
kaolinite. Acidic meteoric waters would account
for both calcite dissolution and kaolinite precipi-
tation. Calcite dissolution by an acidic fluid
Generation of secondary porosity could cause the pH of the fluid to rise to the point
The generation of secondary porosity requires a where it is supersaturated and the mineral
fluid to leach away unstable components. The kaolinite is precipitated (Curtis 1983). Feldspar
setting of the Welsh Borderland shelf sandstones dissolution in the Shropshire sandstones could
updip from a basin of organic-rich mudrocks also be a product of freshwater leaching.
invites speculation that large volumes of fluid Reservations about freshwater leaching and
migrating from the compacting mudrocks might clay precipitation have been expressed by some
pass through the shelf sandstones and leach workers on the basis that fluids tend to migrate
them. The 20 m thick Trecoed Beds sandstones upwards in sedimentary basins (Burley et al.
pass laterally into over 1000m Llandovery 1985). However, the Carboniferous sandstones
mudrocks within 10 km distance (Ziegler et al. were deposited on a terrane of considerable
1968). Extensive secondary porosity in these relief. The distribution of hydrocarbons in the
beds was developed prior to the migration of region (Bath et al. 1986) suggests that migration
hydrocarbons from the Lower Palaeozoic black was updip from the Carboniferous into uplifted
shales. Schmidt & McDonald (1979a) suggest topographically higher Lower Palaeozoic rocks.
that secondary porosity is created particularly by Hydrocarbon seeps are still active in Carbonifer-
acidic fluids produced by CO2 generation from ous sandstones in Shropshire (Bath et al. 1986),
organic-rich sediments prior to liquid hydrocar- so it is important to emphasize that hydrocarbon
bon generation. However, Giles & Marshall migration into the Lower Palaeozoic sandstones
 Sandstone porosity, Welsh Borderland 3IX
was not recent. Hydrocarbons have been record- Conclusion
ed in a borehole through Cambrian sandstones at
over 50 m depth below drift (Parnell 1987). Some The deposits of each of the four main transgres-
hydrocarbon-bearing Lower Palaeozoic rocks sions in the Welsh Basin locally contain hydro-
occur at distances over 10 km from the nearest carbons. The composition of the sandstones was
Carboniferous outcrop. Recent or sub-recent important to reservoir potential, as their porosity
seeps have left viscous hydrocarbon residues is predominantly secondary, after alteration and
unlike the brittle solids found in Lower Palaeo- leaching of feldspars in subarkoses. Alteration in
zoic sandstones. These aspects do not absolutely arkoses was too extensive to allow the preserva-
rule out a sub-recent origin, but petrographic tion of a framework around secondary porosity.
evidence is more conclusive. Hydrocarbons in Precambrian volcanic and plutonic rocks which
the Welsh Borderlands of suspected Lower formed at the unstable southern margin of the
Palaeozoic origin have high reflectivity, low H/C Iapetus Ocean were a source of unstable grains,
ratio and low solubility in organic solvents, all particularly feldspars. However the grains were
indicative of high thermal maturity. They are made available by long-term erosion of basement
clearly distinguishable from hydrocarbons of rocks, rather than direct fall-out of volcanic
younger origin which have lower reflectivity, detritus or penecontemporaneous reworking of
higher H/C ratio and ready solubility (Bath et al. marginal basin volcanic and plutonic rocks.
1986, Parnell 1987). The hydrocarbons from the The secondary porosity was generated by
Trecoed Beds, Hollybush Sandstone and the fluids from two sources; fluids within the basin
Black Grit exhibit one or more of the features of and fluids which post-dated the basin history.
high thermal maturity. Hydrocarbons do not Limited secondary porosity, after dissolution of
migrate in this state and high maturity reflects calcite, was generated by the meteoric fluids. The
their emplacement for a geologically very long porosity was occupied by hydrocarbons from two
period. The hydrocarbons were generated during sources: organic-rich rocks within the basin and
the Lower Palaeozoic (Bath et al. 1986), and source rocks which were deposited at a much
hence secondary porosity in the Trecoed Beds later stage. Hydrocarbon migration within the
was generated before exposure to meteoric basin occurred as the basin subsided during the
fluids. Lower Palaeozoic, before secondary porosity was
Hydrocarbons of probable Carboniferous ori- generated by meteoric fluids. Hydrocarbon mi-
gin were also emplaced in the geological past. In gration from Carboniferous rocks could take
samples where some pore space is isolated by a advantage of secondary porosity of both origins.
thick pore-lining of clay, hydrocarbons do not
occur in the isolated pore space despite the fact
that sub-recent weathering has broken the pore- ACKNOWLEDGMENTS: The manuscript benefited con-
linings. The hydrocarbons were therefore em- siderably from reviews by J. D. Marshall, G. Strong
and E. K. Walton. E. Lawson and E. McKelvey
placed pre-weathering. Furthermore, some hy- provided valuable technical support. P. Eakin kindly
drocarbons occupying fracture porosity below supplied Fig. 7(c). Fieldwork was supported by the
the sub-Carboniferous unconformity are coeval Queen's University of Belfast and NERC grant
with calcite veins (Parnell 1983). F60/A1/64.

References
ANDERTON,R., BRIDGES,P., LEEDER,M. & SELLWOOD, BEVINS, R. E. & ROWBOTHAM,G. 1983. Low-grade
B. 1978. A Dynamic Stratigraphy of the British Isles. metamorphism within the Welsh sector of the
Allen & Unwin, London. paratectonic Caledonides. Geological Journal, 18,
BASSETT, M. G. 1980. The Caledonides of Wales, the 141-67.
Welsh Borderland and Central England. In: BOURGEOIS, J. 1980. A transgressive shelf sequence
OWEN,T. R. (ed). United Kingdom: Introduction to exhibiting hummocky stratification: the Cape
General Geology and Guides to Excursions 002, 005, Sebastian Sandstone (Upper Cretaceous), South-
093 & 151. Institute of Geological Sciences, western Oregon. Journal of Sedimentary Petrology,
London. 50, 681-702.
BATH, A. H., BRASSELL, S. C., EGLINTON, G., HILL, BRENNER,R. L. 1978. Sussex Sandstone of Wyoming--
R. I., HOOKER,P. J., O'NIONS, R. K., PARNELL,J.. example of Cretaceous offshore sedimentation.
ROBINSON,N. & SPIRO, B. 1986. Deep source gases American Association of Petroleum Geologists Bulle-
and hydrocarbons in the U.K. crust. Report of the tin, 62, 181-200.
Fluid Processes Unit, British Geological Survey, No. BRIDGES, P. H. 1975. The transgression of a hard
86/2. substrate shelf: the Llandovery (lower Silurian) of
 312 J. Parnell
the Welsh Borderland. Journal of Sedimentary their palaeo-oceanographic significance. Journal
Petrology, 45, 79-94. of the Geological Society of London, 137, 139-56.
BURLEY, S. D., KANTOROWICZ, J. D. & WAUGH, B. LOUCKS, R. G., DODGE, M. M. & GALLOWAY, W. E.
1985. Clastic diagenesis. In: BRENCHLEY, P. J. & 1979. Importance of secondary leached porosity in
WILLIAMS, B. P. J. (eds). Sedimentology: Recent lower Tertiary sandstones reservoir along The
Developments and Applied Aspects, pp. 189-226. Texas Gulf Coast. Transactions of the Gulf Coast
Geological Society of London Special Publication, Association Geological Society, 29, 164-7 I.
18. BlackweU Scientific Publications, Oxford. NAGTEGAAL, P. J. C. 1978. Sandstone-framework
CURTIS, C. D. 1983. Geochemical studies on develop- instability as a function of burial diagenesis.
ment and destruction of secondary porosity. In: Journal of the Geological Society of London, 135,
BROOKS, J. (ed.). Petroleum Geochemistry and 101-5.
Exploration of Europe, pp. 113-25. Geological ODIN, G. S. & MATTER, A. 1981. De glauconiarum
Society of London Special Publication, 12. Black- origine. Sedimentology, 28, 611-41.
well Scientific Publications, Oxford. PARNELL, J. 1983. The distribution of hydrocarbon
DAVIES, D. K., ALMON, W. R., BONIS,S. B. & HUNTER, minerals in the Welsh Borderlands and ajacent
B. E. 1979. Deposition and diagenesis of Tertiary- areas. Geological Journal, 18, 129-39.
Holocene volcaniclastics, Guatemala. Society of -- 1987. The occurrence of hydrocarbons in Cam-
Paleontologists and Mineralogists Special Publica- brian sandstones of the Welsh Borderland. Geolog-
tion, 26, 281-306. ical Journal, 22.
DEMAISON, G. J. & MOORE, G. T. 1980. Anoxic ROBINSON, N., EGLINTON, G., BRASSELL, S. C.,
environments and oil source bed genesis. Bulletin GOWER, A. P. & PARNELL,J. 1986. Hydrocarbon
of the American Association of Petroleum Geologists, compositions of bitumens associated with igneous
64, I 179-209. intrusions and hydrothermal deposits in Britain.
EARP, J. R. & HAINS, B. A. 1971. British Regional Organic Geochemistry, 10, 145-52.
Geology: the Welsh Borderland. HMSO, London. SCHMIDT, V. & MCDONALD, D. A. 1979a. The role of
FOLK, R. L. 1974. Petrology of Sedimentary Rocks. secondary porosity in the course of sandstone
HemphiUs, Austin, Texas. diagenesis. Society of Economic Paleontologists and
GALLOWAY, W. E. 1979. Diagenetic control of reser- Mineralogists Special Publication, 26, 175-207.
voir quality in arc-derived sandstones: implica- & --1979b. Texture and recognition of
tions for petroleum exploration. Society of Eco- secondary porosity in sandstones. Society of
nomic Paleontologists and Mineralogists Special Economic Paleontologists and Mineralogists Special
Publication, 26, 251-62. Publication, 26, 209-25.
--& HOBDAY, D. K. 1983. Terrigenous Clastic SEEMANN, U. & SCHERER, M. 1984. Volcaniclastics as
Depositional Systems. Springer-Verlag, New York. potential hydrocarbon reservoirs. Clay Minerals,
GILES, M. R. & MARSHALL,J. D. 1986. Constraints on 9, 457-70.
the development of secondary porosity in the SURDAM, R. C. & BOLES, J. R. 1979. Diagenesis of
subsurface: re-evaluation of processes. Marine and volcanic sandstones. Society of Economic Paleon-
Petroleum Geology, 3, 243-55. tologists and Mineralogists Special Publication, 26,
HEALD, M. Z. & LARESE, R. E. 1973. The significance 227-42.
of the solution of feldspar in porosity develop- SWIFT, D. J. P. 1968. Coastal erosion and transgressive
ment. Journal of Sedimentary Petrology, 49, 1217- stratigraphy. Journal of Geology, 76, 444-56.
28. THORPE, R. S., BECKINSALE,R. D., PATCHETT, P. J.,
INGERSOLL,R. V. & SUCZEK,C. A. 1979. Petrology and PIPER, J. D. A., DAVIES, G. R. & EVANS, J. A.
provenance of Neogene sand from Nicobar and 1984. Crustal growth and late Precambrian-early
Bengal fans, DSDP sites 211 and 218. Journal of Palaeozoic plate tectonic evolution of England
Sedimentary Petrology, 49, 1217-28. and Wales. Journal of the Geological Society of
JAMES, D. M. O. 1983. Observations and speculations London, 141, 521-36.
on the northeast Towy "axis", mid-Wales. Geolo- VESSELL, R. K. & DAVIES, D. K. 1981. Nonmarine
gical Journal, 18, 283-96. sedimentation in an active fore arc basin. Society
JONES, O. T. 1949. Geology of the Llandovery district of Economic Paleontologists and Mineralogists
Part 2: the northern area. Quarterly Journal of the Special Publication, 31, 31-45.
Geological Society of London, 105, 43-64. WEDD, C. B. 1982. Notes on the Ordovician rocks of
KEELING, G. 1978. The paratectonic Caledonides of Bausley, Montgomeryshire. Summary of Progress,
mainland Britain. Geological Survey of Canada Geological Survey of Great Britain, 1931 (Part 2),
Paper, 78-13, 89-95. pp. 49-55.
LEE, Y. I. & KLEIN, G. DE V. 1986. Diagenesis of WHITTARD, W. F. 1932. The stratigraphy of the
sandstones in the back-arc basins of the Western Valentian rocks of Shropshire: the Longmynd-
Pacific Ocean. Sedimentology, 33, 651-75. Shelve and Breidden outcrops. Quarterly Journal
LEGGETT, J. K. 1978. Eustacy and pelagic regimes in of the Geological Society of London, 88, 859-902.
the Iapetus Ocean during the Ordovician and ZIEGLER, A. M., COCKS,L. R. M. & MCKERROW, W. S.
Silurian. Earth and Planetary Science Letters, 41, 1968. The Llandovery transgression of the Welsh
163-9. Borderland. Palaeontology, l l , 736-82.
- - 1980. British Lower Palaeozoic black shales and

JOHN PARNELL, Department of Geology, The Queen's University of Belfast, Belfast BT7
1NN, Ireland.
 Carbonate cements in clastic reservoir rocks from offshore
Norway relationships between isotopic composition, textural
development and burial depth
G. C. Saigal & K. Bjorlykke

S U M M A R Y : Many Jurassic and Lower Cretaceous sandstone reservoir rocks from

offshore Norway contain carbonate-cemented intervals which typically make up 1-10~ of
the total sequence. Being nearly impermeable to fluid flow, their distribution may strongly
influence reservoir properties. It is therefore important to understand the genesis of these
carbonate cements.
Studies of the Jurassic and Lower Cretaceous reservoir rocks from the Haltenbanken and
Viking Graben area show that the sandstones contain dominantly two types of calcite
cements: (i) microsparry and sparry calcite (Type I) and (ii) poikilotopic calcite (Type II).
Type I calcite with 6~sO values between -4.5 and -8.4%~ PDB precipitated relatively
rapidly at shallow burial and at temperatures between 15 and 40~ possibly from porewater
of meteoric origin. Type I! calcite with hlsO values between - 1 0 and -12%o PDB have
precipitated later at temperatures corresponding to 50-60~ (at ~ 1.5 km burial depth).
Fluid inclusion data also indicate that Type II calcites have crystallized at temperatures
between 56 and 68~
Silicate grains are often corroded by Type I calcite while grains contained in Type II
calcite are uncorroded, suggesting that the porewater was in near equilibrium with the main
silicate phases (quartz and feldspar) during the precipitation of Type II calcite.
The degree of supersaturation in the porewater with respect to low Mg calcite probably
decreased with increasing burial depth as sources of more soluble carbonate (aragonite, high
Mg calcite) were exhausted, causing slower precipitation rates and increased crystal size.
The Type II (poikilotopic) calcite is thought to have formed at the expense of only slightly
more soluble low Mg calcite crystals.
The fact that the Type II calcite shows nearly the same h180 value ( - 10 to - 12%oPDB) in
sandstones ranging in present burial depth from 1.6 to 5 km suggests that this type of
carbonate cementation probably took place early (at temperatures less than 60-70~ and
that very little dissolution and reprecipitation occurred at higher temperatures (greater burial
depth).

Carbonate cemented intervals in reservoirs form that can precipitate is limited by the availability
impermeable barriers to fluid flow during hydro- of cations such as Ca ++, Mg ++ and Fe ++. Unless
carbon production. It is therefore important to significant amounts of calcium are released from
try to understand how these form diagenetically. non-carbonate minerals, very little 'new calcite'
Only then can predictions about the lateral will precipitate. The Jurassic and Lower Creta-
distribution of carbonate cemented layers en- ceous sandstones examined in the present study
countered in the cores be made. The first from the Norwegian continental shelf, contain
question to ask is: what is the origin of carbonate very little plagioclase or other possible sources of
cements in sandstones? There are three possible Ca ++ from non-carbonate minerals.
sources of c a r b o n a t e cements: (1) Biogenic The calcite cements formed during deep burial
carbonate within the sandstone sequence-- in the sandstones therefore represent redistri-
deposited as scattered fossils or enriched in lag buted carbonate which was buried with the
deposits (coquinas). (2) Early marine carbonate sandstone and adjacent shales. The initial car-
cement precipitated during periods of slow bonate distribution is facies controlled, and the
sedimentation. (3) Carbonate cement introduced later redistribution is caused by dissolution and
from outside the sandstone sequence. reprecipitation of essentially the same volume of
Very little carbonate may be precipitated from carbonate. Bjorkum (1984), in a study of the
porewaters without the dissolution of other Jurassic reservoir sandstones from the Oseberg
carbonate minerals in the sequence. Although field, argued that the precipitation of calcite
carbon is released from maturing kerogen as must have taken place in a semi-closed system.
CO2, or organic acids, the amount of carbonate He observed high Sr/Ca ratios in the poikilotopic

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 313

Geological Society Special Publication No. 36, pp. 313-324.
314 G. C. Saigal & K. Bjorlykke

•o Arctic 2's~ I
6~ 7~ g 9" 1~
/ ~;,o~
STUDY ../~,~"(% ~' (~l
Haltenbanken

Statfjord,
GulLfaks
~r
Bra(

Hil(
0 OSLO

,tavanger

50 100
I J
Km

FIG. 1. Map showing wells used in the present study. The inset shows the location of study areas from
offshore Norway.

calcite and also in samples of the formation on the work of Bjorlykke et al. (1986). This is also
water. He suggested that precipitation of poikilo- supplemented by point-count analyses of 73
topic calcite took place in porewater that was not samples examined during this study. The frame-
very mobile and that the Sr/Ca ratio of the work grain mineralogy and average bulk compo-
porewater would build up in the process. sition of Lower Cretaceous sandstones is based
In the present study we have compared on the point-count analyses of 82 thin sections.
isotopic data with textural analysis of the Approximately 250 points per slide were
different types of calcite cements present in the counted.
Jurassic (Draugen, Midgard, Gullfaks, Brage The Jurassic sandstones are feldspathic and
and Gudrun) and Lower Cretaceous (Agat) micaceous, and range from fine to coarse sand.
clastic reservoir rocks from offshore Norway The grains are subrounded to subangular and
(Fig. 1). poorly sorted. The feldspar content is typically
between 10 and 20%. Potassium feldspar is
nearly always dominant over plagioclase. Quartz
Petrology content varies between 40 and 50%, while mica
constitutes 5-10~o of the total rock. More
The average bulk composition and detrital quartzitic facies occur locally. Cements and
mineralogy of Jurassic sandstones is partly based porosity make up the remaining proportion of
 Carbonates in reservoir rocks, Norway 315

Ft~. 2. Photomicrograph showing sparry (S) calcite (Type I) cement often replacing detrital grains; the arrow
shows relict of a quartz grain. Other grains seen in this picture are of quartz. Scale bar = 200 ktm.

the rock volume. Poikilotopic calcite and diagen- between < 10 and 200 ~tm. The sandstones are
etic kaolinite are the most common cements, and normally grain supported (quartz and feldspar
are found in variable quantities in these grains) but may also show a floating texture. The
sandstones. contacts between calcite crystals are commonly
The Lower Cretaceous sandstones are of fine sharp and straight but sutured (microstylolitic)
to coarse grain size. The grains are subrounded to contacts are also seen. Microstylolitic contacts
subangular and poorly sorted. The sandstones are often more common in samples showing
contain 35-55% quartz (mono and polycrystal- floating texture. Remains of shell debris and/or
line), 6-15% feldspars (dominantly K-feldspars ghosts of shell fragments are enclosed in sparry
with some plagioclase), 3-8% rock fragments calcite cement in some samples. The calcite
(dominantly quartzitic with occasional volcanic) cement commonly partially replaces detrital
and 5-10% micas (mainly biotite with some grains (Fig. 2), and the floating texture can, to
muscovite). Cements and porosity make up the some extent, be attributed to similar replacement
remaining proportion of the rock volume. Authi- of detritals by calcite cement. Under cathodolu-
genic clay minerals (kaolinite, chlorite and illite), minescence the calcite cement shows patches of
quartz overgrowths (rare to common forming more and less bright luminescence. Individual
euhedral outlines) and calcite (dominantly poiki- sparry calcite crystals show uniform lumines-
lotopic with some microsparry to sparry) are the cence without any zoning. SEM studies show
common cements found in these sandstones. that calcite commonly occurs with authigenic
kaolinite as pore filling cement (Fig. 3). The
relative timings of calcite and kaolinite forma-
Calcite textures tion in these pores is not clear. Microsparry and
sparry calcite is hereafter referred to as Type I
The sandstones examined show dominantly two calcite.
types of calcite cement: (i) microsparry and
sparry calcite and (ii) poikilotopic calcite.
Samples may contain either one type of Type H.
calcite cement (pure), or a mixture of both in
Poikilotopic calcite is ferroan in composition
varying proportions (impure).
and makes up between 5 and 45% of the
sandstones examined. The size of poikilotopic
crystals varies between approximately 300 and
Type I. 3000 ~m. These crystals enclose a few to several
Microsparry and sparry calcite constitutes 28- detrital grains (Fig. 4). The contact between the
47~o of the total rock and its crystal size varies calcite crystals and detrital grains is usually
316 G. C. Saigal & K. Bjorlykke

FIG. 3. SEM photograph showing sparry calcite (S) and kaolinite filling a pore. Scale bar = 10 ~m.

FIG. 4. Typical poikilotopic calcite (Type II) crystal (P) enclosing several quartz grains. Note that the detrital
grains are not corroded by poikilotopic calcite. Scale bar = 500 ~tm.

sharp and, unlike sparry calcite, it (poikilotopic cite crystals. Kaolinite occurring as vermicular
calcite) seldom shows any significant replace- stacks of pseudo-hexagonal crystals and discrete
ment of detrital silicate grains (Fig. 4). Like booklets is often partly or almost completely
sparry calcite, the contacts between poikilotopic engulfed in poikilotopic calcite crystals (Fig. 5).
calcite crystals are either straight and sharp or Poikilotopic calcite is hereafter referred to as
sutured (microstylolitic). Microstylolitic type Type II calcite.
contacts are more common in samples showing a
floating texture. Under cathodoluminescence
poikilotopic calcite crystals commonly appear to Fluid inclusions
be made up of more and less bright laths oriented
parallel to each other. Often individual poikiloto- Fluid inclusions present in Type II calcite
pic calcite crystals also show uniform lumines- cement (poikilotopic) have been studied using a
cence without any zoning or laths. SEM studies Linkam TH 600 stage. Generally, fluid inclu-
show that detrital grains (including quartz and sions are rare and < 10 ~tm in diameter. Most of
feldspar overgrowths) and authigenic kaolinite the inclusions have a very tiny vapour bubble
are commonly enclosed within poikilotopic cal- (high liquid to vapour ratio) and are globular or
 Carbonates in reservoir rocks, Norway 3 I7

FIG. 5. SEM photograph showing poikilotopic calcite engulfing vermicular stacks of kaolinite. Scale
bar = 20 gm.

elongate rounded in shape. It was not possible to The 61sO values of Type I calcite cement in
record the temperature of melting of the first or Jurassic and Lower Cretaceous sandstones range
last ice crystal but the temperature of homogeni- from - 4 . 5 to -8.4%0 PDB, whereas 6180 values
zation (Th) was recorded from a total of 19 fluid of Type II calcite cement vary between - 9 . 2 and
inclusions. Out of these, eight inclusions were -12.1%o PDB (Tables 1 and 2). Samples contain-
present in the Jurassic sandstones of the Draugen ing a mixture of Types I and II calcite show
field while 11 were present in the Lower intermediate values.
Cretaceous sandstones of the Agat field. The The 613C values of Type I calcite cement
present burial depth of these Jurassic and Lower present in the Jurassic sandstones of the Halten-
Cretaceous samples is about 1680 and 3600 m, banken, Brage and Gudrun areas range from
respectively. The Th measurements were made - 0 . 9 to - 15.6%o PDB, whereas 613C values of
on a doubly polished, approximately 70 ~tm thick Type II and mixed calcite samples vary between
slice. The range of measured Th values is from 56 - 2 . 7 and -34.4%0 PDB (Table 1). In the
to 68~ but the majority of inclusions (from both Jurassic sandstones of the Gullfaks area, 613C
Jurassic and Lower Cretaceous sandstone sam- values of Type I calcite cement range from + 5.3
pies) show values between 64 and 68~ Similar to + 12.6%o PDB, whereas 613C values of Type II
values have been recorded by Olav Walderhaug and mixed calcite samples are +9.8 and +5.8%0
in poikilotopic calcite from the Haltenbanken PDB, respectively (Table 1). In the Lower
area (personal communication). Cretaceous sandstones (Agat), 613C values of
Type I calcite range from -5.1 to 1.8%o PDB,
whereas 613C values of Type II and mixed calcite
Stable isotope geochemistry samples vary between - 1 0 . 5 and 0.6%0 PDB
(Table 2).
On the basis of petrological studies, samples
containing different types of calcite cement were
selectively chosen and analysed for stable isotope Discussion
compositions. Organic-free, powdered calcite
cement from 24 samples of Jurassic sandstones Stable isotope analyses of calcite cements present
and 19 samples of Lower Cretaceous sandstones in the Jurassic and Lower Cretaceous sandstones
have been analysed for carbon and oxygen show that Type II cements consisting of coarse
isotopes. The results of the analyses are given in poikilotopic calcite crystals are characterized by
Tables 1 and 2. Petrographic and XRD studies the most negative 6180 values ( - 9 . 2 to - 12.1%o
confirmed that calcite cement comprised > 90% PDB) and that the finer sized Type I calcite have
of the carbonate material in all the samples. more positive 6180 values ( - 4 . 5 to -8.4%o
Carbon dioxide was extracted from samples by PDB, Fig. 6). The most negative 6180 values
reacting with phosphoric acid using the proce- ( - 11.1 to - 12.1%o PBD, for Type II calcite) are
dure described by McCrea (1950). found in the shallower Draugen field (in the
318 G. C. Saigal & K. Bjorlykke

TABLE 1. 61 s o and 613C jbr calcite cement types present in Jurassic sandstones

Well and Depth Calcite Average calcite ~180 ~13C

Field sample no. (m) type* Calcite % crystal size (~tm) (PDB) (PDB)

Draugen 6407/9-1
33 1669.7 I 8 120 -8.37 -15.58
18 1670.1 I + II 1--28 1--100 -9.57 - 15.03
II--5 11--450
12 1676.3 II 50 1400 -12.10 - 1 8 . 5 1

7 1681.2 II 45 2000 - 10.36 -34.36

44 1682.3 II 43 1500 -11.14 - 32.93
3 1666.3 I + II 1--18 1--180 - 10.34 - 15.94
11--12 II--650
38 1679.9 II 42 1750 -10.57 - 31.40
Midgard 6407/2-1
15 3088.9 I 45 < 15 -5.1 - 11.7
43 3495.6 I 42 <20 -4.8 -0.9
Midgard 6407/2-2
69 2501.6 II 45 1000 -10.4 -27.2
65 2510.1 II 47 800 -9.8 -21.1
Gullfaks 34/10-3
10 1930.9 I 55 100 -6.24 12.56
11 1942.1 I 40 200 -6.38 11.33
34/10-4
12 1852.00 I + II 1--28 1--150 -5.08 5.81
II--10 11--400
13 1863.70 I 40 200 - 5.80 5.33
14 2351.30 II 40 2500 -9.22 9.84
Brage 31/4-2
7 2222.76 I 45 150 -4.52 -8.55
8 2494.30 II + I II--20 I1--1200 - 11.46 -9.11
1--15 1-200
Gudrun 15/3-3
1 4262.95 II 42 2200 -9.61 -2.69
2 4268.10 II 40 1600 -10.87 - 6.40
3 4268.28 II 35 1300 -10.45 - 3.66

4 4269.20 I + II I 15 1--150 - 11.53 -5.77

II--7 11--1200
5 5000.70 II 14 1500 - 11.57 -8.89
15/9-3
6 3523.40 II 45 ll00 -9.91 -8.20

* I--microsparry and sparry calcite, II--poikilotopic calcite.

H a l t e n b a n k e n region) at about 1680 m, w h i c h at together suggest that the heaviest water from
present is at its m a x i m u m depth of burial, and w h i c h Type II calcite in the D r a u g e n field could
has a bottom hole t e m p e r a t u r e (at 1700 m) of have precipitated had 6180 = -4%o ( S M O W )
about 60~ The H a l t e n b a n k e n region experi- (Bjorlykke et al. 1986 and Fig. 7). Therefore,
enced rapid subsidence and s e d i m e n t a t i o n in late Type II calcite in the D r a u g e n field either
Tertiary and Q u a t e r n a r y time (Ellenor & Moze- precipitated from porewater near present reser-
tic 1986), totalling about 1000 m, and at present voir conditions or from more negative (less than
is exposed to the m a x i m u m burial d e p t h and -4%0 gi~so S M O W ) formation water at lower
t e m p e r a t u r e yet experienced. The bottom hole (<60~ temperatures. Precipitation from
t e m p e r a t u r e (60~ can therefore be t a k e n to be porewater with a more negative oxygen isotope
the m a x i m u m t e m p e r a t u r e of crystallization for composition would imply that it took place
Type II calcite. Hence, the most negative 6~80 during meteoric water flushing.
value ( - 1 2 % o PDB) of Type II calcite and However, since Type II calcite can be shown
m a x i m u m t e m p e r a t u r e of crystallization (60~ to form late and clearly post-date the formation
 Carbonates in reservoir rocks, Norway 3 I9

TABLE 2. 5180 and 6 ~3C for calcite cement types present in Lower Cretaceous sandstones

Well and Depth Calcite Average calcite 6180 (~13 C

Field sample no. (m) type* Calcite % crystal size ~ m ) (PDB) (PDB)

Agat 35/3-2
7 3614. I II 9 2000 - I 1.75 -2.14
II 3638.5 I 45 < 20 - 6.16 -0.56
13 3698.3 II 25 1750 - I 1.50 - 10.48
19 3950.0 II I0 1500 - I 1.38 - 10.26
22 4008.7 I 47 200 - 8.27 -5.14
35/3-4
I 3403.3 I 47 < 20 - 5.42 1.55

5 3448.7 I 45 150 - 6. I 0 1.46

I0 3470. I I 40 80 - 6.73 -0.015

15 3500.3 ]I 20 2000 - 9.61 - 7.54
35/3-5I-
1 3230.2 I 38 200 - 8.03 -0.26
8 3240.3 II + I II--20 11--1200 -7.60 - 1.86
1--13 1--220
11 3253 I + II I--6 1--50 - 8.65 -0.93
I1--3 11--700
14 3255.2 II 35 2250 - 9.37 -0.91
18 3265.2 I + II 1--30 1--150 -7.55 -0.12
II--15 II--1150
23 3272.1 II 5 1100 - 11.78 -1.13
29 3282.1 II 8 900 - 10.81 - 1.26

31 3284.6 I + II 1--30 1--250 -7.93 0.65

II--10 11--800
34 3290.4 I 48 < 20 - 4.80 1.81
37 3292.2 I + II 1--8 1--200 - 7.72 0.10
1I--4 11--800

* I--microsparry and sparry calcite, II--poikilotopic calcite.

i 35/3-5--dry well.

2~00
Ui

2000 9 9

1600
m N
._
9 9

1200

u ,, 9 mm 9
ii .

~9 800 ',:iT 9
u
:I:',; [
,If:
400 ,t t i
,i 1 ~ i
9 9 ooo Y-'-o-I
9 9 ' o', I
9 ~ 1 9 , , J i , , b h i

-2 -4 -6 -8 -10 -12 -14

18
6 0 (PDB)

FIG. 6. Plot of oxygen isotope ratio against calcite crystal size for all samples from Jurassic and Lower
Cretaceous sandstones. Circles and squares represent Types I and II calcite, respectively. Broken lines
represent impure samples containing both Types I and II calcite cements.
32o G. C. Saigal & K. Bjorlykke
160
1500
9 ~ll i 9
Iz.O

120 2000

100
(1.)
L
2500
8o

Q- 60
E
3000
~-- 40
~11 A 9
9 9 9
3500 9
~ : n g e in c~t,cites (Types T & EI )
0 -2 -4 -S -8 -10 -12 -14

6 mO Catcite (PDB) 4000

FIG. 7. Temperature dependence of the calcite-water 9 9 9 9

isotopic equilibrium (after Bjorlykke et al. 1986). The

4500
dark bar represents the observed range of calcite
cements (Types I and II) for Jurassic and Lower
Cretaceous sandstones examined in this study.
5000

-2 -4 -~ -'8 -lo -12 -14

of kaolinite (Fig. 5) we consider the Type II
calcite to have formed near present-day condi- 8 ~80 (PDB)
tions and may continue to form at present right
below the oil/water contact. It is probable that FIG. 8. Plot of oxygen isotope ratio versus depth for
most of the Type I calcite formed during all samples from Jurassic and Lower Cretaceous
sandstones. Circles and squares represent Types I
meteoric water flushing at lower temperatures. and II calcite, respectively. Triangles represent
Oxygen isotope values for Type I calcite as impure samples containing both Types I and II
negative as - 8%~(PDB) would at 20-30~ imply calcite cements. Note that Type II calcite crystals
porewater compositions of about - 4 to -6%o (squares) show similar 6180 values ( - 10 to - 12%o
(SMOW) (Fig. 7). PDB) at all depths between 1600 and 5000 m.
In the deeper wells of the Haltenbanken
(4300 m) and Hild fields (4-5 km) analyses of
formation water gives more positive values:
~51sO= +1.3 and +2.5 to +3.5%0 SMOW, composition of the formation water would have
respectively (Bjorlykke et al. 1986 and Lenoy et been more negative and the calculated tempera-
al. 1986). The porewater probably experienced ture of crystallization would have been lower.
61sO enrichment by isotopic exchange with Calcite cemented sandstones buried to a much
minerals during burial diagenesis (Craig et al. greater depth (4-5 km) analysed in this study do
1956). The rate at which porewater becomes not show 6 ~sO values below - 12%0PDB (Fig. 8),
more positive at depth does not quite correspond indicating that the formation of Type II calcite
to the temperature-dependent fractionation of was probably limited to a narrow temperature
61sO between porewater and calcite. For exam- range (50-60~ assuming 6180 of porewater to
ple, in the Haltenbanken area the present be -4%o SMOW) (Figs 9 and 10). Irwin & Hurst
formation water with 6180 = + 1.3%o SMOW at (1983), analysing carbonate cement from Juras-
4300m gives a maximum temperature of sic sandstones from the North Sea, also found
crystallization for poikilotopic calcite that poikilotopic ferroan calcite had 61sO values
(61so = - 11.75%o P D B ) o f about 95~ while the mostly between - 1 0 and -12%o PDB. Similar
bottom hole temperature is nearly 120~ Simi- values for calcite cement were observed by
larly, in the Hild field the formation water with Lonoy et al. (1986) from the Hild field which has
6lsO = +3.5%0 SMOW gives a maximum tem- been buried relatively deeply (4-5 km) in the
perature of crystallization for poikilotopic calcite Viking Graben. In the Oseberg field, however,
( 6 ~ s O = - 1 1 to - 1 4 ~ o P D B ) o f about 120~ the calcite cements have 6180 values of about
(Leney et al. 1986) while the bottom hole - 14 to - 15%o PDB (Bjerkum 1984). We do not
temperature is nearly 150~ If the Type II understand the reason why the poikilotopic
calcite formed at shallower depth then isotopic calcite cement in sandstones from the Hild and
 C a r b o n a t e s in reservoir rocks, N o r w a y 321

-12

-10

-8 40
T ~
6 180
(POB)

-4 2O
0 9 , , i I i , I

10 -14 -18 -22 -26 -30 -34

G !3C (PDB)

FIG. 9. Plot of oxygen isotope ratio against carbon isotope ratio for Haltenbanken samples. Circles and
squares represent Types I and II calcite, respectively. Triangles represent impure samples containing both
Types I and II calcite cements.

-12

10

40
T ~
6 180
(PDB)

20

0 -2 -4 -6 -8 -10 -12

6 13C (PDB)

FIG. 10. Plot of oxygen isotope ratio against carbon isotope ratio for Agat samples. Circles and squares
represent Types I and II calcite, respectively. Triangles represent impure samples containing both Types I and
II calcite cements. The temperature scale has been determined using 61sO of water = -4%~ (SMOW).

Oseberg fields have more negative 6lsO values (i) The Type II calcite precipitated at shallow
than the samples analysed in the present study. depths (1500-1700 m) did not dissolve, and
The temperature and burial depth we infer remained stable on burial.
from the ~5~sOvalue of the Type II calcite cement (ii) The Type II calcite did dissolve and repre-
depend on the porewater composition at the time cipitated at greater burial depths but the
of crystallization. Earlier work and available porewater composition changed towards
data of present-day formation water show that more positive values in such a way that the
the porewater becomes isotopically heavier with 6180 values of the calcite precipitated
increasing burial (Craig et al. 1956, Bjorkum remained the same despite the increasing
1984, Bjorlykke et al. 1986 and Loney et al. 1986). temperature.
The fact that the oxygen isotope composition of
Type II (poikilotopic) calcite cement does not Alternative (ii) seems less probable as the
become more negative with increasing burial can published and available preliminary data on the
theoretically be explained in two ways: porewater compositions indicate that the 6180
322 G. C. Saigal & K. Bjorlykke
versus depth gradient is not sufficiently steep to remained below the oil-water contact. This
produce calcite cement with the same 5 ' 8 0 trapped meteoric water was not flushed by
values at different depths. Also the fluid inclu- deeper compactional water which would have
sion data indicate that crystallization of Type II had more positive 6180 values. The initial 6180
calcite took place at temperatures between 56 composition of the trapped meteoric water may
and 68~ (between approximately 1.5 and 2 km also depend on when the structures were sealed
burial depth). The negative 6180 values found in off from meteoric water influx. The 51sO
the porewater of shallow reservoirs show that composition of the trapped waters may also
meteoric water was trapped in the structures and change during burial and hence produce carbon-
ate cements having different 51sO values than
those formed from otherwise open porewaters at
2400 the same temperatures.
Samples of Type I calcite with smaller crystals
(<200 ~tm) all have 6180 values that are signifi-
2000 cantly more positive ( - 4 . 5 to -8.4%0 PDB)
indicating that they have formed at somewhat
3
1600
lower temperatures, between 15 and 40~ (as-
.N
_
U3
suming 61s o of formation water = - 4 to - 6%0
SMOW). Within this range the carbonate
ta
1200 cement with the smallest crystals tends to have
the most positive 5180 value (-4.5%o PDB) and
800 are therefore probably formed at lower tempera-
m tures (about 15~
The Type I calcite cements from the Gullfaks
400 area with 513C values between +5.3 and
+ 12.6%o (PDB) and 6180 values of about -6%0
0 " eA l 9 , ,
PDB (Table 1) were produced in a bacterial
-10 -14 -18 -22 -26 -30 -34 fermentation zone (Irwin & Hurst 1983). Type II
6 13C (PDB)
calcite cement from the same well (Gullfaks,
sample no. 14) shows a similar 613C composition
FIG. 11. Plot of carbon isotope ratio against calcite (9.84%0 PDB) but a comparatively lighter 51sO
crystal size for Haltenbanken samples. Circles and value, -9.22%0 PDB. This 51 sO value indicates
squares represent Types I and II calcite, respectively. formation at a somewhat higher temperature,
Triangles represent impure samples containing both probably below the bacterial fermentation zone.
Types I and II calcite cements. The positive ,~' 3C value in this sample of Type II

2400

200C

1600
_N

'~
:,,.,
1200
o
5 800
t~
s

~00
A A ~e

0
2 0 -2 -z. -6 -8 -10 -12
~3
5 C (PDB)

FIG. I2. Plot of carbon isotope ratio against calcite crystal size for Agat samples. Circles and squares
represent Types I and II calcite, respectively. Triangles represent impure samples containing both Types I and
II calcite cements.
 Carbonates in reservoir rocks, Norway 323
calcite may have been inherited from dissolving Norway is composed of two main types:
calcite precipitated in the bacterial fermentation (1) Microsparry and sparry calcite cement (Type
zone. I) with crystals smaller than 0.2 mm and
Type II calcite from the Haltenbanken and 6180 values between - 4.5 and - 8.4%o PDB.
Agat areas shows comparatively lighter carbon The least negative values are associated with
(613C from - 0 . 9 to -34.4%0 PDB, Tables 1 and finer crystalline cement indicating precipita-
2 and Figs 11 and 12), suggesting an influence tion at low temperatures (about 15~ and
from organically derived carbon (CO2) derived shallow burial, possibly in porewater of
from the source rock (Irwin et al. 1977) probably meteoric origin. The coarser crystals have
at the time of oil migration. The Type II calcite more negative 6180 values and may have
cement from the Haltenbanken area with 613C of precipitated at temperatures up to 40~
about -34%0 PDB can not have come from (2) Poikilotopic calcite cement (Type II) with
decarboxylation, but may have formed by disso- 6180 values between - 9 . 2 and -12.1%o
lution of calcite precipitated at very shallow PDB may have precipitated at about 1.5 km
depth under the influence of oxidation of of burial depth from isotopically light
methane (Irwin & Hurst 1983). In the dry well of porewater (6180 = -4%~ SMOW) at tem-
the Agat region (35/3-5), there is less influence peratures corresponding to 50-60~
from negative carbon in the Type II calcite
The fluid inclusion data indicate that crystalliza-
cement than in wells with hydrocarbons (Table
tion of Type II calcite took place at temperatures
2). Thus it is suggested that the early carbonate between 56 and 68~ This includes the deeply
cement (Type I) was dissolved by formation buried samples that are at present at much
water rich in organic acids associated with oil
greater temperatures.
migration, and that simultaneous precipitation In our study, sandstones buried to greater
of Type II calcite took place. The difference in depths (4-5 km) show no isotopic evidence of
solubility of the Types I and II calcite was late carbonate cementation.
probably very small, producing a very low degree
The Type I calcite cements probably precipi-
of supersaturation and a very slow precipitation tated relatively rapidly from porewater super-
of coarse poikilotopic calcite (Type II). Similar
saturated by dissolution of more soluble phases
relationships are also found in the Jurassic
such as aragonite and possibly also high Mg
sandstones of the Oseberg field (Bjorkum & calcite. This cementation was associated with
Bj~rlykke, personal communication).
corrosion of silicate minerals.
During the formation of Type II calcite, the
The Type II calcite cements precipitated
porewater was in near equilibrium with silicate slowly from porewater, a process controlled by
phases as there is little evidence of their the dissolution of the only slightly more soluble
corrosion by the Type II cement. However, in the low Mg calcite phases. The Type II calcite
finer sized Type I calcite-cemented sandstone, cements do not seem to have corroded silicate
there is evidence of corrosion of feldspar and minerals, suggesting that the porewater then was
quartz. Type I calcite cement is believed to have in near equilibrium with the silicate minerals.
formed by relatively rapid precipitation at
numerous nucleation sites from solution with ACKNOWLEDGMENTS: This research has been sup-
a higher (i.e. more than Type II calcite precipi- ported by VISTA, a research cooperation between
tating solutions) degree of supersaturation. This the Norwegian Academy of Science and Letters and
highly super-saturated porewater was probably Den Norske Stats Oljeselskap (Statoil). Support from
produced due to dissolution of more soluble Norwegian Research Councils (NAVF and NTNF) is
also appreciated. G. C. Saigal acknowledges the
phases such as aragonite and possibly also high receipt of post-doctoral fellowship from NTNF and is
Mg calcite. Corrosion of silicate grains by Type I grateful to the Oil and Natural Gas Commission of
calcite cement suggests that porewater from India for granting leave of absence. We also thank
which Type I calcite precipitated was not in Arco, Saga, Shell, BP, Norsk Hydro and Conoco for
equilibrium with silicate phases. Burley & their cooperation in providing samples used in this
Kantorowicz (1986) also observed that finer study. The isotopic analyses have been carried out in
carbonate crystals are more corrosive than the Mass Spectrographic Laboratory at the Depart-
coarsely crystalline spar or poikilotopic ment of Geology, University of Bergen. Fluid inclusion
carbonates. studies were carried out in the Department of Geology,
St Andrews University, Scotland and we thank Judy
Kinnaird for her help during this work. Thanks are also
due to Berit Leken Berg for her help with SEM studies.
Conclusions We thank Eli B. Bjerke for typing the manuscript.
Constructive reviews by S. Rainey, J. Gluyas, D.
Carbonate cement in the Jurassic and Lower Stow and J. D. Marshall were very beneficial to the
Cretaceous clastic reservoir rocks from offshore final manuscript.
324 G. C. Saigal & K. Bjorlykke
References
BJORKUM, P. A. 1984. Et studium av de mellomjuras- discovery. In: SPENCER,A. M. et al. (eds). Habitat
siske sandsteiner: kjerne 30/6-70seberg, med of Hydrocarbons on the Norwegian Continental
s~erlig vekt p~ diagenese. Hovedoppgave. MSc Shelj; pp. 313-6. Graham & Trotman, London.
Thesis. University of Bergen. IRWIN, H., CURTIS, C. • COLEMAN, M. 1977. Isotopic
BJORLYKKE,K., AAGAARD,P., DYPVIK, H., HASTINGS, evidence for source of diagenetic carbonates
D. S. & HARPER, A. S. 1986. Diagenesis and formed during burial of organic-rich sediments.
reservoir properties of Jurassic sandstones from Nature, 269, 209-13.
the Haltenbanken area, offshore mid Norway. In: --& HURST, A. 1983. Application of geochemistry
SPENCER, A. M. et al. (eds). Habitat of Hydrocar- to sandstone reservoir studies. In: BROOKS,J. (ed.).
bons on the Norwegian Continental Shelf, pp. 275- Petroleum Geochemistry and Exploration of Europe,
86. Graham & Trotman, London. pp. 127-46. Geological Society of London Special
BURLEY, S. D. & KANTOROWICZ, J. D. 1986. Thin Publication, 12. Blackwell Scientific Publications,
section and S.E.M. textural criteria for the Oxford.
recognition of cement-dissolution porosity in LONOY, A., AKSELSEN, J. & RONNING, K. 1986.
sandstones. Sedimentology, 33, 587-604. Diagenesis of a deeply buried sandstone reservoir:
CRAIG, H., BOATO, G. & WHITE, D. E. 1956. Isotopic Hild field northern north sea. Clay Minerals, 21,
Geochemistry of Thermal Waters, pp. 29-38. 497-511.
National Academy of Sciences, National Re- MCCREA, J. M. 1950. On the isotope chemistry of
search Council Publication, 400. carbonates and a paleotemperature scale. Journal
ELLENOR, D. W. & MOZETIC, A. 1986. The Draugen oil of Chemical Physics, 18, 849-57.

GIRISH C. SAIGAL, KDM Institute of Petroleum Exploration, Oil and Natural Gas
Commission, Kaulagarh Road, Dehradun--248 195, India. Present address: Institutt
for geologi, Universitetet i Oslo, Oslo, Norway.
KNUT BJORLYKKE, Institutt for geologi, Universitetet i Oslo, Oslo, Norway.
 Aspects of the diagenesis of the Sherwood Sandstones of the
Wessex Basin and their influence on reservoir characteristics

G. E. Strong & A. E. Milodowski

S U M M A R Y : The Sherwood Sandstone Group sandstones of the Wessex Basin are

important potential reservoirs for both hydrocarbons and geothermal brines. Analyses of
conglomerate, sandstone and siltstone samples from outcrop and deep boreholes indicate a
complex diagenetic history that has had important positive and negative effects on the
reservoir properties. Over most of the basin, early diagenesis (eodiagenesis) is dominated by
calcrete development but, towards the basin centre, non-ferroan dolomite associated with
early evaporitic sulphate cements that appear to have been deposited in an inland sabkha or
playa environment are important. Later diagenesis (mesodiagenesis) is characterized by
selective framework grain dissolution of feldspars, and anhydrite cementation and
subsequent dissolution--processes that have locally yielded significant secondary porosity.
However, in some cases this secondary porosity has been destroyed by the precipitation of
late manganiferous ferroan calcite and ferroan dolomite. Quartz cements are locally
important. Rocks near the present-day outcrop have been affected by weathering processes
(telodiagenesis) that have resulted in decalcification, and the precipitation of kaolinite, illite
and iron hydroxides. The key factors determining the permeability of the sandstones are the
original grain size and degree of sorting, the precipitation and subsequent removal of
anhydrite, and the extent of early and late carbonate cementation, framework grain
dissolution and overgrowth.

Petrological investigations of the Sherwood thin sections and where possible they were
Sandstone Group sandstones of the Wessex prepared from core plugs that had been selected
Basin were undertaken to identify primary and for determination of porosity and permeability
diagenetic features for correlation with hydraulic values. The samples were also studied by SEM as
properties, as part of the investigation of the rough stubs suitable for analysis by secondary
geothermal potential of the basin by the British electron imaging. SEM identification of min-
Geological Survey for the Department of Energy erals was aided by qualitative energy-dispersive
(Milodowski et al. 1986). The diagenesis of the X-ray microanalysis (EDS). Selected samples
Sherwood Sandstone in the Wessex Basin has were also analysed by electron microprobe and
been discussed previously (Lott & Strong 1982, by X-ray diffraction analysis (XRD).
Burley 1984, Knox et al. 1984, Burley &
Kantorowicz 1986), based on material from
outcrop and from the Winterborne Kingston and Regional setting
Marchwood boreholes. For this study, material
from a wider number of borehole localities was The lower Triassic Sherwood Sandstone Group
available for analysis, resulting in significant (Warrington et al. 1980) in the Wessex Basin
new information regarding the diagenesis of this consists mainly of fluvial sediments dominated
formation. Samples were taken from the outcrop by feldspathic sandstones, but lithologies range
at the western margin of the basin (east Devon to from conglomerates (lithic and intraformational)
west Somerset) and from cores and drill-chip- to ferruginous mudstones and siltstones. At
pings from deep boreholes at Southampton outcrop in Devon and west Somerset, the
(Marchwood and Western Esplanade), Seabor- Sherwood Sandstone consists of a lower conglo-
ough, Nettlecombe, Winterborne Kingston, meratic unit (Budleigh Salterton Pebble Beds)
Wytch Farm, Shrewton and two confidential overlain by thick cross-bedded sandstones (Otter
wells. Figure 1 shows the isopachytes of the Sandstone). In east Devon the sediments are only
Sherwood Sandstone Group for the Wessex weakly cemented with patchy calcite cements,
Basin and locations of non-confidential bore- and the pebble beds are dominated by pebbles
holes. In total, 122 samples were examined by and cobbles of quartz and quartzite with subordi-
petrological microscope, cathodoluminescence nate slate and igneous rocks. North of Welling-
(CL), and scanning electron microscope (SEM) ton the sediments become increasingly calcar-
using backscattered electron imaging techniques eous and well cemented and the Budleigh
(BSI). The samples were examined as polished Salterton Pebble Beds in particular contain so

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentao' Sequences, 325

Geological Society Special Publication No. 36, pp. 325-337.
326 G. E. Strong & A. E. M i l o d o w s k i

~ .~:;YtoC~ne~*DEVIZES 1 ~" KINGSCLERE 1 Speculative , ~ h y ~ s

--50-- for Sherwood Sondstor~e

v• ~(Q.
[] Palaeozoic basement
~ Inferred subsurface limits of
~ Sherwood Sandstone Group

s,lts, ~s lot._ -T,~,'e___~r Triossic sandst. . . . . t of this

ones 'e~//.~WESTERN~(~j" ~ b o u n d a r y are probably the
or~ rn,,~ " E S P L A N A D E ~ - / . same age as, and a
t:~ ~ ~ ~ ~ 'U%to~ - \ ~',~. marginal facies of, the

I::~, / / ~ ~ ~ ~-- "- o, ~f ~ 9 '< . . . . . . . sg~\"~ A ~ ' ( - ' : wessex Das,n

[~'~ / _-MARS0.WOOD
24U~ '. 1" I ,7o~ 9WINTERBORNE ":" ~So/,- \, \ ~ .,er.~k.-~%,-2~ K, ..a/
u~::it ./ / ./ "NFTTLI:(~OmBE 1/ KINGSTON\l\ )~r _ "' \'~'~ ~ " " ~ "~"

FIG. 1. Isopachytes for the Sherwood Sandstone Group of the Wessex Basin (after Milodowski et al. 1986).

much detrital carbonate that locally the deposits Schmidt & McDonald (1979a), according to
have been worked for lime. The limestone whether diagenetic reactions are controlled by
pebbles and cobbles have been identified as near-surface environments prior to burial (eodia-
Devonian and Carboniferous limestones (Ed- genesis), by conditions prevailing during burial
monds & Williams 1985) and are presumably (mesodiagenesis) or by near-surface processes
derived from the nearby Quantock Hills or following uplift of the sediments (telodiagenesis).
Mendip areas. In the subsurface this division In general, eodiagenetic and early mesodiagene-
into a conglomeratic lower unit and a sandy tic changes have tended to reduce the primary
upper unit can be recognized, although the intergranular pore space through cementation,
conglomeratic unit is not present everywhere, syntaxial grain overgrowths, early compaction
probably because of erosion. and pressure solution. Later mesodiagenetic
In general the sequence consists of a series of processes caused the removal of earlier cement
fining-upward cycles about 1-5 m thick; each and the dissolution of some framework grains,
cycle has an erosional base, above which a resulting in the production of a secondary
conglomerate grades upwards through coarse porosity which has had a major influence on the
cross-bedded sandstone into laminated siltstone aquifer properties of the sediments. In some
and mudstone. Concretionary calcites may be cases, however, the secondary porosity has been
developed towards the tops of cycles, producing in turn reduced by the precipitation of late
mottled and irregularly nodular horizons that are carbonates. Telodiagenesis affects only the sedi-
interpreted as calcretes. Many of the carbonate ments near the outcrop where the influx of
clasts seen in the intraformational conglomerates meteoric water (low pH) has locally resulted in
closely resemble these calcretes in microfabric the dissolution of carbonate cements (particu-
and are considered to be derived penecontem- larly in the southern part of the outcrop),
poraneously by erosion of earlier material. The dissolution of feldspar grains and precipitation
sediments may be cemented by carbonate (gener- of kaolinite and illite. An outline of the temporal
ally calcite, but dolomite and ferroan dolomite relationships of the diagenetic events is shown in
may occur as replacement cements), sulphate Table 1, and the distribution of authigenic
(mainly anhydrite, minor gypsum), or less com- phases is given in Table 2.
monly silica, K-feldspar or albite. At outcrop
limonitic cements are also important.
Eodiagenesis
Diagenesis Iron oxide surface veneers
The sediments have been subject to diagenetic The earliest diagenetic product preserved ap-
processes which have modified both the detrital pears to be a fine veneer of reddish-brown or
grains and the pore space. The diagenetic black iron oxide on many grain surfaces which is
regimes encountered can be classified, following ubiquitous throughout the area of study. SEM
 Diagenesis o f S h e r w o o d Sandstones, W e s s e x Basin 327

TABLE 1. Temporal relationships of the diagenetic erents in the Sherwood Sandstone of the Wessex Basin
>-
Z
ua < ~- ua 9
Z
I~ Or] ~ N O0 <
I~ uJ 0 rr"

r = z < o = z < < _S = ,,, ~:o

LL oV-

EODIAGENESIS
-I Ii 5
5
o=
~T
oo I
I
(J ,,, I
n- Z I- I
%1
~J I
Q3 ~-
m
Z
MESODIA-
GENESIS e Early uJ
Cretaceous r
c)
Z
i

1
? Eocene
! + I

GENESIS

I MAJOR CEMENTS I--1 MINOR CEMENTS

Q HYDROCARBON MIGRATION IIINVASION BY RECENT M E T E O R I C

AT WYTCH FARM BB WATERS AT OUTCROP

TABLE 2. Distribution of authigenic phases in the Sherwood Sandstone of the Wessex Basin
LOCALITY He At Ba An
( I ) ( l l An
) Ct(I) Ct(ll) 0o(I) ~ [
Oz(I)([IQ~ Oh K, Ab Ka ,l Mn Fe
/
W. SOMERSET
OUTCROP 9 9 9 9 9 9 9 O 9 9 9

E. DEVON OUTCROP 9 9 9 9 9 9 9 O 9 O 9

MARCHWOOD

WESTERN
ESPLANADE

SHREWTON

VVYTC H FARM 9 9 9 9 O 9 9
WlNTERBORNE 9 9 -~ 9 9 9 9 9 9 9
KINGSTON
NETFLECOMBE,
SEABOROUGH

He: hematite, At: anatase, Ba: baryte, An(I): eodiagenetic anhydrite, An(ll mesodiagenetic anhydrite,
Ct(I): early calcite, Ct(ll): later calcite, Do(I): early non-ferroen dolomite. Do(ll): later ferroan dolomite,
Qz(I): early quartz, Qz(ll): later quartz, Ch: chalcedony, KP. K-feldspar, Ab: albite, Ka: kaolinite,
I1: iltite, Mn: Mn-oxide/hydroxide, Fe: late Fe-oxide/hydroxide

observations reveal that this veneer is composed Some ferromagnesian grains have been replaced
of tiny (1 ~tm or less) plates, probably of by fine haematite; in particular biotites com-
haematite, and its absence at points of grain monly exhibit haematite plates developed along
contact indicates that it formed after deposition. cleavage planes (Fig. 2a). Some interstitial clays
328 G. E. Strong & A. E. Milodowski

FIG. 2. (a) Haematite plates (bright) replacing biotite grain deformed by compaction. Marchwood borehole,
1680 m. (b) Open framework sandstone cemented by early calcite (C~) with later calcite (C2) filling remaining
porosity in K-feldspar grain. Quartz (Q), K-feldspar (F). Marchwood borehole, 1695 m. (c) Stellate calcite (C1)
with early iron oxide films (Y) and expanded grain framework. Road cutting, La~gford Budville (NGR ST
1235 2218). (d) Euhedral terminations of early non-ferroan dolomite (D1) into open pore space. Quartz
overgrowth absent at (Y), but present at (X) against later ferroan dolomite (D2) and open pore space.
Winterborne Kingston borehole, 2429 m. (e) Euhedral quartz overgrowths (X) developed into pore space, and
later enclosed by barytes cement (B) and late calcite (C,). Both barytes and calcite show some replacement of
adjacent quartz. Matchwood borehole, 1696 m. (f) Authigenic K-feldspar Cadularia') coating grain surface of
detrital K-feldspar. Marchwood borehole, 1686 m. (a, b, d, e)--SEM backscattered electron micrographs;
(c)--plane polarized light; (f)--SEM secondary micrograph.

also show partial replacement by haematite. Earl)' calcite cementation

These features are characteristic of eodiagenesis Early cementation by calcite provides a major
in present-day sediments in hot arid or semi-arid cement in many parts of the Sherwood Sand-
environments (e.g. Walker et al. 1978). stone. Optically, most of the calcite appears as
 Diagenesis o f Sherwood Sandstones, Wessex Basin 329

neomorphic near-equigranular xenotopic spar, grown by syntaxial late ferroan dolomite. Early
although calcite intraclasts, some rim cements dolomite cut by luminescent late calcite-filled
and neomorphic stellate or radial-fibrous struc- fractures was also noted in a confidential
tures (cf. Bathurst 1976, p. 484) can be seen (Fig. borehole 15 km west of Winterborne Kingston.
2c). Stellate or radial-fibrous calcite is well The early dolomite is believed to have precipi-
developed in the Otter Sandstone at outcrop and tated before significant burial of the sediment,
in samples from Marchwood and Western for dolomite-cemented sandstones show little
Esplanade. It is coarser than the rim cement and evidence of compaction. The early dolomite
often disrupts the surrounding detrital grain appears to be restricted to sediments from
fabric (indicating growth under minimal over- boreholes near the basin depocentre (e.g. Win-
burden), and some stellate structures show terborne Kingston, Nettlecombe, Seaborough).
concentric fine banding by iron oxide (goethite
or haematite) which presumably precipitated
during intermittent calcite growth in a manner
Earl), anhydrite
similar to the haematite coatings described pre- Some siltstones and mudstones from boreholes at
viously. SEM, EDS and cathodoluminescence Nettlecombe, Seaborough and a confidential
analyses reveal that the early calcites occur as borehole to the NW of Winterborne Kingston
non-ferroan, non-luminescent or faintly lumines- contain anhydrite that occurs as felted masses of
cent (often with fine, bright multiple zoning) subparallel or radiating laths or irregular no-
rims around detrital grains. They are commonly dules. The anhydrite has clearly disrupted and
best developed around carbonate intraclasts expanded the detrital fabric of the sediments,
(Fig. 3a), which appear to have acted as and it is usually intimately associated with
nucleation sites. In many cases these cements nodular non-ferroan dolomite, suggesting a near-
appear to have been irregularly and repeatedly surface evaporitic origin. The textures are simi-
fractured producing a micropseudo-breccia tex- lar to those of sabkha sediments described by
ture with the microfractures filled by calcite of Shearman (1966), Kinsman (1969) and Dean et
slightly different luminescence. Early calcite al. (1975). Anhydritic siltstones and mudstones
cementation is usually patchy and cemented dominate the lower parts of the Sherwood
areas may coalesce into nodular horizons, but in Sandstone Group sequences immediately above
many cases it is completely pore-filling, produc- the ?Budleigh Salterton Pebble Beds of the
ing a rock with an uncompacted 'open' frame- Nettlecombe 1 and Seaborough 1 boreholes,
work texture (Fig. 2b). These early cement located close to the basin depocentre, and
fabrics are typical of calcretes. We have found no gypsiferous sandstones and shales are recorded
evidence of dissolution of either detrital carbon- above ?Budleigh Salterton Pebble Beds in the
ate intraclasts ('cornstones') or early calcite Marchwood 1 borehole.
cements (except at outcrop where affected by
telodiagenesis). Early rim cements very com- Mesodiagenesis
monly show euhedral terminations into open
pore space, thereby precluding calcite dissolu-
tion as a mechanism for secondary porosity
Burial history
generation away from the outcrop. The burial history of the Sherwood Sandstone in
the Wessex Basin is discussed in detail by
Milodowski et al. (1986). At Winterborne Kings-
Early dolomite
ton, maximum depths of about 2.5 km may have
In the Winterborne Kingston borehole early been reached at the end of the Jurassic, and also
calcite appears to be absent and an early (pre- at the end of the Cretaceous. Maximum depths
quartz overgrowth) non-ferroan dolomite re- for Marchwood and Southampton (Western
places detrital cornstone and siltstone clasts. Esplanade) are estimated at 1.8 km (early
This dolomite is found only in .the lower cored Oligocene), and Wytch Farm about 1.9 km (end
section (2421.44-2436.65 m), where it is the of Cretaceous).
major cement phase in cemented horizons. It
occurs as idiotopic rhombs completely replacing
the intraclasts, and locally it replaces quartz and
Earl), quartz and K-feldspar syntaxial
feldspar. Limited overgrowths of dolomite may
overgrowths
penetrate into adjacent pore-space, causing Authigenic overgrowths of quartz and K-feld-
localized nodular cementation and inhibition of spar on detrital grains developed extensively into
authigenic quartz and feldspar overgrowths (Fig. pore-spaces unoccupied by early carbonates
2d). In open pore-space the dolomite exhibits (Figs 2e and 4b) throughout the Wessex Basin. In
euhedral crystal faces and is commonly over- some sandstones these quartz overgrowths have
330 G. E. Strong & A. E. Milodowski
produced pore-filling cements (e.g. Winterborne also occurs in compacted sandstones and silt-
Kingston, Wytch Farm). The apparently 'com- stones where it post-dates compaction effects
pacted' textures of some of these sandstones, as (e.g. it fills fractures) and encloses late ferroan
seen under the optical microscope, are shown by dolomite. Some feldspar dissolution sites may be
SEM analysis to be the result of mutual interfer- completely filled by anhydrite (Fig. 4a), leaving
ence of overgrowth surfaces. Overgrowths tend only the more stable pure K-feldspar over-
to be inhibited by early ferruginous or clay growths. This has created a grain-supported
coatings, and in the more argillaceous sediments fabric with anomalously oversized areas occu-
K-feldspar with rhombic adularia morphology pied by anhydrite cement--a fabric typical of
(Fig. 2f) forms in preference to syntaxial over- framework grain dissolution or replacement
growths. Microprobe analyses show that authi- (Schmidt & McDonald 1979b). Anhydrite is a
genic K-feldspar is essentially pure, whereas the major cement in sandstones in the Wytch Farm,
detrital grains may contain significant Na § or Winterborne Kingston and Shrewton boreholes,
Ba z+. Doubly terminated quartz crystals are and in other boreholes in the Winterborne
sometimes developed around silt grains resting Kingston area. It occurs as a minor cement in
in open pores. Quartz and K-feldspar authigene- Western Esplanade, and in a confidential well
sis appears to have begun before significant between Southampton and Salisbury, but was
burial compaction, for quartz overgrowths are not found in samples from Marchwood or the
commonly found compacted into less competent outcrop. The source of the mesodiagenetic
grains (e.g. carbonate clasts, micas and mud- anhydrite is presently unresolved. Early eodia-
stone clasts). genetic anhydrite and gypsum that formed in the
basin depocentre may have provided sulphates
for mobilization during burial and diagenesis.
Anhydrite cementation
The Permian Aylesbeare Group (Smith et al.
In general, anhydrite cementation seems to have 1974) beneath the Sherwood Sandstone, and the
occurred relatively early during mesodiagenesis Mercia Mudstone Group overlying the sequence,
because the anhydrite-cemented sandstones gen- are both anhydritic and potential sources of
erally lack features indicative of compaction sulphate.
(Fig. 4a). The mesodiagenetic anhydrite is
usually pore-filling and poikilotopic, and does
Compaction, framework grain dissolution
not show any features indicating near-surface
precipitation (expanded fabrics, felted lath or Compaction has affected most sandstones that
nodular habits, etc.). Its relationships to other remained uncemented by early calcite, dolomite
diagenetic phases is complex and probably due to or anhydrite. Those most affected are normally
repeated dissolution and reprecipitation of ear- argillaceous. Deformation of clasts (especially of
lier anhydrite. On account of its high solubility, micas and mudstone clasts), pressure-welding
remobilization of anhydrite is common during and grain point-contact fracturing occurred.
diagenesis (e.g. Holliday 1970, 1973, Mossop & Cemented frameworks remained relatively unaf-
Shearman 1973). There is a variation in the fected. The dissolution of framework grains
timing of quartz and feldspar overgrowth preci- occurred after the main effects of compaction,
pitation with respect to anhydrite cementation, since skeletal remnants and overgrowths of
which in some cases inhibited authigenic quartz partially dissolved grains, particularly feldspar
and feldspar but in other cases, even within the overgrowths, remain preserved (e.g. Figs 2b and
same borehole (e.g. Winterborne Kingston), 4a). Microperthites, plagioclases and K-feld-
anhydrite post-dates overgrowths. Both quartz spars are the grains most susceptible to grain
and feldspar grains and overgrowths locally dissolution, particularly the sodic laminae of
display signs of etching and embayment against microperthites, but quartz and unstable heavy
anhydrite. At Winterborne Kingston anhydrite minerals such as garnet (Morton 1982), magne-

FIG. 3. (a) Cathodoluminescence micrograph showing early non-luminescent calcite fringes with euhedral
terminations (C1) around cornstone clast (X) with brightly luminescent late calcite (C2) filling remaining
porosity. Quartz (Q) shows dull luminescence, K-feldspar (F) blue. Scale bar = 400 ~tm. Marchwood borehole,
1694 m. (b) Micrograph showing poikilotopic calcite partly filling pore space in sandstone and invading stress
fracture in K-feldspar grain (crossed polarizers). Scale bar -- 200 ~tm. Western Esplanade, 1752 m.
(c) Cathodoluminescence micrograph of the same area as (b) showing early non-luminescent calcite (C1)
seeded on a detrital calcite intraclast (X), with later luminescent calcite (C2). Calcite C2 post-dates the
feldspar overgrowths at (Y), fills stress fractures at (Z), post-dating compaction. Calcite Cz has euhedral
terminations into open pore space (P). Quartz (Q), K-feldspar (F). Scale bar = 200 pm. Western Esplanade,
1752 m.
~Wr
 Diagenesis of Sherwood Sandstones, Wessex Basin 331

FIG. 4. (a) Anhydrite cemented sandstone showing minimal compaction. Anhydride (A) occupies oversized
pores (X) and infills K-feldspar dissolution sites (Y). Quartz (Q), K-feldspar (F). Wytch Farm X14 borehole,
1598 m. (b) Anhydrite cemented sandstone showing some cement-dissolution porosity (P), extensive
framework grain dissolution (X), residual anhydrite (A). Quartz (Q), K-feldspar (F). Wytch Farm X14
borehole, 1600 m. (c) Barytes laths (B) occupying K-feldspar dissolution site (X), in sandstone largely
cemented by late calcite (C2) and some early calcite (C1). Marchwood borehole, 1694 m. (d) Euhedral
authigenic anatase in open pore space. Wytch Farm X14 borehole, 1600 m. (e) Deeply corroded anhydrite in
open pore space. Wytch Farm X14 borehole, 1636 m. (f) Detrital Ba-rich K-feldspars (F) with early pure K-
feldspar overgrowths (X), and later albite overgrowth (Ab). Winterborne Kingston borehole, 2318 m. (a, b,
c, f)--SEM backscattered electron micrographs; (d, e)--SEM secondary electron micrographs.

tite and ilmenite are also affected. Some local and titanium (in ilmenites and other heavy
collapse of partly dissolved grains was noted, minerals) liberated by grain dissolution were
particularly where grain dissolution has been rapidly reprecipitated as authigenic barytes (Figs
locally extensive. Although this secondary com- 2e and 4c) and anatase (Fig. 4d). The barytes
paction may locally reduce porosity, overall commonly occupies K-feldspar grain-dissolution
framework grain dissolution may account for up sites and, less commonly, plagioclase grain-
to 10% of the porosity in some samples. Barium dissolution sites. Anatase is commonly seen
(detected by probe analysis in many K-feldspars) resting on early calcite surfaces but enclosed
33 2 G. E. S t r o n g & A. E. M i l o d o w s k i

within later calcite. Although common, barytes fled by microprobe and SEM-EDS to be manga-
and anatase are volumetrically insignificant as nese-bearing ferroan calcite. It is a major cement
cements. at outcrop and in the Matchwood and Western
Esplanade wells, but it was identified only in
Development of cement dissolution porosity minor amounts in Winterborne Kingston (upper
cored section only, 2315.96-2334.11 m) and
In samples from Wytch Farm, Winterborne
Wytch Farm. These calcites are generally poiki-
Kingston and Western Esplanade there is strong
lotopic and pore-filling (Figs 2e, 3b, c and 4c)
evidence for secondary porosity arising from the
dissolution of anhydrite cements. In these wells although they also commonly replace framework
grains or fill framework grain dissolution sites
anhydrite is strongly etched and corroded against
(Figs 2b and 5b). Stress fractures in framework
open pore space (Fig. 4b, e). In some samples
gypsum occurs as haloes around remnants of grains may also be filled by late calcite (Fig. 3c).
anhydrite cement. Framework grain dissolution The late calcite cements fill minor fractures in
dolomite-cemented sandstone in a well near
appears to be closely associated with anhydrite
removal and areas devoid of anhydrite cement Winterborne Kingston and in early calcite-
cemented sandstone in Marchwood. Intense
usually exhibit a greater degree of framework
grain dissolution. Most of the cement dissolution veining by late calcite at Wolston Quarry (NGR
porosity appears to have developed relatively ST 0945 4015) in the Budleigh Salterton Pebble
late during mesodiagenesis as it is largely Beds at the west Somerset outcrop appears to be
unaffected by compaction. At Wytch Farm there associated with local faulting. These cements are
appears to be a relationship between the pres- clearly a late mesodiagenetic event since they
ence of hydrocarbon and the extent of anhydrite cross-cut other phases. They post-date quartz
dissolution. Anhydrite is most abundant in the and feldspar overgrowths which they commonly
oil-saturated sandstones, whereas it occurs to a etch or replace (Figs 2e and 5c) and they fill or
much lesser extent in oil-free sandstones. In both partly fill the remaining intergranular space not
cases the anhydrite is etched and corroded but it occupied by earlier cements. In open pore-spaces
would appear that hydrocarbon invasion may late calcites show uncorroded euhedral termina-
have suspended further anhydrite dissolution in tions (Figs 2e and 5b). The relationships between
the now oil-saturated sandstones. anhydrite dissolution, framework grain dissolu-
tion and late calcite cementation are complex,
Late quartz and albite and spatio-temporal overlap of all three pro-
cesses is evident. Samples in which anhydrite
A late mesodiagenetic generation of quartz and dissolution clearly preceded the precipitation of
albite overgrowths was noted in some samples late calcite have varying degrees of calcite
from the Winterborne Kingston borehole. cement (e.g. upper cored section of Winterborne
Quartz is developed as an overgrowth around Kingston, Western Esplanade, Wytch Farm and
grains with pressure-solution grain boundaries outcrop), whereas the continued presence of
and it also heals fractured grains. Albite similarly anhydrite appears to have inhibited late calcite
occurs as overgrowths on K-feldspar grains, both in some sandstones (e.g. lower cored section of
on external grain surfaces and intragranular Winterborne Kingston, Wytch Farm).
dissolution surfaces. The albite rims commonly Oxygen isotope analyses of late calcites in the
develop on earlier K-feldspar overgrowths (Fig. Wessex Basin indicate precipitation at a range of
4f) or fill stress fractures cutting both grains and burial temperatures up to 70~ (Bath et al. this
early overgrowths (Fig. 5a). In places the albite volume) if a basinal brine with similar isotopic
overgrowths have coalesced and filled pore composition to the present is assumed. This
spaces, forming a rigid framework cement. This temperature is close to present-day temperatures
generation of quartz and albite clearly post-dates but could have been achieved since maximum
the main effects of compaction, but appears to burial (i.e. from late Cretaceous). However, these
precede the late dolomite. The common associ- deep burial calcites are not in carbon isotopic
ation of authigenic albite with sites of feldspar equilibrium with the present formation fluid
dissolution suggests that dissolution of detrital (Knox et al. 1984, Bath et al. this volume) and
plagioclase and K-feldspar may locally have must therefore pre-date the present hydro-
supplied ions for albite authigenesis and that this geological regime.
process may have occurred concomitantly with
framework grain dissolution.
Late dolomite
Late calcite
A late replacive ferroan dolomite occurs in wells
Late calcite is characterized by bright orange in the Winterborne Kingston area, Shrewton,
luminescence under CI_, analysis and was identi- and between Southampton and Salisbury. The
 Diagenesis of Sherwood Sandstones, Wessex Basin 333

FIG. 5. (a) Compacted sandstone framework of detrital quartz (Q) and K-feldspar (F) with authigenic
overgrowths (X). Late authigenic albite forms overgrowths on K-feldspar and infills stress fractures (Ab).
Winterborne Kingston borehole, 2318 m. (b) Euhedral late calcite (C:) occupying secondary porosity in
feldspar dissolution site. Minor barytes (B). K-feldspar (F), Quartz (Q). Marchwood borehole, 1694 m.
(c) Detrital perthitic K-feldspar (F) with authigenic overgrowth (X) corroded by late calcite (C2). Early calcite
(CI). Marchwood borehole, 1694 m. (d) Detrital K-feldspar (F) and quartz (Q), with late ferroan dolomite
(D2) replacing framework grains. Winterborne Kingston borehole, 2316 m. (e) Corroded scalenohedral calcite
lining pore space in sandstone. Budleigh Salterton Pebble Beds, Holywell Quarry (NGR ST 1269 2703).
(f) Pore-filling and grain-replacive books of authigenic kaolinite, seen in intimate association with ragged
flakes of authigenic illite. Otter Sandstone, Budleigh Salterton (SY 060 816). (a, b, c, d)--SEM backscattered
electron micrographs; (e, f)--SEM secondary electron micrographs.

late dolomite is best developed in the Winter- and locally as poikilotopic cements (Fig. 5d).
borne Kingston borehole, occurring as euhedral Small isolated rhombs of late ferroan dolomite
overgrowths on earlier dolomite (Fig. 2d), as a are occasionally seen on quartz and feldspar
non-selective replacement of framework grains overgrowths from Marchwood and Western
334 G. E. Strong & A. E. Milodowski
Esplanade core samples. In the Winterborne under the present telodiagenetic regime (Walton
Kingston borehole, some late anhydrite occurs 1981).
which post-dates the ferroan dolomite.

Discussion
Fibrous illite
Effects of diagenesis on reservoir properties
In the porous zones of some sandstones in the
upper unit of the Marchwood sequence very The reservoir properties of the Sherwood Sand-
minor amounts of filamentous illite were noted. stone in the Wessex Basin have been described in
The illite clearly post-dates quartz overgrowths detail previously (Burgess et al. 1981, Burgess
but the relationship to other authigenic phases is 1982, Downing et al. 1982, Knox et al. 1984,
uncertain. It is associated with concentrations of Milodowski et al. 1986) and hence are only
detrital clay and clay coatings on grains, and briefly discussed here. The main factors that
locally appears to be replacive. Most of the illite determine the intergranular permeability of the
collapsed on to pore surfaces as a result of core- Sherwood Sandstone at the present time are the
drying (cf. McHardy et al. 1982). sedimentological characteristics (grain size, sort-
ing) and the extent of cementation and grain
dissolution. In general, the sediments were
deposited in a series of upward-fining fluvial
Telodiagenesis cycles (Henson 1970, Lott & Strong 1982). The
mudstones, siltstones and muddy sandstones
The diagenetic mineralogy of the rocks from forming the upper parts of these cycles are
outcrops in Somerset and Devon differs consid- considered to represent overbank flood deposits,
erably from that of the sandstones sampled from and vertical accretion deposits of temporary
the boreholes, presumably because of the direct lakes and back-levee swamps. They possess low
influence of recent weathering processes. The porosities (5-10%) and permeabilities generally
most obvious effects are the dissolution of calcite less than 1 mD. The primary porosity in these
cements (Fig. 5e), framework grain dissolution, lithologies has been reduced by compaction and
the precipitation of kaolinite (Fig. 5f) commonly also in some cases by the replacement or
occupying sites of feldspar dissolution, and the cementation of the matrix by calcite. The
precipitation of minor amorphous or chalcedonic moderately sorted grainstones, representing la-
silica on detrital quartz grain surfaces. Flakey teral accretion deposits at point or channel bars,
illite presumably also of authigenic origin may be possess the best reservoir characteristics, but
intimately associated with kaolinite in these porosities vary considerably from < 5-28% and
rocks. Some pore walls may also be lined with permeabilities range from 1 mD to about 7 D.
spongy precipitates of manganese oxides or Some sandstones from the east Devon outcrop
hydroxides. Framework grain dissolution is may have porosities up to 35% and permeabili-
considerably more extensive at outcrop com- ties up to 9 D (Milodowski et al. 1986).
pared with the deep boreholes. However, it is not Conglomerates, representing channel bedload
possible to distinguish telodiagenetic from meso- deposits, also show considerable variation in
diagenetic dissolution sites except where the porosity and permeability depending mainly on
presence of late calcite in the secondary pores the degree of cementation of the sandy matrix.
indicates early features. Telodiagenetic effects Knox et al. (1984) established an inverse relation-
are most obvious in the outcrop south of ship between the extent of calcite cementation
Wellington where the sediments may have and the permeabilities in sandstones of the
originally had less calcite cementation. In many Marchwood well, with slight modifications due
parts of the east Devon outcrop hard concretion- to compaction and detrital clay contents. This is
ary bands of limonite or goethite are developed also true elsewhere in the basin, although the
near the surface ('iron pans'). No anhydrite or effects of other cements (anhydrite, dolomite) are
gypsum was found in outcrop samples but, as significant.
stated before, their presence in wells 30 km to the An important feature of the Sherwood Sand-
east suggests that they may have been originally stones throughout the Wessex Basin is the
present as mesodiagenetic cements which have occurrence of discrete high permeability zones
subsequently been removed by meteoric ground- that have porosities ranging from 18 to 28% and
waters. The groundwaters in the Sherwood permeabilities generally ranging from 100 mD to
Sandstone at outcrop are saturated with respect about 6 D. They form significant aquifers at
to both kaolinite and illite and indicate that K- Marchwood, Western Esplanade, Winterborne
feldspar alteration and dissolution is occurring Kingston and Wytch Farm and may account for
 Diagenesis of Sherwood Sandstones, Wessex Basin 335
as much as 95~ of the total transmissivity of the lation. More extensive dissolution of anhydrite
formation in the Marchwood and Western followed only in sandstones below the oil-water
Esplanade wells (Milodowski et al. 1986). They contact.
have typically only minor or absent early and late As noted earlier, parts of the Sherwood
carbonates, extensive framework grain dissolu- Sandstone Group sequences in the basin depo-
tion of feldspars (which may account for as much centre contain eodiagenetic (near-surface) anhy-
as 10~ of the total porosity of some samples), and drite which may represent an inland sabkha or
an extensive secondary porosity due to cement playa environment of deposition with minimal
dissolution inferred here to be after anhydrite. In development of calcretes. In areas marginal to
contrast, sandstones that have extensive calcrete the playa environment some calcretes have
development, or have been affected significantly developed that have been later replaced by early
by late calcite (possibly due to early anhydrite non-ferroan dolomite (e.g. lower part of the
removal), are tightly cemented and offer little or Winterborne Kingston sequence). Near the
no aquifer potential. Knox et al. (1984) consi- basin margins, however, calcretes appear to have
dered that the poresity of these permeable been more extensively developed, but there is no
horizons is mainly primary porosity enhanced by evidence of early dolomitization. Isotopic analy-
framework grain dissolution and that it largely ses of the early dolomite (Bath et al. this volume)
escaped calcite cementation. However, their indicate that this dolomitization occurred near
study was limited to material from the March- the surface under highly evaporitic conditions.
wood sequence, where anhydrite is absent. Dolomitized calcretes have been described as an
Burley (1984), and Burley & Kantorowicz (1986) intermediate lithofacies between calcretes and
have suggested that the secondary porosity in the playa deposits (gypsites) in Australian inland
highly permeable zones may be due to calcite palaeodrainage systems (Arakel & McConchie
dissolution, but the ubiquitous presence of well 1982). They consider that the dolomitization
preserved detrital cornstone clasts and euhe- occurs in zones of fluctuating groundwater level
dral terminations into open pore-space of both where repeated solution and reprecipitation of
early and late calcites would appear to preclude the calcrete have taken place.
this mechanism for the Wessex Basin sand- The exceptionally high porosities and permea-
stones, apart from those affected by telodiagene- bilities in sandstones of the east Devon outcrop
sis near the outcrop. We consider that much of are due to extensive decalcification and feldspar
the secondary porosity in the highly permeable grain dissolution. Outcrop groundwater chemis-
bands is due to the late removal of anhydrite, tries indicate that these processes (together with
which has protected pore-space from the effects the precipitation of authigenic kaolinite and
of compaction and has prevented extensive illite) are occurring at the present time (Walton
precipitation of late calcite in these zones. The 1981). However, it is unlikely that calcite
subsequent dissolution of the anhydrite thereby dissolution extends very far into the subcrop
rejuvenated considerable primary and secondary since (according to Walton 1981) the ground-
porosity. Although anhydrite has not been waters rapidly achieve calcite saturation with
reported from the Marchwood well, its presence depth. Any anhydrite originally present in the
at Western Esplanade and elsewhere in the rocks now at outcrop would have dissolved under
subcrop suggests that it may well have been the present telodiagenetic regime. The extent of
present at Marchwood, and in rocks now at anhydrite dissolution eastwards into the basin is
outcrop. There is no evidence at present to unknown, although anhydrite is present in deep
suggest that the highly permeable zones are boreholes about 30 km east of the outcrop.
laterally extensive and hence correlatable. In a
variable fluviatile sequence this seems highly
unlikely. However, it is possible that they may be ACKNOWLEDGMENTS:This work formed part of the
hydraulically interconnected via fissures, faults BGS Low Enthalpy Geothermal Energy Project fund-
etc. ed by the Department of Energy. The authors would
Anhydrite removal at Wytch Farm appears to like to thank colleagues of the British Geological
have been restricted by the migration of hydro- Survey for support and encouragement, in particular
carbons into the sandstones. Oil migration at the Project Manager, Dr R. A. Downing, and Dr S.
Wytch Farm is believed to have begun during the Holloway for discussions relating to the regional
geology. We are grateful to Drs N. A. Chapman, R. W.
early Cretaceous and could have continued until O'B Knox and G. Warrington for critically reading the
the Alpine (Eocene) uplift (Colter & Havard manuscript, and to Dr M. Scherer and an anonymous
1981). The relatively limited extent of anhydrite reviewer for further constructive comments. This
corrosion in the oil-saturated sandstones appears paper is published by permission of the Director,
to indicate only minor dissolution by an initial British Geological Survey, (NERC). Figures and
water flush preceding hydrocarbon accumu- illustrations are NERC copyright reserved.
336 G. E. Strong & A. E. Milodowski

References
ARAKEL, A. V. & MCCONCHIE, D. 1982. Classification American Association of Petroleum Geologists, 53,
and genesis of calcrete and gypsite lithofacies in 830-40.
paleodrainage systems of inland Australia and KNOX, R. W. O'B., BURGESS, W. G., WILSON,K. S. &
their relationship to carnotite mineralization. BATH, A. H. 1984. Diagenetic influences on
Journal of Sedimentary Petrology, 52, 1149-70. reservoir properties of the Sherwood Sandstone
BATH, A. H., MILODOWSKI, A. E. & SPIRO, B. 1987. (Triassic) in the Marchwood Geothermal bore-
Diagenesis of carbonate cements in Permo- hole, Southampton, U.K. Clay Minerals, 19, 441-
Triassic sandstones in the Wessex and East 56.
Yorkshire-Lincolnshire Basins, UK: a stable LoTr, G. K. & STRONG, G. E. 1982. The petrology and
isotope study. This volume. petrography of the Sherwood Sandstone (?Middle
BATHURST, R. G. C. 1976. Carbonate Sediments and Triassic) of the Winterborne Kingston borehole,
their Diagenesis. Developments in Sedimentology Dorset. In. RHYS, G. H., LOTT, G. K. & CALVER,
12. Elsevier, Amsterdam. M. A. (eds). The Winterborne Kingston Borehole,
BURLEY, S. D. 1984. Patterns of diagenesis in the Dorset, England, pp. 135-42. Report of the
Sherwood Sandstone Group (Triassic), United Institute of Geological Sciences No. 81/3.
Kingdom. Clay Minerals, 19, 403-40. MCHARDY, W. J., WILSON, M. J. & TAIT, J. M. 1982.
--& KANTOROWICZ, J. D. 1986. Thin section and Electron microscope and X-ray diffraction studies
S.E.M. criteria for the recognition of cement- of filamentous illite clay from sandstones of the
dissolution porosity in sandstones. Sedimentology, Magnus Field. Clay Minerals, 17, 23-39.
33, 587-604. MILODOWSKI, A. E., STRONG, G. E., WILSON, K. S.,
BURGESS, W. G. 1982, Hydraulic characteristics of the ALLEN, D. J., HOLLOWAY,S. & BATH, A. H. 1986.
Triassic Sherwood Sandstone and the Lower Diagenetic influences on the aquifer properties of
Jurassic Bridport Sands intervals, as derived from the Sherwood Sandstone in the Wessex Basin.
drill stem tests, geophysical logs and laboratory Investigation of the Geothermal Potential of the UK.
tests. In: RHYS, G. H., LOTT, G. K. & CALVER, M. British Geological Survey. J
A. (eds). The Winterborne Kingston Borehole, MORTON, A. C. 1982. Heavy minerals from the
Dorset, England, pp. 164-75. Report of the sandstones of the Winterborne Kingston bore-
Institute of Geological Sciences No. 81/3. hold, Dorset. In. RHYS, G. H., LOTT, G. K. &
- - B U R L E Y , A. J., DOWNING, R. A., EDMUNDS, W. CALVER, M. A. (eds). The Winterborne Kingston
M. & PRICE, M. 1981. The Marchwood Geother- Borehole, Dorset, England. Report of the Institute
mal borehole--a preliminary assessment of the of Geological Sciences No. 81/3.
resource. Investigation of the Geothermal Potential MOSSOP, G. D. 8s SHEARMAN,D. J. 1973. Origins of
of the United Kingdom. Institute of Geological secondary gypsum rocks. Transactions of the
Sciences. Institute of Mining and Metallurgy, 82, 147-
COLTER, V. S. & HAVARD,D. J. 1981. The Wytch Farm 54.
Oil Field, Dorset. In." Petroleum Geology of the SCHMIDT, V. & MCDONALD, D. A. 1979a. The role of
Continental Shelf of North- West Europe, pp. 494- secondary porosity in the c o u r s e o f sandstone
503. Institute of Petroleum, London. diagenesis. In : SCHOLLE,P. A. & SCHLUGER,P. R.
DEAN, W. E., DAVIS, G. R. & ANDERSON,R. Y. 1975. (eds). Aspects of Diagenesis, pp. 175-207. Society
Sedimentological significance of nodular and of Economic Paleontologists and Mineralogists
laminated anhydrite. Geology, 3, 367-72. Special Publication No. 26.
DOWNING, R. A., ALLEN, D. J., BURGESS, W. G. & ---& 1979b. Texture recognition of secondary
EDMUNDS, W. M. 1982. The Southampton (Wes- porosities in sandstones. In: SCHOLLE, P. A. &
tern Esplanade) Geothermal Well--a preliminary SCHLUGER, P. R. (eds). Aspects of Diagenesis, pp.
assessment of the resource. Investigation of the 209-25. Society of Economic Paleontologists and
Geothermal Potential of the United Kingdom. Mineralogists Special Publication No. 26.
Institute of Geological Sciences. SHEARMAN,D. J. 1966. Origin of marine evaporites by
EDMONDS, E. A. & WILLIAMS,B. J. 1985. The geology diagenesis. Transactions of the Institute of Mining
of the country around Taunton and the Quantock and Metallurgy, 75, 208-15.
Hills. Memoirs of the British Geological Surt~ey SMITH, D. B., BRUNSTROM,R. G. W., MANNING, P. I.,
Sheet, 295. SIMPSON,S. & SHOTTON,F. W. 1974. A correlation
HENSON, M. R. 1970. The Triassic rocks of South of Permian rocks in the British Isles. Journal of the
Devon. Proceedings of the Ussher Society, 2, Geological Society of London, 130, 1-45.
172. WALKER, T. R., WAUGH, B. & GRONE, A. J. 1978.
HOLLIDAY, D. W. 1970. The petrology of secondary Diagenesis in first-cycle desert alluvium of Ceno-
gypsum rocks: a review. Journal of Sedimentary zoic age, southwestern United States and north-
Petrology, 40/2, 734-44. western Mexico. Bulletin of the Geological Society
-- 1973. Early diagenesis in nodular anhydrite rocks. of America, 89, 19-32.
Transactions of the Institute of Mining and Metal- WALTON, N. R. G. 1981. A detailed investigation of the
lurgy, 82, 81-4. geochemistry of groundwaters from the Triassic
KINSMAN,D. J. J. 1969. Modes of formation, sedimen- sandstone aquifer of south-west England. Report
tary associations and diagenetic features of shal- of the Institute of Geological Sciences No.
low water and supratidal evaporites. Bulletin of the 81/5.
 Diagenesis of Sherwood Sandstones, Wessex Basin 337
WARRINGTON, G., AUDLEY-CHARLES,M. G., ELLIOTT, TAYLOR, F. M. 1980. A correlation of Triassic
R. E., EVANS,W. B., IVEMEY-COOK,H. C., KENT, rocks in the British Isles. Geological Society of
P. E., ROBINSON, P. L., SHOTTON, F. W. & London Special Report No. 13.

G. E. STRONG & A. E. MILODOWSKI,British Geological Survey, Keyworth, Nottingham

NGI2 5GG, UK.
Deposition and diagenesis in an extensional basin: the CoraHian
Formation (Jurassic) near Oxford, England

C. B. de Wet

S U M M A R Y : The mixed carbonate-siliclastic Corallian Formation was deposited on

horsts and in grabens which were produced during Jurassic extensional tectonics. The high
areas became sites for sandstone and carbonate deposition in shallow subtidal environments
while clays and allochthonous debris accumulated in the depressions. Tectonic activity,
perhaps confined to single fault blocks, produced depositional variation distinct to each area.
The basin has been subdivided into four major east-west trending fault blocks. The most
northerly of these is the Oxford Shallows, interpreted as a small accretionary-margin slope.
The depositional facies and diagenesis of this area are discussed relative to the basin as a
whole.
The diagenesis of the Corallian sediments has been divided into three phases of
cementation: pre-burial, shallow burial and deep burial. Phase 1 spars include marine
precipitates and non-ferroan, neomorphic and drusy spars. Phase 2 cements are ferroan,
precipitating during shallow burial, including an intermediate ferroan (mauve) spar which
may represent development of a redox front. Phase 3 spar post-dates vuggy dissolution which
probably took place during Barremian uplift. The cement precipitated during the ensuing
burial phase.
Geochemistry and stable isotope data show an overall fluid evolution path for the first two
spars, although Sr ~§ values in replaced corals reveal variability in the degree of meteoric
flushing. The chemistry of the late spars suggests low water/rock ratios, where local
dissolution provided the cations for the cement. The oxygen isotope values suggest, however,
that meteoric and connate waters were mixed at slightly elevated temperatures.

The Oxfordian Corallian Formation is a mixed Structure

carbonate-siliclastic sequence found at outcrop
and in the subsurface of southern England. The Chadwick (1986) at the British Geological
study area has traditionally been divided into the Survey produced a synthesis of the basin's
Hampshire Basin to the west and the Wealden structural evolution based on seismic and bore-
Basin to the east, but as recent structural surveys hole data. He proposes that the Wessex Basin
(Stoneley 1982, Whittaker 1985) and this study experienced episodic extensional tectonics pro-
demonstrate, east and west are depositionally ducing a series of grabens and fault block highs.
and diagenetically linked, forming the Wessex Chadwick invokes McKenzie's (1978) model of
Basin (Fig. 1). Lithofacies range from shale to rifting, followed by subsidence, to explain the
sandstone to oolite and packstone (Fig. 2) with repeated pattern of shallow water deposits being
preserved thicknesses varying from 4 m on the followed by laterally widespread deeper water
edge of the basin to over 100 m near its centre. sediments. The Corallian Formation is one of the
Previous studies of the Corallian Formation units deposited during an extensional phase.
have explained the lithologic variability in terms Extension produced highs where predominantly
of shallowing-upward cycles (Talbot 1973, carbonates and coarse clastics were deposited
Brookfield 1973, 1978, Wright 1981 and Chowd- and intervening lows that collected clays and
hury 1982). Many depositional models have been allochthonous debris (Fig. 3). These highs
proposed, including an analogy with the present experienced local tectonic activity, moving them
Gulf of Mexico (Brookfield 1977) and a carbon- above and below sea-level independently of
ate regime like the Bahamas (Chowdhury 1980). movements elsewhere in the basin. This helps to
In this study, diagenesis, coupled with a recent explain much of the lithologic variation, and
structural synthesis (Chadwick 1986), has eluci- difficulty in correlation, found north-south
dated the basin's fluid pathways and has led to a across the area.
new depositional model. This paper discusses Reports in the literature on other extensional
recent developments in the understanding of the regimes are revealing similar patterns to that
basin's structural history, proposes a new deposi- seen in the Wessex Basin. Profiles across China's
tional hypothesis and examines the diagenesis in Jianghan and Bohai Bay basins, which are of
the Oxford area as it relates to the whole basin. roughly the same dimensions as the Wessex

From: MARSHALL,J. D. (ed.), 1987, Diagenesis of Sedimentary Sequences, 339

Geological Society Special Publication No. 36, pp. 339-353.
340 C. B. de Wet

. WELSH 9 ~..... ~~~./..::::.:.:. I..LONDON.. .

9. .-.-.-. .......v ZS~-.,.. ~',"" " "
. . MASSIF.
. . . . . -. .~ -. " - ~ : . ..........
i~ OXFORD PLATFORM
\ \Xi .... : 99
' ......i ~ B ~ SHALLOWS ' . " "

SAMPLE LOCATIONS
9 " " 4 ' ~:::81
. ' . " . "t~' g I--'''.~
_PEWSEY FAULTS ~ _~
~ l k . /
,^ II *IP ~-. " -
* - - ~ .

o• A.~A: 9 9 ~"'//:::..,f~,:..,,,,=.,=-v '# - " - - " - - - - "~'[3~14

, HEAD,NGTON
2 LYE HILL
"." o.@'~'~i/PEWSEY BASIN ""
' WHEATLEY : : f , ~ / ~
4 COTHILL
s .... ELL WEALD BASIN
6 CHALGROVE

PEWSEY HIGH: 0 ,g. .... 15

CAL.E ~ , ~ . BASIN I'~.._..~.
WA RDOUR
9I0 COLLENDEAN
KINGSCLERE FARM . I~POr~,., 18.,~ t ~ ' - . . . ~ . FAULTS
II PALMERS WOO0 1,2,3
12 TATSFIELD
13 SLETCHINGLEY 1,2,3
14 ASHOUR
WEALD BASIN:
15 BOLNEY
WARDOUR HIGH:
16 WINCHESTER
17 FORDINGBRIDGE D
Ia MARCHWOOD CORALUAN OUTCROP
19 PORTSDOWN BASIN
PORTLAND-WIGHT AREA:
20 ARRETON
~, ABBO~SBU~Y
22/23 WEYMOUTHAREA
~ WESSEX BASIN (:500 km)
FIG. 1. Map of southern England showing location of the Corallian Formation outcrop and borehole sites
used in this study. Major east-west trending faults and basins dissect the Wessex Basin.

KIMMERIDGE basin, show an almost identical arrangement of

OXFORD tilted fault blocks stepping down into the basin's
CLAY AREA interior (Hefu 1986). While at present on the
UPPER South China and Red Sea fault-bounded highs,
CALCAREOUS carbonates and coarse clastics are accumulating
~ GRIT and thicker deposits of argillaceous and allochth-
~~ o
0
GLOS
OOLITE
I Im onous debris are filling in the lows (Ru & Pigott
1986).
SERIES
O
c0
OSMINGTON ~ orals Observations: Oxford Shallows
OOLITE LL
rr
E3
~5 lebris An examination of one structural high, the
SERIES Oxford Shallows block, reveals many of the
--i depositional and diagenetic features predicted
Ld E
BERKSHIRE
J
< ~i!lq~ar'z by the proposed model. The Oxford Shallows is
[3_ rF sand the most northerly section within the Wessex
s OOLITE O
O Basin and between two land masses, the London
SERIES
Platform and the Welsh Massif.
LOWER
r
O CALCAREOUS i quartz
sand Depositional environments
GRIT At outcrop the Formation (maximum 10 m thick)
consists of a predominantly cross-bedded sand-
o O OXFORD stone (exposed in quarries at Shellingford and
CLAY Cothill, NGR SU 960 430, SU 980 450) followed
by coral patch reefs and inter-reef shelly debris.
FIG. 2. Corallian Formation stratigraphy and Oxford The sandstones are well-sorted quartz sands with
area lithological log. minor feldspar, containing occasional flaser
 Diagenesis in the Corallian Formation 34I

TECTONIC FRAMEWORK OF THE WESTERN WESSEX BASIN

DEPOSITION OF THE OXFORD CLAY meon seo level
~ - - ............. . .............. " ~ : . . . . . . . . _,-_:. . . . . . :_:_:__:_:_:_:_=_:_:;
:-=-:. . . . . . . . . . . ~ ~ = - _ = - - = : ~ - - - ~

DEPOSITION OF THE CORALLIAN

~.~ '--- ' . . . . --, i _

............ -=::-=--'---- ::=-:~=_- -=----~:------ -=:::-=-:-. . . . . . . . . . . . . . . . . . ::: ~5:--=-:-_-_-_--_ __ :~ :----=r = : - : - ~ : - = = - : = = ~ = = -

.... 'r.._ IX .... 'l .........

EXTENStON , 2>

DEPOSITION OF THE KIMMERIDGE CLAY

20kin "

FIG. 3. Tectonic model for the western Wessex Basin (after Chadwick 1986). The Corallian Formation is
bounded by the Oxford and Kimmeridge Clay Formations. The Corallian Formation rocks onlap the London
Platform to the north. Equivalent deposits are reported from Normandy to the south (Ffirsich 1975).

bedding and mud filled burrows. Shell fragments ooids and pisoids in a micritic-clay matrix. At
and Rhaxella sponge spicule moulds are locally the top of the Corallian there are reworked coral
abundant. fragments and shell fragments. The Chalgrove
The overlying limestone contains cross-lami- borehole contains mudstones and wackestones
nated sands of heavily micritized shell and coral which are extensively bioturbated, with bivalve
fragments (Fig. 4a), pisoids, intraclasts, oncoids debris concentrated in four main horizons.
and ooids. Well preserved Girvanella oncoids are
present plus oncoids showing multiple stages of
encrustation and reworking (Fig. 4b). Echinoids, Diagenetic features
bivalves, and in situ corals comprise most of the The coarse-grained deposits which formed on the
patch reef assemblage (Sellwood 1978). Disarti- block crests acted as conduits for diagenetic fluid
culation of Rhaxella sponges liberated round flow. The Corallian's stratigraphic position
opaline spicules which often form ooid nuclei or between two thick and impermeable clay forma-
cement together to form grapestone clasts. tions, the Oxford and Kimmeridge Clays, en-
Rounded intraclasts containing neomorphic cor- abled it to function as an aquifer, transporting
al remains with non-ferroan and ferroan cements meteoric water in east-west conduits, isolated by
are common. Patchily distributed within the the surrounding intra and interformational clays.
grainstones are micrite clumps, which form Lithification is divided into three phases; pre-
meniscuses and bridges between grains partially burial, shallow burial and deep burial. Phase 1
occluding porosity (Fig. 4c). includes marine precipitates, vadose features,
The sediments thicken to 40 m in the Harwell non-ferroan neomorphic and drusy spars. The
4 and Chalgrove boreholes to the east. The phase 2 cements are ferroan and precipitated
sequence at Harwell 4 consists of laminated under reducing conditions, probably accom-
mudstones which coarsen upward into bioturbat- panying shallow burial. Phase 3 deep burial
ed siltstones. These are interrupted by a 10 cm stylolites are associated with a blocky spar that
thick, graded, cross-laminated fine sandstone has a patchy response to staining (Dickson 1966)
bed with an erosional base and abrupt upper and a characteristic mottled cathodolumines-
contact. Terrigeneous sands are gradually super- cence (CL) pattern. The Corallian Formation
seded by wackestones and packstones containing also contains subordinate amounts of dolomite,
342 C. B. de Wet

FIG. 4. Depositional components from the Oxford

Shallows fault-bounded (?) high. (a) Well preserved
microborings inside a thick micrite envelope. The
envelope was filled with non-ferroan phase 1 spar.
(b) An oncoid showing multiple generations of
encrustation by foraminifera and algae around a
quartz and echinoid nucleus. (c) Ooids, a coated
echinoid spine fragment and multiply encrusted
oncoid cemented by phase 1 vadose micrites. Arrows
indicate meniscus and bridge features suggesting that
the pores contained both air and water while the
micrite was deposited. Ferroan phase 2 equant
cement has occluded the remaining porosity.

siderite, silica and hydrocarbon not discussed gic transformations create complications only
here. distantly related to tectonic movements. As
Basin burial curves are often used to delineate this study and many others (Moore
diagenetic trends. They may, however, present & D r u c k m a n 1981, Sailer 1986) demonstrate,
an oversimplified view as local permeability simple burial curves may belie the complications
barriers, water-table fluctuations and mineralo- of diagenesis.

FIG. 5. Cements from diagenetic phases 1, 2 and 3. All samples were stained according to Dickson (1966)
except D, which was polished for cathodoluminescence. (a) Serpulid encrustations, micrite envelopes and an
echinoid spine fragment showing phase 1 and 2 cements. LMC discrete peloids, p, may have had a HMC
precursor, representing marine cements that precipitated during phase 1. Non-ferroan phase 1 spar is
particularly well developed on the echinoderm as a syntaxial overgrowth (arrow 1). It precipitated from
oxidizing and/or iron-poor meteoric water. The first ferroan (mauve) spar that precipitated during phase 2 is
also syntaxial to the echinoderm spine (arrow 2a). It may represent the development of a redox front. The
second ferroan spar, 2b, also precipitated during shallow burial from iron-rich, reducing formation water. The
pale blue to the right of the photograph is dyed impregnating plastic. (b) Early diagenetic cements. Phase 1
cements are represented by peloids, p, now LMC. They probably had a HMC marine precursor. Spar 1 (1) is
a non-ferroan syntaxial overgrowth on part of an echinoderm shell wall, e. The field of view is inside an
echinoid test cavity that was partially filled by micrite envelopes and shell debris. Phase 2 cements are
represented by 2a and 2b. 2a represents the first ferroan (mauve) spar precipitated during shallow burial. 2b
was subsequently precipitated in the remaining pores as the formation water became reducing and iron-rich.
(c) Phase 1, 2 and 3 spars are shown. Phase 1 non-ferroan equant spar forms isopachous rims around ooids
(arrow 1) and micrite envelopes. Phase 1 spar is present inside some micrite envelopes, indicating that
aragonite dissolution preceded its precipitation. Phase 2 precipitation (2) is represented by ferroan spar that
occluded the remaining porosity. Phase 3 dissolution, probably the result of undersaturated Barremian
meteoric water entering the Corallian Formation preferentially dissolved a coral(?) fragment. Blocky phase 3
spar (3) precipitated in the void. (d) Photomicrograph showing phase 1 non-ferroan equant spar (1) and a
micrite-filled coral mould which has been partially dissolved by phase 3 fluids. Phase 3 (3) crystals contain
relics and inclusions of the previous fabric. (e) Cathodoluminescence photomicrograph showing cements from
phases 1 and 3. Phase 1 spar (1) is dully zoned adjacent to the brightly luminescent substrate. The zones
become more brightly luminescent towards the pore centre. The remaining porosity is filled by phase 3 (3)
blocky spar which exhibits a characteristic mottled luminescence.
ih
3 mm
L I

~r I

0.25 mrr~
immgullililr~: m

| 1
l

2 mm
i ,
9

0.15 mm
1 .I

0.15 m m
 D i a g e n e s i s in the Corallian F o r m a t i o n 343

Phase I cements (Fig. 5e). The most striking aspect of the spar is
its association with dissolution cavities (Fig. 6d).
Phase 1 cements are most common in carbonate Though it post-dates all previous spars, it occurs
horizons. Micrite cemented grapestone and in non fabric-selective vugs and fractures. The
limeclasts constitute about 5~ of the carbonate contact with pre-existing fabrics is marked by
allochems in the grainstones. The earliest pore-
corroded crystal faces, and ghosts and relics of
occluding cements are columnar fringing spars. precursor crystals (Fig. 5d).
These are particularly common on prismatic
bivalve fragments that escaped micritization
(Fig. 6a). Peloidal micrites are usually found
inside protected cavities (Fig. 5a, b), however,
Interpretations: Oxford Shallows
silt-filled fractures, micritic meniscuses between
grains and clotted muds are locally common (see Depositional environment
Fig. 4c). Extensive mouldic porosity, up to 50%,
The sandstones at outcrop have been interpreted
was created by the dissolution of aragonite,
as representing deltaic input (Brookfield 1973),
leaving open micrite envelopes (Fig. 6b). This
or nearshore shoals and beaches (Talbot 1973),
fragile framework was stabilized by neomorphic
possibly derived by the reworking of an older
replacement and the precipitation of the first
sandstone (Triassic?) from topographically sub-
non-ferroan druse spars in primary and dissolu-
dued landmasses to the east and west. Marine
tion pores (Fig. 5c). The cements are dog-tooth to
shells and Rhaxella spicule moulds concur with a
equant, usually isopachous, and line both the
nearshore marine setting.
interior and exterior of micrite envelopes and
The presence of the carbonate units indicates a
coral chambers. Under CL most of the phase 1
diversion of the terrigeneous supply, allowing
spars begin with a non-luminescent zone fol-
pisoids and oncoids to form. Photic zone
lowed by brightly luminescent and non-lumines-
conditions supported blue-green algal oncoids
cent subzones (Fig. 5e).
and scleractinian corals. Micritization, boring,
rounding, and encrusting indicate slow sedimen-
Phase 2 cement tation rates on the high. Patch reefs developed on
stabilized sandstone shoals near the escarpment
The first ferroan (mauve) cement is particularly where currents were most active (Headington,
well developed as syntaxial overgrowths on NGR SP 030 430, Shellingford, Cothill). The
echinoderms (Fig. 5a, b). Although volumetrical- abundance of Rhaxella spicule moulds attests to
ly rare the spar is both spatially and chemically quiet water conditions in the back-reef and
transitional between phase 1 non-ferroan and lagoon (McKerrow 1978). Intraclasts containing
later phase 2 ferroan cements. Under CL the equant cements in coral vugs imply that the
overgrowths show complex zones of non-lumi- marine sediments were subaerially exposed. The
nescent and brightly luminescent bands. The presence of both non-ferroan and ferroan equant
second phase 2 spar precipitated during shallow cements within the clasts suggests that the spars
burial as an equant or granular ferroan cement precipitated in a well-developed meteoric sys-
(Fig. 5a, c). It is the final pore-occluding cement tem. Additional evidence of exposure includes
in the grainstones and is the first and often only micritic clots and muds forming meniscuses
spar in the stratigraphically lower clastic hori- between grains. These features suggest that small
zons. Rhaxella sponge spicule moulds are almost fluctuations in sea-level resulted in subaerial
exclusively filled with this ferroan spar. The exposure and the development of a phreatic lens
cement post-dates minor compaction, filling below the shoal tops.
small fractures, broken micrite envelopes and Eastward, across the slope break, the much-
spalled ooid jackets (Fig. 6c). The crystals are thickened facies at Harwell 4 represents deeper
usually dark under CL, but may show three to water deposition (Milodowski & Wilmot 1984).
five dull-luminescing zones not detected by The sediments mirror the shallow water equiv-
staining. alents, progressing from terrigeneous clastics,
including a storm deposit or small turbidite (the
10 cm thick erosional sand bed), to pisoids and
Phase 3 cement
ooids, through to shelly debris and coral frag-
Phase 3 is the only spar to show a consistent ments. All of the allochems are allochthonous,
crystal morphology, stain and CL response swept off the platform and encased in clay and
across the entire basin. The crystals are blocky, micrite. The finer grained, more uniform sedi-
often poikilotopic, and are up to 1 mm across. ments of Chalgrove represent the deeper subtidal
They have a patchy iron distribution based on zone, more distal to the high (Stanley & Moore
staining (Fig. 5c, d) and a mottled CL response 1983).
344 C. B. de Wet

,-Q O,--~

eD

~ ~ ~ 0 o-

tMJ cIJ ' 4 . . . ~ ~.~

r ,.r r

~ ~ ,.~ tt~ " .~.

~ :~ E ~ ~ .=_
 Diagenesis in the Corallian F o r m a t i o n 345

Diagenesis may also affect luminescence (ten Have & Heijen

1985).
Phase 1 cements
Phase 2 cement
Micritic marine cements bind together the
grapestone and intraclasts, while peloidal preci- Through shallow burial the meteoric waters
pitates are preserved in cavities. These spherical gradually became more reducing and/or iron-
peloids, often surrounded by an interlocking rich, precipitating the first ferroan (mauve-
crystal mosaic, are distinct from clotted micrites staining) spar. This cement's intermediate iron
which occur in depositional layers and burrows. values and its position between phase 1 and the
According to Macintyre (1985), spherical shape later phase 2 ferroan cements provide a record of
and isolated formation site are two criteria useful transition as the redox front developed and
for distinguishing chemically precipitated pe- progressed through the sediments. This phase
loids from biologically induced pellets. Corallian was short-lived as most of the cement crystals are
peloids meeting these criteria are interpreted as too small to be observed except at Cothill (Fig. 5)
being originally magnesium calcite precipitated and in the more rapidly precipitating echinoid
as a cement. Other possible marine cements overgrowths.
include columnar fringes around ooids and on As burial continued, marine waters mixed
some bivalve fragments. They are presumed to with, and were displaced by, reducing meteoric
have had a high magnesium calcite (HMC) or fluids. This was accompanied by minor grain
aragonite precursor based on their columnar breakage and overpacking. The precipitation of
habit and relation to the substrate. The crystals ferroan granualar cement occluded the remain-
in Fig. 6(a) have displaced the surrounding ing porosity and prevented further compaction.
micritic matrix indicating that the muds must Talbot (1971) states that this ferroan cement is a
have been soft at the time of growth. This implies deep burial precipitate, however, it is cut by
that precipitation commenced before compac- stylolites and pre-dates hydrocarbon migration.
tion could solidify the sediments. It is possible Non-fabric-selective solution voids filled with
that the sediment was introduced after the the youngest cement cut across the ferroan
growth of the cement, however because of its crystals. Therefore the ferroan spar must have
geopetal distribution this would be physically the formed prior to Barremian uplift and dissolution
more difficult alternative. Talbot (1971) and (see below).
Chowdhury (1982) interpret the columnar spars The Lower Calcareous Grit lacks phase 1
on ooids as originally HMC or aragonite marine cements because a minimum of shelly material
phreatic cements. Neither the distinct peloids was available for dissolution and reprecipitation.
nor these fringes could be separated for chemical At outcrop the sandstone is cemented by phase 2
analysis. ferroan spar, but in the Harwell 4 borehole there
Silt-filled fractures, small collapse breccias, are friable decalcified sands in discreet beds and
clotted muds and micritic meniscus cements along fractures, suggesting dissolution of the
indicate local subaerial exposure, presumably on original ferroan cement (Milodowski & Wilmot
shoal tops. These vadose features may represent 1984). Abundant intra- and interformational
pedogenic alteration (Martin, Wilkinson & clays may have provided the iron for the ferroan
Lohmann 1986), or beachrock cementation of cement, for even at shallow burial depths iron
limited duration. may be mobilized from iron hydroxides (Emery,
Common aragonite dissolution attests to the this volume).
presence of a well developed, initially undersa-
turated meteoric water system (Longman 1980,
Moldovanyi & Lohmann 1984). As the oxidizing Phase 3 cement
ground waters became saturated with calcite This spar post-dates all others, cross-cutting pre-
they precipitated the non-ferroan druse and existing fabrics in vugs and fractures. Its blocky,
neomorphic spars (Talbot 1971, Chowdhury often poikilotopic, ferroan crystals are typical of
1982). As these cements are non-ferroan and burial precipitates (Scholle & Halley 1985).
non-luminescent the initial cement probably However, its association with solution and its
lacked iron or manganese. Over time, however, patchy response to staining are more enigmatic.
sufficient Mn 2+ was incorporated to promote Dickson & Coleman (1980), and Meyers &
luminescence. The cement remained non-fer- Lohmann (1978)'report patchy staining and CL
roan, suggesting that Mn 2+ may be activating response, Oickson & Coleman attributing it to
luminescence regardless of iron content. It homogenization of formerly zoned cements
should be noted that other factors, such as rate of while Meyers & Lohmann suggest that corrosion
precipitation and trace element concentration, during the stabilization of HMC to LMC left
346 C. B. de Wet

inclusions and relicts. In the Corallian, previous- Summary

ly stabilized allochems and cements are dissolved
The three cementation phases represent resi-
and phase 3 calcite is precipitated. In vugs where
dence time in distinct diagenetic environments.
the host rock is predominantly non-ferroan, the
Phase 1 cements precipitated in surface marine,
new spar contains little iron, while vugs in
vadose and freshwater phreatic environments.
ferroan hosts contain more iron-rich cement.
Phase 2 cements formed in reducing fluids
This suggests that local chemistries dictated the
during initial burial. Minor compaction effects,
composition of the precipitate even though the
such as spalled ooid cortices indicate that
uniformity of the CL response indicates a
cementation was not complete before burial.
homogeneous source. The stable isotopes and
Phase 3 involved tectonic uplift during the
geochemistry discussed in the next section help
Barremian, allowing meteoric recharge through
to resolve this apparent contradiction.
the Corallian Formation, causing vuggy dissolu-
To what environment should these phase 3
tion. Phase 3 cements precipitated during the
cements be ascribed? Aggressive, undersatur-
ensuing deep burial (Table 1).
ated solutions capable of vuggy dissolution are
commonly attributed to active meteoric circula-
tion or to brines preceding hydrocarbon migra- Geochemistry
tion. In the Corallian the phase 3 spars clearly
post-date some initial burial, but only the centre
Trace elements
of the basin reached burial depths sufficient for
hydrocarbon generation. However, Barremian Within the carbonate system, analysing for
uplilft exposed the Formation's margins (Fig. 7) cations whose distribution coefficients are less
as there are Corallian clasts in the Cretaceous than and greater than unity may be useful in
Faringdon Sponge Gravel (Talbot 1971) and in determining the amount of diagenetic alteration
the Lower Greensand Formation (Kelly 1985). a given component has experienced (Moore &
This period of exposure at the surface would Druckman 1981, Veizer 1983, Majid & Veizer
almost certainly have allowed undersaturated 1986). By comparing Mg 2+ and Sr z+ (with
meteoric waters to enter the Formation and be distribution coefficients of 0.013-0.06 and 0.13
driven basinward by gravity flow. During the respectively, values from Veizer 1983) with
ensuing phase of deep burial, the trapped Mn 2+ and Fe z+ (with distribution coefficients of
meteoric water would have mixed with connate 15-30 and 10-20, Veizer 1983) in specific cement
water that was being expelled from clays and phases, general trends in the 'openness' of the
slowly spread throughout the basin. At present, fluid system may become apparent. Phase 1-3
basinal fluid flow is still converging on the cements from the Corallian Formation were
Corallian in the Harwell 4 borehole (J. Black, separated and analysed on a Phillips 50 MHz
personal communication, 1987) due to cross- inductively coupled argon plasma spectrometer
formational flow from the Oxford and Kimmer- (1-3% precision).
idge clays. Therefore it is suggested that such a As magnesium and iron have opposing distri-
mix of meteoric and formation waters produced bution coefficients relative to 1 they should act
the phase 3 spars. antithetically during mineralogic transforma-

BURIAL CURVE- WESSEX BASIN

PHASE! .P___~SE2 PHASE
0 #

-r
b--
I1

N 5oo

rn
I000
U.,.~ BARREMIAN CRETACEOUS TERTIARY
BURIAL._ U P L I F T _ BURIAL UPLIFT .....

,co ,4o ,oo ao go 40 20 o

TIME (Mo)

FIG. 7. General burial curve for the western Wessex Basin indicating the relative timing of cementation
phases I-3. Barremian uplift exposed the Formation's western and northern margins to meteoric recharge.
 Diagenesis in the Corallian Formation 347
TABLE 1. Carbonate cement summary

PHAS~CL RESPONSE STAIN RESPONSE DESCRIPTION DIAGENETICENVIRONMEN'I"

. . . . ,,.

non-luminescent brownish micritic peloids, marine phreatic

to bright grapestone, inlraclasts ,.

non-luminescent non-ferroan isopachous fibrous marine phreatic

to bright
zoned dull to non-ferroen isopachous dogtoolh, oxidizing meteoric phreatic
bright equant, microspar nearsurface
complex zoning ferroan (mauve) syntaxial overgrowths mildly reducing
equant meteoric phreatic
2
non-luminescent ferroan equant, granular reducing metegric',connate,
or dully zoned shallow Dur~aJ
dull with dark variabte blocky, poikilotopic C r e t a c e o u s meteoric +
5 patches connate
dg.eP burial

tions involving a fluid phase. Partitioning theory Mg vs Fe PHASE I o

predicts that Mg 2+ will be preferentially retained 5' PHASE 2 +
in the fluid while Fe 2+ will become incorporated
into the crystal (Pingitore 1982). Therefore, % 4. PHASE :5 •
decreases in Mg 2+ and increases in Fe 2+ from an x
5 o
original marine value should indicate increasing o o o~ ~176 o
~o x
diagenetic alteration. The Corallian cement •
+
+
x
sequence data on phase 1 and 2 cements shows a
steady decrease in Mg 2§ away from a marine :E
value, as Fe z+ increases (Fig. 8a). This is the
expected result assuming aragonite and H M C , b) Mn vs Fe
both containing seawater-derived Mg 2§ stabi- 8 §

lized to L M C in a relatively Mg-poor freshwater % x +

x
+ +
system. The iron content rises due to the x +

reduction of Fe 3+ to Fe z+ and its incorporation o

into calcite during shallow burial. Mobile iron o oo

may have been derived from primary glauconite, x o

chamositic ooids and berthierine found to the

south (Talbot 1973), suggesting that collodial o
iron complexes were available for reduction even Fe (pprn) xlO3
during shallow burial.
The phase 3 cements exhibit variable magne- FIG. 8. Geochemical data for phase 1, 2 and 3
sium concentrations and a decrease in iron cements from the Corallian Formation, Oxford
relative to phase 2 cements, probably reflecting a Shallows area. (a) Plot of magnesium against iron.
multi-sourced origin. Barremian meteoric water Phase 1 cements are depleted in Mg 2+ relative to
present seawater (16300-75400 ppm) and are
dissolved previously stabilized phase 1 and 2
enriched in Fe 2+ (2-39 ppm, marine calcite values
cements, creating a fluid with a new compo- from Veizer 1983). Phase 2 spars show a significant
sition. The phase 3 cement that precipitated increase in iron concentration accompanied by
from this fluid contained similar concentrations continuing magnesium depletion. Phase 3 cements
of Mg 2+ and less Fe z+ than the adjacent phase 2 contain intermediate iron concentrations and
crystals. The majority of the samples analysed variable magnesium values. (b) Plot of magnesium
exhibit this small chemical gradient, suggesting against iron. Phase 1 cements are depleted in both
that the bulk solution began to approach equilib- cations relative to phase 2 and 3 cements. Phase 2
rium with the new fluid. In the example where spars contain the highest concentrations of
manganese and iron, indicating that cations were
the phase 3 spar chemically matches the host
being supplied from source sediments, probably
phase 1 cement, local dissolution and reprecipi- intra- and interformational clays, into reducing
tation without chemical homogenization with waters. Data from the phase 3 cements show two
the new fluid must have occurred. In this case, groups; one containing Mn2+-rich samples and a
bulk solution disequilibrium (Veizer 1983) was second containing two Mn2+-depleted samples.
348 C. B. de W e t

probably the dominant mechanism for cement chemical equilibration with the bulk solution.
growth. Strontium has been widely used as a tracer for
Mn 3§ is more readily reduced to Mn 2§ than evaluating diagenetic alteration, particularly in
Fe 3§ is reduced to Fe 2+. Therefore the incorpora- corals (Pingitore 1982, Pingitore & Eastman
tion of Mn 2§ into calcite should precede the 1986, Yoshioka et al. 1986) since it is readily
incorporation of Fe 2+. Corallian Formation incorporated into the aragonite structure. How-
phase 1 cements luminesce brightly under catho- ever, the original coral structure may have a
doluminescence excitation although they have a significant effect on the amount of alteration
non-ferroan response to staining. In Fig. 8(b), the experienced during stabilization (Walter 1985,
most depleted phase 1 spar contains negligible Constantz 1986). Talbot (1971) discussed differ-
iron but 80 ppm Mn z+. This may reflect the ences in textural preservation relative to coral
initial development of the redox front. species within the Corallian Formation. Good
The trend from phase 1 to phase 2 cements in textural preservation, accompanied by high Sr 2§
Fig. 8(b) is consistent with a shallow burial values, may indicate areas where the dominant
model, demonstrating that initially oxygenated, mechanism for cation movement was diffusion
and/or iron and manganese-poor marine and rather than mass transport (Pingitore 1982). In
meteoric waters became progressively reduced this study, neomorphosed Theocosmilia corals
and enriched in both Fe 2+ and Mn 2+. The iron have Sr 2+ values ranging from 950 to 400 ppm
and manganese were probably derived from (Fig. 9), suggesting differing amounts of flushing
colloids and hydroxides. by phase 1 meteoric waters. All of the phase 1
In Fig. 8(b), the phase 3 cements form Mn-rich calcites have relatively high Sr 2§ concentrations
and Mn-poor categories. The Mn-rich group is and low Mn 2+ values indicative of early stabili-
depleted in Fe 2+ compared to the phase 2 zation and precipitation in an oxygenated and/or
cements, suggesting that Fe2+-depleted Barre- Mn 2+ poor environment.
mian meteoric water mixed with dissolving The cements that precipitated during shallow
phase 2 calcites, thus diluting the overall iron burial are depleted in Sr 2§ and enriched in Mn 2+
concentration. The slight increase in Mn 2§ may relative to the phase 1 spars. This is the expected
account for the similarity of CL response in result assuming that strontium's distribution
almost all of the phase 3 spars. Although staining coefficient is less than one and manganese's is
showed a variable iron content, the pattern of greater than one. During burial, Mn 3§ became
luminescence and intensity are uniform in all of reduced to Mn 2§ and, as predicted by partioning
the blocky phase 3 spars. This indicates that theory, was substituted into the calcite lattice
Mn 2+ predominates over Fe 2§ in dicating CL while Sr 2+ remained in the fluid. The most
response. If the phase 3 waters were only slightly enriched phase 2 cement contains 1200 ppm
reducing or a redox potential fluctuated then Mn 2§ and only 300 ppm Sr -'+ (Fig. 9).
Mn 2+ would be more likely to be incorporated Phase 3 cements are generally intermediate
into the precipitating calcite than would Fe z+ . between phases 1 and 2, a consistent trend seen
As the Oxford Shallows are adjacent to the area
exposed during Barremian uplift, it is likely that
phase 3 waters were only midly reducing when Sr vs Mn PHASE I (Coral) o
(Cement) 9
the sequence was reburied during the rest of the o
Cretaceous. The phase 3 cement would have PHASE 2 +
precipitated before the fluids became fully PHASE 5 *
reduced. An alternative explanation could be
that there was an increase in the concentration of %
Mn 2§ to the system. This would suggest that the ~6
phase 3 cement precipitated during maximum Q. o
burial as compaction released Mn 2+ from adja- &
cent clay horizons. In this environment iron ,74 x

would also be released, but authigenic clays and + x

pyrite could have incorporated it (Milodowski & +

x
x
§

Wilmot 1984), leaving the calcite relatively

depleted.
The two Mn-poor samples in Fig. 8(b) contain o
Mn 2+ and Fe 2+ concentrations that are very o i 4 6 /3 lb (2
Mn (ppm) x 102
similar to the host phase 1 cements. As in the
previous discussion of Mg 2+ and Fe 2§ in phase 3 FIG. 9. Plot of strontium against manganese for
cements, this may imply that localized dissolu- phase 1, 2 and 3 cements from the Corallian
tion and reprecipitation has occurred, without Formation, Oxford Shallows area.
 Diagenesis in the Corallian Formation 349
in all of the geochemical plots. This probably carbon and oxygen isotopic compositions
reflects mixing between cations introduced in (Hudson 1977). Using the average 6180 of - 2 %o
Barremian waters and cations released during palaeotemperature determinations were made
phase 3 dissolution. In Fig. 9, one sample using the equation of Shackelton & Kennett
contains Sr 2+ and Mn 2+ concentrations that (1975):
closely match phase 1 values, indicating incom-
T~ = 16.9 - 4.38 (8c - 6 w ) + 0.10 (6c - 6w) 2
plete homogenization with the bulk solution
relative to other phase 3 cements. (where 6c is the measured isotopic value of CO,
and 6w is the isotopic composition of CO, in
equilibrium with formation water). Assuming a
Stable isotopes
value of w = - 1.2 for Jurassic sea water (Mar-
Stable isotope compositions were measured on shall & Ashton 1980) this gives a palaeotempera-
individual cements collected from serial thin ture of 20.5~ for the average phase 1 cements.
sections and rock chips. Analysis of CO: was This is a reasonable estimate of seawater
carried out using a VG SIRA 12 mass spectro- temperature. It is slightly lower than Marshall &
meter. Each run of 10 samples contained two Ashton's (1980) estimate of 22-25~ from British
internal standards of Lincolnshire Limestone. Jurassic oysters, but is higher than values
Precision was better than 0.1 m1-1. Corrected calculated from their marine cements and ooids.
results (after Craig 1957) are expressed relative The 6180 values for phase 2 cements are similar
to the PDB standard. to those for phase 1 but the 613C values are about
In Fig. 10, the carbon isotopic compositions 3 %o 'lighter'. As discussed above the petrogra-
indicate a gradual evolution in phase 1 and 2 phic and geochemical evidence suggest that
cements from values near to Jurassic marine phase 2 cements formed under the influence of
compositions (Hudson 1977, Walls et al. 1979, meteoric water. The measured carbon isotopic
Marshall & Ashton 1980), through to values with values are consistent with minor input from light
more negative 613C probably caused by the (photosynthetically-fixed) soil-derived CO2 but
incorporation of organogenic carbon during the oxygen values would normally be expected to
shallow burial. Phase 1 cements have 'marine' show more negative 'freshwater influenced'

glSc PDB
:5.0

2.5

/ o / 2.0
\ o
o / 1.5
PHASE I \ o
o /
\ / I.O
\ / 0.5
\oF 180 PDB
- 6 0 -55 -50 -45 -4.0 -35 -5,0 -25 -2.0 -I.5 -I.0 - 0 5
/ \ -0.5
/ ~• \ /
I-
*\ -I.0
/~, \ i, +1
/ \ 1\,, ] PHASE 2 -15

I \ ~- q -2.0

t \\ -2.5
.I \ PHASE B
-:30
•
/ -:35
11
x _..-t/
-4,0

FIG. 10. Plot of 613C PDB against 3180 PDB. Stable isotope data for cement phases 1, 2 and 3 from the
Corallian Formation, Oxford Shallows area.
35o C. B. de Wet
values. This apparent discrepancy may be The model presented in Fig. 11 relies on
explained by suggesting that in the context of the Chadwick's (1986) tectonic interpretation for the
Corallian limestones the 'fresh' water may not Wessex basin and predicts lateral variation in
have been isotopically very different from sea the north-south direction. Facies belts trend
water. Near-coastal rainfall in low latitudes has east-west and seismic profiles show sediment
an isotopic composition which approaches local thickening dramatically in the troughs (R.
sea water (Anderson & Arthur 1983). In the Benmore, personal communication, 1987). The
proposed tectonic setting for the Corallian the steeper, generally south-facing, fault-controlled
fresh water involved in phase 2 cements would slope break marks an abrupt facies change from
have returned as rain without considerable predominantly clastic and carbonate sands to
transport as vapour, thereby limiting fractiona- predominantly clays. Allochthonous debris, pro-
tion. It is conceivable that it could have been duced by storms and gravity sliding, collected in
virtually indistinguishable, in terms of filsO, the muds at the base of the slope. The gentler,
from the sea water particularly if any marginal usually north-facing slope is characterized by
depletion in 61sO from the meteoric water cycle lagoonal micrites and sands, grading into clays.
was compensated for by an opposite marginal The Wessex basin remained dissected by east-
enrichment by local evaporation on the exposed west trending faults throughout its early diagene-
highs. tic history; phases 1 and 2. Corridors of coarse-
The difference in 6180 between phases 1 and 2 grained sediments transported fluids from mar-
and phase 3 cements is about 2.5 Too.If precipita- ginal meteoric recharge sites along fault block
tion again took place from water with a constant highs into the basin. During phase 2 shallow
oxygen isotopic composition the temperature burial meteoric water became reducing and/or
would have to have increased by 12.8~ The iron and manganese-rich as colloids and hydrox-
Wessex Basin has a geothermal gradient of 30~ ides released cations. Local facies distribution
km-1 and in the Marchwood borehole waters in produced permeability differences that in-
the Corallian Formation are presently at 40~ at fluenced phase 1 and 2 cementation. The low
850 m depth (Burgess et al. 1981). However, the permeability wackestones and packstones at
isotopic change in the calcites could be due, at Harwell contain very little phase 1 cement
least in part, to isotopic changes in the precipitat- relative to the grainstones on the Oxford high.
ing water. As argued above, the phase 3 cements The highly permeable sandstone beds both on
probably formed in response to the introduction the Oxford Shallows high and off it, across the
of Barremian meteoric water in a groundwater slope break at HarweU, are almost exclusively
system. Such water would almost certainly have cemented by phase 2 spar or are decalcified. As
had a more negative 6180. The 6180 values in the the fault highs were surrounded by both inter-
cements may therefore reflect both temperature and intraformational clays they were essentially
effects and changing isotopic composition of the isolated from each other during diagenetic
porewater. Most of the phase 3 samples have phases 1 and 2. The cements that precipitated
fil3C values which lie between - 0 . 4 and during that time can be correlated by CL
- 1 . 4 Too; however, in two samples, the carbon internally within a block, but not across to an
values are significantly lower at - 3 . 3 and adjacent high. This is unlike the broad cement
- 3 . 9 %o PDB. These carbon values probably correlations that can be made in more structur-
reflect minor input from a source of isotopically ally uniform basins (Meyers 1978).
light organogenic carbon into a pore fluid whose It is only the phase 3 spar, which probably
carbon values were dominated by the dissolution precipitated during the deepest burial phase, that
of marine carbonate. Possible sources might has a basin-wide CL pattern. The chemistry of
include thermal breakdown of organic matter the late spars is generally intermediate between
('decarboxylation') or surface derived carbon the two preceding phases of cementation, sug-
from the recharge area. The latter is unlikely gesting that introduced Barremian meteoric
except for the limestones adjacent to the recharge water was not in equilibrium with the host rock.
area and soil zones so thermal breakdown is the Initially the waters were undersaturated, creat-
preferred alternative for the phase 3 cements. ing secondary porosity as non-fabric selective
vugs and fractures which attest to this disequilib-
rium. Phase 3 cements precipitated in the voids
as the fluids became supersaturated. Cations
Discussion were supplied both from the introduced water
and from the dissolution of previously stabilized
The depositional environment of the Corallian allochems and cements. Hundreds of pore
sediments near Oxford consisted of a rimmed volumes of fluid must have passed through the
accretionary shelf adjacent to the Welsh Massif. voids, yet relics of precursor fabrics and phase 3
 Diagenes& in the Corallian Formation 35I

.'6"6,o

FIG. 11. Proposed depositional model and facies distribution along block crests and fault-bounded slopes for
the Corallian Formation within the Wessex basin.

cement chemistries that almost perfectly match (2) Early diagensis (cements 1 and 2) was
the host rock suggest that in particular cavities strongly influenced by depositional trends. Ma-
diffusion was the dominant mass transfer mecha- rine cements, probably HMC or aragonite, grew
nism. In the majority of phase 3 cements the as fringes, peloids and grapestone micrites in
chemistry is internally more uniform, implying areas of moderate circulation. Muds were depos-
increased homogenization with the bulk solution ited in meniscuses and bridges between grains
via mass transport. during shoal top exposure. During burial, waters
moved along conduits created by coarser grained
sandstones and grainstones in a freshwater
phreatic system. The fluids were initially oxidiz-
Conclusions ing but became reducing and iron-rich during
shallow burial. Early Cretaceous regional uplift
exposed Corallian margins, admitting undersa-
(1) A depositional model, consistent with, and turated meteoric water and initiating phase 3
dependent on, the internal structure of the vuggy dissolution.
Wessex Basin, is proposed. Fault-bounded (3) Geochemical and stable isotopic data
blocks step down into the basin, trending show fluid evolution in cements 1 and 2, with
longitudinally east-west across it. Syndeposi- Mn 2+ and Fe 2+ increasing and Sr 2+ and Mg 2+
tional tectonics regulate local relative sea-level decreasing through time. The late burial spar 3
independently over each block, producing vari- occurs throughout the basin and has a depleted
able facies distributions. In the Oxford Shallows oxygen isotopic value suggesting increased tem-
area moderate to high energy currents create perature as fluids sourced from connate water
sand shoals and permit patch reef development were expelled from clays. In some examples the
in a shallow subtidal shelf environment. spar contains ionic concentrations which almost
352 C. B. de Wet
m a t c h those of the host rock, in these cases local isotope laboratory is supported by NERC Research
dissolution and reprecipitation in a relatively Grant GR3/5339 to Dr Marshall.) I have benefitted
closed system have dictated the new spar's greatly from discussions with Dr J. A. D. Dickson. Drs
chemical composition. M. Talbot, P. Gutteridge, J. A. D. Dickson and J. D.
Marshall made useful suggestions to improve the
manuscript. The British Geological Survey provided
ACKNOWLEDGMENTS: I wish to thank the staff at Royal the Harwell and Chalgrove borehole samples and a
Holloway and Bedford New College for running the grant from Sigma Xi Research contributed towards
ICP samples, and Dr J. D. Marshall and H. P. field expenses. Texaco (Houston) has kindly supported
Attenborough at the University of Liverpool for laboratory expenses for the Cambridge carbonate
allowing me to run the stable isotopes there. (The stable group.

References
ANDERSON, T. F. & ARTHUR, M. A. 1983. Stable FORSlCH, F. T. 1975. Trace fossils as environmental
isotopes of oxygen and carbon and their applica- indicators in the Corallian of England and
tion to sedimentologic and paleoenvironmental Normandy. Lethaia, 8, 151-72.
problems. In: Stable Isotopes in Sedimentary HEFU, L 1986. Geodynamic scenario and structural
Geology, pp. 1- 1-151. Society of Economic Paleon- styles of Mesozoic and Cenozoic basins in China.
tologists and Mineralogists, Dallas, Short Course, Bulletin of the American Association of Petroleum
no. 10. Geologists, 70, 377-95.
BROOKFIELD, M. 1973. Palaeogeography of the upper HUDSON, J. D. 1977. Stable isotopes and limestone
Oxfordian and lower Kimmeridgian (Jurassic) in lithification. Journal of the Geological Society of
Britain. Palaeogeography, Palaeoclimatology and London, 133, 637-60.
Palaeoecology, 14, 137-67. KELLY, S. R. A. 1985. Jurassic and Cretaceous deposits
--1977. The use of coarse fraction analysis in of Upware, Cambridgeshire, Bulletin of the Geolo-
palaeonvironmental interpretation: upper Oxfor- gical Society of Norfolk, 35, 3-38.
dian and lower Kimmeridgian (Jurassic) sedi- LONGMAN, M. W. 1980. Carbonate diagenetic textures
ments in Dorset, England. Sedimentary Geology, from near-surface diagenetic environments. Bulle-
20, 249-80. tin of the American Association of Petroleum
-- 1978. The lithostratigraphy in the upper Oxfor- Geologists, 64, 461-87.
dian and lower Kimmeridgian beds of south MACINTYRE, I. G. 1985. Submarine cements--the
Dorset, England. Proceedings of" the Geologists' peloidal question. In. SCHNEIDERMANN, N. t~r
Association, 89, 1-32. HARRIS, P. (eds). Carbonate Cements, pp. 3-16.
BURGESS,W. G. etal. 1981. The MarchwoodGeothermal Society of Economic Paleontologists and Minera-
Borehole--a Preliminary Assessment of the Re- logists Special Publication no. 36.
source. Institute of Geological Sciences, Geophy- MAJID, A. H. & VEIZER, J. 1986. Deposition and
sics and Hydrology Division. chemical diagenesis of Tertiary carbonates, Kir-
CHADWICK, R. A. 1986. Extension tectonics in the kuk oil field, Iraq. Bulletin of the American
Wessex Basin, Southern England. Journal of the Association of Petroleum Geologists, 70, 898-913.
Geological Society of London, 143, 465-88. MARSHALL, J. D. & ASHTON, M. 1980. Isotopic and
CHOWDHURY, A. 1980. Geochemistry and sedimento- trace element evidence for submarine lithification
logy of the Corallian sediments of southern of hardgrounds in the Jurassic of eastern England.
England. UnpublishedPhD Thesis, p. 579. Univer- Sedimentology, 27, 271-89.
sity of Southampton, England. MARTIN, G., WILKINSON,B. & LOHMANN,K. C. 1986.
--1982. Variations in diagenetic environments of The role of skeletal porosity in aragonite neomor-
the Coral Rag and underlying oolite of the phism--Strombus and Montastrea from the Pleis-
Osmington Oolite Formation (Upper Oxfordian) tocene Key Largo Limestone, Florida. Journal of
of the Oxford-Berkshire area, U.K., and their Sedimentary Petrology, 56, 194-203.
implications. Sedimentary Geology, 32, 247-63. MCKENZIE, D. P. 1978. Some remarks on the develop-
CONSTANTZ, B. R. 1986. The primary surface area of ment of sedimentary basins. Earth and Planetary
corals and variations in their susceptibility to Science Letters, 40, 25-32.
diagenesis. In: SCHROEDER,J. H. & PURSER, B. H. MCKERROW, W. S. (ed.) 1978. The Ecology of Fossils.
(eds). ReefDiagenesis, pp. 53-76. Springer-Verlag, Duckworth & Co., London.
Berlin. MEYERS, W. J. 1978. Carbonate cements: their
CRAIG, H. 1957. Isotopic standards for carbon and regional distribution and interpretation in Missis-
oxygen and correction factors for mass-spectro- sippian limestones of southwestern New Mexico.
metric analysis of carbon dioxide. Geochimica et Sedimentology, 25, 371-400.
Cosmochimica Acta, 12, 133-49. --& LOHMANN, K. C. 1978. Microdolomite-rich
DICKSON, J. A. D. 1966. Carbonate identification and syntaxial cements: proposed meteoric-marine
genesis as revealed by staining. Journal of Sedi- mixing zone phreatic cements from Mississippian
mentary Petrology, 36, 491-505. limestones, New Mexico. Journal of Sedimentary
- - & COLEMAN, M. L. 1980. Changes in carbon and Petrology, 48, 475-88.
oxygen isotope composition during limestone MILODOWSKI, A. E. & WILMOT, R. D. 1984. Diagene-
diagenesis. Sedimentology, 27, 107-18. sis, porosity and permeability in the Corallian
 Diagenesis in the Corallian Formation 353
Beds (Upper Oxfordian) from the Harwell Re- TALBOT, M. R. 1971. Calcite cements in the Corallian
search Site, South Oxfordshire, U.K. Clay Min- Beds (Upper Oxfordian) of southern England.
erals, 19, 323-41. Journal of Sedimentary Petrology, 41, 261-
MOLDOVANYI, E. & LOHMANN, K. C. 1984. Isotopic 73.
and petrographic record of phreatic diagenesis: -- 1973. Major sedimentary cycles in the Corallian
Lower Cretaceous Sligo and Cupido Formations. Beds (Oxfordian) of southern England. Palaeogeo-
Journal of Sedimentary Petrology, 54, 972-85. graphy, Palaeoclimatology and Palaeoecology, 14,
MOORE, C. H. & DRUCKMAN,Y. 1981. Burial diagene- 293-317.
sis and porosity evolution, Upper Jurassic Smack- TEN HAVE, T. & HEIJEN, W. 1985. Cathodolumines-
over, Arkansas and Louisiana. Bulletin of the cence activation and zonation in carbonate rocks:
American Association of Petroleum Geologists, 65, an experimental approach. Geologie en Mijnbouw,
597-628. 64, 297-310.
PINGITORE, N. E. 1982. The role of diffusion during VEIZER, J. 1983. Chemical diagenesis of carbonates:
carbonate diagenesis. Journal of Sedimentary theory and application of trace element technique.
Petrology, 52, 27-39. In . Stable Isotopes in Sedimentary Geology, pp. 3-1-
--& EASTMAN, M. P. 1986. The coprecipitation of 100. Short Course Society of Economic
Sr 2§ with calcite at 25~ and 1 atm. Geochimica et Paleontologists and Mineralogists, Dallas, no.
Cosmochimica Acta, 50, 2195-203. 10.
Ru, K. & PIGGOTT, J. 1986. Episodic rifting and WALLS, R. A., MOUNTJOY, E. W. & FRITZ, P. 1979.
subsidence in the South China Sea. Bulletin of the Isotopic composition and diagenetic history of
American Association of Petroleum Geologists, 70, carbonate cements in Devonian Golden Spike
1136-55. reef, Alberta, Canada. Bulletin of the Geological
SALLER, A. H. 1986. Radiaxial calcite in Lower Society of America, 90, 963-82.
Miocene strata, subsurface Enewetak Atoll. Jour- WALTER, t. 1985. Relative reactivity of skeletal
nal of Sedimentary Petrology, 56, 743-62. carbonates during dissolution: implications for
SCHOLLE, P. A. & HALLEY, R. B. 1985. Burial diagenesis. In: SCHNEIDERMANN,N. & HARRIS,
diagenesis: out of sight, out of mind! In: SCHNEI- P. (eds). Carbonate Cements, pp. 3-16. Society of
DERMANN, N. & HARRIS, P. (eds). Carbonate Economic Paleontologists and Mineralogists Spe-
Cements, pp. 309-34. Society of Economic Paleon- cial Publication no. 36.
tologists and Mineralogists Special Publication WHITTAKER, A. (ed). 1985. Atlas of Onshore Sedimen-
no. 36. tary Basins in England and Wales. Post-Carbonifer-
SELLWOOD,B. 1978. The Jurassic. In: MCKERROW, W. ous Tectonics and Stratigraphy. Blackie,
S. (ed). The Ecology of Fossils, pp. 204-79. Glasgow.
Duckworth & Co., London. WRIGHT, J. K. 1981. The Corallian rocks of north
STANLEY, D. J. & MOORE, G. T. 1983. The ShetJbreak: Dorset. Proceedings of the Geologists' Association,
Critical Interface on Continental Margins. Society 92, 17-32.
of Economic Paleontologists and Mineralogists YOSHIOKA, S., OHDE, S., KITANO, Y. & NANAMORI,K.
Special Publication no. 33. 1986. Behavior of magnesium and strontium
STONELEY, R. 1982. The structural development of the during the transformation of coral aragonite to
Wessex Basin. Journal of the Geological Society of calcite in aquatic environments. Marine Chemis-
London, 139, 543-54. try, 18, 35-48.

C. B. DE WET, Department of Earth Sciences, Downing Street, University of Cambridge,

Cambridge CB2 3EQ, UK.
 Index

Pages on which figures appear are printed in italic, and those with tables in bold

Agat field 319, 323 Bridport Sands

albite in Sherwood Sandstone Group 332, 333 cementation 109-10, 117
albitization 197, 198 outcrop 103-4
alluvial fan 126-7, 128, 129 petrography 104-7
aluminium mobility 63-4 stable isotopes 107-9
amphibole stability 269 Budleigh Salterton Pebble Beds
anatase in Sherwood Sandstone Group 327, 331-2 cementation 332
anhydrite in Sherwood Sandstone Group 327, 331 outcrop 175, 325-6
dissolution 332, 335 stable isotopes 179
eodiagenesis 329 Burlington Formation
mesodiagenesis 330 diagenesis 251-4
stable isotope composition 181, 182, 185-6 dolomitization
ankerite modelling 254-6
Bothamsall oilfield 56, 60 petrography 239-45
Viking Formation 284-5, 286, 287, 291 timing 245-51
Wilcox Formation 226, 228, 232 stratigraphy 237-40
apatite in turbidites 146, 147, 148, 149
aragonite
Bridport Sands 106, 107, 109, 110 ~13C
Catalan Basin beach ridge systems 130 Bridport Sands 107-9
Corallian Formation 345 Burlington-Keokuk Formations 241,243
Lincolnshire Limestone 204 Catalan Basin beach ridge systems 132, 134
arkose 306, 309 Corallian Formation 349-50
Aylesbeare Group 330 Great Estuarine Group 272-3, 274
Norwegian reservoir rocks 317, 318, 319, 321,
322
bacteria and carbonate precipitation 112-3, 115-6 Sherwood Sandstone Group 179-85
Barton sand 30, 31, 32, 33, 34, 35 Viking Formation 281,282-3, 289
barytes in Sherwood Standstone Group 327, 331-2 Viking Group 112-4
Basal Permian Sands 177, 184-5 Yingtang Formation 164-6
beach-ridge systems calcite cements 21
deposits 123-4 Bridport Sands 109-10
diagenesis 127-30 Burlington-Keokuk Formations 245, 247
dolomitization 134-7 Catalan Basin beach ridge systems 127, 130
reef associations 125-6 Corallian Formation 343, 344, 345-6
stable isotopes 134 Great Estuarine Group 272
trace elements 132-4 Lincolnshire Limestone
beekite 90, 91, 96, 98 ferroan 205-6, 207-8
Beer (Devon) stratigraphy 87, 88, 89 non-ferroan 205, 206-7
berthierine 104, 105, 106, 109 Los Monegros playas 120
bicarbonate 112-3, 115-6 North Coles Levee 194-5
bioclast silicification Norwegian reservoir rocks 315-6
description 88-92 Sherwood Sandstone Group 178, 327, 328-9, 332
discussion 99-100 Viking Formation 281,286, 287, 290
ff80 values 96-8 Viking Group 114-6
role of drusy quartz 101 Welsh Borderland sandstones 304, 309
biodegradation of hydrocarbons 16-7 Wilcox Formation 244-5, 226, 229-31
biotite alteration 147 Cambrian see Comley Sandstone; Hollybush
bitumens 21 Sandstone
Black Grit porosity and hydrocarbons 299, 230, Cannonball Doggers 42-4
231, 306, 307 carbon in Jet Rock concretions 43
Blaen Twrch turbidite 150 carbon isotopes
bloedite 119, 120 fractionation in concretions 47-8
Bog Quartzite 299, 300, 303, 306 see also 613C
Bothamsall oilfield carbonate cements
diagenesis 65-8 growth 1
petrography 55-60 influence of sulphur 21
Brage field 318 isotopic composition 179-85

355
356 Index

sources 313-4 formation

see also calcite; dolomite; magnesite early theories 41-2
Carboniferous see Bothamsall oilfield; Burlington- role of methane 48-52
Keokuk Formations in Great Estuarine Group 264, 266-8
Carboniferous Limestone, role in Jurassic Corallian Formation
diagenesis 214-5, 216 depositional environment 340-1,343
Catalan Basin 123, 124, 125 diagenesis 341-3, 345-6
catalysis and sulphate-hydrocarbon reactions 23 facies analysis 350-2
cathodoluminescence (CL) geochemistry 346-50
Burlington-Keokuk Formations 239, 241,243, stratigraphy 339-40
244 Cretaceous see Cenomanian; Folkestone sand; Small
Corallian Formation 343, 345-6, 347, 348 Cove Hardground; Soreq Formation; Viking
Lincolnshire Limestone 204, 205, 206 Formation; Woburn Sands
Sherwood Sandstone Group 177-8, 182, 329 Cretaceous sandstones of Norway
Viking Group 112, 113 cementation studies 315-6
Welsh Basin turbidites 142, 147 petrology 314-5
Wilcox Formation 222, 224 crude oil thermal cracking 16-17
Yingtang Formation 161-62 Curling Stones 42-4
cement
fringing 109-10, 115
mosaic 110, 115-6 Degree of Pyritization (DOP) 52
pendant 115 Devonian see Dongganglin Formation; Tangding
see also anhydrite; calcite; carbonate; dolomite; Formation; Yingtang Formation
illite; kaolinite; phosphate; quartz; siderite; dickite 285
smectite doggers 42-4, 266
cementation modelling dolomite cements and dolomitization
mathematical basis 3-6 Burlington-Keokuk Formations
results 6-12 cathodoluminescence 239, 241-5, 246
cementation rate constant (fl) 6, 12 mixed-water model 254-6
Cenomanian in Devon 87 petrography 239, 240, 241
chalcedony timing 245-51
Cenomanian Catalan Basin beach-ridge systems 126-7, 128,
bioclasts 90, 94 129, 130, 134-7
nodules 93, 94-5 Lincolnshire Limestone 205
Chicksgrove Plant Bed nodules 76-9, 81-3 Los Monegros playas 120
Sherwood Sandstone Group cements 327,. 334 North Coles Levee 194
Yingtang Formation 160 Sherwood Sandstone Group 182, 186, 327, 328,
Chalgrove borehole 341,343 332-4, 335
chert Viking Formation 284-5, 286, 287, 290
Burlington-Keokuk Formations 249-51 Yingtang Formation
Cenomanian geochemistry 162-7
bioclasts 99-100 mixed-water model 167-8
nodules 94, 95, 100 petrography 160-2
Chicksgrove Plant Bed 74-80 stratigraphy 158-60
Chicksgrove Plant Bed Dongganglin Formation 158, 159
chert petrology 74-80 Dowsing Fault Zone 173
silica circulation 80-3 Draugen field 318
stratigraphy 73-4, 75 Duntulm Formation
chlorite cementation 266, 270
Viking Formation 281,286, 287, 290 depositional environment 262, 263, 264
Welsh Basin turbidites 147, 148, 149 stratigraphy 261
Chondrites 144, 151 vitrinite reflectivity 271
clay minerals
Viking Formation 280, 291
Welsh Borderland sandstones 302, 304 East Yorkshire-Lincolnshire Basin
Cleethorpes drillcore Permo-Triassic petrology 176-7
log 177, 189-90 stable isotopes 182-7
stable isotopes in cements 183-5 stratigraphy 177
Comley Sandstones hydrocarbons 303, 304, 305, structure 173, 176
306-8 Elgol Sandstone Formation 261
compaction effects epidote stability 269
diagenesis 198 evaporite precipitation 119
porosity 37-8 Exogyra silicification 91, 96, 98
compressive strengths in UK sands 37
concretions
 Index 357

Fayetteville Shale concretions 42 diagenesis 300-1,304, 310

feldspar hydrocarbons 303
diagenesis Hollybush Sandstone hydrocarbons 300, 303, 306,
Bothamsall oilfield sandstones 56, 57, 65, 66, 308
67 hydrocarbons
Sherwood Sandstone Group 327, 329-31,331, redox chemistry 16-19
333 geological significance 20-5
Viking Formation 284, 286 Welsh Basin occurrences 298-9, 301,303, 304-6
Wilcox Formation 225-6, 227 porosity and migration 310-11
occurrence in beds of Llandovery age 147, 302, see also methane
304, 306, 307, 309 hydrogen sulphide 20-1
see also albite; plagioclase
florencite 147
fluid flow modelling 3 illite
fluid inclusions 316-7 Bothamsall oilfield sandstones 56, 58, 59, 60, 66,
fluorescence in chert nodules 78-9 67
Folkestone sands 30, 31, 32, 33, 34 North Coles Levee 196
Folly Sandstone 298, 300, 303 Sherwood Sandstone Group 327, 335
fossil silicification see bioclast Viking Formation 284, 285-6, 287, 291
francolite 147 Wilcox Formation 229, 232
Frio Formation cementation Inoceramus silicification 90, 91, 92, 93, 94
comparison with Wilcox Formation 232-33 iron in carbonates
ankerite 226, 228 Burlington-Keokuk Formations 239
calcite 224-5 Corallian Formation 346-9
feldspar 226 Lincolnshire Limestone Formation 206, 209, 210,
quartz 222-4 211
Yingtang Formation 166
iron oxides
garnet stability 269 Catalan Basin beach-ridge systems 126-7
geothermal gradient in Texas 229 Sherwood Sandstone Formation 326-8
G irvanella 341 Viking Formation 281,286
glauconite 281,286, 30I, 306 iron oxyhydroxides 150, 151, 152
goethite 335 isotopes see 613C; •180; ~345; 87Sr/86Sr
Grantham sand 30, 31, 32, 33, 34, 35
Great Dirt Bed 75, 85
Jet Rock concretions
Great Estuarine Group
chemistry 42-4
depositional environment 262
formation 52-3
diagenesis
Jurassic see Bridport Sands; Chicksgrove Plant Bed;
concretions 266-8
Corallian Formation; Grantham sand; Great
early changes 262-4
Estuarine Group; Jet Rock; Kellaways Sand;
late changes 264-6
Lincolnshire Limestone; Viking Group
shell mineralogy 269-70
Jurassic sandstones of Norway
water chemistry 272-4
cementation 315-6
Inner Hebrides basin 259-60
petrology 314-5
mineralogy 260-2
Tertiary volcanic effects 270-2
griestoniensis Zone turbidites 142, 143-4 kaolinite
Guallar playa 120, 121 Bothamsall oilfield sandstones 56, 58, 59, 60, 61,
Guangxi region palaeogeography 158 65, 66, 67
Gudrun field 318 North Coles Levee 196, 197
Guilin Formation 158 Norwegian reservoir rocks 316, 320
Gullfaks field 318, 322 Sherwood Sandstone Group 327, 333, 334
gypsum 119, 120, 332 Viking Formation 284, 285, 286, 287, 290, 291
Keg River Formation 24-5
Kellaways Sand 30, 31, 32, 33, 34
haematite Keokuk Formation
Catalan Basin beach-ridge systems 126 diagenesis 251-4
Sherwood Sandstone Group 326-8 dolomitization
halite ll9, 120 modelling 254-6
Harwell 4 borehole 341,343, 345 petrography 239-45
Haven Cliff (Devon) stratigraphy 87, 88, 89 timing 245-51
heat of reaction in sulphate-hydrocarbon stratigraphy 237-8
chemistry 24 Kilmaluag Formation 261,264
Hild field 320 Kingella silicification 90, 92
Hoar Edge Grit 299, 300 kyanite stability 269
358 Index

Lealt Shales Formation 261, 262, 263, 265-6 Neomiodon preservation 263, 266, 267, 269
lepispheres 75-9 neomorphism in shelly limestone 269-70
limonite 334 Nettlecombe borehole 329
Lincolnshire Limestone Nisku Formation 24, 25
diagenesis nodules
pre-burial 204 siliceous 92-6
post burial 204-6 formation 100-1
stratigraphy 202-3 3180 99
litharenite 306 see also concretions
lithification measurement 34 North Coles Levee reservoir 191, 193
Longmynd 298, 301 Norway, oilfield sandstones
Los Monegros playas 119 calcite textures 315-6
lutecite rosettes 89, 90 fluid inclusions 316-7
petrology 314-5
stable isotopes 317
Norwich Crag sand 30, 31, 32, 33
nucleation of concretions 41-2

magnesite 120-21
magnesium ~18O
Corallian Formation 346-9 Bridport Sands 107-9
Lincolnshire Limestone 209-10, 212 Burlington-Keokuk Formations 241,243
see also dolomitization Catalan Basin beach-ridge systems 132, 134
manganese in carbonates Cenomanian 96-9
Burlington-Keokuk Formations 239 Corallian Formation 349-50
Corallian Formation 346--9 Great Estuarine Group 270, 271,273, 274
Lincolnshire Limestone 208-9 Norwegian reservoir rocks 317, 318, 319, 320,
Yingtang Formation 166 321, 322
manganese oxide 327, 334 Sherwood Sandstone Group 179-85
manganese oxyhydroxides 150, 151, 152 Viking Formation 280-91
Marchwood drillcore Viking Group 112-4
cements 329, 332, 334 Wilcox Formation
stable isotopes 179-80 ankerite 226, 228
log 173, 187 calcite 224-5, 226
reservoir properties 334-5 feldspar 225-6, 227
marine water and dolomitization 135, 136, 137, quartz 221-2
167-8 Yingtang Formation 164-6
matrix geometry modelling opal
method 4-6 chert nodules 75-6, 80, 81
results 6-12 role in silicification 100
Mercia Mudstone 171,330 Ordovician see Black Grit; Hoar Edge Grit;
meteoric water in diagenesis Stiperstones Quartzite
Catalan Basin beach-ridge systems 126-7, 135 organic matter decay
Cenomanian 100-1 concretion control 42
Chicksgrove Plant Bed 81, 82, 83 phosphate diagenesis 150, 151,152
Corallian Formation 345-6, 350 Oseberg field 321
Lincolnshire Limestone 204 Otter Sandstone 173, 177, 329
Viking Formation 288-9, 291,292-3 Oxford Shallows block
Welsh Borderland sandstones 310-11 Corallian Formation environment 340-1
Wilcox Formation 230 facies analysis 350-2
Yingtang Formation 167-8 oxidation states of sulphur 18-19
methane and methanogenesis 20, 45-6 oxygen isotopes see d18o
and concretions 48-52
mica 147, 148, 149
see also biotite; chlorite; muscovite Peclet number 3
microbes permeability modelling 1-3
concretion control 42 Permo-Triassic see Budleigh Salterton Pebble Beds;
phosphate diagenesis 150, 151 Mercia Mudstone; Otter Sandstone;
microfabrics in sands 32-4 Sherwood Sandstone Group
Midgard field 318 phosphate cements
mineral stability, heavy 268-9 appearance 144-7
mirabilite 119, 120 development 148-53
muscovite 56, 57, 66, 67 Pine Point (Canada) 25
Pito playa 120
plagioclase dissolution 196, 198
 Index 359

porosity silicification
modelling and hydrodynamics 3, 4, 7, 8, 9, 10, 11 bioclasts 88-92
North Coles Levee sandstone 196-7 d'8o 96-8, 99-100
UK uncemented sands 32, 34, 37-9 nodules 75-9, 92-6
Welsh Borderland sandstones 300, 301,302, 304 d180 99, 100-1
primary 308-9 silica sources 80-1
secondary 306-8, 309, 310 see also quartz
Portlandian stratigraphy 75 Silurian see Bog Quartzite; Folly Sandstone;
potassium mobility 64-5 griestoniensis Zone; Trecoed Beds
Praeexogyra spp. 269, 271 Skudiburgh Formation 261, 263, 264
Praemytilis spp. 269 Small Cove Hardground 94, 96
pressure solution 39 smectite
pyrite North Coles Levee 196
Jet Rock 43, 44 Viking Formation 284, 285-6, 287, 291
North Coles Levee 197-8 Welsh Basin turbidites 147
Viking Formation 281,286 Wilcox Formation 229, 232
Welsh Basin turbidites 147, 148, 149 sodium in dolostones 166
solute supply modelling 3
solution-mineral equilibria modelling 60-3
quartz Soreq Formation 168
Bothamsall oilfield sandstones 56, 60, 66, 67 Southampton Western Esplanade drillcore
Cenomanian 101 cements 329, 332
Chicksgrove Plant Bed 75, 80, 81 stable isotopes 181-2
Sherwood Sandstone Group 327, 329-30, 331, log 189
332, 333 reservoir properties 334-5
Viking Formation 281,284, 286, 287, 290 Spergen Formation 239, 249
Welsh Borderland sandstones 304, 305, 309 sphene stability 269
Wilcox Formation 221-4, 229-31 stable isotopes see fi~3C; fi~80; t534S
quartz arenite 300-1,304 staurolite stability 269
quartzine 90 Stevens sandstones 193-8
Stiperstones Quartzite 299, 300
diagenesis 306
redox reactions and sulphur 16-19 hydrocarbons 303, 304, 306-8
reefs strontium
development 125-6 concentrations
diagenesis 130-2 Corallian Formation 346-9
replacement reactions 79-81 Lincolnshire Limestone 211-4
reservoir properties Yingtang Formation 166
Sherwood Sandstone Group 335-6 isotope ratios
Welsh Borderland sandstones 306, 309, 311 Burlington-Keokuk Formations 239
Reynolds number 3 Corallian Formation 346-9
Rhaxella 341,343 Lincolnshire Limestone 211-4
rip-up clasts 147, 150 Wilcox Formation 225, 227, 229
rutile stability 269 subarkose 301-4
sublitharenite 301-4
sulphate-hydrocarbon redox reactions 16-19, 44-5
~34S in anhydrite 185-6 bacterial sulphate reduction (BSR) 15
St Genevieve Limestone 239 geological significance 20-5
St Louis Formation 239, 247 thermochemical sulphate reduction (TSR) 15
San Joaquin basin 191,192 Sussex Sandstone 297
Seaborough borehole 329
shear strength in UK sands 36
shell preservation 269-70 Tangding Formation 159-60
Sherwood Sandstone Group temperatures, diagenetic
diagenesis 186-7 Lincolnshire Limestone 205
early 326-9 North Coles Levee 195-6, 198
late 334 Norwegian reservoir rocks 320, 322, 323
middle 329-34 Sherwood Sandstone Group 332
outcrops and stratigraphy 173-4, 325-6 sulphate-hydrocarbon reactions 21-2
petrology 175-8 Viking Formation 287-8
reservoir properties 334-5 Wilcox Formation 228-9
Shrewton borehole 330, 332 Tertiary see Barton sand; Catalan Basin; Frio
siderite Formation; North Coles Levee; Norwich
North Coles Levee 193, 194 Crag; Thanet sand; Wilcox Formation
Viking Formation 281,284, 286, 287, 290 Texas Gulf Coast stratigraphy 220
360 Index

Thalassinoides 93, 95 porosity 306-10

Thanet sand 30, 31, 32, 33 stratigraphy 298, 300
thenardite 121, 122 Wessex Basin
thermal cracking 16 environments in Corallian 343, 350-2
titanium dioxides 147, 148 Permo-Triassic sediments 175-6, 179-82
tourmaline stability 269 structure 173, 174, 339-40, 341
trace elements Western Esplanade drillcore see Southampton
Catalan Basin beach-ride systems 132-4 Wilcox Formation
Corallian Formation 346-50 diagenesis
Lincolnshire Limestone 214-5 ankerite 226, 228
Trecoed Beds 300 calcite 224-5, 226
hydrocarbons 298, 303, 304, 306-8, 310 feldspar 225-6, 227
Troll Field 110-11 quartz 221-4
turbidites in Welsh Basin 143-4, 298 reaction sequence 228-32
cementation 144-7 stratigraphy 220
mineralogy 147 Wilmington (Devon) stratigraphy 87, 88, 89
phosphate diagenesis 148-53 Winterborne Kingston drillcore
cements 329, 330, 332
stable isotopes 181, 185
Valtos Sandstone Formation 261, 263, 266-8, 270 log 175, 189
Viking Formation reservoir properties 334-5
deposition 278-9 Wistow drillcore 182-3
diagenesis 288-93 Woburn Sands 30, 31, 32, 33
temperature 287-8 Wytch Farm drillcore
mineralogy 281-6 cements 329, 330, 331, 332
Viking Group sandstones stable isotopes 182, 185
cementation 112, 114-6, 117 log 189
stable isotopes 112-4 reservoir properties 335-6
petrography 111-2
stratigraphy 110-11
vitrinite reflectivity 22, 271-2 xenotime 147
volcanism and burial 270-1,274

Yingtang Formation
Warsaw Shale 239, 255 dolomitization chemistry 166-7
Welsh Basin, Palaeozoic mixed-water model 167-8
palaeogeography 143-4 stable isotopes 164
sediments 298, 299 petrography 162
Welsh Borderland sandstones stratigraphy 159, 160
diagenesis 300-6
hydrocarbons 298-9, 303
zircon stability 147